Evaluation of Projects Supported by Byrne Memorial Funds

Abt Associates proposes a 36-month study to evaluate the Fortune Society's suite of
services that are tailored to needs of state and local prisoners released to NYC. This will include an assessment of whether the program design and operation as implemented are aligned with its public safety objectives; whether the program is transferable/replicable; whether its programs
reduce clients recidivism, drug use, employment and homelessness; and whether its benefits outweigh its costs. Implementation and transferability will be assessed using information obtained largely from interviews, observation, and Fortune Society MIS data. To test program effectiveness, we will seek to implement an experimental design with random referral/non-referral over a 24-month period by probation/parole officers in NYC to Fortune (n~2400 in each group). For clients identified by Fortune's outreach to prisons/jails (about 2500 over 24-month period), a non-experimental comparison population will be constructed of other state/local prisoners released to NYC. Subjects will be limited to recently
released prisoners (although Fortune accepts persons who have been out for years).
In both pairs of samples, subjects' outcomes will be compared. Recidivism will be
measured primarily as an arrest in the state's criminal history records but alternative measures (e.g., reincarceration, parole violation) will be obtained. Drug use will be measured using probation/parole testing information; employment will be measured using wage reports to the
Social Security Administration; homelessness will be measured by matching NYC homeless
MIS records to state prison and NYC jail records. Information about subjects will be obtained from matched and merged data from several agencies. Models will be developed to estimate effects when numerous intergroup differences are controlled statistically; attention will be given to possible biases associated by selection and self-selection. Expected sample sizes will be sufficient to afford a high probability of detecting relatively small program effects. The economic analysis will probably take the form of a benefit-cost analysis, although
alternatives (e.g., cost-effectiveness analyses) will be considered. Costs of Fortune's programs will be identified. Costs and savings associated with changes in recidivism-principally criminal justice processing costs but also estimated costs to victims-will be estimated, as will
costs/savings associated with changes in drug use and homelessness. Benefits associated with employment-including tax payments-will be estimated.

This research seeks to evaluate the effectiveness of a program intervention in achieving specific outcomes.