A Matched-Samples Comparison of Pre-Arrest and Post-Booking Diversion Programs in Florida’s Second Judicial District

Abstract

Pre-arrest diversion programs have the potential to reform the front-end of the criminal justice system and reduce the stigma associated with a low-level arrest, but little evaluation work has been conducted in this area. The current study was designed with two objectives: 1) to examine within-program indicators of completion and post-program arrest between the Adult Civil Citation (ACC) pre-arrest and the Pre-Trial Intervention (PTI) post-booking program in Florida’s second judicial district, and 2) to conduct a matched samples comparison of subsequent arrest between the two programs. A prospective longitudinal evaluation was conducted from 2010–2017 to assess within-program indicators of successful completion in addition to a between-program comparison of follow-up arrest. Propensity score matching was used to compare post-program arrest rates between 7,410 adults who entered the PTI program and 1,279 adults who participated in the ACC program. Demographic background, offense type, and completion status were collected within the respective programs while follow-up arrest data was drawn from statewide records. Successful program completion and post-program arrest were associated with participants’ sex, racial background, and offense type. Although participants in both programs experienced similar post-program arrest rates, further research is needed on additional program features, including qualitative differences between programs.

Keywords:

• Adult
• diversion
• intervention
• pre-arrest
• pre-trial
• recidivism

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 This tripartite coding scheme was adopted to account for limitations in the available data. Participants in the PTI program were not coded according to Hispanic background, so that information was not available and could not be used to match to ACC participants. The ACC participants’ ethnic background did include a Hispanic indicator, but it was decided to collapse this group into an “other” group for two empirically oriented reasons. When compared to the other ethnic groups, Hispanic ACC participants were no more likely to complete the conditions of the program (χ² (2) = 3.75, p = .44, Cramér’s = .054). In addition, the Hispanic group was comprised of 48 participants or 3.8% of the sample. Distributing this group across offense types, completion status, and rearrest yielded unstable multivariate results with large standard errors.