classification of offenders as an aid to efficient

Journal of Criminal Law and Criminology

Volume 62 | Issue 2 Article 7

1971

Classification of Offenders as an Aid to EfficientManagement and Effective TreatmentMarguerite Q. Warren

Follow this and additional works at: https://scholarlycommons.law.northwestern.edu/jclc

Part of the Criminal Law Commons, Criminology Commons, and the Criminology and CriminalJustice Commons

This Criminology is brought to you for free and open access by Northwestern University School of Law Scholarly Commons. It has been accepted forinclusion in Journal of Criminal Law and Criminology by an authorized editor of Northwestern University School of Law Scholarly Commons.

Recommended CitationMarguerite Q. Warren, Classification of Offenders as an Aid to Efficient Management and Effective Treatment, 62 J. Crim. L.Criminology & Police Sci. 239 (1971)

https://scholarlycommons.law.northwestern.edu/jclc?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol62%2Fiss2%2F7&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc/vol62?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol62%2Fiss2%2F7&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc/vol62/iss2?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol62%2Fiss2%2F7&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc/vol62/iss2/7?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol62%2Fiss2%2F7&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol62%2Fiss2%2F7&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/912?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol62%2Fiss2%2F7&utm_medium=PDF&utm_campaign=PDFCoverPages




TinE Jounmz.L op CnumHAr, waw, Cn=On.OGr AND POLICE ScIENCECopyright C 1971 by Northwestern University School of Law

Vol. 62, No. 2Printed in U.S.A.

CLASSIFICATION OF OFFENDERS AS AN AID TO EFFICIENT MANAGEMENTAND EFFECTIVE TREATMENT

MARGUERITE Q. WARREN*

Recent years have brought an increased impetusto thinking about classification systems and ty-pologiesi of criminals and delinquents. Amongthe several forces contributing to this development,two stand out. One force has come from developingresearch programs. As in other fields, scientificprogress in the field of corrections depends on re-ducing the infinite variety of problems throughconceptualization.

Research efforts attacking the problems of thefield systematically have required some sort oftheoretical framework, either a framework whichfocuses on the etiology of criminal and delinquentbehavior, or at least a framework which charts inan organized fashion signs, symptoms, or dynamicsof patterns covering the universe of offenders.

The second impetus to offender categorizationhas come with the switch from custody emphasis totreatment emphasis in handling offenders and withthe disappointments regarding the total effective-ness of some attempted treatment programs. Likethe humanitarian reform movement itself, tradetraining, increased facilities for socially acceptableoutlets of aggression, and individual and groupcounseling have each been thought of as the answerto the crime problem. While movements in behalfof these causes have undoubtedly made importantcontributions to the field of corrections, they havetended to be viewed as cure-alls, and it is a matterof record that we do not cure all delinquents andcriminals.

RATIONALE R CL ssIricATioN

One of the few facts agreed upon in the field ofcorrections is that offenders are not all alike. Thatis, they differ from each other not only in the formof their offense, but also in the reasons for and themeaning of their crime. Some individuals violate

* B.A. Western Reserve University; M.A., Ph.D.University of California at Berkeley; Program Director,Center for Training in Differential Treatment, Sacra-mento, California.

' The terms "classification system," "typology," and"taxonomy" have been used somewhat interchangeablyin this paper, even though a case may be made fordifferentiating among the terms for some technicalpurposes.

the law because the peer group, upon which theydepend for approval, prescribes criminal behavioras the price of acceptance, or because the values,which they have internalized, are those of a deviantsubculture. Other individuals break laws because ofinsufficient socialization, which leaves them at themercy of all but the most protected environments.Still others delinquently act out internal conflicts,identity struggles, or family crises. This list is ofcourse illustrative, not exhaustive.

Much of the literature in this field is still writtenas if all offenders are alike. Many causal theoriespurporting to explain delinquency have describedonly one segment of the total offender populationand have concluded, for example, that delinquencyis a peer group phenomenon. Differential associa-tion theories, 2 social disorganization theories,8 roletheories,4 and psychogenic theories5 all appear tohave a certain amount of validity when applied tosome segment of the offender population, but noneof these theories alone is sufficiently complex toaccount for the total observable range of causalfactors.

Program prescriptions as well have tended to bemade in an across-the-board fashion, with increasedstaff-offender ratios, improved job opportunities,or insight therapy recommended for all. Although

2 See generally H. SuTHERLAND & D. CREssEY,Pp-mci'LEs or CRnuNoLoGY (1960).8See generally C. SHAW & H. MCKAY, JU-ENIZEDELINQUENCY AND URBAN AREAS (1942); R. MERTON,SocAL TnEoRY AN SocIA STRUCTURE (1957); A.CLowARD & L. OnLN, DELINQUENCY AND OPPOR-TuoNir: A TBEORY or DELINQUENT GANGs (1960).

4 See generally Gough & Peterson, The Identificationand Measurement of Predispositional Factors in Crimeand Delinquency, J. CONSULnIG PSYCHOLOGY 207-12(1952); Sarbin, A Preface to the Psychological Analysisof the Self 59 PsYcHoLoGICAL Rxv. 11-23 (1952);T. PARsoNS, Tax SoCIAL Syszsr (1951); Cohen, TheSociology of the Deviant Act: Anomie Theory and Be-yond, 30 Am. SOcIoOGICAL ERv. 5-14 (1965).

5 See generally K. FRLEDI NER, Tax PsYcHoANA-LYTIC APPROACH TO JUVENILE DELINQUENCY (1947);Formation of the Anti-Social Character, in PsycHo-ANALYTIC STUDY or E Cmm, (1945); W. H EALY &A BRoNNER, NEW LIGHT ON DELINQUENCY AND ITSTREATmENT ?1936); Redl, New Perspectives for Researchon Juvenile Delinquency, QHIDREN's BUREAU Pumai-CATION No. 356 (H. Witmer & R. Kotinsky ed. 1956);E. EUESON, CmnHooD mm SocmY (1950).

MARGUERITE Q. WARREN

some action programs have been aimed at specificsegments of the heterogeneous offender population(for example, psychiatric treatment for the emo-tionally disturbed delinquent), few programs in-deed have based their goals for intervention andtheir treatment and management prescriptions ona specified rationale for handling differentially thevarieties of offender problems which appear in acorrectional setting.

A comment should perhaps be made with regardto an extreme opposite position taken by sometreatment-oriented people who have emphasizedthe great differences between offenders and haveresisted any schematization on the basis of loss ofmeaningful information about individuals. Al-though this position guards against the mistake ofadministering the same kind of treatment to alloffenders, it requires an infinite variety of treat-ments to fit the uniqueness of each case. This posi-tion almost precludes conceptualizing the delin-quency problem, developing intervention theoriesand practices, and instigating research investiga-tions. As such, the position must be rejected.

Theoreticians, practitioners and researchers in-creasingly seek some classification system, somemeaningful grouping of offenders into categories,which offers (1) a step in the direction of explana-tory theory with the resulting aid to predictionwhich follows from understanding, (2) implicationsfor efficient management and effective treatmentdecisions, and (3) greater precision for maximallyeffective research.

TYPOLOGIES OF CnRnALs AND DELINQUENTS

Systems of offender classification might begrouped in several ways. One such grouping, basedon the nature of the underlying dimensions crucialto the classification system, follows. 6

1. Prior probability approaches represented bythe Borstal studies,7 the California Youth Au-

6 Several reviews of the large number of recent con-tributions in this area are available. See e.g., Moles,Lippitt & Withey, A Selective Review of the Researchand Theory on Delinquency, INTxR-CENTE PROGRAMOF RESEARCH ON CHILDREN, YoU AND FAMILY LIFE(1959); Grant, Interaction Between Kinds of Treatmentsand Kinds of Delinquents, 2 CALIFORNIA STATE BOARDOr CoRxECTIoNS MONOGRAPH 5-14 (1961); Glueck &Glueck, Varieties of Delinquent Types, 5 BRIIsH J.Cans. 236-48, 388-405 (1965); Kinch, Continuities inthe Study of Delinquent Types, 53 J. Cans. L. C. & P. S.323-28 (1962); Lejins, Pragmatic Etiology of Delin-quent Behavior, in THE JuvENILE DELINQUENT (C.Vedder ed. 1954); Roebuck, Criminal Typology: ACritical Overview, 9 ALA. CORREcTIONAL J. 34-66 (1962).

7 MNHIM & WILKINS, PREDICTION METHODS IN

RELATION To BORSTAL TJn;ININ (1955).

thority,8 the Department of Corrections Base Ex-pectancy studies, 9 the Glueck prediction tables,10

and the configuration analysis procedures repre-sented by Glaser.i

2. Reference group typologies represented by

Schrag' and Sykesli and the social class typologiesrepresented by W. Miller.1 4

3. Behavior classifications (covering a widerange of specificity from offense types to conform-ity-nonconformity dichotomies) represented by

Roebuck,I5 McCord, McCord and Zola, 6 Oh l in,n

and Reckless.s4. Psychiatric-oriented approaches represented

by the work of Jenkins and Hewitt, 9 Redl,2° Erik-son,21 Aicborn,2 Makkay,23 Reiss, 2 4 Argylei 5 Blochand Flynn,2 6 and the Illinois State Training SchoolTreatment Committee.?

5. Social perception and interaction classifica-

8 Beverly, A Method of Determination of Base Expec-tancies for Use in The Assessment of Effectiveness ofCorrectional Treatment, REsEARCH REPORT No. 3,CALiFORNIA YOUTH AUTHORr Y, DIVIsION OF RE-SEARCH (1959).

Gottfredson & Bonds, Systematic Study of Experi-ence as an Aid to Decisions, RESEARCH REPORT No. 2,CALIFORNIA DEPARTMENT OF CORRECTIONS (1961).

10 S. GLUECx & E. GLUECK PREDICTING DELN-QUENCY AND CRIHE (1959).

1 1 D. GLASER, TnE EFFEcTivENEss oF A PRIsON ANDPAROLE SY sTEM (1964).12 Schrag, A Preliminary Criminal Typology, 4PAC. SOCIOLOGICAL REV. 11-16 (1961).

I G. SYxs, TEE SOCIETY OF CAPIvEs (1958).14 Miller, Some Characteristics of Present-Day Delin-

quency of Relevance to Educators (unpublished paperpresented at the 1959 meetings of the American Associ-ation of School Administrators).

25 Roebuck & Cadwallader, The Negroe Armed Robberas a Criminal Type: the Construction and Application ofa Typology, 4 PAC. SOcIOLOGICAL REv. 21-26 (1961).

'6 W. McCoRD, J. McCoRD, & I. ZOLA, ORIGINS OFCaI (1959).

"7L. ORLIN, SELECTION FOR PAROLE (1951).7SW. REcaxss, THm CRIME PROBLEM (1961).19 Jenkins & Hewitt, Types of Personality Structure

Encountered in Child Guidance Clinics, 14 AmL. J. ORTHo-PSYCHIATRY 84-94 (1944).

