juvenile recidivism report · adult criminal activity, society may first have to address juvenile...

16
Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania

Upload: others

Post on 03-Jul-2020

6 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Measuring and Analyzing Juvenile Recidivism

in Rural and Urban Pennsylvania

Page 2: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal
Page 3: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Measuring and Analyzing Juvenile Recidivismin Rural and Urban Pennsylvania

David E. Kalist, Ph.D., and Daniel Y. Lee, Ph.D.Department of Economics

Shippensburg University of Pennsylvania

January 2009

This project was sponsored by a grant from the Center for Rural Pennsylvania, a legislative agency ofthe Pennsylvania General Assembly.

The Center for Rural Pennsylvania is a bipartisan, bicameral legislative agency that serves as a resourcefor rural policy within the Pennsylvania General Assembly. It was created in 1987 under Act 16, the RuralRevitalization Act, to promote and sustain the vitality of Pennsylvania’s rural and small communities.

Information contained in this report does not necessarily reflect the views of individual board membersor the Center for Rural Pennsylvania.

For more information, contact the Center for Rural Pennsylvania, 200 North Third St., Suite 600,Harrisburg, PA 17101, telephone (717) 787-9555, email: [email protected].

Page 4: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Table of Contents

This research provides a comprehensive analysis of juvenilerecidivism in Pennsylvania, with special reference to differencesbetween rural and urban counties. Pennsylvania is one of the fewstates that do not report on juvenile recidivism. Recidivism isdefined as a relapse into criminal activity and is one of the mostimportant measures used to gauge the success of juvenile justiceintervention programs. The study analyzed data from 1997 to2005 on approximately 190,000 juveniles with a prior convictionto determine the factors leading to a reconviction.

According to the research analysis, recidivism rates were lowestin rural counties and highest in urban counties. The lower ruralrecidivism rates were especially pronounced for blacks, males, andthose with a prior felony. Other estimates indicated that the lengthof time to relapse into crime was from 3 percent to 30 percentlonger in rural counties.

The research identified juvenile characteristics that deservecloser scrutiny to help reduce recidivism. The characteristicsassociated with increased recidivism are: living in an urbancounty, being male or Hispanic, living with a single mother,having at least one deceased parent, committing a prior felony,and attending alternative education. According to the research,special attention should be given to juveniles living with singlemothers or those with at least one deceased parent. More frequentcontact with probation officers and closer court supervision maybe beneficial for juveniles living in single-parent families as ameans of reducing recidivism.

Another interesting result is that an increase in the number ofpolice per capita is associated with a slight decrease in recidivism.The socio-economic status of the juvenile’s county also affectedrecidivism. Juveniles living in counties with a higher per capitaincome had modestly improved recidivism rates. Lower socio-economic counties, as measured by Temporary Assistance for NeedyFamilies per capita, had higher rates of recidivism.

This research demonstrated the feasibility, importance, andongoing need to study juvenile recidivism in Pennsylvania. Manystates already report on recidivism to better understand whichintervention programs are successful and which need to bechanged.

The researchers recommend that Pennsylvania begin to docu-ment juvenile recidivism and regularly issue formal reports. Thesereports could begin to document recidivism rates from juvenilesreleased from specific placement facilities, such as group homes,secure detention facilities, drug and alcohol treatment programs,and Outward Bound-type programs.

The researchers also suggest linking the criminal records ofadults with their juvenile records. By doing so, future researchcould help to determine what sorts of juvenile records are mostlikely to be followed by serious criminal offenses when thejuvenile reaches adulthood, and what sorts of juvenile programs ordispositions are most effective in reducing the likelihood of futurecrime.

Executive Summary

Introduction ................................ 5

Goals and Objectives ................. 6

Methodology .............................. 6

Results ......................................... 8

Conclusions .............................. 12

Policy Considerations .............. 13

References ................................ 14

Page 5: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania 5 The Center for Rural Pennsylvania

IntroductionJuvenile crime remains a serious problem in the

United States and continues to affect millions ofpeople despite a downward trend in recent years. In2003, more than 2.2 million juveniles (under age 18)in the U.S. were arrested for various crimes, including92,300 arrested for violent crimes, such as murder,non-negligent manslaughter, forcible rape, robbery,and aggravated assault (Snyder and Sickmund, 2006).According to Paternoster, Brame, and Farrington(2001), documentation shows a strong associationbetween involvement in adolescent delinquency andinvolvement in adult criminality. In fact, many hard-ened criminals and serial offenders began their crimi-nal careers as juveniles (Farrington, 1992). To alleviateadult criminal activity, society may first have toaddress juvenile crime.

