Baltimore IMC :
Baltimore IMC

Commentary :: Crime & Police

Racial/Ethnic Profiling

This is a review of the legal, sociological, and journalistic writings about racial/ethnic profiling. It concludes with a set of patches designed to reduce stops and searches which are intentionally discriminatory.
Racial Profiling

Howard J. Ehrlich

Profiling--the deliberate institutional policy of a law enforcement agency which authorizes its agents to stop someone because of her or his race/ethnicity

In October 2001, the United States Supreme Court refused to take on the issue of "racial profiling." The case involved a robbery and attack on a 77-year-old White woman in 1992 in Oneonta, NY. Although she did not see the perpetrator, she glimpsed his forearm and so identified him as Black. She said that he might have been cut or scratched in the struggle and that, further, she deduced that he was young from listening to the pace of his footsteps as he left the scene. The local police requested and received a list of all Black students at the state university in the town. The city, campus, and state police questioned almost 100 students and examined them for scratches. In addition, they also stopped and examined 200 Black residents who weren't students. No one was ever arrested. Students, townspeople and others brought suit claiming that their constitutional rights had been violated. The case made its way to the Supreme Court and was rejected for consideration. Was this a case of racial profiling? What constitutional issues were involved?

In the beginning, long before the currency of the term "racial profiling" the police routinely stopped cars for petty offenses in order to gain the opportunity to look around the stopped car, closely evaluate the behavior of the car's occupants, and maybe intimidate them to reveal their hidden cache or nefarious mission. Later, as the drug traffic began to grow in the early 1970s, the Drug Enforcement Administration, the DEA, began to profile people by their behavior. The ACLU identifies its start at the Detroit Metropolitan Airport, but by 1979 it had spread to more than 20 airports. The behaviors profiled were: Was the person nervous? Did he pay for his ticket in cash and with large bills? Was his destination or origin a known place for drug smuggling or dealing?

By the 1980s, the drug market changed as crack began to flood the cities. Skin color most often became the sole criterion. Further, profiling now spread to all places of public transportation, not just airports. By the 1990s, it spread to the interstate highways, bursting into public awareness on I-95 especially around Baltimore and throughout New Jersey.

What touched off this spread were the growing use and acceptance of so-called "pretext stops?" Under the Fourth Amendment, the police may stop you if they have reasonable suspicion or probable cause to make the stop. And so, in the absence of reasonable suspicion, they began using any pretext to stop the car. Cars and drivers, as before, would be stopped presumably for a minor traffic infraction in order to give the police a closer look at the inside of the car and maybe even to search it.

Pretextual stops were approved in 1996 in a case called Whren v. US. The Supreme Court gave the police the authority to stop and search any car. All they had to do was establish that the driver was in some violation of the state's motor vehicle regulations. "Sorry sir, but I noticed that your tire tread was quite low. Would you mind stepping out of the car for a moment?"

In 1997, an Ohio court ruled that a police officer did not have to tell a driver that s/he could refuse to give consent to a search and could leave the scene, if not otherwise under arrest. In that year, too, a Maryland court said it was ok to order passengers to get out of the car even if there was no reasonable suspicion that they were offending. In 1999, in a Wyoming court, the police were given approval to search the purse of a passenger, after the arrest of the driver. Even though the passenger had done nothing to warrant the stop nor presented herself in any suspicious manner, she could still be searched.

Further, in April 2001, the Supremes ruled that a police officer who observed someone breaking a law can place them under full custodial arrest. What makes this so relevant to our concerns here is that this ruling applies especially where the law involved is a minor infraction and where the maximum penalty is a small fine. It is too early to know, of course, how this will affect stops and searches in racially profiled traffic incidents. See David A. Harris, Driving While Black: Racial Profiling on Our Nation's Highways. Special Report of the American Civil Liberties Union, June 1999.

The comments of Tim Wise, a writer and antiracist activist, seem to fit the pattern of court decisions.

