Breaking the pattern of racial profiling: when law enforcement officers make discretionary judgments based on race, can litigation put an end to stop-and-search violations?

On a summer evening in 1991, four young African-Americans were returning to Delaware from a church service in Philadelphia. Although the driver had committed no traffic violations, police officers stopped the car just south of the Philadelphia International Airport and ordered the driver and passengers out. The officers conducted an intrusive search of the car and its occupants using a narcotics-trained police dog and detained the group for almost an hour until convinced that they were not transporting drugs.

To justify the stop, the police issued a warning for obstruction of the car's windshield by a thin piece of string hanging from the rear-view mirror (which could not have been observed by the officers before the stop). When one of the car's occupants asked why they had been stopped, an officer answered with surprising candor, "Because you are young, black, and in a high drug-trafficking area driving a nice car." (1)

Virtually everyone agrees that it is impermissible to stop or search someone solely on the basis of race, but many law enforcement officials and courts continue to assert that race is a legitimate factor in making policing decisions--even without specific racial description of a suspect.

Litigation and government investigations concerning allegations of racial profiling on the New Jersey Turnpike provide data demonstrating a long-standing pattern of profiling. In criminal cases arising from drug arrests, a court found substantial evidence that stops and searches were highly disproportionate based on race. (2) The court determined that blacks and whites violated traffic laws at about the same rate, but that 42 percent of stops and 73 percent of arrests were of African-American motorists. A study by New Jersey's attorney general found that searches of cars on the turnpike were even more racially disparate than the initial stops: 77 percent of all searches consented to were of minorities. (3)

A New York attorney general study of 175,000 pedestrian stops in New York City found a highly disproportionate number of minorities stopped. (4) It determined that African-Americans were stopped six times more frequently than whites; that in precincts where African-Americans constituted 10 percent or less of a largely white population, they were stopped 30 percent of the time; and that, adjusting for crime rates by race, African-Americans were stopped twice as often as whites.

Rationales for racial profiling

Police defend their racially disparate practices by saying that generally, minorities commit more crimes than whites. They also hold that enforcement of criminal laws that are violated by whites and minorities in roughly even numbers (for example, narcotics violations) is disproportionate because the location and social impact of the same types of crimes justifies a more aggressive response in minority communities. They argue that current practices work: Aggressive policing and targeting of minority communities have led to significant seizures of contraband, weapons, and fugitives--and a reduction in crime.

In the wake of September 11 events, advocates of racial and ethnic profiling point to the need for heightened suspicion regarding the activities of young Arabic men. The arrests and detention of hundreds of people after the terrorist attacks has created considerable controversy--many of these people would not have been subject to this treatment were it not for ethnic characteristics, and the government has not yet provided evidence linking them to terrorist activities. Furthermore, it is not likely that ethnic profiling will be any more useful or constitutional than racial profiling.

In the area in which racial profiling has been most controversial--narcotics enforcement-proponents' arguments do not withstand empirical and legal scrutiny. For example, the data do not indicate a minority-dominated drug trade. National drug abuse studies show that minorities possess and use drugs only slightly more frequently than whites do. (5) "The typical cocaine user is white, male, a high school graduate employed full time, and living in a small metropolitan area or suburb," former drug czar William Bennett has said. (6)

Arrest statistics are also misleading. New Jersey's attorney general pointed out that these statistics are "a self-fulfilling prophecy where law enforcement agencies rely on arrest data that they themselves generated as a result of the discretionary allocation of resources and targeted drug enforcement efforts." (7)

Empirical evidence from reviews of car stops and searches supports this view. On the New Jersey Turnpike, 10.5 percent of contraband seized during traffic stops came from white drivers and 13.5 percent from African-American drivers. (8) In Maryland, searches on 1-95 resulted in "find rates" that were roughly equal by race. (9) In both states, mostly small amounts of drugs were seized, indicating possession for personal use.

Millions of drivers use these highways each day, yet so few stops or searches of motorists--black or white--result in contraband seizures that it's hard to justify stopping large numbers of African-Americans so the police can make the occasional seizure.

