Greenness and Crime in the City: An Investigation Using Remote Sensing and Spatial Panel Models

• Published in: Applied spatial analysis and policy Vol. 16; no. 1; pp. 229 - 257
• Main Authors: Bhatia, Shaun, Jason, Leonard
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.03.2023; Springer Nature B.V
• Subjects: Censuses; Criminal statistics; Estimation; Geography; Human Geography; Land use; Landscape/Regional and Urban Planning; Neighborhoods; Regional/Spatial Science; Remote sensing; Social Sciences; Violence; Violent crime; Greenness; Spatial panel model; Chicago; NDVI; Crime
• ISSN: 1874-463X; 1874-4621
• DOI: 10.1007/s12061-022-09477-9

Evidence has suggested that healthy environmental greenness is associated with lower rates of crime. The goal of this study was to explore this phenomenon from a longitudinal perspective while accounting for spatial dependency, and utilizing precision estimation for greenness and refined unit of place. Spatial panel fixed effects estimation was implemented to determine the association between greenness and crime rates in Chicago. Greenness was measured seasonally during the study period (2014–2018) using remote-sensed normalized difference vegetation index (NDVI). Using a spatial error specification with census block group and seasonal effects, higher levels of NDVI were significantly associated with lower overall crime rate (β = -0.24; p  < 0.001) and non-violent crime rate (β = -0.31; p  < 0.001) after adjusting for neighborhood and land-use covariates. Extensions of the current work may focus on more granular units of geography (e.g., city blocks or street segments) using ground-truthing, and cross-validation using alternative measures of greenness.