Protesting the police: anti-police brutality claims as a predictor of police repression of protest

Abstract

Police face a unique dilemma when policing protests that explicitly target them, such as the anti-police brutality protests that have swept the United States recently. Because extant research finds that police response to protests is largely a function of the threat – and especially the threat to police – posed by a protest, police may repress these protests more than other protests, as they may constitute a challenge to their legitimacy as a profession. Other research suggests police agencies are strongly motivated by reputational concerns, suggesting they may treat these protests with special caution to avoid further public scrutiny. Using data on over 7,000 protests events in New York over a 35-year period from 1960 to 1995, I test these competing hypotheses and find that police respond to protests making anti-police brutality claims much more aggressively than other protests, after controlling for indicators of threat and weakness used in previous studies. Police are about twice as likely to show up to anti-police brutality protests compared with otherwise similar protests making other claims and, once there, they intervene (either make arrests, use force or violence against protesters, or both) at nearly half of these protests, compared to about one in three protests making other claims.

Keywords:

• Protest policing
• social movements
• police brutality

Acknowledgements

I would like to thank Jennifer Earl for her helpful feedback and guidance. I would also like to thank the participants and discussants at the Young Scholars Conference at Notre Dame in 2015 for their feedback on an earlier version of this paper.

Notes

1. Models run on the entire data-set, with events from across the U.S., were also run. The findings were very similar to those of the models restricted to New York state.

2. I also ran a regular logistic regression on police intervention, which dropped all cases where police were not present. The results were comparable to the Heckman model results. While no variables changed in their significance, the magnitudes of some of the coefficients were larger without the Heckman correction.

3. The codebook defines force as ‘any physical tactics [by police] during their activity at the event’, whereas violence is defined as ‘attacking protesters, or us[ing] equipment such as guns, tear gas, nightsticks, or riot control equipment’ (McAdam, McCarthy, Olzak, & Soule, 2009).

4. Police are reported to have made arrests at only about 3% of protests, used violence at about 8% of protests, but used some sort of physical force at about 26% of protests.

5. Logged to adjust for heteroskedasticity.