the myth(?) of the police sub-culture - boas práticas no...

30
PIJPSM 26,1 88 Policing: An International Journal of Police Strategies & Management Vol. 26 No. 1, 2003 pp. 88-117 # MCB UP Limited 1363-951X DOI 10.1108/13639510310460314 The myth(?) of the police sub-culture John K. Cochran and Max L. Bromley Department of Criminology, University of South Florida, Tampa, Florida, USA Keywords Police, Employee attitudes, Organizational culture, Self-efficiency Abstract This study examines empirically the extent to which there is evidence of an endemic sub-culture of policing among a sample of sheriffs’ deputies. While failing to observe widespread adherence to the sub-cultural norms and values suggested in the literature, such adherence is observed among a subset of our sample. Advanced statistical techniques (i.e. cluster analysis and discriminant function analysis) are then used to create, replicate, and validate a numerical taxonomy of policing. The taxonomy reveals three types of law enforcement orientations: ``Sub- Cultural Adherents,’’ ``COP Cops,’’ who represent a nouveau sub-culture strongly committed to public service, and ``Normals,’’ who, on average, are quite average and are not especially committed to either sub-cultural form. Both those working in the sociology of work and occupations and those in industrial/occupational psychology acknowledge that employees tend to adopt job-specific sub-cultural responses (i.e. shared beliefs, attitudes, values, and norms) to the contingencies they experience in their organizational and occupational environments (DiMaggio and Powell, 1991; Manning, 1995; Ouchi and Wilkins, 1985; Ritti, 1994; Sackmann, 1991; Schein, 1985). Criminologists working within these perspectives have consistently noted the unique sub- cultural responses of criminal justice practitioners, especially law enforcement and correctional officers, given the particular characteristics of these fields (Bayley and Bitner, 1984; Bitner, 1974; Black, 1980; Brown, 1981; Chan, 1997; Crank, 1998; Drummond, 1976; Farkas and Manning, 1997; Fielding, 1988; Kappeler et al., 1994; McNamara, 1967; Muir, 1977; Reiner, 1985; Reuss-Ianni, 1983; Skolnick, 1966; Van Maanen, 1974; Westley, 1970). That is, the occupational environment of criminal justice includes exposure to human misery, exposure to great situational uncertainty, and exposure to intrinsic danger, all coupled with high levels of coercive authority and ``invisible discretion’’ granted to these officers which enable them to carry out their mandates. Moreover, most criminal justice employees work in unique organizational environments which expose them to rigid, militaristic authority structures with fixed lines of command and communication that are coupled with often vague and conflicting The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/1363-951X.htm Support for this research was provided by an Advancing Community policing grant (#97OCWX0090) from the Office of Community Oriented Policing Services (COPS) of the National Institute of Justice, United States Department of Justice. The findings, interpretations, and opinions contained within are those of the authors and do not necessarily reflect the position of the Department of Justice. The authors wish to thank the sheriff, his staff, and the officers of the Hillsborough County, FL Sheriff’s Department for their cooperation with this study; finally they would also like to acknowledge the anonymous reviewers and Drs Lawrence Travis and Leonard Territo for their helpful comments on an earlier version of this manuscript.

Upload: duongnga

Post on 14-May-2018

219 views

Category:

Documents


0 download

TRANSCRIPT

PIJPSM261

88

Policing An International Journal ofPolice Strategies amp ManagementVol 26 No 1 2003pp 88-117 MCB UP Limited1363-951XDOI 10110813639510310460314

The myth() of the policesub-culture

John K Cochran and Max L BromleyDepartment of Criminology University of South Florida

Tampa Florida USA

Keywords Police Employee attitudes Organizational culture Self-efficiency

Abstract This study examines empirically the extent to which there is evidence of an endemicsub-culture of policing among a sample of sheriffsrsquo deputies While failing to observe widespreadadherence to the sub-cultural norms and values suggested in the literature such adherence isobserved among a subset of our sample Advanced statistical techniques (ie cluster analysis anddiscriminant function analysis) are then used to create replicate and validate a numericaltaxonomy of policing The taxonomy reveals three types of law enforcement orientations ` Sub-Cultural Adherentsrsquorsquo ` COP Copsrsquorsquo who represent a nouveau sub-culture strongly committed topublic service and ` Normalsrsquorsquo who on average are quite average and are not especiallycommitted to either sub-cultural form

Both those working in the sociology of work and occupations and those inindustrialoccupational psychology acknowledge that employees tend to adoptjob-specific sub-cultural responses (ie shared beliefs attitudes values andnorms) to the contingencies they experience in their organizational andoccupational environments (DiMaggio and Powell 1991 Manning 1995 Ouchiand Wilkins 1985 Ritti 1994 Sackmann 1991 Schein 1985) Criminologistsworking within these perspectives have consistently noted the unique sub-cultural responses of criminal justice practitioners especially law enforcementand correctional officers given the particular characteristics of these fields(Bayley and Bitner 1984 Bitner 1974 Black 1980 Brown 1981 Chan 1997Crank 1998 Drummond 1976 Farkas and Manning 1997 Fielding 1988Kappeler et al 1994 McNamara 1967 Muir 1977 Reiner 1985 Reuss-Ianni1983 Skolnick 1966 Van Maanen 1974 Westley 1970) That is theoccupational environment of criminal justice includes exposure to human miseryexposure to great situational uncertainty and exposure to intrinsic danger allcoupled with high levels of coercive authority and ` invisible discretionrsquorsquo grantedto these officers which enable them to carry out their mandates Moreover mostcriminal justice employees work in unique organizational environments whichexpose them to rigid militaristic authority structures with fixed lines ofcommand and communication that are coupled with often vague and conflicting

T h e E m e r a ld R e s e a r c h R e g is te r fo r th is jo u r n a l i s a v a i la b le a t

httpwwwemeraldinsightcomresearchregister

T h e c u rr e n t i s s u e a n d fu l l te x t a r c h iv e o f th is jo u r n a l is a v a i la b le a t

httpwwwem eraldinsightcom1363-951Xhtm

Support for this research was provided by an Advancing Community policing grant(97OCWX0090) from the Office of Community Oriented Policing Services (COPS) of theNational Institute of Justice United States Department of Justice The findings interpretationsand opinions contained within are those of the authors and do not necessarily reflect theposition of the Department of Justice The authors wish to thank the sheriff his staff and theofficers of the Hillsborough County FL Sheriffrsquos Department for their cooperation with thisstudy finally they would also like to acknowledge the anonymous reviewers and Drs LawrenceTravis and Leonard Territo for their helpful comments on an earlier version of this manuscript

The myth()of the policesub-culture

89

guidelines for policies and procedures (Bitner 1974 Lundman 1980Neiderhoffer 1967) As a result these employees are faced with tremendous job-related stressors In an effort to cope with these working conditions theseemployees are said to adopt a unique sub-cultural response

For the police and other law enforcement officers this sub-cultural responsethe police sub-culture is presumed to be made manifest in the manner by whichofficers perceive their role as police and the scope of this role their beliefsregarding how their role should and should not be performed and theirattitudes toward the criminal law criminal procedures including departmentalpolicies the police and other criminal justice practitioners criminal offendersvictims and witnesses even citizens as a whole[1]

More specifically adherence to this police sub-culture includes

negative attitudes toward the various legal restrictions placed upontheir efforts to effectively fight crime plusmn ` handcuffing the policersquorsquo

negative skeptical attitudes toward legal institutions and otherelements of the criminal justice system plusmn ` the courts are too lenientrsquorsquo

negative and suspicious attitudes toward police administration and thepolice bureaucracy plusmn ` departmental policies and procedures are tooburdensome and ineffectiversquorsquo and

negative cynical attitudes toward the citizenry plusmn ` they are all a bunch ofliars and crooksrsquorsquo

This culture idealizes aggressive and authoritativetake-charge approaches topolicing such that crime fighting is accorded very high priority and the ordermaintenance and service elements of the police role are devalued and are seenas ` baby-sittingrsquorsquo andor ` social workrsquorsquo

This strong orientation toward crime-fighting and the accompanyingnegative attitudes toward due process limitations placed on the police by theappellate courts give rise to additional elements of the police sub-culture Theseoutlooks for instance can lead to abuses of police authority (Chevigny 19691995 Kappeler et al 1994 Shearing 1981 Skolnick 1966 Skolnick and Fyfe1993 Westley 1970) The sub-culture of policing stresses the individualautonomy of police officers especially their need to take the initiative on thestreets and during police-citizen encounters in order to effectively control crime(Manning 1995 Sykes and Brent 1980) These outlooks are also the basis forthe sub-cultural preference among many police officers for selectiveenforcement of the law in which the focus is on the more serious violations of thelaw (Brown 1981) Selective enforcement and the use of illegitimate tactics as ameans of fighting crime can in turn give rise to tensions within the communityand to citizen complaints against these officers (Brown 1981 Westley 1970)

The uncertainty and danger associated with policing have also contributedto this sub-culture It is argued that the police respond to these work conditionsby being highly suspicious every situation is read in terms of its potential fordanger and every person is viewed as a liar and a crook (Van Maanen 1974)

PIJPSM261

90

Such an outlook leads sub-cultural adherents to ` maintain an edgersquorsquo while onthe street (Van Maanen 1974 p 118) Their world becomes an insular dualismof ` us versus themrsquorsquo (Rubenstein 1973 Sparrow et al 1990 Tauber 1970) Theycome to see themselves as the ` thin blue linersquorsquo that prevents society fromslipping permanently into moral decay and unrest In essence a primaryelement of the police sub-culture is a great level of cynicism and suspicionwhich is carried by them into their encounters with the citizens they have beenhired to serve and protect This in turn only adds to the tensions between thepolice and their community

Finally the police sub-culture stresses an orientation which simultaneouslygrants autonomy to each officer yet demands from them unyielding loyalty tothe group (Broderick 1977 Brown 1981 Manning 1995 Muir 1977 White1972) These strong norms of group loyalty are said to produce a ` code ofsilencersquorsquo among fellow officers under which they are not to ` snitchrsquorsquo on oneanother (Bitner 1967 Brown 1981 Chevigny 1995 Reiner 1985 Reuss-Ianni1983 Rubenstein 1973 Shearing 1981 Westley 1970) The code of silenceimpedes efforts to detect and investigate corruption and other forms of policedeviancemisconduct Hence not only are sub-cultural adherents suspiciousand cynical in their orientations toward the citizenry they are also suspiciousand distrusting of the police bureaucracy especially upper administrationThese officers tend to view the police bureaucracy as too politically motivatedas ineffective and burdensome and as punitive they see police officials as moreinclined to punish them for their procedural errors (Walker 1977) than toreward them for the substantive successes produced from these ` shortcutsrsquorsquo (iearrests confessions evidence tips etc) Thus sub-cultural adherents areencouraged by their group to ` lie lowrsquorsquoand ` fly under the radarrsquorsquo (Paoline et al2000 p 578) that is to remain as much as possible invisible to upperadministration They are expected to adopt a ` cover your assrsquorsquo attitude whichalso covers their partnersrsquo and other officersrsquo (Chevigny 1995 Westley 1970)These attitudes may also lead police officers to confront and negate any reformefforts especially community-oriented policing through group resistance(Greene et al 1994 Lurigio and Skogan 1994 Sadd and Grinc 1993)

No doubt there are additional elements of this police sub-culture which wehave neglected to include in our brief description Nevertheless our account isfairly complete and representative and is consistent with other efforts atdetailing it (Britz 1997 Crank 1998 Drummond 1976 Farkas and Manning1997 Fielding 1988 Herbert 1998 Muir 1977 Paoline et al 2000 Reuss-Ianni 1983) More to the point our review of the research literature revealsseveral interesting contradictions regarding the police sub-culture Firstmost of the research on the police sub-culture has been exploratory anddescriptive in nature These studies appear to have accepted the existence ofthis sub-culture without question and thus have sought to describe itscharacteristics A key element of these descriptive efforts has been the oft-repeated claim that the sub-culture is endemic to policing and that adherenceto it is widespread However a few studies have been more explanatory in

The myth()of the policesub-culture

91

nature and question whether or not such a sub-culture exists and if found tobe present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks (Jermier et al 1991 Paoline et al 2000) Studieswhich have observed its presence among some but not most officers typicallyproceed by developing a typology of officers based at least in part on thedegree of sub-cultural adherence (Brown 1981 Muir 1977 Reiner 1978Wilson 1968) Rarely however are these typologies the product of analyticstatistical procedures designed for their construction (eg cluster analysis)nor are these typologies subsequently tested for their empirical validity(Bailey 1994 plusmn see Jermier et al (1991) for an exception) This brings us to thethreefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture in terms ofthe deputiesrsquo workrole orientations toward their crime control andservice functions their level of cynicism traditionalism and receptivityto change

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Data and methodsThe data for this study were obtained from self-administered questionnairesdistributed to patrol deputies and other sworn employees in Districts I and II ofthe Hillsborough County FL Sheriffrsquorsquos Office (HCSO) Districts I and II are twoof four decentralized semi-autonomous operational units within the HCSO TheHCSO is a full service sheriffrsquos office which includes law enforcement jailoperations and bailiff and other court-related functions At the time of theadministration of the data instrument (Summer 1997 and Spring 1999) theHCSO comprised 2561 personnel consisting of 970 law enforcement deputies768 jail deputies and 849 civilian employees the rest being administrative linesand court personnel making it one of the largest sheriffrsquos offices in the country(Reaves 1992)

The HCSO provides law enforcement services to the residents of theunincorporated portions of Hillsborough County Florida Hillsborough County islocated in west central Florida and has a total population of approximately900000 The county seat is Tampa with a population of nearly 300000 Theunincorporated area of the county is about 100 sq miles in size has a population

PIJPSM261

92

of approximately 500000 and comprises diverse communities including severalhigh-crimelow-income public housing areas migrant-worker and rural-agricultural communities upper-middle class ` bedroomrsquorsquo communities and anumber of working- and middle-class areas These areas also include severallarge shopping malls many smaller retail and service-oriented enterprises auniversity hospitals and other health-care facilities several parks andrecreational areas and a wide range of commercial and industrial facilities

The two HSCO Districts we examined were purposively selected becausetheir operational areas are most similar to those typical of metropolitanmunicipal police departments (see Weisheit et al 1995 Christensen and Crank2001)[2] Again both districts are semi-autonomous multi-faceted operationalunits within which the SOrsquos effort to integrate community-oriented-agencygroups wide had been well established since 1994 District I which servesnorthern Hillsborough county an area of about 79 sq miles with a populationof approximately 50000 comprises 137 sworn employees of whom six were inupper-administrative positions 24 in mid-level management (ie sergeants andcorporals) six were detectives and 101 were patrol deputies These personnelare 84 percent male 80 percent white 13 percent black and 7 percent HispanicDistrict I is dominated by the presence of a large and unstable public housingarea which almost entirely comprises poor black households and is plaguedwith a significant crime problem

District II serves northeastern Hillsborough County a much larger landmass (259 sq miles) and larger population (approximately 155000) than that ofDistrict I District II comprises 193 sworn personnel of whom five hold upper-administrative positions 24 are in mid-level management 14 are detectivesand 150 are patrol deputies The demographic profile of these sworn personnelin District II very closely resembles that of District I 84 percent male and 78percent white However District II has a higher percentage of Hispanic officers(11 percent) and a lower percentage of black officers (9 percent) This variationin the racialethnic composition of District II personnel may have beenpurposefulstrategic in that this District contains a large rural-agriculturalcommunity with a significant number of low-wage and migrant laborers whoare disproportionately of Hispanic background Importantly no District IIpersonnel had ever worked in District I and likewise no District I personnelhad ever worked in District II

After the purpose of the study and the questionnaire were discussed andapproved by the top command of both the HCSO and each District a 149-itemquestionnaire was distributed and completed during roll call for each of the fiveshifts during consecutive days in August 1997 at District I and in March 1999at District II[3] Respondents were assured that the information they providedus would be kept confidential that they would remain anonymous and thattheir participation was voluntary Approximately one-third (38 percent fromDistrict I and 33 percent from District II) of the sworn employees wereunavailable for participation on the days of the survey administration due tocourt training illness military duty day-off or other personal leave This

