When Police Shootings Don’t Kill: The Data That Gets Left Behind


Julie Ward is no stranger to the realities of gun violence. She worked for almost 15 years as a nurse in hospitals and emergency rooms from Arizona to Oregon and Washington. In that time, she said, she probably treated dozens of people harmed by gun violence, including teenagers, adults, and first responders.

Those experiences led Ward to pursue a doctorate in health policy with a focus on gun violence prevention in the fall of 2019. But while working as a research assistant at Johns Hopkins Bloomberg School of Public Health, Ward was surprised to learn that there is no federal reporting requirement for police-involved shootings or use of lethal force.

“I couldn’t believe that coming from a health care background, where we have tons of mandated documentation,” said the public health researcher, who is now an assistant professor of medicine, health, and society at Vanderbilt University. “I know there are a lot of things that police need to document too, but it didn’t seem right that a major piece of government wouldn’t have such a basic reporting check.”

While efforts to document both gun-related homicides and fatal police shootings have gained ground in recent years, leading to more academic research, less is known about the cases in which people survive shootings by police officers. Victims may endure serious injuries, multiple surgeries, and long-lasting physical and mental complications.

Now, intriguing new research seeks to quantify the extent of those injuries. One recent paper, led by Ward, suggests that police shoot and injure about 800 people each year in the United States. Black and male victims were among those disproportionately injured. Other research has looked into the context of those nonfatal shootings and found they are more likely to happen in communities with high rates of violent crime, poverty, food insecurity, and mental health crises.

“It didn’t seem right that a major piece of government wouldn’t have such a basic reporting check.”

Still, the lack of high-quality, official federal data on police shootings remains a pressing concern, researchers say, and challenges efforts to hold police departments accountable. To add insult to these injuries: Some data show that many of the people injured by officers are not charged with crimes, raising the question of whether those shootings were in self-defense or the result of excessive force.

Police shootings may receive the most attention, but many civilians are injured and killed by other means too, such as George Floyd, who was killed in 2020 by a Minneapolis police officer who kneeled on his neck. Alfreda Holloway-Beth, an epidemiologist at the University of Illinois Chicago, says it is important to study these injuries, too. “People are also dying from being manhandled, and choked out, and under duress in other ways,” said Holloway-Beth, whose research investigates injuries at the hands of police officers.

Firearm-related injuries and deaths, she said, are “literally the tip of the iceberg.”

At least 50 million Americans come into contact with law enforcement each year through traffic accidents, stops, arrests, and house calls, according to data collected by the Law Enforcement Epidemiology Project at the University of Illinois Chicago School of Public Health. That means about one out of every seven Americans interact with police each year.

It’s a very significant number of people, said Holloway-Beth, who directs the initiative, which seeks to analyze and share surveillance data on civilian injuries and deaths, as well as advocate for policy changes around police use of force.

Of those 50 million civilian encounters with law enforcement each year, about one million “experience police threat of or use of force during these interactions,” according to the project, and an estimated 250,000 are injured. Those injuries can be serious — the researchers estimate that at least 75,000 people are hospitalized each year because of police force — and have long-term effects, said Holloway-Beth.

To add insult to these injuries: Some data show that many of the people injured by officers are not charged with crimes.

Gunshot wounds, for example, may require multiple surgeries, can impact mobility and the central nervous system, and may involve long and expensive rehabilitation. People may also lose income as a result of their injuries, and can suffer psychological effects such as PTSD, anxiety, or complications with cognition and memory.

The Law Enforcement Epidemiology Project evolved from Holloway-Beth’s graduate school research on occupational injuries at UIC, when she had to comb through pages and pages of worker compensation data. She kept encountering a billing code that hospitals and insurance companies use to invoice and pay claims. The code was called “Legal intervention” and referred to injuries related to interactions with law enforcement. Curious to see who else had written about it, Holloway-Beth dug into the literature, but could find only one other research paper that focused on this data. She thought the topic deserved more attention, and it became the subject of her dissertation.

Today, Holloway-Beth collaborates with injury epidemiologist Lee Friedman on the project. The researchers analyze national data on deaths and injuries related to police encounters, as well as statistics specific to Illinois and Chicago. They list the sources — which include data from hospital records, death records, and media reports, among others — so that others can replicate their research for other locations, said Holloway-Beth, who is also the director of epidemiology at the Cook County Department of Health.

