Research shows that more cops mean less crime

I’ve written a fair bit recently on the rise in violent crime in Minneapolis and, to a lesser extent, Saint Paul. As Figure 1 shows, there have been 56 homicides in Minneapolis so far this year, eight more than in all of 2019 and just one off of the 20-year high of 57, recorded in 2006. Figure 2 shows that, while Saint Paul’s total for 2020 has not surpassed last year’s yet, it has already exceeded the total for 21 of the previous 23 years.

Figure 1: Homicides in Minneapolis

Source: Federal Bureau of Investigation, City of Minneapolis, and Star Tribune

Figure 2: Homicides in Saint Paul

Source: Federal Bureau of Investigation, City of Saint Paul, and KSTP

How can we tackle this? Research offers a clear suggestion: more cops means less crime. Last year, Matthew Yglesias wrote a surprisingly good summary of the research for Vox: 

In a 2005 paper, Jonathan Glick and Alex Tabarrok found a clever instrument to measure the effects of officer increases through the terrorism “alert levels” that were a feature of the early to mid-aughts. During high-alert periods, the Washington, DC, police force would mobilize extra officers, especially in and around the capital’s core, centered on the National Mall. Using daily crime data, they found that the level of crime decreased significantly on high-alert days, and the decrease was especially concentrated on the National Mall.

…

About a year ago, Stephen Mello of Princeton University assessed the Obama-era increase in federal police funding. Thanks to the stimulus bill, funding for Clinton’s Community Oriented Policing Services (COPS) hiring grant program surged from about $20 million a year in the late-Bush era to $1 billion in 2009. The program design allowed Mello to assess some quasi-random variation in which cities got grants. The data shows that compared to cities that missed out, those that made the cut ended up with police staffing levels that were 3.2 percent higher and crime levels that were 3.5 percent lower.

This is an important finding because not only does it show that more police officers leads to less crime, but that actual American cities are not currently policed at a level where there are diminishing returns. Instead, reductions in crime seem to be about proportional to increases in the size of police forces.

A larger historical survey by Aaron Chalfin and Justin McCrary looked at a large set of police and crime data for midsize to large cities from 1960 to 2010 and concluded that every $1 spent on extra policing generates about $1.63 in social benefits, primarily through fewer murders.

It is crucial to note that more cops don’t mean less crime because more arrests are made. Instead, the mere presence of more police officers reduces crime by deterring crooks from committing them in the first place. As Yglesias notes of Glick and Tabarrok’s research: 

Critically, the finding was not that adding police officers leads to more arrests and then locking up crooks leads to lower crime in the long run. It’s simply that with more officers around, fewer people commit crimes in the first place. That seems to be the criminal justice ideal, in which fewer people are getting locked up because fewer people are being victimized by criminals.

This sounds a little paradoxical, but the reality is the size of the prison population is driven largely by the harshness of the sentencing, not the number of police stops. The criminologist Lawrence Sherman has observed that the United States is very unusual in spending much more money on the prison system than on our police departments. This suggests the possibility of switching to a formula Tabarrok has summarized as “more police, fewer prisons, less crime”: uniformed officers patrolling the streets stopping crime before it starts rather than working in prisons surveilling convicts.

A study by John MacDonald, Jeffrey Fagan, and Amanda Geller that looked at localized policing surges in New York City (dubbed “Operation Impact” by the NYPD) found that they led to both less crime and more “stop and frisk”-type incidents where officers stopped citizens (typically young black or Latino men) without probable cause. This suggests that a problematic trade-off between crime reduction and civil rights exists. But, Yglesias notes:

…the same study, by looking at the covariance of “stop and frisks” and crime reduction, found that the additional stops were doing nothing to reduce crime. All of the anti-crime impact, in other words, came from putting more cops on the beat rather than from the use of aggressive tactics. New York City, not coincidentally, has continued to enjoy low and falling crime rates since stop and frisk tactics were curtailed. What’s helpful is more officers, not more harassment.

Research also suggests that increasing the quantity of policing could increase its quality. As Yglesias notes, “The key mechanism here is fatigue…Tired officers, across a variety of studies, generate more complaints from the civilians they interact with.”:

What all this shows quite clearly is that proposals to ‘defund’ the police are especially bad ideas. If more cops means less crime, what does fewer cops mean? Faced with rising violence in the Twin Cities, we need more cops, not less.

John Phelan is an economist at the Center of the American Experiment.