I have been working with Chicago crime data and its possible correlations with (and effects on) travel behavior for a few months. My recent findings suggest that in Chicago, CTA train stations have higher "crime attractiveness" compared to their surrounding areas. In other words, as moving away from a CTA train station, number of crimes decreases. Note that this only implies correlation and not causation.
The figure on the left illustrates a map of City of Chicago color coded with number of crimes in each Traffic Analysis Zone (TAZ) laid over a map of CTA train stations.
To test the hypothesis that train stations attract crime, I created three buffer zones (with 0.25, 0.5, and 1 mile radius) around each CTA stop and counted the number of "battery" crimes at each circle for every stop. Results show that the average crime density (ratio of # of batteries to area) around CTA stops decreases as moving away from the stop. Average crime density is 42% higher in a 0.25-mile circle around CTA stops compared to a 1-mile circle. There are of course many other factors such as economic activities, crowdedness, etc around transit stations that may play a role. Commonly accepted, certain types of crime like battery occurs where people are and train stations are by nature where people are. However, we can't yet answer the question whether all CTA stops have a high crime rate, or high crime rate areas have a CTA stop?
*Update 1: The map on the right shows crime density (# of battery crimes per area).
*Update 2: The figure below shows crime density (# of battery crimes per area) for three circles with radius of 0.25, 0.5, and 1 mile around a CTA station: As moving away from the CTA train station, crime density decreases.