AMHERST, Mass. – In the five years following the beginning of the Black Lives Matter (BLM) movement in 2014, census places that have had BLM protests experience a 15-20% decrease in police homicides – approximately 300 fewer deaths – according to new research by University of Massachusetts economist Travis Campbell.
Campbell found a benchmark 16.8% average reduction of lethal use-of-force due to BLM protests, a number that is, he writes, “remarkably stable.” He also found that the gap in police homicides between places with and without protests has widened over the years, and is most prominent when protests are large or frequent.
“The payoff for protesting is substantial,” writes Campbell, who is currently pursuing his doctorate in economics at UMass Amherst. “Every five of the 1,654 protests in the sample correspond with approximately one less person killed by the police over the following years. The police killed one less person for every 4,000 participants.”
Campbell presents his findings in “Black Lives Matter’s Effect on Police Lethal Use-of-Force” – an early, non-peer-reviewed version of which is available online via SSRN – and has made the data supporting the findings of his study openly available.
“The effects of Black Lives Matter protests have been debated often in public discourse,” Campbell says. “Many diverse and often polarizing opinions have been expressed about the impact of Black Lives Matter protests on policing and crime, but many such opinions are not supported by data. Despite the size and persistence of this social movement, there is a paucity of rigorous research about its impacts. I therefore decided to apply scientific methods to analyze available data to establish objective measures of how this movement is shaping our society, for better or worse.”
Using a stacked difference-in-difference design that leverages variation in BLM protests’ location and timing, Campbell contrasts four different estimators to overcome the potential challenges in uncovering the BLM’s effect on police lethal use-of-force, such as population-driven variance from the media neglecting protests or police homicides in less populated areas, and time-variant controls.
“While this reduction is robust to specification, estimator choice, choice of data, and population screens, it did not hold if lethal use-of-force is normalized by violent or total crime,” Campbell notes. “The fall in lethal use-of-force may be partially explained by expansions in police body cameras and community policing, tarnished community cooperation/reporting, and a so-called ‘Ferguson Effect’ – low police morale/effort leading to a rise in non-police homicides and fall in low-level arrests. Fewer property crimes, but more murders, are reported to agencies with local protests, while the share of total property crimes cleared by arrests falls.”
“Black Lives Matter protests were associated with a reduction in police officer-involved fatalities,” Campbell says. “While reduced police-related fatalities is a very desirable effect, it is important to understand the factors that contributed to this effect. My research found that decreased police-related fatalities after Black Lives Matter protests can be attributed to three factors: 1) Improved policing practices; 2) Damaged community trust of the police; and 3) Decreased policing effort. While improved policing practices are desirable and perhaps one of the intended effects of protests, decreased trust and reduced police effort have the potential to adversely affect communities by increasing crime. I hope that understanding both the desirable and undesirable effects of protests will lead to efforts by both the protest movement leaders and police leadership to maximize the good outcome of improving police practices while attenuating the undesirable effects of mistrust of police and reduced police effort. Such an approach would improve the benefit of protests for society.”