Who gets a second chance? an investigation of Ohio's blended juvenile sentence.
|Abstract:||actors differentiating blended sentencing cases (Serious Youthful Offenders or SYOs) from conventional juvenile cases and cases transferred to the adult criminal court in Ohio were investigated using a two-stage probit. Conventional juvenile cases differed from cases selected for non-conventional processing (i.e., SYO or transfer) according to offense seriousness, number of prior Ohio Department of Youth Services placements, age and gender. Controlling for probability of selection for nonconventional processing, transfers differed from SYOs according to age, gender, and race. Minorities were significantly more likely than Whites to be transfers rather than SYOs, suggesting possible bias in the decision-making process. Objective risk and needs assessments should be used to identify the most suitable candidates for blended sentences and adult transfer and enhanced services should be provided to juvenile offenders given blended sentences.|
Juvenile offenders (Cases)
Social service (Analysis)
Cheesman, Fred L., II
Waters, Nicole L.
Hurst, Hunter, IV
|Publication:||Name: Journal of Health and Human Services Administration Publisher: Southern Public Administration Education Foundation, Inc. Audience: Academic Format: Magazine/Journal Subject: Government; Health Copyright: COPYRIGHT 2010 Southern Public Administration Education Foundation, Inc. ISSN: 1079-3739|
|Issue:||Date: Winter, 2010 Source Volume: 33 Source Issue: 3|
|Topic:||Event Code: 980 Legal issues & crime Advertising Code: 94 Legal/Government Regulation Computer Subject: Company legal issue|
|Product:||Product Code: 9105130 Social Service Support Programs NAICS Code: 92313 Administration of Human Resource Programs (except Education, Public Health, and Veterans' Affairs Programs)|
|Geographic:||Geographic Scope: United States Geographic Code: 1USA United States|
During the early 1990s, many state legislatures made sweeping changes in the dispositional and sentencing options available to juvenile courts, including the introduction of a new juvenile sentencing innovation, blended sentencing (2) Blended sentencing enables some courts to impose juvenile and/or adult correctional sanctions on certain young offenders (Sickmund, Snyder, & Poe-Yamagata, 1997). Sixteen states had blended sentencing statutes in place at the end of 1995 and at least 26 at the end of 2004. Using data from the 2000 census, these states encompass 60% of the nation's juvenile population (aged 10 to 17).
The purpose of the present article is to examine the practice of blended sentencing in Ohio. The objective is to identify the factors that influence the probability that juvenile offenders in Ohio will be processed in one of three possible dispositional alternatives or tracks: 1) Conventional juvenile; 2) Blended sentence (referred to as a Serious Youthful Offender or SYO); or 3) Transfer to the adult criminal justice system.
Blended sentencing emerged during a period of steadily increasing violent juvenile crime as a compromise between those who wanted to emphasize public safety, punishment, and accountability of juvenile offenders and those who wanted to maintain or strengthen the traditional juvenile justice system. It offers a means of resolving these disparate views because blended sentencing combines opportunities for rehabilitation in the juvenile justice system with the possibility of sanctions in the adult criminal justice system. Blended sentencing offers juvenile offenders a "last chance" at rehabilitation within the juvenile system by providing "an incentive to respond to treatment in order to avoid the consequences of an adult sentence" (Redding & Howell, 2000, p. 147).
States have adopted one of three blended sentencing regimes. The juvenile or criminal court may: (1) impose a juvenile or an adult sentence, (2) impose both a juvenile and adult sentence, with the adult sentence suspended under specified conditions, or (3) impose a sentence past the normal limit of juvenile court jurisdiction; typically, a hearing is held when the juvenile reaches the age of adult jurisdiction (varies by state, 18-21 years) to determine if an adult sentence will be imposed.
The National Center for Juvenile Justice (NCJJ) used these three general regimes as the basis for a widely used typology of blended sentencing practices in the states (Torbet, Gable, Hurst, Montgomery, Szymanski, and Thomas, 1996), summarized in Table 1. Of the 20 states with blended sentencing laws at the end of 1997, nine gave blended sentencing authority to juvenile court judges for cases involving some specified category of juvenile offender adjudicated delinquent. In nine other states, criminal court judges exercise blended sentencing authority following a juvenile's conviction. Two states, Colorado and Michigan, gave blended sentencing options to both juvenile and criminal court judges.
Regardless of the forum in which it is exercised, blended sentencing authority may be exclusive or inclusive, and under some circumstances, it may be contiguous (Torbet, Griffin, Hurst, and MacKenzie, 2000):
* An exclusive blended sentencing model allows a judge to impose either a juvenile or an adult sanction and makes that sanction effective immediately.
* Under an inclusive blended sentencing model a judge may impose both a juvenile and an adult sanction-with the latter usually remaining suspended and becoming effective only in the event of a subsequent violation.
* Finally, some states have enacted contiguous blended sentencing laws, under which a juvenile court may impose a sanction that begins in the juvenile system but lasts beyond the maximum age of extended juvenile court jurisdiction-at which time the offender must be moved into the adult correctional system to serve the remainder of the sentence.
Since January 2002, Ohio has been practicing a juvenile-inclusive blend, similar to Minnesota. The Ohio statutes governing the use of blended sentencing were modeled in large part on the Minnesota statutes (see Cheesman, Waters, Hurst, Halemba, Maggard, and Sohoni, 2009).
Despite the lack of empirical data on its use, especially in the context of the variety of different types of blended sentencing, Table 1 shows that the use of blended sentencing is widespread and increasing. Yet, blended sentencing is still a recent sentencing innovation and there remain opportunities to improve its practice. Efforts in this regard should be based on careful analysis of empirical data relevant to the issues listed above and as described in this article.
EMPIRICAL RESEARCH ON BLENDED SENTENCING
The two major studies that conducted multivariate statistical analyses of blended sentencing (Podkopacz and Feld, 2001, and Cheesman, Green, Cohen, Dancy, Kleiman, & Mott, 2002) both examined juvenile-inclusive blended sentencing in Minnesota (where it is referred to as Extended Juvenile Jurisdiction or EJJ). Podkopacz and Feld compared the characteristics of juvenile offenders in Minnesota's most populous county (Hennepin) who were (1) motioned for either transfer or a blended sentence after the introduction of EJJ with those of juvenile offenders (2) motioned for transfer before the introduction of EJJ. The post-EJJ EJJs and transfers were more likely than the preEJJ transfers to be: younger, African-American, female, charged with an offense that entailed a presumptive commitment to prison, and charged with fewer offenses. Compared to the pre-EJJ group, the post-EJJ group also had fewer prior offenses, began their criminal careers later, and had fewer out-of-home placements. When they compared post-EJJ juvenile offenders against whom prosecutors filed transfer motions with those motioned for EJJ, they found several similarities and differences. The two groups did not differ significantly on the basis of race, gender, prior offenses, out-of-home placements, or age-of-onset of delinquency but did differ according to seriousness of the current offense and age. They also found evidence that widening-of-the-net had occurred after the introduction of blended sentencing.
Cheesman et al. (2002) examined a statewide random sample of juvenile cases and employed a two-stage statistical analysis, controlling for selection bias. They found that both intended factors (age, offense seriousness, number of charges, whether there was an adult codefendant, involvement of a firearm, and number of out-of home placements) and unintended factors (race and judicial district) influenced the prosecutor's decision to motion for EJJ or transfer. When they compared juveniles transferred with those given EJJ status, they found that only one intended factor, prior offense record, strongly distinguished the two in the direction expected. Clearly, the designers of the blended sentence in Minnesota did not intend an extralegal factor such as race to influence decisions about motioning.
Age and current offense factors also distinguished the two but were much less influential. Most current offense factors failed to distinguish the two, including whether the offense involved a firearm, number of charges, victim injury, and type of offense. Thus offenders with charges against another person, offenders who used firearms, and offenders who injured their victim were just as likely to be EJJs as to be transferred. Unexpectedly, the probability of transfer was found to increase as the seriousness of the offense decreased, suggesting that this processing alternative is not being targeted as intended. Moreover, judicial district and race were found to influence the odds of transfer vs. EJJ more than any intended factor, except prior offense history. Consequently, minorities were under-represented among the blended sentencing cases but over-represented among transfers. The results indicate that the factors influencing the decision to motion are different than the factors that influence the selection of processing alternatives and that the latter selection is not influenced by the probability of motioning.