The logic behind racial profiling is faulty, as Ira Glasser, former executive director of the Americal Civil Liberties Union, has said:

   Even if most of the drug dealers in the Northeast corridor or in any
   particular neighborhood or city are black or Latino, it does not follow
   that most blacks and Latinos are drug dealers.... Think about it for a
   minute. Most players in the NBA are black. But if you were trying to get a
   team together, you wouldn't go out in the street and round up random
   African-Americans.

      It's a very simple, logical fallacy. The fact that most drug dealers are
   X does not mean that most X are drug dealers. (10)

In policing, as in many areas of contemporary American life, race matters--and it matters a lot. The substantial racial disparities documented in stop, frisk, and search practices cannot be fully explained or rationalized by crime patterns, police deployment, or policing tactics.

Automobile stops

A series of U.S. Supreme Court decisions over the past 30 years has delineated the rules for law enforcement stops and searches of vehicles. The Court has permitted stops if officers have cause to believe that a crime has been committed, including any traffic violation. (11)

Once a car has been stopped, the driver and passengers can be ordered to stand outside the vehicle (for the officer's protection) without any objective showing of harm or danger. (12) At any time during the encounter--even after a ticket or warning has been issued--the police can secure consent to search the car or its occupants without advising the driver or passengers that consent need not be given or that they are free to leave. (13) Any contraband that is observed in "plain view," including anything pinpointed by flashlight, can be seized and used as probable cause to arrest. (14)

During the stop, anyone who appears dangerous may be frisked for the officer's protection. (15) If cause is established to arrest any of the occupants, a full-scale search of the car--including suitcases and other private containers--and of all other passengers is permissible. (16)

In Whren v. United States, officers made a traffic stop and observed two bags of crack cocaine in the hands of a front-seat passenger. (17) The police testified that they stopped the driver for violating several traffic laws. The defendants claimed that the stop was pretextual; that the police were suspicious because they observed two African-American men in a Nissan Pathfinder in Southeast Washington and that they alleged traffic violations so that they could conduct a drug investigation.

The Supreme Court ruled that from a Fourth Amendment perspective, the "constitutional reasonableness of traffic stops [does not depend] on the actual motivations of the individual officers." (18) The only relevant question was whether the officer had legal cause for the stop.

In response to the claim that pretextual stops could be racially motivated, the Court stated that the Equal Protection Clause of the Fourteenth Amendment would prohibit any intentional race discrimination in a car stop. Intentional race discrimination may be shown by a law that "expressly classifies persons on the basis of race." When there is no direct proof of such a policy or practice, statistical evidence must show that the police acted with the intent to discriminate. The Supreme Court has sustained attacks on racially discriminatory jury selection procedures and racially based peremptory challenges, using statistical evidence demonstrating that race played an impermissible role in these proceedings. (19)

However, in McCleskey v. Kemp, the Court rejected, as insufficient under Eighth and Fourteenth Amendment standards, a statistical analysis concerning the application of the death penalty in Georgia. (20) The study found that the race of the victim was consistently the most important factor in influencing a jury to impose the death penalty. If the victim was white, the odds of a defendant's receiving the death penalty were 4.3 times higher than if the victim was African-American.

The Court questioned how far the logic of the argument would carry: The claim, Justice Lewis Powell stated, "throws into serious question the principles that underlie our entire criminal justice system." (21) Justice William Brennan, dissenting, commented that the Court's concern with the implications of the statistics for other aspects of the criminal justice system displayed "a fear of too much justice." (22)

The Court took a similar hands-off approach in United States v. Armstrong, reversing a ruling that would have permitted discovery in support of a motion to dismiss federal crack cocaine prosecutions against African-American defendants on grounds of racially selective prosecution. (23) The Court stated that prosecutors were entitled to a presumption that they performed their duties properly. It relied on arrest statistics to reject the argument that blacks and whites commit drug offenses at rates that would not justify the highly disproportionate federal prosecution of African-American crack cocaine offenders.

McCleskey and Armstrong are troubling, but they are not dispositive of racial profiling claims. In McCleskey, the Court--in requiring proof beyond statistical patterns for the whole system to show that the judgment in an individual case was infected by intentional discrimination--stressed the unique discretion given juries to decide between life and death.

