HOW CRIES OF RACIAL PROFILING AGAINST POLICE CAN HURT US

March 8, 2003

When science and politics face off, science often gets the worst of it. Today's cautionary tale is the myth of racial profiling on the New Jersey Turnpike.

"Myth?" you're saying. "Myth? Didn't New Jersey Governor Christie Todd Whitman admit state troopers' guilt? Didn't she fire the head of the state police for failing to stop it?"

Yes, she did, but that was politics. Myths demand sacrifices, and Whitman sacrificed her state's cops to preserve her own political viability.

If you want to know what the social science says, though, read Heather Mac Donald's new book, Are Cops Racist?, a collection of her cop-reporting articles from the magazine City Journal. Her focus is not on exonerating the police from charges of racism, although she does an excellent job of demonstrating that the charges are vastly exaggerated. She's much more concerned about how the war against the police harms black Americans, the subtitle of her book.

The rap against the Jersey cops, as the myth has it, is that state troopers were disproportionately stopping black drivers on the Turnpike, with "disproportionately" meaning "in higher numbers than their percentage of the population." The implication, obviously, was that the officers were motivated by racist stereotypes, fueled by the U.S. Drug Enforcement Agency's descriptions of which ethnic groups dominated particular parts of the trade in illegal drugs.

But race, as it turns out, was not the significant factor. Speeding was. After the speed of the car was controlled for, black drivers were actually less likely to be stopped than white ones.

The New Jersey attorney general commissioned a research study that was designed and carried out by the Public Service Research Institute in Maryland. The researchers took photos of nearly 40,000 cars whose speed was simultaneously measured with a radar gun. A team of three evaluators looked at the photographs and identified the race of the driver. (the evaluators had no information about the speed of the car).

When the researchers combined the driver identifications with the speeding information, Mac Donald says, they found that "black drivers speed twice as much as white drivers, and speed at reckless levels even more."

This information was not welcome. "The medieval Vatican could not have been more threatened," Mac Donald observes, "had Galileo offered photographic proof of the solar system."

Consequently -- re-enter politics -- the Bush Department of Justice tried to discredit the study. First they insisted that the racial identifications, made if two out of the three evaluators agreed, were unreliable. But rerunning the numbers with three out of three didn't change the results. Then they tried to argue that including the cars where the driver's race couldn't be determined would change the results. Finally, they just stalled on releasing the study, until the Bergen County, N.J., Record posted it on the Internet.

Mac Donald does not claim that no officers ever use race inappropriately in deciding whether to stop a motorist or ask to search his car. Her point, rather, is that if racial profiling were systematic and widespread, there would be statistical evidence for it, and there is none.

Race-baiters charge otherwise at every opportunity, citing disparities in stops and searches as evidence. It happened here when Denver cops released their statistics, and again when the Colorado Highway Patrol figures came out. But disparities by themselves are not evidence of racial animus unless differences in racial behavior can be ruled out, and they cannot. Mac Donald calls the pretense that all groups commit crimes at the same rate a "collective fairy tale."

But believing in fairy tales can be dangerous. After New Jersey signed a consent decree with the Clinton administration, state troopers pulled back on discretionary searches. "At the height of the drug war in 1988, troopers filed 7,400 drug charges from the turnpike, most of those from consent searches," Mac Donald says. In 2000, the number was 370.

The number of consent searches in 1999 was 440, while in a six-month period in 2001 it was 11.

Since it is unlikely that the drug trade has contracted by that much, the impact on the communities where those drugs are headed must be horrendous. The murder rate in Newark, N.J., rose 65 percent from 2000 to 2001, for example, which is probably no coincidence.

Defendants routinely claim that they are victims of racial profiling, and often it's enough to get them off. New Jersey dismissed drug and weapons charges against 128 defendants who were stopped on the turnpike, without regard to the facts in individual cases.

If the accusations of racial profiling keep cops from doing their job, remember, the people who suffer the most are the victims of crime, not its perpetrators.