The Myth of Racial Profiling

Heather MacDonald (City Journal, Spring 2001, Vol 11, No. 2)
http://www.city-journal.org/html/11_2_the_myth.html

 

The anti-"racial profiling" juggernaut must be stopped, before it obliterates the crime-fighting gains of the last decade, especially in inner cities. The anti-profiling crusade thrives on an ignorance of policing and a willful blindness to the demographics of crime. Yet politicians are swarming on board. In February, President George W. Bush joined the rush, declaring portentously: "Racial profiling is wrong, and we will end it in America."

Too bad no one asked President Bush: "What exactly do you mean by 'racial profiling,' and what evidence do you have that it exists?" For the anti-profiling crusaders have created a headlong movement without defining their central term and without providing a shred of credible evidence that "racial profiling" is a widespread police practice.

The ultimate question in the profiling controversy is whether the disproportionate involvement of blacks and Hispanics with law enforcement reflects police racism or the consequences of disproportionate minority crime. Anti-profiling activists hope to make police racism an all but irrebuttable presumption whenever enforcement statistics show high rates of minority stops and arrests. But not so fast.

Two meanings of "racial profiling" intermingle in the activists' rhetoric. What we may call "hard" profiling uses race as the only factor in assessing criminal suspiciousness: an officer sees a black person and, without more to go on, pulls him over for a pat-down on the chance that he may be carrying drugs or weapons. "Soft" racial profiling is using race as one factor among others in gauging criminal suspiciousness: the highway police, for example, have intelligence that Jamaican drug posses with a fondness for Nissan Pathfinders are transporting marijuana along the northeast corridor. A New Jersey trooper sees a black motorist speeding in a Pathfinder and pulls him over in the hope of finding drugs.

The racial profiling debate focuses primarily on highway stops. The police are pulling over a disproportionate number of minority drivers for traffic offenses, goes the argument, in order to look for drugs. Sure, the driver committed an infraction, but the reason the trooper chose to stop him, rather than the speeder next to him, was his race.

But the profiling critics also fault both the searches that sometimes follow a highway stop and the tactics of urban policing. Any evaluation of the evidence for, and the appropriateness of, the use of race in policing must keep these contexts distinct. Highway stops should almost always be color-blind, I'll argue, but in other policing environments (including highway searches), where an officer has many clues to go on, race may be among them. Ironically, effective urban policing shows that the more additional factors an officer has in his criminal profile, the more valid race becomes—and the less significant, almost to the point of irrelevance.

Before reviewing the evidence that profiling critics offer, recall the demands that the police face every day, far from anti-police agitators and their journalist acolytes.

February 22, 2001, a town-hall meeting at P.S. 153 in Harlem between New York mayor Rudolph Giuliani and Harlem residents: a woman sarcastically asks Giuliani if police officers downtown are paid more than uptown officers, "because we don't have any quality of life in Harlem, none whatsoever. Drug dealers are allowed to stand out in front of our houses every day, to practically invade us, and nothing's done about it." Another woman complains that dealers are back on the street the day after being arrested, and notes that "addicts are so bold that we have to get off the sidewalk and go around them!" She calls for the declaration of a state of emergency. A man wonders if cop-basher congressman Charles Rangel, present at the meeting, could "endow the police with more power," and suggests that the NYPD coordinate with the federal Drug Enforcement Administration, the INS, and the IRS to bring order to the streets.

The audience meets Giuliani's assertions that the police have brought crime down sharply in Harlem with hoots of derision. No one mentions "police brutality."

Valentine's Day, 2001, a police-community meeting at Harlem's 28th Precinct: an elegant man in an angora turtleneck, tiny blue glasses, and a shadow of a goatee breaks a local taboo by asking what the precinct is doing "to address dealing on the corners." Most residents shrink from mentioning the problem at precinct meetings for fear of retaliation from dealers. A tense silence falls. The man, a restaurant investor, tells me, "If this was 59th and Park, the police wouldn't allow these individuals to hang out on the corner." He can't understand why there's no "immediate result," if the police have in fact been cracking down on drug-dealing. "I don't think it should be so hard to dismantle," he says impatiently.