If blended sentences offer juvenile offenders a "last chance" at rehabilitation within the juvenile system, to who is this chance offered? Does the probability of a blended sentence and transfer vary according to extralegal factors, particularly race? It has been well documented that minority juveniles stand a higher probability of transfer to the adult criminal court than white juveniles and that they are consequently disproportionately represented among the population of transfers relative to their representation among all juvenile offenders (e.g., Bishop and Frazier, 1991; Champion and Mays, 1991; Poulos and Orchowsky, 1994; McNulty, 1996; Office of Juvenile and Delinquency Prevention, 1999; Males and Macallair, 2000). Further, as described above, Cheesman et. al. found evidence that race is also a factor in the decision to transfer and to impose a blended sentence in Minnesota. Consequently, we sought to determine whether this key finding from the Minnesota study would be replicated in Ohio. Such a finding would strengthen the argument that current juvenile sentencing practices in states that have adopted blended sentencing should be scrutinized to ensure that unintended extralegal factors such as race do not influence selection of processing alternatives (i.e., transfer, blended sentence, or conventional juvenile processing).
In the current study, the National Center for State Courts (NCSC) examined the practice of blended sentencing in Ohio. To compare the alternative processing tracks, the NCSC collected data on the use of processing alternatives for juvenile offenders from five counties in Ohio (Cuyahoga, Hamilton, Lucas, Summit, and Delaware) who were adjudicated for felony offenses between 2002 and 2004.
Staff from NCJJ conducted site visits to each of these counties, interviewing judges, court administrators, prosecutors and other stakeholders in the juvenile justice system. State-level officials involved in the writing and enabling of the legislation that created the blended sentencing option were also interviewed. Relevant documents and legislation were examined. NCJJ also made inquiries about the availability of data and facilitated its eventual collection. NCJJ's analysis of the practice of blended sentencing in Ohio (available in Cheesman, Waters, Hurst, Halemba, Maggard, and Sohoni, 2010) directly informed the quantitative analysis described in the following.
Although the sample from these five counties does not constitute a random sample of juvenile adjudications across (N=28,628) all counties in Ohio, it should be noted that these five counties collectively accounted for a very significant proportion (75%) of Ohio's statewide juvenile adjudications between 2002 and 2004.
The Ohio Department of Youth Services (ODYS) provided NCSC with two databases from which the overall sample was selected. One database provided information on all juvenile felony adjudications for SYOs and conventional juveniles from 2002 through 2004 while the other provided data on all transfers from the same period. The participating counties provided the NCSC with a listing of their SYO cases as SYO juveniles were not separately identified in the database provided by ODYS.
Data on case processing, offense history, and offender characteristics were obtained primarily from automated databases maintained locally at each of our sites. All data were scanned for obvious errors and our team communicated freely with each of these sites to seek clarification if the need arose. Beyond this, we depended on the sites to maintain valid datasets. Our NCJJ partners had had extensive experience working in most of these sites and expressed general confidence in their data.
The final sample included all blended sentencing (SYO) and transfer cases from the five counties, adjudicated or sentenced between 2002 and 2004 (139 and 164 cases, respectively). The NCSC also drew a proportionate3 random sample of 340 conventional juveniles from each of the five participating counties.
The five counties shown in Table 1 produced a total of 28,628 felony adjudications, 139 Serious Youthful Offender adjudications (blended sentences), and 164 transfers. The ratio of SYOs to conventional adjudications was about 205 to 1 while that for transfers was 174 to 1. Clearly, both phenomena are rare events in the juvenile justice system in Ohio.
To what types of juveniles do judges give a second chance? Who do judges and prosecutors recommend for adult criminal sanctions? The next few sections will describe the characteristics of Ohio juveniles within each processing track. The results of bivariate analyses of processing track with other relevant variables are first presented, followed by the results of a multivariate analysis of the determinants of processing track.
Processing county.--The distribution of juvenile dispositions by county is shown in Table 2. Hamilton and Cuyahoga County processed the largest number of conventional juvenile cases, together comprising just over half of all such cases, while Delaware County accounted for only six percent. Cuyahoga County processed the largest percentage of conventional juvenile cases in the sample, followed by Hamilton, Summit, and Lucas Counties, respectively.
Though the total number of cases processed by Hamilton and Cuyahoga is similar, cases from the two counties were distributed differently among the three processing alternatives. Cuyahoga County processed proportionately more conventional juvenile cases than Hamilton. While nearly half of all juveniles transferred to adult court originated in Hamilton County, only one-fifth originated in Cuyahoga County.
Lucas and Summit Counties each accounted for approximately one-third of the juveniles receiving Serious Youthful Offender (SYO) dispositions. Proportionately, very few juveniles received SYO dispositions in Hamilton County, which more often transfers juvenile felony offenders.
Age, gender, race and ethnicity.--Juveniles ranged in age from 12 to 19 at the time of filing. Transfers (Mean=16.5) are significantly older than other juveniles, while SYOs (Mean=15.8) are significantly older than conventional juveniles (Mean=15.6) (F(2,607)= 37.31, p<.046). (4) The sample includes a total of 62 female juveniles, most of whom were processed as conventional juveniles. Females comprised 15% of the conventional sample, but only 1% of the transfers, and 6% of the SYO adjudications ([chi square]=25.569, df=2, p<.000). This finding confirms expectations, as females tend to commit less serious crimes than their male counterparts.
Black juveniles constituted the largest racial or ethnic group identity within each juvenile processing track. When all non-whites were collapsed into a single category, "Minority, (5)" and Race/Ethnicity was cross-tabulated with juvenile track, it was significantly associated with processing track ([chi square]=16.086, df=2, p<.000). Minorities were under-represented among SYOs (62%) and disproportionately over-represented among the transfers (81%), in comparison to their representation in the counterfactual group, conventional juveniles (65%).
To put these percentages in context, it is insightful to consider the proportionate representation of non-caucasians in each of the counties according to 2009 estimates from the U.S. Census. Non-caucasians account for almost a third of the population of Cuyahoga (33.2 %) and more than a fourth of the population of Hamilton County (28.4%). Around one-fifth of the populations of Lucas and Summit Counties (21.7% and 17.6%, respectively) and about one-tenth (9.7%) of the population of Delaware County were non-caucasian. Clearly, minorities are over-represented with respect to their proportionate representation among the general population in all three processing tracks in all five counties, particularly for transfers.
Enrollment status in school.--A juvenile's enrollment in school provides evidence of a juvenile's stability within the community and failure to bond to school has been identified as a risk factor for delinquency (Wasserman, Keenan, Tremblay, Coie, Herrenkohl, Loeber, and Petechuk, 2003). Two counties (Lucas and Summit) were able to provide comprehensive data on the enrollment in school status for the sample of juveniles. (6) These data show that those transferred to the adult criminal court (28%) were least likely to be enrolled ([chi square]=33.801, df=2, p=.000), as compared to SYOs (43%) and conventional juvenile cases (65%).
Only 18% of Summit transfers were enrolled in school, in contrast to 39% of their counterparts from Lucas. Juveniles processed as SYOs in Lucas County were more likely to be enrolled in school than their counterparts from Summit County (57% vs. 29%, respectively). On average, Summit juveniles had completed ninth grade at the time of their arrest.
Family structure.--It is commonly held that juveniles from nuclear families are less likely to engage in delinquency than juveniles living under alternative family structures. Research supports this belief, though the relationship between family structure and delinquency is moderated by the nature and quality of family interactions. Using data from the 1995 National Longitudinal Survey of Adolescent Health, Demuth and Brown (2004) found that adolescents in single-parent families are significantly more delinquent than their counterparts residing with two biological, married parents, although these differences are reduced once the authors account for various family processes. A lack of parental interaction and involvement increases the risk for violence, particularly among males (Hawkins, Herrenkohl, Farrington, Brewer, Catalano, Harachi, and Cothern, 2000). Failure to set clear expectations, inadequate supervision and monitoring, and severe or inconsistent family discipline practices can also contribute to delinquency and violent behavior (Wasserman et. al., 2003).
