May 1 2000

Statistical Bias

Journalists find 'comfort' after the Diallo killing

When New Yorkers went into shock over the 41 bullets fired at Amadou Diallo, journalist Elizabeth Kolbert found “comfort”–her word–by recalling the sodomizing of Abner Louima. The business with the broomstick, she explained in the New Yorker (3/29/99), was not what we hire the police to do, whereas we do pay them to accost characters who fit a certain pattern. That fusillade, she said, “may not be racism at all but something new, a form of racial bias that is statistically driven and officially sanctioned.”

She meant, of course, profiling, and it’s hardly new. Nor was there anything new in the thought that bias backed by statistics is not racism. It was a notion that she shared with her former colleagues at the New York Times, such as Malcolm Browne and Peter Passell, whose reviews helped make a bestseller of The Bell Curve (New York Times, 10/16/94, 10/27/94; see Extra! Update, 12/94). It’s not racism, stupid, it’s the stats.

The Diallo shooting brought out a statistical pattern at the Times, which was visible one night on its cable TV panel show on New York 1. Two white Metro columnists, John Tierney and Clyde Haberman, urged viewers to grant the police the presumption of innocence (a presumption they had denied to Diallo); while Bob Herbert and David Ramirez, black and Latino respectively, were critical.

When an appellate court later ruled that the policemen could not find a fair jury in New York City, Herbert was on vacation and Ramirez had been reassigned, so there was nobody to contest the view of four white columnists that the court was quite right. Joyce Purnick (2/17/00) also thought it was excessive to accuse the nocturnal manhunters of the Street Crimes Unit (“We own the night”) of murder by “depraved indifference.”

Early on, the Times’ Haberman charged Diallo protesters with a sort of depraved indifference to truth and justice (3/23/99). He said it was they who were stirring up racism, and practicing McCarthyism by applying that foul word to Mayor Giuliani, and he reminded them how soft New York’s only black mayor, David Dinkins, had been on black racist crime. Haberman and Tierney alternated in the Metro column in warning that public safety would suffer if the hands of the police were tied.

Actually, stats occasionally published in the Timescontradicted them. The steep decline in crime began midway through the Dinkins administration. A recent front-pager by Fox Butterfield (3/4/00) reported, with stats, that other cities had obtained as good results as New York or better with community outreach and foot patrolling, as promoted by Dinkins.

Giuliani’s statistically driven strategy was to sweep “high-crime” neighborhoods by paramilitary force, stopping and frisking every male deemed suspicious. A state audit of some 40,000 such accostings found that about one out of ten resulted in an arrest. Though it was not widely noted, the Diallo trial revealed a large flaw in this stat: The defendant cops, though required to report every stop-and-frisk, couldn’t remember how many they had made even that night, and had no notes to refresh their recollection.

Further light was shed by one of those TV real-life cop shows, which was screened on Court TV during a pause in the trial. The cameras followed a chase through a suburb of three youths who had been waving a pistol and a rifle around. The perps were forced off the road, made to straddle prone, the works. Scary. Turned out the weapons were toys, the youths clean-cut and blond. They were let off with a scolding. If they had been black, they surely would have figured in the stats by getting shot or arrested, or both. But bias did not figure in the Diallo trial; the judge ruled it irrelevant.

Even kibitzers on Court TV who were sympathetic to the defense voiced astonishment at the laxity of the prosecution and at the judge’s marathon instruction to the jury that it must acquit if the cops might have thought that Diallo was a criminal or that any of them was in danger. They were embarrassed after the verdict when his honor went calling on defense counsel for a farewell embrace.

A final statistic: A poll that weekend found that a majority of the public, upstate and down, thought the verdict was wrong. The editorialists for all three New York City dailies thought it was correct. But they all felt bad about Diallo.

John L. Hess, a former New York Times reporter, does regular commentary on New York’s WBAI.