February 12, 2001, the fifth floor of a hulking yellow apartment building on Lenox Road in Flatbush, Brooklyn: two officers from the 67th Precinct investigate an anonymous call reporting a group of youths smoking marijuana in the hallway. The boys have disappeared. As officers check the stairwell, a gaunt middle-aged man sporting a wildly patterned black-and-white tie courteously introduces himself as Mr. Johnson, the building superintendent. After slowly bending down to pick up a discarded cigarette butt, he asks politely if anything more can be done about the kids who come from the next building to smoke pot in his hallway.

This is the demand—often angry, sometimes wistful—that urban police forces constantly hear: get rid of the drugs! These recent appeals come after the most successful war on crime that New York City has ever conducted. A decade and a half ago, when drug-related drive-by shootings became epidemic, inner-city residents nationwide were calling even more frantically for protection from drug violence. When New Jersey, a key state on the drug corridor from Central America to New England, sent its state highway troopers to do foot patrols in Camden and Trenton, residents met them with cheers.

In New York, the mayhem eventually led to the development of the Giuliani administration's assertive policing that strives, quite successfully, to prevent crime from happening. Outside of New York, the widespread pleas to stop drug violence led the Drug Enforcement Administration to enlist state highway police in their anti-drug efforts. The DEA and the Customs Service had been using intelligence about drug routes and the typical itineraries of couriers to interdict drugs at airports; now the interdiction war would expand to the nation's highways, the major artery of the cocaine trade.

The DEA taught state troopers some common identifying signs of drug couriers: nervousness; conflicting information about origin and destination cities among vehicle occupants; no luggage for a long trip; lots of cash; lack of a driver's license or insurance; the spare tire in the back seat; rental license plates or plates from key source states like Arizona and New Mexico; loose screws or scratches near a vehicle's hollow spaces, which can be converted to hiding places for drugs and guns. The agency also shared intelligence about the types of cars that couriers favored on certain routes, as well as about the ethnic makeup of drug-trafficking organizations. A typical DEA report from the early 1990s noted that "large-scale interstate trafficking networks controlled by Jamaicans, Haitians, and black street gangs dominate the manufacture and distribution of crack." The 1999 "Heroin Trends" report out of Newark declared that "predominant wholesale traffickers are Colombian, followed by Dominicans, Chinese, West African/Nigerian, Pakistani, Hispanic and Indian. Mid-levels are dominated by Dominicans, Colombians, Puerto Ricans, African-Americans and Nigerians."

According to the racial profiling crowd, the war on drugs immediately became a war on minorities, on the highways and off. Their alleged evidence for racial profiling comes in two varieties: anecdotal, which is of limited value, and statistical, which on examination proves entirely worthless.

The most notorious racial profiling anecdote may have nothing to do with racial profiling at all. On April 23, 1998, two New Jersey state troopers pulled over a van that they say was traveling at 74 miles an hour in a 55-mile-an-hour zone on the New Jersey Turnpike. As they approached on foot, the van backed toward them, knocking one trooper down, hitting the patrol car, and then getting sideswiped as it entered the traffic lane still in reverse. The troopers fired 11 rounds at the van, wounding three of the four passengers, two critically.

Attorneys for the van passengers deny that the van was speeding. The only reason the cops pulled it over, critics say, was that its occupants were black and Hispanic.

If the troopers' version of the incident proves true, it is hard to see how racial profiling enters the picture. The van's alleged speed would have legitimately drawn the attention of the police. As for the shooting: whether justified or not, it surely was prompted by the possibly deadly trajectory of the van, not the race of the occupants. Nevertheless, on talk show after talk show, in every newspaper story denouncing racial profiling, the turnpike shooting has come to symbolize the lethal dangers of "driving while black."