20 RE DL, supra note 5, at 42.21 ERIXSON, supra note 5, at 42.22 A. ACHORN, WAwARD YOUTH (1935).2 Makkay, Delinquency Considered as a Manifesta-

tion of: 1) a Serious Disorder of Development in EarlyChildhood, and 2) Other Delinquency-Prone Disturb-ances of Emotional Development (unpublished manu-script 1960).

2 Reiss, Social Correlates of Psychological Types ofDelinquency, 17 Am. SOCIOrOGICAL REv. 710-18 (1952).

2 Argyle, A new approach to the classifcation ofdelinquents with implications for treatment, 2 CALIFORNIASTATE BOARD OF COl uxCnONs MONOGRAPH 15-26(1961).

6 H. BLOCH & F. FLYNN, DELINQUENCY (1956).27 Illinois State Training School for Boys, Treatment

Committee (Report on Diagnostic Categories) (1953).

[Vol. 62

CLASSIFICATION OF OFFENVDERS

tions of Gough and Peterson,21 Hunt and lardt,29

Sarbin,'0 Peterson, Quay and Cameron," Gibbons,uStudt 33 MacGregor," Sullivan, Grant and Grant,"Warren,"8 and Russon.n

Several of the investigators listed under socialperception and interaction classification systemsmight equally well be grouped together on the as-sumption that their typologies all represent de-velopments in ego psychology, with importantunderlying concepts identified as stage of ego inte-gration, level of psychosocial development, level ofinterpersonal maturity, complexity of perceptualdifferentiation, level of cognitive complexity, etc.Such investigators as Hunt, MacGregor, Makkay,Sarbin, and Warren are currently working ontypologies of offenders, utilizing primarily egopsychology concepts.

In addition to the five groupings, some investi-gators, using a more eclectic approach by includingmeasures of several of the above areas of dimen-sions, have produced empirical-statistical typol-ogies. Among these investigators are Hurwitz,8Jesness,39 and Palmer.40 In a recent paper, theGluecks make a case for this approach and appear

28Gough & Peterson, supra note 4, at 42.2 Hunt and Hardt, Developmental Stage, Delinquency,

and Differential Treatment, J. ,sEARc IN CRn s &DELrNQ. 20-31 (1965).

10 Sarbin, supra note 4, at 42.3 Peterson, Quay and Cameron, Personality and

Background Factors in Juvenile Delinquency as InferredFrom Questionnaire Responses, 23 J. CONSrLTINPSYCHOLOGY 395-99 (1959).32 D. GIBBONS, CHANGING THE LAwBREA ER (1965);

Gibbons and Garrity, Some Suggestions for the Develop-ment of Etiological and Treatment Theory in Criminology,38 Socr FoRcEs 51-58 (1959); Gibbons and Garrity,Definition and Analysis of Certain Criminal Types, 53J. CRan. L. C. & P.S. 27-35 (1962).

33 E. STn r, S. MESSINGER & T. WiLsoN, C-UNIT:SEAuca Yon CoImmuI IN PRIsoN (1968).34 R. MacGregor, Developmental considerations in

psychotherapy with children and youth (paper pre-sented at the Annual Conference of the AmericanPsychological Association, St. Louis, 1962).

U Sullivan, Grant and Grant, The Development ofInterpersonal Maturity: Application to Delinquency, 20PsycAHuRx 373-85 (1957).

8 WARREN et al, INTERPERSONAL MATURY LEVELCLASSIICATION (JUvENILE): DIAGNOSIS AND TREAT-MENT or Low, MIDDLE, AND HIGH MATURTY DELIN-QUENTS, (1966).

'3 Russon, A Design for Clinical Classifcation ofOffenders, 4 CANADIAN J. CORRECTIONS 179-88 (1962).

"8Hurwitz, Three Delinquent Types: A MultivariateAnalysis, 56 J. Camn. L.C. & P.S. 328-34 (1965).

w Jesness, The Fricot Ranch Study, REsExcHr RE-PoR No. 47, California Youth Authority (1965).

40 Palmer, Types of Probation Offenders and Types ofYouth on Probation: Their Views and Ineractions,Youth Studies Center, Project Report (1963).

to be proceeding to develop a typology in thiseclectic manner.4

Each of the above classification systems is notequally relevant for all purposes. Some systemsconcern themselves solely with etiology, otherssolely with treatment. Some consider precipitatingfactors, others maintenance factors. Some focus onsocial organization, some on family organization,some on intrapsychic organization. Some are spe-cific to offender population; others have manydomains of applicability. Some are empirical-sta-tistical; some are empirical-observational; someare theoretical models. Some systems representcontinua or hierarchies; some are developmental.Some have many more direct treatment implica-tions than do others; some are more fruitful thanothers in producing research hypotheses.

Clearly, the last word on typologies has not beenwritten yet. Sociologists continue to accuse psycho-logical typologists of taking insufficient cognizanceof environmental factors; psychologists continue toaccuse sociological typologists of having insuffi-cient regard for intra-psychic factors. Nevertheless,it is now possible to find investigators who are at-tempting to theoretically link the sociological,psychological, and situational variables which areall relevant to a completely satisfactory taxonomy.

Cloward and Ohlin, in their book Delinquencyand Opportunity: A Theory of Delinquent Gangs,4-

note that, when identifying the cause of failure inthe legitimate system, some individuals blame thesocial order and others blame themselves. Cloward& Ohlin suggest that this differential perceptionlargely determines what the individual does abouthis failure. These authors note that attributingfailure to the social system is supportive of thedelinquent subculture, while attributing failure toself is supportive of the legitimacy of conventionalnorms. Cloward & Ohlin therefore indicate theneed to "... . identify the types of personality thatcharacteristically attribute causality (for failure)to themselves or to the world without."

In a recent article,4 3 Cohen notes that anomietheory must establish a more complete and success-ful union with role theory and theory of the self.He suggests that anomie theory is concerned with

4 Glueck and Glueck, Varieties of Delinquent Types,5 BrriS J. Cns. 236-48, 388-405 (1965).

42A. CLowARD AND L. OnxN, DELINQUENCY ANDOPPoRTuNITY: A THEORY or DELINQUENT GANGS 112(1960).

43 Cohen, The Sociology of the Deviant Act: AnomiTheory and Beyond, 30 Am. SOCIOLOGICAL REV. 5-14(1965).

19711

MARGUERiTE Q. WARREN

only one structural source of deviance and thatother deviant behavior is directly expressive ofroles. In seeking a general theory of deviance, heasks:

Is it possible to make any general statementsabout the kinds of deviance that may be attributedto anomie and the kinds that may be attributed torole validation through behavior culturally signifi-cant of membership in the role? Or may twoinstances of any sort of deviant behavior, identicalin their manifest or 'phenotypic' content, differin their sources or 'genotypic' structure?

In a recent paper,4" Warren has attempted toidentify within the delinquent population thosesubgroups for which sociological factors (socialdisorganization, differential association, inadequateaccess to the legitimate opportunity structure, etc.)appear to have the greatest causal significance,those subgroups for which psychological factors(internal conflict, identity struggles, inadequatesocialization, etc.) appear to have the greatestrelevance, and those subgroups for which situa-tional factors (acute family crisis, etc.) appear mostimportant in leading to the delinquent act.

As in all science, criminological investigatorsapproached the problem by first looking for thesimplest explanation of events. However, as ourknowledge has grown, it has become necessary tolook at the subject matter in an increasingly com-plex fashion in order to handle the data that hasaccumulated.

It is a well accepted principle in psychology thata single behavioral event may stem from a numberof different causes or motives, and that any singlecause or motive may lead to any one of several dif-ferent behaviors. That is, the delinquent act as abehavioral event may occur because of a strongyouth's agitation of a weak youth, because of anadolescent's need to conform to a peer group'sprescription for acceptance, because of the anxietyand despair which a family member feels in a fam-ily crisis, because of a youth's need for a car totransport his girl friend to the dance, etc. The be-havioral event-a car theft, for example-mighthave risen from any of the listed causes or stillothers. With regard to the second part of thepsychological principle-that is, that any singlecausal factor may result in different kinds of

4"Warren, The Community Treatment Project: AnIntegration of Theories of Causation and CorrectionalPractice, (Paper presented at the Illinois Academy ofCriminology Conference, Chicago, 1965).

behavior-it is possible for one to know muchabout the causal factors in a particular delin-quency and still be unable to ascertain with cer-tainty why the individual committed.an act whichled to his appearance in the delinquent systemrather than committing an act which, for ex-ample, led to his appearance in a mental hygieneclinic. There are at least two reasons for belaboringthis fairly obvious point. First, there are still thosewho persist in discussing the cause for delinquencyor who persist in seeking a cause which will explain"most" of delinquency. Secondly, when the focusis on the management and treatment of offenders,distinguishing among the varieties of causal factorsbecomes crucial to the establishment of differentialgoals and methods for transforming offenders intonon-offenders.

A classification system for offenders need notserve all purposes in order to be adequate for somepurposes. However, certain factors are importantin all taxonomies. In addition to the usual criteriaexpected of a good typology, such as completecoverage of the relevant population, clear-cut,non-overlapping categories, internally meaningfuland consistent categories, and parsimoniousness,it is especially important to any classification sys-tem used for scientific purposes that the types besufficiently well defined so that the abstractionscan be used with high reliability by trained raters.Beyond these general requirements, it is possiblefor certain purposes to use a classification systemwhich, for example, has no etiological referents, onewhich has no implications for treatment, or onewhich is specific to an institutional setting.

Classification systems which are useful solely formanagement purposes are distinguishable fromthose which are more relevant for establishingtreatment goals. For purposes of this paper, theterm "management" means efficient and effectivecontrol over the behavior of the offender so thatfurther law violations are not committed duringthe period of agency responsibility for the offender.In contrast with "management," the term "treat-ment" refers to attempts to change the individualoffender or the relevant aspects of his environmentso that long-term non-violation behavior is as-sured beyond the period of direct agency responsi-bility for the offender.

CLASSIFICATION FOR ANAGEMENT PURPOSES

Efficient and effective management in an institu-tional setting involves protecting those who are

[V61. 62

CLASSIFICATION OF OFFENDERS

weak from those who are strong, those with rela-tively nondelinquent attitudes from those withstrong delinquent orientations, those who are easilyagitatable from those who agitate, and those whoare non-homosexuals from those who are homo-sexuals. Since a correctional agency has a mandateto protect the community from offenders, inmateswith high escape potential must be identified andplaced in maximum security facilities. All of thesediscriminations imply the need for a classificationof offenders on a variety of dimensions. Other areasof management decision which require some clas-sification of inmates in an institutional setting in-lude: open versus closed institutions, single versusdormitory rooms, amount and kinds of punish-ment, job assignment, time in the institution, useof tranquilizers, custody security level.

