Predicting Property Crime in Chicago: A Time-Series Analysis of Macroeconomic, Social, and Seasonal Factors

In 2025 alone, roughly 108,000 reported instances of property crime in Chicago have caused over 865 million dollars in estimated societal costs [1]. This paper aims to answer three questions central to criminology and economics studies in Chicago: (1) What factors have been associated with property crime historically? (2) How accurately can models forecast property crime on city and police district levels? and (3) In which police districts are future crimes concentrated? Our study uses Random Forest, Ridge Regression, and Autoregressive Distributed Lag OLS models to analyze how well socioeconomic factors, weather trends, and public unrest predict property crime. Namely, we discovered that property crime had significant positive associations with temperature, and negative relationships with unemployment, protest interest, and precipitation. Our SARIMA forecast anticipates persistent crime trends through 2029. Damped ETS suggests sizable movements of districts in their relative rankings of crime risk, although wide prediction intervals make projections uncertain. Using predictive analysis results in conjunction with forecasts identifying future high-risk districts, we propose recommendations for Chicago to mitigate victimizations and damages of property crime.