Consequently, juveniles who resided with their parents were expected to be processed in a less punitive track. A large majority (75-86%) of the juveniles for whom data was available lived with their parents or stepparents at the time of their adjudication. (7) As expected, juveniles transferred to adult court were least likely to reside with their parents, though the difference is not statistically significant ([chi square]=3.40, df=2, p=.066). About 75% of transfers lived with their parents, compared to 86% of SYOs, and 88% of conventional juveniles. The highest percentage of juveniles living with his or her parents was reported for Summit County.
Offense seriousness.--Prosecutors determine processing tracks based on offense seriousness, among other factors. Juvenile offenders in the sample committed a wide range of crimes, yet all faced, at minimum, a fifth level felony charge.
Prosecutors most often charged conventional juveniles with less serious felony charges (fourth or fifth level). Prosecutors recommended a juvenile for a transfer to the adult court often when the charge was a serious felony (first or second degree felony, and murder/attempted murder), as shown in Figure 1. SYOs were distributed among felony levels similarly to transfers with both groups being most commonly charged with a first or second level felony. Overall, the seriousness of the offense was significantly related to the processing track ([chi square]=277.084, df=2, p<.000).
[FIGURE 1 OMITTED]
The prosecutor's original felony charge was almost always consistent with the level charged at adjudication. Prosecutors adjusted only 12% of the charges from the original charging level. In all but a few situations, prosecutors lowered the level of the charge (e.g., from a 2nd level felony to a 3rd).
Number of charges.--In addition to the felony level and type of offense, the number of charges the juvenile faced acted as a proxy to measure overall offense seriousness. Most often, juveniles faced approximately three charges per adjudication. Transfer juveniles faced more felony charges (5 charges on average) than other juveniles (3 for SYOs and conventional juveniles). The difference in the number of charges between juveniles transferred to adult court and the other processing tracks was significant (F(2,600) = 14.4, p = .000). Post-hoc analyses did not significantly distinguish SYOs from conventional juveniles.
Co-defendants.--Association with deviant peers has been identified as a risk factor for delinquency (Wasserman et. al., 2003). How do situational factors such as the presence of co-defendant(s) during the commission of a crime, gang involvement, or a crime against another person as compared to a property offense affect case processing? Summit County provided data on the participation of both adult and juvenile co-defendants in the adjudicated offense, helping to answer this question in that jurisdiction. (8) In Summit, 36 of the 38 juvenile SYOs committed the offense with an adult co-defendant, whereas only 15 (44%) of the 34 transfers, and none of the conventional juveniles committed their offense with an adult co-defendant. Summit juveniles were charged with another juvenile codefendant in 9% of conventional juvenile cases, 68% of transfers, and 95% of SYOs ([chi square]=75.32, df=2, p<.000). In Lucas, juvenile co-defendants in cases transferred to the adult court were also rare, but more evenly split among conventional and SYO juveniles than in Summit (42% conventional, 38% SYO, 12% transfer, [chi square]=5.34, df=2, p<.069). Collectively, this evidence suggests that the participation of other adults or juveniles during the commission of the crime may have influenced the charging prosecutor to seek an SYO designation.
Gang-related offense.--Most juveniles charged with a gang-related offense were transferred to the adult criminal court. In Summit County, prosecutors charged over one-third of the juveniles with gang-related offenses, none of which were processed as conventional juveniles. Of those charged with a gang-related offense, 58% were SYOs and 68% were transfers. (9) This finding suggests that, much as with the presence of co-defendant(s), gang involvement influences prosecutors' decisions aboutr how to process the case.
Victims.--Whether the juvenile harmed a victim or not during the commission of their offense varied widely by jurisdiction; just over one-third (35%) of the juvenile offenses involved a victim in Hamilton compared to 92% in Delaware. In Hamilton and Summit Counties, a significantly higher percentage of SYO offenses involved a victim ([chi square]=22.669, df=2, p < .000 and [chi square]=16.162, df=2, p < .000, respectively) than offenses committed by offenders processed in the other two tracks. This was not the case in Delaware ([chi square]=0.156, df=2, p=.925) and Lucas ([chi square]=5.379, df=2, p=.068) Counties. SYOs and transfers most often committed crimes against male victims while conventional juveniles were more likely to have targeted female victims.
History Of Juvenile Justice Involvement
Age-at-first-arrest.--Age-at-first-arrest speaks to the juvenile's amenability to reform in the juvenile justice system and is a predictor of chronic delinquency (Patterson, Forgatch, Yoerger, and Stoolmiller, 1998). A younger ageat-first arrest increases the probability of chronic delinquency (see, e.g., Loeber, Farrington, and Waschbusch, 1998) and consequently reflects a diminished amenability to reform in the juvenile justice system. Juvenile ages at the time of their first arrest are differentially distributed across the Ohio counties. Cuyahoga juveniles were the youngest at first arrest (M=11.1) and Delaware juveniles were the oldest at first arrest (M=15.4). Juveniles transferred to adult court were, on average, significantly older at the time of their first arrest than the other two groups (13.8 years compared to 12.7 and 12.8 years for conventional and SYOs, respectively; F(2,580)=5.14, p=.006).
Number of prior referrals to juvenile court and DYS.--Another key predictor of processing track is how often the juvenile has been previously referred to court. Cuyahoga juveniles, on average, had 17 previous court referrals (Median =15), while the average number of previous referrals hovered near seven in the other counties. Transfers from all counties except Delaware, reported more court referrals than the other two processing tracks. SYOs had more prior court referrals in their past than conventional juveniles, but fewer than transfers (F(2, 580)=11.579, p=.000).
Prosecutors and judges are influenced by the offender's past placement history with the Ohio Department of Youth Services (DYS) in determining how the juvenile will be processed. DYS is the juvenile corrections system for the state of Ohio. DYS is statutorily mandated to confine felony offenders, ages 10 to 21, who have been adjudicated and committed by one of Ohio's 88 county juvenile courts. Placement with DYS means confinement in a secure facility.
More to the point, it is an explicit factor to be considered when considering transfer or SYO, by statute. Juveniles were typically placed with DYS one or two times prior to the current adjudication. (10) As expected, juveniles transferred to the adult court had been placed in DYS custody more often than SYOs or conventional juveniles and SYOs tended to have more DYS placements than conventional juveniles ([chi square]=68.095, df=16, p<.000). Juveniles processed in Summit County provide an exception since their SYOs had been placed in DYS facilities more frequently than offenders transferred to the adult criminal court. Hamilton transfers had a DYS history with two to three prior placements. Delaware juveniles were rarely placed in DYS.
Several processing variables are considered in this section including whether the juvenile offender was detained after arrest, whether the case was plea-bargained, motioning, and amenability hearings. Differences among the processing tracks for these variables are discussed.
Detention.--Research has consistently shown that juveniles who are detained prior to adjudication tend to fare worse at the hands of the juvenile justice system than those released (see, e.g., Frazier and Bishop, 1985). The range of detention rates across sites was extensive ([chi square]=9.686, df=2, p=.008). (11) Delaware detained only about half of its juveniles prior to adjudication, whereas, Lucas held 75% and 81% in Summit. In Hamilton, virtually all juveniles (99%) were detained.
Juveniles transferred to the adult court fared the worst. Delaware detained its only transfer juvenile and approximately half of the SYO juveniles. (12) Lucas detained virtually all juveniles (94%) transferred to adult court, 83% of its SYOs and 66% of the conventional juveniles ([chi square]=8.165, df=2, p=.017). Summit detained both transfers and SYOs at a high rate (91% and 95%, respectively), and detained conventional juveniles two-thirds of the time ([chi square]=15.404, df=2, p=.000).
Plea bargaining.--Prosecutorial discretion allows prosecutors to offer a plea agreements to some juveniles before trial. Such an offer was provided to a large majority (88%) of juveniles in Cuyahoga, two-thirds (66%) of juveniles in Delaware, and to only one-quarter in Hamilton and Summit (29% and 24%, respectively).
A breakdown by processing track reveals that virtually all of the conventional and SYO juveniles in Cuyahoga were offered a plea agreement by the prosecuting attorney, but only half of juveniles transferred to the adult court. In Delaware, the prosecutor offered a plea to three-fourths of conventional juveniles, and just over half of the SYOs, but not to the lone transfer. Hamilton prosecutors offered a plea to 37% of the conventional juveniles and 39% of the SYOs, but only 19% of the transfers. A similar pattern emerged in Summit County where prosecutors offered 34% of conventional juveniles a plea and 29% of SYO juveniles, but no plea was offered to any of its 34 transfer juveniles.