Armstrong's requirement that the defendant show that similarly situated white offenders were not subjected to federal prosecution, while questionable, should not bar selective policing claims. Evidence that shows a statistical disparity in the rates at which similarly situated black and white drivers are stopped and/or searched for alleged traffic violations establishes the factual basis for the claim. (24)

When a racial profiling challenge is based on the theory that official policy contains an express racial classification, there should be no need to prove the existence of a similar nonminority group or person who was not subjected to the practices. (25)

Pedestrian stops

Pedestrian stops can be triggered by any number of factors, although legally, police officers cannot forcibly stop or detain people without reasonable suspicion of criminal conduct. The complexity of on-the-street investigations, from the perspectives of both race and crime control, makes legal and political judgments in this field controversial.

The seeds of constitutional controversy were planted in 1968 in Terry v. Ohio. (26) The Supreme Court ruled that a stop is reasonable under the Fourth Amendment if the facts and circumstances provide grounds for an officer to believe, objectively, that criminal activity is afoot. The Court decided that if the officer had a reasonable belief that the person stopped was armed and dangerous, he or she could conduct a pat-down frisk, again on less than probable cause.

Since Terry, the Court has significantly expanded police powers (often in deference to the war on drugs), and complaints concerning racial bias and arbitrary stops and frisks continue to mount.

While the issues of both complaints are often addressed as distinct constitutional matters, police investigative detentions often include both. For example, in Brown v. City of Oneonta, a police manhunt followed the assault of an elderly woman in that New York town. (27) She told the police that her attacker was "a young black man" who had cut his hand during the incident. Fewer than 300 African-Americans lived in town, and approximately 2 percent of the students at a nearby state university were black. The police made a sweep of the town, stopping and questioning more than 200 African-American men and inspecting their hands for cuts.

In a civil rights suit, the Second Circuit ruled that if police had a racial description and a relatively small number of possible suspects--even though there was no description beyond race and age--investigating all African-American men did not constitute intentional race discrimination.

Having isolated and rejected race as a constitutional problem, the court proceeded to analyze the Fourth Amendment claims to determine whether the police "seized" any of the plaintiffs. The court recognized that the description of "a young black man" was too vague to justify a stop. It also found that a police officer's pointing a spotlight at the plaintiff and saying, "What, are you stupid? Come here. I want to talk to you," and instructing him to show his hands constituted a seizure.

Brown's equal protection analysis finesses the difficult issues. The fact that a description of a suspect includes a racial characteristic should be merely the start of the analysis. Race is surely an appropriate consideration in such an investigation, but when it becomes the predominant factor, strict scrutiny applies.

A court should consider what other descriptive characteristics are known, the number of potential suspects, and the intrusiveness of the police investigation. Possibly, the court should distinguish between benign forms of investigation (for example, a written request that a person speak with the police) and a physical confrontation or forcible stop. The disquiet that Brown produces results from the artificial doctrinal lines that the Supreme Court has drawn around the Fourth and Fourteenth Amendments.

Remedies

Legislation and the courts have limited some of the traditional legal remedies in racial profiling cases.

Suppression of evidence. The exclusionary rule provides a remedy for some individual victims of illegal stops but does not provide systemic relief from racial profiling and random stop-and-frisk practices. Racially discriminatory stops are not subject to the Fourth Amendment exclusionary rule in selective prosecution cases, and the Supreme Court has yet to recognize a suppression remedy under the Fourteenth Amendment. Moreover, since most people Subjected to these practices are not arrested, the matter never reaches criminal adjudication.

A notable exception is State v. Soto, the opening wedge in the New Jersey Turnpike racial profiling scandal. (28) Invoking liberal state discovery procedures in lawsuits alleging selective prosecution, the African-American defendants moved to suppress evidence that police obtained in a search of their vehicle. They presented evidence that African-Americans were being stopped at rates that far exceeded their use of the Turnpike. Police documents and statistical studies were sufficient to prove an extraordinary pattern of racial profiling, and the trial court suppressed the evidence in the cases.

