Working Paper: CEPR ID: DP15214
Authors: Randi Hjalmarsson; Matthew Lindquist
Abstract: This paper studies the effect on mortality of two Swedish early release reforms in 1993 and 1999 that held prison sentences constant but increased the share of time inmates were required to serve from one-half to two-thirds. Contrary to previous correlational evidence, we find that reform exposure, and the corresponding increase in time served, did not harm post-release prisoner health. Rather, the overall risk of death decreases, with especially large and significant effects for those who are positively selected in terms of their criminal careers and connection to society. We also find (i) significant and persistent reductions in the chance of suicide, (ii) short-term reductions in violent death, and (iii) long-term improvement in general health (circulatory death). These cause-specific effects are driven by particular at-risk populations – individuals with pre-incarceration mental health problems, violent offenders, and older offenders, respectively. We argue that these findings are primarily driven by a direct in-prison health treatment and services mechanism: we demonstrate that health care utilization and program participation increases with time served. We also find that reform exposure decreases recidivism and has some very short-term beneficial labor market effects. Our main findings, however, do not appear to be driven by these life-style changes.
Keywords: prison; health; mortality; crime; recidivism
JEL Codes: K42; I14
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| exposure to the two-thirds reform (C93) | increase in time served (C41) |
| increase in time served (C41) | access to healthcare and treatment services in prison (I14) |
| access to healthcare and treatment services in prison (I14) | improvements in general health (circulatory death) (I12) |
| exposure to the two-thirds reform (C93) | decreased overall risk of death (I12) |
| exposure to the two-thirds reform (C93) | reduced risk of suicide (I12) |
| exposure to the two-thirds reform (C93) | short-term reductions in violent death (I14) |
| exposure to the two-thirds reform (C93) | decreased recidivism (K14) |
| exposure to the two-thirds reform (C93) | short-term positive labor market effects (J68) |