This research, conducted in 2007, estimated theextent of juvenile criminal recidivism in Pennsylvania,examining how and whether various characteristics ofboth the juvenile justice system and juvenile offenderaffect recidivism. Recidivism is defined as a relapse orreturn to criminal activity by juvenile offenders.

Pennsylvania provided a unique setting to studyjuvenile recidivism given that it is one of the largestrural states in the U.S. Approximately 28 percent ofthe state’s 12.4 million people live in a rural county(defined as less than 274 people per square mile).

In Pennsylvania, unlike other states, little is knownabout juvenile criminal recidivism. Pennsylvania,Alabama, Indiana, Kansas, and Tennessee are the onlystates that do not track juvenile recidivism (VirginiaDepartment of Juvenile Justice, 2005). This paucity ofinformation is unfortunate since recidivism is animportant measure of the effectiveness of the juvenilejustice system, capturing the system’s ability to detercriminal relapse and rehabilitate juveniles. Of course,recidivism is not only affected by the juvenile justicesystem but depends on the decisions and choices madeby juveniles, who are likely influenced by socioeco-nomic and sociodemographic factors.

The direct costs of recidivism are the psychologicaland monetary costs to victims, including increasedcosts of the juvenile justice system. A relapse intocrime may have other costs, too, such as reducing thejuvenile’s educational attainment and labor marketopportunities. In fact, research suggests that it mightbe possible to increase educational attainment, even of

non-juvenile offenders, by reducing juvenile recidivism(Grogger 1997), since reductions in violence at schooland in the surrounding area make it easier for students toconcentrate on their studies, while making it less likelythey will miss school because of fear of violence.

As previously mentioned, Pennsylvania does notreport data on juvenile recidivism, although thePennsylvania Department of Corrections has beenpublishing Recidivism in Pennsylvania State Correc-tional Institutions since 1999, which providies recidi-vism statistics of adult offenders released fromPennsylvania’s correctional institutions. However,many states publish reports on the recidivism ofjuvenile offenders as well as adult offenders, and usethem as outcome measures of juvenile justice programsand initiatives. For example, the Maryland Departmentof Juvenile Services publishes an annual report thatincludes juvenile recidivism rates for youths who werereleased from different programs, such as securedetention and substance abuse programs. These dataare used for planning and assessing prevention efforts.

There are no national statistics on juvenile recidi-vism, although the U.S. Department of Justice, inJuvenile Offenders and Victims: 2006 National Report,presents average rates of recidivism for a select groupof states. Juveniles who were released from incarcera-tion have a one-year recidivism rate of 55 percent. Inother words, 55 percent of juveniles are rearrestedwithin one year of their release.

The data used in this study were from a largedatabase of juvenile activity maintained by the Centerfor Juvenile Justice Training and Research in Pennsyl-vania (CJJTR).1 The micro-level data were rich anddetailed in terms of sociodemographic variables. Overthe analysis period of 1997 to 2005, the final datasample used in this study contained about 190,000observations from both rural and urban counties.

This study explored juvenile recidivism, defined asthe time between the juvenile’s first and secondreferral to the juvenile justice system. A referral occurswhen an arresting officer informs a county’s JuvenileProbation Department of the charges against a juve-nile. The first and second referral must have led to asubstantiated charge; therefore, this study exploredrecidivism on the basis of reconviction, one of themost commonly used measures of recidivism (Mbuba,2005).

1 Founded in 1982, CJJTR is housed at Shippensburg University, but it is not part of the university. It is a division of the JuvenileCourt Judges’ Commission (JCJC), established by the Pennsylvania General Assembly in 1959, with the responsibility of overseeingthe juvenile justice system. The CJJTR provides professional training for juvenile justice practitioners, such as probation officers andcourt judges, and collects and disseminates data on juvenile delinquency.