Recently, a Pennsylvania appellate court sustained a suppression order in a criminal case in which it appeared that the police relied on the driver's race as a principal reason to stop a car. As the court stated, "driving while black" is not a violation of the motor vehicle code. (29)

The New Jersey Supreme Court has ruled that there can be no valid consent to a vehicle search after a traffic stop unless the police officer has reasonable grounds to believe that contraband or weapons are in the car. (30)

Civil damages actions. The Civil Rights Act--also known as [section] 1983--allows damages for harm caused by state or local officials' unconstitutional stops and searches. Claims against federal defendants (usually regarding border, customs, and airport searches) can be made under the Constitution as so-called Bivens actions or under the Federal Tort Claims Act. (31) Case law and commentary on [section] 1983 and Bivens actions are extensive. (32)

Briefly, damages actions can provide redress to those unlawfully stopped or searched, and they have the potential to trigger institutional reforms. When unconstitutional practices affect many people, as racial profiling and random stops and searches do, potential damages are substantial, and government agencies may seek to avoid large judgments by changing practices and policies. Proof of racial bias has led to significant damages in police misconduct cases. (33) In the highway racial profiling cases, significant settlements (often joined with affirmative relief by way of consent decree) have been effective in changing practices. (34)

Discovery rules broader than state or federal criminal discovery can be used to develop proof of patterns and practice in racial profiling. Section 1983 liability depends on proof that an individual officer violated a plaintiffs constitutional rights. In the racial profiling context, this requires evidence of intentional discrimination (often accompanied by Fourth Amendment violations regarding the stop and/or search). Qualified immunity may be introduced as a defense, but only where the officer reasonably believed that he or she was acting lawfully. Once intentional discrimination is proved, this defense is highly questionable. Claims against supervisors or municipalities are sustainable only if the constitutional violation was caused by a municipal law, policy, practice, or custom. (35)

Even if no contraband is found and there is a strong claim for damages, the amount may be too modest to justify full-scale litigation. The prospects change radically, however, if claims can be joined, and particularly if a class action can be built. Drug interdiction and related car and pedestrian stops may involve subjecting thousands of people to unlawful practices. The case for liability becomes exponentially stronger with the supportive allegations and testimony of hundreds or thousands of victims that the government has overreached, and the damages claims become substantial enough to sustain federal litigation.

Injunctive relief. Federal injunctive relief against state or local officers who violated federal constitutional or statutory rights is potentially the most effective remedy for preventing rights violations in the future. However, federal courts' powers are circumscribed by federalism, comity, and standing doctrines. (36)

In City of Los Angeles v. Lyons, the African-American plaintiff was stopped by city police officers for a traffic violation. (37) After Lyons got out of his car, the officers drew their guns, ordered him to place his hands on top of his head, and, without provocation, placed him in a chokehold, damaging his larynx. The Supreme Court ruled that Lyons had not proved a sufficient likelihood that he would again be subject to a police chokehold to request an injunction.

The Court distinguished Lyons in Friends of the Earth v. Laidlaw Environmental Services, granting federal standing under Article III to plaintiffs in an environmental lawsuit challenging the pollution of a river. (38) The plaintiffs alleged that they had suffered "injury in fact" because they were deterred from using the river for recreational purposes. The Court ruled that there was "nothing `improbable' about the proposition that a company's continuous and pervasive discharge of pollutants into a river would cause nearby residents to curtail their recreational use of that waterway and would subject them to other economic and aesthetic harms." (39)

The Court ruled that Lyons did not bar the action because the unlawful conduct was occurring as the complaint was filed, a holding that confirms the significant difference between allegations of occasional unlawful conduct and practice-and-policy cases. Of course, a polluted river can affect everyone using it, while racial profiling on highways randomly affects thousands in a much larger class of drivers. The question becomes: Once a policy of racial profiling is established, is the probability that minorities will avoid certain highways or that certain members of the minority group will be stopped on those roads sufficient to establish standing?

In Lyons, the Court may have been reluctant to make a definitive ruling on chokeholds (which are permissible in some circumstances). But racial bias in policing cannot be justified, and permitting a challenge to such practices by someone who might not otherwise have standing to challenge police policies would be consistent with the Court's willingness to grant broad standing to challenge racially discriminatory practices. (40) Thus, while Lyons unjustifiably denies standing, it does not close the door to injunctive relief in the context of racial profiling.