The bottom line is that racial profiling doesn't happen because data justifies the practice, but rather because those with power are able to get away with it, and find it functional to do so as a mechanism of social control over those who are less powerful....No conspiracy here, mind you: just the system working as intended, keeping people afraid of one another and committed to the maintenance of the system, by convincing us that certain folks are a danger to our well-being. ("Racial Profiling and its Apologists," Z Magazine, March 2002)

Of course, racial profiling and discrimination take place in a community context. Police organizations act out the norms and patterns of prejudice of the community they serve. On the assumption that arrest rates are a direct function of profiling, we can assert the following implications from recent criminological research. Profiling is greatest where the Black population is substantial in size but less than a majority; where the governing elites are White; where residential segregation is characteristic of the population; and where the segregated communities are poor and economically disorganized. (See Howard J. Ehrlich, Notes on Black Male Incarceration on Race Relations, Perspectives, No. 26, April 2002, and Steven R. Cureton, "Justifiable Arrests or Discretionary Justice," Journal of Black Studies 30: 703-719, 2000.)

Demonstrating the practice of police profiling is not simple. As sociologists have discovered, the findings of social research are not necessarily the findings that are acceptable to jurists who are quite antagonistic to the social sciences. Moreover, there are some difficulties in the demonstration of racial profiling which I will try to illuminate. One set of findings is clear: Blacks are more likely to be stopped and searched than Whites and Latinos are more likely to be stopped and searched than Whites. However, Whites are more likely to be carrying illegal drugs, guns, or stolen property.

In Texas, an analysis of more than 16 million driving records in 1995 by the Houston Chronicle found that minority drivers who drove through the white enclaves in and around the major cities were twice as likely as Whites to be ticketed for traffic violations. In Bellaire, a White suburb surrounded by Houston, Blacks were 43 times more likely than Whites to receive traffic citations.

In New Jersey, 13% of the drivers on the NJ turnpike were Black, and Blacks and Whites were traffic code violators to the same degree, that is 73%. However, three out of four of those arrested were Black. In a 2000 study in New Jersey, hit rates among those who consented to be searched were 25% for Whites, 13% of Blacks, and 5% for Latinos.

In Maryland, 17% of the drivers on a section of I-95 were Black, but 73% of those stopped and searched were Black. Throughout the state, equal percentages of both groups when searched were carrying contraband. In a subsequent Maryland study covering 27, 308 State Police stops in 2000, 52% of those searched were Black as compared to 37% of Whites.

In Missouri, data from 634 police departments covering more than 450,000 stops over a four- month period showed that Black motorists were stopped at a rate 30% higher than Whites, and were searched 70% more often.

In Philadelphia, not only were similar results obtained for all of its 23 police districts, in 2000, but the researchers concluded that 36% of all minorities stopped by the police were stopped and detained without legal justification. In New York City, the police in a 1998 and 1999 study stopped and frisked disproportionately more Black men but the hit rates for Whites were actually higher. During the span of New York's "zero tolerance policy" under the mayoralty of Rudy Giuliani, the number of minorities stopped and frisked rose dramatically and even more so for their now disbanded "Street Crimes Unit." (For references to these studies and a more comprehensive discussion, see Sandra Bass, Social Justice 28, No.1, 2001.)

These are the gross numbers. They demonstrate that unequal treatment by the police is commonplace. What we will examine now are some of the problems associated with judgments about these data. A recent and comprehensive study in New Jersey is particularly revealing of the problems associated with the "numbers game." The study focused on speeders and nonspeeders on the New Jersey Turnpike. Speeding was defined as exceeding the speed limit by 15 mph. The study conducted from March to June 2001 employed photographs of drivers and radar-gun readings at 14 locations along the Turnpike. The ethnic identities of 26,334 drivers were agreed upon by two researchers. Sixteen percent of the drivers were identified as African American. What makes this study so interesting for our purposes-and it should be clear that this is a study of speeding not a study of profiling-is the complexity of the results. In the 65 mph zone where motorists are entering the Turnpike from Pennsylvania, African American drivers were significantly more likely to be speeding than were White drivers. This was apparently true even when age and gender were taken into account. In the 55 mph speed zone at the northern, more urbanized end of the Turnpike, there was no statistically significant ethnic difference between drivers. To add to this was the observation that drivers less than 45 years-of-age were three times more likely to speed than older drivers. There is even more to the complexity of these data. What is critical to me is the consequence of this report. It is not surprising that police and conservatives have taken this as supportive of their claim that stopping and searching Black drivers is proportionate to their rates of offending. Heather MacDonald, writing in City Journal, (Spring 2002) a publication of the conservative Manhattan Institute, celebrated the Turnpike study.