Less notoriously, black motorists today almost routinely claim that the only reason they are pulled over for highway stops is their race. Once they are pulled over, they say, they are subject to harassment, including traumatic searches. Some of these tales are undoubtedly true. Without question, there are obnoxious officers out there, and some officers may ignore their training and target minorities. But since the advent of video cameras in patrol cars, installed in the wake of the racial profiling controversy, most charges of police racism, testified to under oath, have been disproved as lies.

The allegation that police systematically single out minorities for unjustified law enforcement ultimately stands or falls on numbers. In suits against police departments across the country, the ACLU and the Justice Department have waved studies aplenty allegedly demonstrating selective enforcement. None of them holds up to scrutiny.

The typical study purports to show that minority motorists are subject to disproportionate traffic stops. Trouble is, no one yet has devised an adequate benchmark against which to measure if police are pulling over, searching, or arresting "too many" blacks and Hispanics. The question must always be: too many compared with what? Even anti-profiling activists generally concede that police pull drivers over for an actual traffic violation, not for no reason whatsoever, so a valid benchmark for stops would be the number of serious traffic violators, not just drivers. If it turns out that minorities tend to drive more recklessly, say, or have more equipment violations, you'd expect them to be subject to more stops. But to benchmark accurately, you'd also need to know the number of miles driven by different racial groups, so that you'd compare stops per man-mile, not just per person. Throw in age demographics as well: if a minority group has more young people—read: immature drivers—than whites do, expect more traffic stops of that group. The final analysis must then compare police deployment patterns with racial driving patterns: if more police are on the road when a higher proportion of blacks are driving—on weekend nights, say—stops of blacks will rise.

No traffic-stop study to date comes near the requisite sophistication. Most simply compare the number of minority stops with some crude population measure, and all contain huge and fatal data gaps. An ACLU analysis of Philadelphia traffic stops, for example, merely used the percentage of blacks in the 1990 census as a benchmark for stops made seven years later. In about half the stops that the ACLU studied, the officer did not record the race of the motorist. The study ignored the rate of traffic violations by race, so its grand conclusion of selective enforcement is meaningless.

Only two studies, both by Temple University social psychologist John Lamberth, have attempted to create a violator benchmark. The ACLU used one to sue, successfully, the Maryland state police; a criminal defense attorney in New Jersey used the other to free 17 accused black drug traffickers. Lamberth alleged that blacks in Maryland and southern New Jersey were stopped at higher rates than their representation in the violator population would seemingly warrant. But he defined violator so broadly—in Maryland, traveling at least one mile, and in New Jersey, traveling at least six miles, over the speed limit—that he included virtually the entire driving population. Lamberth must not have spent much time talking to real cops, for his definition of violator ignores how police actually decide whom to stop. Someone gliding sedately at 56 mph in a 55 mph zone has a radically different chance of being pulled over than someone barreling along at 80. An adequate benchmark must capture the kind of driving likely to draw police attention. Despite his severely flawed methodology, Lamberth is in great demand as a racial profiling guru.

Do minorities commit more of the kinds of traffic violations that police target? This is a taboo question among the racial profiling crowd; to ask it is to reveal one's racism. No one has studied it. But some evidence suggests that it may be the case. The National Highway Traffic Safety Administration found that blacks were 10 percent of drivers nationally, 13 percent of drivers in fatal accidents, and 16 percent of drivers in injury accidents. (Lower rates of seat-belt use may contribute to these numbers.) Random national surveys of drivers on weekend nights in 1973, 1986, and 1996 found that blacks were more likely to fail breathalyzer tests than whites. In Illinois, blacks have a higher motorist fatality rate than whites. Blacks in one New Jersey study were 23 percent of all drivers arrested at the scene of an accident for driving drunk, though only 13.5 percent of highway users. In San Diego, blacks have more accidents than their population figures would predict. Hispanics get in a disproportionate number of accidents nationally.