Support Undark Magazine

Undark is a non-profit, editorially independent magazine covering the complicated and often fractious intersection of science and society. If you would like to help support our journalism, please consider making a donation. All proceeds go directly to Undark’s editorial fund.

The initiative’s latest report, which has not been peer-reviewed, analyzes deaths and injuries related to law enforcement in Illinois between 2016 and 2022. In that time, they found that up to 140 people were killed and nearly 5,200 people were injured during contact with law enforcement — whether through shootings, manhandling, blows, or “unspecified means,” among others. Many of the victims were male and African American. The most common injuries were to the head and upper torso. And about 13 percent of those injured experienced a traumatic brain injury, with potential long-term effects, according to the report.

Injuries after encounters with law enforcement “leave people feeling jarred, and marginalized, and feeling like a piece of crap,” said Friedman. Not to mention, researchers can’t track physical injuries that aren’t treated at a hospital. “We can’t see all those that self-manage their wounds or go to urgent care or free clinics.”

Public awareness and academic research on fatal police encounters has grown considerably since 2014. That’s when a series of high-profile cases — Eric Garner in New York City, Michael Brown near St. Louis, Tamir Rice in Cleveland — attracted widespread outrage and spurred the Black Lives Matter protest movement.

The public outrage revived in 2020, when millions of people protested police violence nationwide after Minneapolis police officers killed the unarmed George Floyd.

At the time, Julie Ward was a research assistant at the Johns Hopkins Center for Gun Violence Solutions and wanted to determine the number of police shootings — a considerable challenge due to the lack of comprehensive federal and state data.

Part of that gap is because there is no federal law that mandates police departments to report how many civilians were killed or injured by their officers. The few governmental databases around lethal force that exist rely on voluntary reporting, which can lead to gaps in tracking. According to the FBI, fewer than 8,300 law enforcement agencies — representing 63 percent of sworn officers — reported their 2021 data to the bureau’s National Use-of-Force Data Collection, for example. In 2019, the year the database was launched, the reported data represented only 41 percent of officers.

Julie Ward worked as a nurse in hospitals and emergency rooms before becoming a public health researcher focused on gun violence. Her recent paper suggests that about 1,000 people are killed and nearly 800 are shot and injured by police officers each year in the U.S.

Visual: Courtesy of Julie Ward

At least 34 states require police departments to report crime statistics, but “most of the laws are vague about whether police shootings must be included,” according to a 2022 report in The Washington Post.

California, Illinois, New Jersey, and Wisconsin are among a number of states that have enacted laws that require independent investigations after civilian deaths due to deadly force. Meanwhile, a 2021 paper published in the Lancet estimated that the Centers for Disease Control’s National Vital Statistics System — which is said to provide “the most complete data on births and deaths in the United States” — undercounted the number of deaths due to fatal police encounters by about 17,000 cases over four decades.

It is even more difficult to track how many civilians are injured and survive police shootings specifically, says Ward. But new datasets suggest about 1,000 people are killed and around 800 are shot and injured by police officers each year, according to an analysis led by Ward. The researchers believe this is the first national, multi-year data analysis of non-fatal police shootings.

Ward’s team analyzed data on all shootings by police and law enforcement officers between January 1, 2015, and December 31, 2020. The datasets were sourced at the Gun Violence Archive which collects data from more than 7,500 government, law enforcement, commercial, and media sources.

The team identified about 10,300 cases of police-involved shootings during that six-year period, some of which involved multiple victims, said Ward. In those years, almost 5,900 people were killed and about 4,740 were injured. In more than half of the injurious shooting incidents, victims were armed with a firearm. About ten percent had no weapons at all.

The researchers believe this is the first national, multi-year data analysis of non-fatal police shootings.

Once again, the lack of high-quality data complicated the data collection, said Ward. Almost one-third of the cases did not include the race of the victims.

“We flagged cases that had a photo but no explicit description,” said Ward. “We looked into what kind of difference it would make if we had two coders do a social assessment of race.” This gave them more to work with.

Altogether, they found that Black and Latino individuals were disproportionately shot and injured by officers.

Juveniles were more likely to survive police shootings than adults, the researchers noted. About three percent of the shootings involved juveniles, said Ward, and again, the lack of official data leaves many questions unanswered, such as why they were stopped and if they were armed.

It appeared that more juveniles were unarmed prior to injury, said Ward, adding that non-fatal injuries among juveniles “can have lasting life consequences, too, for potentially many decades to come.”

