del Pozo, Brandon --- "Guided by Race: An Ethical and Policy Analysis of Racial Profiling in Law Enforcement Decisionmaking" [2001] QUTLawJJl 19; (2001) 1(2) Queensland University of Technology Law and Justice Journal 266

• Introduction
• Exploring a contradiction in stance: affirmative action
• The nature of crime investigation
• A sufficiency test for police policy
• Profiling as a police practice
• Racial Profiling
• The empirical basis of racial profiling
• Racial profiling’s red herrings
• The legal climate
• The ethical debate
• Policy recommendations
• Conclusions
• Appendix

Guided by Race: An Ethical and Policy Analysis of Racial Profiling in Law Enforcement Decisionmaking

Brandon del Pozo[*]

Introduction

State Trooper David Riggins[1] has been watching interstate traffic from his concealed position for twenty minutes. Two thirds of the cars he tracks with his radar are breaking the speed limit, but be hasn’t pulled them over. Instead, he waits for the right “type” of car to drive by. In the next minute he sees one: a late model SUV being driven by a young black man. Indeed, the SUV is speeding. He takes off after it, hoping to get the driver to consent to a search and to find contraband.

Trooper Riggins doesn’t merely hunt for speeders, he uses speeding to hunt for drugs and guns. He uses observed violations to target a certain group of people: likely drug dealers.[2] For this reason, he doesn’t waste the time in his tour pulling over little old ladies with a lead foot, or white men in family sedans. He prefers the rental cars and the flashy, high-end imports driven by the black men who, in his experience, are more likely to be couriers of illegal narcotics and firearms. He has built up a 25% success rate[3] thus far, which he sees as a clear vindication of his means: fewer than one percent of the people on the interstate ferry contraband, but he has filtered out a sub-category of driver which is over 25 times as likely to be engaged in such crimes.

Some have asserted that he is intruding on the lives of fifteen good people for every five he arrests, and that his success rate is higher than most. He points out to them that every single person he pulls over is violating a traffic law, so he has the right to stop them in the first place. If things check out and he does not suspect drugs, he is courteous and often issues only a warning. When he finds a big cache he feels like he’s doing his part to help poor, crime-ridden inner cities for which the drugs were destined: he is making it incrementally harder for dealers to addict the citizens around them. He is helping to spread the word that being a drug courier is a risky proposition and hopes it gives people pause when they consider doing it.

Trooper Riggins doesn’t know precisely why a higher percent of black men run drugs than others, but supposes it has to do with poverty and lack of opportunity. All he knows is that if he was forced to stop black drivers only in rough proportion to their representation in the population of interstate travelers, he would take fewer drugs and guns off the street. The drugs would end up in the veins of addicts, and the guns in the hands of violent criminals. He thinks he is doing good work.

Few people would dispute the fact that the fictional trooper above, working a nondescript highway somewhere between New York and Florida, practices racial profiling in the course of his duties. What is not as clear, however, are his ethical, moral and legal rights to do so. In the last three years, the practice of incorporating race as a factor in determining who the police interact with has come under strenuous criticism, if most often from expected quarters. The American Civil Liberties Union released a report, ‘Driving While Black’, which condemned racial profiling as a pervasive practice which systematically denies minorities of their civil rights.[4] New Jersey’s Chief of Troopers, Carl Williams, was fired in 1999 by the state governor because he openly expressed his belief in a link between race and highway drug trafficking.[5]

His termination occurred in the context of the media’s attack on the practice of racial profiling: “the current uproar... has spotlighted one clearly abusive practice that moderates, conservatives, and, indeed, police chiefs should join liberals in assailing: racial profiling”.[6] To a great extent, they have. A writer at The Nation suggested that

it’s no mean feat to find an issue on which Bill Clinton, the Rev. Al Sharpton, Attorney General Janet Reno, many of the nation’s police chiefs and NAACP president Kweisi Mfume agree, but at a Washington conference in June, they all expressed that racial profiling... needs to end.[7]

It is possible that the nation’s police chiefs are afraid not to agree for fear of meeting the same fate as Chief Williams.[8] Racism—broadly defined as considering the race of a person as a factor in determining their rights and treatment[9]—is certainly an immoral, unethical and largely illegal practice. It is easy to then conclude that any practice which includes race in its decision-making process must fall under the umbrella of racism and is similarly unjustified.

Former President William J Clinton relied on such an argument to ground a Presidential directive ordering an investigation into the extent to which federal agencies practice racial profiling. He called the practice “morally indefensible” and “in fact the opposite of good police work where actions are based on hard facts, not stereotypes”.[10] The president made this statement despite extremely limited personal experience in police work, and also despite an extensive body of Supreme Court case law to the contrary. Cases from Terry v Ohio[11] through more recent ones such as Whren v U.S..[12] recognize that certain forms of police work which rely on grounded suspicion in the absence of hard facts are necessary, vital to effective policing, and constitutionally protected.

This suggests that such brisk conclusions are unwarranted and smack of folly. Instead, the issue ought to be discussed more carefully and with a good-faith effort to understand why it has often been put into practice by people who strenuously maintain that they are not racists and are in fact acting in the best interests of justice, among them some minority police officers.[13] There are some who have taken this charge seriously. Even while eventually concluding that racial profiling is not a sound practice, Harvard Law School professor Randall Kennedy acknowledges that to some police officers, “racial profiling is a sensible, statistically-based tool that enables [the police] to focus their energies efficiently for the purpose of providing protection against crime to law-abiding folk.”

Exploring a contradiction in stance: affirmative action

The case that all race-informed decision-making models should not be considered immoral and racist without careful consideration is exemplified by the widespread support—indeed by many of the same parties who condemn racial profiling—of the practice of affirmative action in the workplace and at universities. Off the cuff, some of its defenders would suggest that since affirmative action renders a benefit to minorities and is supposed to be rectificatory in nature, it merits certain dispensations. Meanwhile, they would suggest, racial profiling does not confer a benefit to minorities and exacerbates the racial tensions that affirmative action itself supposedly fights, so it is not akin to affirmative action and in a larger sense unjustified in its practice.

The most evident hazard in such a consequentialist argument is the presumption that race can be a factor in the treatment of people merely if the results are beneficial to the race in question. By extension this would reduce arguments about profiling into statistical ones bent on demonstrating its relative public safety benefits for different races, especially minority ones. While not relying primarily on such an argument, this paper will suggest later that racial profiling, despite any negative implications, has a very compelling statistical—and therefore consequentialist—foundation.

Still, defenders of affirmative action continue to rely on this form of reasoning. Some loosely maintain that using skin-color as a stand-in for a lack of traditional opportunity, they are able to redistribute the benefits of employment and education among those who have systematically lacked the opportunity—through no fault of their own—to receive those benefits. However, racial profiling uses also skin-color; not as a stand-in for lack of opportunity but for one of its results: a propensity to commit crime. It then redistributes police efforts along those lines to enhance the benefits of public safety among the larger class of law-abiding people of all races.

In both cases, the fit is rough. Affirmative action asks its recipients to acknowledge that despite their best efforts they are second-class citizens who are not in a position to succeed as easily as others. Racial profiling asks its subjects to acknowledge that they are part of a class of citizens who are more likely to commit certain crimes. Like most generalizations, both models work in the aggregate but do not offer insight at the individual level. In any event, individual citizens have both a pre-existing right not to be stopped by the police for inadequate reasons as well as a pre-existing right to be judged as independent, fully-enfranchised and capable members of society in a competition of merit with others. In both cases, the reasons for acting otherwise should be compelling, consequentialist or otherwise.

The larger issue here is that because they are born of such similar logic, racial profiling cannot be rejected out-of-hand by individuals who laud affirmative action. A December 18, 2000 New Yorker article by Nicholas Lemann serves to frame the point. At present, the University of Michigan is the subject of two lawsuits brought forth by white students who were rejected by the institution. Their claim is that they were rejected to make room for students with demonstrably poorer academic credentials who were accepted solely because of their race.

The plaintiffs were able to discover documents which provided formulas for acceptance and rejection based on race. One plainly stated that whites would be rejected from a selective pre-medical program if their SAT’s were below 1320, but minorities would not face categorical rejection until their scores fell below 1170. Another formula for undergraduate admissions provided for the automatic acceptance of any minority with a GPA of 3.5 or better and an SAT score above 1200. At the law school, the acceptance rate for blacks with undergraduate GPA’s between 3.25 and 3.49 with LSAT’s between 156 and 158 was 100%; for whites in the same range the acceptance rate was less than two percent.[14]

It is clear then that at least at this university, race plays a prominent role in the decision-making process of admission and the concurrent distribution of the benefit of college acceptance. Ceteris paribus, minorities seem to consistently beat out whites for admission solely because of the colour of their skin, as proscribed by the systematic process used by the admissions staff.

When asked to explain this practice, the university’s president at the time, Lee C. Bollinger (now the president of Columbia University), stated that affirmative action “helps students to understand the full complexity of life—to make the emphatic leap... It’s exciting to be in an environment where people are different from you.” In sum, he contends that minorities are different from whites in some substantive way, and that this difference in practice can be represented in proxy by skin color. It is not clear that a black from Beverly Hills is more different from an affluent white than a poor white from the isolated foothills of western Kentucky. Yet the university, and many others like it, are content to broad categorizations of skin color to be a stand-in for cultural difference in the essential practice of admitting students to their schools.

It will be argued that police officers who use racial profiling actually render much more subtle and complex analyses in their decisionmaking than the law school of the University of Michigan does in its admissions process. Yet few consider if a black person feels belittled for being invited to attend a university because its administration needs to expose whites to cultural differences and to prompt “emphatic leaps,” and not because she was among the best-qualified to engage in a course of study. This lack of analysis is likely because unlike police racial profiling, affirmative action yields the benefit of an academic degree instead of the temporary seizure of a person and the threat of incarceration.

It is not likely that the people who seek to stamp out the race-based decision-making of racial profiling, if successful, will in stride turn their energies to ending the race-based decision-making of affirmative action as the logical extension of their work. This may be due to many factors. It might be because of the belief that while racial generalizations about cultural diversity have firm empirical bases, there are no such bases viz criminality. It might be because they would choose to ignore such processes if they are rendering benefits for minorities (which implies that racial profiling does not). Or it might be because of an inconsistency in their reasoning which stems from political considerations. Regardless of the rationale, the point remains that it is exceedingly difficult to rest the condemnation of racial profiling solely on the fact that it employs race in its calculations and is therefore akin to racism itself. Not only do other, more revered processes do the same, but it is simply not the case the definition of racism casts so wide a net as to include, ipso facto, any act informed by race at its inception.

The nature of crime investigation

There is a segment of police work that involves police-community relations and devising creative approaches to managing low-level disorder such as rowdy youths and public consumption of alcohol. Another includes civil mediation such as resolving interpersonal disputes or documenting the facts surrounding automobile accidents. These facets of policing will not be addressed here. Instead, the topic of discussion requires a closer look at police work as it is focused against crime, more specifically the perpetration of felonies, misdemeanours and to a more limited extent, traffic violations.

In this regard, there are three methods most generally at the disposal of a patrol force. The first tactic is a response to crimes in progress or in the past at the behest of victims or other complainants. Sometimes police officers stumble upon these crimes before they are alerted to them by others, but in all cases the response is basically the same: to ascertain who is committing or has just committed a crime and to apprehend that person. In cases where the police arrive at the scene after the perpetrators have fled, they will most often engage in a search for the person. The search is directed based on information about the criminal’s physical characteristics and direction and method of flight as supplied by witnesses. The “canvass” consists of no more than stopping the group of people who match the given description[15] to see if any in that group can be identified as the person in question, or if they sport evidence of the offence.

