Evaluating the Effects of Realignment Practices on Recidivism Outcomes

California's Public Safety Realignment is considered one of the dramatic criminal justice policy experiments of our time. It entails transferring the correctional responsibility of non-violent and non-serious felony offenders from the state- to the county-level. The shift is premised on the idea that local practitioners may be more effective at managing these populations. California Public Safety Realignment did not mandate specific programs or strategies to supervise these offenders, but in fact encouraged counties to come up with their own plans to meet the individual counties' realigned population. In other words, since they are more knowledgeable about the offender's environment and community, they would be able to enact more responsive programming to the offenders' needs, thus reducing recidivism at a more cost-effective rate.
In order to substantiate these claims and evaluate the overall effectiveness of California Public Safety Realignment, The California Public Policy Institute (PPIC) is partnering with the State of Californias Board of State and Community Corrections (BCCS) to evaluate the outcomes from Realignment in a sample of 11 California counties. PPIC will the evaluate the impact of the Public Safety Realignment natural experiment with these counties realigned population to see if this policy does in fact lower recidivism.
In 2013, BCCS partnered with PPIC to track and report on realigned offenders in a sample of 11 counties in order to fulfill its statutory requirements. The 11 counties were chosen to be representative of the state in terms of demographic and economic characteristics, to include the majority of the realignment population, and to capture the diversity of the realignment strategies and practices employed by counties. PPIC is collecting data from these participating counties that include demographic and economic characteristics, current offense characteristics, criminal histories, and risk and needs assessments, as well as how the offenders are being supervised and what services and sanctions they are referred to or receive. For the purposes of this study, the PPIC data will be merged with offender-level data from the California Department of Justice (DOJ) and the California Department of Corrections and Rehabilitation (CDCR).
Prior to the analysis of the impact of Public Safety Realignment on offender outcomes, PPIC will conduct a genetic non-parametric matching technique to match realigned offenders to non-realigned offenders who remain the responsibility of the state. This will control for any differences within the two correctional populations. Once the matched populations are created, PPIC will conduct a regression analysis to assess the presence of any significant differences in recidivism between the realigned and the non-realigned population. The regression model will also control for individual and county-level fixed effects.
PPIC will then evaluate the effectiveness of individual county supervision and reentry plans in reducing recidivism. To accomplish this, PPIC will run a series of regressions that estimate the cross-county and within-county impact of various plans and treatment modalities on recidivism. Taken together, these two analyses will help correct for biases that may exist due to unobserved factors at both the county- and the individual offender-level.
In an effort to understand, and correct for, any unobserved bias introduced by the tailoring of supervision strategies to the individual offender, PPIC will augment the regression analyses with a probation officer survey to understand the decision-making process regarding the individual offender-level choices to provide services or administer sanctions. All probation officers in the sample counties will be surveyed.ca/ncf