The myth()of the policesub-culture

93

attrition reduced the total number of eligible respondents from both Districts to218 of whom only four declined to participate giving us a response rate of 98percent[4] However several respondents chose not to answer every questiontypically these were the socio-demographic and other ` identifierrsquorsquo items such asage raceethnicity gender rank shift and assignment The questionnaire tookapproximately 35-40 minutes to complete

Measures of adherence to the police sub-cultureRespondentsrsquo degree of adherence to the sub-culture of policing is measured byfive distinct workrole orientation scales crime control service cynicismtraditionalism and receptivity to change The first of these crime controlorientation measures the importance deputies place on the law enforcementand crime control functions of their jobs The 13-item additive scale assessesthe degree to which these deputies agree or disagree (1 = strongly disagree 5 =strongly agree) with a series of Likert-type statements such as `All lawsshould be fully enforced at all times otherwise people lose respect for the lawrsquorsquo` Law enforcement officers should not forget that enforcing the laws is by fartheir most important responsibilityrsquorsquo ` Most law enforcement officers have tospend too much of their time handling unimportant non-crime calls forservicersquorsquo and ` If law enforcement officers act in a service capacity it detractsfrom their ability to fight crimersquorsquo Item wording and frequency distributions forthese 13 items are presented in the Appendix

Several of the 13 items which comprise this scale were reverse-coded beforethey were entered into a principal components factor analysis from which fourfactors yielded eigenvalues greater than 100 however the scree discontinuitytest suggests that a single factor solution best represents these data Loadingson this first factor ranged from 038 to 075 (with three items loading at 016019 and 027) Despite these three weak loadings the Cronbachrsquos Alphareliability coefficient for the additive was 076 and does not improveappreciably if any or all of these three items are deleted High values on thisscale indicate a strong preference among these respondents for the crime-fighting aspects of their roles as law enforcement officers

Respondentsrsquo work orientation toward service-related activities an indicatorof disagreement with the police sub-culture is measured by a 14-item additivescale which was designed to assess the extent to which these deputies agreeand disagree (1 = strongly disagree 5 = strongly agree) with a series of Likert-type statements such as ` Law enforcement officers should ask citizens whattypes of services they wantrsquorsquo `Assisting citizens is just as important asenforcing the lawrsquorsquo ` Law enforcement should be seen primarily as a service-oriented profession rather than a crime control professionrsquorsquo and ` Crimes areonly one of several problems about which law enforcement officers should beconcernedrsquorsquo Item wording and frequency distributions for these 14 items arealso presented in the Appendix

Again several of these 14 service orientation items required reverse codingbefore they were entered into a principal components factor analysis This

PIJPSM261

94

factor analysis produced five factors with eigenvalues greater than 100however the scree discontinuity tests suggest that a single factor solution bestrepresents these data Loadings on this first factor ranged from 036 to 063(with two items loading at only 020 and 028) The Cronbachrsquos Alpha reliabilityfor the additive scale produced from all 14 items was 074 and does not increaseappreciably if either or both of the poorly loading items isare removed Highvalues on this scale are indicative of a strong pro-service work orientationcommensurate with the ideals of community-oriented policing and antagonisticto the police sub-culture

Cynicism is a nine-item additive scale which measures the extent to whichthese deputies agree or disagree (1 = strongly disagree 5 = strongly agree)with another series of Likert-type statement such as ` Most people areuntrustworthy and dishonestrsquorsquo ` Law enforcement officers will never trustcitizens enough to work together effectivelyrsquorsquo ` Citizens will not trust lawenforcement officers enough to work together effectivelyrsquorsquo and ` Most peoplelack the proper level of respect for law enforcement officersrsquorsquo Item wordingand frequency distributions for these nine items are also presented in theAppendix

As with the other scales several of the items comprising this cynicism scalerequired reverse coding so that high values on it demonstrate a cynicalperception of the public and their effectiveness in contributing to the objectivesof law enforcement and community-oriented policing Prior to scaling theseitems were also entered into a principal components factor analysis This factoranalysis produced a single factor solution with loadings ranging from 049 to073 the Cronbachrsquos Alpha reliability coefficient for this nine-item additivescale was 083

Traditionalism is a six-item additive scale which measures the extent towhich officers participating in this study agree or disagree (1 = stronglydisagree 5 = strongly agree) with a short series or Likert-type statements suchas ` In an efficient organization power is distributed at the toprsquorsquo ` Participatorymanagement schemes really do not work within law enforcement agenciesrsquorsquo` The quasi-military structure is the most effective organizational type for lawenforcement agenciesrsquorsquo and ` Communication works best when it follows clearestablished channels from the top downrsquorsquo Item wording and frequencydistributions for these six items are also presented in the Appendix

Once again several items required reverse coding so that high values on thisadditive scale are indicative of a traditionalistauthoritarian orientation towardthe organizational hierarchy and distribution of power within a lawenforcement agency and is counter-intuitive to the ideals of decentralizationand officer-empowerment under community-oriented policing Prior to scalingthese six items were entered into a principal components factor analysis fromwhich a two-factor solution was produced however the scree discontinuity testsuggests that a single factor solution best fit these data Loadings on this firstfactor ranged from 035 to 064 The Cronbachrsquos Alpha reliability coefficient forthis additive traditionalism scale was only 054

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

89

guidelines for policies and procedures (Bitner 1974 Lundman 1980Neiderhoffer 1967) As a result these employees are faced with tremendous job-related stressors In an effort to cope with these working conditions theseemployees are said to adopt a unique sub-cultural response

For the police and other law enforcement officers this sub-cultural responsethe police sub-culture is presumed to be made manifest in the manner by whichofficers perceive their role as police and the scope of this role their beliefsregarding how their role should and should not be performed and theirattitudes toward the criminal law criminal procedures including departmentalpolicies the police and other criminal justice practitioners criminal offendersvictims and witnesses even citizens as a whole[1]

More specifically adherence to this police sub-culture includes

negative attitudes toward the various legal restrictions placed upontheir efforts to effectively fight crime plusmn ` handcuffing the policersquorsquo

negative skeptical attitudes toward legal institutions and otherelements of the criminal justice system plusmn ` the courts are too lenientrsquorsquo

negative and suspicious attitudes toward police administration and thepolice bureaucracy plusmn ` departmental policies and procedures are tooburdensome and ineffectiversquorsquo and

negative cynical attitudes toward the citizenry plusmn ` they are all a bunch ofliars and crooksrsquorsquo

This culture idealizes aggressive and authoritativetake-charge approaches topolicing such that crime fighting is accorded very high priority and the ordermaintenance and service elements of the police role are devalued and are seenas ` baby-sittingrsquorsquo andor ` social workrsquorsquo

This strong orientation toward crime-fighting and the accompanyingnegative attitudes toward due process limitations placed on the police by theappellate courts give rise to additional elements of the police sub-culture Theseoutlooks for instance can lead to abuses of police authority (Chevigny 19691995 Kappeler et al 1994 Shearing 1981 Skolnick 1966 Skolnick and Fyfe1993 Westley 1970) The sub-culture of policing stresses the individualautonomy of police officers especially their need to take the initiative on thestreets and during police-citizen encounters in order to effectively control crime(Manning 1995 Sykes and Brent 1980) These outlooks are also the basis forthe sub-cultural preference among many police officers for selectiveenforcement of the law in which the focus is on the more serious violations of thelaw (Brown 1981) Selective enforcement and the use of illegitimate tactics as ameans of fighting crime can in turn give rise to tensions within the communityand to citizen complaints against these officers (Brown 1981 Westley 1970)

The uncertainty and danger associated with policing have also contributedto this sub-culture It is argued that the police respond to these work conditionsby being highly suspicious every situation is read in terms of its potential fordanger and every person is viewed as a liar and a crook (Van Maanen 1974)

PIJPSM261

90

Such an outlook leads sub-cultural adherents to ` maintain an edgersquorsquo while onthe street (Van Maanen 1974 p 118) Their world becomes an insular dualismof ` us versus themrsquorsquo (Rubenstein 1973 Sparrow et al 1990 Tauber 1970) Theycome to see themselves as the ` thin blue linersquorsquo that prevents society fromslipping permanently into moral decay and unrest In essence a primaryelement of the police sub-culture is a great level of cynicism and suspicionwhich is carried by them into their encounters with the citizens they have beenhired to serve and protect This in turn only adds to the tensions between thepolice and their community

Finally the police sub-culture stresses an orientation which simultaneouslygrants autonomy to each officer yet demands from them unyielding loyalty tothe group (Broderick 1977 Brown 1981 Manning 1995 Muir 1977 White1972) These strong norms of group loyalty are said to produce a ` code ofsilencersquorsquo among fellow officers under which they are not to ` snitchrsquorsquo on oneanother (Bitner 1967 Brown 1981 Chevigny 1995 Reiner 1985 Reuss-Ianni1983 Rubenstein 1973 Shearing 1981 Westley 1970) The code of silenceimpedes efforts to detect and investigate corruption and other forms of policedeviancemisconduct Hence not only are sub-cultural adherents suspiciousand cynical in their orientations toward the citizenry they are also suspiciousand distrusting of the police bureaucracy especially upper administrationThese officers tend to view the police bureaucracy as too politically motivatedas ineffective and burdensome and as punitive they see police officials as moreinclined to punish them for their procedural errors (Walker 1977) than toreward them for the substantive successes produced from these ` shortcutsrsquorsquo (iearrests confessions evidence tips etc) Thus sub-cultural adherents areencouraged by their group to ` lie lowrsquorsquoand ` fly under the radarrsquorsquo (Paoline et al2000 p 578) that is to remain as much as possible invisible to upperadministration They are expected to adopt a ` cover your assrsquorsquo attitude whichalso covers their partnersrsquo and other officersrsquo (Chevigny 1995 Westley 1970)These attitudes may also lead police officers to confront and negate any reformefforts especially community-oriented policing through group resistance(Greene et al 1994 Lurigio and Skogan 1994 Sadd and Grinc 1993)

No doubt there are additional elements of this police sub-culture which wehave neglected to include in our brief description Nevertheless our account isfairly complete and representative and is consistent with other efforts atdetailing it (Britz 1997 Crank 1998 Drummond 1976 Farkas and Manning1997 Fielding 1988 Herbert 1998 Muir 1977 Paoline et al 2000 Reuss-Ianni 1983) More to the point our review of the research literature revealsseveral interesting contradictions regarding the police sub-culture Firstmost of the research on the police sub-culture has been exploratory anddescriptive in nature These studies appear to have accepted the existence ofthis sub-culture without question and thus have sought to describe itscharacteristics A key element of these descriptive efforts has been the oft-repeated claim that the sub-culture is endemic to policing and that adherenceto it is widespread However a few studies have been more explanatory in

The myth()of the policesub-culture

91

nature and question whether or not such a sub-culture exists and if found tobe present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks (Jermier et al 1991 Paoline et al 2000) Studieswhich have observed its presence among some but not most officers typicallyproceed by developing a typology of officers based at least in part on thedegree of sub-cultural adherence (Brown 1981 Muir 1977 Reiner 1978Wilson 1968) Rarely however are these typologies the product of analyticstatistical procedures designed for their construction (eg cluster analysis)nor are these typologies subsequently tested for their empirical validity(Bailey 1994 plusmn see Jermier et al (1991) for an exception) This brings us to thethreefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture in terms ofthe deputiesrsquo workrole orientations toward their crime control andservice functions their level of cynicism traditionalism and receptivityto change

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Data and methodsThe data for this study were obtained from self-administered questionnairesdistributed to patrol deputies and other sworn employees in Districts I and II ofthe Hillsborough County FL Sheriffrsquorsquos Office (HCSO) Districts I and II are twoof four decentralized semi-autonomous operational units within the HCSO TheHCSO is a full service sheriffrsquos office which includes law enforcement jailoperations and bailiff and other court-related functions At the time of theadministration of the data instrument (Summer 1997 and Spring 1999) theHCSO comprised 2561 personnel consisting of 970 law enforcement deputies768 jail deputies and 849 civilian employees the rest being administrative linesand court personnel making it one of the largest sheriffrsquos offices in the country(Reaves 1992)

The HCSO provides law enforcement services to the residents of theunincorporated portions of Hillsborough County Florida Hillsborough County islocated in west central Florida and has a total population of approximately900000 The county seat is Tampa with a population of nearly 300000 Theunincorporated area of the county is about 100 sq miles in size has a population

PIJPSM261

92

of approximately 500000 and comprises diverse communities including severalhigh-crimelow-income public housing areas migrant-worker and rural-agricultural communities upper-middle class ` bedroomrsquorsquo communities and anumber of working- and middle-class areas These areas also include severallarge shopping malls many smaller retail and service-oriented enterprises auniversity hospitals and other health-care facilities several parks andrecreational areas and a wide range of commercial and industrial facilities

The two HSCO Districts we examined were purposively selected becausetheir operational areas are most similar to those typical of metropolitanmunicipal police departments (see Weisheit et al 1995 Christensen and Crank2001)[2] Again both districts are semi-autonomous multi-faceted operationalunits within which the SOrsquos effort to integrate community-oriented-agencygroups wide had been well established since 1994 District I which servesnorthern Hillsborough county an area of about 79 sq miles with a populationof approximately 50000 comprises 137 sworn employees of whom six were inupper-administrative positions 24 in mid-level management (ie sergeants andcorporals) six were detectives and 101 were patrol deputies These personnelare 84 percent male 80 percent white 13 percent black and 7 percent HispanicDistrict I is dominated by the presence of a large and unstable public housingarea which almost entirely comprises poor black households and is plaguedwith a significant crime problem

District II serves northeastern Hillsborough County a much larger landmass (259 sq miles) and larger population (approximately 155000) than that ofDistrict I District II comprises 193 sworn personnel of whom five hold upper-administrative positions 24 are in mid-level management 14 are detectivesand 150 are patrol deputies The demographic profile of these sworn personnelin District II very closely resembles that of District I 84 percent male and 78percent white However District II has a higher percentage of Hispanic officers(11 percent) and a lower percentage of black officers (9 percent) This variationin the racialethnic composition of District II personnel may have beenpurposefulstrategic in that this District contains a large rural-agriculturalcommunity with a significant number of low-wage and migrant laborers whoare disproportionately of Hispanic background Importantly no District IIpersonnel had ever worked in District I and likewise no District I personnelhad ever worked in District II

After the purpose of the study and the questionnaire were discussed andapproved by the top command of both the HCSO and each District a 149-itemquestionnaire was distributed and completed during roll call for each of the fiveshifts during consecutive days in August 1997 at District I and in March 1999at District II[3] Respondents were assured that the information they providedus would be kept confidential that they would remain anonymous and thattheir participation was voluntary Approximately one-third (38 percent fromDistrict I and 33 percent from District II) of the sworn employees wereunavailable for participation on the days of the survey administration due tocourt training illness military duty day-off or other personal leave This

The myth()of the policesub-culture

93

attrition reduced the total number of eligible respondents from both Districts to218 of whom only four declined to participate giving us a response rate of 98percent[4] However several respondents chose not to answer every questiontypically these were the socio-demographic and other ` identifierrsquorsquo items such asage raceethnicity gender rank shift and assignment The questionnaire tookapproximately 35-40 minutes to complete

Measures of adherence to the police sub-cultureRespondentsrsquo degree of adherence to the sub-culture of policing is measured byfive distinct workrole orientation scales crime control service cynicismtraditionalism and receptivity to change The first of these crime controlorientation measures the importance deputies place on the law enforcementand crime control functions of their jobs The 13-item additive scale assessesthe degree to which these deputies agree or disagree (1 = strongly disagree 5 =strongly agree) with a series of Likert-type statements such as `All lawsshould be fully enforced at all times otherwise people lose respect for the lawrsquorsquo` Law enforcement officers should not forget that enforcing the laws is by fartheir most important responsibilityrsquorsquo ` Most law enforcement officers have tospend too much of their time handling unimportant non-crime calls forservicersquorsquo and ` If law enforcement officers act in a service capacity it detractsfrom their ability to fight crimersquorsquo Item wording and frequency distributions forthese 13 items are presented in the Appendix

