US police hospitalized 54,000 people in 2012

US police killed or injured 55,400 people during "legal" stops, searches and arrests in 2012, according to a study in Injury Prevention, one of Britain's peer-reviewed BMJ medical journals. Researchers also found that police disproportionately targeted black people, Native Americans and Latinos for stops and arrests and, though the use of deadly or debilitating force did not vary by ethnicity, the increased contact with police created risks for members of those minorities.

"This is nowhere near a new problem or a new public health problem," said Ted Miller, a scientist at Pacific Institute for Research and Evaluation in Maryland who conceived "Perils of Police Action: A Cautionary Tale From US Data Sets." "Police use of excessive force without due process of law has been with us forever as a problem, and since the Civil War it's been viewed particularly as a problem for the black community."

Miller, along with US- and Australia-based colleagues, examined data from medical and legal sources and tallies of police killings conducted by the Guardian and Washington Post newspapers. For the purposes of the study, "legal interventions" are defined as "arrests, stop-and-search incidents on the street and traffic stops involving a search." One of the more challenging aspects of the research, Miller said, is how little official documentation of police violence can be found in law enforcement databases and how often police involvement in injuries is left out of medical records.

"The surprise that had dropped on all of us around the time of starting this was that the data sets that I've worked with since I came into the field of injury, from the vital statistics on the number of people killed in the United States and the intent of those deaths, that the coroners and medical examiners were failing to code police involvement in almost half of the police-involved deaths," Miller said. "And there are two separate reporting systems that the police are supposed to report police-involved deaths to - and both of those had even worse underreporting."

"When you correct those firearm accounts, what you find is that one in every 13 people who died because someone else fired a firearm, the police pulled the trigger," he added. "A police officer pulled the trigger. And that's scary."

The lack of consistent record-keeping, a lack of research of the type conducted by Miller and his team, and media coverage of attacks on police, such as two recent shootings in which a total of eight US police officers were killed, can lead to the impression that policing is more dangerous for officers than it is for the public. "The fact is, though, that police use force more often, at higher levels, and with more deadly results than they face it," said Kristian Williams, the author of "Our Enemies in Blue: Police and Power in America," a comprehensive critical history of US law enforcement. "You wouldn't know from the TV news, and certainly not from the cop shows on later, that policing is less dangerous than driving a truck," he added. "You also wouldn't know, from those same sources, that the cops kill about three people a day. In fact, literally no one knew that until last year because, until The Guardian started a database, no one had bothered to count."

By the numbers

"Perils of Police Action" estimates that in 2012 one in 291 stops or arrests led to an injury requiring hospitalization or, at the extreme, death, but that the likelihood of life-threatening injuries as a result of police conduct was lower than that of other forms of assault. This led the researchers to determine that "police weren't usually out of control when they physically confronted a suspect." However, the research found that gunshot wounds were significantly more likely to prove fatal if they were inflicted by police - 40 percent to 26 percent.

Though the risk of injury was equally likely once police contact was initiated, the singling out of black, Latino and Native American men for stops leaves those groups especially vulnerable to death or injury. Police stopped black people at a rate nearly three times that of whites (1,404 stops and arrests per 10,000 population versus 503 for whites), Native Americans were stopped or arrested at a rate of 1,141 per 10,000 population, and Hispanics 979. "Excess per capita death rates among blacks and youth at police hands are reflections of excess exposure," the researchers concluded.

"There's a real tendency to understand police brutality as a kind of mistake, as an imperfection rather than as a feature of the institution," said Williams, who was not involved in the study. "The assumption is that the police are behaving in some way contrary to their purpose. I think it makes more sense to start with the behavior of the institution and - seeing that across the country and throughout its history it has reliably produced violence and disproportionately targeted people of color - conclude that that is the real purpose."

The study didn't examine the reasons behind the stops and arrests, for example whether police might be consciously or unconsciously profiling by skin color. However, it did conclude that "given a national history of racism, the excess per capita death rate of blacks from US police action rightly concerns policy analysts, advocates and the press." The authors add that "it would seem prudent to train at-risk groups about appropriate behavior during police stops." A longer-term fix, however, would be to implement accountability measures and address the violence at the institutional level.