THE ROLE OF NEIGHBORHOOD CONTEXT IN YOUTH CO‐OFFENDING

• Date: 01 February 2014
• Author: SCOTT H. DECKER,DAVID R. SCHAEFER,NANCY RODRIGUEZ
• Published date: 01 February 2014
• DOI: http://doi.org/10.1111/1745-9125.12032

THE ROLE OF NEIGHBORHOOD CONTEXT

IN YOUTH CO-OFFENDING

DAVID R. SCHAEFER,1,* NANCY RODRIGUEZ,2

and SCOTT H. DECKER2

1School of Human Evolution and Social Change, Arizona State University

2School of Criminology and Criminal Justice, Arizona State University

KEYWORDS: juvenile delinquency, co-offending, neighborhoods, social networks,

trust

Despite co-offending being a core criminological fact, locating suitable peers has

many challenges. Chief among these, given the risky nature of co-offending, is finding

trustworthy accomplices. We propose that neighborhoods serve as youths’ most ready

source of accomplices, and as such, their composition affects the likelihood of iden-

tifying suitable co-offenders. In particular, youth are more likely to co-offend in con-

texts with more peers of their race/ethnicity, less disadvantage, and greater residential

stability—all of which promote trust among neighbors. We test our hypotheses using

multilevel models applied to census data and official court records for 7,484 delinquent

youth in a large metropolitan area. The results offer support for our hypotheses and

provide greater insight into how individual and contextual factors combine to affect

co-offending behavior. An implication of these findings is that many of the same neigh-

borhood characteristics that reduce crime lead to a greater proportion of co-offending.

“I don’t have friends. Just associates. Friends you trust.” (Lafayette Rivers, quoted in

Kotlowitz [1992: 55])

Co-offending is a core criminological fact (Reiss, 1988; Smangs, 2010). Although more

than half of all offending occurs in groups (Felson, 2003; Reiss, 1988), the group context

for delinquency is even higher, with some estimates of delinquency occurring in groups as

high as 90 percent (Klein and Crawford, 1967; Reiss, 1988; Warr, 2002). Beyond its sheer

prevalence, co-offending is important to study given its implications for delinquency onset

and criminal trajectories. In many cases, co-offending plays a role in recruiting others into

criminal behavior who may not have participated in crime otherwise (Felson, 2003; Reiss,

1988; Warr, 1996). Co-offending also can increase the frequency, seriousness, and dura-

tion of criminal careers (Alarid, Burton, and Hochstetler, 2009; Bouchard and Ngyuen,

2010; Carrington, 2009; Felson, 2003; McCord and Conway, 2002; McGloin and Piquero,

2009). Moreover, co-offending complicates attempts to estimate changes in crime rates

achievable through the removal of offenders via incarceration (Andresen and Felson,

2010; Zimring, 1981). In this sense, the study of co-offending has broad social implications

and influence on a wide range of sociological/criminological theories (Sarnecki, 2001;

∗Direct correspondence to David R. Schaefer, School of Human Evolution and Social Change,

Arizona State University, P.O. Box 873701, Tempe, AZ 85287 (e-mail: david.schaefer@asu.edu).

C2013 American Society of Criminology doi: 10.1111/1745-9125.12032

CRIMINOLOGY Volume 52 Number 1 117–139 2014 117

118 SCHAEFER, RODRIGUEZ, & DECKER

Smangs, 2010). These issues call for developing a deeper understanding of co-offending

that can inform criminological theory and practice (McGloin and Nguyen, 2013).

Co-offending departs from solo-offending through more than simply the addition of

one or more offenders. The shift to co-offending transforms the nature of delinquency

into a social act requiring cooperation and making group processes possible. However,

even with willing co-offenders, cooperation is not a given. Offending entails several forms

of risk, many of which are exacerbated when others are involved. Consequently, trust

between co-offenders becomes an especially crucial ingredient (McCarthy, Hagan, and

Cohen, 1998). But how do potential co-offenders come to trust one another enough to

act together? Answering this question requires expanding the focus beyond individual

attributes to a broader set of causal factors.

We propose that the neighborhoods in which youth reside can either facilitate or hinder

youth co-offending. Neighborhood factors have long been argued to help explain crime

and juvenile delinquency (Shaw and McKay, 1942). More recent research has found that

co-offending rates vary across neighborhoods (Daly, 2005), although the causal mecha-

nisms have not been articulated. Our motivation to focus on neighborhoods is twofold.

First, neighborhoods provide a pool of proximate others from which co-offenders are

drawn. Second, neighborhoods determine, in part, how well youth can find “suitable” co-

offenders (Tremblay, 1993) who can be trusted (McCarthy, Hagan, and Cohen, 1998).

We propose that neighborhood characteristics that promote trust among residents more

generally also will promote trust, and subsequently co-offending, among youth. This line

of reasoning leads to a seeming contradiction: The positive effect of neighborhood fac-

tors (e.g., disadvantage and mobility) on crime is the opposite of their hypothesized effect

on co-offending. Our argument parallels that of D’Alessio and Stolzenberg (2010), who

found that urbanization inhibits co-offending. We place our empirical focus on youth co-

offending and neighborhood context.

CURRENT STUDY

We investigate several neighborhood factors (i.e., racial/ethnic similarity, disadvantage,

and instability) that affect the likelihood of co-offending. As von Lampe (2007) observed,

understanding the context in which criminal networks develop and function is a key item

on the co-offending research agenda. This focus serves to differentiate our approach

from most other work on co-offending as we attempt to account for the likelihood of

co-offending in a geographic context and merge two units of analysis, networks and geog-

raphy (Tita and Radil, 2011). To do so, we develop three hypotheses drawn from social

network theory, social psychology, and the neighborhoods and crime literature. We test

these hypotheses using multilevel models applied to census data and delinquency records

from a large metropolitan area. Thus, we evaluate the effect of neighborhood factors net

of controls for youth sociodemographic indicators that have been the focus of prior inves-

tigations. The results offer support for a structural theory of co-offending that emphasizes

the linkage between neighborhoods and group processes.

THEORETICAL BACKGROUND

It has long been recognized that it is impossible to understand delinquency with-

out understanding co-offending. Much research has investigated the correlates and