Criminal Justice Section
Criminal Justice Magazine
Volume 17 Issue 2
Racial Profiling Revisited:
"Just Common Sense" in the Fight Against Terror?
By David A. Harris
We have all heard, many times, that the events of September 11 "changed everything." Many political issues that had stood front and center in the national debate before that day abruptly assumed back-burner status; others disappeared altogether.
Racial profiling—an issue of broad national concern before September 11—did not become less important nor did it disappear. Instead, it was recast and dramatically changed. Before September 11, polling data had shown that almost 60 percent of Americans—not just African Americans and Latinos, but all Americans—understood what racial profiling was, thought it was an unfair and unwise law enforcement tactic, and wanted it eliminated. After September 11, that same percentage, including those minority citizens most often subjected to past profiling, said that they thought some racial or ethnic profiling in the context of airport searches was acceptable and even necessary as long as the group profiled consisted of Arabs, Muslims, and other Middle Easterners.
This is not particularly hard to understand. After all, the 19 suicide hijackers who attacked the World Trade Center and the Pentagon on September 11 were all young Arab Muslims from the Middle East. And the group claiming responsibility for the attacks, al-Qaeda, draws its legitimacy from an interpretation of Islam that cites the wrongs done to Muslims in the Middle East by nonbelievers. Therefore, most people say it simply makes sense to focus our law enforcement energies on Arab and Middle Eastern men; like it or not, they are the people who constitute the real threat to us, and no amount of political correctness will change that.
But just because it is easy to understand this reasoning does not mean it is correct. No matter how intuitively obvious it may seem, the use of racial and ethnic profiling, even in today’s circumstances, is a huge mistake that jeopardizes the strength of our antiterrorism initiatives. When we look at what we have learned about racial profiling on highways and city streets over the last few years, we see there is good reason to avoid using this tool to fight al-Qaeda—to refuse to take the easy, quick-fix path of racial profiling.
In times of great national fear and real threat, it is tempting to say that we must do everything we can to make ourselves safe, even if that means sacrificing individual liberties we normally seek to preserve. But we should not restructure our society and the principles that make it great without asking whether these changes would, in fact, make us any safer.
What we knew before September 11
Before September 11, a national consensus had emerged about racial profiling: It is a biased and unfair police practice that Americans wanted eliminated. This consensus resulted, in no small part, from the data released in the mid- 1990s that substantiated, for the first time, the stories told by African Americans and Latinos of being stopped, searched, and treated like suspects by the police. The data proved to a majority of Americans that these were not just stories or excuses used by criminals trying to escape wrongdoing, but manifestations of a real phenomenon, something we could observe and measure.
For example, in data submitted by the Maryland State Police to a federal court, African Americans comprised more than 70 percent of all drivers stopped and searched, even though they represented just 17 percent of all drivers on the road. (Report of John Lamberth, Wilkins v. Maryland State Police et al., Civil No. MJG–93–468 (D. Md. 1996).) To be sure, not every police department, and not every police officer, used racial or ethnic profiling. But it was common enough that most minority men and women, of all ages, professions, and classes, had either experienced it first-hand, or had close relatives or friends who had.
The availability of these data brought about an important shift in the public debate. Fewer defenders of the police denied the existence of racial profiling. Rather, they insisted that such profiling was not racist; it was just common-sense crime fighting. Defenders of profiling argued that because African Americans and Latinos were disproportionately represented among those arrested and imprisoned, smart police officers focused their enforcement efforts on these racial and ethnic groups. The fact that this resulted in disproportionate numbers of stops, searches, and arrests of African Americans and Latinos was, according to a spokesman for the Maryland State Police, simply "the unfortunate byproduct of sound police policy." (Michael Fletcher, Driven to Extremes: Black Men Take Steps to Avoid Police Stops, Wash. Post., March 29, 1996, at A1.)
In order to advance the debate, researchers and scholars needed to confront the assumption head-on. Did racial profiling, in fact, make for better and more productive law enforcement? It was already clear that racial profiling existed, given the data from the mid-1990s. Many Americans were already aware of the high social costs of profiling: the distrust of police, cynicism about the courts and the law, an unwillingness to believe police officers testifying as witnesses in court. These factors were often dismissed as just the costs of fighting crime and drugs, since it was assumed that profiling did, indeed, fight crime. Thus the importance of testing the assumption. Did racial profiling really help police officers find the bad guys more often?
