Abstract
The ignition interlock has been touted as a "high-tech" intervention to reduce drunk-driving recidivism. However, it has been difficult to evaluate because outcomes may be a result of program selection. In the study reported here, a random sample of interlock cases was contrasted against a comparison group of impaired drivers in a retrospective analysis. Program effects were tested through a series of logit regressions, and program and comparison group drunk-driving survival rates were also compared. Program participants had lower recidivism rates and higher survival rates, indicating that the program has strong potential to protect the public and to change offender behavior. (A)