Page 6: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

The Center for Rural Pennsylvania 6 Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania

Goals and ObjectivesThe goal of this research was to understand one

important aspect of juvenile crime, namely, juvenilerecidivism in rural Pennsylvania.

The two major objectives of the research were tocalculate basic descriptive statistics on juvenile recidi-vism for Pennsylvania, the most fundamental of whichwas the recidivism rate or the percentage of juvenileswho return to crime within a specified time period;and to examine how sociodemographic variables of the

juvenile, juvenile’s family, and county of residenceaffected recidivism. For example, this study comparedrecidivism rates by sex, race, whether the juvenilelived in a single- or two-parent household, andwhether the juvenile committed a felony or non-felonycrime.

The achievement of these objectives may make itpossible to identify juveniles who are at increased riskof recidivism.

MethodologyThe researchers used data from CJJTR. CJJTR is part

of the Juvenile Court Judges’ Commission (JCJC), andwas created in 1959 by the Pennsylvania GeneralAssembly. JCJC is responsible for advising juvenilecourts on court procedures and care of delinquents,overseeing administrative practices in probationoffices, and collecting and reporting juvenile courtstatistics, such as the Pennsylvania Juvenile CourtDispositions, published annually.

This study used a large database of juvenile activityfor Pennsylvania collected and maintained by CJJTR.It included demographic variables, charges, substanti-ated charges, number of counts, disposition outcomes,and criminal or civil sanctions, such as fines, consentdecrees, community service, probation, informaladjustment, and type of placement of all juveniles inPennsylvania who were referred to their county’sjuvenile justice system. Each record or observationprovided information on an individual juvenile’sdisposition. The period of analysis was 1997 to 2005.CJJTR reported that, prior to 1997, the data wereneither reliable nor consistent with more recent data,owing to changes in how the variables were collectedand reported. For each year, CJJTR collects andcompiles data, which are initially entered by thejuvenile court in each of Pennsylvania’s 67 counties,on each juvenile referred to the juvenile court systemof Pennsylvania. A juvenile may be referred to thecourt by the police, a school, a probation officer, arelative, a social agency, another juvenile court, or adistrict justice. For every juvenile record, there wasinformation on charges filed and whether the chargeswere substantiated. In addition, the juvenile’s outcomeat disposition was reported, which may includedismissal of the case, fines, probation, informaladjustment (a juvenile who was not adjudicated adelinquent may nonetheless be required to visit aprobation officer periodically and perform other

duties), consent decree (a juvenile guilty of an act mayhave his case expunged if he fulfills several courtrequirements, such as victim restitution), warning andcounseling, and placement, such as secure detention,drug and alcohol treatment program, and group home.Along with these variables, information was alsoavailable on the juvenile’s race, age, ethnicity (His-panic or non-Hispanic), county of residence, familystatus (whether the juvenile’s parents were currentlymarried, divorced, separated, never married, or one orboth deceased), and living arrangements (whether thejuvenile was living with both parents, mother, father,relative, father and stepmother, mother and stepfather,or foster parents). The researchers used the county ofresidence information to determine whether a juvenilewas living in a rural or urban county. A rural county,as defined by the Center for Rural Pennsylvania, hasfewer than 274 persons per square mile. Furthermore,the county information allowed the researchers to addadditional variables to the analysis, such as the numberof police officers per capita, real (inflation adjusted)income per capita, and other sociodemographicvariables in the surrounding community that mayaffect recidivism.

A recidivism study requires longitudinal data, whichincludes a juvenile’s record of encounters with thejuvenile justice system over a period of time. In itspresent form, the CJJTR data set was not suitable forstudying recidivism because it did not contain acomplete set of unique IDs that would allow for thecreation of a juvenile criminal history; however, therewas information in the data to help create unique IDs.For 33 counties, the data contained unique IDs forjuveniles, making it rather straightforward to compiletheir histories. To create juvenile crime histories, thejuvenile IDs were matched within each county over theanalysis period. Unfortunately, some of these countieschanged their ID system over the years, making it

Page 7: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania 7 The Center for Rural Pennsylvania

difficult to collect the entirecriminal history of a juvenile. Forexample, Columbia and Greenecounties started using a new IDsystem in 2004, so there was noway to determine whether juvenileswho had contact with the county’sjuvenile justice prior to 2004 werereferred in 2004 or later.