Data that became available in the late-1990s made such testing possible for the first time. These data, from police departments of all kinds from all over the country, allowed the study of "hit rates"—the success rates for police stops and searches using racial profiling as opposed to those stops and searches using traditional, nonracial criteria that focus simply on observation of suspicious behavior. Hit rates indicate the rate at which police find what they seek— drugs, guns, people with arrest warrants, and the like—when they execute a stop and search. In these studies, stops and searches of Caucasians were not counted as profiling because they were based not on race, but on observed behavior that appeared suspicious. Stops and searches of African Americans and Latinos, on the other hand, may have been based on some suspicious behavior, but were driven, overwhelmingly, by race or ethnic appearance. This is why, in all of these studies, African Americans and Latinos were stopped in numbers greatly disproportionate to their presence on the highways, roads, or city sidewalks. With these data broken down by racial categories for each of the police departments studied, I was able to make an apples-to-apples, side-by-side comparison of police behavior within particular police departments, both with and without the variable of race or ethnicity. ( See David A. Harris, Profiles in Injustice: Why Racial Profiling Cannot Work, chapter 4 (The New Press, 2002).)
The results of these hit rate studies were striking, all the more so for their consistency across many different jurisdictions and law enforcement agencies. The data on hit rates show that targeting law enforcement using racial or ethnic appearance does not, in fact, improve policing. It actually makes policing worse—less successful, less productive, less likely to find guns, drugs, and bad guys. Contrary to what the proponents of profiling might expect, hit rates were not higher using racial profiling. In fact, hit rates for race-based stops were lower—significantly lower—than the hit rates for traditional, nonprofile-based policing. That is, when police used racial and ethnic profiling to target black and brown populations as suspicious, the results they got were uniformly poorer than the results they got when they stopped whites simply on the basis of suspicious behavior. Racial profiling, then, doesn’t improve policing; it pulls it down, delivering less bang for the law enforcement buck. Even if we ignore the high social costs—distrust of all government, including police and the legal system; exacerbation of existing problems such as residential segregation and employment discrimination; and destruction of valuable law enforcement initiatives such as community policing—racial profiling as a means to crime reduction simply does not deliver.
A little history
The use of race or ethnic appearance as a proxy to tell police officers who the likely criminals might be did not begin recently. Some trace profiling back all the way to the slave patrols, which were empowered to stop and demand an explanation from any black person unaccompanied by a white person, as well as to search blacks’ dwellings for contraband, such as books. But it was in the 1980s that certain segments of law enforcement certainly honed the tactic to perfection. Racial profiling was not an accident, and it was not the result of a few bad apples in policing. It was the predictable outcome of a number of decisions by the U.S. Supreme Court that allowed police to use traffic enforcement as a convenient and universally available means to make end runs around the requirements of the Fourth Amendment. And it was the federal government, through the Drug Enforcement Administration, that deliberately made use of these opportunities.
As far back as 1973, the U.S. Supreme Court confronted the possibility that police officers might use arrests, traffic stops, and other traditional enforcement mechanisms as pretexts to do what would otherwise be forbidden by the Fourth Amendment in the absence of probable cause or a warrant or both. In Robinson v. U.S., 414 U.S. 218 (1973), the Supreme Court allowed police to perform a full search incident to a traffic arrest, despite the lack of any reason to fear the presence of weapons or the destruction of evidence, the two traditional rationales for searches incident to a valid arrest. The real intentions of the police officer in making the arrest, that is, whether the arrest might be nothing but an excuse for a search or seizure for which there was no supporting evidence at all, did not matter. All that was important, the Court said, was that there was valid probable cause for the arrest.