In a field setting, management primarily involvescontrol of offenders to prevent further law viola-tions in a way that protects both society and theoffender at a "reasonable" price. This means, forexample, assigning to high surveillance conditionsonly those who require constant external controlsto prevent crime, and assigning to low surveillanceconditions individuals who represent low threat inthis regard. It also involves decisions regarding ex-tent of the parolee's freedom to determine his ownliving arrangements, his job, and his obligations.

All of these management decisions require animplicit or explicit classification system. The diffi-culty, of course, with an implicit grouping is thatthere is no way of checking the accuracy or thevalue of the system there is no built in self-correct-ing process. Currently, in the reception and diag-nostic centers of many correctional programs, de-cisions are made with regard to "rehabilitating" aparticular offender, using the variety of conditionsand programs available to the correctional system.Recommendations for decisions are typically madeby intake workers using a subjective weighting ofnumerous opinions, impressions, and perhaps a feweducational and aptitude measures. The basis ofthe intake worker's judgements may be clear orunclear in his own mind. In either case they arelikely based on uncorrected personal biases, sincehe rarely finds out whether or not his recommenda-tions were, in fact, carried out and, if carried out,whether or not they led to a "rehabilitated" of-fender. Even if feedback to the intake worker werecomplete with regard to the effectiveness of hisrecommendations, as long as the basis for judge-ments remained implicit and intuitive, the cor-

rectional system would benefit only when experi-enced intake workers were on the job. It is onlywhen recommendations are made on explicit di-mensions and expectations that the system has thebenefit of checking out expected relationships andpassing along relevant information to new and in-experienced workers.

The prior probability approaches, supra at 240-41 are examples of classification systems usefulfor management purposes. Decisions regardingwhether a particular offender is to be handled inthe community or in an institutional setting canmost rationally be made by considering, amongother things, the offender's risk of parole violation.Surveillance level on parole and related aspects ofcaseload size may be determined in part by knowl-edge of probability of violation. In an interestingexperiment in the California Department of Cor-rections, 45 parolees who represented low risk ofparole failure (as predicted by Base Expectancyscore) were assigned to minimum supervision case-loads (one contact with parole agent every threemonths). Violation rates of this experimental groupwere no higher during a 12-month follow-up thanviolation rates of a comparable control group whichreceived regular parole supervision. 4

1

Prior probability classification systems may beused, not only as an aid to administrative decision-making, but also as a check on whether or notmanagement decisions have the desired effect. In astudy reported by Gottfredson,47 a correctionalagency planned to release from an institution some-what earlier than would be expected a group pa-roled to special reduced caseloads. The goal in-volved was that of decreased confinement costs forthe selected group without any increase in paroleviolations. Two prediction classification schemeswere needed to control known biases in selectingcandidates for special parole programs: (1) a clas-sification of offenders by parole violation riskgroup, and (2) a classification of expected prisonterms under an indeterminate sentence law. Using

4 5Havel, Special Intensive Parole Unit IV: The HighBase Expectancy Study, Research Report No. 10, De-partment of Corrections (California) (1963). See alsoInteraction Between Treatment Method and OffenderType, 1 CAL. ST. Bn. o ConxciroNs M0NOoGRApn27-30 (1960).

4SCompared with minimal supervision, regularsupervision involved one third more office contacts,twice as many field contacts, and more than twice asmany collateral contacts.

4 Gottfredson, A Strategy for Study of CorrectionalEffectiveness. (Paper presented at the fifth Interna-tional Criminological Congress, Montreal 1965).

19711 ,


these two classification systems, the study showedthat: overall, men selected for the special programdid not serve shorter terms; first termers selectedfor the special program tended to serve less time,while recidivists selected for the special programtended to serve more time; for the total selectedgroup, no differences in parole violation werefound; first termers selected tended to have mark-edly fewer violations during the first year on pa-role, while recidivists selected tended to have moresuch violations. Provided with these classificationand accounting procedures, it was possible for theadministrator to test whether or not paroling de-cisions had been made consistently with policy ob-jectives.

There are a number of studies using prior prob-ability and psychiatric-oriented classification sys-tems which have implications for the kind of settingin which various subgroups of offenders may bestbe handled. The Borstal studies48 and Week's studyof Highfieds49 are examples of research showing arelationship between kind of inmate and kind ofcorrectional setting. Both studies show the mainadvantage of open institutions over closed institu-tions to be for the better risk inmates. A study byReiss50 suggests that all delinquents with relativelystrong personal controls should be assigned to homeand community placement; whereas, assignmentsto short terms in institutions or to communityplacement contingent on case progress should bemade for delinquents with relatively weak personalcontrols; and assignment to closed institutionsshould be made for those with marked social de-terioration or very immature personalities. Beck5'suggests that Socialized type delinquents should beplaced in an open, relaxed, institutional atmospherebest suited to the diversion of their delinquentenergy. Unsocialized Aggressive type delinquentsshould be placed in a controlled institutional en-vironment, since permissiveness will only make thisgroup more difficult to handle. Argyle,52 among hismany recommendations, suggests that the Deviant

4 8 H. MIANNimim AND L. WILKiNs, PxEDICTioNMIETHODS IN RELATION TO BORSTAL TRAINING (1955).

49 H. WEEKS, YOUTHrUL OPPENDERs AT HIGIn'ELDS(1958).50 Reiss, Delinquency as a Failure of Personal andSocial Controls, 15 Ams. Soc. REv. 196-207 (1951).51 Beverly, A Method of Determination of Base Expec-tancies for Use in the Assessment of Effectiveness of Cor-rectional Treatment, Research Report No. 3, CaliforniaYouth Authority (1959).

52 Argyle, A New Approach to the Classification ofDelinquents with Implications for Treatment, 2 CAL.ST. BD. OF CORRECTIONS MONOGRAPH 15-26 (1961).

Identification type delinquent should be separatedfrom his peer group and installed in an essentiallynondelinquent environment.

Several of the social perception and interactionclassification systems have been used in makingmanagement recommendations or decisions. Gib-bonsn bases his typologies of juvenile and adultoffenders on patterns of social roles as defined byoffense behavior and career, and by self conceptand attitudes. Among other management recom-mendations, Gibbons suggests that PredatoryGang Delinquents be segregated from other boysin order to minimize victimization; that Non-Gang,Casual Delinquents be kept out of the correctionalsystem, i.e. merely threatened and released in asmuch as no intervention is required in such cases;that the Automobile Thief-"Joyrider" be di-verted from the "tough guy" pose in an institutionby recreational and athletic programs; that HeroinUsers be placed in protective environments typi-fied by milieu-management programs such asSynanon; that Overly Aggressive Delinquents beforcibly controlled initially in a residential setting;etc.

Using Warren's Interpersonal Maturity Classi-fication System: juvenile, 54 Jesness conducted astudy55 in which inmates of a boys' training schoolwere assigned to living units on the basis of delin-quent subtype, and an attempt was made to de-velop and describe the management techniquesmost useful in dealing with each subtype. Warrenand the staff of the Community Treatment Pro-ject have developed a treatment model which de-fines nine delinquent subtypes and prescribes bothdifferential management and treatment techniquesin the community for the various subtypes.5" Thenature of controls to be used by the treatmentagent, characteristics of a suitable placement,school, job, and leisure time recommendations aredescribed.

CLAssiricATIoN TrOR TR4ATIIENT

The function of treatment in a correctional pro-gram is to modify the characteristics of the offenderand/or the aspects of his environment which areresponsible for his involvement in deviant activ-ities. From many treatment prescriptions, it is clearthat, in addition to the long-term prevention of

53 D. GIBBONS, CHANGING THE LAWBREAxERS (1965).54 Warren, supra note 36.55 C. JEsNEss, THE PRESTON TYPOLOGY STrMnY

(1970).66 Warren, supra note 36.

[Vol. 62


law violations, there is also the intent to bringabout changes in the offender and in his societywhich will reduce his cost to society in other waysby, for example, decreasing the chances of the in-dividual's depending on welfare or unemploymentrolls, or by increasing the individual's responsi-bility as a family member and as a citizen.

One source of evidence for the importance of aclassification system which differentiates amongsubgroups of the delinquent population is providedby treatment studies. Studies of the impact oftreatment of client populations have been gener-ally discouraging. No one has yet empiricallyanswered Eysenck's challenge that the proportionof mental patients improved following treatment isapproximately the same as the spontaneous re-mission rate.5 Reviews of the correctional litera-ture tell a similar story--some studies showing thetreated to be considerably improved followingtreatment, some showing negative effects, andmost showing no difference. Bailey,3 in a review ofone hundred correctional outcome studies con-ducted between 1940 and 1959, noted that thosestudies which exhibited the most rigorous experi-mental designs reported either more harmful ef-fects of treatment or no change. A fairly typicalstudy is the one which produces contradictory evi-dence about improvement, with the treated sub-jects looking improved on some measures of changeand either unimproved or in worse condition onother behavioral measures (see, for example,O'Brien 9).

How should these negative and inconclusivestudies be viewed? One possibility is that, in ourpresent state of knowledge, treaters simply don'tknow how to bring about changes in individualsvia a treatment process. However, another pos-sible explanation is available, an explanation illus-trated by the PICO I study and by the CampElliott study. Adams" reported on a three-yearfollow-up of youthful offenders who had taken partin the Pilot Intensive Counseling study, a programof individual interview therapy. Subjects in the

67 H. EYSENCK, TxM SCIENTrIC STUnY Or PER-soNAmT= (1952).

8 Bailey, Correctional Outcome: An Evaluation of100 Reports, 57 J. Cnr. L.C. & P.S. 145 (1966).

N O'Brien, .Personaity Assessment as a Measure ofChange Resulting from Group Psychotherapy with MaleJuvenile Delinquents, California Youth Authority(1961).

60 Adams, Interaction Between Individual InterviewTherapy and Treatment Amenability in Older YouthAuthority Wards, 2 CAL. ST. BD. oF CORXaCTONSMONOGRAPH 27-44 (1961).

study were classified as "amenable" and "non-amenable" to treatment; both groups were thenrandomly assigned to treatment or nontreatmentconditions. Parole performance of the four sub-groups was compared on many criteria of per-formance. The treatment amenable group had asignificantly better parole record than the non-treated amenable group. Furthermore, the treatednonamenable group had the poorest parole recordof the four subgroups, poorer than either the non-treated amenables or the nontreated noname-nables.