Often, industrious police officers will seek to place themselves in situations where the chances of stumbling upon a crime in progress are maximized or the response time to the scene of a complainant-originated report is minimized. These measures might include heightened alertness at certain times of day and spending more time in neighbourhoods with higher complainant-reported crime rates. In any case, the issue of arrest is still predicated upon a complainant’s assertion that a person has committed a course of action against them which qualifies as a crime; whether the police notice this fact first or are alerted to it, the route of their investigation is determined by the statements of victims and witnesses.

The second tactic is the proactive apprehension of people committing crimes for whom no civilian complainant is readily apparent, and the apprehension of people who are about to commit a crime. In these cases, the police must make their determinations based on facts aside from stated allegations by others. These include cases where the state is the only complainant, such as all traffic infractions and crimes such as prostitution. They also include cases where the state is the only complainant but there is a strong presumption that the crime in question is a prelude to future complainant-driven crimes: unlicensed firearms possession is unlawful under the presumption that it is a prelude to armed robbery or felonious assault. A police radio scanner, an armoured vest, and lock-picking tools might be legal by themselves, but when possessed all at once at night near a commercial strip, they strongly indicate an impending burglary. Finally, some proactive apprehensions are of criminals for whom the suitable complainant is still to be located. This includes apprehending the driver of a stolen car before its owner has discovered it stolen, or arresting a burglar before the resident of the building in question gets home.

The third measure is deterrence by mere presence, in which the police prevent would-be criminals from acting because the spectre of capture and incarceration overshadows the benefits of committing a crime. This tactic is illustrated by the fact that it is extremely rare that sane people knowingly commit crimes in the presence of uniformed officers, and that highway patrolmen must usually conceal themselves to observe people speeding. Deterrence is akin to “winning without fighting” in the sense that criminals shy away from their intended actions because the probability of engagement by the police is so high, and the odds are so skewed in favour of the law.

As the tactics outlined move from reacting to the statements of crime victims through measures which are predictive in nature such as proactive enforcement and deterrence, the police’s reliance on conjecture (“that car might be stolen; those guys look like they’re scoping out a store for a robbery”) versus hard, situational facts (“I am looking for three men in red jackets who fled northbound in a black Honda two minutes ago”) necessarily increases. Despite less reliance on supplied facts, citizens depend on the police to take these proactive steps. Lacking such initiative, they would make no arrests in state-complainant cases where perpetrators take reasonable steps to hide their illegal actions (such as concealing illegal weapons). The police would find themselves only encountering criminals either through dumb-luck chance encounters or after the state-driven crime has yielded a complainant-driven one, and they would begin their investigation only after an innocent person has been harmed and the trail of the perpetrator has started to cool.

It can be presumed that good police practices are ones which expose as few citizens as possible to criminal acts resulting in bodily harm and the loss of property while respecting human rights and freedoms. What follows is that the ideal state of law enforcement is one in which most crimes are deterred, and those which are not are intercepted before an innocent person has been harmed by a criminal. But since the police cannot be everywhere at once, where should they be to maximize this deterrence? In what enclaves should they loiter to minimize response time to crimes? If certain groups are much more likely to commit crime, should they preemptively keep a closer watch on them? This would suggest—contrary to President Clinton’s assertions—that good police work is based minimally on hard situational facts, but instead more often on presumptions, predictions, intuition, and inference which officers then parlay into levels of suspicion, action, and only towards the end, hard facts and evidence.

A sufficiency test for police policy

As suggested, police policy must respect people’s rights and freedoms.[16] It would be very effective to protect all citizens from crime by implementing widespread random searches, or allowing for arrest and interrogation on less than probable cause. Yet those policies infringe on the rights of innocent citizens, and the loss of trust and attendant fear of the police among the innocent would greatly outweigh the benefits of reduced crime[17] if not be catastrophic for democratic government.

Beyond this, not every end in policing is a clear moral imperative. For example, documenting motor vehicle accidents and verifying pawn broker licenses are both regular functions of the police officer, but they are not morally required goals of policework. At the same time, preventing citizens from being robbed, raped or killed are the clear moral goals of an organization acting as the government’s sole or primary agent in this regard. In this way, many goals of police policies are themselves moral ends.

This yields one of the three facets of a test for otherwise effective police policies. The first task would be to ensure that the policy is a moral end, and if it is not, that it does not interfere with what are ostensibly moral ends. The implication is that, for example, policies which address crimes against persons would take precedence over those which address property, and still over those which concern civil or administrative functions of the state. Thus, police agencies ought to carefully consider the moral ends which fall under the scope of their organization when constructing their plans and policies.

The second task is to determine if the practice being formulated, although producing a desired effect, is still a moral one in that it respects the rights and freedoms mentioned above. Even if the goal of the policy is moral, such as to prevent rape, a program of warrantless searches and torture would not be a morally permissible method of preventing such a crime. Involved in this formulation is the calculus of civil rights, occasions when such rights may be curtailed, and issues of exceptions and due process.

The third facet develops from the fact that while protecting citizens from crime might be a generally moral end of the police, it is not a procedure in and of itself. To this end, the police must still enact various specific crime reduction policies. Some will be more effective than others, but as long as they are good-faith efforts to achieve these moral ends, none of them have more of an intrinsic moral value than other variations of the policy. Unless, that is, the policies are formulated in a manner that causes the agency to neglect certain other moral duties. For example, an agency might decide that since thousands of people shoplift each year, it is one of the most frequent and troublesome crimes in their jurisdiction. The department might commit most of its resources to arresting these violators, even as less frequent violent crimes such as rape and murder occur and receive almost no attention. This is type of negligence is serious and serves to give the policy an added negative moral dimension. Another type of negligence of a similar nature is financial. If a police department hell-bent on catching murderers spends all if its budget doing so while becoming too cash-poor to provide the remainder of its service obligations, then it is negligent. The same idea applies to a government that lavishly funds police efforts while ignoring other public services necessary for a safe and healthy community.

As long as a department’s goals are moral and properly ordered, and barring these types of negligence, policies may be enacted, repealed or changed to meet the sensibilities and expectations of the citizens they serve. One policy might concentrate the police in high-crime areas, but the next week scatter them about. Another might combat domestic violence as a means to lowering the murder rate, the next one might fight street crime to the same end. Policies are political tools enacted by a government to fulfill its goals. An implication of this adaptability is that if an otherwise good policy is divergent from the expectations of the citizenry, causing dismay and disapproval, then the policy may be changed or discarded so long as the effect of doing so does not border on negligence. Less charitably, if the policy does not fulfill the goals of the government—no matter whether the satisfaction of citizens is a calculation in them or not—then it may be readily changed or modified as the government sees fit. It is simply reality that police policy is written under the umbrella of larger political considerations. While law enforcement agents would ostensibly like to incorporate the measures that most effectively reduce the amount of crime in a city and the number of offenders walking its streets, the overseeing politicians must take into account other considerations. These include matters of budget, but they also include public sentiment. If the public prefers certain types of enforcement over others, and the end result is less-efficient crime fighting, greater public approval, and a net difference in result that is not immoral in its negligence, then one should not be surprised if the police department in question receives political guidance to that effect.

It will be presumed that the ends of policing discussed in context with racial profiling are moral ends. Indeed, they typically are: to arrest those dealing dangerous drugs, to avert robberies and murders, to apprehend those carrying illegal weapons intended for use in crimes, etc. The merit of the argument that racial profiling is actually meant to support a racist government’s oppression is not considered here but is addressed later as a red herring. This allows for the omission of the first test of police policies in the assumption that their ends will be moral ones that are properly ordered. To succinctly restate the remaining two elements of the test, it is sufficient for police policies to meet two criteria:

1)	The policy must try to meet its stated goal without violating the moral rights and freedoms of citizens in its practice.
2)	The policy must not in and of itself take on an added negative moral significance by neglecting certain other moral duties.

If these criteria have been met, then the policy may be constructed, reconstructed or adapted inside these parameters for the widest range of reasons, including those of political and economic expedience.

Profiling as a police practice

Having constructed a sufficiency test for the soundness of a given policy, we will set it aside for a time and in the interim consider the nature of profiling, and then of racial profiling. Profiling can be defined as a broad method of targeting police resources based on where they are most likely to encounter crime.[18]

Profiling can be executed geographically and against people.[19] Certain areas have much higher crimes rates, and through the use of statistical analysis, it is possible to determine to within a few city blocks where crimes are occurring with the most frequency. For example, a feud between youth gangs may yield a higher frequency of felony assaults at the border of their perceived turfs. A certain senior-citizen complex might suffer more mailbox break-ins as social security checks arrive each month. Thieves might evince a predilection for a certain industrial neighborhood for dumping and stripping stolen cars. In all three of these cases, the police would be inefficient if not foolhardy not to concentrate their resources at certain places and times to combat such prevailing trends. What such crime-mapping does is speculate about the future based on past facts. It speculates that certain past geographic trends will continue and that directed efforts informed by these trends will yield more arrests than patrolling which ignores these facts. The public does not seem to mind these assumptions and in fact often calls for an increased police presence in certain areas in the wake of crime trends.

This speculation can extend to people. If the youth gang which complainants say is responsible for numerous recent shootings sports purple bandannas, the police will keep a closer watch on youths with purple bandannas. If packs of youths frequently descend upon high-end department stores in midtown Manhattan to commit larcenies en-masse, then the police would do to discreetly tail such throngs.

What profiling generally does, then, is collect categorical data for use in speculation about the future. Often the data can be contradictory, as with the profiles presented by the US Drug Enforcement Agency in the fight against drug smuggling. They include “acting too calm... acting too nervous... traveling alone... traveling with a companion...”.[20] This suggests that some instances of profiling are more successful than others and that some nets are cast more widely than others, but in each case the profile is presumably drawn with the aim of singling out sub-groups more likely to be engaged in crime than a random person from the population at large. The sub groups may be very large ones such as the DEA list above, but as long as they are smaller than the relevant population, they will increase the efficiency of enforcement efforts.[21]

This can be rephrased in more precise language: a profile identifies categorical data which correlate with criminal activity. For these correlations to be of use, they must not be due to chance. If they are not, then the police have constructed an “instrumentally rational” enforcement tool.[22]

It might be possible that the correlation is spurious, but that does not detract from a profile’s usefulness, and therefore instrumental rationality. Applbaum explains that when “[such] Bayesian decision analysis expresses a degree of certainty about some event occurring, it makes no statistical commitment to any one of the many underlying causal mechanisms compatible with the statistical inference made” (emphasis added). If the youth gang discussed sports the purple bandanna not as a conscious display of their colors but because they need a bandage handy in case of injury and purple bandannas happen to be the most cost-efficient solution at present, does this diminish an officer’s suspicion of people with these cloths? No, but it does suggest that profiles might be periodically reevaluated to test their validity and timeliness. Perhaps there will be a sale on orange bandannas the next week.