But though the numbers to date are incapable of telling us anything about racial profiling, that does not mean that it was not going on in some locations, at some times. Hard racial profiling in car stops—pulling over one speeder among many just because he happens to be black or Hispanic—has surely been rare. But conversations with officers in strong interdiction states such as New Jersey suggest that some troopers probably did practice soft racial profiling—pulling someone over because driver and car and direction and number and type of occupants fit the components of a courier profile.

Over time, officers' experience had corroborated the DEA intelligence reports: minorities were carrying most of the drugs. An example of the patterns they noticed: a group of young blacks with North Carolina plates traveling south out of Manhattan's Lincoln Tunnel into New Jersey? Good chance they're carrying weapons and drugs, having just made a big buy in the city. Catch them northbound? Good chance they're carrying big money and guns. Some officers inevitably started playing the odds—how many, the numbers cannot yet tell us.

Despite the hue and cry, there is nothing illegal about using race as one factor among others in assessing criminal suspiciousness. Nevertheless, the initial decision to pull a car over should be based almost always on seriousness of traffic violation alone—unless, of course, evidence of other law-breaking, such as drug use, is visible. If the result is that drug couriers assiduously observe the speed limit, fine. But compared with most other policing environments, highways are relatively cueless places. In assessing the potential criminality of a driver speeding along with the pack on an eight-lane highway, an officer normally has much less to work with than on a city street or sidewalk. His locational cues—traveling on an interstate pointed toward a drug market, say—are crude, compared with those in a city, where an officer can ask if this particular block is a drug bazaar. His ability to observe the behavior of a suspect over time is limited by the speed of travel. In such an environment, blacks traveling 78 mph should not face a greater chance of getting pulled over than white speeders just because they are black and happen to be driving a car said to be favored by drug mules.

Soft racial profiling was probably not widespread enough to have influenced traffic-stop rates significantly. Nor will eliminating it quickly change the belief among many blacks that any time they get stopped for a traffic violation, it is because of their race. Nevertheless, state police commanders should eliminate any contribution that soft profiling may make to that perception, unless strong evidence emerges (as it has not so far) that soft profiling has had an extremely high success rate in drug interdiction. Far more is at stake here than the use of race in traffic stops. Specious anti-racial profiling analysis threatens to emasculate policing in areas where drug enforcement is on a far stronger basis.

The most important victory of the anti-racial profiling agitators occurred not on the traffic-stop battlefield, but on the very different terrain of the searches that sometimes follow a stop. And here is where people who care about law enforcement should really start to worry. On April 20, 1999, New Jersey's then-attorney general Peter Verniero issued his "Interim Report of the State Police Review Team Regarding Allegations of Racial Profiling." It was a bombshell, whose repercussions haven't stopped yet.

"The problem of disparate treatment [of blacks] is real, not imagined," the report famously declared. Governor Christine Todd Whitman chimed in: "There is no question that racial profiling exists at some level." The media triumphantly broadcast the findings as conclusive proof of racial profiling not just in the Garden State but nationally. The New York Times started regularly referring to New Jersey's "racial bias" on the highways as incontrovertible fact. Defense attorneys and their clients celebrated as well. "Whenever I have a state police case, I file a suppression motion . . . alleging that the stop was based on color of skin and therefore illegal," a Trenton criminal defense attorney told the New York Times. "And now guess what? The state agrees with me!"

Yet the report's influential analysis is shoddy beyond belief. Contrary to popular perception, Verniero did not reach any conclusions about racial profiling in stops. His finding of "disparate treatment" is based on the percentage of "consent searches" performed on minorities after a stop has occurred. (In a consent search, the motorist agrees to allow the trooper to search his car and person, without a warrant or probable cause.) Between 1994 and 1998, claims the report, 53 percent of consent searches on the southern end of the New Jersey Turnpike involved a black person, 21 percent involved whites, and overall, 77 percent involved minorities. But these figures are meaningless, because Verniero does not include racial information about search requests that were denied, and his report mixes stops, searches, and arrests from different time periods.

