Cost matters: a randomized experiment comparing recidivism between two styles of prisons

This study drew on a cost–benefit method to compare recidivism between two groups of inmates. Subjects were randomly assigned to serve a 6-month ‘early release’ term in one of two prisons which diverged dramatically in terms of structure, freedom, and staff–inmate interaction style (boot camp versus traditional prison). The study was motivated, in part, by the state of Maryland’s asking whether the ‘extra’ cost of running an early release program at the boot camp rather than a traditional prison was justified. Cost–benefit tools were employed, not only because they addressed this specific question, but also because they provided a platform for describing differences both in the amount and the quality (harm) of recidivism associated with facility assignment. Two themes are emphasized in this paper. First, even without consideration of differences in post-release impact, the boot camp cost less to run per inmate. This may be an important observation for policy makers now wrestling with budget declines and potential cuts; they should not assume that programs which offer more services are necessarily more expensive to operate than alternatives offering less. Second, the boot camp generated substantially less recidivism-harm relative to the traditional prison. This finding held across numerous sensitivity analyses standard to the field, as well as new approaches presented here.