Alternative Solutions to the Problem of Selection Bias in an Analysis of Federal Residential Drug Treatment Programs

摘要

In an evaluation of prison-based residential drug treatment programs, we use three different approaches to modeling post-release outcomes. Two of the approaches, the instrumental variable and the Heckman approach, attempt to minimize selection bias as an explanation for treatment outcomes. The results from these analyses are compared to regression analysis in which treatment effects are measured by a dummy variable. We discuss the advantage of using more than one method to increase confidence in findings within the context of selection bias issues. The advantages of the Heckman approach, which models selection bias and controls for its consequences, are discussed. Three-year outcome data for 2,315 Federal inmates are used in the analyses where we separately examine the outcome measures of recidivism and drug use for men and women.