But most important: Verniero finds culpable racial imbalance in the search figures without suggesting a proper benchmark. He simply assumes that 53 percent black consent searches is too high. Compared with what? If blacks in fact carry drugs at a higher rate than do whites, then this search rate merely reflects good law enforcement. If the police are now to be accused of racism every time that they go where the crime is, that's the end of public safety.

The hue and cry over the alleged New Jersey search rate makes sense only if we assume that drug trafficking is spread evenly across the entire population and that officers are unable to detect the signs of a courier once they have pulled over a car. There are powerful reasons to reject both these assumptions.

Judging by arrest rates, minorities are vastly overrepresented among drug traffickers. Blacks make up over 60 percent of arrests in New Jersey for drugs and weapons, though they are 13.5 percent of the population. Against such a benchmark, the state police search rates look proportionate.

The attorney general's report dismissed this comparison with an argument that has become de rigueur among the anti-racial profiling crowd, even in Congress: the "circularity" argument. Arrest and conviction data for drugs and weapons are virtually meaningless, said Verniero. They tell you nothing about the world and everything about the false stereotypes that guide the police. If the police find more contraband on blacks and Hispanics, that is merely because they are looking harder for it, driven by prejudiced assumptions. If the police were to target whites with as much enforcement zeal, goes this reasoning, they would find comparable levels of criminality. David Harris, a University of Toledo law school professor and the leading expert for the anti-profiling forces, makes this preposterous argument. An enforcement effort directed at 40-year-old white law professors, he assures a Senate subcommittee, would yield noticeable busts. The disproportionate minority arrests then reinforce the initial, racist stereotypes, and the vicious cycle begins all over again—too many minorities arrested, too many whites going free.

The circularity argument is an insult to law enforcement and a prime example of the anti-police advocates' willingness to rewrite reality. Though it is hard to prove a negative—in this case, that there is not a large cadre of white drug lords operating in the inner cities—circumstantial evidence rebuts the activists' insinuation. Between 1976 and 1994, 64 percent of the homicide victims in drug turf wars were black, according to a Heritage Foundation analysis of FBI data. Sixty-seven percent of known perpetrators were also black. Likewise, some 60 percent of victims and perpetrators in drug-induced fatal brawls are black. These figures match the roughly 60 percent of drug offenders in state prison who are black. Unless you believe that white traffickers are less violent than black traffickers, the arrest, conviction, and imprisonment rate for blacks on drug charges appears consistent with the level of drug activity in the black population. (And were it true that white dealers are less violent, wouldn't we expect police to concentrate their enforcement efforts on the most dangerous parts of the drug trade?)

The notion that there are lots of heavy-duty white dealers sneaking by undetected contradicts the street experience of just about every narcotics cop you will ever talk to—though such anecdotal evidence, of course, would fail to convince the ACLU, convinced as it is of the blinding racism that afflicts most officers. "The hard-core sellers are where the hard-core users are—places like 129th Street in Harlem," observes Patrick Harnett, retired chief of the narcotics division for the NYPD. "It's not white kids from Rockland County who are keeping black sellers in business."

The cops go where the deals are. When white club owners, along with Israelis and Russians, still dominated the Ecstasy trade, that's whom the cops were arresting. Recently, however, big shipments have been going to minority neighborhoods; subsequent arrests will reflect crime intelligence, not racism.

There's not a single narcotics officer who won't freely admit that there are cocaine buys going down in the men's bathrooms of Wall Street investment firms—though at a small fraction of the amount found on 129th Street. But that is not where community outrage, such as that Mayor Giuliani heard in Harlem, is directing the police, because they don't produce violence and street intimidation.

Ultimately, the circularity argument rests on a massive denial of reality, one that is remarkably vigorous and widespread. In March, 2000, for example, New Jersey senator Robert Torricelli asserted before then-senator John Ashcroft's Judiciary Subcommittee: "Statistically it cannot bear evidence [sic] to those who suggest, as our former superintendent of the state police suggested, that certain ethnic or racial groups disproportionately commit crimes. They do not." Needless to say, Torricelli did not provide any statistics.

