Under Pressure? Performance Evaluation of Police Officers As an Incentive to Cheat: Evidence from Drug Crimes in Russia
Abstract: This paper provides an empirical analysis of manipulations of seized drug amounts by police officers, based on a unique dataset that contains full information on drug crimes in Russia reported during 2013-2014. First, using a bunching estimator, I document a significant excess mass of heroin cases above the punishment cliff. The mass is 6.325 times greater than the average number of cases in a counterfactual scenario without manipulation. Next, I employ an event study approach to investigate the incentives for police officers to manipulate and find that the motivation arises from the officers' performance evaluation system. One of the main indicators applied for evaluation is the number of serious and most serious drug crimes, which can be easily increased by moving offenders from below to above the threshold. Exploring the dynamics of this indicator during a calendar year, I document that it increases by 23% in the month when a police station is close to reaching the previous year's level of performance, current target. Comparing the performance evaluation systems of two separate drug control agencies, I find further evidence of this response to performance requirements. Finally, applying a novel bunching technique, I determine that police officers are more likely to manipulate the drug amounts seized from repeat offenders. The overall effect of manipulation on the sentence length of drug users is an additional year of incarceration, which is a 67% increase, compared to the average sentence length without manipulation.
