Skip to main content
ABC News
Cities That Reduced Arrests For Minor Offenses Also Saw Fewer Police Shootings

In response to nationwide protests last summer over the murder of George Floyd by police, many cities and states have tried to change their approach to policing. One such strategy is to make fewer arrests for low-level offenses in an effort to reduce the number of potentially violent encounters between the police and the public. Virginia, for instance, banned police from pulling people over for exclusively minor traffic violations earlier this year. Meanwhile, Oregon decriminalized drug possession. Louisiana restricted police from making arrests for certain misdemeanors, asking police to instead issue summons. And cities like San Francisco and Portland, Oregon, began sending clinicians instead of police to help people suffering from a mental health crisis. 

These efforts are all part of a shift that has been underway in America’s largest cities for a number of years now. But despite the effect these changes have had in reducing violent encounters with the police, these efforts have still experienced significant backlash, particularly from some in law enforcement who have blamed rising murder rates on police “pulling back” and being “defunded.” This criticism has intensified as murder rates have risen, even though most cities’ police budgets weren’t cut by much in 2021 and murders still ticked up in cities that increased their police budgets. 

I’m a data scientist and founder of the Police Scorecard, a research group that analyzes policing data to better understand how to end police violence. We are still waiting for the federal government to publish arrests data for 2020 and 2021, but what we do know from previous years is that low-level arrests are in decline, and that appears to have helped reduce the number of shootings by police — not made violent crime worse. 

Data from the FBI Uniform Crime Report shows arrest rates have generally been declining since the 1990s, when crime rates were much higher than they are now. These declines have been accelerated, in more recent years, by changes in policing in America’s largest cities. For example, police departments serving 86 of America’s 100 most populous cities reported 30 percent fewer total arrests in 2019 than they did in 2013.Florida Department of Law Enforcement, which administers the state’s Uniform Crime Reporting program.

">1 This decline was particularly pronounced among low-level offenses, or offenses not involving crimes against people, sex offenses, weapons offenses or serious property and financial crimes. In America’s largest cities, arrests for low-level offenses declined during this period by 38 percent, with arrests for disorderly conduct, curfew and loitering violations, gambling, prostitution, drunkenness and liquor law violations all falling by more than 50 percent — the largest reductions of any offenses reported. It wasn’t just changes in policing that reduced low level arrests — policy changes enacted by city councils, local prosecutor’s offices and state legislatures likely contributed to these declines as well.

While national data hasn’t been published for 2020 yet, preliminary data suggests arrests in big cities declined even further beginning with the lockdowns in March and April 2020.

There has been, in other words, a shift away from “broken windows” policing, or the debunked idea that aggressive policing of minor crimes deters more serious crimes. But what has this shift meant for crime more generally and for reports of police violence?

Only 27 percent of the nation’s law enforcement agencies report data on police shootings to the FBI’s National Use-of-Force Data Collection program, and no agency-level data from this program has been made public. But the data that is available suggests that in cities where there were reductions in low-level arrests, there were also reductions in police shootings. 

Police misconduct trials are rare. Instead, cities pay millions to settle claims

Searching open data portals, internal affairs publications and media databases, I obtained data on fatal and nonfatal police shootings from 2013 to 2019 in 86 of America’s 100 largest cities.2 These cities reported a decline from 749 police shootings in 2013 to 464 shootings in 2019, a 38 percent decrease over this period.

And cities that cut low-level arrests by 50 percent or more, such as Milwaukee, Philadelphia and Newark, New Jersey, also saw some of the larger reductions in police shootings. There were 57 percent fewer police shootings in these jurisdictions in 2019 than there were in 2013. Notably, police shootings actually increased in some cities that made more low-level arrests, like Jacksonville, Florida, and Louisville, Kentucky.

To be sure, reducing low-level arrests isn’t the only reform that might have contributed to fewer police shootings. In this time period, the U.S. Department of Justice also initiated reforms in 12 major cities through consent decrees, MOAs or the Collaborative Reform Initiative,3 with cities that underwent these reforms, like Newark and Baltimore, making far larger reductions in low-level arrests and police shootings. These cities would have been more likely to implement policies intended to prevent unlawful, excessive or unconstitutional arrests, as well as new use-of-force policies restricting the amount of force used by police.

An orange background with a word bubble that reads: “The rarity of these convictions and our surprise even in the face of overwhelming evidence tells us how broken the whole system is.”

related: What Has — And Hasn’t — Changed Since George Floyd Was Murdered Read more. »

But it wasn’t just police shootings that decreased. Reported crime fell in jurisdictions that cut low-level arrests; in fact, it fell by just as much as those cities that made more low-level arrests. Consistent with recent research, cities that reduced low-level arrests did not experience an uptick in violent crime — or murder, specifically — compared to other cities during this period. Moreover, cities that made fewer arrests for low-level offenses did not see a substantial reduction in violent crime arrests, suggesting a more lenient approach to low-level offenses has not resulted in police being less responsive to serious public safety threats.

There is still substantial room for improvement, though. Despite the changes that have taken place, low-level arrests still made up 55 percent of all arrests reported in the nation’s largest cities, and 69 percent of all arrests nationwide, in 2019. In other words, there are still plenty of opportunities for cities to make fewer low-level arrests, which could help police shootings fall even further without compromising public safety. But unfortunately, the politics surrounding the increase in murder and violent crime in some of America’s cities could make it harder to pass these types of reforms, even though they’ve so far seemed effective.

Remembering George Floyd: A year of protest

Why serious changes to policing are unlikely to happen


  1. Population figures are from the American Community Survey’s 2019 five-year estimates. Police departments are limited to those for which we were also able to obtain data on police shootings in 2013 and 2019. New York City and Washington, D.C. did not report arrests data to the FBI Uniform Crime Report so individualized arrests data published online by these cities was used instead. And Philadelphia did not report 2019 arrests data to the UCR, so 2018 arrests data was used for this city. Six Florida municipalities — Jacksonville, Miami, Tampa, Orlando, St. Petersburg and Hialeah — did not report any 2013-18 arrests data to the UCR; 2013 arrests data for were obtained from the Florida Department of Law Enforcement, which administers the state’s Uniform Crime Reporting program.

  2. Data is collected at the incident level, so some incidents may involve multiple officers discharging multiple weapons. Some police departments and organizations release the total number of “officer-involved shootings,” while others provide a breakdown of incidents by the level of force used, including if and whether any firearms were discharged. In instances where a topline number of police shootings was not provided, we used the sum of total incidents where firearms such as handguns, rifles and pistols were discharged.

  3. Those cities are: Albuquerque, New Mexico; Baltimore; Philadelphia; Cleveland; Miami; Memphis; New Orleans, Newark, New Jersey; Milwaukee; San Francisco; Chicago; and Portland, Oregon.

Samuel Sinyangwe is a data scientist, policy analyst and co-founder of Campaign Zero.