Does Juvenile Delinquency Reduce Educational Attainment?

• Date: 01 December 2015
• Published date: 01 December 2015
• DOI: http://doi.org/10.1111/jels.12090

Does Juvenile Delinquency Reduce

Educational Attainment?

Shannon Ward and Jenny Williams*

This article investigates the effect of delinquency in youth on subsequent educational

attainment. To do so, we focus on delinquent acts committed by age 16 and examine their

impact on two measures of educational attainment: high school graduation and college

graduation. Using information on males from the extremely rich National Longitudinal

Survey of Youth 1997, we find plausible evidence that delinquency by age 16 reduces the

likelihoodof graduating from high school and college.This effect is driven by earlyinitiators,

those who offend intensely, and by those whose delinquent activities involve income-

generatingacts. Importantly,the impact of delinquencyon education is not confinedto those

who have interaction with the criminal justice system, or gang members. Further analysis

suggests thata channel through which delinquency impactseducation is expected returns to

crime, asreflected by subjectivebeliefs about the probability of arrestfor a property crime.

I. Introduction

Education is a key determinant of economic success. It is also an important determinant

of wellbeing more broadly defined, as demonstrated by studies showing that poor educa-

tional attainment is causally related to crime and incarceration in the adult population

(Lochner & Moretti 2004; Oreopoulos & Salvanes 2009; Merlo & Wolpin 2008; Buo-

nanno & Leonida 2006; Machin et al. 2011; Meghir et al. 2012).

1

This suggests that an

avenue through which crime, incarceration, and the social disadvantage they bring may

be reduced is through improving educational outcomes.

2

Doing so is, however,

*Address correspondence to Jenny Williams, Department of Economics, FBE Building, University of Melbourne,

3010 Victoria, Australia; email: jenny.williams@unimelb.edu.au. Ward is at the University of Melbourne. Williams

is Professor of Economics, University of Melbourne.

The authors acknowledge the Faculty of Commerce and Business, University of Melbourne for supporting this

research. They are grateful for useful feedback received from their discussant at the CELS 2014, and to the edi-

tor of JELS and two anonymous reviewers for helpful comments on a previous version of this article.

1

Early school leaving can be viewed as a turning point that attenuates social bonds, resulting in a greater likeli-

hood of crime in adulthood (Sampson & Laub 2005).

2

This is a task not easily achieved. Despite the many programs that target raising the level of education in the

United States, the high school graduation rate has changed little since the 1970s, increasing to just 84 percent

for those born in 1986--1990 from 81 percent for those born in 1946--1950 (Heckman & LaFontaine 2010;

Murnane 2013).

716

Journal of Empirical Legal Studies

Volume 12, Issue 4, 716–756, December 2015

complicated by the fact that criminal behavior often begins with delinquency while still

in school. The onset of delinquency typically occurs between the ages of 13 and 15 and

peaks in the late teenage years (Welsh et al. 2008; Blumstein et al. 1986; Farrington

1986; Gottfredson & Hirschi 1986; Grogger 1998; Lochner 2004). In the United States,

for example, males younger than 16 account for 11 percent of arrests for property crime

and 8 percent of those for violent crime, while males younger than 18 account for 19

percent of arrests for property crime and 14 percent of arrests for violent crime (Federal

Bureau of Investigation 2011). Given that the minimum school leaving age ranges from

16 to 18 across U.S. states, it is clear that not only are crimes being committed by boys

who are still in school, but their participation in crime is peaking at the same time as

they are making important decisions about their education. This suggests that in addi-

tion to being a cause of crime, low educational attainment may be a consequence of

(juvenile) crime.

The focus of previous research studying the relationship between education and

crime has been on establishing the impact of educational attainment on crime commit-

ted by adults. Several studies have, however, investigated the contemporaneous effect of

school attendance on crime, finding that attending school reduces property crime

(Anderson 2014; Jacob & Lefgren 2003; Luallen 2006). Few studies examine the effect

of juvenile delinquency on educational attainment. Hjalmarsson (2008) investigates the

effects of juvenile arrest, charge, conviction, and imprisonment on high school gradua-

tion; Kirk and Sampson (2013) and Webbink et al. (2013) study the impact of being

arrested on graduating from high school; while Aizer and Doyle (2015) investigate the

effect of incarceration on high school graduation. These studies conclude that interac-

tions with the criminal justice system in the form of arrests and incarceration decrease

educational attainment. Whether juvenile crime per se leads to early school leaving is an

open question, and one we seek to answer.