With this in mind, certain profiles have already been written and have stood the test of time. Men are more likely to commit violent crime than women. Men from the ages of 14 to 30 are more likely to do so than children, the middle-aged and the elderly; single men commit more crime than married men.[23] A survey of precinct-by-precinct crime statistics in New York City suggests people in poor neighbourhoods are subject to more violent crime than those in exclusive, wealthy ones.[24] More violent crimes are committed between the hours of dusk and two in the morning than at all other times. The population which has committed violent crimes in the past is more likely to commit them again than the population at large.

This yields numerous categorical variables suitable for profiling. They include sex, manner of dress, age group, criminal history, education, marital status, level of education, location, and time of day. Some of these factors are readily apparent, others only known of the usual suspects, and others only known under bizarre circumstances. These are then complemented with cues such as mannerisms, traditional indicia of criminal activity (such as “casing a joint”) garnered through the practice of “social decoding”.[25] The result is a set of statistical stereotypes for use in streamlining proactive policework. As with all stereotypes, their use is limited and exceptions will abound, but there is enough of a basis in truth to warrant their use.

Moreover, when any number of disparate statistical probabilities which describe the categorical attributes of the same dependent variable (in this case crime) come into confluence at a certain point in time, the attendant likelihood of a positive hit increases accordingly. A sixteen-year-old male with a known arrest record for assault bearing a gang bandanna and walking the street in a poor, violence-prone neighbourhood at midnight bears more watching than the elderly female who crosses his path, or the middle-aged man who walks several paces behind him: several factors correlated to crime have come together at that moment which are together much more suggestive than any smaller group of them standing alone. This suggests that “among the set of search strategies with positive net benefits, some are better than others... If a refined search strategy is available, not to use it is inefficient”.[26]

Now it is entirely possible that the man behind him is a mugger who is actually using the unwitting youth to run interference. Even then, all is not lost: if he strikes, the complainant will disseminate his description and the canvass will focus on people who look like him. If enough people like him commit muggings viz youths, the profile will shift.

As an aside, it is also a practice to profile victims. It is instrumentally rational for plainclothes officers to trail a drunk, disoriented man stumbling home at 3am sporting an expensive watch and gold chain. They might also do well to follow an old woman home from the bank at dusk as she dangles her purse distractedly. In each case, the officers have profiled the person involved as more likely to be a crime victim than a member of the population at large. These presumptions have been formulated and disseminated by police officers who have informally collected data from numerous calls for assistance from drunk people who have been robbed at night and older women who have had their purses snatched while walking home from the bank.

The instances outlined thus far involve measures such as positioning officers in certain places as certain times, or keeping certain people under more careful observation than the average citizen. As presented, the measures are only mildly controversial except in that if they are not carefully formulated they might not be of more use than patrolling in a random manner. However, as will be discussed later, these measures might also include using the profile to formulate a level of suspicion of a person which a police officer then uses to stop and possibly frisk them. In this case, the police officer is conducting a proactive investigation with the hope of finding the evidence of a crime.[27] Profiling might also be used to pick out who to apprehend from a known set of violators (such as all those who are speeding down a highway) when it is impossible to apprehend everyone.[28] The expectation is that profiling what type of speeder is also more likely than average to be a criminal will not only deter speeding but also parlay itself into the apprehension of such criminals,[29] to include the drug dealers mentioned in the introduction.

The emergent point about profiling is that it is a practice of data collection and analysis with an eye toward two things. This first is providing the criteria for separating a certain sub-population from a larger one which has already committed a minor violation and are legally eligible to be stopped when various factors prevent stopping them all. The second is to provide the focus for police enforcement efforts in the formulation of a level of suspicion legally adequate for police action. In these cases the conduct of the police seems more controversial than before, though not because of profiling per se but instead because of the level of intrusiveness it is used to justify.

Racial Profiling

The use of race in profiling has been conspicuously absent up to this point, but not by accident. In an attempt to establish the rationale for profiling in general, an early introduction of race would only serve to prematurely confuse the issue by drawing too much of the focus away from necessary theory. The groundwork having been laid, racial profiling merely expands the categorical data collected for use in speculation about the future to include data concerning race.

In addition to supposing that certain types of dress, attitude, location, and other factors such as gender and age have a non-chance relationship with crime, the basis of racial profiling is that race itself has the same type of non-chance relationship. Thus, for racial profiling to be an instrumentally rational tool, it must have an empirical basis. Once this is established, it must also meet the requirements proposed earlier for a sound policy. It must not be inherently immoral or unjust. Secondly, if its application as policy does not assume a moral weight due to negligence of certain moral duties, then it should also be a policy which meets with the desires, expectations and sensibilities of the public it serves.

If racial profiling were to somehow be proven as empirically true, morally just and a politically sound policy from the outset, then its usefulness could not be underestimated. Race is an incontrovertible piece of data not subject to easy change or manipulation. Age is hard to estimate, the fashion trends of street criminals change, and it takes a very keen eye to be adept at social decoding.[30] A person’s race is most often clear and hard to disguise. As long as race’s relationship to crime could be maintained, it would be a readily available for use in calculations.

The empirical basis of racial profiling

There is strong and compelling evidence to support a relationship between race and crime that is not due to chance. As Randall Kennedy maintains, “Statistics abundantly confirm that African Americans—and particularly young black men—commit a dramatically disproportionate share of street crime in the United States”. This assertion is based on several sources, but the ultimate foundation can be found in statistics compiled by the Department of Justice. These include the National Crime Victimization Survey (NCVS) and the FBI’s annual Uniform Crime Reports (UCR).

Certain authors maintain that the use of arrest data is unreliable because racist officers will look to arrest fewer whites and more minorities, then point to arrest data as proof that minorities commit more crimes than whites and use this as a reason to go on patrol and target more minorities.[31] While this criticism is most tenable in the discussion of drug offences and other crimes where the state is the complainant, it stands on shaky ground in crimes with victims. In such cases, the police will make an arrest based on descriptions provided by the victim and ultimately on her identification of a suspect. To say that the police seek to arrest minorities when victims have instead described their perpetrators as white is not only counter-intuitive but involves widespread collusion by crime victims, a considerable number of whom are minorities themselves.[32]

It is unfortunate that many of the arguments in support of the empirical link between race and crime published in peer-reviewed journals have been authored by the white separatist Jared Taylor.[33] Still, he uses data based on government-supplied statistics and subjects them to straightforward analysis to present relative rates of offence. While it is easy to find his conclusions and policy recommendations fatally flawed and extreme, it is useful to consider the data he presents.

Based on NCVS data, individual blacks are 50 times more likely to commit crimes against whites than vice-versa; groups of blacks are up to 250 times as likely to do so.[34] In fact, NCVS data suggests that blacks are responsible for 90% of all violent interracial crime.[35] What this implies is that in racially-mixed situations, blacks account for the vast majority of violent, interracial crime. A further implication is that if the police are patrolling such areas populated by a mix of white and black citizens, the sub-group of blacks among them contains significantly more criminals.

Taylor also notes that the data are skewed by the misleading characterization of Hispanics as white in federal crime reports. If a person of Mexican or Colombian descent, for example, is arrested for a robbery, it is recorded as a white person committing the crime. Hispanics may very well be closer to white than black in their racial makeup. Still, categorizing them as white is the bizarre exception to a general rule of considering them as a separate racial/ethnic group for the purposes of government data collection, and perhaps more importantly in the eyes of racists who discriminate against them. In fact, the only time Hispanics get their own category in federal crime statistics is not when they perpetrate a crime but instead when they are the victim of a hate crime.[36] The net effect is representing the population of whites as being larger than it actually is when comparing rates of offense based on race.

The main flaw with Taylor and Whitney’s methodology is its focus on interracial crime. It thereby neglects the analysis of crimes committed intraracially, which are just as important to solve, deal with, and account for as interracial ones. The rate at which blacks commit crimes among themselves can yield useful data about relative offence rates and the degree to which chance might explain racial disparities.

To this end, this author conducted three Chi-Square tests for fit. The tests used the categorical variables of race and type of offence viz the total number of arrests of blacks and whites for those offences in 1997. The first crime chosen was murder, for several reasons. Murder is the crime least likely to be downplayed by the police. There are great efforts made to close as many murder cases as possible regardless of the race of victims or offenders. Also, it is as extremely difficult to doctor murder statistics as it is for the police to remove a dead body from a street or hallway and act as if nothing happened despite the inquiries of aggrieved family and friends.[37]

The second offence chosen was the aggregate category of arrests for violent felony crimes. These include robbery, aggravated assault, rape, and of course murder. This choice was made under the presumption that such crimes provide a more compelling case for proactive police measures such as profiling than property crimes, and because, in the case of all violent crimes, a witness or complainant must describe and identify a suspect prior to arrest, minimizing the potential for police bias.

The third offence chosen was arrest for illegal weapons possession, in light of the criticism that the NYPD has encountered for disproportionately targeting minorities for frisks in an effort to find illegal firearms. One criticism of this statistic, as mentioned earlier, is that one can find things only where one looks for them: “If blacks are stopped at rates that are shockingly disproportionate to any other group in the population... then it should not be a surprise that they are subsequently arrested, prosecuted and convicted more frequently than whites”.[38] Still, a certain amount of these arrests are incidental to the investigation of other crimes such as disputes and assaults, and enforcement efforts searching for weapons are often driven by a neighborhood’s reported rate of violent crime.

It was assumed that the black population in the United States is approximately 12 percent.[39] It was also assumed that whites make up 85% of the population, which is somewhat of an overestimate, but since Hispanic offenders cannot be separated from this group without guesswork, it was estimated that they offend at a rate equal to or less than whites for a margin of safety.

The data considered were taken from the 1997 FBI UCR,[40] which is the most recent year for which the FBI posts detailed race and crime data on the internet. It is summarized in Table 1, below (The complete table is published in Appendix A).

	White Qty.	Black Qty.	Other Qty.	Total Qty.	White %	Black %	Other %
Murder	5,345	7,194	220	12,759	41.9	56.4	1.7
Violent Crime	284,523	205,823	10,275	500,621	56.8	41.1	2.0
Weapons	89,305	60,322	2,614	152,251	58.7	39.6	1.7

Table 1: 1997 UCR extract by race for selected offences

If there were no differences in the arrest rates of people of different races, then it would be expected that blacks would account for 12% of the total volume of crime in each category and whites would account for 85%. It is obvious that in 1997 this was not the case. Still, one may construct a null hypothesis that these deviations are not socially significant but in fact due only to chance. The Chi-Square test for fit determines the probability that a given deviation from the expected, equally-distributed outcome is due to such a random chance. These distributions are illustrated below in Table 2, below.

	% Pop	Murder Expctd.	Murder Obsvd.	% Obsvd.	Violent Expctd.	Violent Obsvd.	% Obsvd.	Weapns Expctd.	Weapns Obsvd.	% Obsvd.
White	85	10,845	5,345	41.9	425,528	284.523	41.1	129,413	89,305	58.7
Black	12	1,531	7,194	56.4	60,075	205.823	56.8	18,270	60,322	39.6
Other	3	383	220	1.7	15,018	10,275	2.0	4,568	2,614	1.7
Total	100	12,759	12,759	100.0	500,621	500,621	≈100	152,251	152,251	100.0

Table 2: Differences in observed and expected frequencies by race for selected offences, 1997.

In all three cases, the most rigorous application of Chi-Square analysis at df=2 suggested that there was extremely less than a .01 probability that the offence rates between blacks, whites and all others differed by chance (Table 3, below). This one percent probability is the most stringent measure imposed by social scientists when examining the validity of findings,[41] and the Chi-Square scores surpass the cutoff for this level to such a large degree that it suggests a .0001 significance level. The null hypothesis can most certainly be rejected. We cannot say that the disproportionate distribution of arrest rates between races is only due to chance.

