Evaluating the (F)utility of Mandatory Minimum Sentencing Laws in Pennsylvania

Abstract

In the current criminal justice policy sphere mandatory minimum sentencing serves two important purposes 1) they are used as a punitive response to immediate crime concerns and 2) their removal is viewed as a tool to conserve resources, decarcerate, and promote fairness in sentencing. Though much research explores how the passage of these laws relates to crime, the literature has not focused on the public safety implications of removing mandatory minimum sentences. Using a comparative interrupted time-series approach, the present work investigates whether a Pennsylvania Supreme Court decision that invalidated several mandatory minimum sentencing provisions impacted the state’s crime rate. We find little to no evidence of a discrete shift in overall or type-disaggregated crime rates, or changes in the slope of any crime trend when the state reduced their use of mandatory minimums. These findings tentatively suggest that many mandatory minimums can be repealed without risking public safety.

Keywords:

• crime trends
• interrupted time-series
• mandatory minimum sentencing

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1 Many scholars and practitioners also commend MMs for their impact on court functioning. These sentences are often seen by prosecutors as tools to leverage plea deals, which vastly improve the efficiency of the court system.

2 These values exclude DUI mandatories.

3 This study provides a macro-level assessment of notice MMs in Pennsylvania. It is possible that the policy in question could impact individual behavior (e.g., by changing the costs associated with committing weapons or drug related offenses), a micro-level analysis is beyond the scope of this study.

4 We do not include public order and “other” offenses in our measure of the total crime rate because each state includes different crimes in these categories, rendering them incomparable.

5 State data on opioid-related hospital admissions were not used as a measure of drug use because in Pennsylvania hospital coding requirements for labeling opioid-related incidents expanded during the study period, shortly after the Hopkins decision. Given that values before the new coding may be underestimated, and the bias in these estimates changes around the time of our intervention, we chose to rely on UCR data of drug offenses as a proxy for drug use, in which the error is stable over the study period (Pennsylvania Health Care Cost Containment Council, 2018).

6 Analyses conducted without the natural logarithm do not vary substantively from the findings presented below and are available upon request.

7 As a sensitivity test, models including a trend variable were estimated as a second to address non-stationarity. These models’ findings do not vary substantively from those presented below and are available upon request.

8 It is important to note that our study was conducted immediately following a decade long national crime decline. In 2010, both Pennsylvania and Ohio reported lower crime rates than at any point in the two decades prior. A low base rate could obstruct our ability to observe a policy effect. Over the study period, both violent and property crime continued to mirror the national trends, suggesting the potential influence of national factors, though it is agreed that there was more variation in state and city level crime trends after 2000. Further, our analytic strategy removes the potential influence of national and regional trends by adding Ohio as a comparison group.