3

This article contributes to the literature by empirically investigating whether juve-

nile crime leads to lower educational attainment and, if it does, determining why this is

the case. To this end, our analysis allows for the possibility that different offending pro-

files (as characterized by the type of delinquent acts committed, the age of first offend-

ing, and intensity of offending) have differential effects on educational attainment. This

is an important contribution since knowledge of who among delinquents are at greatest

risk of early school leaving is critical in targeting interventions to ensure they are

received by those who will benefit most. A further contribution of this study is that we

investigate several important mechanisms via which delinquency may lead to early

school exit. Understanding the mechanisms underlying the link between delinquency

and school leaving is critical for formulating policy interventions that are effective at

reducing the educational consequences of delinquency. For example, if it is contact

with the court system or incarceration that leads to reduced educational attainment,

3

Previous studies have shown that juvenile delinquency is negatively associated with education but do not attempt

to discern whether this association reflects common confounders or a causal relationship (Tanner et al. 1999;

Hannon 2003).

717Does Juvenile Delinquency Reduce Educational Attainment?

then diverting young offenders from the criminal justice system, through youth confer-

ences, for example, may improve their educational outcomes. If, however, the effect of

delinquency on education is not confined to those subject to criminal justice interac-

tions, then diversion programs will have no educational benefit for the majority of delin-

quents who do not come into contact with the criminal justice system.

The key empirical challenge faced in this research is that the decisions to partici-

pate in delinquency and to leave school are likely to be affected by difficult to observe

factors such as cognitive and noncognitive abilities (Heckman et al. 2006).

4

The poten-

tial for common unobserved confounders renders youthful delinquency endogenous to

the schooling decision. We address this issue using the approach developed by Altonji,

Elder, and Taber (AET) (2005).

5

This approach is based on a bivariate probit setup but

assumes that the model is underidentified and uses selection on observables to learn

about selection on unobservables. Requiring a rich set of observable characteristics that

explain the outcome of interest, this method is particularly appropriate given our use of

the National Longitudinal Survey of Youth 1997, a survey that was designed with a focus

on the educational experiences of youth, and that also provides detailed information on

their delinquent behavior. The AET method is attractive because it allows us to evaluate

the plausibility and potential strength of a relationship running from delinquency to

education in the absence of plausible exclusion restrictions.

Focusing on males and making use of the AET approach, we find plausible evi-

dence that delinquency by the age of 16 lowers the probability of graduating from high

school and college. We find that this effect is driven by early initiators (by age 14) and

those who commit a greater number of delinquent acts. Especially in the case of high

school graduation, we find that it is engaging in income-producing delinquent activities,

either alone or in combination with non-income-producing activities, that substantially

reduces educational attainment. Our investigation into the potential mechanisms

through which delinquency impacts education shows that this effect is not solely attrib-

utable to contact with the criminal justice system, nor is it specific to gang members. We

do find evidence that the impact of delinquency on education works, at least in part,

through subjective beliefs about the probability of being arrested for a property crime, a

key determinant of the expected returns to crime (Becker 1968). Overall, our findings

are consistent with delinquency building criminal capital and increasing expected

returns to crime relative to work, culminating in early school exit.

The rest of the article is laid out as follows. Section II presents the empirical

framework underlying our study of the impact of delinquency on schooling, and

describes the AET approach to addressing endogenous selection into delinquency in

4

These cognitive and noncognitive abilities are shaped at a very young age with parents playing an important

role in their development (Heckman & Carneiro 2002; Heckman & Masterov 2007; Cunha et al. 2006). Differen-

ces in these abilities tend to begin early and persist throughout life.

5

Altonji, Elder, and Taber (2005) investigate the impact of attending a Catholic school on education outcomes.

The AET approach has also been used to examine the effects of psychiatric disorders on labor market outcomes

(Chatterji et al. 2011); the effects of alcohol on youthful risky sexual behaviors (Markowitz et al. 2005); and the

effects of Swan-Ganz catheterization on mortality (Altonji et al. 2008), among others.

718 Ward and Williams