Got Used to Violence? The Micro-Level Effects of Homicides on Mobility

• In Event: Method Meets Mission: Empirical Applications Across Crime, Justice, and Community Response

Abstract

Theory predicts that the risk of victimization alters individuals' behaviors, exerting physical and emotional effects on their daily routines. Here, we test whether violent crime impacts daily visits to Points-of-Interest (POIs) in two American cities: Baltimore and Chicago. To do so, we leverage data on the timing and location of homicides in both cities during the 2022–2023 period, matching them with fine-grained, hour-level data on visits to POIs located within a radius of 820 feet from each crime incident. As a result, we assembled two datasets containing over 41 million hour-homicide-POI observations for Chicago and one counting around 15 million measurements for Baltimore. Via a computational framework based on Bayesian modeling, we estimate the effects of homicides on hourly visits to POIs, distinguishing between short-term and long-term effects, focusing on two-day and two-week windows after the homicide, respectively. Our findings challenge criminological and victimization theories, demonstrating that only a small fraction of POIs experience statistically significant reductions in visits, while the vast majority remain unaffected. We further explore whether socioeconomic and demographic characteristics of block groups can explain reductions in visits, contributing to the empirical, theoretical, and policy discourse on the effects of violence in American cities.

Authors

• Gian Maria Campedelli, Fondazione Bruno Kessler

• Simone Centellgher, Fondazione Bruno Kessler

• Massimiliano Luca, Fondazione Bruno Kessler

• Ariadna Albors-Zumel Albors-Zumel, University of Trento

• Bruno Lepri, Fondazione Bruno Kessler