New Research Highlights the Cost of Police Pursuits
Just a few weeks ago, two papers were released providing new and interesting insights into the impact of police vehicle pursuits. There is, of course, a rich history of research and evaluation of police pursuit policy going back decades, and this newsletter edition isn’t going to wade into that history. But both of these new papers help contextualize this issue and the degree to which law enforcement agencies should probably err on the side of caution when establishing and enforcing pursuit policies.
The first paper finds that fatalities from police chases have risen a ton over the last 15 years. The paper, titled Police Pursuit Fatalities in the US, 2009 to 2023 by Andrew Hendrix, Tamriage Martin, Joshua Gazzetta, and Jade Nunez, was published on April 1st. The paper shows that, between 2009 and 2023, deaths from police chases “increased 2% annually after adjusting for population and crash frequency. More fatal pursuit crashes occurred at night, in urban areas, on noninterstate roads, and in southern states.”
It’s an approachable study that uses data from the National Highway Traffic Safety Administration’s Fatality Analysis Reporting System (FARS). FARS has a handy tool called the Fatality and Injury Reporting System Tool (FIRST), and you can use the NHTSA’s FARS’ FIRST to evaluate traffic accident characteristics through 2024.
The tool only has police pursuit data on fatal crashes, but expanding out to 2024 shows just how much fatalities in crashes involving a police pursuit have increased over the last few years. Pedestrians and cyclists make up less than 10 percent of all fatalities in accidents involving a police pursuit
Of critical importance, there are limitations with this tool. We can’t use the FARS data to evaluate (as far as I know) the degree to which the pursuit contributed to the crash, whether the officer was injured or killed, any information on non-fatal crashes, and many other factors related to the pursuit. It’s simply a binary yes/no involved a police pursuit. Still, the dataset makes clear that more people are dying each year in crashes involving a police pursuit.
The second paper of interest, published in late March, came from John Hall and Justin Nix and it is titled Fast and Spurious: A Causal Evaluation of NYPD’s Pursuit Policy, 2022–2025. This paper looked at a massive increase in pursuits in New York City between 2022 and early 2025.
They write that “Monthly pursuits surged from roughly 8 to over 160 events — a rise exceeding 2,000%. Pursuit-related collisions increased proportionally with pursuit volume and declined sharply following the February 2025 partial re-restriction that reduced pursuits by approximately three-quarters.” The surge in collisions in New York City was in “direct proportion” to the surge in pursuits which suggests a strong (and obvious) causal link between increasing pursuits and increasing crashes.
Hall and Nix then looked at whether the surge in pursuits potentially produced a drop in crime — specifically looking at shootings at robberies. There they find that more pursuits “produced clear and attributable collision harms while providing no credible evidence of crime reduction.” Moreover, there was no subsequent increase in crime when pursuits were partially restricted again in February 2025.
There are obviously times when a police pursuit is essential for public safety purposes, but employment of that tactic must always be balanced with the potential public harms. Even very brief pursuits can lead to erratic driving and eventual loss of life long after the pursuit has been called off. In one case where I am, for example, a woman was chased after brazenly shoplifting from a local mall. Police pursued her vehicle for only a short time, but she drove erratically for several miles and eventually crashed into another car killing two teens.
I’m only scratching the surface on both papers and I’d encourage anyone interested to give them both a read. Both papers offer the sensible next step for police departments to strongly consider their pursuit strategy and requirements given the very real-world potential harms.
The paper from Hendrix, et al concludes that their findings “support risk-based restrictions, mandatory national pursuit reporting, and investment in nonpursuit alternatives among US police.”
Hall and Nix are a bit more specific, writing “At least in the context of the NYPD escalation, the evidence does not resolve the trade-off in favor of the pursuit policy. The crash costs are demonstrably real, and the crime prevention benefits are not demonstrated.” Agencies, therefore, “should require affirmative evidence of crime-prevention benefits before expanding pursuit authority.”
There are steps that agencies and policymakers can take to ensure that police pursuits only occur when absolutely necessary. Serendipitously, while prepping this newsletter I came across a piece in the local paper highlighting the various ways that law enforcement agencies in Louisiana are using technology to limit chases.
I don’t have much trust that increasing the punishment of fleeing from an officer will do much to reduce police chases, but spending resources to enhance officer training and purchase new technologies that might make pursuits safer seems like the right track that governments everywhere should be considering.
New On The Jeff-alytics Podcast
I talk a lot about crime analysis and analyzing crime data, but I realized that I’ve never actually talked to a crime analyst yet. To correct this, I’m talking with Carlee Ruiz, a former crime analyst and policing researcher, discussing how she bridges the gap between academic research and real-world law enforcement through her platform, Police Research Hub.
And while you’re here, be sure to check out these other recent great episodes:
Council on Criminal Justice President Adam Gelb
Baltimore Mayor Brandon Scott
Filmmakers Ferne Pearlstein and Bob Edwards
Professor and researcher Ian Adams
I'm very open to these sorts of conclusions but a little skeptical that we can draw them from the research here. A few things seem hard to establish in the data:
1) Are the long-term trends from driven by increases in reporting and associating deaths properly with police chases? I know, for example, that police killing databases and records have gotten much more thorough, and by default include many cases (such as chase fatalities) that tended to be ignored in the past. I see the first paper claims to mitigate this to a degree but it's hard to say if they truly do.
2) Would the NYC apply nationally? The transportation and crime environments in NYC are atypical to say the least. I'm also not at all surprised to find it may not have done anything from a deterrence pov, but it's also weird that they only looked at harms and deterrence. I would think arrest and conviction rates would also matter.
It makes a lot of sense that any individual police pursuit is net negative in terms of crime/safety.
But taken collectively, my concern is that a blanket policy of fewer (or no) pursuits would lead to a significant increase in people fleeing police once word gets out that you can just drive dangerously for a few minutes and the police will back off.
I feel like we saw the same thing with shoplifting, once thieves figured out that the security guard at CVS won't lift a finger and you can just load up a trash bag of stuff and walk out the door. Or fare jumping on the subway. Does it really make sense for the police to chase down one guy over a $3 fare? Individually probably not, but unless you do it sometimes, people will start doing it everywhere.
I think technology is the answer here. There's no issue with letting a guy go if a drone is able to follow him home.