A New Paper Rebuts the Notion of a ‘Ferguson Effect’

St Louis County police officers watch as anti-police demonstrators march in protest in Ferguson, Missouri
Photo: Lucas Jackson/Reuters/Corbis

At the moment, there’s an “escalating dispute,” as The Hill put it, between FBI Director James Comey and former Attorney General Eric Holder over the so-called “Ferguson effect.” Believers in the Ferguson effect argue that police, under fire after the violent deaths of Eric Garner, Michael Brown, and other unarmed African-Americans — many of them caught on video — are now too demoralized and/or worried about the repercussions of using force that’s later deemed to be unnecessary to do their jobs properly, which has led to an increase in crime. In a couple of recent speeches, Comey has endorsed this notion, while Holder disputed it to the Associated Press, calling it “anecdotal.”

If the Ferguson effect exists, it would be a difficult thing to prove the existence of. For one thing, while 2015 has seen a spike in murders in certain cities, there just isn’t evidence of a broader post-Ferguson uptick in crime. So any argument about the Ferguson effect would have to center around a few cities, and neither the timing nor the location of the upticks lines up in a manner that clearly supports the notion. As of August, for example, New York City had seen just a 9 percent uptick in murders since 2014, and New York is where Garner’s death — arguably the most dramatic such incident caught on video — occurred.

For another thing, as any researcher can tell you, when crime does go up or down, it’s incredibly difficult to figure out why. As Dana Goldstein laid out last year at the Marshall Project, even when it comes to the heavily studied decades-long drop in violent crime that led to today’s generally low rates (by U.S. standards), there are a great deal of theories, and experts debate them fiercely. So most researchers respond with skepticism to the idea that localized upticks in murders can be laid at the feet of a single national phenomenon. It’s likely that things are much more complicated than that.

Still, some researchers are trying to test the notion of a Ferguson effect. In one recent study, published online earlier this month in Law and Human Behavior, Scott E. Wolfe of the University of South Carolina and Justin Nix of of the University of Louisville don’t attempt to study a direct connection between law enforcement personnel’s perception of how they’re viewed and crime rates, but rather between their perception of how they’re viewed and their willingness to work with communities to reduce crime — which the researchers think, for common-sense reasons, might itself be tied to crime rates.

Wolfe and Nix conducted an online survey of 567 deputies who worked at a metropolitan jurisdiction in the Southeast, asking them about four different types of things: to what extent they felt like “recent negative publicity has harmed their motivation”; their assessment of their department’s level of organizational fairness (which included issues like how well commanders treat their subordinates); their assessment of their own and their department’s legitimacy (to what extent they felt confident in their own authority and that they occupied an important place in society); and how willing they were to “engage in community partnership.”

At first glance, the relationship that popped out indicated a Ferguson effect: The more the deputies felt like negative publicity had harmed their motivation, the less desire they expressed to work with communities. But once the researchers controlled for the fairness and legitimacy issues, this relationship was no longer statistically significant — “washed away,” as they put it. In other words, the deputies’ views of their own departments were what mattered, not the extent to which they sensed they were being criticized. The more they felt their departments were fair places and that their work had a sense of legitimacy, the more likely they were to report wanting to partner with communities to lower crime. 

As with any study of a complicated behavioral issue, there are caveats aplenty here. The paper was entirely based on self-reported survey items (rather than actual crime rates or indicators of real-world police behavior) and only covered one jurisdiction. Still, given the lack of other evidence for the Ferguson effect — and the importance of understanding the difference between effective, trusted police departments and ineffective, distrusted ones — it’s a useful piece to add to the broader puzzle.