Custodial Versus Non-Custodial Sentences: Long-Run Evidence from an Anticipated Reform

Given that offenders who received a custodial sentence did not have to wait as long as those who received a non-custodial sentence (because the latter were tried closer to the date of the reform), it is possible that our results (showing that custodial sentences foster crime compared to noncustodial ones) underestimate the true negative impact of incarceration. However, two elements allow us to minimize the extent of this concern. First, the percentage increase in trial waiting time in the 12 months prior to the signing of the legislation remains limited. It only increases sharply in the couple of weeks preceding the signature. Second, while theoretical arguments suggest that punishment celerity should deter recidivism, criminologists now tend to agree that it has no impact (see Pratt and Turanovic (2018) for a review of the evidence).^30
30 In Pratt and Turanovic’s own words: “While exceptions exist […] the general pattern revealed in this body of work is that celerity effects of punishment are nonexistent, and that even when present it can be difficult to disentangle such effects from other potentially confounding influences, like the perceived (or actual) certainty or severity of punishment. So again, the pattern is pretty clear that faster punishments appear to have little to no consistent, independent effect on one’s future criminal behavior.” While a few studies on the impact of the Hawaii Opportunity Probation with Enforcement (HOPE) project fostering swift-and-certain punishments found significant positive effects (Hawken and Kleiman, 2009; Kilmer et al., 2013), replication studies carried out in other states found no impact (see, for example, the experimental studies by Lattimore et al. (2016) and O’Connell et al. (2016)). Criminology & Public Policy devoted an issue to this topic, see for example Nagin (2016) and Cullen et al. (2016) in addition to the articles just cited.