The juvenile IDs for the 34remaining counties were created bymatching within a county thejuveniles’ first and last names alongwith date of birth. One potentialproblem with this matching tech-nique was that the names of juve-niles were not always consistentlyentered by the county’s juvenilejustice system. For example, ajuvenile named Joseph Smith maysometimes be entered into thesystem as Joe Smith, Joseph E.Smith, or Joey Smith Jr. Theresearchers attempted to minimizethese occurrences by manuallycorrecting the obvious cases.Furthermore, if a juvenile wasreferred to several different countyjuvenile justice systems, there wasno effective way to construct thejuvenile’s complete criminal historybecause each county used its ownID system that did not follow ajuvenile across county lines. Thesame problem will arise if juvenileswere to move across counties. Afterdeletion of observations withmissing variables and juvenileswhose charges were not substanti-ated, the final dataset containedapproximately 187,000 observa-tions.

Given these data limitations, inaddition to not having criminalhistories prior to 1997, the re-searchers could not determinewhether the juvenile’s first substan-tiated charge listed in the datasetwas indeed the juvenile’s actualfirst offense. However, the fre-quency of multiple offenders in thedataset indicated that there were

relatively few juveniles whocommitted three or more offenses.Therefore, the researchers expectedthat the majority of cases examinedwere for the juveniles’ first andsecond referrals.

There were other shortcomings ofthe data. First, there was noinformation on whether juvenilesliving in Pennsylvania have com-mitted crimes in neighboring states,so the data might not have capturedthe complete criminal history of ajuvenile. Second, the data did notinclude information on juvenileswho were convicted of a delinquentact in Pennsylvania and then movedout of state. As long as juvenilesmoving out of state were notsubstantially more or less likelythan nonmovers to commit asubsequent offense, the estimatedresults will be unbiased. Takentogether, all the shortcomings andweaknesses of the data will have theeffect of understating the extent ofrecidivism.

Pennsylvania’s Act 33 of 1996created more conditions in which ajuvenile under age 18 could beexcluded from juvenile courtjurisdiction. Prior to Act 33, onlyjuveniles charged with murder wereexcluded from the juvenile court’sjurisdiction. Act 33 amended thelaw by expanding the transfermechanisms. Under this act,juveniles could now be excluded ifthey were aged 15 or older andcharged with “rape; involuntarydeviate sexual intercourse; aggra-vated assault; robbery; robbery of amotor vehicle; aggravated indecentassault; kidnapping; voluntarymanslaughter; or an attempt,conspiracy, or solicitation tocommit murder or any of thecrimes listed.”2 In addition, thejuvenile must either have used adeadly weapon or have previouslybeen adjudicated to have committedone of the excluded offenses. Even

if there is a prima facie case (a casein which, on first impression, theevidence appears strong) and theexclusion criteria are met, it ispossible that some cases may returnto juvenile court following adecertification hearing in criminalcourt whereby a juvenile files arequest that the case be tried injuvenile court. As a result of theforegoing exclusions, the datasetused in this study did not containinformation on juveniles whocommitted some of the most violentcrimes, and this will lead to anunderreporting of recidivism.However, this omission was aminor problem since the number ofjuveniles excluded from juvenilecourt was small. In 1996, only 473cases were excluded from juvenilejurisdiction, leading to 109 convic-tions in criminal court.

The researchers used threedifferent survival techniques toexamine recidivism: Kaplan Meier,parametric survival models, andCox regression. These modelsestimated how recidivism wasaffected by the juvenile’ssociodemographic factors, and thepunitiveness of the juvenile justicesystem. Moreover, the results werecompared to juvenile recidivismrates in urban areas of Pennsylva-nia. Special attention was given tohow the juvenile’s family andliving arrangement affected recidi-vism. In addition, the researchersexplored whether characteristics,such as income, Temporary Assis-tance to Needy Families, thenumber of police, and the arrestrates of the juvenile’s county ofresidence, were related to recidi-vism.

2 Office of Justice Programs, U.S.Department of Justice available at: http://www.ncjrs.gov/html/ojjdp/summary/08_2000/jtcc_10.html.

