Investigating the Influence of Land Use and Alcohol Outlet Density on Crime in Juja Sub-County, Kenya

The main objective of this study was to investigate the linkage between land use and alcohol outlet density and crime in Juja sub-county, Kenya. Crime data (n = 1560) was obtained from the Juja Police Station for the years 2017, 2019, 2020, and 2021. Land use land cover classification of our study area was performed to obtain land use classes (commercial, agricultural, forest, grassland, industrial, residential, and waterbody), and zonal operations at the zone level (n = 233) were performed to obtain summary values of each land use type per zone. Alcohol outlet density was also calculated at the zonal level. Population was identified as a crime determinant factor in addition to land use and alcohol outlet density. From these factors, the 4 most significant ones (residential, agricultural, population, and off-premise outlet density) were identified using ordinary least squares (OLS) model. This revealed that off-premise alcohol outlet density had a statistically significant negative relationship with crime, while residential areas had the highest statistically significant positive relationship with crime. While on-premise alcohol outlet density did demonstrate the highest positive coefficient with crime, this relationship was not statistically significant. This infers that while on-premise alcohol outlet density may explain crime in areas where such establishments are dense, with high crime rates, they may not explain crime in other areas of the sub-county that equally recorded high-density crime rates. The random forest algorithm was then adopted to predict crime from the most significant variables.