To Detain or Not to Detain? Using Propensity Scores to Examine the Relationship Between Pretrial Detention and Conviction

• Published date: 01 February 2019
• DOI: 10.1177/0887403416668016
• Date: 01 February 2019
• Subject Matter: Articles

https://doi.org/10.1177/0887403416668016

Criminal Justice Policy Review

2019, Vol. 30(1) 128 –152

© The Author(s) 2016

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0887403416668016

journals.sagepub.com/home/cjp

Article

To Detain or Not to Detain?

Using Propensity Scores to

Examine the Relationship

Between Pretrial Detention

and Conviction

Jacqueline G. Lee1

Abstract

Recent work has found that individuals who are detained prior to their sentencing are

more likely to be convicted, be sentenced to prison as opposed to jail or probation,

and to receive longer sentences if convicted. Is this effect due to detention, or

is it merely a result of the same underlying criminal propensity being considered

separately at each stage of the sentencing process? This study uses propensity score

methods to create more comparable comparisons between detained and released

criminal defendants. The results of this study indicate that detention itself has an

independent effect on the likelihood of conviction. Even after matching with available

covariates, individuals who are detained are more likely to be convicted than those

who are released prior to their trial.

Keywords

criminal justice policy, criminal court, sentencing disparity, sentencing

Introduction

Bail and pretrial release decisions are a distinct and important part of criminal justice

processing. Although bail and pretrial release decisions have received limited research

attention as compared with later sentencing outcomes, previous studies have displayed

that these early decisions can be the site of meaningful disparity within the United

States (Demuth, 2003; Schlesinger, 2005). Several factors make the pretrial process

one worthy of further academic study. First, a substantial number of criminal

1University of Maryland, College Park, MD, USA

Corresponding Author:

Jacqueline G. Lee, University of Maryland, 2220 LeFrak Hall, College Park, MD 20742, USA.

Email: jgclee@umd.edu

668016CJPXXX10.1177/0887403416668016Criminal Justice Policy ReviewLee

research-article2016

Lee 129

defendants (approximately 30%-50%) are detained prior to their trial or sentencing,

depending on the sample (see Franklin, 2010; Schlesinger, 2005; Spohn, 2009).

Second, judges are given wide discretion and are often forced to make split-second

decisions with incomplete information (Goldkamp, 1979). Third, unlike other criminal

justice processes, bail decisions are made by a number of different decision makers

and these decisions are rarely reviewed by appellate courts (Clarke & Kurtz, 1983;

Demuth & Steffensmeier, 2004). And finally, pretrial detention involves incarceration

before conviction. While preventive detention and community protection are both

constitutionally permissible reasons for detention, the fact remains that individuals

who are detained while awaiting trial or sentencing have not yet been convicted of any

crime (Goldkamp, 1979; Gottfredson & Gottfredson, 1988).

Furthermore, the significance of pretrial detention to defendants’ lives both before

and after trial or conviction cannot be overstated (Demuth & Steffensmeier, 2004).

While detained, defendants often lose their jobs, are unable to financially provide for

their families, cannot meet family obligations such as child care, and may lose their com-

munity ties (Frazier, Bock, & Henretta, 1980). It is more difficult for them to meet with

their attorneys and this may lead to a lessened ability to aid in preparing a defense (Foote,

1954; Gottfredson & Gottfredson, 1988). Some work has found that lawyers spend less

time with their detained clients than their clients who are not in jail (Allan, Allan, Giles,

Drake, & Froyland, 2005). Kellough and Wortley (2002) have further found that detained

individuals reported pleading guilty as a way to control certainty in their lives. In qualita-

tive interviews, detained respondents stated that they pled guilty to get out of jail soon

because it took longer to go through a trial (Kellough & Wortley, 2002).

Most importantly to the research at hand, defendants who are detained while await-

ing trial tend to have harsher punishment outcomes. Extant research indicates that

detained defendants are more likely to be convicted (Phillips, 2008), to be sentenced to

an incarcerative as opposed to non-incarcerative sentence (Tartaro & Sedelmaier,

2009), and to receive longer sentences if convicted (Sacks & Ackerman, 2014; Spohn,

2009). Previous research on this topic clearly demonstrates that detained individuals

face harsher outcomes than released individuals, but prior studies generally rely on

traditional regression methods that may not fully account for potential selection effects.

Strict comparisons between individuals who received bail compared with those who

did not may overestimate the observed impact of this decision because similar factors

may be driving both the early detention decision and the later conviction and sentencing

decisions. To help address this issue, propensity score matching is a useful strategy

because it provides for closer comparisons of more comparable individuals. The current

study will investigate whether being detained prior to trial is a disadvantage for crimi-

nal defendants when it comes to their likelihood of conviction, or whether the relation-

ship between detention and conviction is actually the result of selection effects. This

article makes a unique contribution to the literature by analyzing a well-documented

relationship with a newer and more precise methodology. This article proceeds by first

reviewing literature on predictors of pretrial detention, then outlines the advantages of

propensity score matching to the question at hand, and then tests the hypothesis that

pretrial detention may have independent effects on conviction.