note: this is the second in a series of posts about our ongoing work looking at 287(g) and the dynamics of racial profiling in traffic stops
After looking at the relationship between age, perceived race, and likelihood to be stopped by the police in North Carolina, I decided to break the traffic stop data we had access to down by time-of-day. Stops were coded either by a string giving either the 12 or 24-hour time, and after reconciling those two different formats I was able to pull out stop totals by hour throughout the day, distinguished by the perceived race of the driver. I divided each count by the total number of people in the state who reported that race on the census, and then divided each of those estimated likelihoods by the average likelihood for the whole population to get ‘normalized stop likelihoods’ — essentially an index of the degree of racial profiling, where a value greater than 1 means that somebody perceived as being from a given race is more likely than average to be stopped, whereas a value less than one means that person is less likely than average to be stopped.
What’s most interesting about the data here is up to you — I’m intrigued that the 9-5 workday is the period of greatest racial equality on the streets of NC (is capitalism really color-blind? do the police perceive drivers first as abstract labor selling itself on the open market and second as unique individuals?).