Several of the 13 items which comprise this scale were reverse-coded beforethey were entered into a principal components factor analysis from which fourfactors yielded eigenvalues greater than 100 however the scree discontinuitytest suggests that a single factor solution best represents these data Loadingson this first factor ranged from 038 to 075 (with three items loading at 016019 and 027) Despite these three weak loadings the Cronbachrsquos Alphareliability coefficient for the additive was 076 and does not improveappreciably if any or all of these three items are deleted High values on thisscale indicate a strong preference among these respondents for the crime-fighting aspects of their roles as law enforcement officers

Respondentsrsquo work orientation toward service-related activities an indicatorof disagreement with the police sub-culture is measured by a 14-item additivescale which was designed to assess the extent to which these deputies agreeand disagree (1 = strongly disagree 5 = strongly agree) with a series of Likert-type statements such as ` Law enforcement officers should ask citizens whattypes of services they wantrsquorsquo `Assisting citizens is just as important asenforcing the lawrsquorsquo ` Law enforcement should be seen primarily as a service-oriented profession rather than a crime control professionrsquorsquo and ` Crimes areonly one of several problems about which law enforcement officers should beconcernedrsquorsquo Item wording and frequency distributions for these 14 items arealso presented in the Appendix

Again several of these 14 service orientation items required reverse codingbefore they were entered into a principal components factor analysis This

PIJPSM261

94

factor analysis produced five factors with eigenvalues greater than 100however the scree discontinuity tests suggest that a single factor solution bestrepresents these data Loadings on this first factor ranged from 036 to 063(with two items loading at only 020 and 028) The Cronbachrsquos Alpha reliabilityfor the additive scale produced from all 14 items was 074 and does not increaseappreciably if either or both of the poorly loading items isare removed Highvalues on this scale are indicative of a strong pro-service work orientationcommensurate with the ideals of community-oriented policing and antagonisticto the police sub-culture

Cynicism is a nine-item additive scale which measures the extent to whichthese deputies agree or disagree (1 = strongly disagree 5 = strongly agree)with another series of Likert-type statement such as ` Most people areuntrustworthy and dishonestrsquorsquo ` Law enforcement officers will never trustcitizens enough to work together effectivelyrsquorsquo ` Citizens will not trust lawenforcement officers enough to work together effectivelyrsquorsquo and ` Most peoplelack the proper level of respect for law enforcement officersrsquorsquo Item wordingand frequency distributions for these nine items are also presented in theAppendix

As with the other scales several of the items comprising this cynicism scalerequired reverse coding so that high values on it demonstrate a cynicalperception of the public and their effectiveness in contributing to the objectivesof law enforcement and community-oriented policing Prior to scaling theseitems were also entered into a principal components factor analysis This factoranalysis produced a single factor solution with loadings ranging from 049 to073 the Cronbachrsquos Alpha reliability coefficient for this nine-item additivescale was 083

Traditionalism is a six-item additive scale which measures the extent towhich officers participating in this study agree or disagree (1 = stronglydisagree 5 = strongly agree) with a short series or Likert-type statements suchas ` In an efficient organization power is distributed at the toprsquorsquo ` Participatorymanagement schemes really do not work within law enforcement agenciesrsquorsquo` The quasi-military structure is the most effective organizational type for lawenforcement agenciesrsquorsquo and ` Communication works best when it follows clearestablished channels from the top downrsquorsquo Item wording and frequencydistributions for these six items are also presented in the Appendix

Once again several items required reverse coding so that high values on thisadditive scale are indicative of a traditionalistauthoritarian orientation towardthe organizational hierarchy and distribution of power within a lawenforcement agency and is counter-intuitive to the ideals of decentralizationand officer-empowerment under community-oriented policing Prior to scalingthese six items were entered into a principal components factor analysis fromwhich a two-factor solution was produced however the scree discontinuity testsuggests that a single factor solution best fit these data Loadings on this firstfactor ranged from 035 to 064 The Cronbachrsquos Alpha reliability coefficient forthis additive traditionalism scale was only 054

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

90

Such an outlook leads sub-cultural adherents to ` maintain an edgersquorsquo while onthe street (Van Maanen 1974 p 118) Their world becomes an insular dualismof ` us versus themrsquorsquo (Rubenstein 1973 Sparrow et al 1990 Tauber 1970) Theycome to see themselves as the ` thin blue linersquorsquo that prevents society fromslipping permanently into moral decay and unrest In essence a primaryelement of the police sub-culture is a great level of cynicism and suspicionwhich is carried by them into their encounters with the citizens they have beenhired to serve and protect This in turn only adds to the tensions between thepolice and their community

Finally the police sub-culture stresses an orientation which simultaneouslygrants autonomy to each officer yet demands from them unyielding loyalty tothe group (Broderick 1977 Brown 1981 Manning 1995 Muir 1977 White1972) These strong norms of group loyalty are said to produce a ` code ofsilencersquorsquo among fellow officers under which they are not to ` snitchrsquorsquo on oneanother (Bitner 1967 Brown 1981 Chevigny 1995 Reiner 1985 Reuss-Ianni1983 Rubenstein 1973 Shearing 1981 Westley 1970) The code of silenceimpedes efforts to detect and investigate corruption and other forms of policedeviancemisconduct Hence not only are sub-cultural adherents suspiciousand cynical in their orientations toward the citizenry they are also suspiciousand distrusting of the police bureaucracy especially upper administrationThese officers tend to view the police bureaucracy as too politically motivatedas ineffective and burdensome and as punitive they see police officials as moreinclined to punish them for their procedural errors (Walker 1977) than toreward them for the substantive successes produced from these ` shortcutsrsquorsquo (iearrests confessions evidence tips etc) Thus sub-cultural adherents areencouraged by their group to ` lie lowrsquorsquoand ` fly under the radarrsquorsquo (Paoline et al2000 p 578) that is to remain as much as possible invisible to upperadministration They are expected to adopt a ` cover your assrsquorsquo attitude whichalso covers their partnersrsquo and other officersrsquo (Chevigny 1995 Westley 1970)These attitudes may also lead police officers to confront and negate any reformefforts especially community-oriented policing through group resistance(Greene et al 1994 Lurigio and Skogan 1994 Sadd and Grinc 1993)

No doubt there are additional elements of this police sub-culture which wehave neglected to include in our brief description Nevertheless our account isfairly complete and representative and is consistent with other efforts atdetailing it (Britz 1997 Crank 1998 Drummond 1976 Farkas and Manning1997 Fielding 1988 Herbert 1998 Muir 1977 Paoline et al 2000 Reuss-Ianni 1983) More to the point our review of the research literature revealsseveral interesting contradictions regarding the police sub-culture Firstmost of the research on the police sub-culture has been exploratory anddescriptive in nature These studies appear to have accepted the existence ofthis sub-culture without question and thus have sought to describe itscharacteristics A key element of these descriptive efforts has been the oft-repeated claim that the sub-culture is endemic to policing and that adherenceto it is widespread However a few studies have been more explanatory in

The myth()of the policesub-culture

91

nature and question whether or not such a sub-culture exists and if found tobe present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks (Jermier et al 1991 Paoline et al 2000) Studieswhich have observed its presence among some but not most officers typicallyproceed by developing a typology of officers based at least in part on thedegree of sub-cultural adherence (Brown 1981 Muir 1977 Reiner 1978Wilson 1968) Rarely however are these typologies the product of analyticstatistical procedures designed for their construction (eg cluster analysis)nor are these typologies subsequently tested for their empirical validity(Bailey 1994 plusmn see Jermier et al (1991) for an exception) This brings us to thethreefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture in terms ofthe deputiesrsquo workrole orientations toward their crime control andservice functions their level of cynicism traditionalism and receptivityto change

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Data and methodsThe data for this study were obtained from self-administered questionnairesdistributed to patrol deputies and other sworn employees in Districts I and II ofthe Hillsborough County FL Sheriffrsquorsquos Office (HCSO) Districts I and II are twoof four decentralized semi-autonomous operational units within the HCSO TheHCSO is a full service sheriffrsquos office which includes law enforcement jailoperations and bailiff and other court-related functions At the time of theadministration of the data instrument (Summer 1997 and Spring 1999) theHCSO comprised 2561 personnel consisting of 970 law enforcement deputies768 jail deputies and 849 civilian employees the rest being administrative linesand court personnel making it one of the largest sheriffrsquos offices in the country(Reaves 1992)

The HCSO provides law enforcement services to the residents of theunincorporated portions of Hillsborough County Florida Hillsborough County islocated in west central Florida and has a total population of approximately900000 The county seat is Tampa with a population of nearly 300000 Theunincorporated area of the county is about 100 sq miles in size has a population

PIJPSM261

92

of approximately 500000 and comprises diverse communities including severalhigh-crimelow-income public housing areas migrant-worker and rural-agricultural communities upper-middle class ` bedroomrsquorsquo communities and anumber of working- and middle-class areas These areas also include severallarge shopping malls many smaller retail and service-oriented enterprises auniversity hospitals and other health-care facilities several parks andrecreational areas and a wide range of commercial and industrial facilities

The two HSCO Districts we examined were purposively selected becausetheir operational areas are most similar to those typical of metropolitanmunicipal police departments (see Weisheit et al 1995 Christensen and Crank2001)[2] Again both districts are semi-autonomous multi-faceted operationalunits within which the SOrsquos effort to integrate community-oriented-agencygroups wide had been well established since 1994 District I which servesnorthern Hillsborough county an area of about 79 sq miles with a populationof approximately 50000 comprises 137 sworn employees of whom six were inupper-administrative positions 24 in mid-level management (ie sergeants andcorporals) six were detectives and 101 were patrol deputies These personnelare 84 percent male 80 percent white 13 percent black and 7 percent HispanicDistrict I is dominated by the presence of a large and unstable public housingarea which almost entirely comprises poor black households and is plaguedwith a significant crime problem

District II serves northeastern Hillsborough County a much larger landmass (259 sq miles) and larger population (approximately 155000) than that ofDistrict I District II comprises 193 sworn personnel of whom five hold upper-administrative positions 24 are in mid-level management 14 are detectivesand 150 are patrol deputies The demographic profile of these sworn personnelin District II very closely resembles that of District I 84 percent male and 78percent white However District II has a higher percentage of Hispanic officers(11 percent) and a lower percentage of black officers (9 percent) This variationin the racialethnic composition of District II personnel may have beenpurposefulstrategic in that this District contains a large rural-agriculturalcommunity with a significant number of low-wage and migrant laborers whoare disproportionately of Hispanic background Importantly no District IIpersonnel had ever worked in District I and likewise no District I personnelhad ever worked in District II

After the purpose of the study and the questionnaire were discussed andapproved by the top command of both the HCSO and each District a 149-itemquestionnaire was distributed and completed during roll call for each of the fiveshifts during consecutive days in August 1997 at District I and in March 1999at District II[3] Respondents were assured that the information they providedus would be kept confidential that they would remain anonymous and thattheir participation was voluntary Approximately one-third (38 percent fromDistrict I and 33 percent from District II) of the sworn employees wereunavailable for participation on the days of the survey administration due tocourt training illness military duty day-off or other personal leave This

The myth()of the policesub-culture

93

attrition reduced the total number of eligible respondents from both Districts to218 of whom only four declined to participate giving us a response rate of 98percent[4] However several respondents chose not to answer every questiontypically these were the socio-demographic and other ` identifierrsquorsquo items such asage raceethnicity gender rank shift and assignment The questionnaire tookapproximately 35-40 minutes to complete

Measures of adherence to the police sub-cultureRespondentsrsquo degree of adherence to the sub-culture of policing is measured byfive distinct workrole orientation scales crime control service cynicismtraditionalism and receptivity to change The first of these crime controlorientation measures the importance deputies place on the law enforcementand crime control functions of their jobs The 13-item additive scale assessesthe degree to which these deputies agree or disagree (1 = strongly disagree 5 =strongly agree) with a series of Likert-type statements such as `All lawsshould be fully enforced at all times otherwise people lose respect for the lawrsquorsquo` Law enforcement officers should not forget that enforcing the laws is by fartheir most important responsibilityrsquorsquo ` Most law enforcement officers have tospend too much of their time handling unimportant non-crime calls forservicersquorsquo and ` If law enforcement officers act in a service capacity it detractsfrom their ability to fight crimersquorsquo Item wording and frequency distributions forthese 13 items are presented in the Appendix

Several of the 13 items which comprise this scale were reverse-coded beforethey were entered into a principal components factor analysis from which fourfactors yielded eigenvalues greater than 100 however the scree discontinuitytest suggests that a single factor solution best represents these data Loadingson this first factor ranged from 038 to 075 (with three items loading at 016019 and 027) Despite these three weak loadings the Cronbachrsquos Alphareliability coefficient for the additive was 076 and does not improveappreciably if any or all of these three items are deleted High values on thisscale indicate a strong preference among these respondents for the crime-fighting aspects of their roles as law enforcement officers

Respondentsrsquo work orientation toward service-related activities an indicatorof disagreement with the police sub-culture is measured by a 14-item additivescale which was designed to assess the extent to which these deputies agreeand disagree (1 = strongly disagree 5 = strongly agree) with a series of Likert-type statements such as ` Law enforcement officers should ask citizens whattypes of services they wantrsquorsquo `Assisting citizens is just as important asenforcing the lawrsquorsquo ` Law enforcement should be seen primarily as a service-oriented profession rather than a crime control professionrsquorsquo and ` Crimes areonly one of several problems about which law enforcement officers should beconcernedrsquorsquo Item wording and frequency distributions for these 14 items arealso presented in the Appendix

Again several of these 14 service orientation items required reverse codingbefore they were entered into a principal components factor analysis This

PIJPSM261

94

factor analysis produced five factors with eigenvalues greater than 100however the scree discontinuity tests suggest that a single factor solution bestrepresents these data Loadings on this first factor ranged from 036 to 063(with two items loading at only 020 and 028) The Cronbachrsquos Alpha reliabilityfor the additive scale produced from all 14 items was 074 and does not increaseappreciably if either or both of the poorly loading items isare removed Highvalues on this scale are indicative of a strong pro-service work orientationcommensurate with the ideals of community-oriented policing and antagonisticto the police sub-culture

Cynicism is a nine-item additive scale which measures the extent to whichthese deputies agree or disagree (1 = strongly disagree 5 = strongly agree)with another series of Likert-type statement such as ` Most people areuntrustworthy and dishonestrsquorsquo ` Law enforcement officers will never trustcitizens enough to work together effectivelyrsquorsquo ` Citizens will not trust lawenforcement officers enough to work together effectivelyrsquorsquo and ` Most peoplelack the proper level of respect for law enforcement officersrsquorsquo Item wordingand frequency distributions for these nine items are also presented in theAppendix

As with the other scales several of the items comprising this cynicism scalerequired reverse coding so that high values on it demonstrate a cynicalperception of the public and their effectiveness in contributing to the objectivesof law enforcement and community-oriented policing Prior to scaling theseitems were also entered into a principal components factor analysis This factoranalysis produced a single factor solution with loadings ranging from 049 to073 the Cronbachrsquos Alpha reliability coefficient for this nine-item additivescale was 083

Traditionalism is a six-item additive scale which measures the extent towhich officers participating in this study agree or disagree (1 = stronglydisagree 5 = strongly agree) with a short series or Likert-type statements suchas ` In an efficient organization power is distributed at the toprsquorsquo ` Participatorymanagement schemes really do not work within law enforcement agenciesrsquorsquo` The quasi-military structure is the most effective organizational type for lawenforcement agenciesrsquorsquo and ` Communication works best when it follows clearestablished channels from the top downrsquorsquo Item wording and frequencydistributions for these six items are also presented in the Appendix