Page 8: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

The Center for Rural Pennsylvania 8 Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania

A First Look at Juvenile Recidivism inPennsylvania

Table 1 provides summary statistics for the numberof juvenile offenders, number of reoffenders, andpercent of reoffenders by county and by rural andurban areas. From 1997 to 2005, there were 187,173juveniles who were charged with substantiated of-fenses. Of these juveniles, 35,337, or about 19 percent,were later charged with at least one more substantiatedoffense during the period. The majority of offenderscame from urban counties. There were 140,144offenders in urban counties compared to 47,029 inrural counties, with the percent of reoffense higher inurban counties than rural counties (20 percent vs. 16percent, respectively). The percent of juveniles whoreoffend varied greatly among the counties, rangingfrom 1 percent in Adams to 36 percent in Fayette. Byway of comparison, Philadelphia’s rate was 25 percent.Again, it is important to note that the data in Table 1did not include those juveniles who were convicted incriminal court, which will likely result in anunderreporting of the percent of reoffenders. Further-more, since violent crime is more prevalent in urbancounties, the percent of reoffenders is likely biasedlower in urban counties.3

As shown in Table 2 on Page 10, those juvenileswhose first offense was a felony were more likely tocommit another offense. Of the 32,456 felony offend-ers, 24 percent committed a second offense, while 18percent of the 154,717 nonfelony offendersreoffended. The rate of reoffense was lower in ruralcounties than in urban counties regardless of theoffense type. For example, the percent of reoffenderswho initially committed a felony in rural and urbancounties were 19 and 25 percent, respectively.

The data in Table 3 on Page 10 indicate that, asmight be expected, there were substantially more maleoffenders and reoffenders than female offenders andreoffenders. The difference in the percent ofreoffenders between urban and rural counties also waslarger for males than females.

Table 4 on Page 10 compares the number of offend-ers and reoffenders in urban and rural counties by race.The results are presented separately for whites, blacks,

and juveniles of other races (Asians, Pacific Islanders,and Native Americans). The lowest rate of reoffensewas among other races, followed by whites and thenblacks. For all races, the percent of reoffenders waslower in rural counties than in urban counties. Inparticular, for the other race category, the percent ofreoffenders was 15 percent in urban counties comparedto only 7 percent in rural counties, the largest differ-ence among the races.

Table 5 on Page 10 compares non-Hispanic andHispanic juvenile offenders and reoffenders. Thisethnicity classification did not distinguish amongraces. For Hispanic offenders, 7,733 resided in urbancounties and 353 resided in rural counties. The percentof reoffenders was substantially different betweenHispanics who resided in urban and rural counties. Inurban counties, the rate of reoffense for Hispanics was30 percent and in rural counties was 13 percent. Thestatewide percent of reoffense for Hispanics was 29percent, compared to 18 percent for non-Hispanics, orabout 1.6 times higher. It should be noted that the datadid not include information about citizenship orimmigration status.

Tables 1 through 5 seem to suggest that living in arural county is correlated with lower rates ofreoffense; however, caution should be exercised ininferring causation.

A Cursory Examination of MultipleReoffenders

Although the main objective of this study was toexamine the timing to the first reoffense (the timebetween the juvenile’s first and second referral), theresearchers also explored the issue of juveniles whocommitted multiple reoffenses. Tables 6 through 8 onPage 11 present the frequency distributions ofreoffenses at the statewide level and for urban andrural counties. Table 6 shows that, of the 187,173juveniles who were charged with a substantiatedoffense, 24,755 (approximately 13 percent) committedone more offense during 1997-2005, while 7,139 (4percent) committed exactly two more subsequentoffenses. Approximately 2 percent (3,443 juveniles)committed three or more subsequent offenses.

For juveniles in rural counties, 3 percent of thosecharged with a substantiated offense later committedtwo reoffenses, and about 4 percent in urban countiesdid so. Overall, juveniles who lived in rural countiesappeared to commit fewer multiple reoffenses com-pared to their counterparts in urban counties.

Results

3 The FBI’s Uniform Crime Reporting data for 2002 indicate thatthe violent crime rate in urban areas was 545.6 per 100,000population compared to 212.6 per 100,000 population in ruralcounties (available at: http://www.fbi.gov/ucr/cius_02/html/web/offreported/02-nviolent02.html).