Federal intervention 

The Violent Crime Control and Law Enforcement Act of 1994 authorizes the U.S. attorney general to bring civil actions for declaratory or equitable relief against police departments engaged in a pattern or practice of violating constitutional or statutory rights. (41) Since 1994, the Department of Justice has taken legal action against five law enforcement agencies. It has reached consent decrees with Pittsburgh, Pennsylvania; Steubenville, Ohio; and New Jersey (over racial profiling on the turnpike) and a settlement agreement with Montgomery County, Maryland (over racial profiling on 1-95). It has also initiated litigation against the Columbus, Ohio, police department.

The New Jersey agreement prohibits troopers from considering the race, nationality, or ethnicity of turnpike motorists or passengers in determining which vehicles to stop or search, unless those factors were part of a suspect's description. The consent decree requires officers to record data on motor vehicle stops, including the gender and race or ethnicity of the driver and any passenger subjected to a "procedure," as well as the reason for the stop and any subsequent search. Supervisors must review troopers' reports on stops and may also review corresponding audio or video tapes. When a review reveals a possible violation of the decree, nondisciplinary action--such as counseling or more training--may be ordered. Similar provisions are incorporated in the Maryland agreement.

Three developments of the past few years have the potential to alter the legal and political landscape of contemporary policing practices.

First, there is almost universal condemnation of racial profiling. While debate over the exact definition of the term continues, there is growing support for the principle that race cannot be used to support a discretionary judgment by a police officer to stop, frisk, or search someone, except where a description of a suspect, including race, has been provided. Police departments will be pressed to adopt and implement policies prohibiting racial profiling, and courts are likely to intervene where patterns of racial profiling are proved.

Second, for the first time, reliable data are being collected and analyzed with respect to both racial disparities in stops and searches and whether adequate cause existed for the police intrusions. These data are essential to any serious effort to reconsider constitutional doctrine and address remedies for racial profiling and other arbitrary police conduct.

Third, police administrative measures that reflect a more progressive police management philosophy have been adopted. Ultimately, abuses will be significantly reduced only if these fundamental reforms in police policies and practices are fully implemented.

Notes

(1.) Wilson v. Tinicum Township, No. Civ. A. 92-6617, 1993 WL 280205, at *2 (E.D. Pa. July 20, 1993) (opinion on class action certification).

(2.) State v. Soto, 734 A.2d 350, 360 (N.J. Super. Ct. Law Div. 1996).

(3.) PETER VERNIERO & PAUL H. ZOUBEK, INTERIM REPORT OF THE STATE POLICE REVIEW TEAM REGARDING ALLEGATIONS OF RACIAL PROFILING 26-28 (1999), available at www.state.nj.us/lps/intm_419.pdf [hereinafter NEW JERSEY INTERIM REPORT].

(4.) OFFICE OF THE ATTORNEY GENERAL OF NEW YORK STATE, REPORT ON THE NEW YORK CITY POLICE DEPARTMENT'S "STOP & FRISK" PRACTICES 88-89 (1999), available at www.oag. state.ny.us/press/reports/stop_frisk/stop_frisk.html.

(5.) A 1997 government survey found that the rate of illicit drug use for blacks was 7.5 percent; for whites, 6.4 percent; and for Hispanics, 5.9 percent. OFF. OF APPLIED STUD., U.S. DEP'T OF HEALTH & HUMAN SERVS., NATIONAL HOUSEHOLD SURVEY ON DRUG ABUSE (1997).

(6.) Ron Harris, Blacks Feel Brunt of Drug War, L.A. TIMES, Apr. 22, 1990, at A1.

(7.) NEW JERSEY INTERIM REPORT, supra note 3, at 67-68.

(8.) Id. at 26-28.

(9.) Report of John Lamberth, Md. State Conf. of NAACP Branches v. Md. Dep't of State Police, 72 F. Supp. 2d 560 (D. Md. 1999), available at www.aclu.org/court/lamberth.html.

(10.) Ira Glasser, ACLU Biennial Speech (June, 1999), available at www.aclu.org/issues/racial/ bispeech99.html.

