Research on District Attorneys'' Pretrial Diversion Programs: A Proposal for a Comprehensive Multi-Method Study

The purpose of this research is to conduct a 30 month national, multi-method study of prosecutor led diversion programs (PDPs). The goals of the study are: (1) to produce a rich understanding of 17 existing PDPs in 11 selected district attorneys; offices through in-depth case studies; (2) to conduct a prospective process evaluation to determine the risks and needs of program participants, the types and dosage of services used, the completion rates of assigned services, and participants, satisfaction with services they received; and (3) to conduct an impact and cost-benefit analysis examining the PDPs, influence on case outcomes, reoffending, pre- and post-disposition incarceration, and quantifiable cost savings.
To accomplish the goals of this project, Center for Court Innovation (CCI), along with their research partners, RAND Corporation (RAND) and the Association of Prosecuting Attorneys (APA), will collect case study data through intensive four day in-person site visits to each of the 11 sites by two members of the research team. The products of the site visits will be: (1) a logic model for each PDP program; (2) a final cross site typology of alternative PDP models; (3) rich information about specific promising or innovative practices; and (4) rigorous evaluability assessments to ensure that the impact evaluation will be successful.
During site visits, researchers will conduct in-depth interviews with program stakeholders (e.g. police, pretrial services, the court, probation, and community based service providers), specifically with lead attorneys in the prosecuting agency as well as with those attorneys doing the hands-on work of reviewing and determining the eligibility of cases referred to the PDP. The impact evaluation will focus on four programs selected through site visits that meet the following criteria: have at least 250 participants, have the ability to design a quasi-experiment with a comparison group; have quality data on PDP participants; have the capacity to obtain cost and impact data for both participants and comparisons; commitment from staff; and have a breadth of selected program types within the site. Baseline and follow-up interviews of the defendants will collect data on factors such as antisocial personality, criminal thinking patterns, antisocial peers, family or marital problems, school or work problems; lack of pro-social leisure/recreational activities; substance abuse; and criminal history, which will be obtained through administrative data. Further, the interviews will include brief inventories in three attitudinal domains, procedural justice, legal pressure and readiness to change.
The analysis plan for this study includes multiple methods comparing PDP participants and the comparison group on case outcomes: filing rate, dispositions, sentences, re-arrests, jail/prison time, etc. Logistic regressions will be used to predict dichotomous outcomes and Poisson or Negative Binomial specification when predicting count outcomes. Cox Regression and Survival Analysis will be used to estimate group differences on days to re-arrest. Data from site visits will be analyzed qualitatively. ca/ncf