Page 9: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania 9 The Center for Rural Pennsylvania

Table 1. Juvenile Recidivism Rates by County, 1997-2005

Page 10: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

The Center for Rural Pennsylvania 10 Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania

Estimates of JuvenileRecidivism

This section provides the firstestimates of juvenile recidivism inPennsylvania. Figure 1 shows thesurvival rate for all juveniles inPennsylvania, where the survivalrate measured the percent ofjuveniles who did not receive asecond substantiated referral at aspecific time period. The survivalat day 1,000 was approximately 75percent, meaning that 1,000 daysafter the juvenile’s first substanti-ated referral, roughly 75 percent ofthe juveniles did not return to thejuvenile justice system. An alterna-tive interpretation is that 25 percentof the juveniles did receive areferral leading to a second substan-tiated charge during the first 1,000days. At day 2,000, the survivalrate was lower (only about 60percent of the juveniles did notcommit a second offense by thistime), implying that the recidivismrate was higher. More specifically,the one-, two-, and three-yearrecidivism rates were 11, 20, and28 percent, respectively.

Figure 2 on Page 12 comparesthe survival rates between juvenilesliving in rural and urban counties.The survival curves were statisti-cally significant from one another.As juveniles in rural counties havelower recidivism rates, the survivalcurve for juveniles who lived inrural counties lies completely abovethe survival curve for juveniles wholived in urban counties. Thedifference in recidivism ratesbetween rural and urban countiesbecame larger over time. The one-,two- and three-year recidivism ratesin rural counties were 10, 17, and23 percent, respectively; for urbancounties these rates were 11, 21,and 29 percent, respectively.Restricting the analysis to theperiod 2002-2005 revealed that the

Table 2. Recidivism Ratesby Severity of First Offense, 1997-2005

Table 5. Offenders and Reoffenders by Ethnicity, 1997-2005

Table 4. Offenders and Reoffenders by Race, 1997-2005

Table 3. Recidivism Rates by Gender, 1997-2005

Page 11: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania 11 The Center for Rural Pennsylvania

one-, two-, and three-year recidi-vism rates in rural counties were11, 20, and 27 percent, respec-tively, which implied a slight rise inrecidivism since the 1997-2001period. The modest rise in recidi-vism may be attributable to recentimprovements in data collectingand reporting.

Analysis of JuvenileRecidivism

To explore how various demo-graphic characteristics may affectrecidivism, the researchers con-ducted a separate survival analysis.The estimates (Kaplan-Meier) areinformative but they cannot controlfor the effect that a particularvariable may have on recidivism.For example, the initial analysissuggested that juveniles residing inrural counties had lower recidivismrates. These results, however, didnot control for other confoundingfactors, such as race, familycharacteristics and othersociodemographic variables.Controlling for these omittedvariables may cause the differences

in recidivism rates between ruraland urban counties to disappear.

The initial model suggested thatjuveniles living in rural countieshad a longer expected time to theirsecond referral than juveniles wholived in urban counties. To test thisresult, the researchers reran themodel and excluded juveniles fromPhiladelphia. Despite the exclusionof Philadelphia, the model showed

that living in a rural county wascorrelated with reduced recidivism.

The first model containedvariables that characterize thejuvenile’s sociodemographic status.There was a variable indicatingwhether the juvenile was male totest for differences in recidivismbetween males and females. To helpcontrol for the effect of age onrecidivism, the juvenile’s age and

Table 6. FrequencyDistribution of Reoffensesin Pennsylvania, 1997-2005

Table 7. FrequencyDistribution of Reoffenses

in Rural Counties,1997-2005

Table 8. FrequencyDistribution of Reoffenses

in Urban Counties,1997-2005

Figure 1. Survival Estimates, All Juveniles, 1997-2005

Page 12: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

The Center for Rural Pennsylvania 12 Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania

age squared were used as explana-tory variables. A variable wasincluded indicating whether thejuvenile was black and anothervariable indicated whether thejuvenile belonged to another race,such as Asian or Native American.

Juveniles residing in ruralcounties had longer times to theirsecond referral than juvenilesresiding in urban counties, holdingconstant the sociodemographicvariables.

