Working upstream: A statewide analysis of individual and contextual risk factors for future juvenile justice involvement among youth receiving prevention services

• Published date: 01 May 2024
• Author: Keller G. Sheppard,Jorge L. Hernandez,Jovontae Butts,Orlando Mendonca,Julie C. Orange
• Date: 01 May 2024
• DOI: http://doi.org/10.1111/1745-9133.12656

DOI: ./- .

ORIGINAL ARTICLE

Working upstream: A statewide analysis of

individual and contextual risk factors for future

juvenile justice involvement among youth

receiving prevention services

Keller G. Sheppard1Jorge L. Hernandez2Jovontae Butts2

Orlando Mendonca2Julie C. Orange2

College of Criminology and Criminal

Justice, Florida State University,

Tallahassee, Florida, USA

Florida Department of Juvenile Justice,

Tallahassee, Florida, USA

Correspondence

Keller G. Sheppard, Florida State

University, Tallahassee,FL, USA.

Email: ksheppard@fsu.edu

Abstract

Research Summary: Juvenile justice agencies are

increasing their emphasis on early intervention and pre-

vention services as a growing body of research evinces

their effectiveness and cost efficiency. The presentstudy

analyzed the relationship between individual risk fac-

tors, contextual risk factors, and future juvenile justice

involvement for , Florida youth residing in 

census tracts served by prevention programs. A series of

two-level logistic regressions indicated that several dis-

tinct criminological domains (e.g., aggression, relation-

ships, family, substance use, and attitudes/behaviors)

predict future juvenile justice system involvement.How-

ever, education-related risk factors are among the most

consequential for all youth, especially older youth.

Concerning community context, neighborhood disad-

vantage directly affects system involvement, but only for

youth under .

Policy Implications: These findings indicate the

importance of addressing highly consequential risk

factors—especially educational deficits—of youth in

early intervention and prevention programs while also

recognizing the impact of their social environments.

Criminology & Public Policy. ;:–. ©  American Society of Criminology. 227wileyonlinelibrary.com/journal/capp

228 SHEPPARD .

Agencies attempting to work upstream with prevention

services may benefit from prioritizing educational ser-

vices and allocating resources to highly disadvantaged

communities for early intervention programming.

KEYWORDS

community context, early intervention, juvenile justice, risk

factors

The Risk–Need–Responsivity (RNR) model is among the most influential strategies in juvenile

justice for optimizing youth rehabilitation (Andrews & Bonta, ). Risk/need assessment instru-

ments are an integral component of the RNR model, and they facilitate systematic classifications

of risk and the identification of criminogenic needs. Determining individual needs has yielded

opportunities for substantial improvement of treatment efficacy, as prior scholarship indicates

that identifying juvenile justice-involved youth’s top criminogenic needs and delivering targeted,

need-aligned interventions improve treatment outcomes (Baglivio,Wolff, Howell, et al., ;Nel-

son & Vincent, ). However, there is a growing appreciation that the relative influence of risk

factors on delinquency varies, and agencies can optimize outcomes by prioritizing the treatment

of high-impact risk factors (Baglivio et al., ; Baglivio, Wolff, Piquero, et al., ; Hay et al.,

). While many juvenile justice agencies have adopted risk and need assessments to optimize

youth processing and facilitate case planning, there are further opportunities to improve system

efficacy by prioritizing interventions that target high-impact risk factors (Skeem et al., ).

Juvenile justice agencies have historically focused on providing treatment to youth after they

have committed a delinquent act, been arrested, and been referred to juvenile courts (Hay et al.,

; Loeber et al., ). As such, policy and research efforts to incorporate the RNR model

and address the most consequential risk factors—as identified by risk and need assessment

instruments—have focused on youth who have already enteredthe juvenile justice system. How-

ever, there are increasing efforts to broaden the scope of juvenile justice functions and actively

engage at-risk youth before formal system contact. Numerous early intervention and prevention

programs have been designed for and administered to younger children (i.e., under ) before

the onset of delinquency (Augimeri et al., ; Farrington & Welsh, ). Research on these

programs has subsequently demonstrated their effectiveness in reducing system contact in a

cost-efficient manner (Cohen et al., ;Loeberetal.,).

