Harvard researchers say police killings ought to be treated as a public health matter. Writing in the journal Public Library of Science, Nancy Kreiger and her collaborators propose that in-custody deaths be deemed a "reportable condition" requiring prompt notification of national public health authorities. She uses Baltimore as an example: "[I]n late April 2015, in the wake of the death of Freddie Gray, a 25-year old African American man who was fatally injured while in the custody of the police, the resulting civil unrest, which occurred prior to charges being brought against the six police officers involved, led to immediate and long-term public health harms, including medication crises linked to the destruction of a dozen pharmacies, opioids from these pharmacies entering the illicit drug street market, mental health trauma, and further damage to the economies of neighborhoods already burdened by high rates of unemployment and premature mortality."
While it may be a stretch to attribute a "public health harm" to Gray's death, as if the resulting riot was a predictable outcome, there is no good case to be made for the current lack of public data about police killings.
Currently, these statistics are compiled and reported only months and years after the fact. Epidemiologists would not stand for it if, say, measles deaths took a year or two to make it into a usable database.
The Harvard researchers have already examined the long-term trends in "deaths due to legal intervention" (the term of art) between 1960 and 2010 in several cities, and the results are striking: In every city—except Baltimore—deaths by police fell precipitously in those decades.
Repeat: By the available data, police were killing fewer people in the 2000s than they did 20 or 30 or 50 years before. They are killing fewer black people, and they are killing fewer white people—although the rate of African-American deaths by police is still much higher than the rate of white deaths by police.
Counting births and deaths is a first-order responsibility of a civilized nation. But in the U.S., the deaths of people in police encounters are obscured by gaps in the statistical reporting requirements and recalcitrance by law enforcement agencies. The Guardian sought to remedy this last June with its "The Counted" open data project. They found 1,134 deaths in 2015—far more than the FBI has in its own database, according to the PLoS researchers. FBI Director James Comey said last fall that it was "ridiculous and embarrassing" that the British newspaper had more comprehensive data than the Bureau.
It is unclear why the U.S. has never focused on this subset of deaths (in-prison deaths are yet another mystery, by the way, although the Bureau of Justice Statistics keeps somewhat of a steady count, at least). The country already has a "National Violent Death Reporting System," but only 32 states report, so it is incomplete. The Justice Department also reportedly is trying to find a way to count deaths by police officer in closer to real time.
The Harvard researchers say their way is easiest: "No act of Congress is needed. No police department need be involved," they write. "Public health agencies can do the job."