	Chi-Square Score	df	.01 sig cutoff	Sig. level
Murder	23,805	2	9.211	.01+
Violent crime	401,820	2	9.211	.01+
Weapons poss.	110,035	2	9.211	.01+

Table 3: Chi-Square test for independence; results and significance.

It is important to note here, as it has been noted earlier, that this does not indicate that race and crime are related because people are biologically condemned by their race to commit crimes. It could well be that this link between race and crime is spurious and actually reflects a link between crime, lack of education, and poverty,[42] but that “the relationship of crime to poverty is obscured,” and that “white communities tend to be much wealthier than communities of color”.[43] Still, the functional difference is nil in that a police officer may still come away with an instrumentally rational tool for use in focusing her efforts. For the tool to be useful, race need only be a more reliable proxy for these other factors than other categories that are just as easily discernable.

Having presented a reasonable case for the empirical bases of racial profiling, it can still be argued that the average police officer does not go to any similar length to formally collect empirical data for her particular racial profiles. Police agencies will not do so in the present political climate, and this decision is a prudent one. The data collection for crime rates in various areas and at various times commonly used to target enforcement usually does not extend to race.

Instead, police officers often rely on two other sources of information: the informal training which more experienced peers impart on them, and their own personal experiences over time. Some protest that these perceptions are by their nature merely based on the type of anecdotal evidence that constitutes stereotyping. However, they are anecdotal only in the sense that any data set is a composition of individual instances, and with a large enough data set this becomes a trivial consideration. If a police officer continues to work where she has traditionally worked or even possibly grew up, then her own past experiences responding to crimes and dealing with local criminals will provide her with a nuanced, adaptable set of guidelines as to what are the most likely attributes of an offender. The more time she has on patrol, the better she will get, as “there is likely to be an enormous difference in the degree of discrimination displayed by a rookie from the suburbs and a veteran officer who has spent years patrolling the neighborhood in which [s]he grew up”.[44]

It is possible that these impressions are informed by racism, but this only means that some officers wrongly allow racism to color their impressions of what makes a person more likely to be a criminal. The driving force behind profiling in general ought to be the successful apprehension of criminals, and a racism-driven profile will not work towards this end for long unless it only incidentally is concurrent with empirical data. In any case it is immoral, but not to be confused with the mere use of skin color as categorical data.

What this shows is that racial profiling has empirical foundations in two ways. The first derives from analysis of the available data provided by government agencies. The second suggests that even absent formal techniques of data analysis, officers will use their experience in responding to crimes and working in a particular neighborhood to draw their own conclusions about the use of race as a factor of potential criminality. Such data, though assailable as anecdotal, is actually based on immersion in a community and its changing social customs. Regardless of the method employed, these findings suggest that racial profiling—especially when combined with other data—offers utility as a police policy.

Racial profiling’s red herrings

There are several arguments against racial profiling which do not stand up to ethical scrutiny, mainly because they do not attack the practice at its heart, but instead are red herrings. Nonetheless, many of them are rather commonly used. They are discussed below.

Racial profiling is wrong because the majority of the people of all races don’t commit crimes. David Cole has made this claim, and others are quick to quote him: “In any given year, Cole points out, 98 percent of blacks and 99.5 percent of whites are not arrested for anything”.[45] Another author, paraphrasing Cole, states that “the correlation of race and crime remains a stereotype, and most blacks will not conform to the stereotype. A police officer who relies on race in stopping and questioning individuals is therefore likely to stop many more innocent than guilty individuals”.[46] Cole himself writes “In fact, race is a particularly bad basis for suspicion, since most black people, like most whites, don’t commit any crimes”.[47]

What is immediately apparent from Cole’s own statistics is that 300% more blacks are arrested than whites. If this has anything to do with their likelihood to be involved in crime, then race is indeed a factor worthy of consideration. What fisherman would not want to fish from a pond where, although the vast majority of the fish she caught would need to be thrown back, she was four times as likely to encounter ones she could keep than when fishing elsewhere?

Some grudgingly make small, vague concessions to this rebuttal: “Even critics acknowledge that racial profiling is not entirely irrational in treating young black inner-city men as presumptively more worthy of attention than, say, grandmothers”.[48] This hits upon what gives racial profiling even more power, and that is its use in confluence with other types of profiling. It is highly unlikely that police officers can use any isolated piece of profile data with ongoing effectiveness. Yet in reluctantly pitting “black inner-city men” against “grandmothers,” the door is opened by critics for supporters of profiling to pit “black-inner city teenagers with gang bandannas in established high-crime locations at 2am on a school night” against “white grandmothers on their porches with their grandchildren in a low-crime neighborhood at noon.”

As already argued, in combining such data sets the likelihood of a positive hit rises. If the police are to maximize their efforts at keeping citizens safe, should manpower be directed to ensuring the safety of the night-shift worker who must commute home on payday by passing the street-corner hangout of the youth, or the cash-carrying delivery boy who rides his bike past grandma? Is this merely “not entirely irrational” to watch this black youth even though 98% of blacks are not arrested for anything each year, or is it eminently rational nonetheless?

The mission of profiling is to find out as much as possible about the nature and characteristics of that two percent of blacks and that half-percent of whites who are arrested yearly, and use it to protect the other 98+% of the population that is law-abiding. If a much greater percent of people were arrested than at present, there would hardly be a need for profiling of any type. The police could just throw themselves into the mix and frequently find criminals. It is precisely because criminals are such a small percentage of the population that the police must seize upon what factors might distinguish them and use those factors to protect the vast majority of citizens.

Police wrongly use racial profiling to justify their intensive, aggressive presence in minority neighbourhoods. First of all, for a police officer patrolling a geography consisting almost entirely of one race, as many NYPD precincts do, it is almost impossible to utilize racial profiling. If an officer who patrols Brownsville, Brooklyn—almost 100% black and Hispanic—coalesces various factors in her mind which will lead her to pay attention to certain citizens more than others, one of these factors cannot be that the person is black or Hispanic. It is not for her a question of choosing between blacks, Hispanics and whites to watch, but between different categories of blacks and Hispanics. In truth she must make subtle, non-race related distinctions between members of the population he patrols: who works for a living, who malingers on a street corner to look for potential victims; who she might be involved in a foot chase with versus who will call 911 when they see her in pursuit, or otherwise in need of help. She may use sex, age and dress in these distinctions, and it is counter to both her survival and productivity to be bad at making them. Indeed, she might use racial profiling when a white person drives up to a corner store known to deal drugs, goes in for a moment, and comes out with no groceries. If it is a very uncommon sight in her experience, is not her suspicion rightly aroused more so than when a black youth who goes in and comes out, possibly with candy in his pocket? Racial profiling does not mean by definition that only minorities are subject to such profiles; the method is ostensibly applicable to any race.[49] Some officials openly acknowledge that certain illegal drugs such as methamphetamines are sold predominately by white criminals.[50]

This leads to the broader question of why minority neighbourhoods are more often patrolled by greater numbers of police than white neighbourhoods. It could be of a racist government’s desire to oppress and harass minorities, but that is probably not their stated policy. Usually, their stated policy is instead that they wish to concentrate police resources where there is the most crime. One might remove the precinct numbers and neighborhood names from their attached crime statistics, jumble them up and present them to police managers who would be tasked with assigning them appropriate levels of manpower. These managers would likely argue that lacking particulars such as racial demographics or race-associated place-names such as “Brownsville” or “Harlem,” but instead only dealing with numerical data, they would make substantially the same staffing decisions. Every citizen deserves protection from crime, and those innocent citizens in high-crime areas (ostensibly at least 98% of the neighbourhood) deserve the according level of protection. To complain however that the police in these neighbourhoods are overly-aggressive and heavy-handed may or may not be true, but

Such rationales reflect the tendency... to confuse racial profiling with a different phenomenon: the policies of police in places like New York City to patrol (and stop, and search) most aggressively in high-crime neighborhoods. When done with respect and sensitivity, this can produce safer communities and better community relations.[51]

Sometimes it is not executed well, and that is a different story. In such cases, officers may need to tone down their actions. Nonetheless what brought so many of them there in the first place were not decisions based on race but instead the desire to keep citizens safe in areas where it is undeniable that they are more likely to be victimized.

Racial profiling is wrong because it gives racist police officers license or guise by which to harass and brutalize minorities. This argument incorrectly confuses racism with racial profiling, or assumes they are inseparable. Simply being a police officer effectively gives a racist a substantial opportunity to actively manifest his racism in negative actions against minorities. Suppose for a particular police officer race stops being a piece of categorical data for use in enforcement and starts being a trait with certain negative value-related connotations which a police officer uses to systematically accord some citizens (minorities) fewer rights than others independent of their criminality. This police officer has stopped treating the use of racial profiling as an instrumentally rational tool and is instead using race as an irrational basis for committing immoral acts. It is important that all police officers not be racists independent of the suitability of racial profiling as a police policy. That race with other factors might contribute to a level of suspicion which results in an innocent person being temporarily detained, causing inconvenience and some irritation, would not make the practice unethical per se, but instead a questionable policy.[52]

Police officers who are racists will continue to be racists, and to think that racial profiling alone gives them a license to act on their perceptions is naive. Some police officers become racists over time because they wrongly allow their persistent negative contacts with criminals of a certain race to overtake their initially benign perceptions. They will fall victim to such pitfalls as long as they are weak-willed and their work yields predominantly negative contacts with criminals who in certain neighbourhoods are mostly of one race. This is extremely common by policing’s nature. These problems underscore the need for carefully-selected and well-trained officers but not the need to deny such people a potentially useful tool.

Racial profiling alienates minorities from law enforcement because it subjects innocent people to repeated negative contacts with the police due only to their skin colour. There are many reasons why this argument is false, though it will be discussed later in another form as one of the most potent reasons for racial profiling being an ill-advised policy. Primarily, it confuses the issue of racial profiling being ethical versus being employed by rude and discourteous police officers.

It is possible to stop a person and give them a ticket, then have them shake your hand and thank you. The author has in fact been heartily thanked by people who he has sent to jail, and whose licenses have been revoked by his tickets. He has towed cars out from underneath drivers and been thanked; he has been thanked by people he has sprayed with mace. This suggests that police encounters involving criminals need not be ultimately negative, which strongly implies that encounters with innocent people need not be negative, but possibly even affirming, positive encounters.

The first step lies in having reasonable suspicion to stop a person. Race alone rarely provides this. It could be a combination of factors including race and should ideally involve an observed violation. It would be exceedingly difficult to justify intruding on a person’s freedom solely because of race, and the narrow case of exceptions will be discussed later. To stop a person only because you are itching to know what they are up to but cannot articulate a founded suspicion for doing so is unlawful and unethical. To stop people only because they are black and you want to know what they are up to is also wrong, not to mention poor, hamfisted policework. These things will irritate people and alienate them from the police, though not because the use of race was wrong, but because it was insufficient.