Robinson set the stage for a series of decisions through the 1970s, 1980s, and 1990s that increased police discretion and the ability to use stops and searches as pretexts to do what would otherwise be impermissible under the Fourth Amendment. These cases reached their high-water mark in Whren v. U.S., 517 U.S. 806 (1996). In Whren, the Court resolved an argument that had split the federal circuits and state courts: Could police officers use traffic enforcement as an excuse to stop drivers for investigation of other crimes when there was no evidence of these crimes? Since the mid-1980s, the use of traffic enforcement as a pretext to investigate drugs and drug trafficking on the roads had been championed by the federal government itself. The federal Drug Enforcement Administration had been using a profile to spot potential drug couriers in airports; the Supreme Court’s cases and those from lower courts encouraged the DEA to try to engraft this method onto traffic enforcement on highways across the country. A DEA program called Operation Pipeline was set up in the 1980s to spread the gospel of highway drug courier profiling nationwide. The DEA trained tens of thousands of state and local police officers in the tactics of profiling through Operation Pipeline; these officers then went back to their own departments to train others and to set up specialized drug interdiction units. The DEA says that none of its training encouraged the police to use race as part of its profiles, but training materials sometimes showed otherwise. For example, one training video shows officers making several mock stops; in each one, the driver stopped has a Hispanic surname. The DEA and other federal agencies were also disseminating intelligence in the 1980s and 1990s that blamed trafficking in particular drugs on identified ethnic groups.
The Whren case can, thus, be seen as the Supreme Court’s final word and official blessing of these practices. The parties confronted the Court with the question of pretext stops and their use as tools in racial profiling—the briefing of the case is replete with references to these issues—but the justices were unimpressed. In a unanimous opinion, the Court said that unlimited police discretion to make traffic stops—even the use of this power in a racially biased manner—had no place in the Fourth Amendment analysis. The Court dismissed them as nothing more than the possible grounds for a lawsuit based on the Equal Protection Clause.
As a result, racial profiling became a national topic of conversation in the mid- and late-1990s not simply because of occasional overzealous law enforcement. On the contrary, profiling was an entirely predictable outcome of U.S. Supreme Court decisions that broadened police power and discretion to make traffic and pedestrian stops and searches to the point that this discretion was for all practical purposes unlimited. That police took advantage of the authority these decisions gave them is understandable. Indeed, it would have been surprising if they had not.
More than one way to skin a cat
What Whren and recent cases such as Atwater v. Lago Vista, 533 U.S. 924 (2001) (Fourth Amendment is no barrier to arrest and accompanying search of woman for offense that carries no possible sentence of incarceration), make clear is that any changes in police use of racial profiling will almost certainly not come from the courts. The few successful private lawsuits regarding these practices, such as those in New Jersey and Maryland, are not nearly so numerous as those that have failed. ( E.g., Nater et al. v. Vogel, 106 F.3d 415 (11th Cir. 1997); Chavez v. Illinois State Police, 251 F.3d 612 (7th Cir. 2001); see also Daryl Kelly, Federal Racial Profiling Suit Is Dismissed, L.A. Times, March 19, 2002.) Thus, traditional litigation strategies, although certainly a tool that should not be ignored, do not represent the best method for attacking these issues.
Yet the inhospitability of the courts to opponents of racial profiling is, perhaps surprisingly, where the story grows more encouraging. Lack of success in using judicial routes to create change has forced opponents of profiling to find other avenues: the use and mobilization of public opinion and concerted pushes for legislative measures. These efforts have met with surprising success. By the beginning of 2002, 15 state legislatures had enacted some kind of law concerning racial profiling. These laws ranged from simple requirements that police departments collect data on some or all of their traffic stops to laws that require new antiprofiling policies, training, and other initiatives, in addition to data collection. Besides these new laws, hundreds of police departments have, on their own and without any state law requirement, begun to collect data to track their traffic and other stops and searches, and have revised their policies. In a way, this is far more important than state legislative action. It shows that these police departments have recognized the fact that it is in their own interest to face the issue of profiling head-on and to forthrightly collect information on police practices, regardless of what it may or may not show.
