Working Paper: NBER ID: w20837
Authors: Howard Bodenhorn
Abstract: Prison overcrowding is a perennial problem and several states are under court order to reduce crowding. The long-term solution to crowding has been more prisons. The short-term solution is early release. Early release programs can be effective when they balance the savings of reduced prison costs against the costs of recidivism by released convicts. This paper uses historical data to investigate how prison officials altered their early release policies in the face of prison crowding and rising real per prisoner detention costs. The empirical evidence is consistent with the hypothesis that prison officials make use of information about the risks of recidivism revealed at trial and during incarceration to make informed decisions about whom to release and when.
Keywords: Prison Crowding; Recidivism; Early Release; Rhode Island
JEL Codes: K14; N00
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| increased prison crowding (K14) | higher probability of prisoners being selected for early release (K14) |
| higher costs of incarceration (K14) | higher probability of prisoners being selected for early release (K14) |
| prison population exceeds design capacity (H76) | higher likelihood of early release (J26) |
| observable characteristics (length of sentence, age at release) (K40) | selection into early release (C24) |
| future recidivism risk (K14) | fewer days off through early release program (J22) |