It is also wrong to use a reasonable suspicion founded in racial profiling and other factors such as an observed violation to execute a long, overly-intrusive search that goes far beyond the limited scope of a roadside or sidewalk investigation. Critics often confuse the use of racial profiling with a concurrent unethical disregard of the extent to which an officer can detain and search a person and his belongings. For example, the ACLU began its special report on racial profiling with the now oft-cited story of a black US Army sergeant who was traveling with his son when he was stopped in Oklahoma by state troopers. The man and his son were then detained for two-and-a-half hours while his car was comprehensively searched and he and his son were placed in the back of a cruiser.[53]

My intuition is that this seizure of his person and the search of his belongings was unlawful, and it makes an excellent case for disciplining the officers involved for Fourth Amendment violations. It is unclear, however, what this intrusive and likely unlawful search has to do with racial profiling. If the officers had found cocaine inside his car’s rear bumper, would defence attorneys rest their case on the on the fact that the defendant was black, or instead on the egregiousness of the prolonged and intrusive search itself? It is all speculation, because the paper does not cite what set of factors led to his stop in the first place. It is possible for a stop to include race as a factor, to be justified, and then to veer off into lawlessness only in its later stages. This should not be confused with it being unjustified from the outset, however.

Obeying the laws of search and seizure by combining race data with other elements to produce sufficient reasonable suspicion is only the first step of avoiding an alienating police-citizen interaction, because it merely justifies the stop by setting the stage for the legal temporary seizure of a person. The next ingredient is the extent to which the police officer is courteous and respectful to the person being stopped. Sometimes this is impossible. There are nasty, hostile citizens of all races in the population, and this sub-group makes hard for a police officer to leave a good impression on them.

Barring that category of person, an officer can briefly explain the circumstances for a stop, the proposed course of action, and pepper her words with “please” and “thank you.” They can explain that they were stopped for a traffic violation. Providing that there was sufficient reason for the stop in the first place, the grounds for a subject’s protests would become arguably slim and the effects of the use of race in decisionmaking would be mitigated.[54]

What critics contest is that even good officers tire of such courtesy over time and it devolves into rudeness and racism:

One can easily imagine... that racially discriminatory conduct which starts off as courteous will degenerate, that disciplined racial selectivity will give way to undisciplined racial selectivity, and that the emotions that always seem to coalesce around racial distinctions will lead to the sort of conflicts which have often vexed relations between police departments and black communities.[55]

Of course this is an argument against the prudence of racial profiling as a policy and not of its ethical nature, and it is based in the assumption that most police officers are incapable of not letting their work turn them into racists. As asserted earlier, if this is true it will happen regardless of whether racial profiling is employed, and in fact prohibiting racial profiling will cause racist officers to instead hide their racism in elaborate disguises.

It is also argued that minorities, for fear of being brutalized by the police or inadvertently harmed by nervous and jumpy officers informed by racial profiling and the racism it entails, have had to adopt “protective mechanisms whether to avoid vehicle stops by police or to minimize the potential for harm during these stops,” and that these methods constitute an “altered public persona”.[56] These men

[sit] erect while driving, travelling at the precise posted speed limit...structuring their encounters with police during car stops: placing both hands on the steering wheel, responding to an officer’s questions with “sir” or “ma’am”....[57]

As well they should. These steps are merely sound advice for any person who both wants to drive safely and who might be stopped by a police officer, and they make for a more positive interaction. One should never make an officer feel as if she is in danger, especially if the person is innocent aside from a mere traffic violation. Russell has described the expected standard of conduct for the members of a civilized society toward police officers who are courteous and respectful to the person stopped, not a reluctant kow-tow to the police out of fear. It is therefore strange to note that this display of courtesy, obedience to the traffic law and a respect for the perception of the personal safety of police officers is characterized as an “altered public persona.” This leaves readers to wonder if she presumes these minority drivers would otherwise be aggressive, discourteous speeders who would take actions which would serve to intimidate police officers if it were not for their perception of “the potential for harm.”

Racial profiling leads the police to make “naked racial generalizations” which are too superficial in nature treat a wide range of minorities like criminals. An example of such generalizations is a police officer’s choice of shadowing a group of black kids into a store for fear of them shoplifting while leaving a group of Chinese kids in the next store alone.[58] In doing so, the police officer ostensibly does not rely on any factor other than race in his decisionmaking. However, such unwanted attention, when foisted upon innocent minorities, can be injurious to their self-esteem and harmful to police-citizen relationships.

People who use this argument have set up naked generalizations as a straw man to attack in place of racial profiling. To say that a police officer can make no finer a set of observations than the race of the people on his beat is to call him a poor police officer. Of course such broadly-based nets of suspicion are harmful, but they are also of very limited practical application. One ethnographer states that “many [blacks] are disturbed by the inability of some whites to make distinctions—particularly between people who are out to commit crime and those who are not... sales personnel [in stores] pay particular attention to people until they feel they have passed inspection, and black males are almost always given extra scrutiny” .[59] Unlike sales clerks, police officers need to learn to make distinctions “between people who are out to commit crime and those who are not” and to say that racial profiling leads them to waste their efforts on innocent people while hurting them at the same time is to say that they are not profiling as well as they could, not necessarily that it is wrong. Nonetheless, the issue of how racial profiling harms the relationship between citizens and the police is an important consideration that will be discussed later.

The legal climate

It is entirely possible for the legal actions surrounding an issue to fly in the face of certain ethical analyses. Nonetheless, briefly considering the relevant decisions of the Supreme Court and selected other courts regarding racial profiling is worthwhile. This is because the moral principles which guide the higher courts should not only ostensibly transcend political maneuverings, but also use as a grounding the same moral principles which are used to buttress and justify ethical analyses: egalitarianism, principles of justice, etc.

The foundation for all stop-and-frisk law as we know it can be traced back to the decision of Terry v Ohio.[60] The case involved a Cleveland, Ohio plainclothes detective, Martin McFadden. The detective had 39 years of experience as a police officer and 30 years of experience patrolling the downtown area that was the setting for the case. His attention on the day in question was drawn to three men, two of whom had separately walked back and forth from a store window to a street corner, where they conferred with the third man. McFadden stated that after observing them, “they didn’t look right” and that his suspicion was aroused. He watched them individually look into the store window and confer for several minutes and decided after a time to confront them about their conduct. When he did, they mumbled an unclear and evasive answer, and he executed a quick pat-down for his safety which yielded a loaded handgun from the coat of one of the men. He then frisked the other two and recovered another loaded handgun. The two with guns were placed under arrest for illegal firearms possession.

The issue the Court had to address was McFadden’s questionable Fourth Amendment grounds for seizing a person on a level of suspicion less than probable cause and conducting a limited search for his own safety, namely a “frisk.” The Court, in its decision to affirm McFadden’s actions, introduced the idea of “reasonable suspicion.” It “was a revolutionary decision because it was always assumed that any search or seizure without probable cause was unreasonable on its face and therefore violated both the common law and the Fourth Amendment”.[61] The Court had thus ruled that if an officer had a grounded suspicion that “criminality was afoot,” but which was short of probable cause, she was able to conduct a seizure of the person to include a quick, focused investigation. She could also then conduct a pat-down of the outermost garments of the suspect in the name of her safety.

In many senses it was the Court’s acknowledgement of a police officer’s expertise and feel for the street which those outside law enforcement often lack. “The Court treated McFadden’s largely unexplained suspicions as the ‘specific reasonable inferences’ of a highly ‘experienced’ officer rather than a mere hunch by transforming McFadden into an expert”.[62] Whether he was “transformed” into an expert by the Court or had become one over almost four decades of practice, the decision was a recognition of the fact that vital and judicious policework must sometimes take place in the absence of hard facts.

The opinion itself also made no mention of the race of the defendants, two of whom were black. This left the door for racial analyses of the decision open to swing two ways. One was that the Court deliberately left race out to make it clear that it was not interested in exploring the connection between race and suspicion in a manner that would detract from the main focus of the decision.[63] Other, less charitable commentary suggested that the lack of a racial dimension meant that the Court could not discuss the issue of race without detracting from the integrity of its opinion.[64]

This tack has yielded criticisms of the Terry decision which have been renewed recently in light of the maelstrom against racial profiling. After all, there exists the possibility that race was an unarticulated factor in McFadden’s mind when he concluded that the racially-mixed group of men “didn’t look right.” Some criticisms are more direct than others. One commentator, L Darnell Weeden, professed “doubt that three white males would have been rather routinely characterized as planning a robbery for engaging in that great inner-city past time of window-shopping by those who are either unemployed or under-employed”.[65] The suggestion shows no concern for the detective’s observation that the behavior in question was not leisurely window-shopping but instead in his experience closer to “casing a job, a stick-up.”

Weeden then proposes that Terry should be narrowed to provide for the exclusion of evidence recovered from frisks that is not directly linked with the initial motive for the stop “to deter police from expanding a reasonable stop and frisk into a general search for crime by abusing the Terry inquiry”.[66] Under this rationale, Weeden makes the case that the defendant Terry’s gun never should have been admitted as evidence against the defendant because “the nexus between the gun and the [theory of] robbery is too remote to justify the admissibility of the gun... there was not sufficient evidence to present probable cause for robbery by mere possession of a gun.” It is possible that there is no nexus between robbery and the illegal, public possession of a handgun, but there are only one or two other things that can be done with a handgun in such circumstances.

Weeden spends the rest of his essay trying to dismantle the powers given to the police by reasonable suspicion. While the courts are supposed to render decisions devoid of concessions to present or impending political debates, it can be hypothesized that Weeden and others attack the Court because its race-neutral language empowers the police to conduct stops which might quietly have race as a factor in the calculus of suspicion. An undiminished Terry thereby buttresses the practice of racial profiling and makes politically assailing it more difficult.

They are correct, and the Court is unambiguous in following up on the logic set by Terry. The powers it granted to police officers have remained intact to the present day, and reaffirmed in the decision of Whren v. U.S.,[67] “the culmination of a sequence of doctrinal and conceptual moves that began in Terry”.[68]

In Whren, the police conducted a traffic stop for the observed violation of a traffic infraction, namely turning without signaling and travelling at an unreasonable rate of speed. Upon executing the stop, the officers observed crack cocaine in plain view and effected the arrest of the two men inside the vehicle, from which an additional quantity of crack was recovered.[69] The men were convicted, but it was argued on appeal that the race of the men, who were black, made the officers suspicious in the first place, and the traffic violation was merely an excuse for investigating these men.

The issue before the Court in Whren was whether the stop of a car, prompted by the police observation of a traffic violation that under ordinary circumstances would be sufficient to justify a stop, should be deemed improper because the traffic rationale was a mere “pretext” to conduct an investigatory search.[70]

The Court held that not only was the stop legal, but “‘the actual motivations of the individual officers’ are irrelevant to Fourth Amendment analysis of the validity of a search or seizure [and] the Court specifically stated that this rule applies even when a search or seizure is prompted by ‘considerations such as race’”.[71] In short, if a stop was facially justified, other motivations such as race are Constitutionally irrelevant. “By removing the subjective motivation of the arresting officer from the Fourth Amendment calculus, the Whren Court effectively stripped defendants of their ability to establish that unlawful considerations such as race played a part in the decision to stop and arrest them”.[72] It is worth noting here that the author presumes that race is an “unlawful consideration,” but the Court does not specifically say as much. Nor does the author consider that the Federal Court of Appeals has ruled that stops of suspects in which race is a factor but not the sole factor in the stop do not violate “plaintiffs’ 14^th Amendment rights to equal protection because ‘they were not questioned solely on the basis of their race’”.[73]

What this does is leave the door open wide for the use of race as a factor in developing levels of suspicion. It also makes it clear that once an objective reason for a stop has been established, it no longer makes a difference if race was a factor in the first place. It is not entirely clear how this would apply to personal stops such as the one in Terry where only reasonable suspicion—not the probable cause of a traffic violation—has been developed. Some argue that in this vein the Court’s holding is too readily applicable to vehicles because it is always possible to find some technical violation committed by almost any driver an officer wishes to investigate.[74] However, it is clear how the ruling would apply to cases where a police officer has chosen to pay particular attention to a person on the street due to his race, among other factors, and is able to apprehend that person committing a crime only because he has chosen to pay such focused attention. The Court would uphold the ruling if the officer had refrained from making a seizure of the person until he had been observed committing a crime.

