What happens when a city increases the size of its police force?
According to research from University of Pennsylvania criminologist Aaron Chalfin and colleagues from the University of Oregon, the University of California, Los Angeles, and Barnard College, homicides and other serious violence decreases, but arrests for low-level offenses like liquor violations and drug possession increase.
Specifically, the researchers found that an additional 10 to 17 officers hired prevented one new homicide per year, a decline that’s twice as large for Black victims in per capita terms. Yet with each extra officer came anywhere from seven to 22 new low-level arrests. The researchers will publish their findings in a forthcoming American Economic Review: Insights paper.
Chalfin, who studies the costs and benefits of policing in the United States, has been interested in this subject for some time. As public attention on police shootings increased, and with it calls to defund the police, he started to wonder what tradeoffs might occur if a city were to spend less on its police force.
“There are some people who say, ‘We can invest less in police. They’re not doing much to control crime,’” Chalfin says. “But it’s a complex task to figure out what the effects of doing that will actually be.”
The research team knew that the FBI collects data on what are called “index crimes,” which include murder, rape, robbery, aggravated assault, burglary, larceny, and motor vehicle theft. The agency also has statistics on lesser crimes—known as “quality of life” crimes—that include disorderly conduct, liquor violations, loitering, and drug possession.
“Going back to 1977, there’s good data on arrests of particular types,” Chalfin says. “We know the racial demographics of the arrestees; we know the racial demographics of murder victims. We wanted to know, when the size of the police force increases, what happens to each of these variables?”
Overall, Chalfin and colleagues analyzed police employment data for 242 cities across the U.S. during the 38-year period from 1981 to 2018, factoring in federal grants to departments earmarked for hiring more officers. The study included cities with populations greater than 50,000 and only full-time sworn officers. For three categories—index crimes, quality of life crimes, and other—the researchers tracked total arrests as a whole and broken down by race.
Using an economic model, the team then drew a conclusion: “When a city increases the size of its police force, you get fewer crimes, fewer homicides, and fewer arrests for serious crimes, but more arrests for less serious crimes,” Chalfin says. Beyond that, “the homicides reduced are disproportionately where the victims would have been Black—which is maybe not surprising given that the demographics of homicide victimization skew that way.”
For context, the FBI reported 16,425 murders in the U.S. in 2019, or 5 homicides for every 100,000 people. Of the close to 14,000 where race was known, 7,484—nearly 54%—were Black. Yet Census data for that same year showed that Blacks and African Americans made up 13.4% of the total U.S. population.
“We live in a very unequal world,” says Chalfin. “Homicide victimization rates are not equal by race. When you throw more police into the mix, it’s not hard to imagine that the murders they abate are murders that would have had Black victims. But that’s not something we would have known definitively without the research.”
Though the work didn’t incorporate police management styles and training or account for composition and quality of the police forces, Chalfin says he hopes these findings can help guide policymaking.
At the very least, he says, they provide concrete evidence that in public safety redirecting funds from the police to other measures has, in the past, had consequences. “The research suggests that money spent on police is effective at reducing crime and violence,” he says. That being said, “when it comes to reducing violence, it’s crucial to have a portfolio of strategies.”
Aaron Chalfin is an assistant professor in the Department of Criminology in the School of Arts & Sciences at the University of Pennsylvania.
Other researchers on the paper included Benjamin Hansen of the University of Oregon, Emily K. Weisburst of the University of California, Los Angeles, and Morgan C. Williams, Jr. of Barnard College, Columbia University.
