An Analysis of the New York City Police Department's “Stop-and-Frisk” Policy in the Context of Claims of Racial Bias

Gelman, Fagan, and Kiss (2007) produce estimates that-by the authors' calculations-imply 32% of black-civilian stops made by the NYPD could not be explained even by differential criminality between racial groups of suspects, as proxied by prior arrest rates.

“Using the SQF data and controlling for precinct, suspected crime, and prior local arrest rates by race, Gelman, Fagan, and Kiss (2007) produce estimates that -by our calculations -imply 32% of black-civilian stops made by the NYPD could not be explained even by differential criminality between racial groups of suspects, as proxied by prior arrest rates.”

These stops and searches are less likely to lead to a finding of contraband or arrest, indicating poorer accuracy (lower hit rates) when young minority men are targeted.

“These stops and searches are less likely to lead to a finding of contraband or arrest, indicating poorer accuracy (lower hit rates) when young minority men are targeted (Gelman, Fagan, and Kiss 2007;Goel et al. 2017).”

Black individuals are more likely to be stopped by police than White individuals, even after a variety of statistical controls have been applied.

“Both recent and decade-old data show that Black individuals are more likely to be stopped by police than White individuals (Fryer, 2016; Gelman et al., 2007; Miller et al., 2017; U.S. Department of Justice, 2016), even after a variety of statistical controls have been applied.”

If a civilian's race affects whether officers choose to stop that civilian, then analyzing administrative police records amounts to conditioning on a variable that is itself affected by suspect race.

“If a civilian ' s race affects whether of fi cers choose to stop that civilian (Gelman, Fagan, and Kiss 2007; Glaser 2014), then analyzing administrative police records amounts to conditioning on a variable that is itself affected by suspect race, namely, whether a suspect appears in the data at all.”

The analytic approach of Gelman, Fagan, and Kiss (2007) most likely underestimates the proportion of racially discriminatory stops, so the implied value of r is conservative.

“This analytic approach most likely underestimates the proportion of racially discriminatory stops -the number of prior arrests in a precinct and racial group is not a direct measure of criminality, but is itself likely contaminated by discrimination in previous detainments and arrests. We,therefore, regard the value of r implied by Gelman, Fagan, and Kiss (2007) as conservative.”

Racial disparities persist even after leading candidate omitted variables, such as differential criminal activity across racial groups, are accounted for.

“A raft of studies have also found that racial disparities persist even after leading candidate omitted variables, such as differential criminal activity across racial groups, are accounted for (Gelman, Fagan, and Kiss 2007).”

These policing efforts generated racially disparate impacts on police stops and arrests of Black and Hispanic residents.

“Although there is evidence that these NYPD strategies were associated with modest crime control gains (MacDonald et al., 2016;Weisburd et al., 2016), numerous studies suggested that these policing efforts generated racially disparate impacts on police stops and arrests of Black and Hispanic residents (e.g., Fagan & Davies, 2000;Gelman et al., 2007;Jones-Brown et al., 2010).”

Data from New York indicated that blacks and Hispanics are more likely to be stopped and frisked than whites, after controlling for precinct and rates of crime commission.

“Data from New York, one of the jurisdictions that collected stop and frisk data because of a court order, indicated that blacks and Hispanics are more likely to be stopped and frisked than whites, after controlling for precinct and rates of crime commission (Gelman, Fagan, and Kiss 2007;Goel et al. 2017).”

The corresponding values of r from Gelman et al. and Goel et al. are 0.32 and a lower bound of 0.34, respectively, when considering black civilians.

“The corresponding values of r from these two studies are 0.32 and a lower bound of 0.34, respectively, when considering black civilians.”