no124-131

[131] The relationship between ambient population and crime count

Empirical analysis using the COVID-19 state of emergency as a natural experiment

○Mayuko Yamane¹, Mamoru Amemiya², Tomoya Ohyama³, Takahito Shimada⁴ (1. Geospatial Information Authority of Japan, 2. University of Tsukuba, 3. Center for Data-driven Science and Artificial Intelligence Tohoku University, 4. National Research Institute of Police Science)

Keywords:ambient population, crime, COVID-19, natural experiment, panel data analysis

Identifying the relationship between the “ambient population” and crime has been challenging due to a lack of data and the difficulty of causal inference. This study aims to reveal the relationship between the ambient population and crime through using the COVID-19 state of emergency as a natural experiment. The analysis uses the Poisson regression model with fixed-effects and the instrumental variable method. The results show that the effect of the ambient population on crime varies by crime type and neighborhood land use pattern. Furthermore, there was little difference in the effects of the day and night ambient populations on crime. This study is novel in that it verified the causal relationship between the ambient population and crime using the instrumental variable method.