For the county-level variables,per capita income and the numberof police per capita were associatedwith lengthening the time to thejuvenile’s second referral. Percapita TANF, arrest rate, andpercent of cases dismissed injuvenile court reduced the time tothe second referral.

The researchers further investi-gated the effects of living in a ruralcounty on the length of time to thesecond referral. The rural variablewas interacted with all thesociodemographic variables,creating 11 interaction variables.Some findings were that malesliving in rural counties had longertimes to second referral, as well asthose who committed a felony inrural counties. Juveniles who were

ConclusionsThe research data show that Pennsylvania’s one-,

two-, and three- year recidivism rates were 11, 20, and28 percent. In other words, the three-year recidivismrate estimated that approximately 28 percent ofjuveniles who were convicted of a first offense wouldexperience another conviction within three years. Therecidivism rates were lower in rural counties, with one-,two-, and three-year recidivism rates of 10, 17, and 23percent, respectively. In contrast, the one-, two-, andthree-year recidivism rates in urban counties were 11,21, and 29 percent, respectively. For male juvenilesliving in rural counties, the rates of recidivism weremuch lower than their urban counterparts. For ex-ample, the three-year recidivism rate for males in rural

counties was 25 percent compared to a rate of 33percent for males in urban counties.

Two research models were used to determine factorsassociated with increased risks of recidivism. The firstshowed that juveniles with certain characteristics havean increased risk of recidivism. The characteristicsinclude being male, black, Hispanic, a felony offender,living with a single mother and having at least onedeceased parent. The second model indicated thatliving in an urban county, being male or Hispanic,living with a single mother, having at least one de-ceased parent, committing a prior felony, and enroll-ment in alternative education, were all associated withincreased recidivism.

Figure 2. Survival Estimates,Rural and Urban Counties, 1997-2005

represented by a private attorney inrural counties also had longersurvival times. There were twocases in which living in a ruralcounty appeared to adversely affectrecidivism: juveniles who livedwith their mother only, and thosewho received alternative educationhad quicker relapses into criminalactivity.

In models estimated but notdescribed, the effects of the percentof cases in a county leading toplacement were examined. The

most common types of delinquencyplacement were private institutions,group homes, drug and alcoholtreatment, and boot camps. Theresults were mixed. An examinationrevealed that juveniles who re-ceived placement were slightlymore likely to have a quickerrelapse into criminal activity. Theseresults did not suggest that place-ment was ineffective but that thejuveniles who were placed weresimply more troubled and morelikely to return to crime.

Page 13: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania 13 The Center for Rural Pennsylvania

Policy ConsiderationsThis research demonstrated the feasibility of measur-

ing juvenile recidivism in Pennsylvania by providing acomparison of recidivism between rural and urbancounties. Recidivism is an important measure of thesuccess and failure of juvenile justice interventions.

According to the Virginia Department of JuvenileJustice: “Recidivism is the key statistic in determiningwhether or not criminal justice interventions, fromdiversion through incarceration, are making a differ-ence in keeping offenders from committing morecrimes.”4 Unfortunately, Pennsylvania and a handful ofother states do not measure juvenile recidivism (Vir-ginia Department of Juvenile Justice, 2005).

The researchers recommend that Pennsylvania beginto document juvenile recidivism and issue formalreports on a regular basis. These reports could gobeyond the analysis of this study by examining recidi-vism rates of juveniles released from specific place-ment facilities, such as group homes, secure detentionfacilities, drug and alcohol treatment programs,Outward Bound programs, and supervised independentliving. If there are benefits to these programs, in termsof reduced recidivism, it would be possible to con-struct evidence-based measures of program effective-ness by cost-benefit analysis. Different types ofintervention have different costs. This type of analysismight reveal that the most successful interventions, orthose showing the greatest reduction in recidivism,may be the most costly, and thus should be used for thejuveniles whose criminal behavior predicts the greatestcost to society.

It would also be extremely useful to link the criminalrecords of adults with their juvenile records. By doingso, one could determine what sorts of juvenile recordsare most likely to be followed by serious criminaloffenses when the juvenile reaches adulthood, andwhat sorts of juvenile programs or dispositions aremost effective in reducing the likelihood of futurecrime.

