The Philadelphia Foot Patrol Experiment

Patrolling on foot to convey a sense of omnipresence and deter crime had been a cornerstone of policing dating back to the night watch. The Philadelphia Foot Patrol Experiment would challenge the assumptions head on, and provide new evidence that foot patrols deployed in hot spots can impact crime. The 60 beats that officers would ultimately patrol were identified using block randomization. Given a two-tailed test and a presupposed alpha of 0.10, 60 beats in the treatment condition and 60 beats in the control condition indicated that power was adequate when the effect size was large or medium, but low when effect size was small. The Philadelphia Foot Patrol Experiment’s statistical power was comparable to other randomized experiments testing the impact of hot spots or other place-based strategies. The results suggested that a beat needed to have a threshold level of violence prior to the intervention in order to detect statistically meaningful change.