Working Paper: NBER ID: w26438
Authors: Steven F. Lehrer; Louis Pierre Lepage
Abstract: Using data from the New York City Police Department's Stop-and-Frisk program, we evaluate the impact of a specific terrorist attack threat from Al Qaeda on policing behavior in New York City. We find that after the Department of Homeland Security raised the alert level in response to this threat, people categorized as "Other" by the NYPD, including Arabs, were significantly more likely to be frisked and have force used against them yet no more likely to be arrested. These individuals were in turn less likely to be frisked or have force used against them immediately after the alert level returned to its baseline level. Further, evidence suggests that these impacts were larger in magnitude in police precincts that have a higher concentration of mosques. Our results are consistent with profiling by police officers leading to low-productivity stops, but we cannot rule out that it constitutes efficient policing given important differences between deterrence of terrorism versus other crimes.
Keywords: NYPD; terror threats; stop-and-frisk; racial profiling; policing behavior
JEL Codes: C33; J15; K42
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Department of Homeland Security's alert level increase (Y50) | probability of being frisked for individuals categorized as 'other' (J79) |
| Department of Homeland Security's alert level increase (Y50) | use of force against individuals categorized as 'other' (K49) |
| Department of Homeland Security's alert level increase (Y50) | arrest rates for individuals categorized as 'other' (Y90) |
| alert level returned to yellow (Y50) | likelihood of being frisked (K42) |