The second condition necessary to explain the higher minority search rates on the highway is patrol officers' ability to detect drug trafficking. Unlike the initial decision to pull over a car, the decision to request permission to search rests on a wealth of cues. One of the most frequent is conflicting narratives among passengers and driver. "If a group in a car is carrying drugs, there will always be inconsistencies in their stories," reports Ed Lennon, head of the New Jersey Troopers Union. "It's unbelievable. A lot of times the driver won't know the passengers' first or last names—'I only know him as Bill'—or they'll get the names completely wrong. Sometimes they'll have a preplanned answer regarding their destination, but their purpose in being on the road will vary."

A driver's demeanor may also be a tip-off. "I've stopped white guys in pick-up trucks with a camper compartment on top," recalls Lennon. "Their chest is pounding; they're sweating, though it's the dead of winter. They won't look at you." And they're also hiding drugs.

Once a trooper stops a car, he can see the amount of luggage and its fit with the alleged itinerary, the accumulation of trash that suggests long stretches without stopping, the signs of drug use, the lack of a license and registration, the single key in the ignition and no trunk key, or the signs that the vehicle may have been fitted out with drug and weapon compartments. Some New York narcotics officers recently pulled over an Azusa SUV and noticed welding marks along the rain gutter on top. The occupants had raised the entire roof four inches to create a drug vault. If a car's windows don't roll all the way down, drugs may be concealed in the doors.

The fact that hit rates for contraband tend to be equal across racial groups, even though blacks and Hispanics are searched at higher rates, suggests that the police are successfully targeting dealers, not minorities. Race may play a role in that targeting, or it may not. Most cues of trafficking are race-neutral; it may be that race often correlates with the decision to search rather than causing it. But if race does play a role in the request to search, it is a much diminished one compared with a car stop based on a courier profile. When an officer has many independent indices of suspicion, adding his knowledge of the race of major trafficking groups to the mix is both legitimate and not overly burdensome on law-abiding minorities.

Amazingly, Attorney General Verniero acknowledges that the police merely try to maximize their hit rates in deciding whom to search, but he blames them for doing so. "The state police reward system gave practical impetus to the use of these inappropriate stereotypes about drug dealers" by rewarding big busts, he frets. But if the police were seeking to maximize their contraband yields, and the alleged "inappropriate stereotypes" were not helping them do so, presumably they would abandon those "stereotypes" and find some other set of cues—unless, of course, they were merely out to harass minorities for the thrill of it. But in that case, their hit rates would be lower for minorities than for whites, which they were not.

The bottom line is this: the New Jersey attorney general has branded the state police as racist without a scintilla of analysis for his finding. Yet New Jersey is the wave of the future, for racial profiling data-collection initiatives are sweeping the country. At least 30 states could soon require their state police to collect racial data on all traffic stops and searches, with the stated end of eliminating "racial profiling." Urban forces are under identical pressure. Virtually every major law-enforcement organization opposes these bills, because of their failure to deal with the benchmarking problem. Until someone devises an adequately sophisticated benchmark that takes into account population patterns on the roads, degrees of law-breaking, police deployment patterns, and the nuances of police decision making, stop data are as meaningless as they are politically explosive. Attorney General John Ashcroft has encouraged these data-gathering initiatives; he should instead withhold his support, unless local proponents can prove that they will capture the complex realities of law enforcement.

Unfortunately, the flurry of racial profiling analysis is not confined to the highways. It will wreak the most havoc on urban policing. Despite the racket by protesters, it is in city policing that race probably plays its least significant role, because officers have so many other cues from the environment. In assessing whether a pedestrian is behaving suspiciously, for example, they might already know that he is at a drug corner, about which they have received numerous complaints. They know if there has been a string of burglaries in the neighborhood. As they observe him, they can assess with whom he is interacting, and how.