The anti-racial profiling juggernaut has finally met its nemesis: the truth....Until now, the anti-police crusade that travels under the banner of "ending racial profiling" has traded on ignorance. Its spokesmen went around the country charging that police were stopping "too many" minorities for traffic infractions or more serious violations. The reason explained the anti-cop crowd, was that the police were racist. They can argue that no more.

If you allow yourself to become trapped in the numbers game, as did MacDonald among many others, here are some the "numbers" you have to deal with. I present them in no particular order, because I believe that this randomness may mirror the problems we face here, and you may find other numbers you would like to add to the game. These include number of miles driven by different groups, age distribution of drivers; gender distribution of drivers; mileage, age, and gender distribution of those stopped; mileage, age, and gender distribution of those searched; race/ethnicity, age, and gender distribution of those searched; class by race/ethnicity, age, and gender distribution of those stopped; race/ethnicity, class ,age, and gender distribution of those searched; reasons for stopping by all of the above, and reasons for searching. Now consider issues of methodology: How many observational points are used in space and in time; how accurate are the observers, what was the form of observation, what is the nature and error rate of the recording instruments; how are people classified by race/ethnicity?

Finally, for those who want to more directly monitor the police you need to consider the accuracy and detail of police records themselves, particularly since such records have obviously been designed to monitor the police themselves.

One spinoff of the numbers game is the "proportionality" routine. Police critics and supporters alike point to proportionality as the deciding factor--as if there were some truth out there called proportionality. What is proportionality? In its commonsense meaning it refers to a constant ratio between two quantities. In the application here, we might say that if Danireans were 10% of the Baltimore population, then they should constitute 10% of the drivers stopped on the highway.

Let's question this proposition. Do we really mean that every segment of the population be represented proportionally among all of those stopped by the police? Would this somehow entail greater justice than if there were not this proportionality? What is it we want to be proportional to? Since White males, for example, commit approximately 60% of the violent crimes, do we want to insist that 60% of all stops be of White males? Do we want to extend this to all searches too? And what about arrests? There are many other reasons to target White males: they are a disproportionate number of drunk drivers, killing more than 10,000 people annually and White males are twice as likely as blacks to molest children. (See Tim Wise, "Racial Profiling and its Apologists," Z Magazine, March 2002)

With regard to our common understanding of racial profiling, the numbers game has no end. For liberals, the only way to win is not to play. For the profiler, as long as the economically disadvantaged and sociological minorities are arrested and imprisoned with greater frequency than others, continuing the game is a winning strategy. This is not an argument against profiling. Rather, it is an argument against the use of racial/ethnic identity as the sole or major criterion in the decision of a police person to detain or charge someone.

The real issues are the police departments and the nature of discrimination. The character of police departments is revealed by the response of the New Jersey State Police to their present notoriety. The New York Times reports that senior commanders were aware of racial profiling by troopers as early as 1996, but they refused to do anything about it. They apparently rejected proposals to set up a monitoring program or to require troopers who stopped high numbers of minorities to undergo counseling or to be disciplined. (October 11, 2000) Thus, even if racial/ethnic profiling were not a matter of formal organizational policy, it was a practice acceptable to top policymakers. It was in fact a social norm-a shared, common standard of expected behavior. As such, it had as much power as a written rule or policy.