Once again several items required reverse coding so that high values on thisadditive scale are indicative of a traditionalistauthoritarian orientation towardthe organizational hierarchy and distribution of power within a lawenforcement agency and is counter-intuitive to the ideals of decentralizationand officer-empowerment under community-oriented policing Prior to scalingthese six items were entered into a principal components factor analysis fromwhich a two-factor solution was produced however the scree discontinuity testsuggests that a single factor solution best fit these data Loadings on this firstfactor ranged from 035 to 064 The Cronbachrsquos Alpha reliability coefficient forthis additive traditionalism scale was only 054

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

91

nature and question whether or not such a sub-culture exists and if found tobe present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks (Jermier et al 1991 Paoline et al 2000) Studieswhich have observed its presence among some but not most officers typicallyproceed by developing a typology of officers based at least in part on thedegree of sub-cultural adherence (Brown 1981 Muir 1977 Reiner 1978Wilson 1968) Rarely however are these typologies the product of analyticstatistical procedures designed for their construction (eg cluster analysis)nor are these typologies subsequently tested for their empirical validity(Bailey 1994 plusmn see Jermier et al (1991) for an exception) This brings us to thethreefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture in terms ofthe deputiesrsquo workrole orientations toward their crime control andservice functions their level of cynicism traditionalism and receptivityto change

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Data and methodsThe data for this study were obtained from self-administered questionnairesdistributed to patrol deputies and other sworn employees in Districts I and II ofthe Hillsborough County FL Sheriffrsquorsquos Office (HCSO) Districts I and II are twoof four decentralized semi-autonomous operational units within the HCSO TheHCSO is a full service sheriffrsquos office which includes law enforcement jailoperations and bailiff and other court-related functions At the time of theadministration of the data instrument (Summer 1997 and Spring 1999) theHCSO comprised 2561 personnel consisting of 970 law enforcement deputies768 jail deputies and 849 civilian employees the rest being administrative linesand court personnel making it one of the largest sheriffrsquos offices in the country(Reaves 1992)

The HCSO provides law enforcement services to the residents of theunincorporated portions of Hillsborough County Florida Hillsborough County islocated in west central Florida and has a total population of approximately900000 The county seat is Tampa with a population of nearly 300000 Theunincorporated area of the county is about 100 sq miles in size has a population

PIJPSM261

92

of approximately 500000 and comprises diverse communities including severalhigh-crimelow-income public housing areas migrant-worker and rural-agricultural communities upper-middle class ` bedroomrsquorsquo communities and anumber of working- and middle-class areas These areas also include severallarge shopping malls many smaller retail and service-oriented enterprises auniversity hospitals and other health-care facilities several parks andrecreational areas and a wide range of commercial and industrial facilities

The two HSCO Districts we examined were purposively selected becausetheir operational areas are most similar to those typical of metropolitanmunicipal police departments (see Weisheit et al 1995 Christensen and Crank2001)[2] Again both districts are semi-autonomous multi-faceted operationalunits within which the SOrsquos effort to integrate community-oriented-agencygroups wide had been well established since 1994 District I which servesnorthern Hillsborough county an area of about 79 sq miles with a populationof approximately 50000 comprises 137 sworn employees of whom six were inupper-administrative positions 24 in mid-level management (ie sergeants andcorporals) six were detectives and 101 were patrol deputies These personnelare 84 percent male 80 percent white 13 percent black and 7 percent HispanicDistrict I is dominated by the presence of a large and unstable public housingarea which almost entirely comprises poor black households and is plaguedwith a significant crime problem

District II serves northeastern Hillsborough County a much larger landmass (259 sq miles) and larger population (approximately 155000) than that ofDistrict I District II comprises 193 sworn personnel of whom five hold upper-administrative positions 24 are in mid-level management 14 are detectivesand 150 are patrol deputies The demographic profile of these sworn personnelin District II very closely resembles that of District I 84 percent male and 78percent white However District II has a higher percentage of Hispanic officers(11 percent) and a lower percentage of black officers (9 percent) This variationin the racialethnic composition of District II personnel may have beenpurposefulstrategic in that this District contains a large rural-agriculturalcommunity with a significant number of low-wage and migrant laborers whoare disproportionately of Hispanic background Importantly no District IIpersonnel had ever worked in District I and likewise no District I personnelhad ever worked in District II

After the purpose of the study and the questionnaire were discussed andapproved by the top command of both the HCSO and each District a 149-itemquestionnaire was distributed and completed during roll call for each of the fiveshifts during consecutive days in August 1997 at District I and in March 1999at District II[3] Respondents were assured that the information they providedus would be kept confidential that they would remain anonymous and thattheir participation was voluntary Approximately one-third (38 percent fromDistrict I and 33 percent from District II) of the sworn employees wereunavailable for participation on the days of the survey administration due tocourt training illness military duty day-off or other personal leave This

The myth()of the policesub-culture

93

attrition reduced the total number of eligible respondents from both Districts to218 of whom only four declined to participate giving us a response rate of 98percent[4] However several respondents chose not to answer every questiontypically these were the socio-demographic and other ` identifierrsquorsquo items such asage raceethnicity gender rank shift and assignment The questionnaire tookapproximately 35-40 minutes to complete

Measures of adherence to the police sub-cultureRespondentsrsquo degree of adherence to the sub-culture of policing is measured byfive distinct workrole orientation scales crime control service cynicismtraditionalism and receptivity to change The first of these crime controlorientation measures the importance deputies place on the law enforcementand crime control functions of their jobs The 13-item additive scale assessesthe degree to which these deputies agree or disagree (1 = strongly disagree 5 =strongly agree) with a series of Likert-type statements such as `All lawsshould be fully enforced at all times otherwise people lose respect for the lawrsquorsquo` Law enforcement officers should not forget that enforcing the laws is by fartheir most important responsibilityrsquorsquo ` Most law enforcement officers have tospend too much of their time handling unimportant non-crime calls forservicersquorsquo and ` If law enforcement officers act in a service capacity it detractsfrom their ability to fight crimersquorsquo Item wording and frequency distributions forthese 13 items are presented in the Appendix

Several of the 13 items which comprise this scale were reverse-coded beforethey were entered into a principal components factor analysis from which fourfactors yielded eigenvalues greater than 100 however the scree discontinuitytest suggests that a single factor solution best represents these data Loadingson this first factor ranged from 038 to 075 (with three items loading at 016019 and 027) Despite these three weak loadings the Cronbachrsquos Alphareliability coefficient for the additive was 076 and does not improveappreciably if any or all of these three items are deleted High values on thisscale indicate a strong preference among these respondents for the crime-fighting aspects of their roles as law enforcement officers

Respondentsrsquo work orientation toward service-related activities an indicatorof disagreement with the police sub-culture is measured by a 14-item additivescale which was designed to assess the extent to which these deputies agreeand disagree (1 = strongly disagree 5 = strongly agree) with a series of Likert-type statements such as ` Law enforcement officers should ask citizens whattypes of services they wantrsquorsquo `Assisting citizens is just as important asenforcing the lawrsquorsquo ` Law enforcement should be seen primarily as a service-oriented profession rather than a crime control professionrsquorsquo and ` Crimes areonly one of several problems about which law enforcement officers should beconcernedrsquorsquo Item wording and frequency distributions for these 14 items arealso presented in the Appendix

Again several of these 14 service orientation items required reverse codingbefore they were entered into a principal components factor analysis This

PIJPSM261

94

factor analysis produced five factors with eigenvalues greater than 100however the scree discontinuity tests suggest that a single factor solution bestrepresents these data Loadings on this first factor ranged from 036 to 063(with two items loading at only 020 and 028) The Cronbachrsquos Alpha reliabilityfor the additive scale produced from all 14 items was 074 and does not increaseappreciably if either or both of the poorly loading items isare removed Highvalues on this scale are indicative of a strong pro-service work orientationcommensurate with the ideals of community-oriented policing and antagonisticto the police sub-culture

Cynicism is a nine-item additive scale which measures the extent to whichthese deputies agree or disagree (1 = strongly disagree 5 = strongly agree)with another series of Likert-type statement such as ` Most people areuntrustworthy and dishonestrsquorsquo ` Law enforcement officers will never trustcitizens enough to work together effectivelyrsquorsquo ` Citizens will not trust lawenforcement officers enough to work together effectivelyrsquorsquo and ` Most peoplelack the proper level of respect for law enforcement officersrsquorsquo Item wordingand frequency distributions for these nine items are also presented in theAppendix

As with the other scales several of the items comprising this cynicism scalerequired reverse coding so that high values on it demonstrate a cynicalperception of the public and their effectiveness in contributing to the objectivesof law enforcement and community-oriented policing Prior to scaling theseitems were also entered into a principal components factor analysis This factoranalysis produced a single factor solution with loadings ranging from 049 to073 the Cronbachrsquos Alpha reliability coefficient for this nine-item additivescale was 083

Traditionalism is a six-item additive scale which measures the extent towhich officers participating in this study agree or disagree (1 = stronglydisagree 5 = strongly agree) with a short series or Likert-type statements suchas ` In an efficient organization power is distributed at the toprsquorsquo ` Participatorymanagement schemes really do not work within law enforcement agenciesrsquorsquo` The quasi-military structure is the most effective organizational type for lawenforcement agenciesrsquorsquo and ` Communication works best when it follows clearestablished channels from the top downrsquorsquo Item wording and frequencydistributions for these six items are also presented in the Appendix

Once again several items required reverse coding so that high values on thisadditive scale are indicative of a traditionalistauthoritarian orientation towardthe organizational hierarchy and distribution of power within a lawenforcement agency and is counter-intuitive to the ideals of decentralizationand officer-empowerment under community-oriented policing Prior to scalingthese six items were entered into a principal components factor analysis fromwhich a two-factor solution was produced however the scree discontinuity testsuggests that a single factor solution best fit these data Loadings on this firstfactor ranged from 035 to 064 The Cronbachrsquos Alpha reliability coefficient forthis additive traditionalism scale was only 054

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

92

of approximately 500000 and comprises diverse communities including severalhigh-crimelow-income public housing areas migrant-worker and rural-agricultural communities upper-middle class ` bedroomrsquorsquo communities and anumber of working- and middle-class areas These areas also include severallarge shopping malls many smaller retail and service-oriented enterprises auniversity hospitals and other health-care facilities several parks andrecreational areas and a wide range of commercial and industrial facilities

The two HSCO Districts we examined were purposively selected becausetheir operational areas are most similar to those typical of metropolitanmunicipal police departments (see Weisheit et al 1995 Christensen and Crank2001)[2] Again both districts are semi-autonomous multi-faceted operationalunits within which the SOrsquos effort to integrate community-oriented-agencygroups wide had been well established since 1994 District I which servesnorthern Hillsborough county an area of about 79 sq miles with a populationof approximately 50000 comprises 137 sworn employees of whom six were inupper-administrative positions 24 in mid-level management (ie sergeants andcorporals) six were detectives and 101 were patrol deputies These personnelare 84 percent male 80 percent white 13 percent black and 7 percent HispanicDistrict I is dominated by the presence of a large and unstable public housingarea which almost entirely comprises poor black households and is plaguedwith a significant crime problem

District II serves northeastern Hillsborough County a much larger landmass (259 sq miles) and larger population (approximately 155000) than that ofDistrict I District II comprises 193 sworn personnel of whom five hold upper-administrative positions 24 are in mid-level management 14 are detectivesand 150 are patrol deputies The demographic profile of these sworn personnelin District II very closely resembles that of District I 84 percent male and 78percent white However District II has a higher percentage of Hispanic officers(11 percent) and a lower percentage of black officers (9 percent) This variationin the racialethnic composition of District II personnel may have beenpurposefulstrategic in that this District contains a large rural-agriculturalcommunity with a significant number of low-wage and migrant laborers whoare disproportionately of Hispanic background Importantly no District IIpersonnel had ever worked in District I and likewise no District I personnelhad ever worked in District II

After the purpose of the study and the questionnaire were discussed andapproved by the top command of both the HCSO and each District a 149-itemquestionnaire was distributed and completed during roll call for each of the fiveshifts during consecutive days in August 1997 at District I and in March 1999at District II[3] Respondents were assured that the information they providedus would be kept confidential that they would remain anonymous and thattheir participation was voluntary Approximately one-third (38 percent fromDistrict I and 33 percent from District II) of the sworn employees wereunavailable for participation on the days of the survey administration due tocourt training illness military duty day-off or other personal leave This

The myth()of the policesub-culture

93

attrition reduced the total number of eligible respondents from both Districts to218 of whom only four declined to participate giving us a response rate of 98percent[4] However several respondents chose not to answer every questiontypically these were the socio-demographic and other ` identifierrsquorsquo items such asage raceethnicity gender rank shift and assignment The questionnaire tookapproximately 35-40 minutes to complete

Measures of adherence to the police sub-cultureRespondentsrsquo degree of adherence to the sub-culture of policing is measured byfive distinct workrole orientation scales crime control service cynicismtraditionalism and receptivity to change The first of these crime controlorientation measures the importance deputies place on the law enforcementand crime control functions of their jobs The 13-item additive scale assessesthe degree to which these deputies agree or disagree (1 = strongly disagree 5 =strongly agree) with a series of Likert-type statements such as `All lawsshould be fully enforced at all times otherwise people lose respect for the lawrsquorsquo` Law enforcement officers should not forget that enforcing the laws is by fartheir most important responsibilityrsquorsquo ` Most law enforcement officers have tospend too much of their time handling unimportant non-crime calls forservicersquorsquo and ` If law enforcement officers act in a service capacity it detractsfrom their ability to fight crimersquorsquo Item wording and frequency distributions forthese 13 items are presented in the Appendix

Several of the 13 items which comprise this scale were reverse-coded beforethey were entered into a principal components factor analysis from which fourfactors yielded eigenvalues greater than 100 however the scree discontinuitytest suggests that a single factor solution best represents these data Loadingson this first factor ranged from 038 to 075 (with three items loading at 016019 and 027) Despite these three weak loadings the Cronbachrsquos Alphareliability coefficient for the additive was 076 and does not improveappreciably if any or all of these three items are deleted High values on thisscale indicate a strong preference among these respondents for the crime-fighting aspects of their roles as law enforcement officers

Respondentsrsquo work orientation toward service-related activities an indicatorof disagreement with the police sub-culture is measured by a 14-item additivescale which was designed to assess the extent to which these deputies agreeand disagree (1 = strongly disagree 5 = strongly agree) with a series of Likert-type statements such as ` Law enforcement officers should ask citizens whattypes of services they wantrsquorsquo `Assisting citizens is just as important asenforcing the lawrsquorsquo ` Law enforcement should be seen primarily as a service-oriented profession rather than a crime control professionrsquorsquo and ` Crimes areonly one of several problems about which law enforcement officers should beconcernedrsquorsquo Item wording and frequency distributions for these 14 items arealso presented in the Appendix

Again several of these 14 service orientation items required reverse codingbefore they were entered into a principal components factor analysis This

PIJPSM261

94

factor analysis produced five factors with eigenvalues greater than 100however the scree discontinuity tests suggest that a single factor solution bestrepresents these data Loadings on this first factor ranged from 036 to 063(with two items loading at only 020 and 028) The Cronbachrsquos Alpha reliabilityfor the additive scale produced from all 14 items was 074 and does not increaseappreciably if either or both of the poorly loading items isare removed Highvalues on this scale are indicative of a strong pro-service work orientationcommensurate with the ideals of community-oriented policing and antagonisticto the police sub-culture

Cynicism is a nine-item additive scale which measures the extent to whichthese deputies agree or disagree (1 = strongly disagree 5 = strongly agree)with another series of Likert-type statement such as ` Most people areuntrustworthy and dishonestrsquorsquo ` Law enforcement officers will never trustcitizens enough to work together effectivelyrsquorsquo ` Citizens will not trust lawenforcement officers enough to work together effectivelyrsquorsquo and ` Most peoplelack the proper level of respect for law enforcement officersrsquorsquo Item wordingand frequency distributions for these nine items are also presented in theAppendix

