Photo by Acton Crawford on Unsplash
The ZIP Code Factor in Fatal Police Pursuits
In A Nutshell
- Counties with the fewest resources had a 63% higher rate of fatal police chases than wealthier counties, even after accounting for local crime rates.
- Researchers analyzed five years of data covering more than 3,100 U.S. counties and over 3,000 fatal chases.
- Weak schools, poor health and environmental conditions, and low economic opportunity were each independently linked to higher rates of fatal chases.
- The study cannot prove that living in a low-opportunity county causes fatal chases, only that the two are strongly linked.
A person dies every day in the United States because a police chase went wrong. Sometimes it’s the driver being pursued. Sometimes it’s a passenger. And sometimes, like in a 2023 case in Chicago that ended with a $22 million settlement, it’s a bystander who had nothing to do with the chase at all. A new study suggests that where someone lives is linked to whether their community becomes the backdrop for one of these tragedies.
Researchers examined five years of data from more than 3,100 counties across the country and found a clear pattern: counties with fewer resources, weaker schools, more pollution, and less economic opportunity saw far more fatal police chases than wealthier, better-off counties. Even after accounting for how much violent crime each county actually had, the gap didn’t close. It shrank, but it stayed large.
This isn’t the first time researchers have found that police violence clusters in poorer neighborhoods. Past studies have linked neighborhood poverty to higher rates of fatal police shootings and more aggressive use of force during stops. But this study, published in JAMA Network Open, is among the first to ask whether that same pattern shows up in deadly car chases, a type of police-related death that’s often left out of similar research.
Five Years of Chase Data, Matched County by County
To answer this question, researchers from New York University’s School of Global Public Health pulled data from the Fatal Police Pursuits Database, a public tracker built by combining federal crash records, court documents, news reports, and information from groups that monitor police violence. Every case was manually checked to avoid counting it twice. The final dataset covered 2017 through 2021 and included more than 3,000 fatal chases nationwide, deaths caused directly by the chase itself, such as crashes or drownings, not deaths from an officer’s use of force.
That data was then matched, county by county, with a tool called the Childhood Opportunity Index. Despite its name, the index isn’t just about kids. It scores every U.S. county on 44 different measures, things like poverty rates, employment, housing quality, access to health care, air pollution, and school quality, and combines them into one score of how much opportunity a place offers its residents. Counties were sorted into three groups: low opportunity, moderate opportunity, and high opportunity.
Researchers then ran statistical models to see whether counties with lower opportunity scores also had higher rates of fatal chases, while accounting for other factors that might explain the difference, such as whether a county was rural or urban, the population’s age and sex makeup, and local violent crime rates.
Low-Opportunity Counties Saw a 63% Higher Rate of Fatal Chases
Out of 15,570 county-year records, the large majority, nearly 9 in 10, had no fatal chases at all in a given year. When chases did turn deadly, most counties saw just one that year, though a small number saw five or more.
A clear gap shows up when comparing high- and low-opportunity places. Counties with low overall opportunity had a 70% higher rate of fatal chases than high-opportunity counties before crime was factored in. Accounting for violent crime narrowed the gap, but it stayed large: a 63% higher rate.
Breaking the opportunity score into its three parts told a similar story. Counties with weak educational opportunity, meaning worse schools and fewer resources for kids, had a 59% higher rate of fatal pursuits. Counties with poor health and environmental conditions, think polluted air or limited access to health care, had a 66% higher rate. And counties with weak economic resources, low income and high poverty, saw a 69% higher rate. Every slice of the data pointed the same direction.
Heavier Policing in Disadvantaged Areas May Drive the Gap
Researchers point to a public health idea called fundamental cause theory, which holds that people and places with more money, connections, and influence can better protect themselves from all kinds of harm, while under-resourced communities have fewer tools to shield themselves. The same idea has already been used to explain why fatal police shootings cluster in poorer neighborhoods.
Part of the explanation may also come down to how police work. Many departments concentrate officers in neighborhoods flagged as high-crime areas, a tactic known as hot spot policing. That heavier police presence means more encounters between officers and residents, which creates more chances for a stop to turn into a chase. The study’s authors note this fits with earlier findings on stop-and-frisk encounters in New York City, where greater neighborhood disadvantage was linked to more use of force by officers.
There are limits to what this study can tell us. It looked at counties, not individuals, so it can’t say that any specific person’s risk went up because of where they live. It also couldn’t determine whether low-opportunity counties have more chases overall or whether their chases are simply more likely to turn deadly, because no national database currently tracks how many total chases happen, only the ones that end in death. The authors call the results a documented pattern, not proof that one thing causes the other.
Fatal police chases have been a growing public safety concern for years, and this research adds a new piece to that picture: fatal pursuit incidents aren’t spread evenly. They cluster in the same kinds of communities that show up again and again in research on policing, the ones with the fewest resources.
Disclaimer: This article summarizes findings from a peer-reviewed observational study. It does not establish that living in a particular county causes a fatal police chase, and it should not be used to assess risk for any individual person or community.
Paper Notes
Limitations
The authors describe this as an ecological study, meaning it examined patterns at the county level rather than tracking individual people, so its findings cannot be used to predict any single person’s risk. The observed associations should not be interpreted as proof of cause and effect because of the possibility of unmeasured factors influencing the results. The Fatal Police Pursuits Database, while described as the most comprehensive source available, likely undercounts actual incidents due to incomplete reporting by law enforcement agencies and gaps in local news coverage. Because there is no national database tracking the total number of pursuits (not just fatal ones), researchers could not calculate what share of pursuits turn deadly in each county, only how many fatal incidents occurred. The study also lacked county-level traffic volume data and was restricted to 2017-2021 due to incomplete 2022 data, and it excluded Alaska, Hawaii, Puerto Rico, and U.S. island territories because of limited data availability for the neighborhood opportunity measure used.
Funding and Disclosures
The article lists no conflict of interest disclosures for the authors. The authors thanked Jennifer Gollan and Susie Neilson of the San Francisco Chronicle for making the Fatal Police Pursuits Database publicly available, noting that neither received financial compensation for that contribution.
Publication Details
The study, titled “Neighborhood Opportunity and Fatal Police Pursuits in the US,” was written by Jemar R. Bather, PhD, and Melody S. Goodman, PhD, of the Department of Biostatistics at New York University School of Global Public Health. It was published in JAMA Network Open on August 26, 2026 (2026;9(8):e2630679), with the DOI 10.1001/jamanetworkopen.2026.30679. The study is open access under a CC-BY-NC-ND license.