Juvenile justice agencies seeking to engage in proactive intervention strategies that target high-

impact risk factors (i.e., factors strongly associated with offending) are forced to draw insights

from a body of research with two key limitations. First, early intervention programming typically

serves considerably younger children than juveniles formally referredfor delinquency (Farrington

& Welsh, ). By virtue of their age, the characteristics and salience of variousrisk factors may

differ substantially from the comparatively older juvenile justice-involved population studies in

the extant literature. Indeed, research emerging from developmental criminology demonstrates

that age can moderate the impact of risk factors on delinquency (Tanner-Smith et al., ). Risk

factors identified as consequential for youth in late adolescence are dissimilar from those salient

for younger children in earlier developmental stages and responsive to distinct environmental

influences.

SHEPPARD . 229

Second, prior research on risk/needs assessment-derived individual risk factors has bene-

fited from including youth’s broader social ecology. For example, adverse community conditions’

impact on delinquency is well established in the criminological literature (Antunes & Manasse,

; Kubrin & Weitzer, ; Messner & Zimmerman, ; Shaw & McKay, ). A juvenile’s

ecological context has profound implications for exposure to various risk and protective factors

that may not be fully reflected by risk/needs assessments (Baglivio et al., ; Bronfenbrenner,

; Onifade et al., ). Consideration of these would inform not only the types of services with

the most significant potential impact on youth but also identify the contextual risk factors that

could be mitigated by preventative programming.

Motivated by these limitations, the present study assesses the individual and contextual risk

factors associated with future juvenile justice involvement among Florida youth participating in

prevention programs. Drawing on data from the Florida Department of Juvenile Justice (DJJ),

multi-level analyses explore the most consequential risk factors for future involvement for ,

youth residing in  census tracts. Consistent with a developmental perspective, analyses

explored how these risk factors differentially predicted offending across age groups. The study

begins with an overview of individual and contextual risk factors associated with delinquency

and the unique characteristics of prevention youth and services.

1LITERATURE REVIEW

1.1 Criminogenic risk: Distinguishing individual factors and

quantifying risk

Decades of criminological research have identified an array of risk and protective factors associ-

ated with delinquency and future involvement in the criminal justice system (Farrington et al.,

; Tanner-Smith et al., ). There is no single path to delinquency, as numerous theoretical

models—and their associated risk factors—have received empirical support. Youth’s demograph-

ics, personal traits, family characteristics, peer associations, and community contextscan combine

to exert multiplicative and interactive effects that shape their propensity toward delinquency

(Case and Haines, ). Longitudinal studies investigating multiple risk factors have been

instrumental in establishing causal relationships and quantifying the relative strength of their

associations. A comprehensive review of all known risk factors and their relative effect sizes (see

Farrington et al. () for an overview of systematic reviews and meta-analyses) is beyond the

scope of this study.Instead, this review will focus on research exploring the individual risk factors

as commonly operationalized and classified for use in juvenile justice systems.

Rooted in the RNR model (Bonta & Andrews, ), risk assessment instruments that utilize

various factors to assess overall risk have proliferated in juvenile justice systems. These instru-

ments draw upon the rich body of criminological research referenced above to identify discrete

types or domains of criminogenic needs, which can, in turn, be matched to interventions target-

ing those needs (Brogan et al., ). Andrews and Bonta () identified eight factors—termed

the Central Eight Risk Factors—that have robust statistical associations with recidivism: crimi-

nal history, pro-criminal attitudes, antisocial personality pattern, pro-criminal associates, family

circumstances, education/employment, leisure/recreation, and substance abuse. Many modern

risk assessment instruments incorporate many of these factors, albeit with distinct operational-

izations and frequent inclusion of additional factors (Baird et al., ). Subsequent evaluations

of risk assessment instruments’ efficacy typically examine the association of overall risk scores