Spatiotemporal population mix (SPM) as a criminogenic mechanism: Testing environmental-criminology hypotheses with mobility clusters derived from mobile-device geotracking

Crime pattern theory and contemporary environmental criminology posit that crime risk is shaped not only by where people are but by how they circulate, converge, and anchor in urban space. This study advances that theoretical tradition by introducing the Spatiotemporal Population Mix (SPM)—a multidimensional construct that captures visitor-origin diversity, travel distance, stop frequency, dwell duration, and nighttime-resident share. Using year-long GPS traces from 166 census block groups in Arlington, Texas, three SPM profiles were identified via k-means clustering (Stable-Residential, Moderate-Mobility, and High-Transience) and evaluated with generalized spatial two-stage least-squares models. Block groups exhibiting a High-Transience SPM recorded violent, property, and drug-crime rates two-to-three times higher than Stable-Residential areas, net of social disorganization, land use, and spatial spillovers. Complementary continuous analyses confirmed that transient SPM facets—long travel, frequent stops, and diverse origins—elevate crime risk, while residential anchoring—long dwell and high nighttime-resident share—suppresses it. By demonstrating that the SPM explains crime above and beyond static population counts, the study refines routine activity and crime-pattern theory, offers a replicable behavioral metric for place-based research, and points practitioners to a small set of transient micro-areas that disproportionately drive urban crime.