Critics counter that the inherent bias many police officers have against minorities transforms any type of behavior they evince into “suspicious behavior”.[75] However, this does not account for the fact that most street (not vehicular) encounters between police and suspects focus on male juveniles and adolescents.[76] If the police had a general, prejudicial bias towards minorities that was apart from rational considerations, they would stop as many adult black men and women as young men. Apparently their rationale is more nuanced.

Finally, the Court has made it difficult for plaintiffs to obtain injunctions against profiling. With the difficulty of citing Fourth Amendment violations stemming from the use of race as a possible motivation in police stops, some have tried to seek these injunctions on the grounds that the harms profiling poses to minorities are ongoing and serious. However, even if the plaintiffs were able to show that they had been harmed by profiling, the Court’s City of Los Angeles v Lyons standard would require them to show a “‘virtual certainty of future injury’ – an insurmountable hurdle”.[77] After all, it is extremely difficult for any group of plaintiffs to show that they will most certainly suffer future harm themselves due to racial profiling.

In sum, the law certainly leaves determinations regarding the future of racial profiling in the hands of police officials and policy makers. While it does not do so, it almost comes out and plainly says that race may play some role in police officers’ articulation of suspicion. Even if one does not agree with this strong a statement, it is exceedingly difficult to deny that the Court has had several opportunities to rule on racial profiling as it relates to Fourth Amendment search and seizure concerns. In fact, certain parts of New York State remain the only places where pretext stops are more stringently examined than in Whren, which has otherwise become the national standard.[78] Even in these jurisdictions, however, the focus is mainly on the scope and intrusiveness of the stop. It is still excessively difficult to draw out what unspoken factors might have led to the otherwise legal stop in the first place. Still, in each Federal case, the Court has refused to strip police officers of the right to conduct stops based on founded levels of suspicion, even if the accumulation of such suspicion was motivated in some undetermined part by race.

The ethical debate

Having spent time laying the foundation of profiling in theory, then arguing the empirical utility of profiling along racial lines, it is now possible to explore the ethical dimensions of the practice. By casting aside certain red herrings at the outset and looking at the present legal status of profiling beforehand, the discussion can remain more tightly focused.

The question can be framed as such: Can race be used as a factor in the decisionmaking process of police officers to target their proactive enforcement efforts when race correlates with certain relevant criminal phenomenon to a degree not indicated by mere chance?

There are several approaches to answering this question, but I will start from the position of assuming that the answer is “yes,” and then trying to void that answer. This approach is afforded by the following underlying assumptions: that it has been effectively shown that there is a firm empirical basis for racial profiling, and that this basis indicates the measure’s instrumental rationality and usefulness. It further supposes that the legal climate for the use of racial profiling is if not supportive, at least neutral or not prohibitive. Some strenuous arguments are fundamentally off base, and those have been set aside. What remains is to show that despite all of this, there are other, compelling reasons to consider racial profiling unethical.

A litmus test I frequently return to in these cases is John Rawl’s “veil of ignorance”. It contends that just principles are those that

free and rational persons would accept in a position of initial equality... no one knows his or her place in society, his or her class position or social status; his or her fortune in the distribution of natural assets... deliberating behind a veil of ignorance, people determine their rights and duties.[79]

Not knowing these things, would a person deliberating about rights be inclined to condone racial profiling if they might or might not be subject to it? The answer could easily be “no,” and for the following reason: it places an increased burden squarely on the shoulders of one group of people. Just because a person is black, he will bear the brunt of speeding enforcement measures while whites are more likely to speed freely. Most minorities are hard workers who commit no crimes, yet they will be subject to a disproportionate amount of enforcement action and temporary seizures of the person even though their physical and moral conduct is no different than those of whites.

At this point the argument and its rebuttals spin off in several different directions. The first rebuttal is that it is important to mitigate false-positive hits with courtesy, respect, and explanation. It is clearly for the good of all that criminals are apprehended, and if profiling can be executed with a minimum of intrusion and indignity, then the policy becomes more tenable for those behind the veil.

This underscores the ethical importance of using the most accurate and sophisticated profiles available. Profiling based on race is truly useful only when it is combined with other factors. The fact that 98% of blacks are not arrested annually versus 99.5% of whites each year might indeed be due to chance, and in any event ignores several reliable other factors such as age, geographic location, etc. Applbaum is correct to assume that not using these more finely-tuned methods is inefficient. However, in light of the fact that ignoring them also places an undeserved burden on a large volume of law-abiding citizens, it is also unethical to not employ them. Consider this: a person is stopped for speeding only because he is black. He is delayed and possibly subjected to a ticket. Is it unethical to use race as the main determination for this stop when this person would be driving freely if a more accurate and productive measure were used? It is indeed unethical, under the principle that government measures ought to intrude upon the freedoms of as few people as possible to achieve their goals.

Applbaum effectively argues that police mean nothing personal when they make such a stop, and that the police have the right to stop the speeder anyway for her initial violation. I would argue however that consistently considering race alone for punitive measures and seizures of the person functionally appears no different from racism, and that it is inadequate to merely insist to citizens that it is in fact not. The fact that race alone does not supply reasonable suspicion except in extraordinary circumstances supports this. The burden of proving this in practice is on the police, and the only way to do so is to treat race like the piece of categorical data many officers often refer to it as. This involves using it realistically, judiciously, and with as much a complement of other categorical data as possible. It also involves respecting the rights of all citizens who are ultimately innocent of any wrongdoing beyond minor traffic violations.

This suggests that most of the racial profiling conducted on the highways at present is ethically dubious. While reasonable suspicion is provided by traffic violations apart from any racial considerations, officers sometimes seize on driver errors which would be considered petty if not for the officer’s desire to stop and question a minority person. These include failing to signal a lane change on a desolate highway or a slight crack in a car’s brakelight lens. It is doubtful the veiled deliberator would agree to a system where certain offences would hardly ever be enforced except as pretexts solely to stop people of a certain skin colour; pretext stops should instead be used as tools befitting people of all races or at least in coordination with the more complex calculus above.

This dubiousness can extend to more widely-enforced traffic violations. That a person commits a violation and is black should not be the only cause for a stop among a pool of black and white speeders when there are better indications of criminality to be used, including age, type of car, just how fast the violator was going over the limit, and other measures. Minorities frequently complain that they are pulled over much more often for legitimate violations even if they are middle-aged men, wearing business suits, travelling with their families, and driving modest cars.[80] It is unethical to pull such people over when comprehensive analysis of the available data suggests the likelihood of their criminality is no different from a similar white person’s except by chance.

Hand-in-hand with this desire to limit such intrusions over too-broad a segment of a particular race should be the need to equally distribute sanctions across populations. If a cohort of minorities is still likely to be pulled over more frequently than whites, there is no reason why they need to be sanctioned proportionally more often than whites considering that the issuance of traffic tickets is discretionary. This will result in the whites who are stopped getting less warnings and more tickets, and the majority of blacks and Hispanics getting warnings. Not only does this mitigate the animosity violating blacks might grow to feel as a result of concentrated targeting, but it ensures that the extra burden otherwise innocent minorities bear is minimized in a very important way: as traffic law violators, they will not be subject to traffic adjudication disproportionately more often than anyone else.

The argument can then turn to a second rebuttal: that the factor of race is of such instrumental importance in the calculation of suspicion that to ignore it would be neglecting one’s duty. These cases differ from traffic violation stops in that race might be a factor used in the calculus of the reasonable suspicion in the first place, not of who to choose from amongst a body of people who all could be justifiably stopped for an observed infraction. Cases such as these form a continuum of race’s instrumentality in the calculus of suspicion. This continuum will be narrated, then outlined.

Take the following three examples to start with: In the first, a group of racial supremacists has been entering predominantly black neighbourhoods, physically assaulting residents walking alone at night and committing acts of arson. They strike frequently but at random and have hurt several people and destroyed a considerable amount of property. Late one Saturday night in such a black neighbourhood, an officer sees some men lurking near a bar, looking in its windows and conferring with each other in a quiet manner as intoxicated patrons occasionally stumble out. As cars pass by, the men move away from their headlights and into the shadows.

At this point, several things might already serve to raise the officer’s level of suspicion. Would the men’s race contribute to her level of suspicion? What if the men were black? Would she have less of a reason to suspect an impending act of violence or that these men might be at the heart of the recent crime wave? What if they were white? To deny that race is relevant in this case is idiocy. If the men are white the officer has an obligation to make a limited inquiry into their business in the area, and a failure to do so would border on negligence. It would be exceedingly hard to show that her behavior would be unethical.

Now take the example of a group of black militants who have been ambushing police officers in an effort to combat what they think is social injustice, as happened in New York during the 1970s. They have shot several police officers already, and even killed one of them. Patterns indicate the shootings take place in a small handful of racially-mixed areas. One night, the police get a call of a disturbance at a home in one of those areas. One unit responds to the house, but a second one spots a car hidden in the shadows nearby. It is occupied by two young men who have an excellent vantage point of the house in question. As the officer surveys the men, a call comes over the radio from the first car that the disturbance call was unfounded. Should the officer take relevant precautions and then attempt to investigate the men in the car? What if the men were white? What if they were black? Race seems to be the deciding factor, and the ethical issues seem to be clear.

In a third example, one author takes a step away from the traditional realm of racial profiling to suggest its appropriateness in other circumstances. Instead of looking at instances of street crime, he looks at a case in which nuclear secrets are stolen by China. This is not an incident far removed from reality, either. He proposes that

China is known to have obtained a top-secret warhead design. Among those with clearance to work on that design are people from various kinds of national and racial background [sic]. Which ones should the investigators concentrate on? The Swedes? The answer surely is: They should first check out anyone who has family or friends in China, who has made trips to China, or who has met with Chinese officials.[81]

Are these individuals under investigation being racially profiled? Yes, they are. Asians, and people who consort with Asians would be put under close scrutiny primarily because of their race. However, the issue of race as a proxy for nationality is practically instrumental to the crime and of course the stakes are extremely high. In a case such as this, not to consider race as a jumping-off point for the focus of an investigation would be foolish if not negligent. While extreme, it demonstrates an instance in which the use of racial profiling in the concentration of investigative effort seems both ethical and prudent.

What these three cases also illustrate is that the closer a race-laden profile comes to being an actual description of a person or persons wanted for specific crimes committed in the past, the more ethical and prudent its employment is. The ultimate case of this is of course the description supplied by a victim used in a police canvass. Consider the case of a witness to a shooting in the recent past who states the suspect was, among other things, a dark-skinned black. It is clear that the police should not stop and question any race other than blacks except perhaps non-blacks who could supply the whereabouts of this particular black person. The same logic of course applies to any race. When a person has been shot, robbed or raped and the perpetrator is gaining a lead over law enforcement as every second passes, to stop and question whites and Hispanics—or women if the criminal is a man—is not only bad police work, but an act of negligence.

