Job-Related Programs for People on Supervision: Reframing the Problem

• DOI: http://doi.org/10.1177/00027162221115501
• Published date: 01 May 2022
• Date: 01 May 2022
• Subject Matter: Improving Human Capital and Labor Market Opportunities

98 ANNALS, AAPSS, 701, May 2022

DOI: 10.1177/00027162221115501

Job-Related

Programs for

People on

Supervision:

Reframing the

Problem

By

SHAWN BUSHWAY

1115501ANN The Annals of the American AcademyJob-Related Programs for People On Supervision

research-article2022

Job training programs for people under supervision

have been based on an economic framework that identi-

fies individuals involved in crime as a disadvantaged

group with poor human capital. The best available

research evidence has not found that these programs

consistently improve employment outcomes. This arti-

cle reviews the evidence for the effectiveness of stand-

ard job training programs and then examines new

developments in the field that use alternative frame-

works for understanding the roles of such programs.

The first alternative is signaling: how people under com-

munity supervision use the completion of job training to

signal to employers and others that the behavior that led

to their conviction is either anomalous or no longer

representative of them. The second alternative is a

model of desistance known as identity change: the ways

in which job training can help individuals solidify a new,

more prosocial identity. I make sense of extant work and

new alternatives and provide a set of recommendations

for change in the community supervision system.

Keywords: desistance; job training; identity; parole;

probation; community supervision;

employment; signaling

Introduction

In 2018, 4.4 million people were under com-

munity supervision in the United States: 3.5

million people on probation and almost 900,000

on parole (Maruschak and Minton 2020). Many

of the programs created for this significant

population focus, at least in part, on employ-

ment or employment training. This standard

approach is based on an economic framework

that suggests that people choose to commit

crime because the benefits of crime outweigh

the benefits of work. The solution, then, is

Shawn Bushway is a senior policy researcher at the

RAND Corporation and a professor of public adminis-

tration and policy at the University at Albany. He is an

expert on desistance and the relationship between

crime, criminal justice involvement, and employment.

Correspondence: sbushway@rand.org

JOB-RELATED PROGRAMS FOR PEOPLE ON SUPERVISION 99

either to improve the human capital of these workers so the benefits of their work

improves (supply-side programs) or to incentivize employers with benefits or sub-

sidies, encouraging them to hire people on community supervision (demand-side

programs) (Bushway and Reuter 2001). While it is an empirical fact that many

people currently on community supervision have low levels of education and poor

job skills, decades of research find that employment programs designed to

improve human capital or increase employer demand have failed to achieve

noticeable results (Visher et al. 2005; Bushway and Apel 2012). This suggests that

the framework justifying these programs needs to be re-examined and modified.

One possible modification would be to focus on other problems faced by this

population, including the very fact of their community supervision. This view-

point frames the problem not as the lack of human capital or detachment from

the labor market among people under supervision but, rather, as the stigma and

human capital displacement caused by conviction and supervision itself. As such,

lowering the rates of conviction and community supervision would jointly fix the

problems of both unemployment and crime. This is an attractive solution, at least

in part because the high rate of criminal justice involvement in this country is so

obviously problematic and unsustainable. And there is some evidence to support

this approach. Recent evidence suggests that without any significant increases in

crime, the government can reduce supervision rates by less frequently revoking

parole and probation for technical violations, by applying probation sentences

instead of prison, and/or by using pretrial supervision selectively (Austin and

Jacobson 2013; Lofstrom, Bird, and Martin 2016).

However, the best available evidence simply does not support the claim that

even the harshest form of criminal justice involvement—incarceration—causes

the problems encountered in the labor market by people with records (National

Research Council 2014). While criminal justice involvement certainly carries

stigma, the employment problems for formerly incarcerated people existed before

their spells of detention (Kirk and Wakefield 2018). Bureau of Justice Statistics

data show that over 80 percent of all incarcerated people have at least one prior

felony conviction before their current prison term (Langan and Levin 2002).

Similarly, most people who serve time in prison have poor outcomes across dimen-

sions before entering the criminal justice system (Petersilia 2003).1 As a result, the

notion of reentry for people on parole is misleading in its focus on the need to

reintegrate people into the community. Rather than prison causing people to lose

their integration, many were not integrated even before their incarceration.

This lack of integration is also an issue for many people on other forms of com-

munity supervision, like probation. In most populations, criminal justice involve-

ment peaks in the late teens and early twenties, before employment becomes a

primary way for people to spend time outside the home. In other words, criminal

NOTE: I would like to acknowledge the contributions of past coauthors to the ideas discussed

in this article, including Peter Reuter, Bobby Brame, Robert Apel, Garima Siwach, Megan

Denver, Nidhi Kalra, Brent Orrell, and Raymond Paternoster. I would also like acknowledge

the helpful suggestions from the two editors, David Harding and Bruce Western. All errors

remain my own.