TESTIMONY OF

PROFESSOR JONATHAN TURLEY

SHAPIRO PROFESSOR OF PUBLIC INTEREST LAW

GEORGE WASHINGTON UNIVERSITY LAW SCHOOL

WASHINGTON, D.C.

BEFORE

THE SUBCOMMITTEE ON AVIATION

COMMITTEE ON TRANSPORTATION & INFRASTRUCTURE

"AVIATION SECURITY AND PASSENGER PROFILING

 AT UNITED STATES AIRPORTS"

FEBRUARY 27, 2002

_______________

I.

INTRODUCTION

Thank you, Mr. Chairman, it is a honor to appear before the Subcommittee on Aviation and its distinguished members to discuss the constitutional and policy implications of the use of passenger profiling at the nation's airports.  My name is Jonathan Turley and I am a law professor at the George Washington University Law School where I hold the J.B. and Maurice C. Shapiro Chair for Public Interest Law.   I know that your time is limited today and, with the consent of the Subcommittee, I would like to submit a longer written statement to augment my oral testimony. 

Profiling is a subject that could easily occupy an entire session of Congress, let alone a single congressional hearing.  There is no subject that more polarizes our population or pulls within its vortex a greater array of complex constitutional and policy questions.  While I bring some relevant experience to this subject,[1] I cannot claim to represent the majority of scholars or even guess at what that consensus might be.[2]  In fact, this subject is so heavily laden with deep societal conflicts that it is often difficult to hold a meaningful conversation about the merits and dangers of profiling.  This is why today's hearing is so important.  It is essential that we move beyond stereotypes and generalizations concerning profiling and engage in a good-faith dialogue over the possible use of this security device.  To put it simply, we need to move beyond our inhibitions in the discussion of race and speak frankly about the value and dangers of passenger profiling during these precarious times.

In the current crisis, we have an enormous asset in a constitutional system that was designed to adapt with changing times and emerging dangers.  The United States Constitution is a document grounded in practicality.  Doctrines like the separation of powers were conceived in a frank appraisal of the natural tendencies of people in power to expand and abuse authority.  Another reflection of that practicality is found in the tests governing the violation of core constitutional rights from the freedom of speech of the First Amendment to the search and seizure protections of the Fourth Amendment to the equal protections of the Fourteenth Amendment.  These rights are consistently interpreted as requiring some level of balancing of interests.  Whether the conflict is between an individual and the state or between two citizens, the Constitution often requires a consideration of the rivaling interests in a given area of conflict. In the area of search and seizure such balancing is particularly evident in constitutional tests and analysis.  As will be shown, the Supreme Court has often accommodated the interests of the government in areas like borders and airports as well as those circumstances raising issues of exigency or public interest. 

It has been said that the United States Constitution is not a suicide pact.[3]  While protecting core liberties, it is a document that also allows for the accommodation of transient government needs or exigent circumstances.  In this sense, September 11^th did not create a new reality for constitutional analysis.  Rather, it reflected the fact that the state's interests can ebb and flow with periodic dangers and realities.  Nevertheless, there are limitations on such accommodation and, while the Constitution protects the security and survival of the state, it also guards against the arbitrary and abusive use of power in the name of the public interest.

Profiling may be an inevitable response to the dangers evident at airports.  Every month, over 40 million people travel by air in this country.  It is simply practically impossible to closely scrutinize every passenger while maintaining a functioning air travel system.  Obviously, it is theoretically possible to subject every passenger to an in-depth interview and screening process but the costs of such a system would be prohibitive.  As a result, airlines must make choices and select which passengers to subject to closer scrutiny.  There are only two basic systems for such scrutiny:  random selection and profile selection.  Profile selection can itself be divided into two basic forms: computerized passenger prescreening (like the CAPPS system) or observational profiling by security agents.  Both forms of profiling are valuable and can be used in concert.  However, regardless of its form, profiling is only as accurate and reasonable as the criteria used to screen passengers. 

Due to the erroneous inclination of the public to consider "profiling" and "racial profiling" as synonymous, there has been an unwillingness of airlines to implement any formal profiling systems.  Instead, the primary identification process remains the random selection process.  As an academic, I became aware of this system and its inherent inefficiency while traveling not long after September 11^th.  Early one morning, I was rushing to Chicago for a speech.  I was aware that I fit an easy profile for a suspicious passenger.  I had purchased my ticket not long before the flight; I was traveling with no luggage; and I was visibly nervous as I rushed for a flight to a major U.S. city with some of the "high value targets" identified after the September 11th attacks.  In addition, my Italian complexion could easy be construed as Middle Eastern rather than Mediterranean.  I was not surprised, therefore, to see security coming over in my direction.  However, rather than pull me aside, the security stopped an elderly lady next to me.  The elderly lady gradually disassembled as a team of security officers pulled out every one of her many pill bottles, inhaler, and cosmetic items.  While I watched this scene unfold, I asked the supervisor why they stopped the old lady and not the younger, single male with no luggage. The security official responded that she would have probably preferred to search me but that they are strictly prohibited from using any profile.  The old lady was the choice of a random selection process so they would carry out a search on the least likely individual on the airplane to actually be a terrorist.

Such experiences hardly make one feel safer.  Rather, due to our inhibitions, we appear unable to implement systems that offer the reality and not just the appearance of enhanced security. This is not to say that random selection can never isolate a guilty party.  Rather, random selection rolls the dice on security by selecting an insignificantly small number of the 40 million monthly travelers for special scrutiny.  At the beginning of any given monthly cycle, the airlines are facing the same odds as a lottery player in a multi-million-dollar game.[4]  The airlines have tried to increase these odds with the same strategy as multiple ticket purchasers - increasing the number of randomly selected passengers.  This difference, however, does not change the overwhelming odds against "winning." This is not to say that an airline cannot hit the jackpot and land a Mohamed Atta.  It is simply statistically unlikely that the terrorist lottery will yield a positive result.