What the first three cases do, then, is suggest that racial profiling is most acceptable in instances where race is absolutely instrumental to the crime at hand. They bridge the speculative gap with the above shooting case which involves the ethically-accepted canvassing for particular criminals wanted for a documented crime with known descriptions that include race. One small step removed, they involve proactively looking for people who might commit crimes based on known trends and information to include race. While race is a strong deciding factor in taking proactive action—perhaps the strongest element of the criminal profile—it still seems to be acceptable because the results of a false-positive could be minimally intrusive, the cases are all peculiar, and the total sum of the profile outlines a narrow range of people with a relatively high positive-hit probability. This again implies overall that careful, well-done profiling that involves good-faith efforts at accuracy, success and minimal intrusiveness are ethically acceptable even if race is a factor.

Race in these examples is instrumental in the decision to take police action then for one of two reasons. Either because it is an inexorable element of the crime (ie it is what is commonly called a “hate crime”), or because actual descriptions by witnesses ascribe a particular race to the criminal, thus making race an inexorable factor in the instance. It is possible for both reasons to be present and buttress each other. One step further removed from this are cases in which race need not be inexorably instrumental, but has become so due to a prevalent criminal phenomenon. Consider a fourth, true case presented by the Police Chief of Los Angeles, Bernard Parks, who is black:

We have an issue of violent crime against jewelry salespeople... the predominant suspects are Colombians. We don’t find Mexican-Americans, or blacks, or other immigrants. It’s a collection of several hundred Colombians who commit this crime. If you see six in a car in front of the jewelry mart, and they’re waiting and watching people with briefcases, should we play the percentages and follow them?[82]

Well, what if these Colombians in the car are a husband, wife, grandmother and three children? Parks suggests “We’re not just using race... it’s got to be race, plus other indicators, so that won’t happen.” It is clear that six Colombian men in a car in front of a jewellery store in this climate warrant very close watching, and the ethical dilemmas are small. It is likely that veiled deliberators would accept the possibility of being under police surveillance in such situations, and that there are few attendant ethical harms to speak of.

What has happened here is that the phenomenon of jewellery-store robberies by Colombians has lead to a racially-laden profile of considerable value. There is considerable utility and little debate about looking for the specific suspects of specific crimes in the days following the crime if they have yet to be apprehended. Such is the substance of an “ongoing investigation.” This is merely a prolonged version of the immediate canvass. Close on the heels of this practice is the one of using the fruits of these ongoing investigations to widen the net, using a generalized description to apprehend people who may commit a specific phenomenon of crime. Instead of looking for the particular Colombians who have committed the last three dozen jewellery-store robberies in Los Angeles, officers would instead generally look for Colombians who might be planning to execute the next dozen. There will likely be an overlap of these two populations of suspects, but what is clear is that the officers would be using race and other data to focus on Colombians whose behaviour suggests a nexus with jewellery store robbery. Throughout the literature of racial profiling, no author has effectively addressed the ethical legitimacy or efficacy of a measure such as this one.

The next step, however, is not on as solid a footing. It is the oft-discussed use of generalized descriptions including race in the investigation not of a particular criminal phenomenon or trend but of a broad range of crimes. It is the province of following the young male black late at night while watching the older male black go about his business. When these decisions are based on known trends, current analysis and incorporate numerous significant factors in addition to race, their application is ethically sound if it is done conscientiously and in good faith. Beyond this is the subconscious assimilation of factors including race into a “hunch,” which should not be ignored when utilized by extremely experienced police officers, but is too unrefined to allow for widespread use. In this case, it is not too much to ask an officer to articulate the empirical details of her hunch. If Detective McFadden had been able to, many avenues for the attack of Terry would have been cut off in 1968.

Lowest on this food chain of speculation is the highway stop discussed earlier. When motorists are stopped only because of their race when other, more refined means are available for use, it is extremely difficult to justify the use of such a simplistic racial generalization in light of the harm and dissension it causes. While this objection seems to be placed in policy, it is also a throwback to Applbaum’s assessment that not using the most refined means possible is to sacrifice efficiency. It can be further argued that such a simplistic yet inflammatory practice as simply pairing race with an observed violation or—in the case of merely focusing attention on certain persons—nothing else is negligent enough to be unethical. It is true that blacks might commit crimes at a higher rate than whites, but it is still wrong then to pick the first black person you see and start tailing them.

What this narrative illustrates is a continuum of race-laden profiling that starts out as being ethical and descends into not only unethical practices but also unsound policework. The continuum can be divided into six categories:

Instantial instrumentality: Race is relevant to the formulation of suspicion against a person because witnesses to the crime (including the victim) have identified the perpetrator(s) as being of a certain race in the instance at hand. This case is relevant to the hunt for people who have committed crimes in the past both in the immediate canvass and during ongoing investigations if the suspect is not quickly apprehended.

Ipso Facto instrumentality: Race is relevant to the crime by its very nature: blacks committing hate crimes against whites, whites against blacks, or the Chinese stealing nuclear secrets from the US. While it is conceivable that a white person might be self-hating and attack other whites for that reason or a Swede might steal US secrets for the Chinese, the “very nature of the crime” rationally suggests otherwise and forms the basis for race-laden suspicion.

Phenomenal instrumentality: Particular crime trends indicate that race is for whatever reason instrumental to the crimes under investigation. It could be the above instance of Colombian (Hispanic) gangs who rob jewellery stores, or Mexicans trying to illegally cross into the United States at the Texas border.[83] In any case, a trend presents itself that for some reason incorporates race in a clear, consistent manner. Not only will apprehension efforts target a particular race for reasons of instantial instrumentality, but proactive efforts to head off future crimes will involve singling out suspects from the larger population of a particular race.

Multivariately linked: A conscious examination of a set of variables which includes race suggests that a certain segment of a population is more likely than its remainder to commit certain types of crimes. These variables can include many things such as age, sex, dress, behavior, known history, etc. While these factors may create a level of suspicion, the point at which they become reasonable enough for a stop is open to debate.

Speculatively linked: A lesser version of a multivariate link, this articulation of suspicion incorporates variables including race, as well as others in a more informal, hunch-based manner which has the potential of being largely anecdotal. Its usefulness and ethical appropriateness depends in some part on the expertise and experience of the police officer involved.

Univariately linked: This is akin to a naked racial generalization, relying merely on the crude fact that blacks and Hispanics tend to offend at higher rates than others. Not only is such a link unethical because it fails to account for larger, more complex analyses which would exclude the vast majority of any population from an unnecessary intrusion, but it is the method which can most easily be used to disguise blatant racism. It is also the method most typically decried in its application to selecting the subjects of car stops.

What these categories do is allow for a thumbnail sketch of the ethics of racial profiling. Race ought to be incorporated into profiles used in law enforcement decisionmaking if race is instantially or ipso facto instrumental. Not to would be to engage in an unethical brand of professional neglect. The same thing is substantially so for cases of phenomenal instrumentality because it is akin to instantial instrumentality but encompasses numerous enough cases to allow for speculation about the race of perpetrators of future crimes within the phenomenon.

Multivariately profiling a person to include race among one of many characteristics in the articulation of suspicion opens the door to more subjective measures. However, it is still ethical to do so as long as the profile is conscientiously written and fairly and consistently applied. Race as one small part of a sketch of suspicion, especially in light of complex and well-documented empirical proof of rates of criminality not of whole races in general but certain subcategories (such as those which incorporate age, sex, and behavior) is an ethical and rational tool in the focus of police efforts.

This strength points to the weakness of a speculative link. While it is possible to acknowledge the expertise and experience of a veteran police officer in her ability to formulate correct hunches, too often hunches are used as an excuse for misconduct and for the perpetuation of stereotypes which are incorrect and counterproductive. In light of the better methods available, the speculative hunch that is informed by race may sometimes be a sound tool but is in the largest range of cases ethically dubious by virtue of its fundamental inexplicability.

So it goes with the clumsy univariate link, which simply targets minorities because they are part of a population more likely to offend than the population at large, without refinement or further investigation. The result of this practice is the temporary seizure of a large population of people who are innocent, and who wouldn’t have been stopped if more sophisticated measures were used. This puts such a link on very shaky ethical ground, to be avoided as it provides a clear nexus between latent racism, policework, and the erosion of a community’s trust in its law enforcement.

It is possible then to see that using race in the effort to apprehend criminals is in many cases an ethical and effective method. However, the clear cases are also the most compelling ones and often fairly peculiar. The burden of a good-faith effort rises considerably once these cases are abandoned in favor of a more generalized practice. This does not mean that the profiling in such cases would be unethical per se, however, but instead that it becomes more demanding to undertake ethically. At the far end, the ham-handed, simplified link between race and crime absent other considerations becomes practically impossible to perform ethically in the face of more rigorous alternatives.

Policy recommendations

Regardless of whether there is an ethical justification for racial profiling, it is abundantly clear that many people simply do not like it. This is attributable to the acknowledged history of racism present in the United States[84] and skepticism about the police’s ability to employ racial profiling judiciously and in a good-faith manner.

The first step toward a tenable solution to this problem is to enact police policies which prohibit racial profiling in the cases above where it is clearly wrong. These are cases such as car-stops where officers scrounge for traffic violations only to stop people of a certain race absent other relevant factors. The second step would be to make sure in the case of traffic violations that sanctions for such violations are distributed in proportion to the racial distribution of the offending population. This would ensure that no particular race must shoulder not only the burden of greater intrusiveness due only to skin colour, but also a disproportionate amount of the penalties that all violators of all races ought to bear.

Beyond this, two important points must be stressed in the application of race-based decisions to stop and investigate certain people. Rigorous policing of Fourth Amendment violations must be undertaken lest racial profiling continue to go hand-in-hand with unlawful searches and seizures. Secondly, policework must be conducted with professionalism and respect lest racial profiling be seen as synonymous with rudeness, discourtesy, and officers whose attitudes erode the relationship between the police and the community. Too often this point is given mere lip service, and as a result such relationships continue to deteriorate irrespective of the presence of racial profiling.

When the above conditions are met, it is hard to deny the more compelling applications of racial profiling. Time and time again, this essay has returned to the need for comprehensive, multi-faceted profiles which incorporate race if race is to be used at all. Still, Randall Kennedy rejects even these as being too injurious to the already fragile state of race relations in the United States. He is correct in his analysis of race relations in this country, but whether to take his advice or not is a matter of a policy decision.

Regarding such a decision, it is worth noting that it is possible to exclude race from decisionmaking and still be able to articulate a level of reasonable suspicion against almost the same population of people provided an analysis more refined than the univariate link is employed. Policy makers and other police officers would do well to observe the strategies and methods employed by police officers in racially homogeneous neighborhoods. Such officers are faced with making decisions that cannot incorporate race except in a trivial manner. Their task is to select, for example, black persons to investigate from amongst a pool of black people. The result is the cultivation of an expertise which does not necessarily include racial generalizations. In this way, these officers cannot use race as the crutch for sloppy policework it is often thought to be.

