Primary Submission Category: Design-Based Causal Inference
Crime, Salience, and the Housing Market
Authors: Nency Dhameja,
Presenting Author: Nency Dhameja*
How do housing markets respond to nearby crime? Do buyers price crime based on objective risk, or do they react more strongly to highly visible events? Using 347,455 repeat home sales from 148,349 Chicago properties between 2008 and 2022, matched to 4.7 million geocoded crime incidents, this paper examines how crime is capitalized into housing prices.
Preliminary analysis across 120 distance–time specifications shows that 109 (90.8%) produce negative point estimates, with the largest effects concentrated at the shortest distances and most recent time windows. A near–far identification strategy sharpens this descriptive pattern into a causal test by comparing crime occurring immediately around a property (0–0.05 miles) with crime slightly farther away in the surrounding neighborhood (0.05–0.30 miles), under property and Community Area × Month fixed effects. Crime within the immediate block significantly reduces sale prices by −0.195%, while crime slightly farther away has no detectable effect, indicating highly localized capitalization. Within violent crimes, daytime incidents affect prices, whereas nighttime incidents of the same categories do not. Taken together, the results suggest that housing markets respond disproportionately to visible, unexpected crime events, producing block-level price distortions driven by salience rather than long-run risk.