Community Corrections in Context: A Multi-Method Evaluation of Risk Responsive and Neighborhood-Oriented Probation Models in New York City

A. Statement of the Problem
With over 4.7 million people under the supervision of federal, state, or local probation and parole agencies and widening support for a shift towards rehabilitative approaches to corrections, jurisdictions around the country are seeking efficient ways to integrate evidence-based practice into community supervision. Recent innovations in probation supervision in New York City (NYC)including a restructured supervision model rooted in Risk-Need-Responsivity principles and a neighborhood-oriented approach to supervision (NeON)may serve as a model for other jurisdictions, but have yet to be rigorously evaluated. Accordingly, the proposed multi-method evaluation will fill a gap in the scholarly literature on "what works" in community supervision and have important implications for community corrections agencies nationwide.

B. Subjects
An official records study will retrospectively analyze a sample of more than 5,000 NYC probation clients sentenced between 2010 and 2016; and a minimum of 450 probation clients sentenced after January 1, 2013 will take part in a client survey study. Four NeON neighborhood sites and two comparison neighborhoods will be selected for all study activities.

C. Research Design and Methods
The project will take a quasi-experimental, multi-method approach to evaluating the impact of RNR-based supervision and NeON, including three major research strategies: (1) a comparison of the impact of three supervision models on criminal justice outcomes; (2) in-depth surveys with active probation clients under two different supervision models (NeON and RNR-only); and (3) site-level case studies of post-reform probation supervision in six NYC neighborhoods.

D. Analysis
The analytic plan will employ descriptive, bivariate and multivariate (regression) strategies to assess the impact of supervision model (pre-reform, RNR-only, or RNR-NeON) on outcomes, as well as the potential influence of multiple intervening variables drawn from relevant survey domains, such as client risk-need characteristics, procedural justice, and collective efficacy. Propensity score adjustment techniques may be used to enhance the equivalency of samples.

E. Products, Reports, and Data Archiving
This project will have implications for jurisdictions nationwide. Along with an Executive Summary for NIJ, results will be reported in at least two peer-reviewed research journal articles. An array of practitioner dissemination efforts will also be undertaken, including a white paper summarizing research findings, a Frequently Asked Questions (FAQ) document, conference presentations, and multi-media dissemination efforts. De-identified study data will be archived per NIJ requirements. ca/ncf