The last decade has been marked by significant shifts in crime control policies, particularly those policies designed to address the unique problems created by drug offenders, alcohol offenders, sex offenders, and mentally ill offenders. This article examines the decision to expand the use of prisons and jails for offenders in those categories and the corresponding conversion of community corrections programs to a surveillance/control orientation. Based on their review of current evaluation research, the authors explore the differential impact of this policy shift on women, minorities, and the poor and then discuss the implications of their findings for sentencing and correctional policy. They argue that, unless there is an immediate shift in correctional control policy, the problems of gender-, race-, and class-based bias will remain. Although they find that much criminal justice research concludes that discrimination does not exist, it is still the case that gender-, race-, and class-based differences can be identified at every point in the criminal justice process. Hence, it can be argued that it is the claims of those on either side of the issue that are important in policymaking, and not the empirical data. References