The Camp Elliott study by Grant and Grant"investigated an experimental living group programwith military offenders. Among the several con-trolled conditions in this study were the inter-personal maturity levels of individual prisoners inthe living and treatment groups and the char-acteristics of the supervisory team. The most im-portant finding from this study was that the inter-action between the maturity level of the subjectsand the supervisor characteristics significantly af-fected later success rate of subjects. Not only werethe treatment methods of some internally-orientedsupervisory teams effective in increasing the suc-cess rates of high maturity offenders, but also, thetreatment methods were markedly detrimental tothe success chances of low maturity offenders.Furthermore, the externally-oriented supervisoryteam had the reverse effect on high and low ma-turity subjects. As long as the data of the CampElliott program was used as a study of single var-iables, its findings were comparable to those ofmany other correctional studies: that is, no de-monstrable treatment (supervisory effectiveness)effect, and only a low, though significant, classi-fication (maturity) effect.. In both the Camp Elliott and the PICO I stud-ies, it was only when the interaction of the treat-ment and classification variables was consideredthat one found productive relationships with latersuccess/failure rates. Thus, by lumping togetherall subjects, the beneficial effects of a treatmentprogram on some subjects, together with the detri-mental effects of the same treatment program onother subjects, may each mask and cancel out theother.

It is very likely that, in many treatment studies,this masking effect has occurred, either because-the

6" Grant and Grant, A Group Dynamics Approach tothe Treatment of Nonconformists in the Navy, 322A.NAxs or A.mmECAw AcADEuY or PoLITICAL ANDSocrAL SciENc=E 126-35 (1959).

19711


data have not been viewed in sufficiently complexfashion, or because the crucial dimension, the clas-sification of subjects in a treatment-relevant way,was missing. If one accepts the notion that offend-ers are different from each other in the reasons fortheir law violations, then it appears rather obviousthat attempts to change the offender into a non-offender will vary in ways which are relevant to thecause. Ideally, the goals of the treatment will relatein some direct manner to the causes of the delin-quency, and the treatment methods will relatespecifically to the goals for the various offendersubgroups.

Treatment decisions which must be made by acorrectional organization involve in part some ofthe same issues involved in management decisions.For example, the correctional setting may be atreatment tool as well as a management tool.Mueller'2 conducted a study in which "treatment"was defined as the setting in which the offender washandled. The "treatments" available were (a) re-lease to direct parole in the community, (b) for-estry camp, and (c) training school. Mueller founddifferential effects of these treatments over kindsof delinquents. Conforming and over-inhibitedboys had higher parole success rates when assignedto non-institutional or open institutional programs.Assigning aggressive or insecure delinquents to anyprogram did not lead to greater success. Subjectsleast like socialized delinquents and most like emo-tionally disturbed delinquents were more success-ful on direct parole, almost as successful in andfollowing camp assignment, and more inclined tofail than succeed in and following a training schoolexperience.

Another group of treatment variables which maybe differentially prescribed for various subgroups ofoffenders relates to the characteristics of the treater.An excellent study attempting to match types ofprobation officers with types of youth on probationwas carried out by Palmer.0 Ratings were madefrom recorded interviews with officers and proba-tioners, and the ratings were cluster-analyzed. Theanalyses yielded three distinct empirical groupingsof officers and eight groupings of youths. Theempirical dusters of probationers were labeled:(a) Communicative-alert, (b) Passive-uncertain,

62 Mueller, Success Rates as a Function of TreatmentAssignment and Juvenile Delinquency ClassificationInteraction, 1 CAL. ST. BD. OF CORRECTIONS MONO-GRAPH 7-14 (1960).

63 Palmer, Types of Probation Offenders and Types ofYouth on Probation; Their Views and Interactions,Yourx Sruiis CENTER, PRojxcT REPORT (1963).

(c) Verbally hostile-defensive, (d) Impulsive-anxious, (e) Dependent-anxious, (f) Independent-,assertive. (g) Defiant-indifferent, (h) Wants to behelped and liked. As measured by an index ofyouths' evaluation of the relationship with theirofficer and view of the overall effectiveness of pro-bation, a number of interactions between officertype and probationer type were shown. For exam-ple, Relationship/Self-expression officers achievedtheir best results with youths who were Com-municative-alert, Impulsive-anxious, or Verballyhostile-defensive. Surveillance/Self-control officershad their greatest difficulties with individuals whowere Verbally hostile-defensive or Defiant-indif-ferent. Surveillance/Self-expressing officers seemeduniquely matched with probationers who wantedto be helped and liked.

A third group of treatment variables which maybe differentially prescribed for various subgroups ofoffenders relates to characteristics of programs andspecific therapeutic methods. Many clinical reportscan be found in the literature which suggest dif-ferential programs for specified kinds of offenders.To date, few programs have offered any supportiveresearch evidence for stated hypotheses. In theline of recommendations, Jenkins and Hewitt"have suggested the following treatment programfor the Unsocialized Aggressive delinquent. Thereshould be a warm and accepting attitude on thepart of the therapist. He should, in small steps, es-tablish and effectively maintain pressure towardrequired behavior and against certain objection-able types of behavior. Jenkins and Hewitt believethat the methods suitable for use with the Neuroticchild will make the Unsocialized Aggressive childworse; for example, the encouragement of free ex-pression of aggression for this type of child does nothelp because his well of hostility is bottomless.Jenkins and Hewitt's thinking on the treatment ofthe Socialized or Adaptive delinquent appears tobe based on the assumption that, for this child, thedelinquent behavior is a function of social status,role, peer associates, group identifications, and theattitudes and values learned through social con-tacts. The treatment plan, therefore, is based onthe child's fundamental socialization, capacity forloyalty, capacity to identify with a masculine,socialized adult. The methods the authors suggestare somewhat similar to those suggested by CliffordShaw and his associates in the Area Projects in

64 Jenkins and Hewitt, Types of Personality Struc-ture Encountered in Child Guidance Clinics, 14 Am. J.ORTHOPSYcHATRY 84-94 (1944).-

(Vol 62


Chicago for'use with the group often known asCultural delinquents.

In their book, Origins of Crime, McCord, Mc-Cord and Zola65 suggest six different treatmentplans for six offense types--criminals who commita wide range of anti-social acts, those who commitcrimes against property, those who commit crimesagainst persons, sex criminals, drunkards, andtraffic offenders. The recommended treatment forthose who commit crimes against property, forexample, centers on the giving of attention andrecognition, and on the provision of consistent,nonpunitive discipline.

Also in the line of recommendations for treat-ment, Gibbons"8 offers suggestions for differentialtherapeutic methods for his various subtypes de-fined by social role. For the juvenile subtypes,Gibbons recommends group therapy for Gang de-linquents and Joyriders, intensive individual psy-chotherapy for Overly Aggressive delinquents,depth psychotherapy for Behavior Problem de-linquents, milieu therapy for Heroin Users, groupor individual client-centered counseling and familytherapy for Female delinquents, and no treatmentfor Casual delinquents. For the adult subtypes,Gibbons recommends group therapy for Semipro-fessional Property Offenders and Violent SexOffenders, client-centered counseling for NaiveCheck Forgers and Nonviolent Sex Offenders,intense individual psychotherapy for "psycho-pathic" Assaultists, no treatment but help withcommunity adjustment for Professional "Fringe"Violators, Embezzlers, Personal Offenders, and"One Time Losers," and, lastly, altering societyso that consistent law enforcement is maintainedis recommended treatment for White Collar Crimi-nals.

In an attempt to increase the precision andeffectiveness of social casework practicd, Free-man, Hildebrand and Ayre, working at thePittsburgh Family and Childrens Service, havedeveloped a typology of clients with corollarytreatment techniques. While this typology is notspecific to the offender population, it is an excel-lent example of a treatment model built from clini-cal experience and clinical need. The underlyingdimension relating the types is a continuum oflevels of emotional maturity or ego autonomy.

11 W. McCoRD, J. McCoRD, & I. ZOLA, ORIGINs OFCnRIU (1959).6s D. GIBBONS, CHANGING ThE LAWBREAioR (1965).

97Freeman, Hildebrand, and Ayre, A ClassifiatiomSystem that Prescribes Treatment, 46 SoCmL CAsnwoRx423-29 (1965).

The authors suggest "That the treatment tech-niques most appropriate to the task of strengthen-ing the coping powers of each type are prescribedby the very nature of the ego structure and theparticular stage of ego development." 6 Thistypology has much in common with the typologiesof Hunt,6 9 MacGregor,70 and Warren7' in that theyare all'also based on an underlying developmentalgrowth continuum.

MacGregor, 72 in a research study of the familiesof middle class delinquent youth, has developed atypology of family patterns. Products of the studyare a set of propositions by which families may beclassified for treatment planning. The familydiagnosis, labeled in terms of the arrest in develop-ment of the nominal patient, are:. Type A Infantile functioning in adolescence

(schizophrenia);Type B Childish function in adolescence or

preadolescence (character disorder), theAutocrats;

Type C Juvenile functioning in adolescence orpreadolescence (childhood neurosis),the Intimidated Youth;

Type D Preadolescence functioning' in adoles-cence (adjustment reaction of adoles-cence), the Rebels.

The bases of the diagnosis involves ratings ofsuch factors as family response to crisis, familyrelationship with community, family leadershipand exploitation, sibling interaction, and familycommunicative style. The general stated thera-peutic goal is to help a family allow its youth toadvance beyond the developmental arrest in whichall participated. The major method for achievingthis goal is multiple impact therapy, i.e, two daysof concurrent sessions with varying combinationsof therapeutic team and family members. -

The following are some treatment recommenda-tions made for Type D: The defiant Rebel shouldnot have his responsibilities diminished. Rebellionshould not be encouraged, but respect for theRebel's opinions and standards should be shown bythe treatment team. Identification of the childwith the father should be pointed out to the father,

68Id. at 429.6O. HARvEY, D. HuiN, & H. Scnxonxn, CoNcEp-

TuAL SysT'zs AND PERSONAITY ORGANIZATioN (1961).70 MacGregor, Developmental Considerations, in

Psychotherapy With Children and Youth (paperpresented at the annual conference of the AmericanPsychological Association, St. LOuis, 1962).71 Warren, supra note 36.7' MacGregor, supra note 70.

19711


and he should be encouraged to offer more opensupport to his wife.

Treatment recommendations for Type B include:Help the mother to turn to the father, rather thanthe child, for emotional release, and help the fatheroffer the mother emotional support. Help parentsget over fear of exposing themselves to competitiveevaluation at home. Encourage father to trusthimself to intervene more directly to influence thechildren. Help mother relinquish her aggressivepower role and trust husband's leadership. As amodel, treatment team members should demon-strate healthy and vigorous interaction for the par-ents. Mother should be encouraged to develop in-terests other than child-rearing. Father-child inter-action should be encouraged by having them, in themother's absence, discuss their dealings with her.Ways of decreasing parental dependence on himshould be discussed directly with the child. TheAutocrat should be made to see that he is beingexploited as much as he is controlling others. TheAutocrat should be prepared for the changingbalance of forces in the family, and the parentsshould be prepared to meet the tests of the changewhich the Autocrat will present.

