In a piece that contains the telling (even in context) line “I am a racist,” longtime Washington Post columnist Richard Cohen employed a mishmash of poorly explained statistics and bafflingly ignorant mathematical reasoning to argue that Trayvon Martin was “understandably suspected because he was black” — that is, Americans should assume any young black men they meet are criminals.
Writing that he could “understand why Zimmerman was suspicious and why he thought Martin was wearing a uniform we all recognize” (hoodies presumably being the universal uniform of all black male criminals), Cohen argued that Americans, and especially American politicians, are too afraid of the PC police to acknowledge the reality that black men are more likely to be criminals. “Where is the politician,” Cohen wails, “who will own up to the painful complexity of the problem and acknowledge the widespread fear of crime committed by young black males?”
Set aside for the moment that President Obama, who Cohen claims “perpetuated” the enforced silence surrounding black crime, incessantly talks to black audiences about violence in the African-American community. Is Cohen’s basic claim — that higher rates of black crime make it reasonable to suspect black men of being criminals — true?
No. First, the basic assumption, that black men are more likely to commit crimes than the average member of another demographic, isn’t nearly as well-founded as Cohen wants it to be. For instance, blacks and whites use drugs at rates basically proportionate to the population: blacks are 14 percent of Americans and 14 percent of monthly drug users. Yet blacks are at least four times as likely as whites to be incarcerated for a drug crime. Could that have something to do with attitudes like Cohen’s?
Broader numbers bear the point about race and drug use out. In a piece entitled “The Trayvon Martin Killing and the Myth of Black-on-Black Crime,” Jamelle Bouie cites a litany of recent statistical evidence to show that blacks aren’t “especially criminal,” including the fact that “among black youth, rates of robbery and serious property offenses are at their lowest rates in 40 years, as are rates of violent crime and victimization.” While its true that there are some numbers that go the other way, academic research suggests the blacks are still unfairly treated: in a review of the work on sentencing disparities for the American Bar Association, Marc Mauer cites 2008 research suggesting “only 61 percent of the black incarceration rate is explained by disproportionate engagement in criminal behavior…Thus, nearly 40 percent of the racial disparity in incarceration today cannot be explained by differential offending patterns.” The American criminal justice system has already done what Cohen wants: assume black men are criminals and ask questions later.
But even if every one of Cohen’s prejudices about black men and crime was true, the columnist’s claim that it’s reasonable to assume black men are criminals when you meet them is founded on a laughably simple statistical error. Statisticians call it the “base rate error,” and it goes something like this: suppose taller mountains are disproportionately more likely to be volcanoes than shorter mountains. That doesn’t make it smart to assume that most tall mountains you see are volcanoes, because volcanoes are still a tiny percentage of all tall mountains. The most important question is the basic likelihood of a tall mountain being a volcano (base rate), not whether tall mountains make up a higher percentage of volcanoes as compared to other mountains.
What this means in practice is that, even if blacks are more likely to commit crimes that people of other ethnicities, you still wouldn’t be right to say George Zimmerman “understandably suspected” Trayvon Martin of being a criminal. Violent crime in the United States is extremely rare, and despite a slight uptick in 2013, crime rates are in free fall. In Detroit, a city commonly associated with violent crime, there’s one murder for every 2,005 residents. Even in Memphis, which saw 1,000 more violent crimes in 2012 than in 2011, there’s only one violent crime for every 57 residents — meaning your overall chance of being a victim is less than 2 percent. The bottom line is that violent crime is rare, meaning that it’s mathematically absurd to assume any individual black man you see is going to be a criminal.
But I shouldn’t need to point this out to Cohen. As a columnist at one of the nation’s most famous newspapers, he should have known better than to argue using obvious errors that work in practice to reinforce one of the most vicious racial prejudices in America: that black men are, by in large, criminals and hence should be treated as such by the criminal justice system and society at large. But then again, maybe we shouldn’t have expected better.