Let's play cops and professors--you're it!

Charles Blow writes a column called “By the Numbers,” in the New York Times. Linked below is his commentary on the Henry Louis Gates incident. He recounts a personal episode, a chilling police stop in Louisiana where a cop said he could kill him without cause and without consequences. Inexcusable behavior that’s relevant to the discussion—“so far, so good,” as President Obama said in his remarks about Gates.

Blow then seeks to support his point of view with some graphs that I believe are deeply misleading. The graphs, attributed to a 2008 NY Times / CBS survey, depict that black men are about seven times as likely as white men to answer “yes” to the question, “Have you ever felt you were stopped by the police just because of your race or ethnic background?”

By the numbers, indeed. Blacks are about 12% of the US population, whites about 75%. What does that predict about the likelihood—at random—that when stopped by police, blacks will experience a white police contact, versus whites experiencing a black police contact? Forget good or bad. Just consider, “If I’m stopped, what is the probability that the officer will not be my own race?”

Whites (or anybody) contacted by a police officer have about one chance in eight that the officer will be black--ie, will be a different-race police officer that’s relevant to this discussion. But blacks (or anybody) have six chances in eight that, if they encounter the police, their contact will be with a white officer. So, regardless of what the question is, blacks should report six times--600%--the frequency of observing a white officer doing anything, versus the frequency whites would report of observing a black officer doing the same thing—whether it’s acting aggressively, chewing gum on the job, or trying to sell tickets to the policemen’s ball.

Why? Once more—because, based purely on the racial make-up of the US population, blacks are six times as likely to be stopped at random by a police officer who’s white as whites are to be stopped by a police officer who’s black.

The graphs accompanying the article are very misleading, in a way that supports Mr. Blow’s narrative, and the New York Times’ political agenda, but wouldn’t pass muster in a behavioral statistics class. Police racism is bad enough when it’s accurately reported and depicted. But, to portray some ill-behavior as occurring six times more likely than chance, when it's really only occurring a few percent more than chance, is a serious inadequacy.

Geronimo

http://www.nytimes.com/2009/07/25/opinion/25blow.html?scp=3&sq=charles%20blow&st=cse