Based on a theory of socialization-ConceptualSystems7 3 Hunt and Hardt have related develop-mental stage, i.e., Conceptual Level, to delinquentbehavior and delinquent orientations, and havespeculated about the implications of the theoreti-cal model for differential treatment of delin-quents.74 Five Conceptual Levels are defined, eachlevel characterizing the person's interpersonalorientation, that is, his knowledge about himselfand the relation between himself and others. Amajor application of the Conceptual System modelhas occurred in the field of educationY5 Diagnosesof Conceptual Level were made on students in alower class, junior high school population, andstudents classified at one of three lowest levels wereassigned to classrooms which were homogeneousby developmental stage. Differential managementand teaching methods were reported by teachershandling the various groups. On the basis of thisstudy, Hunt defined optimal environments forindividuals at the three stages. Since research hasshown that these stages bear relationships to

73 Harvey, supra note 69.74 Hunt & Hardt, Developmental Stage, Delinquenay,

and Differential Treatment, J. RasEaacH iN Canen &DEIINQ. 20-31 (1965).

75 Hunt & Dopyera, Personality Variation in Lower-Class Children, 62 J. or PsycuoLoGy 47-54 (1966).

delinquent behavior and orientation, 7 Hunt andHardt have drawn implications from the educa-tional study for the differential treatment ofdelinquents.

The overall change goal in this system is move-ment from a lower to a higher conceptual stage.In the context of this general aim, specific sugges-tions are made regarding treatment methods ateach level. For example, boys classified as Sub I"require activities (rather than discussions)focused on the present and organized verydearly."7 The training agent should offer theSub I boy "controlled experiences in which he istangibly responsible for outcomes." For the Stage Iboy, the training agent initially should exhibitauthority clearly, since persons at this stage arevery dependent on normative expectations. Even-tually the "agent should attempt to encouragegreater self-responsibility and an appreciation ofalternative solutions." In working with Stage IIboys, the training agent should help the boy dis-cuss his behavior and consider alternative solu-tions to his problems. A long-term goal for this boywould be to acquire empathy by beginning tounderstand that some of the feelings of others aresimilar to his own.

The work of Warren and associates at the Cali-fornia Youth Authority's Community TreatmentProject is based on the theory of Levels of Inter-personal Maturity, a formulation describing asequence of personality integrations in normalchildhood development." In many ways similar tothe Conceptual System theory, the InterpersonalMaturity Level Classification system focuses uponthe ways in which the individual is able to perceivehimself and the world, and understand what ishappening among others as well as between him-self and others. According to the theory, sevensuccessive stages of interpersonal maturity charac-terize psychological development, ranging fromthe interpersonal reactions of a new born infantto an ideal of social maturity. Every person doesnot necessarily work his way through each stage,and may become fixed at any particular level. Therange of maturity levels found in an offenderpopulation is from Maturity Level 2 (IntegrationLevel 2 or 12) to Maturity Level 5 (I). Level 5occurs with relative frequency in an adult popula-

7 6 Hunt & Hardt, supra note 74.7Id. at 30.7 8 Sullivan, Grant, & Grant, The Development of

Interpersonal Maturity: Appliations to Delinquency,20 PsYcnrATRY 373-85 (1957).

[Vol. 62


tion, but is rare in a juvenile delinquent popula-tion, so that Levels 2 through 4 are sufficient todescribe the cases in the Community TreatmentProject.

An elaboration of the original classificationsystem was developed by Warren in 1961 for usein the Community Treatment Project (CTP).After assuming that a diagnosis of Maturity Levelis identified a group of individuals with a commonlevel of perceptual differentiation, it became ap-parent that not all of the individuals in this groupresponded to this perceptual level in the same way.An attempt was then made to classify types withineach Maturity Level according to response set.In this manner, nine delinquent subtypes wereidentified, i.e., two 12 subtypes, three 13 subtypesand four 14 subtypes. In the 1961 elaboration, thenine subtypes were described by means of itemdefinitions characterizing the manner in which themembers of each subgroup perceive the world, andare perceived by others. At the same time, manage-ment and treatment plans were prescribed for eachsubtype. These management and treatment pre-scriptions grew primarily from the theory, butalso, to some extent, from previous work withmilitary offenders 9 and with prison inmates."

Based on the 1961 treatment model, the CTPbegan to treat serious delinquents in a communitysetting instead of an institutional setting. In thenine years of the Project's existence, the charac-teristics items for each subtype have increased andbecome more detailed, and the treatment strate-gies have become increasingly specific and realistic.Current descriptions of the nine delinquent sub-types, with predicted most effective interventionor treatment plans, combine to make up the 1966edition of the treatment model." This model ismuch too lengthy and elaborate to review here. Itis possible only to note the various areas coveredby the intervention prescriptions. The specificgoals of intervention for each subtype follow fromthe nature of the problem, as defined in the charac-teristics items. From each goal, a specific interven-tion method follows. The treatment plan pre-scribes: the characteristics of an appropriateplacement, preferred family treatment, school

nGrant & Grant, A Group Dynamics Approach tothe Treatment of Nonconformists in the Navy, 322Ax Ar.s or AmmcA ACADEmy or PomrcAL &SocIAn ScIENcE 126-35 (1959).

80 Grant, A Study of Conformity in a NonconformistPopulation (1961). (Unpublished Ph.D. dissertation,University of California, Berkeley.)

"1 See Warren, supra note 36, at 46.

and/or job recommendations, sources of com-munity support, leisure time activities, recom-mendations regarding peer group variables,required controls, specific therapeutic methods,characteristics of an appropriate treatment agent,and support required by the treatment agentworking with the subtype.

Cnoss-CLAss ICATIoN or TYPOLOGIES

During 1966, a conference on typologies ofdelinquents was sponsored by NIMH and attendedby a number of the investigators whose work isreported in this paper- Hunt, Hurwitz, Jesness,MacGregor, Makkay, Reiss, Quay, and Warren.David Bordua, as well as David Twain and Sey-mour Rubenfeld of the NIMH staff, also partici-pated in the conference. A cross-tabulation of theclassification systems was attempted. Three orfour broad bands across the classification systemswere identified and tentatively agreed upon by theconference participants. A further breakdown intosix cross-classification bands seems also possible 3

Chart A presents a cross-classification of the ty-pologies represented at the NIMH conference plusa tentative cross-tabulation of other classificationschemes.'

4

Within the first of these bands, to be called forpurposes of this paper the Asocial type, are in-eluded Hunt's Sub I type, Hurwitz's Type 11,Jesness's Immature-aggressive and Immature-passive, MacGregor's Schizophrenic youth,Makkay's Antisocial Character Disorder-Primi-tive (aggressive and passive-aggressive), Quay'schildren high on Unsocialized-psychopathic fac-

8It should be noted that, of those who presentedclassification schemes, all but Quay referred to theirsystem as a typology. Quay prefers to view classifica-tions in terms of dimensions of behavior. See Quay,The Structure of Children's Behavior Disorders (1965).(colloquia at the University of Minnesota and theUniversity of Maryland). See also, Personality Dimen-sions in Delinquent Males As Inferred From the FactorAnalysis of Behavior Ratings, 1 J. REsEA~ca nu CRnAwD Dxram. 33-37 (1964).

Since the cross-classification presented here issomewhat more complex than the one discussed at theNIMYH conference, the responsibility for errors ofplacement should be viewed as entirely that of theauthor.

4Several of the classification schemes reviewed forthis paper were not included in the cross-classificationbecause the typology did not make enough discrimina-tions (Aichom, Lejins) because the typology pur-portedly differentiated among disturbance areas withinthe individual rather than among individuals (Redl),or because the nature of the underlying bases of thesystem did not relate to those charted (Ohlin, WalterMiller).

19711

CHART A____________ ________________________ CROSS-CLAeSSIICAIOeop OrPEisuza TYenLooIES

Subtypes

1. Asocial

Aggressive

Passive

2. Conformist

Nondelin-quently-oriented

Delinquently-oriented

3. Antisocial-ma-nipulator

4. Neurotic

Acting-out

Anxious

S. Subcultural-identifier

6. Situational

Types not cross-classified

Jesness

Immature, ag-gressive

Immature, pas-sive

mnmature, pas-sive

3ocialized con-formist

Nfanipulator

Neurotic, act-ing-out

Neurotic,anxious

Neurotic, de-pressed

Zultuml delin-quent

Hunt Hur- Mac- Iwits Gregor Makkay

.1--I-I-I

3tage fTypeIf III

Antisocial Char.acter Disor-der-Primitive

Aggressive

Passive-aggres-sive

Antisocial Char.acter Disor-der-Organized

Passive-aggres-sive

Antisocial Char-acter Disor-der-OrganizedAggressive

Neurotic

Subcultural

Mental Retard-datePsychotc

Quay Warren APA Argyle Gibbons

! - 1 1 - - 1III __________ I __________ _________

Unsocial-ized-psy-chopath

Inadequate-immature

/?Subcul-tural/

Neurotic-

disturbed

Subcultural

/?Rela-tively in-tegrated/

Defectivesuperego

Relativelyweak ego

Relativelyintegrated

I

Asocial, aggres-sive

Asocial, passive

1:

Conformist, Im-mature

Conformist, Cul-tural

IS

Manipulator

I Neurotic

Neurotic, act-ing-out

Neurotic,anxious

14Cultural identi-

fier

ItSItuational,

emotional re-action

Passive-aggres-sive personal-ity

Aggressive

Passive-aggres-slve

Passive-aggres-sivepersonal-ity

Passive-depend-ent

Antisocial per-sonality

Sociopathic per-sonality dis-turbance

Dyssocial reac-tion

Adjustment re-action of ado-lescence

Lack ofsympathy

Inadequatesuperego

Inadequatesuperego

Weak egocontrol

Deviantidentifi-cation

Overly ag-gressive

Gang of- 1/?Social-fenders ized/

joyrider

Behaviorproblems

Gang of-fenders

Casual de-linquent

Heroin userfemale de-linquent

Jeand n McCord

HewittReckless

Unsocial-Ized ag-gressive

Scbrag

Asocial

Anti-social

Pseudo-social

Prosocal

Studt

Isolate

Receiver

Manipu-lator

Love-seeker

Conform-ist

Aggres-sive

(psy-cho-pathic)

Neurotic-with-drawn

Overin-hlbited

Socialized

Psycho-path

NeuroticPerson-ailty

Offendersof themoment

Eruptivebehav-ior

kntisocialjLearner


tor, and Warren's 12 Asocial aggressives and Asocialpassives. To this general classification band tenta-tively 5 can be added: Argyle's Lack of Sympathytype, Gibbon's Overly Aggressive delinquent,Jenkins and Hewitt's Unsocialized Aggressivedelinquent, Schrag's Asocial type, and Studt'sIsolate. Behavioral and family history characteris-tics of offenders who fall in this first classificationband are generally agreed upon. The offenderclassified in this band is described as primitive,under-inhibited, impulsive, hostile, insecure, inade-quate, maladaptive, concretely negativistic, undif-ferentiated, demanding of immediate gratification,non-trusting, thoroughly egocentric, alienated, etc.It is generally agreed that this type of offenderdoes not see himself as delinquent or criminal, butrather seems himself as the victim of an unreason-able, hostile and confusing world. Those typolo-gists who have investigated etiological factorshave consistently shown extreme emotionaldeprivation, generalized and continual parentalrejection, and frequently, physical cruelty orabandonment. Most investigators who relate tothe treatment question for this type recommend asetting which offers a clear and concrete structureof low pressure, warmth, and acceptance from anextremely patient parent substitute, slow andsupportive direction toward conformity, and at-tempts to reduce the fear of abandonment andrejection via teaching rather than psychotherapy.86

The second broad classification band which cutsacross typologies, the Conformist type, includesHunt's Stage I group, Jesness's Immature-passiveuand Socialized conformist, Makkay's AntisocialCharacter Disorder-Organized (passive-aggres-sive), Quay's children high on Inadequate-im-mature factor, and Warren's 13 Immature conform-ists and Cultural conformists. To this classificationband tentatively can be added: Argyle's Inade-quate Superego delinquent, Gibbon's Gangoffenders, McCord's Conformists, and Studt's

85 These cross-classifications have not been checkedwith the authors of the classification systems.

81The definitions of subtype characteristics, thedescriptions of etiological factors, the treatment recom-mendations-none of these for this subtype nor thefollowing subtypes do justice to the detailed and exten-sive work of some investigators. The intent here issimply to indicate in very general terms examples ofapparently agreed-upon and disagreed-upon descrip-tions and prescriptions.