A New York Street Crime Unit sergeant in Queens describes having stopped white pedestrians who had immediately changed directions as soon as they saw his unmarked car or ducked into an alley or a store for eight seconds and then looked for him once they came out. The night I spoke to him, he was patrolling the 102nd Precinct in Woodhaven, a largely white and Hispanic neighborhood. He had earlier questioned a white kid hanging out in front of a factory. "He was breaking his neck looking back at us; we thought he was a burglar." It turns out he was waiting for a friend. Another night in another precinct, the sergeant saw two black kids on bikes. "One guy's arm was hanging straight down, like he was carrying a gun. When they saw us, the other guy took off on his bike and threw a bag away. It was felony-weight drugs." Are you ignoring whites with guns? I asked him. "Of course not; I could see the same thing tonight," he said impatiently. "I don't use race at all. The only question is: are you raising my level of suspicion? Fifteen minutes after a stop, I may not even be able to tell you the color of the guy."

Even car stops on city streets usually have more context than on a highway. "If we pass four or five guys in a car going the opposite direction," explains the Queens sergeant, "and they're all craning their necks to see if we notice them, we may reverse and follow them for a while. We won't pull them over, but our suspicion is up. We'll run their plates. If the plates don't check out, they're done. If they commit a traffic violation, we won't pull everyone out of the car yet; we'll just interview the driver. If he doesn't have paperwork, it may be a stolen car. Now everyone's coming out to be frisked."

Hard as it is to believe, criminals actually do keep turning around to look at officers, though it would seem an obvious give-away. "Thank God they're stupid, or we'd be out of a job," the sergeant laughs.

But urban policing depends on another race-neutral strength: it is data-driven. The greatest recent innovation in policing was New York's Compstat, the computer-generated crime analysis that allows police commanders to pinpoint their enforcement efforts, then allows top brass to hold them accountable for results. If robberies are up in Bushwick, Brooklyn, the precinct commander will strategically deploy his officers to find the perpetrators. Will all the suspects be black? Quite likely, for so is the neighborhood. Does that mean that the officers are racist? Hardly; they are simply going where the crime is. In most high-crime neighborhoods, race is wholly irrelevant to policing, because nearly all the residents are minorities.

Urban police chiefs worry about the data-collection mania as much as highway patrol commanders do. Ed Flynn, chief of police for Virginia's Arlington County, explains why. Last year, the black community in his jurisdiction was demanding heavier drug enforcement. "We had a series of community meetings. The residents said to us: 'Years ago, you had control over the problem. Now the kids are starting to act out again.' They even asked us: 'Where are your jump-out squads [who observe drug deals from their cars, then jump out and nab the participants]?' " So Flynn and his local commander put together an energetic strategy to break up the drug trade. They instituted aggressive motor-vehicle checks throughout the problem neighborhood. Cracked windshield, too-dark windows, expired tags, driving too fast? You're getting stopped and questioned. "We wanted to increase our presence in the area and make it quite unpleasant for the dealers to operate," Flynn says. The Arlington officers also cracked down on quality-of-life offenses like public urination, and used undercover surveillance to take out the dealers.

By the end of the summer, the department had cleaned up the crime hot spots. Community newsletters thanked the cops for breaking up the dealing. But guess what? Says Flynn: "We had also just generated a lot of data showing 'disproportionate' minority arrests." The irony, in Flynn's view, is acute. "We are responding to heartfelt demands for increased police presence," he says. "But this places police departments in the position of producing data at the community's behest that can be used against them."

The racial profiling analysis profoundly confuses cause and effect. "Police develop tactics in response to the disproportionate victimization of minorities by minorities, and you are calling the tactics the problem?" Flynn marvels.

However much the racial profilers try to divert attention away from the facts of crime, those facts remain obdurate. Arlington has a 10 percent black population, but robbery victims identify nearly 70 percent of their assailants as black. In 1998, blacks in New York City were 13 times more likely than whites to commit a violent assault, according to victim reports. As long as those numbers remain unchanged, police statistics will also look disproportionate. This is the crime problem that black leaders should be shouting about.

