The Challenges of Translating Odds Ratios into Legal Policy: A Cautionary Tale of Strangulation Laws and Homicide Prevention

• In Event: Causal Inference and Algorithmic Discovery for Criminology and Public Policy

Abstract

Integrating scientific evidence into legal policy is a complex process that requires careful interpretation and communication of findings. This paper examines the widespread adoption of strangulation laws in the United States, which aimed to address the previously underestimated prevalence of non-fatal strangulation (NFS) in intimate partner violence cases. A significant motivation behind these laws was research evidence indicating an alarmingly high odds ratio between NFS and future homicide risk. We argue that policymakers may have overlooked the low base rate of domestic homicide when translating this finding into legislation. We assess the real-world impact of these laws using a variety of causal inference techniques. We first show strangulation laws lead to a substantial increase in the share of assaults being charged as felonies --- a common stipulation of these laws. However, we do not find an effect of these laws on homicide victimization rates. Through calibrated simulations, we demonstrate how odds ratios from case-control studies, combined with reasonable estimates of prevalence, can provide a more nuanced understanding of the potential impact of policy changes. Our findings highlight the challenges of translating odds ratios, a commonly reported statistic, into effective legal interventions in the presence of low base rates.

Authors

• Zubin Jelveh, University of Maryland

• Xinyi Situ, University of Maryland, College Park