87 According to Jesness, the Immature-passive groupsplits, with about half of the group most similar tothe Asocial, passives in classification band one and theother half most similar to Immature conformists ofclassification band two.

Receiver. Some typologies do not differentiatebetween delinquent behavior which is imitated or"conformed to" from delinquent behavior whichgrows out of an internalized value system thus,it is difficult to know whether Reiss's Relativelyintegrated delinquent, Schrag's Antisocial type,Jenkins and Hewitt's Socialized delinquent, andchildren high on Quay's Subcultural factor belongpartially in this second classification band, orwhether all delinquents classified in these waysbelong in the fifth band, described below. Theoffender classified in this band is described as con-cerned with power, searching for structure, domi-nated by the need for social approval, conformingto external pressure, rule-oriented, unable toempathize, cognitively concrete, having low selfesteem, conventional and stereotyped in under-standing, oriented to short-term goals, havingsuperficial relationships with others, and self-representing as problem-free. This Conformistgroup has been subdivided further by some investi-gators into groups consisting of those individualswhose self perceptions are delinquent and whoconform primarily to a delinquent peer group andindividuals whose self perception is nondelinquentand who conform to the immediate power struc-ture, delinquent or nondelinquent. Investigatorswho have studied etiological factors for the Con-formists have found patterns of family helplessnessor indifference (rather than open rejection),inability to meet the dependency needs of thechildren, inconsistent structure and discipline,and absence of adequate adult models. Treatmentrecommendations for offenders in this classificationband include use of a clear, consistent externalstructure in which concern for the offender can beexpressed via controls of his behavior, use of grouptreatment to increase social perceptiveness, use ofpeer group as a pressure toward nondelinquency,and teaching of skills in order to help change self-definition in the direction of adequacy and inde-pendence.

A third clear-cut cross-classification band-theAntisocial-Manipulator-includes Jesness's Ma-nipulator, MacGregor's Autocrat, Makkay'sAntisocial Character Disorder-Organized (aggres-sive), Reiss's Defective Superego type, and War-ren's I Manipulator. To this classification bandcan tentatively be added: McCord's Aggressive(psychopathic) type, Reckless's Psychopath,

M W. McCoRD, W. McCoRD & I. K. ZOLA, OIGINSor CRm 195-98 (1959), presents evidence for theimportance of this distinction.

1971]


Schrag's Pseudosocial type, and Studt's Manipu-lator. The offender classified in this band is de-scribed as not having internalized conventionalnorms, guilt-free, self-satisfied, power-oriented,counteractive to the authority system, nontrust-ing, emotionally insulated, cynical, callous andextremely hostile. Those typologists who haveinvestigated etiological factors have found dis-trustful and angry families in which members areinvolved in competitive and mutually exploitivepatterns, parents who feel deprived and who ex-pect the children to meet their dependency needs,alternating parental patterns of overindulgenceand frustration of the children, and inconsistentparental patterns of affection and rejection. Ingeneral, investigators report a discouraging pic-ture as far as the treatment of this group of of-fenders is concerned. Treatment recommenda-tions take two distinct paths-one path beingthat of encouraging the Manipulator to develophis manipulative skills in a socially-acceptabledirection,8" and the other path being that ofattempting to allow the offender to work throughhis childhood trauma in a treatment relationshipwhich will revive his capacity to depend on and beconcerned about others." The first path makesthe assumption that it is possible to have a non-destructive, nondelinquent "psychopath," whichmany consider a contradiction in terms. Treatmentrecommendations toward the goal of socially ac-ceptable manipulation include increasing thesocial perceptiveness and ability to predict viagroup treatment, and increasing opportunities forlegitimate accomplishments via training in job,social, athletic, etc. skills. The second path clearlyinvolves a serious and possibly very long-termindividual treatment effort, and one which has noguarantees of success. The latter course is a diffi-cult one to fit into most social agency programs.

The fourth classification band-the NeuroticOffender-includes Hunt's Stage 11 group, Hur-witz's Type III, Jesness's Neurotic (acting-out,anxious, or depressed) types, MacGregor's Inti-midated youth, Makkay's Neurotic, Quay's chil-

8 D. C. GIBBONS, CHANGING THE LAWBREAxER(1965); Gibbons & Garrity, Some Suggestions for theDevelopment of Etiological and Treatment Theory inCriminology, 38 SocuL FoRcEs 51 (1959); Gibbons &Garrity, Definition and Analysis of Certain CriminalTypes, 53 J. CRm. L.C. & P.S. 27 (1962).

90 E. Makkay et al, Juvenile Delinquency Field Dem-onstration and Training Project: Newton-Baker Projectof the Judge Baker Guidance Center. Basic Design.Proposal to National Institute of Mental Health, 1961.

dren high on Neurotic-disturbed factor, Reiss'sRelatively Weak Ego type, and Warren's 14 Act-ing-out Neurotic and Anxious Neurotic types. Tothis classification band can tentatively be added:Argyle's Weak Ego-control type, Gibbon's joy-rider and Behavior Problem types, Jenkins andHewitt's Over-inhibited type, McCord's Neurotic-withdrawn, Reckless's Neurotic personality,Schrag's Prosocial type, and Studt's Love-seeker.As is indicated by the terms "intimidated," "dis-turbed," "overinhibited," "anxious," "depressed,"and "withdrawn," most investigators have identi-fied an offender type in which symptoms of malad-justment are dearly visible. Some investigatorshave identified a second subgroup of neuroticoffenders whose inner dynamics are quite similarto the visibly disturbed offender, but whose innerconflicts and anxieties are "acted-out" rather thanappearing as neurotic symptoms. In addition toJesness's and Warren's Acting-out Neurotic types,Gibbon's Joyrider and Studt's Love-seeker typesappear to be most like the second group of Neuroticoffenders. Investigators of etiological factors sug-gest that this type of offender is often the victimof parental anxiety or neurotic conflicts betweenthe parents, with the offense viewed as a masculineidentity striving. Some investigators have found afairly typical role-reversal phenomenon in whichthe child, at an early age, finds himself expected toplay a mature, responsible role with a child-likeparent. It has been suggested by some authorsthat neurotic delinquency is primarily a middleclass pattern. However, figures from the Com-munity Treatment Project show that, althoughmiddle class offenders make up a larger proportionof the Neurotic subtypes than of other subtypes,by far the largest proportion of the Neurotic sub-groups, as well as other subgroups, is lower class.0'Treatment recommendations for the Neuroticoffender focus on the resolution of the neuroticconflict through insight into family or individualdynamics which lead to the offense behavior. Suchconflict resolution is sought through family grouptherapy and/or by individual or group psycho-therapy for the offender.

The fifth classification band-the Subcultural-Identifier-includes Hunt's Stage II, Hurwitz'sType I, Jesness's Cultural delinquent, MacGregor'sRebel, Makkay's Subcultural type, Quay's chil-dren high on Subcultural factor, Reiss's Relatively

91 M. WA E N & T. PALMER, CommumNT TREAT-IENT PROJEcT, RESEARCH REPORT No. 6, 1965.

[Vol. 62


integrated delinquent, and Warren's 14 CulturalIdentifier. To this classification band can be tenta-tively added: Argyle's Deviant Identification type,Gibbon's Gang Offenders, Jenkins and Hewitt'sSocialized type, Schrag's Antisocial type, andStudt's Learner. The essential characteristic ofthis type of offender is that the individual, al-though developing "normally" 9 2 in most respects,has internalized the value system of a deviantsubculture. Thus violation behavior, for examplestealing from representatives of the larger culture,becomes simply an expression of what the Subcul-tural-Identifier considers "right." Investigatorsdescribe this offender type as interpersonally re-sponsive, psychosocially healthy, loyal to his ownprinciples and his own group, adequate, proud,suspicious of the authority system, capable ofidentifying himself with a mature socialized per-son, and accessible to new experiences. As wasnoted in the description of the second classifica-tion band, those investigators who focused onoffender behavior and delinquent attitudes have notdistinguished between the Subcultural-Identifierand the Subcultural-Conformist. At these levels ofobservation, the two groups appear similar: highlypeer group oriented, distrusting of the authoritysystem, comfortable with "delinquent" label,extensive delinquent histories, problems viewed as"external" rather than "internal," and apparentlyself-satisfied. In addition, both types include highproportions of minority group members. Strikingdifferences between the two groups appear whenthe foci of observation are family stability andconcern, individual capacity for self-knowledge andself-evaluation and differentiated perception ofothers, interpersonal relationship ability, goalorientation, concern with status, time perspective,etc. This series of characteristics becomes crucialto assessment of the individual's potential forbecoming a contributing citizen and for makingmanagement and treatment decisions. Two levelsof treatment appear to be recommended for theSubcultural-Identifier, one focused on stoppingthe violation behavior and one focused on changing

Some investigators have noted that this type ofoffender, while having satisfactory mother-child re-lationship, does not have a strong, authoritive, respectedfather with whom to identify. R. MacGregor, MiddleClass Delinquent Youth, a Study of Families. FinalReport (1965); E. MARKAy, DELrNQUENCY CONSIDEREDAS A MANIFESTATION OF: (1) A SEIuous DisoRDER oFDEVELoPmENT iN EAxRL CnI=HooD, A-m (2) OmERDELmQuENCY-ForNE DIsTUmnANcEs Or EMOTIONAD.EzvopmENT (unpublished manuscript, 1960).

the content of his value system. For the former,suggestions for stopping the violation behaviorinclude demonstrating to the offender through useof the "lock up" that "crime dces not pay",and teaching the individual how to meet status andmaterial needs in ways acceptable to the largerculture. The second level of treatment involvesworking through a relationship with a strong iden-tity model who is a representative of the largerculture and thus enlarging the offender's conceptof his in-group and broadening his self-definition.

