This project identifies and measures the effects of criminal record-sealing laws on ex-offenders' employment outcomes and subsequent crime commission. Existing research finds that long after criminal offenders complete their sentences, they face major socioeconomic obstacles due to their criminal records, such as an inability to find work. Moreover, criminal recidivism research demonstrates that unemployment and low wages are significant, though moderate, predictors of future offending. In theory, then, allowing offenders to seal their records might improve outcomes by offering them an escape from this cycle of poverty and crime, and a number of states have adopted record-sealing laws based on this assumption. However, no existing empirical research tests these laws' effectiveness, and critics suggest that the laws may either fail due to information leaks or even increase crime by hiding information that is relevant to public safety.
The project tests the theory underlying record-sealing laws using data from Michigan, a state that has a law allowing most first-time offenders with a single conviction to seal their records after five clean years. Using a large sample of ex-offenders and matched data from several state agencies, we employ a number of robust statistical approaches to draw causal inferences about the consequences of sealing records, taking advantage of the fact that Michigan's eligibility requirements draw sharp distinctions between groups of offenders that are otherwise quite similar. For instance, first-time offenders who are eligible for sealing can be compared to otherwise-similar offenders who are rendered ineligible by a minor, long-ago prior misdemeanor such as underage drinking. Likewise, offenders' outcomes can be compared just before and after the five-year waiting period ends, to see if becoming eligible for sealing after five years causes a discontinuous, sudden improvement in outcomes.
The project is designed to guide policymaking that could have important effects on the lives of the forty-seven million Americans with criminal records, on their families, and on public safety. In addition, the large matched database created for the project will be useful to future researchers studying other issues concerning reentry and recidivism.
