In the debate over “defunding the police,” two criminologists at George Mason University’s Center for Evidence-Based Crime Policy examined more than 4 million 911 calls to see why people call for help. They found that shifting resources to mental health providers would only address a tiny percentage of calls and not necessarily justify moving large amounts of money away from daily police operations.

The killing of George Floyd by a Minneapolis police officer in May 2020 and other police killings of unarmed individuals — many of them people of color — have spurred calls for defunding the police. Proponents for defunding argue for shifting resources away from law enforcement to other public services to improve the quality of life in marginalized communities and reduce the criminal justice footprint. However, the idea of defunding police has been met with resistance by law enforcement and communities with high rates of crime and poverty, who argue they need more police protection, not less. Ultimately, defunding the police is an idea that needs a lot more thought.

Absent from the debate is adequate research into the scale or nature of issues that police handle and whether other government or nongovernmental agencies could effectively and fairly manage these problems better.

To address this gap, we studied nearly 4.3 million dispatched 911 calls from nine large, medium and small agencies across the United States and found that calls to police are voluminous. Police receive a significant number of calls, with rates around one or two calls per person per year. These calls are geographically and temporally concentrated, which means that for some, calling the police about persistent problems in their neighborhood may be a weekly — or even daily — event. The vast majority of these calls are for everyday concerns and conflicts, with only a very small proportion related to people in serious mental distress, which is a common rallying point for defunding arguments.

The most common types of calls police are dispatched to are traffic-related (about 17 percent of calls on average), followed by “disorders” (16 percent of calls on average). Disorders encompass several events that may or may not be crimes, including noise violations, animal disturbances, civil or other arguments, disorderly conduct, trespassing, abandoned vehicles, graffiti or vandalism, people who were drunk in public, loitering, fireworks, illegal dumping, burning, disturbing the peace, public urination and vagrancy.

People also regularly request that police handle suspicious persons and behaviors (about 13 percent of calls on average); again, some of these are not crimes, but situations where it is unclear who else to call. Another large proportion of calls (11 percent) are from people asking that police follow up on previous incidents or making other requests for service. Property crimes make up about 10 percent of calls on average, followed by calls for alarms (7 percent), violence (6 percent) and domestic disputes (6 percent).

For at least 80 percent of these millions of calls to law enforcement, people believe the police are the best — or perhaps the only — public agency that can respond. While some of these calls are certainly not crimes, they are not obviously transferrable to other organizations without significant expenditures of resources or adjustments in the current scope of work of other organizations.

Furthermore, contrary to public perception, calls related to people in mental distress accounted for only 1.3 percent of the dispatched calls on average across the agencies we studied. This finding is supported by systematic observations we have conducted in 911 call centers. Other research indicates that a majority of people refuse these services when offered. While mental health service providers may be needed to respond with the police to those events, shifting primary responsibility to another agency will have only very minor effects on demands for police service.

We also found that for a large majority of calls for police service, officers rarely made an arrest or wrote a report. Most of the time (between 62 and 83 percent of calls received), they provided assistance, advice or peacekeeping functions and took no further official action. For only 2 to 9 percent of calls across the nine agencies did a call result in an arrest or citation, and citations for traffic infractions made up the bulk of these official actions.

These findings raise several questions. Who besides the police would be able to handle these everyday concerns? What resources, staffing and training would be necessary for these agencies? Would other agencies be able to handle the calls safely, fairly and effectively? If an arrest or use of force is warranted, will other agencies be prepared, or will police still need to show up at these events, doubling commitments of government resources?

If defunding does not happen, our findings do not let the police off the hook from their responsibilities. People are upset about the long-standing disparities and injustices of our criminal justice system, and rightly so. Police will need to more widely adopt evidence-based strategies that are effective in preventing and reducing crime, disorder and traffic problems in the community, and that do not exacerbate justice disparities. How officers speak to people, the empathy and emotional intelligence they convey, and the procedural and distributive justice in their actions are at the heart of defunding discussions and will also need to be addressed if officers retain the main responsibilities they currently have.

The bottom line is that defunding the police is an idea that needs a lot more thought. As part of that work, we need to determine how prevention and investments in other organizations can play a role. And we need to identify better ways to assure accountability for the services that police retain.

Cynthia Lum is a professor of criminology, law and society and director of George Mason University’s Center for Evidence-Based Crime Policy. Christopher S. Koper is a professor of criminology, law and society at George Mason University and principal fellow of the Center for Evidence-Based Crime Policy.