An illustration of how some troopers regarded the attempt to change this norm took place in early April 2001. After the state police superintendent publicly stated that 10% of New Jersey troopers were still practicing racial profiling, some troopers began wearing T-shirts picturing a skull, crossed pistols and the word "10 percenters."

Police profiling has taken on the status of a "social problem" not because it is a new event, but because of the developing public awareness of the corruption of major police departments. Dramatic instances of the police framing people--planting drugs, evidence, and even fingerprints--of shakedowns, of abusive treatment, and of brutality and the killings of unarmed civilians (most of them during routine traffic stops) have aroused deep public concern. In 2000, at least 40 police departments were involved in charges and internal investigations regarding abuse and brutality including such major agencies as the U.S. Customs Service, Los Angeles, Detroit, Indianapolis, Kansas City, Louisville, Miami, Philadelphia, Pittsburgh, Providence, Sacramento, Salt Lake City, Seattle, and Washington, DC. At a more extreme level, almost all studies reveal that Blacks and Latinos are more likely than non-Hispanic Whites to be shot by the police, fatally and non-fatally. This is true regardless of population size. In some cities, Blacks were six to seven times more likely to be shot than in others. (Police Executive Research Foundation, Deadly Force, 1992)

Although systematic data on discrimination within police departments is not easily accessed, it is my impression that virtually every major police department in the country has at least one suit against it claiming discrimination in training, assignments, or promotions. Although one may argue that thiikely true of almost every major American corporation as well, these businesses are not charged with law enforcement.

In summary, racial/ethnic discrimination, the abuse of power, and the use of excessive force are characteristic of police. Racially-profiled traffic stops or the stopping and searching pedestrians are likely to continue as long as these more serious events continue.

It would be instructive now to examine how the public regards the police and racial/ethnic profiling. The most recent data come from The Gallup Poll Social Audit conducted between March and May 2001. The pollsters conducted telephone interviews with 1,003 Black and 895 White respondents. There are two central findings that concern us. First is the gap between the two groups in their perception. Nearly nine out of ten Whites feel that they are treated fairly by local or state police. This compares to just over half of Black respondents. When it comes to racial profiling, 83% of Blacks but only 55% of Whites believe that it is widespread. Obviously, the gap itself is an obstacle to developing policy.

Secondly, are the reports of the bias in police contacts. The researchers asked: "Can you think of any occasion in the last thirty days when you felt you were treated unfairly in the following places because you were Black? How about in dealings with the police?" One out five Blacks said "yes," they were "unfairly treated," and that figure becomes 31% when we look only at the responses of Black men. This question, in good survey practice, limited the respondents to the past 30 days. When the researchers ask about lifetime experiences, reports of profiling increase. "Forty-four percent of Blacks feel that the police have stopped them at some time in their life because of their race or ethnic background."

The Gallup survey is essentially replicated in a Washington Post study which was conducted a few months later in October 2001. It reported that 37% of Black men and 20% of Hispanics felt they had been "unfairly stopped because of their race or ethnicity." Whatever arguments may be invoked to downplay the prevalence of racial/ethnic profiling-- methodological, constitutional, or matters of police procedure--the reports of those victimized remain the strongest evidence of the differential treatment accorded people of color by the police.


Part of the problem in dealing with racial/ethnic profiling is that analysts are conflating three different forms of discrimination. First of all let's define discrimination. There is a considerable consensus among social scientists that discrimination refers to actions which deny equal treatment to a group or to people believed to be members of a group. This is of course what profiling is about. People are being denied the right to walk or drive or fly without being stopped or searched.