It is this last set of changes—police departments themselves deciding that it is important to take some action on profiling—that points the way to the future. By getting beyond the old questions of whether or not profiling exists, as it surely does in many places, and whether anything should be done about it, the public debate can move to instituting better ways of policing. Many individuals involved in policing have long characterized modern law enforcement using a stark dichotomy. We can have good, tough, effective law enforcement that brings down the crime rate or we can have "kinder, gentler" policing that respects citizens and their rights, but we cannot have both. The example used to illustrate this "either-or" approach is New York City. During the 1990s Mayor Rudolph Giuliani and the New York Police Department used tough, zero tolerance law enforcement against low-level, quality-of-life crimes. The crime rate dropped an astonishing seven years in a row. But while New York accomplished much in the way of crime reduction, it was not alone. Although New York City, with its nationally ambitious mayor and its role as media capital of the world, got most of the attention, it was only one of many major American cities that saw a dramatic fall in crime rates. And most of these cities did not use New York City’s zero tolerance model. In fact, a number of large cities, from San Diego to Boston, used entirely different methods that produced drops in crime that were the same or greater than those seen in New York—and did so with less manpower on a per capita basis. And, unlike New York, these cities put their anticrime initiatives together in cooperation with their minority communities, instead of in opposition to them. Now, as New York deals with a backlash of distrust between police and minority communities, these other cities have built partnerships between their police and minority communities that can serve as the basis for strong efforts to reduce crime far into the future.
It is examples such as these that demonstrate that we should reject the "either-or" dichotomy. The idea that we can have either effective policing or policing that respects citizens, but not both, is a myth. We can—we must—have both. Police departments nationwide are taking concrete steps to make this happen. They are not following pie-in-the-sky academic theory, but instead using solid management tools and putting the police accountability at the top of their agenda. Among the approaches employed are:
• the collection of analysis of data on traffic and pedestrian stops and the aftermath of those stops, including consent searches and their results;
• early warning systems, designed to help supervisors manage officers under their charge and to spot officers who may need more training or who should be taken out of policing in order to avoid further problems down the road;
• greater supervisory responsibility of police middle management—chiefly sergeants and lieutenants—for the actions of those officers under their commands doing line policing;
• improvement of the conduct of police officers during traffic stops;
• better training designed to ferret out and combat pre-existing biases and to show officers how bias-free policing improves police performance;
• changes in police organizational incentive structures, to promote the type of policing that is both effective and respectful; and
• the use of technology, such as video and audio recording and real-time data tracking systems, to observe and make records of police actions on the street.
(David A. Harris, Profiles in Injustice: Why Racial Profiling Cannot Work, chapter 7 (The New Press, 2002).)
In short, there is more than one way to skin the law enforcement cat. Avoiding racial and ethnic profiling will improve policing, and, along with other accountability-based reforms, can take it beyond the demonstrably false, "either-or" myth.
What about now?
Since September 11, of course, much has changed in our national discussion of policing strategy, not least the shift in focus from anticrime to antiterrorism. Given all that we now know—a lethal attack upon thousands of innocent civilians by al-Qaeda, a Muslim organization in which the weapons of mass destruction were airplanes hijacked and flown by 19 Muslim Arab men for the Middle East—how is it possible to ignore the fact that being Arab or Muslim means that one is more likely to be involved in terrorism? Doesn’t it just make sense to use a racial or ethnic profile that singles out Arabs, Muslims, and other Middle Easterners for greater attention, including more frequent stops and searches? Isn’t anyone suggesting tactics to the contrary simply suffering from blindness of political correctness? Syndicated columnist Kathleen Parker has given voice to this way of thinking. "When a police officer apprehends and searches an African American only because he’s black, assuming no other mitigating factors, that’s unjustified racial profiling. When an airport security guard searches a male of Middle Eastern extraction following a historical terrorist attack by males of Middle Eastern extraction, that’s common sense." Parker wrote that "a terrorist attack of such enormous proportions, followed by a declaration of war, makes racial profiling a temporary necessity that no patriotic American should protest." (Kathleen Parker, All Is Fair in This War Except for Insensitivity, Chi. Trib., Sept. 26, 2001, at 19.)
To a large extent, that is the way things have gone. In the wake of the attacks of September 11, Arabs and other Middle Easterners were removed from planes. Others were repeatedly searched and questioned. More serious, the U.S. Department of Justice detained hundreds of Arabs and Muslims—not one of whom has been linked to the September 11 attacks. The department also undertook the "voluntary" questioning of 5,000, and then another 3,000, young Arab men who were, according to the U.S. attorney general, not suspects, in the hope of learning something, anything, that might help. (Law enforcement professionals reacted by quietly refusing to take part in the questioning and commenting that this tactic differed markedly from established methods for conducting antiterrorism investigations.) The government has also announced an "absconder initiative" in which it will focus its deportation efforts on 6,000 men from Middle Eastern nations who have violated their visas—despite the fact that they make up only the tiniest fraction of all the hundreds of thousands of violators.