As with the other scales several of the items comprising this cynicism scalerequired reverse coding so that high values on it demonstrate a cynicalperception of the public and their effectiveness in contributing to the objectivesof law enforcement and community-oriented policing Prior to scaling theseitems were also entered into a principal components factor analysis This factoranalysis produced a single factor solution with loadings ranging from 049 to073 the Cronbachrsquos Alpha reliability coefficient for this nine-item additivescale was 083

Traditionalism is a six-item additive scale which measures the extent towhich officers participating in this study agree or disagree (1 = stronglydisagree 5 = strongly agree) with a short series or Likert-type statements suchas ` In an efficient organization power is distributed at the toprsquorsquo ` Participatorymanagement schemes really do not work within law enforcement agenciesrsquorsquo` The quasi-military structure is the most effective organizational type for lawenforcement agenciesrsquorsquo and ` Communication works best when it follows clearestablished channels from the top downrsquorsquo Item wording and frequencydistributions for these six items are also presented in the Appendix

Once again several items required reverse coding so that high values on thisadditive scale are indicative of a traditionalistauthoritarian orientation towardthe organizational hierarchy and distribution of power within a lawenforcement agency and is counter-intuitive to the ideals of decentralizationand officer-empowerment under community-oriented policing Prior to scalingthese six items were entered into a principal components factor analysis fromwhich a two-factor solution was produced however the scree discontinuity testsuggests that a single factor solution best fit these data Loadings on this firstfactor ranged from 035 to 064 The Cronbachrsquos Alpha reliability coefficient forthis additive traditionalism scale was only 054

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

93

attrition reduced the total number of eligible respondents from both Districts to218 of whom only four declined to participate giving us a response rate of 98percent[4] However several respondents chose not to answer every questiontypically these were the socio-demographic and other ` identifierrsquorsquo items such asage raceethnicity gender rank shift and assignment The questionnaire tookapproximately 35-40 minutes to complete

Measures of adherence to the police sub-cultureRespondentsrsquo degree of adherence to the sub-culture of policing is measured byfive distinct workrole orientation scales crime control service cynicismtraditionalism and receptivity to change The first of these crime controlorientation measures the importance deputies place on the law enforcementand crime control functions of their jobs The 13-item additive scale assessesthe degree to which these deputies agree or disagree (1 = strongly disagree 5 =strongly agree) with a series of Likert-type statements such as `All lawsshould be fully enforced at all times otherwise people lose respect for the lawrsquorsquo` Law enforcement officers should not forget that enforcing the laws is by fartheir most important responsibilityrsquorsquo ` Most law enforcement officers have tospend too much of their time handling unimportant non-crime calls forservicersquorsquo and ` If law enforcement officers act in a service capacity it detractsfrom their ability to fight crimersquorsquo Item wording and frequency distributions forthese 13 items are presented in the Appendix

Several of the 13 items which comprise this scale were reverse-coded beforethey were entered into a principal components factor analysis from which fourfactors yielded eigenvalues greater than 100 however the scree discontinuitytest suggests that a single factor solution best represents these data Loadingson this first factor ranged from 038 to 075 (with three items loading at 016019 and 027) Despite these three weak loadings the Cronbachrsquos Alphareliability coefficient for the additive was 076 and does not improveappreciably if any or all of these three items are deleted High values on thisscale indicate a strong preference among these respondents for the crime-fighting aspects of their roles as law enforcement officers

Respondentsrsquo work orientation toward service-related activities an indicatorof disagreement with the police sub-culture is measured by a 14-item additivescale which was designed to assess the extent to which these deputies agreeand disagree (1 = strongly disagree 5 = strongly agree) with a series of Likert-type statements such as ` Law enforcement officers should ask citizens whattypes of services they wantrsquorsquo `Assisting citizens is just as important asenforcing the lawrsquorsquo ` Law enforcement should be seen primarily as a service-oriented profession rather than a crime control professionrsquorsquo and ` Crimes areonly one of several problems about which law enforcement officers should beconcernedrsquorsquo Item wording and frequency distributions for these 14 items arealso presented in the Appendix

Again several of these 14 service orientation items required reverse codingbefore they were entered into a principal components factor analysis This

PIJPSM261

94

factor analysis produced five factors with eigenvalues greater than 100however the scree discontinuity tests suggest that a single factor solution bestrepresents these data Loadings on this first factor ranged from 036 to 063(with two items loading at only 020 and 028) The Cronbachrsquos Alpha reliabilityfor the additive scale produced from all 14 items was 074 and does not increaseappreciably if either or both of the poorly loading items isare removed Highvalues on this scale are indicative of a strong pro-service work orientationcommensurate with the ideals of community-oriented policing and antagonisticto the police sub-culture

Cynicism is a nine-item additive scale which measures the extent to whichthese deputies agree or disagree (1 = strongly disagree 5 = strongly agree)with another series of Likert-type statement such as ` Most people areuntrustworthy and dishonestrsquorsquo ` Law enforcement officers will never trustcitizens enough to work together effectivelyrsquorsquo ` Citizens will not trust lawenforcement officers enough to work together effectivelyrsquorsquo and ` Most peoplelack the proper level of respect for law enforcement officersrsquorsquo Item wordingand frequency distributions for these nine items are also presented in theAppendix

As with the other scales several of the items comprising this cynicism scalerequired reverse coding so that high values on it demonstrate a cynicalperception of the public and their effectiveness in contributing to the objectivesof law enforcement and community-oriented policing Prior to scaling theseitems were also entered into a principal components factor analysis This factoranalysis produced a single factor solution with loadings ranging from 049 to073 the Cronbachrsquos Alpha reliability coefficient for this nine-item additivescale was 083

Traditionalism is a six-item additive scale which measures the extent towhich officers participating in this study agree or disagree (1 = stronglydisagree 5 = strongly agree) with a short series or Likert-type statements suchas ` In an efficient organization power is distributed at the toprsquorsquo ` Participatorymanagement schemes really do not work within law enforcement agenciesrsquorsquo` The quasi-military structure is the most effective organizational type for lawenforcement agenciesrsquorsquo and ` Communication works best when it follows clearestablished channels from the top downrsquorsquo Item wording and frequencydistributions for these six items are also presented in the Appendix

Once again several items required reverse coding so that high values on thisadditive scale are indicative of a traditionalistauthoritarian orientation towardthe organizational hierarchy and distribution of power within a lawenforcement agency and is counter-intuitive to the ideals of decentralizationand officer-empowerment under community-oriented policing Prior to scalingthese six items were entered into a principal components factor analysis fromwhich a two-factor solution was produced however the scree discontinuity testsuggests that a single factor solution best fit these data Loadings on this firstfactor ranged from 035 to 064 The Cronbachrsquos Alpha reliability coefficient forthis additive traditionalism scale was only 054

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

94

factor analysis produced five factors with eigenvalues greater than 100however the scree discontinuity tests suggest that a single factor solution bestrepresents these data Loadings on this first factor ranged from 036 to 063(with two items loading at only 020 and 028) The Cronbachrsquos Alpha reliabilityfor the additive scale produced from all 14 items was 074 and does not increaseappreciably if either or both of the poorly loading items isare removed Highvalues on this scale are indicative of a strong pro-service work orientationcommensurate with the ideals of community-oriented policing and antagonisticto the police sub-culture

Cynicism is a nine-item additive scale which measures the extent to whichthese deputies agree or disagree (1 = strongly disagree 5 = strongly agree)with another series of Likert-type statement such as ` Most people areuntrustworthy and dishonestrsquorsquo ` Law enforcement officers will never trustcitizens enough to work together effectivelyrsquorsquo ` Citizens will not trust lawenforcement officers enough to work together effectivelyrsquorsquo and ` Most peoplelack the proper level of respect for law enforcement officersrsquorsquo Item wordingand frequency distributions for these nine items are also presented in theAppendix

As with the other scales several of the items comprising this cynicism scalerequired reverse coding so that high values on it demonstrate a cynicalperception of the public and their effectiveness in contributing to the objectivesof law enforcement and community-oriented policing Prior to scaling theseitems were also entered into a principal components factor analysis This factoranalysis produced a single factor solution with loadings ranging from 049 to073 the Cronbachrsquos Alpha reliability coefficient for this nine-item additivescale was 083

Traditionalism is a six-item additive scale which measures the extent towhich officers participating in this study agree or disagree (1 = stronglydisagree 5 = strongly agree) with a short series or Likert-type statements suchas ` In an efficient organization power is distributed at the toprsquorsquo ` Participatorymanagement schemes really do not work within law enforcement agenciesrsquorsquo` The quasi-military structure is the most effective organizational type for lawenforcement agenciesrsquorsquo and ` Communication works best when it follows clearestablished channels from the top downrsquorsquo Item wording and frequencydistributions for these six items are also presented in the Appendix

Once again several items required reverse coding so that high values on thisadditive scale are indicative of a traditionalistauthoritarian orientation towardthe organizational hierarchy and distribution of power within a lawenforcement agency and is counter-intuitive to the ideals of decentralizationand officer-empowerment under community-oriented policing Prior to scalingthese six items were entered into a principal components factor analysis fromwhich a two-factor solution was produced however the scree discontinuity testsuggests that a single factor solution best fit these data Loadings on this firstfactor ranged from 035 to 064 The Cronbachrsquos Alpha reliability coefficient forthis additive traditionalism scale was only 054

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

95

Lastly receptivity or openness to organizational change is an additive scalecomprising the following four Likert-type statement to which respondentswere asked to indicate the degree to which they agreed or disagreed (1 =strongly disagree 5 = strongly agree) ` Most changes at work are problematicand ineffectiversquorsquo plusmn this item was reversed-coded ` I often suggest newapproaches for doing my jobrsquorsquo ` Most changes make my work more efficient (iesaves time effort money)rsquorsquo and `Most changes make my work more effective(more arrests faster response times crime reduction)rsquorsquo Item wording andfrequency distributions for these four items are also presented in the Appendix

These four items measuring respondentsrsquo receptivity to organizational changewere entered into a principal components factor analysis prior to scaleconstruction The results suggested that a single factor solution best representedthe data (ie only the first factor yielded an eigenvalue greater than 100)Loadings on this single factor ranged from 046 to 089 The Cronbachrsquos Alphareliability coefficient for the additive scale was 072 High values on this scaleindicate a high level of receptivity to modern organizational changes occurringwithin law enforcement which are antagonistic to the police sub-culture

The existence of the traditional police sub-culture would be established by atleast a sub-set of deputies with high scores on the crime control cynicism andtraditionalism scales and low scores on the service and receptivity to changescales Furthermore based upon arguments raised by Paoline et al (2000) wecontend that female and minority officers since they are relatively new to theranks of law enforcement and thus constitute a form of demographic change tothe composition of personnel in law enforcement agencies should at leastadhere to the police sub-culture We also contend that adherence to the policesub-culture for the same reasons expressed above will be negativelyassociated with deputiesrsquo level of educational experience and rank (as codedhere) but positively associated with their age and level of experience

These demographicwork experience variables are measured as followsrespondentsrsquo age (in years) gender (0 = female 1 = male) raceethnicity (0 =minority 1 = white) level of educational attainment (a seven-point ordinal scaleranging from 1 = high school graduateGED to 7 = advancedgraduate degree) rank(0 = corporal or higher 1 = deputy) work experience with the HCSO (measured inmonths employed) and operational district (0 = District II 1 = District I)

Analytic strategy and techniquesGiven the threefold purpose of this study our analytic strategy and thestatistical techniques we employ need to be carefully described Our primarypurpose is to determine whether or not there is evidence of the traditionalmonolithic police sub-culture among the sheriffsrsquo deputies who participated inour survey and if so the extent of adherence to it As such this study attemptsin part to replicate and corroborate similar research conducted by Paoline et al(2000) That is we examine the frequency distributions and other univariatestatistics for each of the five scales constructed to operationalize keycomponents of the police sub-culture and for several of the representative

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

96

single-items measures which comprise these scales In addition we examine theconditional distributions of these five scales across several important socio-demographic and work-experience characteristics of these sheriffsrsquo deputies toassess the level of consensus for the police sub-culture

The second purpose of this study presuming we find variation in deputiesrsquoadherence to this sub-culture is to construct and replicate a taxonomy of policeofficers by making better use of advanced statistical procedures for doing sospecifically cluster analysis The third and final purpose is to validate thistaxonomy by examining the extent to which significant differences existbetween types across an array of socio-demographic and work experienceindicators In addition we attempt to validate the taxonomy with adiscriminant function analysis in which the taxonomy is employed as anominal scale dependent variable and is modeled by these socio-demographicand work experience characteristics of these deputies

To accomplish taxonomy construction and replication we first parse ourdata according to the operational districts from which the deputies areemployed District I data are used for taxonomy construction while District IIdata are used for replicating the taxonomy Those unfamiliar with themultivariate statistical techniques we employ namely cluster analysis fortaxonomy construction and replication and discriminant function analysis fortaxonomic validation will find very helpful descriptions by Bailey (1994) andAldenderfer and Blashfield (1984) regarding cluster analysis and by Klecka(1980) on discriminant function analysis

Cluster analysis for taxonomy construction Cluster analysis is a multi-dimensional statistical method of synchronic (cross-sectional) empiricalclassification of observations into a numerical taxonomy of polythetic classestypes We derive our dimensions (independent variables) theoretically from thepolice sub-culture literature Because these dimensions are based on fiveadditive scales the cluster analysis groups observations into classes that areempirically similar but not identical in their characteristics on these fiveequally weighted dimensions Hence our analyses yield mutually exclusiveclasses which are exhaustive of the observations in our data In the currentanalysis we employ an agglomerative hierarchical clustering techniqueSpecifically we employ Wardrsquos (1963) hierarchical clustering method

Hierarchical clustering models involve sequential iterative clustering andreclustering of the data until all observationsclusters are clustered into a singlecluster Thus agglomerative cluster analysis produces between one and Nclusters Various diagnostic information is then utilized to determine the ` bestrsquorsquonumber of ` truersquorsquo clusters In sum we employ SAHN clustering methods(Sneath and Sokol 1973) SAHN stands for sequential agglomerativehierarchical and non-overlapping

Cluster analysis produces one or more ` validrsquorsquo clusters with the raw data butit does not assist the researcher with interpretation (Bailey 1994 pp 61-3) It isup to the researcher to give interpretive meaning to the observed clusters Thisis done by examining the ` profilersquorsquo of each cluster and determining the

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

97

underlying character of the type produced The profile of a cluster refers to thewithin-cluster univariate statistics (means standard deviations etc) for each ofthe dimensionsvariables used in its construction From these values theclusters are ` interpretedrsquorsquo In the present case we are employing cluster analysisin a confirmatory manner That is we are examining these data for evidence ofthe police sub-culture Should an endemic police sub-culture be present then asingle cluster solution will be produced and the profile of this cluster willclosely resemble the characteristics of the police sub-culture Should only asubset of these sheriffsrsquo deputies adhere to the police sub-culture then clusteranalysis will produce a multi-cluster solution and the profile of one of theseclusters will closely resemble the characteristics of the police sub-culture Thepresence of a police sub-culture is thus revealed in a single cluster that ischaracterized by high mean values on the crime-control traditionalism andcynicism scales and conversely low mean values on the service and receptivityto change scales In addition this profile should also reveal limited variationaround these means in the form of small standard deviations The profiles ofany other clusters are of secondary importance to our purposes and theirinterpretations will be more challenging for us but we proceed in a similarmanner of identifying the characteristic or dimensional profile of each

Once a clustering solution has been identified it is necessary to validate it(Aldenderfer and Blashfield 1984) We do so through two separate processesreplication and association with external variables (predictive validity) Withregard to replication we apply cluster-analytic methods first to the dataderived from our sample of sheriffsrsquo deputies in District I We then apply thesesame techniques to the data derived from the sample in District II and examinethe degree to which the same method applied to different samples produced thesame clustering solution We also attempt to validate our clustering solution byexamining the predictive validity of the resultant classes We accomplish thistask by two different statistical techniques First we test for significantdifferences between the resultant classes across an array of external variables(variables not used as dimensions in the original construction of the clusters)measuring deputiesrsquo socio-demographic and work experience characteristicsSecond we test the degree to which these socio-demographic and workexperience characteristics accurately predict deputiesrsquo classification Because aclassification scheme is a nominal level dependent variable the method ofanalysis used is discriminant function analysis