There are three types of discrimination relevant here. The first is individual discrimination. This refers to the personal acts of individuals intended to have a differential or harmful effect on the members of an outgroup. Many cops are racist and their activities in stopping people driving while Black reflects that racism. Profiling can also be a function of organizational policy. Call this institutional discrimination. It refers to the policies and practices of institutions -carried out by its representatives- which are intended to treat members of designated groups in some different manner. The policy may appear in memoranda, in other communications, or even in training in the organization. Hubert Williams and Patrick Murphy, two well-known police chiefs, remind us of the history of American police:

The fact that the legal order not only countenanced but sustained slavery, segregation, and discrimination for most of our nation's history and the fact that the police were bound to uphold that order - set a pattern for police behavior and attitudes toward minority communities that has persisted until the present day. That pattern included the idea that minorities have fewer civil rights, that the task of the police is to keep them under control, and that the police have little responsibility for protecting them from crime within their communities. ( Hubert Williams and Patrick Murphy, "The Evolving Strategy of the Police: A Minority View." in V. Kappeler, ed., The Police and Society. Illinois: Waveland Press, 1990.)

What is important to note here is that some departments have shifted away from institutionally discriminatory practices though many older officers are left who are still working though now under different rules. Further, the residues of those practices may remain without the full awareness of the command staff or even those on patrol.

Finally there is a normative form of discrimination which develops as part of the subculture. In every organization there is an informal set of norms, beliefs and values that give the organization its unique character. Like the culture of the larger society, it is transmitted across generations. It is internalized, and it is difficult to remain in good standing within the organization unless you conform to its basic norms.

Why bother with this apparently academic typology? The answer is that sorting these types is a critical matter of policy. Dealing with racist cops calls for a different intervention than dealing with the policies and practices of the institution. Likewise, programs aimed at changing the subculture call for a different strategy. You can isolate the racist cop, and you can change the basic policies, rules and regulations, but changing the norms means changing the subculture. And as Williams and Murphy point out, it has been around a long time.

Raising the Ante

Since September 11, 2001, John Ashcroft, the Attorney General of the United States has established a profile of people he wants his agents to interrogate. So far, through April 2002, he has identified about 8,000 people. The profile is this: male, between 18 and 33 years of age having entered the US after January 2000 on nonimmigrant visas and who lived in or visited countries where al Qaeda had a presence. Most of those summoned were from Middle Eastern and Central Asian countries. From the little information that has been released, none of these men have been charged with terrorist activities and fewer than one percent were arrested, mainly for immigration violations. Obviously, this is an instance of profiling using multiple criteria. Race/ethnicity is not explicitly cited, yet the stipulation of "lived or visited countries where--" almost invariably targets Arab countries. Is this an action of racial profiling?

In April 2002, Ashcroft issued an "absconders apprehension initiative." This identified about 6,000 Arab men for immediate deportation because they have overstayed their visas. Since there are an estimated 300,000 others known to be in the US on expired visas--and they have not been targeted--some immigration lawyers claim that this is another example of racial profiling by the Attorney General. Ashcroft and the FBI Director, Robert Mueller, both assert that they have not and will not target people solely on their ethnicity. The emphasis is, of course, on solely. This is yet another form of discrimination, structural discrimination. Structural discrimination refers to institutional policies and practices which have a discriminatory outcome regardless of the intent of the institution or its representatives. Our emphasis is on outcome.

It is possible to reject racial/ethnic profiling on city streets or interstate highways, but accept its practice at the airport. Stuart Taylor, a conservative writer on legal affairs puts the argument this way:

Stopping hijackings is different. First, preventing mass murder is infinitely more important than finding illegal drugs or guns. Second 100 percent of the people who have hijacked airliners for the purpose of mass-murdering Americans have been Arab men....This movement includes people who have lived legally in America for years--some of whom may be citizens--so the risk of weapons being smuggled onto airliners cannot be eliminated by giving special scrutiny only to foreign nationals.

