ABSTRACT
The practice of parole in the United States has undergone significant change during recent decades. Since 1980, inmates have been increasingly released through mandatory rather than discretionary means. Using a sample of over 38,000 inmates released from fifteen states in 1994, survival analysis was used to determine the differences in recidivism between those released by discretionary and mandatory methods. Results show a statistically significant difference in recidivism, based on re-arrest, between those released from prison by these two methods. Kaplan-Meier survival graphs showed that mandatory release inmates had a median re-arrest date about 10 months earlier than paroled inmates. Additionally, a Cox Proportional Hazard survival analysis was employed using gender, race, age, criminal history, and offense type as control variables. The results demonstrated that offenders released by discretionary parole release had a 31 percent lower hazard of re-arrest than mandatory release offenders.