Police Open Up!

Detecting Racial Bias in Boston Police Vehicle Stops with the Stanford Open Policing Project

There is no national database for reporting all police activity data. Departments who do publish data often self report and there is no consistent format to examine data across departments nationally.

The Stanford Open Policing Project (OPP) is the first attempt to engineer a comprehensive police activity database at the national level. The OPP has cleaned and compiled data from numerous departments across the country and offers a tutorial for using the data to detect racial bias in police stops.

The tutorial works with data for Philadelphia and OPP conducts the tutorial in R. I became interested in the OPP because police bias has been a longtime concern of mine and because I wanted to learn R and this was a great way practice.

I first translated the tutorial from R to Python, a language with which I am much more familiar.

Once I had done so, I downloaded R Studio and replicated the original tutorial there.

My next step was to apply the tutorial to a new city. Since the OPP has not compiled data for New York City, I selected Boston, my original hometown. Before returning to R, I applied the tutorial steps to the new dataset. Because some inconsistencies in the OPP data still exist between location datasets, each step needed careful modifications to function correctly.

The Stanford OPP may be found here and my work on this (and other projects) here.