THIS IS WHAT TRANSPARENCY LOOKS LIKE: AN EMPIRICAL ANALYSIS OF NYPD MISCONDUCT AFTER THE REPEAL OF 50-A

Abstract

This Note presents the first empirical study of the implications of the repeal of Civil Rights Law § 50-a ("50-a"), which made public New York Police Department ("NYPD") personnel records, including disciplinary investigations. These data demonstrate the limited potential of transparency reforms, which are lauded as an important step toward increasing police accountability but do little to impact that actual behavior of police officers. Using a version of regression discontinuity design known as interrupted time series, this Note demonstrates that the repeal of 50-a did not live up to its promise of reducing police misconduct. These findings illuminate the disconnect between the professed purpose of this particular legal change and its actual impact. However, as this Note also demonstrates through a text-mining approach called topic modeling, journalistic coverage of NYPD misconduct did increase after the repeal. This is framed as a second-order effect of the transparency measure-not reducing police misconduct ex ante but instead publicizing it after the fact. These data show that transparency on its own to cannot bring meaningful change in policing, but it can lead to a more informed public, thus playing an important role in conceiving and implementing policies that will reduce the harms of policing.

Keywords: NYPD, Police Misconduct, Transparency, Interrupted Time Series