A Multilevel Examination of Risk/Need Change Scores, Community Context, and Successful Reentry of Committed Juvenile Offenders

• Author: Katherine Jackowski,Kevin T. Wolff,Michael T. Baglivio,Mark A. Greenwald
• DOI: 10.1177/1541204015596052
• Published date: 01 January 2017
• Date: 01 January 2017
• Subject Matter: Articles

Article

A Multilevel Examination of

Risk/Need Change Scores,

Community Context, and

Successful Reentry of

Committed Juvenile Offenders

Michael T. Baglivio

1

, Kevin T. Wolff

2

,

Katherine Jackowski

1

, and Mark A. Greenwald

3

Abstract

Current theory and practice dictates the use of risk/needs assessment to guide programming to

reduce reoffending. Limited research has examined assessment change scores and recidivism, none

examining whether such changes moderate the effects of deleterious community contexts. We

examine a multiyear statewide sample of juvenile offenders returning to the community from

residential placement (N¼12,302). We address whether changes in dynamic risk/needs scores

predict official recidivism upon return, community socioeconomic contexts predict recidivism, and

which risk/needs changes moderate the effects of context. Findings reveal 6 of the 17 change scores

affect reoffending, context matters, and some change scores moderate contextual effects.

Keywords

juvenile offenders, recidivism, multilevel, change scores, risk assessment

Nationwide, approximately 42,000 juveniles were in residential commitment placement for a delin-

quent or status offense on a given day in 2011 (Sickmund, Sladky, Kang, & Puzzanchera, 2013). In

Florida alone, 3,862 youth were released from residential programs from July 1, 2012, to June 30,

2013 (Florida Department of Juvenile Justice [FDJJ], 2015). Yet, little empirical research has exam-

ined youth and community context characteristics most likely to lead to successful reentry. Cer-

tainly, there has been a proliferation of risk assessment tools and subsequent validation studies

(Schwalbe, 2008; Yang, Wong, & Coid, 2010). Theory and practice has informed the results of

risk/needs assessments are intended to be used in decision making of where to place juvenile

1

G4S Youth Services, LLC, Tampa, FL, USA

2

John Jay College of Criminal Justice, City University of New York, New York, NY, USA

3

Florida Department of Juvenile Justice, Tallahassee, FL, USA

Corresponding Author:

Michael T. Baglivio, G4S Youth Services, LLC, 6302 Benjamin Rd., Suite 400, Tampa, FL 33625, USA.

Email: michael.baglivio@us.g4s.com

Youth Violence and JuvenileJustice

2017, Vol. 15(1) 38-61

ªThe Author(s) 2015

Reprints and permission:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1541204015596052

yvj.sagepub.com

offenders, in terms of restrictiveness level of the system, how to manage and supervise offenders,

and what services would be appropriate during placement (Andrews & Bonta, 2003; Gottfredson

& Moriarty, 2006; Howell, Lipsey, & Wilson, 2014). However, studies examining the use of risk/

needs assessments to measure progress of juvenile offenders based on changes in assessed risk/needs

and subsequent recidivism simply do not exist. Furthermore, there have been no multilevel analyses

of risk assessment change scores that examine the context of where committed youth return and

whether inducing change in risk levels during placement may moderate any impact of community

conditions on reentry. This study represents an initial attempt to fill those gaps.

First, we examine the ability of changes in dynamic risk/criminogenic needs to predict recidivism

upon juvenile offenders’ return to the community. We then assess whether characteristics of the

community to which a youth returns affect successful reentry. Finally, we examine whether changes

in criminogenic needs during residential placement moderate any significant effect of contextual

measures, examining whether such changes can suppress a negative contextual effect or increase

a positive, for example. As such, this article is organized into the following sections: First, we briefly

review the limited literature on risk assessment change scores as well as the relevant prior work on

community context and criminal behavior. Next, we discuss the data and methods used in the current

study, followed by the results of the current analysis in the order described above. Finally, we end

with conclusions and the policy implications stemming from the current research.

Risk Assessment Change Scores

The risk-needs-responsivity (RNR) model has become a dominant paradigm of corrections and juve-

nile justice over the last few decades (Andrews, Bonta, & Wormith, 2011; Howell et al., 2014;

McGrath & Thompson, 2012; Peterson-Badali, Skilling, & Haqanee, 2015; Vose, Lowenkamp,

Smith, & Cullen, 2009). The synthesis of validated risk assessment (risk), targeting dynamic risk

factors predictive of offending (criminogenic needs), and using interventions and strategies proven

effective, but individualized in their delivery (responsivity), forms the foundation of the model

(Andrews & Bonta, 2010). Through this process, in theory, risk is reduced, strengths/protective fac-

tors enhanced, and the likelihood of recidivism diminished. The notion of criminogenic needs is

essential in that they are dynamic targets for change that are empirically associated with recidivism

(Flores, Lowenkamp, Smith, & Latessa, 2006; Lowenkamp & Bechtel, 2007; Simourd, 2004). How-

ever, few studies have examined whether risk assessment scores change over time nor whether such

changes actually increase or decrease future offending (with almost no research on changes in risk

assessment scores as the result of targeted intervention and subsequent reoffending). The majority of

notable exceptions have examined changes in adult probationers using the Level of Service Inven-

tory–Revised (LSI-R; Raynor, 2007; Schlager & Pacheco, 2011; Vose et al., 2009).

Raynor (2007), using the LSI-R, examined changes in the composite score and subsequent offi-

cial offending in a sample of 360 adult probationers. Those probationers with an increase in com-

posite LSI-R score evidenced a higher likelihood of new offending in comparison to those who

evidenced a decreased composite LSI-R score. Vose, Lowenkamp, Smith, and Cullen (2009) used

the LSI-R assessment composite scores of 2,849 male and female adult probationers at two points in

time to assess whether each assessment, as well as the change scores from Time 1 to Time 2, pre-

dicted recidivism. Results showed the percentage change in LSI-R from Time 1 to Time 2 was a

significant predictor of reoffending, controlling for initial risk and other pertinent factors, for both

males and females. These studies were valuable in informing practice in that they demonstrate the

total/overall score of a risk assessment can change over time and that those percentage changes are

predictive of future offending, at least for adult probationers. While tying change scores to out-

comes, these studies examined only the composite score leaving questions of which needs to target

for the best chances of success unanswered.

Baglivio et al. 39