(11.) Whren v. United States, 517 U.S. 806, 816-19 (1996).

(12.) Maryland v. Wilson, 519 U.S. 408, 410 (1997).

(13.) Ohio v. Robinette, 519 U.S. 33 (1996).

(14.) Horton v. California, 496 U.S. 128, 133-37 (1990).

(15.) Michigan v. Long, 463 U.S. 1032, 1035 (1983).

(16.) Wyoming v. Houghton, 526 U.S. 295, 302-07 (1999).

(17.) 517 U.S. 806.

(18.) Id. at 813.

(19.) Batson v. Kentucky, 476 U.S. 79, 94-98 (1986).

(20.) 481 U.S. 279 (1987).

(21.) Id. at314.

(22.) Id. at 339 (Brennan, J., dissenting).

(23.) 517 U.S. 456 (1996).

(24.) See Hunter v. Underwood, 471 U.S. 222 (1985) (relying on statistics in finding that a facially neutral disenfranchisement statute was in violation of the Equal Protection Clause of the Fourteenth Amendment).

(25.) Brown v. City of Oneonta, 221 F.3d 329, 337 (2d Cir. 2000); see, e.g., Rodriguez v. California Highway Patrol, 89 F. Supp. 2d 1131 (N.D. Cal. 2000) (allowing equal protection claim based in part on statistical evidence); Md. State Conf. of NAACP Branches, 72 F. Supp. 2d 560 (noting that plaintiffs have standing based on continuing practice of racial profiling by police); Nat'l Cong. for Puerto Rican Rights v. City of New York, 191 F.R.D. 52 (S.D.N.Y 1999) (holding that allegation that police stopped and frisked African-American and Latino men based on their race and national origin was sufficient to state equal protection claim).

(26.) 392 U.S. 1 (1968).

(27.) 221 F.3d 329.

(28.) 734 A.2d 350.

(29.) Commonwealth v. Palmer, 751 A.2d 223 (Pa. Super. Ct. 2000).

(30.) State v. Carty, 790 A.2d 903 (N.J. Mar. 4, 2002).

(31.) Bivens v. Six Unknown Fed. Narcotics Agents, 403 U.S. 388 (1971) (recognizing Fourth Amendment cause of action brought directly under the U.S. Constitution).

(32.) See MICHAEL AVERY ET AL., POLICE MISCONDUCT: LAW AND LITIGATION (3d ed., 2002).

(33.) E.g., Price v. Kramer, 200 F.3d 1237 (9th Cir. 1999); Morgan v. Woessner, 997 F.2d 1244 (9th Cir. 1993); Hall v. Ochs, 817 F.2d 920 (1st Cir. 1987).

(34.) See, e.g., Robert Jackson, Eagle County Must Pay for Stopping Motorists, ROCKY MOUNTAIN NEWS, Nov. 10,1995, at A4 (reporting settlement of racial profiling claim based on litigation, Whitfield v. Board of County Comm'rs, 837 F. Supp. 338 (D. Colo. 1993)); Settlement Agreement, Wilson v. Tinicum Township, No. 92-6617 (E.D. Pa. Jan. 19, 1995).

(35.) Board of County Comm'rs v. Brown, 520 U.S. 397 (1997); City of Canton v. Harris, 489 U.S. 378 (1959); Monell v. Dep't of Soc. Servs., 436 U.S. 658 (1928).

(36.) Lewis v. Casey, 518 U.S. 343 (1996); City of Los Angeles v. Lyons, 461 U.S. 95 (1983).

(37.) 461 U.S. 95.

(38.) 528 U.S. 167 (2000).

(39.) Id. at 184.

(40.) See Shaw v. Hunt, 517 U.S. 899, 904-05 (1996); Powers v. Ohio, 499 U.S. 400 (1991).

(41.) 42 U.S.C. [section] 14141 (1995).

David Rudovsky is a senior fellow at the University of Pennsylvania Law School and a founding partner in Kairys, Rudovsky, Epstein & Messing of Philadelphia. Sections of this article were published in the February 2001 issue of the University of Pennsylvania Journal of Constitutional Law.