In Hamilton, of the 48 juveniles offered a plea, all but the transfer juveniles accepted the plea. Cuyahoga data reveal the same pattern--conventional and SYO juveniles tended to accept offered pleas (90% and 96%, respectively), but transfer juveniles did not (only 7% accepted). All pleas were accepted in Summit and Delaware Counties.
Motioning.--SYO status can be either mandatory (in the case of certain types of offenders whose motion for transfer to the adult criminal court fails) or discretionary. In the case of a discretionary SYO, the prosecutor must file a petition for SYO status with the juvenile court which in turn rules on the motion. Undoubtedly, much discretion is exercised in the decision to file for SYO status. Discretionary transfers require attorneys to file a motion to transfer with the juvenile court.
In both Summit and Delaware, SYO motions were made for approximately one-third of the juveniles (31% and 34%, respectively). These percentages encompass virtually all juveniles processed as SYOs. In Summit County, only two juveniles were motioned for SYO and not granted SYO status (one resulted in a transfer and one remained in the juvenile court). One juvenile in Summit County received a SYO designation, and yet, according to the data, the juvenile did not have a record of any SYO motion made on his behalf.
In Summit and Hamilton Counties, prosecutors motioned juveniles for transfer to the adult criminal court in just under half of the cases (46% and 44%, respectively). It was very unusual for a case that had been motioned for transfer in Summit, Hamilton, or Delaware Counties to ultimately end up as a conventionally processed case. However, 66% of SYOs in Summit and 62% in Hamilton received original transfer motions. Only two of the Delaware SYO cases were originally motioned for transfer (15%); whereas, all transfer cases in Delaware and Summit Counties and a majority of Hamilton transfer cases (71%) were originally motioned as transfers. That such a high proportion of SYO cases in Summit and Hamilton Counties were originally motioned for transfer suggests that prosecutors in those counties identify juveniles for either SYO or transfer, and decide later which of these two processing tracks to select. (13)
Amenability Hearings.--By statute, a major factor for prosecutors and juvenile judges to consider when deciding the appropriate processing track for a juvenile offender in Ohio is the offender's amenability to treatment in the juvenile justice system which refers to the probability that the juvenile offender will respond favorably to interventions designed to reduce the probability of re-offending. An amenability hearing is required when transfer or SYO status is sought by the prosecutor for the purpose of making this determination.
Amenability hearing data were available for Summit and Delaware Counties only. Summit County held amenability hearings in one-third of its cases (75% of SYO cases, 59% of transfer cases, and 3% of conventional juvenile cases). Approximately one-third of the Summit juveniles were judged amenable at the hearing and in another one-fifth of the cases, prosecutors withdrew the motions.
In addition to data on the juvenile, his or her offense, and adjudication, information on the disposition received by each juvenile, including court-ordered treatment, was also collected. Lucas and Delaware provided information on all treatment services, both inpatient and outpatient. Summit provided drug and mental health treatment data.
Overall, approximately 200 juveniles from Lucas, Summit, and Delaware Counties were sentenced to receive treatment. Predominantly, the treatment was outpatient and more often than not, targeted juvenile drug use. Generally, inpatient treatment was rare. Approximately half of those who received drug or delinquency treatment were conventional juveniles and approximately one-third were processed as SYOs. Nearly two-thirds (61%) of juveniles receiving treatment for mental health issues were conventional juveniles.
One reason cases may be retained in the juvenile justice system is for the purpose of providing treatment to the juvenile. For example, one-third of Lucas juveniles receiving mental health treatment were SYO cases. Delaware juveniles receiving treatment were virtually always conventional cases and were almost never transferred to adult criminal court. Summit juveniles, on the other hand, received treatment in approximately equal proportions across processing tracks and approximately one-third of their SYOs received drug treatment (both inpatient and outpatient). Slightly more than one-third of the conventional juveniles received mental health outpatient care. Inpatient care was rare, but in Summit County 23 juveniles received inpatient mental health treatment, over half of whom were transferred to the adult criminal court. Just under one-third of the juveniles who received inpatient treatment for mental health issues were conventional juveniles and only 17% were SYOs.
What is the profile of a typical juvenile processed through each of the three tracks? Several criteria best differentiated how juveniles were processed. Predominantly, the offenders were male; female offenders were particularly rare among the transfers and SYOs. A typical juvenile transferred to adult court tends to be older than juveniles processed as SYOs or as conventional juveniles. There was a high proportion of minorities in all three tracks, but minority juveniles were disproportionately over-represented among offenders transferred to the adult court and under-represented among juveniles given a blended sentence, relative to the proportion of minorities among conventional juvenile offenders.
The county of adjudication accounted for most of the variation in how the case was processed and whether the juvenile was detained prior to adjudication. The odds of a blended sentence were higher in Delaware County than in Cuyahoga. The odds of transfer were much higher for juvenile offenders from Hamilton County than any of the other counties. In Delaware, only half of the juveniles were detained prior to adjudication, but over three-quarters were detained in the other four counties. In particular, transfer juveniles were detained at a higher rate than the other two processing tracks.
Transfer juveniles were least likely to be enrolled in school and were most likely to have been charged with a first degree felony, murder or attempted murder. Conventional juveniles were most likely to face fourth and fifth degree felonies and SYO juveniles fell mid-way between, most likely facing first or second degree felony charges.
The univariate and bivariate results just described are used in the following to inform a multivariate statistical model that identifies factors that influence the probability of a juvenile offender being processed by each of the three processing track alternatives.
Factors Influencing the Selection of Processing Track
Bivariate tests of significance, such as those conducted in the last section, are useful for identifying factors that influence the probability of the selection of processing track, but ignore the possibility that more than one factor at a time is jointly influencing this selection. Multivariate regression allows for and, in fact, anticipates that independent variables will jointly influence the probability of the selection of processing track.
In the present case, the factors that influence the decision to process a juvenile offender by one of three possible tracks, (1) Conventional juvenile court processing, (2) Transfer to the adult criminal justice system, or (3) Serious Youthful Offender (SYO) were examined. Just as was the case in Minnesota, (Cheesman et al., 2002), it is hypothesized that track selection is a two-step process. First, the prosecutor decides whether to process the juvenile conventionally in juvenile court or to select the juvenile for special processing, by motioning for either SYO or transfer. At this early stage of processing, it is hypothesized that the prosecutor hasn't determined whether to process these cases as SYOs or transfers, but has decided that the case merits more than "business as usual" juvenile court processing. The second stage of track selection involves the selection of either the SYO or transfer tracks for the juveniles selected for special processing in the first stage.
There is some empirical support for this approach to the analysis. As noted in the discussion of motioning, 66% of SYOs in Summit and 62% in Hamilton had originally been motioned for transfer. These results suggest that while prosecutors quickly identify candidates for special processing, the decision as to process the case as a transfer or SYO is made at a later stage of processing.
Maximum likelihood probit estimation with selection is an appropriate technique for this problem. Selection is clearly an issue in this analysis because, as described above, we hypothesize that the process of selecting juvenile offenders for unconventional processing is not random but based on formal and informal criteria. Statistical analyses based on non-randomly selected samples can lead to erroneous conclusions and poor policy if appropriate techniques are not applied. Maximum likelihood (14) probit estimation with selection, a two-step statistical approach, offers a means of correcting for non-randomly selected samples, appropriate for dichotomous dependent variables under the assumption of normally distributed errors. Since we hypothesize that the process of selecting processing tracks for juvenile offenders is a two-step process involving two sequential dichotomous decisions (i.e., conventional vs. unconventional processing and then SYO vs. transfer), maximum likelihood probit estimation with selection provides an appropriate model of the processes we seek to explain.
Taking this approach, one first estimates a probit regression to identify the factors that influence the decision to process the case either conventionally or non-conventionally, referred to as the "selection equation." One then calculates a second probit regression to determine the factors that influence the decision to process the case as an SYO or as a transfer, controlling for the probability that the case was selected for special processing.