Discriminant function analysis Discriminant function analysis is a statisticaltechnique for examining differences between two or more groups ofobservations with respect to several exogenous variables (Klecka 1980 p 7) Itcan be used to test the effects of exogenousdiscriminating variables on anominal dependent variable such as a typology andor to classifyobservations into the typology based upon their values on the exogenous ordiscriminating variables For the purposes of the current study we make use ofboth of these functions of discriminant function analysis That is we hope tovalidate the taxonomy of sheriffsrsquo deputies produced during the cluster

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

98

analysis Thus our exogenous variables (ie deputiesrsquo socio-demographic andwork experience characteristics) should effectively discriminate betweenclasses on the typology In addition the discriminant function analysisvalidates the taxonomy if it can successfully classify the same observationsinto these classes based upon first the same dimensions used in the clusteranalysis (ie indicators of adherence to the police sub-culture) andor secondother discriminating external variables (ie deputiesrsquo socio-demographic andwork experience characteristics) In doing so discriminant function analysisprovides evidence on whether or not each of these ` discriminatingrsquorsquo variables` discriminatesrsquorsquo and if so how well they ` discriminatersquorsquo

Observations once classified by discriminant function analysis can becompared with their ` truersquorsquo group classification and measures of classificationaccuracy generated In this way we can validate the classificatory efficacy of theoriginal dimensions (measures of deputiesrsquo adherence to the police sub-culture)used in the cluster analyses If our taxonomy of sheriffsrsquo deputies is valid thenother indicators of these deputiesrsquo characteristics should also effectivelydiscriminate between these groups thus establishing the predictive validity ofthe taxonomy Specifically we anticipate that age gender raceethnicity level ofeducational attainment rank amount of experience and operational districtshould each discriminate sub-cultural adherents if any from the rest of theofficers We anticipate that younger deputies females minorities those withhigher levels of education those with lower rank and fewer years of service willnot be substantially represented in the group of sub-cultural adherents

Results I evidence of a police sub-cultureOur first research question asks whether or not there is any evidence ofdeputiesrsquo adherence to the police sub-culture To answer this question wecompare the observed distributions on the five scales measuring sub-culturaladherence with their hypothetical or ` expectedrsquorsquo distribution which is assumedto be a perfect normal distribution with no skewness or kurtosis Thesedistributions are presented in Table I If sub-cultural adherence is widespreadthen we would expect the observed means to be significantly different from theexpected means for each scale moreover we would also expect the observeddistributions to be leptokurtic (ie highly peaked or narrowly packed aroundthe mean) Evidence of a significant minority of deputies who adhere to thispolice sub-culture but no evidence of widespread adherence would also berevealed by observed means that are significantly different from the expectedmeans on these scales The existence of this sub-set of sub-cultural adherentswould also be revealed in observed distributions that are highly skewed

The values presented in Table I suggest that while there is some evidence ofsub-cultural adherence among these sheriffsrsquo deputies this adherence is by nomeans widespread For instance the observed mean for the crime-control scale(319) is slightly below the ` expectedrsquorsquo mean for this scale (39) and the observedrange of values on this scale is much more narrow than expected (13 to 50 versus13 to 65) While the observed distribution is mesokurtic it is also very slightly

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

99

positively skewed suggestive perhaps of a small group of sub-cultural crime-fighters Overall these deputies do seem to value the crime-fighting aspects oftheir role but not singularly they also value the service order maintenancepeacekeeping and other non-crime control aspects of policing

This is also evident in the observed frequency distributions of the 13 itemswhich comprise this scale For instance only 28 percent of these deputies feelthat ` an aggressive tough bearing is more useful than is a friendly

Expected Observed

13-Item crime-control scaleRange 13-65 13-50Meanmedian 3939 31932Std dev 867 614Percent lt Mean + 1 std dev 84 855Percent lt Mean + 2 std dev 977 967Skewness 00 029Kurtosis 00 plusmn001

14-item service orientation scaleRange 14-70 31-69Meanmedian 4242 48749Std dev 933 604Percent lt Mean + 1 std dev 84 865Percent lt Mean + 2 std dev 977 972Skewness 00 plusmn002Kurtosis 00 057

Nine-item cynicism scaleRange 9-45 9-43Meanmedian 2727 2424Std dev 600 552Percent lt Mean + 1 std dev 84 885Percent lt Mean + 2 std dev 977 972Skewness 00 018Kurtosis 00 102

Six-item traditionalism scaleRange 6-30 9-28Meanmedian 1818 1818Std dev 400 295Percent lt Mean + 1 std dev 84 916Percent lt Mean + 2 std dev 977 981Skewness 00 002Kurtosis 00 087

Five-item receptivity to change scaleRange 5-25 5-23Meanmedian 1212 15716Std dev 433 274Percent lt Mean + 1 std dev 84 854Percent lt Mean + 2 std dev 977 986Skewness 00 plusmn019Kurtosis 00 055

Table IUnivariate statistics onmeasures of adherence

to police sub-culture

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

100

courteous mannerrsquorsquo and only 47 percent agree or strongly agree that ` problemsolving should not be a part of an officerrsquos responsibilityrsquorsquo On the other hand642 percent feel that ` enforcing the law is by far their most importantresponsibilityrsquorsquo and 60 percent agree or strongly agree that ` many of thedecisions by the courts interfere with [their] ability to fight crimersquorsquo Finally476 percent feel that they ` spend too much of their time handling unimportantnon-crime calls for servicersquorsquo

The strong service orientation of these deputies an orientation counter to thepolice sub-culture is also evident in the data provided in Table I While the` expectedrsquorsquo range of this 14-item additive scale is 14 to 70 the observed range ismuch narrower 31 to 69 In addition the observed mean (487) is above the` expectedrsquorsquo mean for this scale (42) and the observed distribution is somewhatpeaked or leptokurtic Thus rather than suggesting widespread adherence to thepolice sub-culture these univariate statistics suggest the contrary that is thesevalues are suggestive of widespread adherence to a strongly service-orientednouveau police sub-culture Evidence of this nouveau police sub-culture is alsofound in the frequency distributions for the 14 items which comprise the serviceorientation scale (see Appendix) For almost all of these items less than 30percent of the deputies oppose (disagree or strongly disagree with) the serviceaspects of the police role However 563 percent disagree or strongly disagreethat ` law enforcement should be seen primarily as a service-oriented professionrather than a crime control professionrsquorsquo

While these data are not very supportive of the crime-fighter elements of thepolice sub-culture other data in Table I do support other aspects of this sub-culture For instance these deputies or at least a subset of them are rathercynical The distribution on this nine-item cynicism scale is both leptokurtic orpeaked and slightly positively skewed The frequency distributions for the nineitems which comprise this scale (see Appendix) are also indicative of asomewhat jaded and cynical view of the public A substantial proportion ofthese deputies believe that ` most people liersquorsquo (439) a small majority see thepublic as uncaring (ie 528 percent disagree or strongly disagree with thestatement ` Most people do not hesitate to go out of their way to help someonein troublersquorsquo) 27 percent feel that `most people would steal if they knew theywouldnrsquot get caughtrsquorsquo and 371 percent feel that ` most people lack the properlevel of respect for law enforcement officersrsquorsquo

Similarly the data in Table I and the Appendix also suggest a degree of sub-cultural adherence with regard to deputiesrsquo support for the traditionalhierarchical organizational structure of the sheriffrsquos office For instance the datain Table I reveal a non-skewed leptokurtic distribution around an observedmean equal to the expected mean value for this traditionalism scale In fact 679percent of the deputies sampled believe that ` an organization should have clearlydefined positions of powerauthority among its membersemployeesrsquorsquo Likewise724 percent support a chain of command and communication that ` follows clearestablished channels from the top downrsquorsquo and 546 percent feel that the ` quasi-military structure is the most effective organizational type for law enforcementagenciesrsquorsquo However less than 15 percent of these deputies feel that ` subordinates

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

101

should not be involved in either the setting or enforcing of policies andproceduresrsquorsquo or agree with the statement ` participatory management schemesreally do not work within law enforcement agenciesrsquorsquo

Lastly data in Table I and the Appendix suggest mixed levels of adherenceto the police sub-culture with regard to deputiesrsquo receptivity to organizationalchange For this scale we observe a leptokurtic distribution surrounding anobserved mean greater than that expected for this scale suggesting thatoverall these deputies are quite receptive to organizational change howeverthis distribution is also somewhat negatively skewed suggesting the presenceof a sub-group of deputies who are not open to such changes and thus mayconstitute a group of sub-cultural adherents The frequency distributions forthe five items which comprise this scale (see Appendix) reveal a similarlymixed level of support for the police sub-culture on this dimension While thevast majority of deputies report being slightly open or at worst neutral to theissue of organizational change many of them are skeptical regarding theefficacy of such changes That is approximately 35 percent of these deputiesdoubt that organizational changes make their work more efficient or moreeffective

In sum we find very mixed evidence of adherence to a police sub-cultureamong these deputies at best there may be a small minority of these deputieswho represent this sub-culture On the whole our respondents are both crime-control- and community-service oriented though they give primacy to thecrime-control aspects of their work While they are open to organizationalchange they are also skeptical of such changes likewise they tend to besomewhat cynical They tend to support the traditional quasi-militaristichierarchical structure of police agencies but they also support a structurewhich provides for their input and participation in decision making Thus asobserved in each of the small number of the previous empirical attempts toestablish the existence of the police sub-culture (Haarr 1997 Herbert 1998Jermier et al 1991 Manning 1995 Paoline et al 2000 see also Worden(1995) we also find that at best it is present in only a small sub-set ofdeputies Before we close this portion of our study we attempt to identify thesocio-demographic and work experience characteristics of those deputies whomost closely adhere to these sub-cultural positions we do so by simplyexamining the bivariate correlations between these characteristics and fivescales of sub-cultural adherence These correlations inform us as to whetheror not the conditional distributions of these scales vary significantly acrossvalues of deputiesrsquo socio-demographic and work experience characteristicsTable II presents Pearsonrsquos zero-order correlations coefficients across thesetwo sets of variables

As is evident in Table II only deputiesrsquo age and rank are significantlyassociated with any of the five measures of sub-cultural adherence suggestingthat but for these exceptions the conditional distributions for the sub-culturaladherence scales do not vary across the socio-demographic or work experiencecharacteristics of these deputies That is there appears to be considerableconsensus across these characteristics with regard to the extent of sub-cultural

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

102

adherence which we noted above was at best rather limited But there isevidence of variation in the conditional distributions of sub-cultural adherencewith regard to deputiesrsquo age and rank Older deputies are less cynical (r =plusmn023) less crime-control-oriented (r = plusmn017) and more service-oriented (r = 017)than are younger deputies we anticipated findings opposite to these andsuggesting that sub-cultural adherence is stronger among older deputiesSimilarly those holding a rank of corporal or higher are less cynical than those atthe entry-level rank of deputy (r = 013) Thus it is those with less experience(based on their age and rank) who are most likely to adhere to elements of thepolice sub-culture Perhaps these younger and lower-ranking officers perceivegreater pressure to express their sub-cultural allegiance than do those with moreexperience Because we have observed only partial evidence of sub-culturaladherence typically by only some of our subjects we move now toward theconstruction of a taxonomy of sheriffsrsquo deputies on the basis of variation in theirdegree of adherence to the five dimensions of the police sub-culture

Results II an empirical taxonomyTable III presents findings from an agglomerative hierarchical cluster analysis(Wardrsquos Method) based on the five sub-cultural adherence scales for thosedeputies deployed at one of the two operational districts sampled (District I)and a replication of this analysis for those deployed in the other operationaldistrict (District II) These findings are indicative of both a three- and a two-cluster structure to these data though as we will defend below we believe thata three-cluster structure ` bestrsquorsquo fits these data Moreover both the original(District I) cluster analysis and its replication (District II) produced strikinglysimilar findings indicating that we were able to successfully replicate thecluster structure in these data

The cluster profiles for the three-cluster solutions from both the originalanalysis (District I) and its replication (District II) are reported in Table IV Forboth analyses slightly over half of the deputies form one cluster (53 percent in

Measures of sub-cultural adherenceCrimecontrol Service Cynicism Traditionalism

Receptivityto change

Socio-demographicsAge plusmn017 017 plusmn023 plusmn010 006Race (non-white = 0) 007 plusmn006 plusmn007 000 plusmn003Gender (female = 0) 000 009 002 010 plusmn004Education plusmn001 002 001 plusmn011 plusmn002

Work experienceRank (Corporal + = 0) plusmn005 plusmn006 013 002 002Experience plusmn008 008 plusmn010 plusmn010 007District (District II = 0) plusmn011 plusmn003 011 005 001

Note p lt 005 (one-tailed t-test)

Table IIZero-order correlationsfor deputiesrsquo socio-demographic and workexperiencecharacteristics and thefive sub-culturaladherence scales

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

103

the original cluster analysis and 54 percent in the replication) The profile for thiscluster in both analyses reveals a group of deputies whose scores on the sub-cultural adherence scales are on average very average That is these deputiesreport an average orientation toward crime control (District I 3242 and DistrictII 3274 compared with the total sample average of 319 on this scale) Likewisethese deputies report an average orientation toward service (District I 4606 and

Cluster 1 Cluster 2 Cluster 3

District I (originalcluster analysis)CRIMCONT 3242 4062 2664SERVICE 4606 4549 5491CYNICISM 2436 3237 2194TRADITIONALISM 1772 1903 1782RECEPTIVITY TOCHANGE 1551 1397 1698LABEL Normals Sub-cultural adherents COP CopsFREQUENCY 45 (53) 14 (16) 26 (31)

District II (clusteranalysis replication)CRIMCONT 3274 4100 2568SERVICE 4607 4784 5538CYNICISM 2399 3100 2050TRADITIONALISM 1767 1896 1735RECEPTIVITY TOCHANGE

1523 1432 1765

LABEL Normals Sub-cultural COP CopsAdherents

FREQUENCY 70 (54) 25 (19) 34 (26)

Table IVCluster profiles forDistrict I (original

cluster analysis) andDistrict II (cluster

analysis replication)

District I Original District II Replication

Number of clustersRMSSTD

Semi-partial

R2 R2

Approxexpected

R2RMSSTD

Semi-partial

R2 R2

Approxexpected

R2

10 297 0016 0751 0768 376 0015 0721 077609 261 0018 0733 0751 322 0024 0698 076208 291 0023 0701 0732 371 0026 0672 074507 304 0027 0683 0708 294 0026 0646 072606 311 0044 0639 0680 321 0035 0611 070205 346 0069 0570 0644 397 0036 0576 067104 399 0071 0499 0596 321 0046 0530 062803 365 0072 0427 0521 363 0065 0465 056502 431 0147 0280 0376 428 0161 0304 044101 493 0280 0000 0000 491 0304 0000 0000

Note Italicised values indicate locations for ` bestrsquorsquo number of clusters within these data

Table IIICluster analysis and

replication plusmndiagnostics for thenumber of clusters

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

104

District II 4607 compared with an overall average of 487) Again this cluster ofdeputies reports an average level of cynicism (District I 2436 District II 2399overall average 249) an average level of traditionalism (District I 1772 DistrictII 1767 overall average 180) and an average level of receptivity toward change(District I 1551 District II 1523 overall average 157) Thus the first cluster inboth analyses is composed of deputies who on average are average with regardto their level of adherence to the police sub-culture and therefore cannot beconsidered as adherents to this police sub-culture Accordingly we label this firsttype in our taxonomy the `Normalsrsquorsquo this type is represented by slightly over halfof the deputies in this agency and closely resembles the `peace-keeping moralentrepreneursrsquorsquo observed by Jermier et al (1991)

