Managerial Accountability and Frontline Worker Safety: Evidence from New York City Policing Reform

• In Event: Evaluating Policing Strategies: Administrative Systems, Training, and Enforcement Outcomes

Abstract

Organizations routinely invest in accountability systems to improve productivity, operational efficiency, and organizational performance. These systems mitigate principal-agent problems by reducing information asymmetries and strengthening incentives for managers and employees to pursue organizational objectives. With tangible performance metrics and regular evaluation, accountability reforms increase the costs of poor performance while rewarding effective management. Although these reforms are typically implemented to improve efficiency, the same mechanisms that improve organizational performance may likewise influence worker safety. If improvements in accountability-oriented management similarly improve workplace safety, then the returns to organizational reform extend beyond efficiency alone. Although the returns to managerial accountability are often studied in private-sector organizations, similar incentive and information problems arise throughout the public sector. These challenges may be particularly pronounced in bureaucratic organizations, where hierarchical structures and delegated decision-making can weaken monitoring and create divergence between managerial and employee incentives, making the design of effective performance incentives particularly important. Police departments, in particular, operate in environments where managerial oversight influences not only organizational performance but also the safety of frontline officers, making them a useful setting for examining whether accountability reforms generate benefits beyond their intended objectives. This paper explores these second-order effects in a law enforcement setting. We leverage New York City's 1994 implementation of a data-driven police management system (Compstat) in a difference-in-differences framework, using data from the FBI's Law Enforcement Officers Killed and Assaulted (LEOKA) program and a control group of comparable large-city police departments to estimate a 30\% decline in officer assaults following Compstat adoption. The effects strengthen over time, are concentrated in team-patrol encounters and evening hours, and are robust to multiple specifications. We supplement the DD approach with the synthetic control method, which constructs a data-driven counterfactual for NYPD by reweighting control cities to match NYPD's specific pre-reform trajectory. The estimated time-varying effects from each approach strengthen over time suggesting that the Compstat program effectiveness increased with officers and supervisors becoming increasingly familiar with managerial practices and integrating them more fully into routine operations. Thus, while reforms are efficiency-driven, the returns to such policies are enhanced by improvements in workplace safety. This paper contributes to multiple strands of literature across labor, public choice, and crime. Although ours is a public sector setting, economically significant improvements in worker safety resulting from greater supervisor accountability may extend to other industries.

Authors

• Kaiyi Wen, University of Texas - El Paso
  Presenting Author

• Hyeran Chung, University of Texas - El Paso
  Non-Presenting Co-Author

• Xiaolong Hou
  Non-Presenting Co-Author

• Yichu Li, Binghamton University (SUNY-Binghamton)
  Non-Presenting Co-Author

• Rhet A Smith, University of Texas - El Paso
  Non-Presenting Co-Author