The predictors for both regression equations were identical with one exception (see Table 3). The selection equation includes a dummy variable that indicated whether the juvenile offender was charged with a second degree felony while the second probit does not. Bivariate analyses revealed that being charged with murder or attempted murder, a first degree felony, or a second degree felony was associated with an increased probability of special processing, and hence their inclusion in the regression. Another set of bivariate analyses examined the factors that influence the selection of either the SYO or transfer tracks among those selected for special processing. This analysis revealed that being charged with murder or attempted murder or being charged with a first degree felony, but not a second degree felony, was predictive of transfer. Consequently, the dummy variable for being charged with a second degree felony was omitted from the second probit.
It should be noted that the analysis was conducted on a state-level, which is to say that the individual counties were not entered as predictor variables. While the authors acknowledge the differences between the counties detailed in the preceding section of this paper, it is also important to recognize that juvenile sentencing policy is made at the state not the local level. Even if the selection process is biased in only one county (or a small number of counties), it is a matter of state-wide concern because such bias will have been shown to be possible and may emerge in additional counties and it flies in the face of the uniform application of justice to similarly situated offenders. The corrective action needs to be taken at the state-level to promote the uniform and fair application of justice. This is especially true in Ohio where the juvenile court is very strongly and historically county-based.
Because the object is to identify the factors that influence the probability of a given processing track for a particular case, it is necessary to weight the cases used in this analysis according to the frequency with which a given dispositional alternative was invoked during 2002- 2004. Only the randomly selected conventional juvenile cases required a weight, since the full population of transfers and SYOs from the counties that supplied us with data was obtained.
Table 3 summarizes some of the key results obtained from the maximum likelihood probit model with selection, using case weights. The table shows that the Wald test of independent equations was non-significant, indicating that the two probit equations can be analyzed independently. In other words, the process of selecting cases for non-conventional processing, as opposed to conventional processing, in the juvenile justice system does not seem to influence the selection of nonconventional cases for transfer as opposed to SYO.
The selection equation identifies the factors that influence the probability of selection for special processing (i.e., as either an SYO or a transfer) as opposed to conventional processing in the juvenile system. Juveniles charged with murder or attempted murder, a first degree felony, or a second degree felony offense (as opposed to being adjudicated for a third, fourth, or fifth degree felony offense) had a significantly higher probability of being processed unconventionally. Likewise, being older, being male, and having more prior DYS placements all increased the probability of special processing. Put another way, being adjudicated for murder or attempted murder, a first degree felony, or a second degree felony, being older, being male, and having more prior DYS placements all increased the probability that a juvenile offender would be processed as either a transfer or SYO, as opposed to being processed conventionally in the juvenile justice system. Race, number of non-DYS out-of-home placements, and the number of prior felonies were not predictive of the selection of non-conventional as opposed to conventional processing.
The second-stage probit identifies the factors that influence the selection of either the transfer or the SYO tracks, given that the juvenile offender had been previously selected for special processing. Table 3 shows that only demographic factors differentiated SYOs and transfers, to the exclusion of any legalistic factors. Being older, being male, and being non-white, all significantly increased the probability of transfer as opposed to SYO. Being charged with a murder or a first degree felony, number of charges, number of prior DYS commitments, number of non-DYS out-of-home placements, and the number of prior felonies were not predictive of the selection of transfer as opposed to SYO.
The selection equation (Pseudo R =.212) was able to correctly classify 57% of the cases (a slight improvement over chance, which would have correctly classified 55% of the cases simply by assigning all of the cases to the most frequently occurring dispositional alternative, conventional juvenile processing). By type of dispositional alternative, 100% of the conventional juvenile offenders were correctly classified while only 5% of the transfers and 4% of the SYOs were correctly classified as non-conventional cases. This equation represented only a marginal improvement over chance, in part because of the relative scarcity of non-conventional cases (only three percent of the weighted cases) and in part because the two tracks were genuinely difficult to distinguish.
The second-stage probit equation (Pseudo [R.sup.2] =.161) was able to correctly classify 68% of the cases, a significant improvement over chance, which would have correctly classified 50% of the cases simply by assigning all of the cases to the most frequently occurring dispositional alternative (transfer). By type of dispositional alternative, 69% of the transfers and 66% of the SYOs were correctly classified.
The information generated by this investigation, in conjunction with other knowledge already generated by the handful of empirical studies of blended sentencing, is useful in addressing several questions about the actual practice of blended sentencing and how that practice aligns with intentions. The censored probit with selection model demonstrated that juvenile offenders who were processed as conventional juvenile offenders could be distinguished from juvenile offenders whose cases were transferred to the adult criminal justice system in ways that were generally expected.
The selection model identified factors that influenced the prosecutors' decision to process a juvenile case as a conventional case or to seek a more significant sanction through designation as a SYO or by transferring the case to the adult criminal justice system (i.e., "special processing"). It was hypothesized, consistent with research on blended sentencing conducted in Minnesota, that this is the initial decision made by a prosecutor when deciding how to process a case, leaving until later, with additional information, the decision whether to process the case as a SYO or as a transfer.
Results showed that this initial decision was influenced primarily by legal factors. The type of offense with which the offender was charged influenced the probability of special as opposed to conventional processing. Offenders charged with murder or attempted murder, a first degree felony, or a second degree felony were more likely to be processed non-conventionally than offenders charged with lesser felonies.
The probability of nonconventional processing for a given offender increases in direct proportion to the number of prior DYS placements for that offender. This result is not unexpected since this factor is a key consideration in determining the juvenile's amenability to treatment in the juvenile justice system, which in turn, influences the selection of processing track.
Age and gender were the only significant demographic variables influencing the initial processing decision. Compared to younger offenders, older juvenile offenders had a higher probability of special processing. This result is not surprising since age is, by statute, a legitimate factor to consider when making decisions about the appropriate processing track for a particular case. Males were more likely to undergo special processing than females, a finding consistent with prior research on transfer (Bishop, Frazier, and Lanza-Kaduce, 1997) and blended sentencing (Cheesman et al., 2002).
The second-stage probit identified the factors that differentiated transfers from SYOs, controlling for the probability of selection for special processing. A mix of legalistic and demographic factors differentiated these two processing tracks Among the legal variables, offense seriousness differentiates the two processing tracks since offenders charged with murder or attempted murder are significantly more likely to be processed as transfers rather than SYOs, in comparison to offenders charged with lower level felonies. Interestingly, number of prior DYS placements was not a significant predictor of processing track at this stage.
Demographic factors played a greater role in the selection of transfer over SYO than in the selection of special over conventional processing. As was the case with the selection model, age and gender were significant predictors of processing track. Males and older juveniles were significantly more likely to be processed as transfers than SYOs.
In addition, race was also a significant predictor of processing track in the second-stage probit. Minorities were significantly more likely than Whites to be processed as transfers rather than as SYOs. Thus, while race is apparently not a factor in the decision to process a case conventionally or non-conventionally, once an offender has been selected for special processing, minorities have a significantly greater chance to be processed by the more punitive track, transfer to adult court, than as SYOs. This result suggests the possible operation of bias during the second stage of the decision-making process.
There has been only limited research on the factors that influence the selection of a blended sentence over transfer (Cheesman, et al., 2002; Podkopacz & Feld, 2001). The results reported herein are generally consistent with the results reported by Cheesman et al. (2002) in Minnesota, another state that practices a "juvenile-inclusive" type of blended sentencing. One exception is the role that offense seriousness played in differentiating transfers from blended sentencing cases. Offense seriousness was not found to differentiate blended sentencing cases (SYOs) from transfers in Ohio although it did in Minnesota, (albeit in an unexpected direction). In Minnesota, blended sentencing cases had more serious charges than transfers. Clearly, however, in both states, transfer was not being reserved for the "worst of the worst" and blended sentencing cases were not the "least worst of the worst" (Feld, 1995b).
More troubling, yet consistent with the results in Minnesota, was the finding that, all other things being equal, minority juvenile offenders were significantly less likely than white offenders to be SYOs as opposed to being transferred, even after controlling for the influence of a variety of legalistic variables. The high transfer rate of minorities has been noted in many studies (e.g., PoeYamagata and Jones, 2000) and while SYO held the promise of impacting this high rate by helping to maintain additional minority juveniles under the control of the juvenile justice system as SYOs, it is not living up to this part of its promise. SYOs contain the highest proportion of whites among the three processing alternatives (38% vs. 35% and 19%, for conventional juveniles and transfers, respectively). If blended sentencing offers juvenile offenders a "last chance" at rehabilitation within the juvenile system by providing "an incentive to respond to treatment in order to avoid the consequences of an adult sentence" (Redding and Howell, 2000, p.147), it is apparently a last chance for white youth more so than minority youth.
