Redemption for Reintegrating Prisoners in the Era of Widespread Background Checks

Building on the approaches developed by Blumstein and Nakamura (2009), the current research project will estimate redemption times for prison releasees. The idea of redemption and redemption time is relevant to the ways in which criminal background checks are currently conducted and used, particularly by employers. Redemption time represents a time point when a prior record loses its relevance in distinguishing those with a criminal record from the general population or those without a record in terms of their risk of committing a crime in the future.

In order to account for heterogeneity among the releasees, this project will consider important factors that would likely affect the recidivism risk, such as the age at first arrest, the age at release, the number of prior records (arrests, convictions), the types of prior offenses, the length of incarceration, and the nature of post-release supervision.

The researchers will use data from a sample of individuals who were released from their first incarceration in New York from 1980 through 1986. This will serve as a function of the length of arrest-free time since the release. To address the possible recidivism heterogeneity within the population of releasees, the probability of recidivism will be examined by their characteristics such as the age at first arrest and age at release, the type of offense for which the releasee was incarcerated, the length of incarceration, criminal history, and the number of prior commitments. Corrections data from the New York State Department of Correctional Services will be linked with criminal history data from the New York state Division of Criminal Justice Services in order to examine the criminal history prior to the release and the post-release criminal records of first-time releases. In order to test the robustness of the New York findings, similar data will be obtained from two other states. Data from the FBI will be requested in order to account for the prevalence of out-of-state arrests.
Survival analysis will be used to model time to recidivism. The hazard function will be used to examine the timing of redemption. Crime-switch matrices and crime-type-specific hazards will be used to address the hazard of a rearrest for any offense and the hazard of a rearrest for a particular offense type. Subsequent results could inform employers about hiring individuals with criminal records and could inform state government pardon processes.
ca/ncf