The sixth classification band-the SituationalOffender-includes Hunt's Stage II, and War-ren's 14 Situational Emotional Reaction type. Tothese may be tentatively added Gibbon's CasualDelinquent and Reckless's Offender of the Mo-ment. Offenders in this group are represented asnormal individuals who give no evidence of long-term psychoneurosis or psychopathy and forwhom crime is ego-alien. These individuals havepresumably found themselves involved in viola-tion behavior as a result of accidental circum-stances or a specific, nonrecurring situation whichtaxed their normal coping capacities. Treatmentis either considered unnecessary or, if offered, isoriented toward helping the individual solve thespecific social or personal problem which led tolaw-breaking.

In summarizing the cross-tabulation chart, itappears that six classification bands can be ten-tatively identified as cutting across various ty-pologies. The minimum number of identified sub-types within any of the included typologies isthree. Of those systems which involve only a three-way breakdown of the offender population, thesingle agreed-upon subtype is the one referred toin this paper as Neurotic. Of the sixteen systemscharted, ten involve either a three-way or a four-way breakdown. Of these ten, the most typicalpattern includes counterparts of the followingsubtypes: Neurotic (10 out of 10), Subcultural-Identifier (8 out of 10), Asocial (7 out of 10) Con-formist (5 out of 10) and Antisocial-manipulator(5 out of 10). Classification systems which involvemore than a five-way breakdown of the offenderpopulation add the Situational type and/or sub-divide the Asocial, the Conformist, and the Neu-rotic categories. Warren's typology involves thelargest number of subgroups, defining-in addi-tion to the Antisocial-Manipulator, the Subcul-tural-Identifier and the Situational-two kinds ofAsocial types (aggressive and passive), two kinds


of Conformist types (delinquency-oriented andnondelinquency-oriented), and two kinds of Neu-rotic types (anxious and acting-out), for a total ofnine subtypes.

It should be noted that most of the typologiesare based on studies of juvenile boys. Only Hunt,Schrag, and Warren have specifically includedgirls or women, but these investigators have foundtheir typologies to be equally appropriate for thefemale population. Schrag's typology is basedprimarily on adult offenders (although institution-alized juveniles have been classified by some ofSchrag's followers), and the original form of War-ren's typology (Interpersonal Maturity Levels,without subtypes) was found to be as appropriatefor an adult as a juvenile population. It is an as-sumption, albeit justified, that the six-band cross-classification system is an adequate way of sub-dividing female juvenile offenders as well as adultoffenders.

One measure of the appropriateness of cross-tabulation of subtypes from various classificationsystems might be the degree of similarity betweenthe proportions of offenders placed in each of thevarious classification bands. Many of the typolo-gies do not report these data. Even for those whodo, the major differences in the nature of the popu-lations studied are so great as to make comparisonsof questionable meaning. Table I presents theestimated proportions in the six classificationbands, using data from five studies of juvenileoffenders. The Jesness data are based on a study ofyoung boys (ages 8 to 14) committed to a statetraining school. The Community TreatmentProject (CTP) data are based on boys and girls

TABLE IEsTriAn D P~oPopRnoNs or DELIQuyEn IN VAnious

SUBTYPES

Jesness CTP PTS Hurwitz ReissSubtype Data N Data N Data N Data N Data N

=210 =400 = 371 - 198 = 511

Asocial 18% 10% 10% 34%Conformist 30 28 40Antisocial- 12 15 15 12%

ManipulatorNeurotic 35 40 33 21 22Subcultural- 14 5 2 45 12

IdentifierSituational 3 1

Total 100% 100% 100% 100% 46%*

* 54% of Reiss's subjects were not classified.

(ages 9 to 18) committed to the State YouthAuthority from juvenile courts and declared eli-gible for participation in an intensive communityprogram. The Preston Typology Study (PTS)data are based on older adolescent boys committedto a state training school. All three of these studygroups contain a population of serious or habitualdelinquents. The Hurwitz data are based on casesappearing before the juvenile court. The Reissdata are based on 46% of a juvenile probationpopulation, the 46% identified as those proba-tioners who were examined by court psychiatrists.These last two study groups may be generallyassumed to include less serious delinquents thanthose in the first three groups.

The higher proportion of the Conformist type inthe PTS data than in the Jesness data and the CTPdata probably reflects the large number of recidi-vists in the PTS population. Warren and Palmer3

have shown a high failure rate for Conformists(compared with most other subtypes) followingtraditional correctional programs. The Hurwitzdata in Table I indicate that Hurwitz's Type II(34%) and Type I (45%) probably contain of-fenders which should more accurately be cross-tabulated in other classification bands. It is likelythat some individuals in Type EI could be classi-fied as Antisocial-Manipulators, and that someindividuals in Type I could be classified as Con-formists. Another possibility is that the Subcul-tural-Identifier group represents a larger propor-tion of a court-appearance population than it doesof the more serious habitual delinquent populationcommitted to a State program. This possibilityis in line with Reiss's assumption that a largenumber of the 54% of the probationers in his study,who were not classified by court psychiatrists,belong in his Relatively Integrated subtype (in-cluded here in the Subcultural-Identifier group).

Based on descriptive data, a cross-classificationof several important offender typologies is ap-parently possible. In the present state of the sci-ence of corrections, this much consistency in thedata of various studies is a most encouraging find-ing, leading us to feel that the identifiable subtypesof offenders reflect at least a partial "truth" aboutthe population rather than simply a convenientfantasy in the mind of the criminologist. The factthat a cross-classification is possible is even moreimpressive when one considers the varieties ofmethods of deriving the subtypes-theoretical

" Warren & Palmer, supra note 91.

[Vol. 62


formulations, empirical-observational methods,multivariate analysis procedures. Additionally, itis important to note that not only is it possible tofind similarities in the descriptions of offendercharacteristics across typologies, but also that con-sistency is evident in descriptions of etiologicaland background factors and in treatment prescrip-tions for seemingly similar subtypes.

Having said that typologies are apparentlyoperating on a common ground, it is necessary toadd that much crucial information is missing whichwould be necessary in order to determine whetheror not any two subtypes are exact counterparts.The ultimate test of such a cross-cl~ssificationwould come from a study in which a typing ofindividuals in a single population was conductedby experts in the use of each of the various classifi-cation systems. Such a study would not only clarifythe extent to which the subtypes in one systemare actual counterparts of those in another system,but also lead typologists to increase the precisionof their subtype definitions.

Until the matter of classification of offenders ishandled in some generally agreed-upon way, it isalmost impossible to compare treatment programsbeing conducted in various parts of the country. If,from the cross-classification study suggested above,a group of the leading typologists could agree on acommon taxonomy, the path would be open for agreat number of significant studies. The nextimportant step would be the determination of themost efficient diagnostic methods. Once the cate-gories had been agreed upon, a number of scien-tists in various parts of the country could work onthe problem simultaneously. Additionally, inter-related studies of management and treatmentmethods could be conducted-trying a variety ofwell-defined treatment approaches to the samecategory of offender. It would then be realisticto attempt the replication of experimental ap-proaches, suggested by Keith Griffiths in hisCorrectional Research Model.

There is evidence that both at the theoretician94

and practitioner 95 levels, the field is ready to

It In the July, 1966 issue of Cnru AN DELIN-QuENcY, vol. 12, no. 3, Glaser, The New CorrectionalEra-Implicatioos for Manpower and Training; Gilman,Problems and Progress in Staff Training; and Nelson,Strategies for Action in Meeting Correctional Manpowerand Program Needs, all point to the importance ofdeveloping treatment-relevant classification systemsand differential treatment methods.

"5 Demands for training in differential treatmentmethods come to the California Youth Authority's

move toward treatment programs which are basedon categorizing the range of problems representedin the correctional population. Not only is there aready ear for such conceptualizing, but it alsoappears that a time of concensus among typologistsmay be approaching in which a rational, correc-tional treatment model may be begun.

To date, little work has been done toward utiliz-ing typologies for building differential treatmentstrategies. The work which has been done has oc-curred largely in small experimental programs. Itis right and proper for experimental programs tobe in the lead and for the rest of the field to beeyeing their exploratory work with hope. But thesize of the gap between these programs and thegenerally undeveloped state of correctional prac-tice is crucial in estimating what programmaticutility the typological concensus has in the fore-seeable future. Are the classification concepts orthe corollary program prescriptions so esotericthat only academicians can understand them?Are the treatment methods which might arise froma rational correctional model such that the aver-age practitioner could not apply them?

While the typologies reviewed here vary con-siderably in the complexity of their derivation,the essence of the correctional model which fol-lows from a treatment-relevant typology is a rathersimple idea. The idea is this: The goals of correc-tional treatment with any offender should relatein some direct manner to the causes or meaning ofthe law violation, and the treatment methods shouldrelate specifically to the goals. This idea, when putforth with examples, makes the greatest kind ofsense to the practitioner who is supposed to "dosomething" about delinquent behavior.

If the idea is simple, what about its implementa-tion? Assuming an agreed-upon taxonomy, whatabout the methods of individual diagnosis? Inorder to move easily beyond small experimentalprograms into large operating programs, it isessential that the classification be done via easy-to-administer-and-score-measures or via an al-ready-established clinical process in the correc-tional agency. Most correctional programs nowhave a time and place set aside for intake andclassification procedures, so that the machinery fortyping offenders may be well available. As formethods of obtaining the diagnosis, work toward

Community Treatment Project from correctionalagencies, large and small, both within and outside,California.

1971]


simpler procedures should continue after an agreed-upon typology is available. Of the sixteen typolo-gists represented in the cross-classification chart,several do not specify diagnostic methods, sincepresumably the major concern in the developmentof the typology was not the classification of in-dividual offenders (Argyle, Gibbons, McCord,Reckless). The Jenkins and Hewitt, Jesness, andHurwitz typologies grew out of factor analyticprocedures, utilizing many tests and clinical judge-ments and thus do not lend themselves to indi-vidual diagnosis. The Reiss typology was based onpsychiatric judgements. The Studt types werederived from a series of intensive interviews withoffenders and with others who knew the inmateswell. Although the diagnosis of Antisocial Charac-ter Disorder is well spelled out by Makkay, dif-ferentiations of subtypes within that category arebased on a fairly lengthy observation period, withthe criteria not well defined as yet. The MacGregordiagnosis is based on a series of interviews withthe entire family of the delinquent. Thus, at thispoint in time, none of the above-listed classificationsystems represents a practical method for thediagnosis of large correctional populations.

