Identifying Changes in the Spatial Distribution of Crime: Evidence from a Referee Experiment in The National Football League.

Between the 2009-10 and 2010-11 seasons, the National Football League (NFL) repositioned one of its officials in order to prevent injuries amongst officials. This creates a quasi-experiment for studying how a change in the extent of policing affects detection of offences. Using play by play data from the 2009-10 and 2010-11 NFL season, I estimate how the detection of offensive holding changes when the positioning of an official changes. I find that there is approximately a 20 percent increase in the number of offensive holding penalties called after the NFL repositioned the official. Penalties called on defensive lineman fell as a result of the repositioning. Overall, there was no change in the total number of penalties called. Using the estimated change in the probability of a penalty, I estimate of the probability of an official calling a penalty. I infer that NFL officials detect approximately 60 percent of crimes committed on the field