But the politics of racial profiling has taken over everything else. Here again, New Jersey is a model of profiling pandering, and it foreshadows the irrationality that will beset the rest of the country. In February 1999, New Jersey governor Christine Todd Whitman peremptorily fired the head of the state police, Colonel Carl Williams, whose reputation for honesty had earned him the nickname "The Truth." It was the truth that got him fired. The day before his dismissal, Williams had had the temerity to tell a newspaper reporter that minority groups dominate the cocaine and marijuana trade.

Of course, this information had constituted the heart of DEA reports for years. No matter. Stating it publicly violated some collective fairy tale that all groups commit drug crimes at equal rates. Whitman's future political career depended on getting Williams's head, and she got it. One scapegoat was not enough, however. The New Jersey state troopers who shot at the van are now on trial for attempted murder—a wildly trumped-up charge—and the attorney general has been prosecuting the case in a flagrantly political fashion.

One way to make sure that nasty confrontations with the facts about crime don't happen again is to stop publishing those facts. And so the New Jersey state police no longer distribute a typical felony-offender profile to their officers, because such profiles may contribute, in the attorney general's words, to "inappropriate stereotypes" about criminals. Never mind that in law enforcement, with its deadly risk, more information is always better than less. Expect calls for the barring of racial information from crime analysis to spread nationally.

The New Jersey attorney general's office has also dropped its appeal of a devastating 1996 trial court decision that had declared the state police guilty of "institutional racism." Using Lamberth's New Jersey traffic study as proof of racial profiling, the court dismissed drug indictments against 17 blacks without so much as glancing at the facts of their cases. The court was wrong on the evidence and wrong on the law, but the case now stands permanently on the books as the most important judicial decision to date on racial profiling.

Next, the New Jersey attorney general himself dismissed en masse drug and weapons charges against 128 defendants. The defendants all alleged that state troopers had pulled them over merely because of race. The attorney general was not willing to defend the state's officers and so let the defendants go free. In one case, the defendants' car allegedly passed a marked cruiser at 75 miles an hour; the occupants were openly smoking pot and drinking; the trooper found cocaine—hardly a case of racial profiling, hard or soft. Numerous requests to the attorney general's office for comment on the case have gone unanswered.

New Jersey will soon monitor the length of traffic stops that individual officers make and correlate it to the race of the motorist. It will also monitor by race the computer checks that individual officers run on license plates, on the theory that racist officers will spend more time bothering innocent black motorists and will improperly target them for background checks. Of course an officer's stop and arrest data will be closely scrutinized for racial patterns as well. And if in fact such investigatory techniques correlate with race because more minorities are breaking the law? Too bad for the cop. He will be red-flagged as a potential racist.

These programs monitoring individual officers are present in all jurisdictions that, like New Jersey, operate under a federal monitor. Along with the new state requirements for racial data collection on a department-wide basis, they will destroy assertive policing, for they penalize investigatory work. The political classes are telling police officers that if they have "too many" enforcement interactions with minorities, it is because they are racists. Officers are responding by cutting back enforcement. Drug arrests dropped 55 percent on the Garden State Parkway in New Jersey in 2000, and 25 percent on the turnpike and parkway combined. When the mayor and the police chief of Minneapolis accused Minneapolis officers of racial profiling, traffic stops dropped 63 percent. Pittsburgh officers, under a federal consent decree monitoring their individual enforcement actions, now report that they are arresting by racial quota. Arrests in Los Angeles, whose police department has been under fire from the Justice Department, dropped 25 percent in the first nine months of 2000, while homicides jumped 25 percent.

The Harlem residents who so angrily demanded more drug busts from Mayor Giuliani last February didn't care about the race of the criminals who were destroying their neighborhood. They didn't see "black" or "white." They only saw dealers—and they wanted them out. That is precisely the perspective of most police officers as well; their world is divided into "good people" and "bad people," not into this race or that.

If the racial profiling crusade shatters this commonality between law-abiding inner-city residents and the police, it will be just those law-abiding minorities who will pay the heaviest price.