In applying the Warren typology, the primaryinstrument for diagnosing individuals is a tape-recorded interview with the delinquent subject. Adisadvantage of this procedure is the trainingrequired to achieve rater reliability. The Warrensystem currently has some advantages over theothers in that hundreds of delinquents have beeninterviewed sequentially over time. Both highinterrater agreement 6 and high reliability overtimes have been shown. In addition, sets of specificcharacteristics items to be rated have been de-veloped for each delinquent subtype.98

The classification systems having the simplestdiagnostic methods are those of Hunt, Quay, andSchrag. Hunt's methods involve a simple T/Finstrument and a rating made from a set of sub-ject-completed sentences. Although the discrimina-

16 Reliability estimates for all subtypes based uponthe independent judgments of two different raters(trained research personnel) made at approximatelythe same point in time have fallen, on the average, inthe mid-80's.

17 Reliability estimates for diagnosis at intake ascompared with followup diagnosis (three to six monthslater for Experimental cases and eight to twelve monthslater for Control cases) have centered in the mid-80'sand low 90's.

8 Warren, Interpersonal Maturity Level Classifration(juvenile): DiAGNosis Am TREATmExT or Low,MmDxLE AND HIGH MATuRITY DELINQUENTS, CTPPUB IcATION (1966).

tions made in Hunt's Conceptual Levels systemare dearly treatment-relevant, further work instudying offenders with the typology is needed todetermine whether or not the three-way classifi-cation of the delinquent population is sufficient forprescribing treatment.

As noted earlier, Quay does not view his work asleading to types of individuals, but rather to aclassification of behavior dimensions. An individualis represented by a profile of behavior dimensionscores. Those individuals who have similar profilesmay presumably be grouped together in terms oftreatment need. The diagnostic instruments de-veloped by Quay are easy to administer and score,involving check lists and ratings of the individual'sbehavioral characteristics. The measurements canbe shown to have adequate reliability. The diffi-culty with using profiles is that, since few indi-viduals have a simple profile-i.e., a high score onone factor and low scores on all other factorse-askilled judgement must be made with regard togrouping for intervention purposes in the majorityof cases.

Schrag's typology has been used primarily tostudy subcultures within the prison walls, andSchrag has not wished to claim more generalapplicability for it in the absence of research data.Within the institutional setting, 50% to 70% ofindividuals can be typed easily using questionnaireand interview data. The remaining individualsare identified as mixed types. Since the types de-scribed by Schrag compare closely with types de-scribed by others, it is very likely that the typologyhas more general applicability than Schrag hasclaimed.

An optimistic note may be made with regard toour present ability to diagnose meaningful sub-types with realistically simple procedures. In astudy previously mentioned (the Preston TypologyStudy), Jesness classified the intake population of alarge California training school for boys, using theWarren typology. Diagnostic procedures includethe Jesness Inventory (consisting of 155 T/Fitems, scored for delinquent subtype using a dis-criminant analysis formula), a sentence completionand a short interview. The final diagnosis is madeusing all three instruments, with the hope thateventually the Inventory alone may be scored toproduce an accurate diagnosis. Using all ofWarren's nine subtypes, the diagnoses on 500subjects from the Inventory alone agrees 62% ofthe time with the final diagnosis. If, instead of

[Vol 62


using the nine subtypes, the three larger categories(Interpersonal Maturity Levels) are used, theagreement is 83%. Within the nine subtypes, someof the groups are identified by the Inventory alonemuch more accurately than other groups. Forexample, the Neurotic, Anxious subtype is diag-nosed accurately from the Inventory alone 84%of the time, and is diagnosed accurately 94% ofthe time as falling within Maturity Level 4. Theaccuracy with other subtypes is lower. It is pos-sible that if the Inventory cannot achieve an ac-ceptable level of accuracy for all subtypes it mayat least identify that proportion of the populationwhich needs further diagnostic instruments ap-plied.

An important point to be made with regard totreatment prescriptions which follow from offendertypologies is that the intervention strategies are notby and large made up of new and unusual treat-ment methods, but rather consist of many of theold alternatives differentially applied to the vari-ous categories of offenders. In this sense a typologywhich leads to differential prescriptions leaves thefield no worse off in terms of the need for skilledtreaters. In another sense, the field is far betteroff. If offenders can be classified by differentialtreatment need, correctional staff can then beassigned differentially. In this way a particularcorrectional line worker need not have the entirerange of specific management and therapeuticskills at his fingertips. Instead, his training canprepare him to handle only those treatment andmanagement methods appropriate for certaintypes of offenders. Further, since correctionalworkers can be characterized as having certain"natural" treatment stances, a matching of workerstyle and offender problem can be accomplished.9

If the field were to move toward a correctionalmodel utilizing differential management and treat-ment of various subtypes of offenders, how wouldthe training of correctional workers be affected?Since a differential model calls for training staffwho work with some types of offenders to utilizedifferent methods than those working with othertypes of offenders, the job for the trainers becomessomewhat more complex; however, the job of the

"Investigations into these "natural" stances arebeing conducted at the Community Treatment Project,see Palmer, Personality Characteristics and ProfessionalOrientations of Five Groups of Community TreatmentProject Workers: A Preliminary Report on DifferencesAmong Treaters, CTP REPORT SER Es, No. 1, CAI~r-BonEI YOUTH AToTHoBI (1967); C.F. Jesness, TamPiRsToN TYPoLoGY STruY (1970).

trainees is considerably simplified, since the workermust no longer learn how to handle the entirerange of problems. Under these conditions, trainingcontent can become less vague, less general, andless oriented toward producing that nebulousentity-the "good correctional worker." Instead,the training content can be specific to characteris-tics of particular types of offenders and preciselyrelevant to the management and treatment de-mands of the offender type. Because of the limi-tation in the range of content that a particular cor-rectional worker needs to learn in order to dealeffectively with his assigned offender population,it is likely that whatever training time is now avail-able in various correctional agencies could be moreeffectively used. This does not imply, of course,that all is now known about how to turn variouskinds of offenders into non-offenders. It does notimply that the need for imaginative and creativeapproaches to the problem is gone. It does imply,however, that treatment and management pro-grams, if based on an offender typology can be-come more rational by better defining the differen-tial problems leading to offense behavior, by pre-scribing differential goals for the correctional effort,and by training workers within the differentialframework.

A case can be made for the importance of utiliz-ing an offender classification system at each stepalong the entire correctional continuum. The advan-tages of using explicit, rather than implicit, classifi-cation systems at each correctional decision pointhas already been made in this paper. To the extentthat the correctional system is free to make deci-sions based, not on retributive justice, but ratheron a goal of turning offenders into nonoffenders,i.e., offender need-to that extent it is importantto have available at each correctional decision pointclassification information which will indicate thesetting and methods most likely to achieve theoverall goal. For example, what is the treatment ofchoice when an individual identified in the cross-classification chart as a Conformist first appearsin the correctional system? Some data are availablefrom the Community Treatment Project whichindicate that such individuals (1) become increas-ingly oriented toward delinquency in the highlydelinquent peer group atmosphere of an institu-tion, and (2) can be satisfactorily managed andtreated in certain kinds of community programs10 0

100 In CTP, the failure rate for Conformists with 24months of community exposure time was only 33.3%for delinquents treated in an intensive community

1971]


Beyond the possibility of sorting out at eachdecision point those individuals who need to moveon through the correctional system, there is afurther advantage in making the differential diag-nosis as early as possible in the correctional career.A typology with its consequent goal specificationallows for a unification among the treatment ef-forts of various segments of the correctional proc-ess. At many points in present correctional prac-tice, it is possible to observe the total irrelevance ofthe goals and methods of treatment in an institu-tional setting to the goals and methods of treat-ment in the after-care program. The goals in thetwo settings may even be at odds with oneanother-the aim of the institutional time being toachieve conformity to a strict control system andthe aim of the parole time being to achieve indi-vidual self-responsibility. While it is true thatthese two aims follow somewhat naturally from thecharacteristics of the two settings, it is possible toaim for conformity in a community setting and toaim for individual self-responsibility in an institu-tional setting--should the nature of the problemwith particular offenders require one approach orthe other. Even assuming that there are institu-tional administration needs to consider and com-munity safety needs to consider, it seems possiblethat the determination of a treatment-relevantdiagnosis early in an individual's correctionalcareer might well contribute to a more consistentand therefore more effective total interventionprogram.

SUMMARY' 0

A rationale for classifying the offender popula-tion into meaningful subgroups was presented.

program compared with 72.7% for comparable indi-Iduals following a period of incarceration.1011n addition to the references already cited, the

following sources were consulted in the preparation ofthis paper: Hayner, Characteristics of Five OffenderTypes, 9 AlABAMA ComRcRTIoNAL J. 75 (1962); Linde-smith and Dunham, Some Principles of Criminal Typol-ogy, 19 SociAL FORCEs 309 (1941); Loveland, TheClassification Program in the Federal Prison System:1934-60, 24 FEDERAL PROBATION 8 (1960); Peters,Treatment Needs of Juvenile Offenders, 1 CALIF. ST. BD.oF CoRacnoNs MONOGRAPH 22 (1960); Topping,Case Studies of Aggressive Delinquents, 11 Am. J. oFORTroPsYvCMARX 485 (1941); Vedder, Theory ofCriminal Types, 9 ALA. CoR crnoNAL J. 1 (1962).

Various classification approaches were describedand their implications for efficient managementpractices and effective treatment strategies wereillustrated with a number of clinical and researchstudies. A cross-tabulation of sixteen typologicalsystems was presented and six cross-classificationbands were identified. The six bands or offendersubtypes were entitled: Asocial, Conformist, Anti-social-manipulator, Neurotic, Subcultural-Identi-fier and Situational offender. It was pointed outthat the consistency in the data of several ty-pological studies which made the cross-classifi-cation possible is an encouraging sign. However,the importance of taking the next step--an actualcross-classification of offenders from a singlepopulation, using the various typological schemes-was noted. It was further suggested that if acommon taxonomy could be agreed upon, the waywould be open for conducting and replicatingnumerous interrelated studies of management andtreatment methods.

In asking whether a typological concensus hasany programmatic utility at the present time, cur-rent interest among practitioners in developingdifferential treatment strategies for various typesof offenders was noted. It was suggested that itmay be possible in the near future to make differ-ential diagnoses of large populations, and tosimplify the training of correctional workers byteaching management and treatment specialtiesrather than the entire range of correctional tech-niques. The use of differential diagnosis in decision-making along the correctional continuum and itspotential value as a treatment-unifying influencewas discussed.

Typologies of offenders represent an importantmethod of integrating the increasing body ofknowledge in the field of corrections. Ultimately,typological approaches will flourish or not depend-ing on their fruitfulness in producing improvedmanagement and treatment methods for the prac-titioner working in this discouraging field. At themoment, the classification studies reported in thispaper appear to represent solid steps in the de-velopment of a systematic science of corrections.

[VOL 62

classification of offenders as an aid to efficient

Documents