George Floyd’s death has elicited many proposals for improving America’s criminal justice system. Some of these, such as banning public employee unions, ending qualified immunity, and reducing our currently absurd levels of criminalization, are good ideas that ought to be pursued. However, another proposal, defunding the police, is an unwise idea that ought to be taken off the table.

Far from solving problems like police misconduct and over-incarceration, spending reductions will make those problems worse and will, moreover, inflict additional harm on the communities that have already been harmed the most by the turmoil that followed Floyd’s death. If we want to improve conditions in those communities and ensure there are fewer instances of police misconduct and less incarceration, we need to hire more police officers.  And we need to pay them more. In the long run, those measures may save money because we’ll have to spend much less on jails and prisons. For now, however, hiring more police officers and paying them more will require more funding, not less. That’s not going to be a popular course of action right now, but it is what we ought to do.

Suggesting that we ought to respond to anti-police protests by spending more on the police may sound paradoxical, but it’s not. Putting more active-duty police officers in the field will mean fewer crimes. Fewer crimes will mean fewer arrests and convictions. And fewer arrests and convictions will mean lower levels of incarceration. Similarly, higher pay scales will attract a larger and better-qualified pool of applicants to police programs. A larger and better-qualified pool of applicants will reduce the incentive for police programs to keep or rehire bad actors and improve the overall level of professionalism. And police officers who maintain higher professional standards will be less likely to misbehave. Finally, lower crime rates and more professional policing will make it easier to rebuild the neighborhoods that have suffered the most from the recent unrest.

There’s nothing new or speculative about what I’m suggesting. In 2006, Jonathan Klick and Alex Tabarrok published an article in the Journal of Law and Economics that described the results of an analysis of crime data for Washington, D.C. “Using a variety of specifications,” they said, “We show that an increase in police presence … leads to a statistically and economically significant decrease in the level of crime.” Tabarrok returned to the subject in 2013 in a piece called, “More Police, Fewer Prisons, Less Crime,” and again in 2016 in response to an Obama administration report called, “Economic Perspectives on Incarceration and the Criminal Justice System.” Most of the Obama report was devoted to data pertaining to crimes, arrests, and incarceration in the United States. However, as Tabarrok noted, it did provide the following chart comparing the U.S. to the rest of the world in terms of criminal justice system employees per 100,000 residents:

The report didn’t comment on the significance of the chart beyond noting that, during the relevant period, the United States simultaneously had “the largest prison population in the world and … employed over 30 percent fewer police officers per capita than other countries.” As Tabarrok pointed out, however, the chart is consistent with his hypothesis. Indeed, to me, the chart suggests an intriguing possibility. Maybe the incarceration rate in the U.S. is higher than in other countries because our crime rate is higher, and perhaps our crime rate is higher because we have fewer police officers on the street discouraging crime.

Tabarrok returned to the subject yet again last year in response to a report that compared the ratio of police to prison spending in the United States and Europe. Summarizing the report’s findings, Tabarrok said, “The US spends less on police and more on prisons than any European country … [and] not because Europe spends less on criminal justice.” He also provided the following chart, which illustrates the extent to which our spending focuses on prisons rather than on police and also shows that Europe is “not noticeably below the US states on total spending as a percent of GDP.”

If it’s true that more police means less crime, then we’re going to need many additional police officers in the next few years, especially in the areas where the recent riots have occurred.  During otherwise peaceful protests, crimes were committed by opportunists for whom the protests provided convenient cover. Moreover, contrary to what some have suggested, the crimes associated with the protests weren’t limited to crimes against property. Homicide and other violent crimes also spiked.

The immediate result is inevitable. Many of the businesses that were vandalized and looted will leave. People who can afford to relocate to safer neighborhoods will leave too. New businesses and new residents will be reluctant to take their place. And, unless crime levels quickly come down and stay down, those neighborhoods will probably suffer a long period of decline similar to those that occurred in Detroit after the 1967 riots and in Los Angeles after the riots that occurred in 1992.

Sadly, the evidence from the cities that have experienced anti-police riots in more recent years suggests that crime levels are likely to remain high, and, incidentally, that evidence also tends to corroborate the more police/less crime theory. In “Policing the Police: The Impact of ‘Pattern-or-Practice’ Investigations on Crime,” Harvard economists Tanaya Devi and Roland G. Fryer Jr. compare crime levels in 27 cities before and after investigations of police misconduct. In 22 of these cities, the complaints that led to the investigations did not receive much national media attention. However, in Baltimore, Chicago, Cincinnati, Riverside, CA, and Ferguson, MO, the investigations were triggered by what the authors call “viral” incidents. As the authors explain, the incidents, each of which involved the use of deadly force against an African American civilian, “caught national media attention and the cities witnessed protests and riots soon after.”

Devi and Fryer found that the 22 investigations that were not preceded by viral incidents led to small reductions in homicide and total crime. “In stark contrast,” they say:

…all investigations that were preceded by “viral” incidents of deadly force have led to a large and statistically significant increase in homicides and total crime. We estimate that these investigations caused almost 900 excess homicides and almost 34,000 excess felonies. The leading hypothesis for why these investigations increase homicides and total crime is an abrupt change in the quantity of policing activity. In Chicago, the number of police-civilian interactions decreased by almost 90% in the month after the investigation was announced. In Riverside CA, interactions decreased 54%. In St. Louis, self-initiated police activities declined by 46%.

Think about that for a moment: 900 excess homicides and 34,000 excess felonies. Those are huge numbers! Moreover, while the authors don’t provide any details about the victims, given the demographic make-up of the cities themselves and differential crime rates in general, we can be quite sure that those victims were disproportionately African American and, to a lesser extent, Hispanic. Nor is the direct suffering of those victims and their families the whole story. Those excess crimes coming on top of the riots that preceded them have undoubtedly harmed all of the residents of those cities, including the African American and Hispanic majorities, by depressing property values and discouraging investment and job creation.

We need to be objective and honest about this. Crime rates in America have declined significantly since the 1990s, but they are still much higher than in other industrialized countries. They are still much higher than they were in the middle of the 20th century. And they are particularly high in African American and Hispanic communities. Those elevated crime rates harm everyone, but they mainly harm the residents of African American and Hispanic communities. And the recent riots will inevitably make both the crime and the harm it causes worse in the short run. If we want to ensure that the riots don’t make those problems worse in the long run as well, we shouldn’t be talking about defunding the police. Instead, we should be talking about how to provide more police funding so that we can hire more police officers and pay them higher salaries.