In 2015 and 2016, the coincidence of a major surge in homicides following mass protests over the police killing of Michael Brown in Ferguson, Missouri, prompted a heated debate about whether the demonstrations, and the anti-police hostility they engendered, helped cause the murder spike. Law enforcement leaders and some public commentators—including City Journal contributing editor Heather Mac Donald—identified a “Ferguson Effect,” whereby public scrutiny reduced police proactivity and led to an increase in violent crime. Supporters of the protests just as fervently derided the idea as imaginary and “debunked.”

Social scientists made a few contributions to this debate, but the research they produced offered limited insight into the causal relationship between scrutiny, proactivity, and crime. Two new studies, however, rely on better data sets and methods to provide strong evidence that highly scrutinized officer-involved fatalities reduce discretionary police activity and lead to an increase in violent crime.

The more-recent study, just published in the Journal of Public Economics by university economists Cheng Cheng and Wei Long, looks at the effect of Brown’s death on police activity and crime on a week-to-week level in St. Louis (which is near Ferguson), and on a month-to-month level in 60 big cities. The St. Louis police department collects high-quality data on self-initiated activity, the authors note, allowing them to assess in specific detail police behavior just before and just after Brown’s death.

Their findings manage somehow to be both unsurprising and shocking. In the immediate aftermath of Brown’s death, self-initiated arrests fell 62 percent. Similar declines are seen across nine out of 11 categories of self-initiated activities, including foot patrol (down 82 percent) and pedestrian checks (76 percent). Notably, the decline in arrests is concentrated among misdemeanor arrests (more discretionary than felonies) and among arrests of blacks (rather than whites). This reduction in police activity persisted for at least the next two years. In the same period, the city experienced a significant rise in homicide and aggravated assault.

The multicity analysis shows a similar pattern. Cheng and Long assume that a Ferguson Effect will be more common in cities with higher black population shares, so they compare those in the top quartile of black population against the other cities as a control. They identify the same phenomenon as in St. Louis: the blackest cities saw larger declines in arrests relative to those with smaller black populations, driven by a decline in misdemeanor arrests, and a simultaneous large (10 percent) increase in homicide.

Cheng and Long’s observations are consonant with a drop in police activity and an increase in violent crime following the events in Ferguson. But what about the effects of officer-involved fatalities in general? A second study, by Deepak Premkumar of the Public Policy Institute of California, compares arrest patterns among police departments in 52 cities that experienced high-profile shooting incidents to 2,688 police departments that did not—and comes to a similar conclusion.

Just as in Cheng and Long’s study, Premkumar finds that police reduce their activity—as measured by arrests—after a high-profile incident, but not evenly. He finds large average declines in arrests for minor offenses (up to 33 percent for marijuana possession) but no significant decline in arrests for major offenses, whether violent crime or property crime. At the same time, cities that experienced high-profile incidents, relative to those that did not, saw a 10 percent to 17 percent increase in murders and robberies, as well as small increases in theft and grand theft auto.

Both studies not only identify a decline in police activity and an increase in homicide after a high-profile incident but also argue that the nature of the decline in activity supports a “public scrutiny” mechanism for the effect. In particular, a drop in police activity for minor offenses drives the overall policing pullback. This fact supports a theory (as articulated by Premkumar) whereby officers reduce arrests that they deem “less important” in order to minimize public contact. That differs from a competing theory under which a loss of police “legitimacy”—which, in turn, reduces community engagement with the cops—drives the Ferguson Effect. If the competing theory were borne out, the study should have found reductions in arrests for offenses of all kinds.

In other words, these two studies provide good evidence that high-profile policing incidents and the ensuing scrutiny drive discretionary police activity down and violent crime, especially homicide, up. That doesn’t settle the debate, of course. It remains an open question why a reduction in enforcement of petty crimes would lead to an increase in serious ones like homicide. But it does give support to those who propose a relationship between 2020’s wave of anti-police protests and the ensuing homicide wave still sweeping the country—and to those who believe that continued hostility to the police will lead to more bloodshed.

Photo by Michael B. Thomas/Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next