Study: Van Dyke’s complaint history could have foretold shooting of McDonald

Jason Van Dyke could have been flagged as a problem officer who needed intervention to avoid a tragic incident long before he fired the 16 shots that killed 17-year old Laquan McDonald in 2014, a study of civilian complaints against Chicago police officers concludes.

The study of more than a decade of civilian complaints by Kyle Rozema of the University of Chicago Law School and Max Schanzenbach of Pritzker Law School at Northwestern University found that officers with the highest numbers of civilian complaints were more likely to be named as defendants in future civil rights lawsuits resulting in high payouts.

Van Dyke is scheduled to go on trial this week for murder charges for killing McDonald. Rozema and Schanzenbach found that the number of complaints against him from 2002 to 2014 put him in the top three percent of officers, a category that the authors found were likely to be the subject of civil rights lawsuits and large payouts.

Investigations that expose, influence and inform. Emailed directly to you.


Their study, “Good Cop, Bad Cop: Using Civilian Allegations to Predict Police Misconduct,” is to be published in a forthcoming edition of the American Economic Journal: Economic Policy.

Van Dyke recently suggested to reporters that shooting 17-year-old Laquan McDonald 16 times in 2014—an act for which he goes on trial tomorrow—was an aberration in his policing career, since he had never before fired his gun on the job.

The study authors, however, suggest that his significant number of complaints by civilians over the years were predictive of his use of force. They found a correlation between officers who, like Van Dyke, are among the three percent of officers with the most complaints, and later major incidents that lead to lawsuits and payouts. The authors, using the Empirical Bayes method of data analysis, found an even stronger correlation for officers who are in the top one percent of subjects of complaints; those officers “generate almost five times the number of payouts and four times the total damage payouts” in civil rights lawsuits.

While Chicago has long allowed civilians to make complaints about police behavior, those allegations rarely result in action by the department. Of the 2002-2014 allegations analyzed by Rozema and Schanzenbach, the Independent Police Review Authority (IPRA)—which has since been replaced by the Civilian Office of Police Accountability—“sustained” the complaint and recommended disciplining the officer only 2.4 percent of the time.

IPRA could only proceed to an investigation when the person making the complaint agreed to make a statement in a sworn affidavit. Rozema and Schanzenbach found that complaints without an affidavit were just as predictive of future officer behavior than those with an affidavit, suggesting that the affidavit requirement might be an unnecessary hurdle in the complaint process.

Representatives of police unions have downplayed the importance of civilian complaints, arguing that officers who work hardest and work in more dangerous neighborhoods are bound to rack up more complaints. Rozema and Schanzenbach’s analysis, however, found that switching districts did not affect an officer’s propensity to rack up complaints. Furthermore, officers with a lot of civilian complaints also had more complaints from supervisors and peers within the department about their performance on the job, suggesting that civilian complaints are not just a result of an officer working harder than everyone else.

The recently released draft of the Chicago consent decree—a court order requiring specific actions from the police department—would compel the department to make changes in its civilian complaint and officer intervention processes. Under this version of the decree, the police department and the city would have to make “best efforts” to ensure that lack of a sworn affidavit does not prevent a complaint investigation. And the department would be required to implement an automated, data-driven system identifying officers at risk of problematic behavior and notifying supervisors.