The second cluster identified in both analyses is composed of less than one-fifth of the deputies (16 percent of the District I deputies and 19 percent of theDistrict II deputies) The profile for this cluster suggests a group of deputieswith a stronger than average crime-control orientation (District I 4062 andDistrict II 4100 compared with an overall average of 319) This group is alsocharacterized by a marginally lower than average service orientation (DistrictI 4549 and District II 4784 compared with a total sample mean of 487) anda higher than average level of cynicism (District I 3237 and District II 3100versus a sample average of 249) Finally this cluster is nominally lessreceptive to organizational change and slightly more traditionalistic Givenits strong crime-control orientation above average cynicism marginallyabove average traditionalism and slightly below average receptivity tochange and pro-service orientation this cluster comes closest to the ideal ofadherents to the police sub-culture though it is by no means ideally orstrongly sub-cultural nevertheless we label this type the ` Sub-culturaladherentsrsquorsquo The profile of this cluster closely resembles Brownrsquos (1981) ` Old-style crime-fighterrsquorsquo and the ` crime-fighting street professionalsrsquorsquo observed byJermier et al (1991)

The last cluster in both the original and replication cluster analyses (31percent of those in District I and 26 percent of those in District II) appears tobe the opposite of the Sub-cultural adherent The profile for this clusterreveals a below average crime control orientation (District I 2664 andDistrict II 2568 versus a sample mean of 319) a strong service orientation(District I 5491 District II 5538 versus a sample mean of 487 on this scale)below average level of cynicism (District I 2194 District II 2050 versus asample mean of 249) a near average level of traditionalism (District I 1782District II 1735 versus a sample mean of 180) and a slightly elevatedreceptivity to organizational change (District I 1698 District II 1765compared with a sample mean of 157) Based on this profile we label thiscluster the ` COP copsrsquorsquo the community-oriented policing cops though theycould also be tagged as the ` service providersrsquorsquo because its profile closelyresembles Brownrsquos (1981) ` servicersquorsquo style and Jermier et alrsquos (1991) ` anti-military social workersrsquorsquo Either way this cluster appears to constitute onceagain a nouveau police sub-culture

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

105

In sum both the original cluster analysis performed on data from theDistrict I deputies and the replication analysis conducted on the District IIdeputies produce evidence of two or three unique clusters of deputies Theprofiles for these clusters suggests a large group of typical sheriffsrsquo deputieswho on average are quite average and are by no means strong adherents to apolice sub-culture we have labeled this group the `Normalsrsquorsquo The remainingtwo clusters constitute sub-cultural opposites in which one-sixth of the deputiesmay be categorized as ` Sub-cultural adherentsrsquorsquo while the remaining one-fourthto one-third of the deputies appear to be sub-cultural rejectors whom we havelabeled ` COP copsrsquorsquo As with the findings reported in Tables I and II the clusteranalysis and its replication both falsify the existence of widespread adherenceto the police sub-culture but at the same time these analyses cannot rule out itsadherence by at least a small component of the force It is important to notethat when a two-cluster structure is imposed on these data in both the originalanalysis and the replication analysis the first and third clusters are combinedjoining the ` COP copsrsquorsquo with the `Normalsrsquorsquo against the ` Sub-culturaladherentsrsquorsquo This is relevant to our purposes because it again indicates theexistence of a small group of adherents to the police sub-culture thoughadmittedly this adherence is not ideal

Given the three-group taxonomy suggested by the cluster analysis we nowattempt to validate this taxonomy by examining the extent to which significantmean differences if any can be observed across these three types on a variety ofsocio-demographic and work experience variables In addition we employ thetaxonomy as a nominal-level dependent variable in a discriminant functionanalysis modeled by these same socio-demographic and work experience variables

Results III validationOur initial efforts at validating this numerical taxonomy of law enforcementofficers involves examining mean and proportional differences between typeson a number of socio-demographic and work experience variables which wereexternal to the construction of this taxonomy but for which the researchliterature suggests we should anticipate observing significant differencesThese between-type differences on these external variables are reported inTable V In addition Table V reports the mean differences between type on thefive sub-cultural adherence scales from which the taxonomy was producedThe differences reported in Table V produce two very clear sets of findingsFirst as we would expect the three types are significantly different across thedimensions of sub-cultural adherence from which the taxonomy wasconstructed ` Sub-cultural adherentsrsquorsquo as expected show a significantlystronger crime control orientation than the other two types they are also morecynical and less receptive to organizational changes Conversely the ` COPcopsrsquorsquo have a significantly stronger service orientation than the other two typesthey also have a significantly weaker crime control orientation are less cynicaland are more receptive to change than the other two types While these findingsconfirm the basis upon which these types are constructed these significant

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

106

differences do nothing to aid in validating the taxonomy which is the primarypurpose of this component of the study

Unfortunately the taxonomy cannot be statistically validated through theexternal variables (the second clear finding reported in Table V) That iswhile the research literature led us to anticipate observing significantdifferences between these types on the external variables we failed to observeany statistically significant mean or proportional differences That is thegender racial age educational and work experience composition of thesethree types of sheriffsrsquo deputies are not significantly different from oneanother There are however several differences which while not statisticallysignificant may have some substantive significance that merits furtherdiscussion For instance the ` COP copsrsquorsquo tend to be both older (byapproximately three years) and to have more experience in law enforcement(by approximately 16 months) than the other two types It was our belief thatsub-cultural adherence would be most manifest among the older and moreexperienced officers and that the younger and less experienced officers wouldbe more inclined to express work-related values and orientations consistentwith the recent organizational shifts toward community-oriented andproblem-oriented policing but this is not what we have observed in thesedata Equally unexpected are the differences in the racial composition of thesetypes The composition of both the ` COP copsrsquorsquo and the ` Sub-culturaladherentsrsquorsquo is slightly more minority than the ` Normalsrsquorsquo (by about 10-13percent) We anticipated lower minority representation among sub-culturaladherents Perhaps some of the minority officers perceive greater pressure toexpress such values and orientations in order to fit into an occupational realmthat has historically denied them an entrance

` Sub-culturaladherentsrsquorsquo vs

`Normalsrsquorsquo

` Sub-culturaladherentsrsquorsquo vs ` COP

copsrsquorsquo`COP copsrsquorsquo vs

`Normalsrsquorsquo

Mean differencesCrime control 851 1398 plusmn547Service plusmn026 plusmn942 917Cynicism 795 1043 plusmn248Traditionalism 126 121 005Receptivity to change plusmn148 plusmn301 154Age plusmn014 plusmn310 296Education 007 plusmn011 018Work experience plusmn043 plusmn1644 1602

Proportional differencesFemale 0007 plusmn0012 0018Minority 0096 plusmn0032 0129Deputy plusmn0016 0028 plusmn0044District I plusmn0025 plusmn0068 0043

Note p lt 005

Table VMeanproportionaldifferences betweentypes on internal andexternal variables

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

107

The lack of any statistically significant mean or proportional differencesbetween the three types across the various ` externalrsquorsquo socio-demographic andwork experience variables as reported in Table V replicated in our final effortto validate the taxonomy via discriminant function analysis (see Models 1 and2 of Table VI) None of these external variables significantly assists indiscriminating between the three groups Moreover as reported in Model 1these external variables cannot effectively classify the observations into thetaxonomy The overall classification error rate is about 41 percent thoughapproximately 82 percent of the `Normalsrsquorsquo can be correctly classified On theother hand only 17 percent of the ` Sub-cultural adherentsrsquorsquo can be correctlyclassified and only 42 percent of the ` COP copsrsquorsquo are correctly classified Thepoor performance of these ` externalrsquorsquo variables suggests that our taxonomy(and perhaps other police typologies available in the literature) lack predictivevalidity

On the other hand with the inclusion of the ` internalrsquorsquo variables used toconstruct our taxonomy we are able to effectively discriminate among thesethree types (see Model 2 of Table VI) The overall classification error rate forthis model is only approximately 7 percent and between 90 percent and 95percent of the observations within each type are correctly classified once theinternal variables are included in the model Of the 12 ` discriminatingrsquorsquovariables employed in Model 2 only three effectively discriminate between the

MODEL 1 (External vars only) MODEL 2 (with internal vars)

Partial R2Wilksrsquo

Lambda

Avgsquared

canonicalcorrelation Partial R2

WilksrsquoLambda

Avgsquared

canonicalcorrelation

Independent variablesAge 228 098 0011 216 023 0480Gender (female = 0) 032 096 0023 088 022 0488Race (minority = 0) 147 096 0019 040 022 0492Level of education 002 095 0027 067 022 0491Rank (non-deputy = 0) 029 095 0024 165 022 0486Level of experience 047 096 0021 024 022 0493District (DII = 0) 057 095 0027 251 023 0483

Crime-control 5070 049 0253Service 3758 031 0404Cynicism 2299 024 0474Traditionalism 063 022 0490Receptivity to change 020 022 0493

Percent correctlyclassified` Sub-culturaladherentsrsquorsquo 1667 9000`Normalsrsquorsquo 8155 9505` COP copsrsquorsquo 4167 9153Classification error rate 4093 684

Table VIValidation of taxonomyplusmn discriminant function

analyses

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

108

three types of deputies These are the crime-control orientation serviceorientation and cynicism scales again each is internal to the construction ofthe taxonomy While such findings should be of no surprise they do ` validatersquorsquoour taxonomy to the degree that we are able to replicate the classificationscheme using an alternative statistical procedure

ConclusionIn our review of the extant literature regarding the police sub-culture wenoted two key observations First most of the research on the police sub-culture has been exploratory and descriptive in nature and they appear tohave uncritically accepted the existence of this sub-culture A key element ofthese descriptive efforts has been the oft-repeated claim that the sub-cultureis endemic to policing and that adherence to it is widespread However a fewstudies have questioned whether or not such a sub-culture exists and if foundto be present address the issue of how prevalent or widespread its adoption isamong law enforcement officers (Haarr 1997 Herbert 1998 Jermier et al1991 Manning 1995 Paoline et al 2000 see also Worden (1995)) Most ofthese studies either fail to verify the existence of this sub-culture or observethat its prevalence is not widespread but is present among only a smallsegment of the police ranks Second studies which have observed its presenceamong some but not most officers typically proceed by developing a typologyof officers based at least in part on the degree of sub-cultural adherence(Brown 1981 Jermier et al 1991 Muir 1977 Reiner 1978 Wilson 1968)rarely however are these typologies the product of analyticstatisticalprocedures designed for their construction (eg cluster analysis) nor are thesetypologies subsequently tested for their empirical validity (Bailey 1994 plusmn foran exception see Jermier et al (1991))

These observations brought us to the threefold purpose of the present study

(1) to examine the degree of adherence to the police sub-culture

(2) to attempt to construct a taxonomy (ie an empirically derived typologysee Bailey (1994)) of police officers by making better use of advancedstatistical procedures for doing so specifically cluster analysis and

(3) to validate the taxonomy by examining the extent to which significantmean differences exist between these types across an array of socio-demographic and work experience indicators through the use ofdiscriminant function analysis

Based upon our analysis of survey data from a sample of sheriffsrsquo deputieswe failed to find any evidence suggestive of a police sub-culture that isendemic and widespread However we did find evidence of sub-culturaladherence by a segment of the deputies studied In addition we foundevidence of a potentially nouveau police sub-culture that is strongly orientedtoward community-service Furthermore we were able to produce andreplicate an empirical taxonomy of sheriffsrsquo deputies based on their degree ofsub-cultural adherence This taxonomy suggests the existence of three types

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

109

of sheriffsrsquo deputies ` Sub-cultural adherentsrsquorsquo ` Normalsrsquorsquo and ` COP copsrsquorsquoThe Sub-cultural adherents come closest in their profile to the ideal of thepolice sub-culture but comprise only one-sixth of the sample We again notethe presence of what may prove to be a nouveau police sub-culture the ` COPcopsrsquorsquo which places strong emphasis on the importance on the communityservice roles associated with policing this sub-culture is composed ofapproximately 25-30 percent of the deputies sampled About half of thesubjects studied make up the primary ` typersquorsquo of deputies who on average arevery average plusmn the ` Normalsrsquorsquo Finally we were able to partially validate thistaxonomy through an alternative statistical classification process(discriminant function analysis) but the taxonomy fails to have anypredictive validity in that ` externalrsquorsquo socio-demographic and work experiencevariables do not effectively ` discriminatersquorsquo between the types

A note of caution regarding this study is necessary Our conclusions may bepremature for a number of reasons At a minimum these include issues ofsampling and measurement First our study is based upon a sample of sheriffsrsquodeputies and thus our findings may not generalize to other law enforcementpersonnel Because sheriffs are elected officials and are thus directlyanswerable to the public it is likely that they are more inclined than policechiefs to adopt an attitude to policing that is strongly service-oriented (Falconeand Wells 1995 Bromley and Cochran 1999) If such is the case then this pro-service orientation is likely also to find expression in the occupational andorganizational culture of a sheriffrsquos agency Second our data are based upon asample of deputy sheriffs who are employed in an agency which has adoptedcommunity-oriented policing agency-wide This too may account for thestronger than anticipated service orientation and receptivity to organizationalchange observed in these data Finally our data were not originally collectedwith this study in mind hence with regard to this study these data constitutesecondary data and are subject to all of the measurement issues common tosecondary data Key among these may be limitations in the validity of our fivemeasures of sub-cultural adherence Do these scales actually measure elementsof the police sub-culture Clearly additional research needs to be conducted inthis area of inquiry before we can be confident that our findings aregeneralizable

Nevertheless since ours is one of only two or three studies which we areaware have attempted to create replicate and validate an empiricaltaxonomy of law enforcement officers (as opposed to the subjective creationof conceptual typologies which have dominated this area of policescholarship) we are concerned about both the validity of these othertypologies and the lack of scholarship critically assessing them (seeLangworthy and Travis 1998 p 253) Importantly however we note thesimilarities of our ` Sub-cultural adherentsrsquorsquo and ` COP Copsrsquorsquo to Brownrsquos(1981) ` Old-style crime-fightersrsquorsquo and ` Servicersquorsquo styles respectively and toJermier et alrsquos (1991) ` crime-fighting street professionals and ` anti-militarysocial workersrsquorsquo respectively In our view we are currently at a critical period

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

110

of scholarship on this subject and a considerable amount of research is nowvery badly needed No longer can we blindly accept the notion of an endemicwidespread police sub-culture In fact we may find that there are multiplepolice sub-cultures in existence In addition we can no longer permit theconstruction of mere conceptual typologies when the technologies necessaryfor the construction of empirical taxonomies are so freely available Lastlywe can no longer continue to presume the validity of these typologiestaxonomies when their validity can be tested empirically It is for thesereasons that we find ourselves at the dawn of scholarship regarding thestudy of sub-cultural responses by law enforcement officers to the uniqueoccupational and organizational contingencies of police work and the policerole

Because of the need for more research in this area we are discomfited bythe notion of offering any policy recommendations from our work Clearlythis research needs to be replicated and its findings corroborated before anypolicies should be offered Moreover it is clear that research is needed notonly on the existence of police sub-cultures but also on their consequences forlaw enforcement (desirable andor undesirable) and their origins (importationvs socialization) If police sub-cultures are found to be the product of thevalues beliefs and orientations that recruits bring with them when they enterthe force then the policy implications of this body of research should bedirected at recruitment and training processes If these sub-cultures are theproduct of professional socialization processes then policies may need to bedirected at academy and field training procedures and the personnel selectedto offer this training On the other hand if these sub-cultures are found to be aconsequence of the unique dangers and stresses of police work then policydirectives should focus on danger- and stress-reduction technologies andprograms To the degree that multiple sub-cultures exist as is suggestedfrom this study and to the degree that they lead to both desirable andundesirable consequences then the policies implicated would correspond toissues of officer assignment and deployment For instance those stronglyoriented toward the service-related aspects of police work should be selectedfor community- or problem-oriented policing units juvenile public relationsetc Conversely those antagonistic toward service roles but strongly orientedtoward crime-fighting should be deployed accordingly to high crime areasandor special tactical or street crime units