In short, the mathematical probability that a randomly chosen Arab passenger might attempt a mass-murder-suicide-hijacking--while tiny--is considerably higher than the probability that a randomly chosen White, Black, Hispanic, or Asian might do the same. In constitutional law parlance, while racial profiling may be presumptively unconstitutional, that presumption is overcome in the case of airline passengers, because the government has a compelling interest in preventing mass-murder-suicide-hijackings and because close scrutiny of Arab-looking people is narrowly tailored to protect that interest. (The Atlantic Online, September 25, 2001)

Taylor's argument center about three propositions which I find questionable. I think that a drug dealer can do even more damage over time than the suicidal murdering hijacker. I suspect that the implicit statistics, a tiny probability as Taylor deems it, are not really meaningful. Finally, on his more serious constitutional presumption, I do not agree that the government has a compelling interest here that supersedes its interest in maintaining the integrity of the fourth amendment. What Taylor could have said was that given the threat of hijacking and our commitment to reasonable search and seizure that we need to increase airport and airplane security for as long as serious threats remain.

Reducing Racial Profiling

We need to think of ourselves as repair persons called in to patch a system which was never quite adequate to the tasks for which it was designed. We know, moreover, that the system itself is damaged, so we will have to innovate any patches. While we can hope ultimately to replace the system, we are trapped in the moment as tinkerers --the more pretentious may prefer the label of "reformers."

  1. First, an obvious starting point is to require police departments to have a clear, written policy statement against race/ethnic profiling.
  2. Such a policy requires direct sanctions for those in violation. Officers engaged in race/ethnic profiling could be subject to fines, required to participate in human relations training, and maybe transferred to jobs where discriminatory acts are blocked. Sanctions could also be applied at the state and federal level through withholding funds from departments which have a high rate of offenses or complaints. In areas of litigation, some portion of settlement costs could be taken from the departmental budget.
  3. The actions of police officers have to be made more visible. "Visibility" is a central organizational mechanism for control. Video cameras mounted on patrol cars are a very direct way of establishing the visibility of the officer's behavior. It would be helpful for the department to maintain traffic stops statistics. Officers would be required to record the race/ethnicity of the person stopped and the grounds for stopping them.
  4. With decent statistics, better training, and a departmental policy statement, pretextual stops could be greatly reduced. (There may be instances where the behavioral profiling of a person or the intuition of a veteran police officer may lead to a stop on a pretext.)
  5. In general, we need to demand better training of police officers. The level of so-called human relations training in police academies is inadequate. There needs to be specific training on the nature of prejudice and discrimination, and cops need to be led to confront their own attitudes. In multicultural city neighborhoods, officers assigned should receive appropriate diversity training and some minimum of language training. Most professional organizations require some form of continuing education credits for its members, aside from any in-house education. CEUs in designated areas might well improve the quality of policing.
  6. Eventually the nature of police organization must be changed. Much police activity is in the nature of public service although the organization itself is fashioned on a military model. Its militarism and its bureaucratic character need to be revamped. Think of the police along the lines of the agricultural extension agency, providing customer service and education.
  7. Finally, we need to improve accountability. One way is to establish a hotline by which people can report their experience of a discriminatory stop. The ACLU maintains one such hotline (1- 877-6-PROFILE), but local, well-publicized hotlines can be very effective. Another technique is a straightforward community survey in which a comprehensive and representative sample of residents could report on their experiences with the police over a specified time period. The initial survey would provide a baseline and progress could be charted over repeated, annual or semiannual, polls. Such polls could be handled by volunteers with assistance from local university sociologists and by the human relations commission in the area.
  8. At a level beyond polling would be a "Citizens Grand Jury." Based on the idea of combining a town meeting and a grand jury, the Citizen's Grand Jury is a way of enhancing democratic participation in policy decisions. It has been use in a variety of contexts.

Basically, a randomly selected jury pool balanced on age, gender, education, and race/ethnicity would hear testimony (on all sides) concerning the operation of the police department. Following hearings based on the identification of the central issues, the jury would deliberate in order to produce an evaluative report and policy recommendations if appropriate.

The proposals I sketched here are patches. They can reduce the incidence of racial profiling; they will not end it. Ending means changing the structure of police organizations. Ending means dismantling the structures of power and inequality that are sustained by and sustain discrimination and its sequel in ethnoviolence.


This site made manifest by dadaIMC software