Blended sentencing statutes are generally quite explicit about the factors that were intended by their respective legislatures to guide the selection of candidates for blended sentencing. In almost every case, these factors were the same as those used to guide selection of juvenile offenders for transfer to adult criminal court (Feld, 1995b; Redding & Howell, 2000), emphasizing the seriousness of the current offense and prior offense history. Though these statutes unequivocally identify the factors that juvenile judges are to use in their deliberations, they generally provide little guidance as to how the factors should be weighted and combined to collectively to determine the appropriate processing track. This lack of precision forces judges to rely on their own informal, subjective assessments of juvenile offenders' risk to public safety and their amenability to treatment in the juvenile justice system which opens the door to potential bias.
By providing the juvenile justice system with an intermediary response to juvenile offending that is not as punitive as the adult criminal justice system but which has more "teeth" than a conventional juvenile disposition, blended sentencing can become a cornerstone of a juvenile justice system that provides a "graduated" response to juvenile offending (National Criminal Justice Association, 1997). However, to be effective in this capacity, blended sentencing must be free from bias and used in a manner which is consistent with public safety. Our research, however, suggests that in states employing juvenile-inclusive blended sentencing, minorities will be disproportionately represented among transfers, the most punitive of the processing tracks, and disproportionately under-represented among blended sentences, the latter providing the last chance treatment in the juvenile justice system. Is there a solution to this problem that allows the juvenile justice system to "save this baby while throwing out the dishwater?"
The most promising solution to "rationalize" the use of blended sentencing and to avoid disparities in its use is to incorporate the principals of "risk and needs" in its application. A growing number of experts have advocated the incorporation of the risk principal throughout the criminal and juvenile justice systems as a means to rationalize decision-making and increase the effectiveness (e.g., Warren, 2007). Importantly, objective (15) risk assessment can reduce or eliminate undesirable bias in decision-making (Coordinating Council on Juvenile Justice and Delinquency Prevention, 1996).
Offender "needs assessments" can also inform the decision-making process by assessing the offender's "criminogenic needs" (16) (Andrews, Bonta, and Hoge, 1990), which have direct bearing on determinations of juvenile offenders' amenability to treatment in the juvenile justice system. Knowledge of the offender's criminogenic needs can assist juvenile justice system officials to determine whether these needs are best met in the juvenile justice system or in the adult criminal justice system.
Juvenile Judges are currently making determinations as to offender needs and risk on an informal basis, but formal risk and needs assessment procedures can improve the validity and fairness of such determinations (Silver and Chow-Martin, 2006). Consequently, our principal recommendation is that objective risk and needs assessments be used to identify the most suitable candidates for blended sentences and adult transfer. The recommendations generated by the risk and needs assessments need not be binding on the juvenile court but will provide valuable information to better inform the decision-making process.
Under this scenario, juvenile court judges would be provided with information about offender risk and needs prior to adjudication. Adult transfer would be reserved for a few of the most serious juvenile offenders that present the greatest risk to society and who are least amenable to treatment in the juvenile justice system, as informed by an objective risk/needs assessment. Blended sentences would be restricted to older juvenile offenders ( on average, not as old as transfers but older than most conventional juvenile court cases) who present less of a risk to public safety than transfers and who have the greatest need for and the greatest potential to respond to treatment in the juvenile justice system. Use of blended sentencing should be expanded to avoid the transfer of inappropriate juvenile offenders to the adult criminal justice system, keeping more juvenile offenders in the juvenile justice system while also holding them accountable. The use of risk and needs assessments to determine the appropriate processing track for each juvenile offender will reduce the odds that extralegal factors such as race will influence their selection.
The state-of-the- art of juvenile risk and needs assessment supports this recommendation but with caveats. In their comprehensive review of juvenile risk and needs assessment instruments, Vincent et. al. found that only two juvenile assessment systems, Structured Assessment of Violent Risk in Youth (SAVRY) and the Youth Level of Service/Case Management Inventory (YLS/CMI) meet the minimum standard that they established for what should be considered an evidence based risk assessment tool. (17) A few other tools came close to meeting their standard, and would therefore, be considered promising, including the Washington State Juvenile Court Assessment/Youth Assessment and Screening Instrument (WSJCA/YASI), Early Assessment Risk List for Boys (EARL-20B),Risk and Resiliency Checkup (RRC), Estimate of Risk of Adolescent Sexual Offense Recidivism (ERASOR), and the Juvenile Sex Offender Assessment Protocol (J-SOAP-II).
A very promising development along these lines is the development of the Ohio Youth Assessment System or OYAS (Latessa, Lovins, and Ostrowski, 2009). Developed by the Center for Criminal Justice Research at the University of Cincinnati, this system is designed to provide assessment information at five different stages of the juvenile justice system, including and especially significant for this paper, the disposition stage. The OYAS-DIS tool was shown to be effective in discriminating between low, moderate, and high risk (of re-arrest) male and female juvenile offenders. As research on this instrument accumulates, it may prove to be the very instrument for which we are calling.
However, it should be noted that at this point, there have not been any randomized studies showing that implementing risk assessment tools in juvenile justice has an impact on recidivism or appropriate intervention planning. Research for the youth in the juvenile justice system lags far behind that for adults on this subject; studies of the LSI-R with adult offenders have indicated a reduction in recidivism after implementation of the instrument (Flores, Lowenkamp, Holsinger, & Latessa, 2006). The expectation is that research will eventually confirm similar results for the juvenile justice system.
It should also be recognized that risk and needs assessments are not necessarily a panacea to racial bias in the selection of processing tracks. We endorse Vincent et. al.'s criteria that the validity of risk and needs assessment instruments should be examined by gender and race/ethnicity to prevent the introduction of bias through the instruments themselves. Further, assessment instruments are only as good as the information that is used to score them so efforts must be taken to ensure the quality of this information. Finally, it is important that staff receives adequate training in the use of these instruments and that they receive periodic "refreshers" of training on an ongoing basis. Scoring should also be monitored by periodic inter-rater tests of reliability.
Consequently, the next steps in this line of research would be to identify the factors that can be used to select the best candidates for transfer and blended sentencing on the basis of minimizing risk to public safety (18) and identifying those juvenile offenders most amenable to treatment in the juvenile justice system. Once these factors have been identified, objective risk/needs assessment instruments can be designed and field-tested, with the explicit purpose of identifying the most appropriate candidates for transfer and also for blended sentencing.
Our second recommendation is to provide enhanced services and supervision to juvenile offenders given blended sentences. Given these juvenile offenders are potentially subject to adult penalties (in addition to whatever requirements are imposed by the juvenile court) and that they have been found by the juvenile court to be amenable to treatment in the juvenile justice system, it follows that they should receive enhanced services, above and beyond those received by conventional juvenile offenders, designed to reduce their probability of re-offending. On the other hand, they likely present a higher risk to public safety than conventional juvenile offenders and consequently will require higher levels of supervision. The risk and needs assessments that are recommended could be used to identify an appropriate level of supervision and also identify services that could be used to address the offender's criminogenic needs. As Vincent, Terry, and Maney (2009) point out, "Arguably, the most dangerous youths should receive the most punitive sanctions and the most intensive interventions." (p. 388).
The best strategy to avoid or delay future reoffending is to address the offenders' criminogenic needs with services that are provided in sufficient dosage as to provide the offenders with a realistic opportunity to change their behavior (see, e.g., Andrews, Bonta, and Wormith, 2006). The level of services and supervision received by juvenile offenders given blended sentences in Ohio is currently a local decision but state-wide standards and additional resources made available by the state to local juvenile courts could provide the basis for the fair and effective distribution of enhanced services for juvenile offenders given blended sentences.
