Informing the Reformers: Identifying Patterns in the Effectiveness of Pro-Rehabilitation Arguments in the U.S. Criminal Justice Reform Movement

Abstract

The advent of mass incarceration in the United States, which has entailed a 500% increase in the nation’s incarceration rate since 1970, has coincided with the rise of ‘tough-on-crime’ rhetoric and the emergence of retributivism as the justifying aim of the criminal justice system. A criminal justice reform movement has emerged in response to America’s mass incarceration crisis, with the aim of reducing incarceration levels by promoting rehabilitation as the goal of criminal justice policy in lieu of retributivism. Little is known about the effectiveness with the different arguments employed by reform groups promote public support for rehabilitation, and how this effectiveness varies for different partisan and demographic groups. Consequently, it has previously been impossible to determine whether or not reform groups’ prevailing campaign strategies are efficient in generating the maximum impact given the groups’ limited resources. This thesis uses an experimental survey design to create a new dataset, from which the effectiveness of fiscal, social, racial justice, and religious arguments in promoting pro-rehabilitation attitudes within a general population sample may be tested. Cross-section analyses are conducted using linear regression models with interaction terms to identify patterns in the effects of these arguments with different groups. This study is unique as no prior study has investigated the effects of pro-rehabilitation arguments on public or partisan opinion. Results from 1,214 subjects indicate that self- identified Republicans have initially more retributivist attitudes than Democrats, and that fiscal, social, and religious arguments increase Republican support for rehabilitation to match the support of Democrats. Arguments rooted in racial justice principles have no effect on Republicans, and none of the four experimental treatments have any effect on Democrats. However, partisanship is shown to be only one of several determinants of treatment effect: treatment effects vary by party, religion, education level, and race, meaning that some of the treatment effects for Republicans are better attributed to underlying demographic effects. For the criminal justice reform movement, these findings suggest glaring inefficiencies in the ways that the use different types of arguments to target their audiences. The results suggest that reform groups may maximise their impact by targeting specific groups with specific arguments, namely, Republicans with fiscal arguments, Christians with social, racial justice, and religious arguments, and people without bachelor’s degrees with fiscal and racial justice arguments. Reform efforts targeting Democrats are ineffective, as Democrat attitudes are unlikely to shift, which means that non-partisan organisations that dedicate resources to rehabilitation advocacy campaigns that have Democrats in their audience are inefficient. Reform groups are shown to be out of alignment with these efficient strategies, suggesting that the findings of this thesis may be used to increase the efficiency of the criminal justice reform movement.