Working Paper: NBER ID: w23436
Authors: A. Mitchell Polinsky; Paul N. Riskind
Abstract: In this article we derive the mix of criminal sanctions—choosing among prison, parole, and probation—that achieves any target level of deterrence at least cost. We assume that prison has higher disutility and higher cost per unit time than parole and probation and that potential offenders discount the future disutility of sanctions at a higher rate than the state discounts the future costs of sanctions. Our primary insight is that there is a “front-loading advantage” of imprisonment due to these differential discount rates. This advantage implies that (a) whenever a sentence includes both a prison term and a parole term, the prison term should be imposed first; and (b) it may be optimal to employ a prison term even if prison has higher cost per unit of disutility than parole and probation and even if prison is not needed to achieve the target level of deterrence.
Keywords: deterrence; prison; parole; probation; criminal sanctions
JEL Codes: H23; K14; K42
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| order of sanctions (K40) | effectiveness of deterrence (K42) |
| prison first followed by parole (K14) | overall deterrence effect (K49) |
| higher disutility per unit time from prison (K14) | deterrence effect (K40) |
| discounting of future disutility by offenders (D15) | preference for imprisonment first (K14) |
| target level of deterrence (K42) | optimal use of sanctions (F51) |