Notes

1 For an excellent discussion of the police sub-culture readers are encouraged to examine thework of Paoline et al (2000)

2 Weisheit et al (1995) in a very complete examination of the conceptual and empiricalissues of crime and policing in rural and small-town USA have built an effective case for amulti-dimensional conceptualization of ` ruralrsquorsquo Based upon this multi-dimensionalconceptualization we are very confident that the unique crime and justice issues whichconfront rural communities and rural law enforcement relative to those of metropolitancommunities are not characteristic of the communities served by the two operational

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

111

districts examined in this study With regard to the ` demographicrsquorsquo dimension of ` ruralrsquorsquocommunities both Districts I and II are densely populated serve over 50000 and 150000residents respectively and are socially and economically tied to the greater Tampa BayMetropolitan Area `Economicallyrsquorsquo these districts are not even secondarily agriculturalbut instead are characterized by a complex division of labor which is heavily service-sector and industrial in nature Finally along the dimensions of ` social structurersquorsquo and` culturersquorsquo these districts do not resemble Weisheit et alrsquos (1995) description of ruralcommunities life in these districts is heavily influenced by secondary-group interactionsand is very heterogeneous That is these two operational districts though located withunincorporated areas are best characterized as ` suburbanrsquorsquo or `metropolitan fringersquorsquo andthus address a work environment similar to that of most municipal and metropolitan policedepartments Thus our examination of the extent if any to which these sheriffsrsquo deputiespossess work orientations consistent with the traditional police sub-culture is not likely tobe biased by our sample For a more complete description of these districts and theircommunities see Bromley and Cochran (1999) Cochran et al (1999) and Halsted et al(2000)

3 The two-year delay between survey administrations was an accident of our researchsuccess That is originally we had planned only to assess District I However oursuccesses there combined with an NIJ COPS grant involving District II allowed us to re-employ our questionnaire to these additional personnel To our knowledge no significantorganizational changes took place in the interim though readers should be cautiousregarding any District I vs District II differences observed herein

4 Comparisons of the socio-demographic profiles within each district of the personnelsampled with those unavailable for participation produced no statistically significantdifferences

References

Aldenderfer MS and Blashfield RK (1984) Cluster Analysis Sage University Paper Series onQuantitative Applications in the Social Sciences 07-044 Sage Thousand Oaks CA

Bailey KD (1994) Typologies and Taxonomies An Introduction to Classification TechniquesSage University Paper Series on Quantitative Applications in the Social Sciences 07-102Sage Thousand Oaks CA

Bayley D and Bitner E (1984) ` Learning the skills of policingrsquorsquo Law and ContemporaryProblems Vol 47 pp 35-59

Bitner E (1967) `The policeman on skid row a study in peacekeepingrsquorsquo American SociologicalReview Vol 32 pp 699-715

Bitner E (1974) `Florence Nightingale in pursuit of Willie Sutton a theory of the policersquorsquo in JacobH (Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 17-44

Black D (1980) The Manner and Customs of the Police Academic Press New York NY

Britz MT (1997) `The police subculture and occupational socialization exploring individual anddemographic characteristicsrsquorsquo American Journal of Criminal Justice Vol 21 pp 127-46

Broderick JJ (1977) Police in a Time of Change General Learning Press Morristown NJ

Bromley ML and Cochran JK (1999) `A case study of community policing in a Southernsheriffrsquos officersquorsquo Police Quarterly Vol 2 No 1 pp 36-56

Brown MK (1981) Working the Street Police Discretion and the Dilemmas of Reform RussellSage New York NY

Chan J (1997) Changing Police Culture Policing a Multicultural Society Cambridge UniversityPress New York NY

Chevigny P (1969) Police Power Police Abuses in New York City Pantheon New York NY

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

112

Chevigny P (1995) The Edge of the Knife Police Violence in the Americas The New Press NewYork NY

Christensen W and Crank JP (2001) ` Police work and culture in a non-urban setting anethnographic analysisrsquorsquo Police Quarterly Vol 4 pp 99-122

Cochran JK Bromley ML and Landis LV (1999) ` Officer work orientations perceptions ofreadiness and anticipated effectiveness of an agency-wide community policing effortwithin a county sheriffrsquos officersquorsquo Journal of Police and Criminal Psychology Vol 14 No 1pp 43-65

Crank JP (1998) Understanding Police Culture Anderson Cincinnati OH

DiMaggio P and Powell W (1991) The New Institutionalism in Organizational AnalysisUniversity of Chicago Press Chicago IL

Drummond DS (1976) Police Culture Sage Beverly Hills CA

Falcone D and Wells L (1995) ` The county sheriff as a distinctive policing modalityrsquorsquoAmerican Journal of Police Vol 15 pp 123-49

Farkas MA and Manning PK (1997) `The occupational culture of correctional and policeofficersrsquorsquo Journal of Crime and Justice Vol 20 pp 51-68

Fielding NG (1988) Joining Forces Police Training Socialization and Occupation CompetenceRoutledge London

Greene J Bergman W and McLaughlin E (1994) ` Implementing community policing culturaland structural change in police organizationsrsquorsquo in Rosenbaum D (Ed) The Challenge ofCommunity Policing Testing the Promises Sage Thousand Oaks CA pp 92-109

Haarr RN (1997) ` Patterns of interaction in a police patrol bureau race and gender barriers tointegrationrsquorsquo Justice Quarterly Vol 14 pp 53-85

Halsted AJ Bromley ML and Cochran JK (2000) ` The effects of work orientations on jobsatisfaction among sheriffsrsquo deputies practicing community-oriented policingrsquorsquo PolicingVol 23 pp 82-104

Herbert S (1998) ` Police subculture reconsideredrsquorsquo Criminology Vol 36 No 2 pp 343-69

Jermier JM Slocum JL Jr Fry LW and Gaines J (1991) ` Organizational subcultures in a softbureaucracy resistance behind the myth and facEumlade of an official culturersquorsquo OrganizationScience Vol 2 pp 170-94

Kappeler VE Sluder RD and Alpert GP (1994) Forces of Deviance Understanding the DarkSide of Policing Waveland Press ProspectHeights IL

Klecka WR (1980) Discriminant Analysis Sage University Paper Series on QuantitativeApplications in the Social Sciences 07-019 Sage Thousand Oaks CA

Langworthy RH and Travis LT III (1998) Policing in America 2nd ed Prentice-Hall NewYork NY

Lundman RJ (1980) Police and Policing An Introduction Holt Rhinehart amp Winston New YorkNY

Lurigio AJ and Skogan WG (1994) `Winning the hearts and minds of police officers anassessment of staff perceptions of community policing in Chicagorsquorsquo Crime and DelinquencyVol 40 No 3 pp 315-30

McNamara JH (1967) ` Uncertainties in police work the relevance of police recruitsrsquo backgroundand trainingrsquorsquo in Bordua D (Ed) The Police Six Sociological Essays John Wiley NewYork NY pp 163-252

Manning PK (1995) ` The police occupational culturersquorsquo in Bailey W (Ed) Encyclopedia of PoliceScience Garland New York NY pp 472-5

Muir WK Jr (1977) Police Streetcorner Politicians University of Chicago Press Chicago IL

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

113

Neiderhoffer A (1967) Behind the Shield The Police in Urban Society Doubleday Garden City NY

Ouchi W and Wilkins A (1985) ` Organizational culturersquorsquo Annual Review of Sociology Vol 11pp 457-83

Paoline EA III Myers SM and Worden RE (2000) `Police culture individualism and communitypolicing evidence from two police departmentsrsquorsquo Justice Quarterly Vol 17 No 3 pp 575-605

Reaves BJ (1992) Sheriffsrsquo Departments 1990 Bureau of Justice Statistics Washington DC

Reiner R (1978) The Blue-Coated Worker Cambridge University Press New York NY

Reiner R (1985) The Politics of Police St Martinrsquos Press New York NY

Reuss-Ianni E (1983) Two Cultures of Policing Transaction Books New Brunswick NJ

Ritti R (1994) The Ropes to Skip and the Ropes to Know Studies in Organizational BehaviorJohn Wile New York NY

Rubenstein J (1973) City Police Farrar Strauss and Giroux New York NY

Sackmann S (1991) Cultural Knowledge in Organizations Exploring the Collective Mind SageThousand Oaks CA

Sadd S and Grinc R (1993) Innovative Neighborhood-Oriented Policing An Evaluation ofCommunity Policing Programs in Eight Cities Vera Institute New York NY

Schein E (1985) Organizational Culture and Leadership Jossey-Bass San Francisco CA

Shearing C (1981) Organizational Police Deviance Butterworth Toronto

Skolnick JH (1966) Justice without Trial Law Enforcement in a Democratic Society John WileyNew York NY

Skolnick JH and Fyfe J (1993) Above the Law Police and the Excessive Use of Force Free PressNew York NY

Sneath PHA and Sokol RR (1973) Numerical Taxonomy Freeman San Francisco CA

Sparrow M Moore MA and Kennedy D (1990) Beyond 911 A New Era for Policing BasicBooks New York NY

Sykes RE and Brent EE (1980) ` The regulation of interaction by the police a systems view oftaking chargersquorsquo Criminology Vol 18 pp 182-97

Tauber RK (1970) ` Danger and the policersquorsquo in Johnston N Savitz L and Wolfgang ME (Eds)The Sociology of Punishment and Correction John Wiley New York NY pp 95-104

Van Maanen J (1974) `Working the street a developmental view of police behaviorrsquorsquo in Jacob H(Ed) The Potential for Reform of Criminal Justice Sage Beverly Hills CA pp 83-130

Walker S (1977) A Critical History of Police Reform The Emergence of ProfessionalismLexington Books Lexington MA

Ward JH Jr (1963) ` Hierarchical grouping to optimize an objective functionrsquorsquo Journal of theAmerican Statistical Association Vol 58 pp 236-44

Weisheit RA Wells LE and Falcone DN (1995) Crime and Policing in Rural and Small-TownAmerica An Overview of the Issues NIJ Research Report US Department of JusticeWashington DC

Westley W (1970) Violence and the Police A Sociological Study of Law Custom and MoralityMIT Press Cambridge MA

White SO (1972) `A perspective on police professionalismrsquorsquo Law and Society Review Vol 7pp 61-85

Wilson JQ (1968) Varieties of Police Behavior The Management of Law and Order in EightCommunities Harvard University Press Cambridge MA

Worden R (1995) ` Police officersrsquo belief systems a framework for analysisrsquorsquo American Journalof Police Vol 14 No 1 pp 49-81

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

114

Appendix

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Crime control items1 An aggressive tough bearing is

more useful to a law enforcementofficer than a friendly courteousmanner 05 23 177 572 223

2 If law enforcement officers act ina service capacity it detractsfrom their ability to fight crime 42 252 224 449 33

3 Law enforcement officers shouldnot become personally familiarwith the residents of the areathey patrol 23 14 61 612 290

4 All laws should be fully enforcedat all times otherwise people loserespect for the law 103 206 271 383 37

5 Problem solving should not bepart of an officerrsquos responsibility 14 33 75 659 220

6 Good law enforcement requiresthat officers concern themselveswith the consequences of crimeand not with its root causes 09 140 168 514 168

7 Law enforcement officers shouldnot forget that enforcing the lawis by far their most importantresponsibility 112 530 223 126 09

8 Most law enforcement officershave to spend too much of theirtime handling unimportant non-crime calls for service 154 322 248 262 14

9 Law enforcement officers shouldnot have to handle calls thatinvolve social or personalproblems where no crime isinvolved 61 234 276 425 05

10 Many of the decisions made bythe courts interfere with theability of law enforcement officersto fight crime 195 405 205 167 28

11 If law enforcement officers in highcrime areas had fewer restrictionson their use of force many of theserious crime problems in theseareas would be significantlyreduced 56 181 247 433 84

(continued)

Table AIFrequency distributionson items measuringpolice sub-culturaladherence (percentagesplusmn values in italicsrepresent sub-culturaladherence)

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

115

Stronglyagree Agree Neutral Disagree

Stronglydisagree

12 Law enforcement officers wouldbe more effective if they did nothave to worry about ` probablecausersquorsquo requirements for searchesas mandated by the courts 65 177 126 474 158

13 Law enforcement officers wouldbe more effective if they did nothave to worry about a suspectrsquosrights during interrogations 65 93 126 553 163

Service items

1 Law enforcement officers shouldbe sincerely concerned about thewellbeing of the citizens in theneighborhoods they patrol 248 659 65 19 09

2 Law enforcement officers shouldmake frequent informal contactswith the people in the area theypatrol 178 692 107 19 05

3 Law enforcement officers shouldtry to work with neighborhoodresidents civic groups and thelocal business community to solvecrime problems on their beat 197 695 89 14 05

4 Law enforcement officers shouldtry to solve the non-crimeproblems identified by citizens ontheir beat 28 358 364 210 47

5 Law enforcement officers shouldask citizens what types ofservices they want 65 442 288 195 09

6 Crimes are only one of severalproblems about which lawenforcement officers should beconcerned 47 659 187 93 14

7 Assisting citizens in need is justas important as enforcing the law 144 665 130 60 00

8 Lowering citizensrsquo fear of crimeshould be just as high a priority forthe HCSO as cutting the crime rate 103 598 206 79 14

9 Community crime problems canbe solved by cooperation betweenlaw enforcement and local non-criminal justice agencies 102 628 207 65 05

(continued) Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

PIJPSM261

116

Stronglyagree Agree Neutral Disagree

Stronglydisagree

10 Law enforcement officers shouldbe required to help settle familydomestic disputes 42 435 294 182 47

11 Law enforcement officers shouldbe required to handle publicnuisance problems 38 507 254 160 42

12 Law enforcement officers shouldbe required to attend to the sickor injured 28 377 269 250 75

13 Law enforcement officers shouldbe required to assist citizens whoare having problems with theircars (locked out dead battery outof gas etc) 23 402 285 238 51

14 Law enforcement should be seenprimarily as a service-orientedprofession rather than a crimecontrol profession 28 202 207 441 122

Cynicism items

1 Most people lie when answeringquestions posed by lawenforcement officers 79 360 304 252 05

2 Most people do not hesitate to goout of their way to help someonein trouble 05 257 210 458 70

3 Most people are untrustworthyand dishonest 05 89 282 592 05

4 Most people would steal ifthey knew they would not getcaught 19 254 263 432 33

5 Most people respect the authorityof law enforcement officers 00 472 238 220 70

6 Most people lack the proper levelof respect for law enforcementofficers 52 319 254 357 19

7 Law enforcement officers willnever trust citizens enough towork together effectively 05 88 212 608 88

8 Most citizens are open to theopinions and suggestions of lawenforcement officers 05 548 253 180 14

9 Citizens will not trust lawenforcement officers enough towork together effectively 05 124 286 567 18

(continued)Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI

The myth()of the policesub-culture

117

Stronglyagree Agree Neutral Disagree

Stronglydisagree

Traditionalism items

1 To be effective an organizationshould have clearlydefined positions of powerauthority among its membersemployees 79 600 247 74 00

2 In law enforcement organizationspower should be evenlydistributed among its personnel 46 449 220 269 09

3 Communication works best whenit follows clear establishedchannels from the top down 131 593 117 131 28

4 Participatory managementschemes really do not workwithin law enforcement agencies 14 117 329 446 94

5 The quasi-military structure is themost effective organizational typefor law enforcement agencies 79 467 252 145 56

6 Subordinates should not beinvolved in either the setting orthe enforcing of policies andprocedures within lawenforcement agencies 28 280 183 592 169

Receptivity to change items1 Most changes at work are

problematic and ineffective 09 150 376 437 28

2 I can usually find some way toget around changes at work 05 164 369 425 37

3 I often suggest new approachesfor doing things at my job 23 449 397 117 14

4 Most changes make my workmore efficient (ie saves timeeffort money) 14 262 374 304 47

5 Most changes make my job moreeffective (ie more arrests fasterresponse times crime reduction) 14 201 411 313 61 Table AI