Nick Selby, an Atlantic reader and police detective in the Dallas-Fort Worth area, responds to our earlier request for empirical data on the question posed above. His email is long and densely packed, highlighting some key points in the debate over police shootings—especially in the context of the tragic shooting in Baltimore County referenced below:
In his response to a letter from Charles Black, Conor Friedersdorf refers to a statistic quoted by The Washington Post that claimed, “Black males are seven-times-more-likely than white-males to die at the hands of police.” The data are there to be examined, but part of the issue that makes the math of that claim so dodgy is that the Post assumed black males are distributed equally around the United States—as if as many black males live in, say, Nebraska or Vermont as in, say, Florida or Mississippi. Which they do not.
Another fatal flaw with this idea of using incidents in which police kill people to draw larger conclusions about race and justice in America is that it considers the wrong cohort. While Mr. Friedersdorf can and rightfully does point to the case of officers Michael Slager and Ray Tensing as being excellent examples of officers both apparently acting terribly and also lying about it—cases we wouldn’t have known about had there not been witness video—he points to these as evidence of the selection by police officers of black men to harass and ultimately to kill.
To those who have never served in or been trained in law enforcement or use-of-force, it is easy to see why the cases of Freddie Gray and Eric Garner look equally bad. So, let’s just assume that these deaths at the hands of police were just as baldly criminal, for the moment, as those of Walter Scott and Samuel DuBose.
Mr. Friedersdorf is correct that, in seeking potentially unjust killings, the segment of the Washington Post cohort of "all people shot by police" is too broad - one quite simply must look past the killings by police of armed people, to the killings by police of the unarmed*.