These actions may seem intuitively correct, but it is not at all clear that they are, in fact, the right way to proceed. If we look at what has happened in the past with racial profiling, we will see that there are real reasons to hesitate to take the profiling path now. In fact, using racial or ethnic profiles to protect ourselves against al-Qaeda operatives will likely have an effect opposite of what we desire. Our antiterrorism cause will be set back, not advanced.
First, think of what we might call the gold standard of traditional policing: the observation of suspicious behavior. Veteran police officers know there is no substitute for this. It’s not what people look like that tells the savvy officer who is suspicious; it’s what people do that matters. Using a racial or ethnic profile to decide whom to regard as suspicious cuts directly against this experience-based principle. Some people do look different than the majority, of course, and we know that those in the majority often tend to see people who are different as threatening or suspicious. The net result is that those who look different, especially in some skin-deep way that we equate with danger, come under immediate suspicion without any suspicious behavior; our law enforcement agents shift their attention from what counts—how people behave—to what doesn’t: what they look like. We could not make a bigger mistake than to take our eyes so completely off the ball.
A second point is related to the first. If we use racial and ethnic profiles, instead of behavioral clues, to decide which people we think should be treated as suspects, we will greatly enlarge our suspect pool. This is true whether our profiles are based only on race or ethnicity (a very unlikely event) or if race or ethnicity is just one characteristic among many in a multifactor profile. This means that our enforcement resources will be spread thinner than they would if they were part of an effort focused on suspicious behavior. This is not a trivial consequence. On the contrary, our enforcement resources will be stretched to address many more individuals, and, therefore, they will be that much less effective. This points to one of profiling’s inherent flaws: It is always overinclusive. Judging whether individuals are suspicious based on their racial or ethnic characteristics means that many people who would not otherwise draw the attention of law enforcement get swept into the dragnet. Of course, this is exactly what happened to African Americans and Latinos profiled during the war on drugs. It wasn’t just the drug dealer on the corner, but the minister, the doctor, and the businessman who were stopped, searched, and treated like suspects. Many more minorities were stopped because of their color than would have been the case had police simply looked for suspicious behavior. In the antiterrorism context, not even the FBI has unlimited manpower. Even with the aid of state and local antiterrorism task forces, little can hold us back more completely than increasing the enforcement burden beyond what focusing on behavior would call for.
If focusing on suspicious behavior is the gold standard in policing, an accompanying principle is the importance of collecting and using information and intelligence. If we are to avoid attacks in the future by al-Qaeda operatives based on our own soil, we need to do much better in the intelligence arena than we did before September 11. And that, of course, means we will have to get information from those likely to know the Arab, Muslim, and Middle Eastern men we might suspect. This information is going to have to come not from the population at large, but from the Middle Eastern communities themselves; there is simply no avoiding this. It stands to reason, then, that what we need most right now are good, solid relations with the Arab and Muslim communities in the United States. Profiling that focuses on Arab and Muslim heritage will effectively communicate to these very same communities that we regard all of their members not as our partners in law enforcement and terror prevention, but just the opposite: as potential terrorists. When every young Arab or Muslim man is effectively labeled a potential terrorist, when many are detained indefinitely on petty immigration violations, and even those who have come forward to help have been rewarded with incarceration, it is not hard to imagine the result: alienation and anger toward the authorities at a time when we can least afford it.