The Pennsylvania government agencies best suited to

report on juvenile recidivism are the PennsylvaniaCommission on Crime and Delinquency (PCCD) andthe Center for Juvenile Justice Training and Researchat Shippensburg University.

The direct policy implications of this research werethe identification of juvenile characteristics thatdeserve closer scrutiny to help prevent a relapse intocrime. The characteristics associated with increasedrecidivism were living in an urban county, being maleor Hispanic, living with a single mother, having atleast one deceased parent, committing a prior felony,and attending alternative education. With respect to thehigh rates of recidivism among Hispanics, it may beimportant for the juvenile justice system to better tailorinterventions to meet the different cultural needs,language barriers, and customs of Hispanic juvenilesand their parents.5

Special attention should also be given to juvenilesliving with their mother only and those with at leastone deceased parent. More frequent contact with aprobation officer may be beneficial. It also may behelpful to reduce the caseloads of probation officersassigned to juveniles who live in single-parent familiesso that they can interact more closely with these at-riskjuveniles.

This research also documented a higher incidence ofrecidivism among juveniles with prior felony convic-tions. As with juveniles living in single-parent fami-lies, there may be a need for more intensive treatmentor intervention for felony offenders in an effort toreduce recidivism. Overall, the juvenile justice systemin Pennsylvania should address all the predictors ofrecidivism documented in this study. One method ofdoing so would be with increased program interven-tions and alternative treatment settings. However, it isimportant that all interventions, both old and new, areassessed for their effectiveness. One of the mostimportant indicators of program effectiveness isrecidivism. Therefore, program effectiveness should bemeasured at least in part by reports on recidivism.

4 Source: Virginia Department of Juvenile Justice, available at: http://www.vaperforms.virginia.gov/i-recidivism.php.5 Federal government mandates require health care organizations receiving federal money to provide culturally competent care, whichincludes calling for organizations to have strategies to recruit diverse staffs, and provide language interpreters and other culturallyappropriate services.

Page 14: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

The Center for Rural Pennsylvania 14 Measuring and Analyzing Juvenile Recidivism in Rural and Urban Pennsylvania

References

Farrington, D. P. (1992). “Criminal Career Research in the United Kingdom.” British Journal of Criminology, 32, 521–

536.

Grogger, J. (1997). “Local Violence and Educational Attainment.” Journal of Human Resources, 32(4), 659-682.

Mbuba, J. M. (2005). “A Refutation of Racial Differentials in the Juvenile Recidivism Rate Hypothesis.” African

Journal of Criminology & Justice Studies. 1(2). 51-68.

Paternoster, R., R. Brame, and D. P. Farrington. (2001). “On the Relationship Between Adolescent and Adult Convic-

tion Frequencies.” Journal of Quantitative Criminology, 17(3), 201-205.

Snyder, H. N., and M. Sickmund. (2006). Juvenile Offenders and Victims: 2006 National Report. Washington, DC: U.S.

Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention.

Virginia Department of Juvenile Justice. (2005). “Juvenile Recidivism in Virginia.” DJJ Quarterly, 3, 1-12. (Available

at http://www.djj.state.va.us/About_Us/Administrative_Units/Research_and_Evaluation_Unit/recidivism_RQ.)

Page 15: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal
Page 16: Juvenile Recidivism Report · adult criminal activity, society may first have to address juvenile crime. This research, conducted in 2007, estimated the extent of juvenile criminal

The Center for Rural Pennsylvania200 North Third St., Suite 600Harrisburg, PA 17101Phone: (717) 787-9555Fax: (717) 772-3587www.ruralpa.org1P0109 – 450

The Center for Rural PennsylvaniaBoard of Directors

Senator John R. Gordner, Chairman

Representative Tina Pickett, Vice Chairman

Senator John Wozniak, Treasurer

Dr. Nancy Falvo, Clarion University, Secretary

Representative Tim Seip

Dr. Theodore R. Alter, Pennsylvania State University

Steve Crawford, Governor’s Representative

Dr. Stephan J. Goetz, Northeast Regional Center for Rural Development

Dr. Keith T. Miller, Lock Haven University

Dr. Robert F. Pack, University of Pittsburgh

William Sturges, Governor’s Representative