Blended sentencing can be important component of a system of graduated sanctions for juvenile offenders. However, in the states that currently permit blended sentences, it has not provided relief for the larger, somewhat intractable, issues that led to its creation. The use of blended sentencing in states that have implemented it has been very modest, not enough to reduce rates of transfer to the adult criminal justice system. Partly in consequence, juvenile offenders that have the potential to succeed with blended sentences or even less restrictive dispositions, are transferred to the adult criminal justice system in unacceptably high numbers and this true for minority juvenile offenders in particular. Provision of risk and needs assessment information to juvenile court judges has the potential to help them make better decisions about appropriate processing tracks for juvenile offenders and should encourage the use of blended sentencing for appropriate juvenile offenders.
Andrews, D. A., Bonta, J., & Hoge, R. D. (1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52.
Andrews, D. A., Bonta, J., & Wormith, J. (2006). The recent past and near future of risk and/or need assessment. Crime and Delinquency, 52, 7-27.
Bishop, D., and Frazier, C. (1991). Transfer of juveniles to Criminal Court: A case study and analysis of prosecutorial waiver. Notre Dame Journal of Law, Ethics, and Public Policy, 5, 281-302.
Bishop, D., Frazier C., and Lanza-Kaduce, L. (1996). The transfer of juveniles to Criminal Court: Does it make a difference. Crime and Delinquency, 42, 171-191.
Champion, D. and Mays, G. (1991). Transferring juveniles to criminal courts: Trends and implications for criminal justice. New York: Praeger.
Cheesman, F., Green, H., Cohen, T., Dancy, D., Kleiman, M., and Mott, N. (2002). Blended sentencing in Minnesota: On target for justice and public safety? An evaluation. Williamsburg, Virginia: National Center for State Courts.
Cheesman, F., Waters, N., Hurst, H., Halemba, G., Maggard, S., and Sohoni, T. (2010). Who gets a second chance? An evaluation of blended sentencing in Ohio and Vermont. Williamsburg, Virginia: National Center for State Courts.
Coordinating Council on Juvenile Justice and Delinquency Prevention. (1996). Combating violence and delinquency: The National Juvenile Justice Action Plan Report. Washington DC: Author.
Demuth, S. and Brown, S. (2004). Family structure, family processes, and adolescent delinquency: The significance of parental absence versus parental gender. Journal of Research in Crime and Delinquency, 41(1), 58-81.
Feld, B. (1995a). The Social context of Juvenile Justice Administration: Racial Disparities in an Urban Juvenile Court. In K. Leonard, C. Pope, and W. Feyerherm (Eds.), Minorities in Juvenile Justice. Thousand Oaks, CA: Sage Publications.
Feld, B. (1995b). Violent youth and public policy: A case study of juvenile justice law reform. Minnesota Law Review, 79, 965-1128.
Flores, A., Lowenkamp, C., Holsinger, A., & Latessa, E. (2006). Predicting outcome with the Level of Service Inventory-Revised: The importance of implementation integrity. Journal of Criminal Justice, 34, 523-529
Frazier, C. and Bishop, D. (1985). The pre-trial detention of juveniles and its impact on case dispositions. Journal of Criminal Law and Criminology, 76, 1132-1152.
Hawkins, J.D., Herrenkohl, T.I., Farrington, D.P., Brewer, D., Catalano, R.F., Harachi, T. W. & Cothern, L. (April 2000). Predictors of youth violence. Washington DC: Office of Juvenile Justice and Delinquency Prevention.
Latessa, E., Lovins, B., and Ostrowski, K. (2009). The Ohio Youth Assessment System. Cincinnati, Ohio: Center for Criminal Justice Research, University of Cincinnati.
Loeber, R., and Farrington, D., and Waschbusch, D. (1998). Serious and violent juvenile offenders. In R. Loeber and D. Farrington, D. (Eds.). Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions. Thousand Oaks, CA: Sage Publications.
McNulty, E. 1996. The Transfer of Juveniles to Adult Court: Panacea or Problem. Law and Policy, 18, 61-75.
Males, M. and Macallair, D. (2000). The color of justice. Washington DC: Building Blocks for Justice.
National Criminal Justice Association. (1997). Juvenile justice reform initiatives in the states: 1994-1996. Washington DC: Office of Juvenile Justice and Delinquency Prevention.
Patterson, G., Forgatch, M., Yoerger, K., and Stoolmiller, M. (1998). Variables that initiate and maintain an early-onset trajectory for juvenile offending. Development and Psychopathology, 10, 531 -547.
Poulos, T. and Orchowsky, S. (1994). Serious juvenile offenders: Predicting the probability of transfer to criminal court. Crime and Delinquency, 40, 3-17.
Podkopacz, M. and Feld, B. (2001). The back-door to prison: waiver reform, "blended sentencing," and the law of unintended consequences. The Journal of Criminal Law and Criminology, 91, 997-1072.
Poe-Yamagata, E. and M. A. Jones (2000). And justice for some. Washington DC: National Council on Crime and Delinquency.
Redding, R. and Howell, J. (2000). Blended sentencing in American juvenile courts. In J. Fagan and F. Zimring (Eds.) The Changing Borders of Juvenile Justice, Chicago: The University of Chicago Press.
Sickmund, M., Snyder, H., Poe-Yamagata, E. (1997). Juvenile offenders and victims: 1997 Update on violence. Washington, DC: Office of Juvenile Justice and Delinquency Prevention.
Silver, E. and Chow-Martin, L. (2002). A multiple models approach to assessing recidivism risk: Implications for judicial decision making. Criminal Justice and Behavior, 29, 538-568.
Torbet, P, Gable, R., Hearst, H., Montgomery, I., Szymanski, L., and Thomas, D. (1996). State responses to serious and violent juvenile crime. Washington, D.C.: National Center for Juvenile Justice.
Torbet, P., Griffin P., Hurst, H., and MacKenzie, L. (2000). Juveniles facing criminal sanctions: Three states that changed the rules. Washington, D.C.: Office of Juvenile Justice and Delinquency Prevention.
Vincent, G., Terry, A., and Maney, S. Risk/Needs Tools for antisocial behavior and violence among youthful populations. In J. Andrade (Ed.) Handbook of Violence Risk Assessment and Treatment: New Approaches for Mental Health Professionals. New York: Springer Publishing Company.
Warren, R. (2007). Evidence-based practice to reduce recidivism: Implications for state judiciaries. Washington DC: The Crime and Justice Institute and National Institute of Corrections, Community Corrections Division.
Wasserman, G., Keenan, K., Tremblay, R., Coie, J., Herrenkohl, T., Loeber, R., and Ptechuk, D. Risk and Protective Factors of Child Delinquency. Child Delinquency Bulletin Series, (April 2003, 1-15). Washington, DC: Office of Juvenile Justice and Delinquency Prevention.
FRED L. CHEESMAN, II
NICOLE L. WATERS
National Center for State Courts
HUNTER HURST IV
(1) Funding for the project was provided by NIJ under grant number 2006-NIJ- 1162.
(2) Between 1992 and 1995, 41 states changed their laws to make waiver to adult court easier, 16 states modified or added statutes requiring mandatory minimum periods of
incarceration for certain violent or serious offenders, and 12 states extended the maximum age of the juvenile court's continuing jurisdiction over juvenile offenders-most often to age 21 (Sickmund, Snyder, & Poe-Yamagata, 1997).
(3) That is, the randomly selected sample of conventional juvenile offenders was proportionately distributed among the five counties according to the proportion that each county represented of the total 2002-2004 adjudications.
(4) Post-hoc analyses indicated that the average ages of all groups were significantly different.
(5) To assess whether the differences in the racial composition of the juvenile tracks were statistically significant, the minority categories (Black, Latino, and Asian) were first collapsed into a single category, "Minority." Small numbers of juveniles fell into the Asian and Latino categories (1 and 17, respectively).
(6) The remaining counties provided limited data for this variable and were consequently excluded from this analysis. Data were missing for 64% of the sample, and entirely missing for sample members from Hamilton and Cuyahoga Counties. The majority of cases from Lucas, Summit, and Delaware Counties provided data for this variable although each county reported substantial amounts of missing data (33.3%, 18.4%, and 19.4%, respectively).
(7) Data on juveniles' family structure were missing for Lucas and Cuyahoga Counties.
(8) Delaware County also provided data on adult co-defendants, but none of the juveniles were charged with an adult co-defendant. Two Delaware County juveniles were charged with another juvenile co-defendant, one received a SYO and one was adjudicated conventionally. Missing data from Hamilton and Cuyahoga on co-defendants inhibited the ability to conduct analyses in this section.