The New York City Police Department’s stop, question and frisk report worksheet (reproduced in full in Appendix B) highlights an officer’s ability to use numerous factors other than race to articulate a level of suspicion. The factors officers can indicate come in three forms: those of background circumstances, the particular actions of the suspect, and environmental factors. They include 26 in all, not counting an opportunity to articulate other factors not included on the form. Some pre-printed ones are “Report From Victim/Witness/Officer,” “Inappropriate Attire for Seasonal Weather,” “Carrying Objects In Plain View Commonly Used in the Commission of a Crime eg, Slim Jim/Pry Bar, etc,” “Furtive Movement(s),” “Knowledge of Suspect’s Prior Criminal Behavior,” and “Time of Day, Day of Week, Season Corresponding To Reports Of Criminal Activity.” Nowhere is race mentioned, nor would it be in the present political and policy climate in New York City. However, it still seems as if police officers there have a considerable arsenal of factors of suspicion to use in the investigation of crime and of potential criminals.

This fact suggests a reasonable course for police policy makers. It would encompass the use of racial profiling in those cases where it is instrumental to the crime at hand. In fact, there seems to be little option to do otherwise in these cases. Beyond that, race should be used with caution. Instead, police officers should be encouraged to cultivate their investigative skills in a direction that leads them away from the proxy of skin colour and towards the multitude of other factors which also can be used in conjunction to articulate suspicion. Such a refinement of ability would enable a police department to formally abandon racial profiling if its citizens demanded it without a considerable loss in policing efficiency. Any policy doing this would at the same time prohibit the simple use of race in traffic enforcement decisions and certainly in the meting out of sanctions. In this way, the goals of the police department could be met without clashing with the expectations of citizens in a nation recovering from decades of racial turmoil.

Conclusions

If police officers are reluctant to abandon racial profiling it is only because they feel it works, and because in the broadest sense, empirical data supports this fact. However, police officers are not tasked with satisfying the sensibilities and expectations of the communities they serve, especially when the community includes the race they are paying particular attention to and has only just begun to recover from a history of racism and oppression. In short, the fictional Trooper Riggins of the introduction should not underestimate the ill effects his targeted car stops have on the numerous otherwise innocent minorities he repeatedly pulls over.

That does not mean, however, that politicians should ignore sound law-enforcement techniques that can help to reduce crime. In a sense, this has been happening with the politically-expedient and near-universal condemnation of all forms of racial profiling. Such a condemnation has served to chill the necessary academic and internal discussion police agencies need to deal with the tough issues of racial profiling without causing divisiveness and friction within police agencies and towards the communities they serve.

Indeed, the issue is much more subtle, as authors such as Applbaum and Kennedy suggest. It is necessarily a balancing act. While crafting a policy that works against harmful and counterproductive generalizations with only marginal benefits, police departments and politicians must also recognize that there are ethical applications of racial profiling which, if neglected, would do more than merely encumber police officers: the neglect would wreck unnecessary harm on all sectors of the citizenry. The place at which policy-makers should stop in the continuum must never be near the margins, but instead somewhere in the middle. Within that middle, the appropriate limit is no longer an ethical one but one of sound and careful policy compromises.

Appendix

A: Distribution of crimes by race, nationwide, 1997: FBI UCR


B: Stop, question and frisk report worksheet, PD 344-151A, 12/ 2000 Revision, NYPD
Front Back

---

[*] Brandon del Pozo is a New York City Police Officer assigned to patrol in the 67^th Precinct, Brooklyn South. He is presently on temporary assignment writing terrorism response training and doctrine for the NYPD, and is completing his Master of Arts in Criminal Justice at the John Jay College of Criminal Justice, New York City, NY. He is also a First Lieutenant, rifle company commander, in the US Army National Guard. Send correspondence to [e-mail].
[1] This story is fictional.
[2] I Applbaum, ‘Racial Generalization, Police Discretion, and Bayesian Constructualism’ in J Kleinig (ed), Handled with Discretion, Roman and Littlefield New York 1996 at 145-157.
[3] The Maryland State Police have reported a 30% success rate. See J Goldberg, ‘The Colour of Suspicion’, The New York Times Section 6 (magazine) June 20 1999 at 51.
[4] D Harris, ‘Driving While Black: An ACLU Special Report’ (1999) <http://www.aclu.org/profiling/report/index.html> retrieved December 13 2000, 18 pages.
[5] Harris, supra n 4; Goldberg, supra n 3.
[6] S Taylor, ‘Racial Profiling: The Liberals are Right’ (1999) 31 No 17 Legal Affairs at 1084.
[7] D Cole, ‘The Color of Justice’ The Nation, October 11 1999 at 12-15.
[8] J Derbyshire, ‘In Defense of Racial Profiling: Where is our Common Sense?’ The National Review, February 19 2001 at 38-40.
[9] Usually this consideration is a negative one, such as for example assaulting a person only because he is Hispanic. However, racism may come in the form of a positive consideration, such as offering discounts to people at a store only because they are white. In this sense, charging others full price would be racism.
[10] S Holmes, ‘Clinton Orders Investigation on Possible Racial Profiling’ The New York Times, June 10 1999, A22.
[11] [1968] USSC 142; 392 U.S. 1 (1968).
[12] [1996] USSC 55; 517 U.S. 806, 812 (1996). These decisions will be discussed later in the work. Briefly, Terry concludes in cases where a police officer is able to articulate a reasonable suspicion that criminality is afoot, that even though such an articulation falls short of the hard facts of probable cause, the officer may nonetheless detain the person(s) under suspicion for investigation and can conduct a “carefully limited” (at 30) search for weapons. Whren concludes that in the calculation of this reasonable suspicion, a factor such as a traffic violation that would normally justify a police stop is sufficient in and of itself to justify a particular stop, even if it is only a pretext for a deeper investigation into other possible crimes, and even if this larger calculus includes race as an unarticulated factor.
[13] Goldberg, supra n 3; R Kennedy, ‘Suspect Policy’ The New Republic, 13 September 1999 at 30-35.
[14] N Lemann, ‘The Empathy Defense’ The New Yorker, 18 December 2000 at 46-51.
[15] Applbaum, supra n 2.
[16] Applbaum, supra n 2; Cole, supra n 7; J Taylor & G Whitney, ‘Crime and Racial Profiling by US Police: Is There an Empirical Basis?’ (1999) 24 No 4 The Journal of Social, Political and Economic Studies at 485-510; P Williams, ‘Road Rage’ The Nation, 22 March 1999 at 10.
[17] H McGary, ‘Police Discretion and Discrimination’ in J Kleinig (ed), Handled with Discretion Roman and Littlefield New York 1996 at 131-144.
[18] Goldberg, supra n 3.
[19] While the police often profile certain motor vehicles as more likely to be driven by criminals, I have not mentioned them here. Cars, as well as all objects, when profiled, are actually profiles suggesting a pattern of behavior by the people who possess them. The officer assumes “the driver of that car is more likely to be a criminal,” not “that car is a criminal car.” Thus, profiles of objects are actually profiles of their owners. J Miller, Profiling Populations Available for Stops and Searches Police Research Series Paper 131 (2000) London: Home Office, Police and Reducing Crime Unit.
[20] S Trende, ‘Why Modest Proposals Offer the Best Solution for Combating Racial Profiling’ (2000) 50 Duke Law Journal:331.
[21] Applbaum, supra n 2.
[22] Applbaum, supra n 2.
[23] D Courtwright, Violent Land: Single Men and Social Disorder from the Frontier to the Inner City, Harvard University Press Cambridge 1996.
[24] ‘Crime In Your Neighbourhood’ New York Daily News, 22 December 2000 at K4-K5.
[25] D Wasserman, ‘Racial Generalizations and Police Discretion’ in J Kleinig (ed), Handled with Discretion Roman and Littlefield New York 1996 at 115-130.
[26] Applbaum, supra n 2.
[27] Harris, supra n 4; L Weeden, ‘It Is Not Right Under the Constitution to Stop and Frisk Minority People Because They Don’t Look Right’ (1999) 21 University of Arkansas at Little Rock Law Review 829.
[28] McGary, supra n 17.
[29] Applbaum, supra n 2; Greenberg, 1999; McGary, supra n 17.
[30] Wasserman, supra n 26.
[31] K Russell, The Color of Crime, New York University Press New York 1998; Williams, supra n 16.
[32] Kennedy, supra n 13; Russell, supra n 32.
[33] Trende, supra n 21.
[34] Supra n 16.
[35] Ibid.
[36] Ibid.
[37] For this reason, William C Heffernan of the John Jay College of Criminal Justice refers to the murder rate as the “gold standard” of crime statistics.
[38] Williams, supra n 16.
[39] Kennedy, supra n 13.
[40] <www.fbi.gov/ucr/ cius_97/97crime4.pdf>.
[41] A Aron & E Aron, Statistics for the Behavioral and Social Sciences, Prentice Hall Press Upper Saddle River 1997.
[42] J Reiman, The Rich Get Richer and the Poor Get Prison, 6^th edn Allyn and Bacon Boston 2001.
[43] Williams, supra n 16.
[44] Wasserman, supra n 26.
[45] Williams, supra n 16.
[46] Taylor, supra n 6.
[47] Cole, supra n 7.
[48] Taylor, supra n 6.
[49] Goldberg, supra n 3; Kennedy, supra n 13.
[50] Williams, supra n 16.
[51] Taylor, supra n 6.
[52] Applbaum, supra n 2; R Kennedy, Race, Crime and the Law, Pantheon Books New York 1997; Kennedy, supra n 13.
[53] Harris, supra n 4.
[54] Applbaum, supra n 2; Kennedy, supra n 53; Kennedy, supra n 13; McGary, supra n 17.
[55] Kennedy, supra n 53.
[56] Russell, supra n 32.
[57] Ibid.
[58] Wasserman, supra n 26.
[59] E Anderson, The Code of the Street: Decency, Violence and the Moral Life of the Inner City, W W Norton New York 1999.
[60] [1968] USSC 142; 392 U.S. 1 (1968).
[61] Weeden, supra n 28.
[62] A Thompson, ‘Stopping the Usual Suspects: Race and the Fourth Amendment’ (1999) 74 New York University Law Review 956.
[63] Ibid.
[64] Weeden, supra n 28; Thompson, supra n 53; A Abramovsky & J Edelstein, ‘Pretext Stops and Racial Profiling After Whren v United States’ (2000) 63 Albany Law Review 725.
[65] Weeden, supra n 28.
[66] Ibid.
[67] [1996] USSC 55; 517 U.S. 806, 812 (1996).
[68] Thompson, supra n 53.
[69] Abramovsky & Edelstein, supra n 55.
[70] Thompson, supra n 53.
[71] Ibid.
[72] Abramovsky & Edelstein, supra n 55.
[73] A Newman, ‘Court Upholds Extensive Stops of Black Men After a Crime’ The New York Times, 27 October 1999 at B5.
[74] Abramovsky & Edelstein, supra n 55.
[75] Thompson, supra n 53.
[76] This observation is admittedly anecdotal in this case, based on this author’s experience as a police officer. However, see Courtwright, supra n 24 and Anderson, supra n 50 re: adolescent men and crime.
[77] B Garrett, ‘Standing While Black: Distinguishing Lyons in Racial Profiling Cases’ (2000) 100 Columbia Law Review 1815.
[78] Abramovsky & Edelstein, supra n 55.
[79] J Rawls, A Theory of Justice, revised edn Belknap Press 1999.
[80] Harris, supra n 4; Russell, supra n 32.
[81] Derbyshire, supra n 8.
[82] Goldberg, supra n 3.
[83] Thompson, supra n 53.
[84] Applbaum, supra n 2; Kennedy, supra n 13; Kennedy, supra n 53.