Twisting statistics to smear cops

False claims about police brutality are poisoning race relations.

This feeling of “unfairness,” as President Obama noted in his weekly address on May 16, is “fuel(ing) the kind of unrest that we've seen in places like Baltimore, Ferguson and New York.”

The riots it incites threaten people's lives and destroy businesses. Police officers' lives are put at greater risk.

And even though, as in Baltimore, many of the police officers involved are black, the debate has inflamed tensions between whites and blacks.

“It's outrageous when officers use excessive force against young, unarmed African-American men, who are 21 times as likely to be shot dead by the police as young white men,” claimed Nicholas Kristof in a recent New York Times op-ed.

If this claim were indeed true, it would surely be outrageous. But the claim originating from ProPublica and further exaggerated by Kristof is completely false. This is not based on national data but comes from a couple hundred police departments (217 in 2012, just 1.2 percent of all the departments in the country). Nor that these few departments are unusual: They are predominantly in urban areas, which have much higher concentrations of blacks. This skews the numbers to over-represent black deaths.

But other factors also must be taken into account. Nationally, black male teenagers were nine times more likely to commit murder than were their white counterparts, thus presumably posing more of a risk to police. Adjusting for that and that the data are from heavily black areas leaves no significant difference in the risk that police pose for young black and white men.

ProPublica justified the flawed claim by quoting Professor David Klinger of the University of Missouri-St. Louis. But it turns out that he never viewed the data as reliable. Indeed, last October, Klinger complained to me that he had strongly warned ProPublica against using the FBI Uniform Crime Report data on justifiable police homicide.

Like other academics who have worked with this data, he labeled it as “no good.” (One ProPublica author, Ryan Gabrielson, denied to me that Professor Klinger told them this.)

Yet, Klinger repeatedly has stated that ProPublica misquoted him. In a later interview with St. Louis National Public Radio, Klinger was very clear: “The ProPublica thing needs to be shut down. They cherry-picked the three years that had the worst disparity instead of being honest about the whole picture. The ProPublica analysis is absolute garbage because it is based on the FBI's supplemental homicide reports. I told them, don't do it because the stats are (expletive deleted).”

Yet, as misleading as ProPublica's outrageous claim already was, Kristof managed to make it even more inflammatory. He added his own twist and put the term “unarmed” in front of “African-American men.”

The actual claim, however, did not refer to unarmed men but to justifiable police homicides. And those justifiable homicides involve people threatening others with death or serious injury — usually attackers armed with some type of weapon.

What terrible reporting. Kristof and The New York Times could have discovered these errors by fact checking on the Internet.

Still, despite being personally told of these errors, Kristof still has not posted a correction. And The New York Times never published a letter pointing out these errors.

Journalists need to be careful about the supposed facts they use. Mistakes that “fuel” feelings of “unfairness” can have dire consequences on the streets.

• Lott is the president of the Crime Prevention Research Center and the author of "More Guns, Less Crime" (University of Chicago Press, 2010).