(9) Data on gang-related offenses for Cuyahoga and Hamilton were not available. In Lucas and Delaware County, gang-related charges were not prevalent (only 5 of 107 juveniles had a gang-related offense).
(10) Cuyahoga provided data on whether the juvenile had previously been in a secure detention, but could not ascertain whether the detention was in DYS. Therefore, Cuyahoga's data are not included in this analysis.
(11) Nearly half of the data on detention prior to adjudication was missing in Cuyahoga County. No further analyses were run for this county.
(12) Delaware did not process adequate numbers of juveniles for significance testing (N=29 with data).
(13) Motioning data for Cuyahoga and Lucas Counties was not available.
(14) In order to avoid the sample selection problem, and to get asymptotically efficient estimators, the model parameters are estimated by maximum likelihood.
(15) Objective risk assessment instruments were created to minimize subjectivity and unreliability associated with clinical decision-making. Objective tools evaluate all offenders using the same set of criteria, using information that can be factually verified. The results are then tabulated in some fashion and pre-determined uniform decision functions such as cutting scores or decision trees decide the outcome.
(16) Criminogenic needs are attributes of offenders that are directly linked to criminal behavior. Effective correctional treatment should target criminogenic needs in the development of a comprehensive case plan.
(17) Their standard is based on the following considerations:
* A manual that contains scoring criteria and/or detailed item descriptions to structure administration
* Empirically-based risk factors
* Known internal structure: The tool should have some reported scale analyses to justify the scoring of the instrument. Ideally, this would be a factor analysis through, factor analyses will not apply to all types of risk assessment instruments.
** At least one test, but preferably two, of reliability that was conducted by an independent party (not the test developer).
** In the case that the tool is self-report only (that is, the tool does not rely on examiner ratings), reliability is assessed by examining internal consistency and test-retest reliability.
** For tools that rely on examiner ratings, evidence for inter-rater reliability is required. The preferred measure of reliability in this case is intra-class correlation coefficients (ICCs). ICCs should be at least above 0.70, and preferably above 0.90.
* Validity: Evidence that the instrument predicts recidivism and /or violence.
** At least one study, preferably two, by an independent party demonstrating good predictive validity (medium to large effects) in a juvenile justice setting
** Examine differences in validity by gender and race/ethnicity
** Prospective studies of the tool's validity for predicting recidivism or antisocial behavior
(18) Risk and needs assessments enhance public safety by identifying offenders most likely to reoffend who can then receive the enhanced supervision and services (designed to reduce the probability of re-offending) associated with transfer and blended sentences.
Table 1 Varieties of Blended Sentencing Used Across States In Type Description Adopted by Year Juvenile- The juvenile court New Mexico 1995 Exclusive imposes either juvenile Blend (delinquency) or adult (criminal) sanctions. Juvenile- The juvenile court Illinois, Kansas 1990 Inclusive Blend imposes both juvenile and adult sanctions, South Dakota 1993 typically suspending the adult sanction. Minnesota 1994 Alaska, Arkansas, Connecticut 1995 Michigan, Montana, 1997 Vermont Ohio 2002 Juvenile- The juvenile court Texas 1987 Contiguous imposes a juvenile sanction that would be Massachusetts, 1990 in force beyond the age Rhode Island of its extended jurisdiction. At that Colorado 1993 point, the juvenile court determines whether South Carolina 1994 the remainder of that sanction should be served in an adult criminal corrections system. Criminal- The criminal court Virginia, West 1985 Exclusive Blend imposes either juvenile Virginia or criminal sanctions. Colorado 1993 Florida 1994 California, Idaho 1995 Michigan 1997 Oklahoma 1998 Criminal- The criminal court Virginia 1985 Inclusive Blend imposes both juvenile and criminal sanctions, Florida 1994 typically suspending the criminal sanction. Arkansas, Michigan 1995 Iowa 1997 Table 2. Number of Juveniles in Sample, by County and Processing Track Processing Track County Conventional Transfer SYO Total N % N % N % N % Hamilton 76 22 79 48 13 9 168 26 Cuyahoga 105 31 32 20 28 20 165 26 Lucas 74 22 18 11 46 33 138 22 Summit 64 19 34 21 38 27 136 21 Delaware 21 6 1 1 14 10 36 6 Total 100 164 100 139 100 643 100 Table 3: Maximum Likelihood Probit Model Predicting Transfer, with Selection Robust Coeff. SE z p>z Selection Equation: Transfers and SYOs (=1) vs. Conventional Juvenile (=0) Charged w/ Murder? (Y=1) 1.624 0.527 3.08 0.002 Charged w/ 1st Degree Felony? (Y=1) 1.281 0.177 7.230 0.000 Charged w/ 2nd Degree Felony? (Y=1) 0.421 0.126 3.340 0.001 Number of Charges (1, 2, 3, 4 or +) 0.107 0.052 2.080 0.038 Age at Filing (12-19) 0.245 0.054 4.560 0.000 Gender (Male=1) 0.462 0.190 2.440 0.015 Race (Non-White=1) -0.051 0.127 -0.400 0.691 # of Prior DYS Placements (0-9) 0.306 0.062 4.930 0.000 # of Non-DYS Out-of-Home Placements (0-5) 0.070 0.096 0.730 0.466 # of Prior Felonies (0-28) -0.007 0.016 -0.400 0.687 Constant -6.750 0.903 -7.470 0.000 Probit: Transfer (=1) vs. SYO (=0) Charged w/ Murder? (Y=1) 1.385 1.194 1.160 0.246 Charged w/ 1st Degree Felony? (Y=1) -0.095 0.784 -0.120 0.903 Number of Charges (1, 2, 3, 4 or +) 0.107 0.101 1.060 0.290 Age at Filing (12-19) 0.397 0.200 1.990 0.047 Gender (Male=1) 1.635 0.611 2.680 0.007 Race (Non-White=1) 0.735 0.201 3.650 0.000 # of Prior DYS Placements (0-9) 0.046 0.190 0.240 0.807 # of Non-DYS Out-of-Home Placements (0-5) 0.050 0.111 0.450 0.650 # of Prior Felonies (0-28) 0.019 0.023 0.830 0.407 Constant -8.823 5.615 -1.570 0.116 Alternative Rho -0.066 0.828 -0.080 0.936 Rho -0.066 0.825 -- -- 95% C. I. Selection Equation: Transfers and SYOs (=1) vs. Conventional Juvenile (=0) Charged w/ Murder? (Y=1) 0.592 2.656 Charged w/ 1st Degree Felony? (Y=1) 0.934 1.628 Charged w/ 2nd Degree Felony? (Y=1) 0.174 0.668 Number of Charges (1, 2, 3, 4 or +) 0.006 0.208 Age at Filing (12-19) 0.140 0.351 Gender (Male=1) 0.090 0.834 Race (Non-White=1) -0.300 0.199 # of Prior DYS Placements (0-9) 0.184 0.427 # of Non-DYS Out-of-Home Placements (0-5) -0.119 0.259 # of Prior Felonies (0-28) -0.038 0.025 Constant -8.521 -4.979 Probit: Transfer (=1) vs. SYO (=0) Charged w/ Murder? (Y=1) -0.955 3.724 Charged w/ 1st Degree Felony? (Y=1) -1.633 1.442 Number of Charges (1, 2, 3, 4 or +) -0.091 0.304 Age at Filing (12-19) 0.005 0.788 Gender (Male=1) 0.437 2.832 Race (Non-White=1) 0.340 1.130 # of Prior DYS Placements (0-9) -0.327 0.420 # of Non-DYS Out-of-Home Placements (0-5) -0.167 0.267 # of Prior Felonies (0-28) -0.026 0.065 Constant -19.829 2.183 Alternative Rho -1.690 1.557 Rho -0.934 0.915 Wald test of independent equations (Rho = 0): [Chi.sup.2](1) = 0.01 Prob > [Chi.sup.2] = 0.94 Note. Observations (N=573): Censored (N=313), Uncensored (N=260). Log pseudolikelihood = -1,401.17, Wald [Chi.sup.2] (9) = 40.28, Prob > [Chi.sup.2] = 0.00.
|Gale Copyright:||Copyright 2010 Gale, Cengage Learning. All rights reserved.|