The interesting thing about the "common sense" of profiling Middle Easterners is how roundly and quickly many law enforcement professionals rejected it. Attorney General John Ashcroft’s directive ordering the "voluntary" questioning of 5,000 Middle Eastern men was met by skepticism in a number of major police departments. (Fox Butterfield, Police Are Split on Questioning of Mideast Men, N.Y. Times, Nov. 22, 2001.) Command staff quickly recognized the damage that this questioning would do to their long-term efforts to build crime-fighting partnerships with their Middle Eastern communities. Some departments quietly refused to take part or were happy to leave the task to FBI agents; a few police officials publicly rejected the Department of Justice’s requests that they take part. Eight former FBI officials, including the former chief of both the FBI and CIA, William H. Webster, went on record with his doubts about the law enforcement value of these tactics; one former official called the wholesale questioning "the Perry Mason School of Law Enforcement." (Jim McGee , Ex-FBI Officials Criticize Tactics on Terrorism, Wash. Post., Nov. 28, 2001.) Little wonder, then, that senior U.S. intelligence officials circulated a memorandum early in the fall that warned about the dangers of profiling. This memorandum, first reported in the Boston Globe, urged law enforcement and intelligence agents to balk at racial profiling. Profiling would fail, the memorandum said; the only way to catch terrorists was the observation of suspicious behavior. (Bill Dedman, Airport Security: Memo Warns Against Use of Profiling as Defense, Boston Globe, Oct. 12, 2001.) Too bad that this warning never seems to have penetrated to the senior levels of the Department of Justice leadership.
Despite all of this, many continue to argue that it doesn’t make sense to screen anyone at airports besides young Arab and Middle Eastern men. After all, why waste time and resources stopping or searching anyone else when we already know exactly who the "real" suspicious people are? This argument shows a serious misunderstanding of the nature of what we are up against in al-Qaeda. The most important thing for us to realize about this organization is that beyond its vicious and murderous elements, it has also shown itself to be intelligent, patient, and thoroughly adaptable. These are the qualities that make it a formidable enemy. The attack on the World Trade Center on September 11 was not the first but the second assault on this landmark. When the first attack in 1993 failed to accomplish their goal, the terrorists pulled back and spent eight years devising an entirely new method of attack—planned to the smallest detail and then practiced so that it could be carried out almost perfectly. Despite our important military successes in Afghanistan, this set of qualities remains very much alive in the al-Qaeda structure. In the aftermath of September 11, we began to harden cockpit doors, to check carry-on bags for even the smallest potential weapons, and to profile Middle Eastern men. Al-Qaeda’s answer was Richard Reid—a non-Arab, non-Middle Easterner from England; a British citizen with a valid British passport and a bomb in his shoe. Clearly, they knew what we were looking for, and they did not repeat what they had done in the past.
Those who insist that it is a waste of time to look at any non-Arab, who want all Arabs carefully screened and searched despite the lack of any suspicious behavior, seem to have missed this point entirely. Terrorists generally, and al-Qaeda specifically, will always look for new weaknesses to use as their point of attack. Just such a weakness became glaringly obvious in mid-January 2002 when the government announced that it could not meet the deadline for having all bags aboard airliners electronically screened for explosives. Instead, the authorities said, they would use a group of other stopgap measures such as partial bag matching. It is much more likely that the next attack on an airplane will come not in the form of a suicide hijacking, but through efforts to bring an explosive aboard a plane—either in the checked baggage of an al-Qaeda suicide soldier who boards the plane and avoids detection through bag matching, or in the luggage of an innocent passenger. Who can forget the foiled terrorist attempt some years ago to bomb a plane by putting explosives in the luggage of the terrorist’s unsuspecting pregnant girlfriend? And, as some federal law enforcement officials have already said, the next face of a terrorist we will see is unlikely to be a Middle Easterner; instead, we may see someone from Asia, Malaysia, the Philippines, or even a European country. Thus, random searches and searches of non-Arabs are, in fact, extremely important security tactics. Without them, we create a weak point that begs to be exploited by our enemies.
Those who insist on the "commonsense" approach of focusing on Middle Eastern men would, of course, have us ignore all of this. They want a fast food solution—something convenient and comforting for "us" that only inconveniences "them." But there is little reason to think that type of profiling will see any greater success than profiling for drugs and other crimes did. If we go down that failed path we will likely reap the same kind of reward: poor results and the alienation of our citizens from the police at a time when we most need cooperation between the two. Ironically, if we keep ourselves focused on traditional methods of police procedures—surveillance of suspicious behavior and the collection and effective use of intelligence—we do not have to fail. We can do a better job of policing, even as we do right by our fellow citizens.
David A. Harris is the author of Profiles in Injustice: Why Racial Profiling Cannot Work (The New Press, 2002). He is Balk Professor of Law and Values at the University of Toledo College of Law, and the author of numerous law review articles on search and seizure, racial profiling, stop and frisk, and other related topics.