Exploring the mobility-crime dynamics across various stages of a Global Health crisis: Temporal evidence from 25 cities in the United States

Contemporary global crises have intensified the conditions of a risk society. In response to perceived risks, human mobility, which refers to physical movement across locations such as stores and schools, has changed. Routine Activity Theory (RAT) posits that such mobility changes can influence crime by affecting the interactions among potential offenders, suitable targets, and capable guardianship. Until now, limited research has examined the long-term mobility-crime dynamics. This study addresses this gap by investigating the temporal associations between mobility patterns across six location types and eight common offense types in 25 major U.S. cities in a risk society exemplified by the COVID-19 pandemic. Descriptive analyses reveal divergent crime trajectories: persistent declines in drug offenses, residential burglary, and robbery, while a sustained rise in motor vehicle theft. Using Toda-Yamamoto Granger causality tests, we find the strongest associations between mobility and crime in the initial stage, with varying relationships across locations and crime types. Mobility shows a greater influence on violent crime than the reverse, whereas bidirectional associations are more common in property crimes. These findings underscore the global health crisis's lasting effects on mobility-crime dynamics and provide insights for targeted crime prevention and effective urban planning strategies in a risk-oriented society. • Some crimes like drug and motor vehicle theft did not return to pre-pandemic levels. • RAT had the strongest explanatory power early and varied across crime types later. • Mobility impacted violent crime more, with a bidirectional property crime link. • Different routine activities are Granger associated with crimes in distinct ways.