Examining differential officer effects in the Minneapolis domestic violence experiment

Reassesses the Minneapolis Domestic Violence Experiment (MDVE) data of the 1980s which said that “arrest works best”. Examines the potential effects of a small number of officers generating a relatively large number of MDVE cases. Expresses concern that a few officers submitting a high rate of cases might have an inordinately large effect on the experimental outcomes. Seeks to determine whether they followed the experimental protocol but finds instead that it was the low rate officers who failed to do so. Finds that officers who submitted experimental cases at a high rate were also different in other ways from the rest of the officers. Finally, excludes findings relating to lower case officers and thereby reveals that recidivism in domestic violence was not reduced by the arrests.