As noted in an earlier post, Kansas City Missouri has opened access to their crime data. It’s tabular, but contains address locations.
We used the addresses in the crime data to map their locations. The map on the right-hand side of the following image show the locations of 98,325 crime incidences occurring between January 1st and October 24th.
After mapping (geocoding) the records we compared crime density to population density on a per square-mile basis.
The comparison is not conclusive, but it does indicate that, generally speaking, areas with higher population densities also have higher crime densities.
To account for the possible relationship between population density and crime, we created another map by dividing the crime density by the population density (on a square mile basis). The result of this is a map that puts to pattern the per-person crime rate.
Looking at crime-per person reveals a very different pattern from the original crime-density map—one in which the central city looks much safer. In this case, the areas around the airport and the east-bottoms stand out as locations where a person is most prone to crime within the city.
No doubt, there’s a bunch of ways to spin and criticize the methodology used herein. That said, there’s no denying that mapping crime data, or any other type of phenomena, is never as simple as dropping the raw data onto a map, pointing to the blobs that stand out, and claiming you’ve figured it out.