Meanwhile, almost one-quarter of those injured or killed by police officers were experiencing mental health or behavioral health crises, according to Ward’s data. This follows trends reported in another recent study, which found about one in three victims in all fatal police shootings were believed to have experienced mental or behavioral health crises.

Police departments — and their legislative or civilian oversight — should consider more tailored responses to resolving encounters with people experiencing mental and health issues, said Ward.

Police officers should be thought of as responders of last resort, “who we call when we don’t have someone else to more directly call to respond to the needs that we’re experiencing,” said Ward. “There is a potential cost that comes with an armed response to the problem that we’re dealing with.”

Ajenai Clemmons, a public policy professor at the University of Denver who focuses on police-community relations, said that Ward and her colleagues completed a careful analysis. Focusing their study on 2015 to 2020, in particular, provided high-quality data since “this timeframe saw massive efforts to generate, collect, and digitize data for police shootings,” said Clemmons, who was not involved in Ward’s research.

Clemmons says there could be an undercount for both fatalities and non-fatalities. One reason, she said, is “because the authors excluded accidental shootings from their data,” she said.

An analysis reported by VICE News — conducted from 2010 to 2016 — estimates for every person killed by an officer, two people were shot and survived. The analysis pulled data from a different dataset than the one referenced in Ward’s study. “They obtained their data from the nation’s 50 largest police departments,” said Clemmons, who added that the VICE News report and additional datasets are among the reasons that she would agree with the authors that their estimates of the impact of nonfatal shootings are conservative.

Ward’s paper did not estimate the rate of police shootings across all arrests. But another recent paper offers new data around the frequency of police shootings and their relationship with poverty and other socioeconomic factors.

The study was funded by the Center for Justice Research at Texas Southern University — a historically Black university in Houston — and published earlier this year.

The researchers collected data from the one hundred largest cities in the United States for 2021 and 2022 and found that police shootings happened about 15 times every 10,000 arrests. Violent crime, food insecurity, and living in communities with higher rates of mental health challenges are among the major factors associated with police shootings, according to their statistical model.

Police shootings tend to happen in under-resourced communities that have higher rates of violent crimes, said Howard Henderson, a criminologist based at Texas Southern University and lead author of the paper.

“What you find is that a lot of times what we ascribe to ‘race,’ is in fact, a byproduct of the fact that African Americans have a greater likelihood of being poor,” said Henderson. “Our proximity to wealth and income is a problem that we oftentimes see actualized in the criminal justice system.”

Food insecurity also can lead to criminal behavior, said Henderson and the co-authors, such as theft and property crime, as well as aggression and violent behavior. And the prices of food, gasoline, and housing have increased in recent years, Henderson added.

Experts noted an additional concern: A lack of follow-up or criminal charges after police shoot and injure civilians.

National numbers are difficult to locate but there are data for some cities and states. The Detroit Free Press, for example, published a first of its kind investigation in April, which found: “More than a third of people shot nonfatally by Detroit police in a recent seven-year period either were not charged with a crime or not convicted of the conduct officers said prompted them to open fire.”

The numbers in Illinois were even starker: Only about one quarter of the patients injured in police shootings were discharged to courts, suggesting there were no charges in 75 percent of the cases, according to the recent white paper by the University of Illinois Chicago Law Enforcement Epidemiology Project.

“Why are the police using force on the individual and then not pressing charges?” asked Friedman, who co-authored the report. “You see the opposite with fatal injuries,” he added, noting that in about 90 percent of the fatal shootings, the person is allegedly armed or threatened the police.

The lack of criminal charges following many police shootings — at least in Illinois — suggests the possibility of excessive force.

“We don’t track these shootings at the federal level, or the use of officers’ weapons in general.”

The new data on non-fatal injuries and police shootings “gets us closer to the true national snapshot” of the total impact of these cases, said Julie Ward, but the paper has its limitations. One of the leading challenges is the lack of comprehensive federal and state data around police shootings. The researchers are hopeful that these types of studies will help inform new policies.

“We don’t track these shootings at the federal level, or the use of officers’ weapons in general. This is an important and unique responsibility that we have given to officers, as the face of our government, to use deadly force when they deem it necessary,” said Ward. “But we don’t track how often that decision is made or the consequences of it nationally.”


Rod McCullom is a Chicago-based science journalist and senior contributor to Undark whose work has been published by Scientific American, Nature, The Atlantic, and MIT Technology Review, among other publications.