If we want to have an effective screening system at our airports for terrorism, some form of profiling must be used. As will be shown below, I believe that a well-tailored and balanced system could pass constitutional muster.  Moreover, such a system could actually reduce ethnic and racial incidents at airports by subjecting stops to greater scrutiny and monitoring. The greatest danger to citizens from profile stops is not the brief interruption in their travel but what occurs after the stop in the tenor and time of questioning. Abuses, if they occur, will be found in the use of illegitimate criteria or in the disrespectful or abusive treatment of individuals selected for airport stops.  I would suggest that answering these insular questions would most serve the individual rights of our citizens as well as the collective interests of air travelers.

II.

PROFILING IS A CONCEPT THAT IS DISTINCT AND INHERENTLY AT ODDS WITH THE ABUSIVE USE OF RACIAL PROFILING.

          This hearing would constitute a great leap forward if it only succeeded in establishing an important threshold distinction between profiling as a general concept and racial profiling as a specific (and corrupted) application.  Racial profiling is to the science of profiling as forced confessions are to the art of interrogation.  Like forced confessions, racial profiling achieves only the appearance of effective police work.  Racial profiling often shields or obscures a racial prejudice and unscientific bias against a particular class or group.  It is the antithesis of legitimate profiling in that it elevates stereotypes over statistics in law enforcement.  It is important to understand what is wrong with racial profiling to understand what might be right with airport profiling.

          Racial profiling is something of a misnomer.  A legitimate profile is an aggregation of reliable factors with statistically proven indicators of criminality.  Abusive racial profiles use a single or outcome-determinative criterion for isolating suspects.  When someone judges individuals entirely and superficially on the basis of race, they are making by definition a racist determination.  Dressing up such bias as a scientific calculation is little more than a post-rationalization.  This is not to say that race cannot be a factor in some legitimate profiles, but its placement and use in a profile is extremely precarious and delicate work. All of the profiling abuses that have been documented follow a common pattern where race is the only or predominant factor used to select suspects.  There are very limited circumstances where race is a relevant criterion and only when it is tied to other criteria of selection.  For example, if police are dealing with an increase in crime in a neighborhood by white or African-American gangs, race may be part of a legitimate profile of possible suspects.  Along the Mexican border, race is an obvious criteria in looking for illegal immigrants.  In some opium smuggling circumstances, the government may profile on the basis of race if it finds that gangs primarily use Asian couriers due to the presence of family members in source areas (for possible retaliation).  Finally, some profiles may be generated by intelligence to indicate a specific threat that includes information on race or nationality.

These are not, however, the type of cases that most concern the majority of Americans. Over the years, we have seen profiling used to isolate African-American males under a general assumption that they are more inclined to criminality.  The social and legal costs of this abusive practice have been extraordinarily high.  Part of the reason for this practice is the common reference to a statistical difference between races in terms of criminal activity.  There is in fact a higher percentage of crime by African-American males in comparison to their percentage of the population.[5]  However, this difference does not support the use of racial profiling for two reasons.  First, the higher percentage of African-American males stopped may explain some of the higher percentage of convictions.  Racially based stops can be a self-fulfilling prophesy as the greater number of stops naturally yield a great number of arrests and convictions.  If whites were stopped with greater frequency, the number of arrests would likely rise with the increased detection and enforcement.  Second, even assuming arguendo that race is a statistical indicator of criminality, its use may still be barred as imposing prohibitive social costs that outweigh any social benefits. Racial profiling reinforces stereotypes of criminality; fuels racial divisions; imposes social stigma on individual citizens; creates de facto travel restrictions or barriers; and drains police efforts from alternative and proven method of enforcement.

          Some of these concerns are clearly present in airport profiling but not to the same degree and not to the same effect.  There are both legal and practical distinctions between racial profiling and airport profiling.[6]  The controversial use of racial profiling concerns the elevation of race as the sole or over-riding criteria; to tie a race to a general propensity for crime.  Conversely, airport profiling is tied to specific criminal conduct and a demonstrated threat. There is clearly a greater threat of terrorism coming from certain countries with extreme Islamic and anti-American movements.  Moreover, an airport is a limited and somewhat unique location for profiling.  In racial profiling, it is the race of the suspect and not his conduct that often triggers a stop.  In the most controversial cases, a black driver is stopped on suspicion of "crime" rather than an articulated suspicion of a specific crime.  Airport profiling is tied more closely to the locus and the specific dangers of terrorism.  Part of the social costs to racial profiling is that every African-American male is always at risk of being stopped for the offense of "driving while black."  An airport, however, is a location that is understood to be subject to more stringent security and restrictions.  One's expectations are manifestly different when entering a terminal as opposed to a neighborhood.  Finally, the specific crime involved in airport profiles presents a greater exigency and risk for society than does a typical drug profile stop on the street.  Quite frankly, society can live with a couple of nickel bags of narcotics making it through the law enforcement net.  The costs of a failure to detect criminality at an airport are far more immediate and deadly.  Failures at airports security are not measured by nickel bags but body bags - a reality that cannot be ignored in any final analysis. 

Given our painful history with racial profiling, it is obviously tempting to argue that we can use profiles with the express prohibition on the inclusion of race or nationality or ethnicity as criteria.  However, it is possible to accept that race may, in limited circumstances, be a relevant criteria without endorsing the use of racial profiling in its colloquial sense.[7]  This is particularly the case where intelligence or witnesses indicate the race of possible suspects.  In the same fashion, profiles that include well-founded factors of nationality or ethnicity cannot be categorically rejected as evidence of prejudice or bias. In the case of terrorism, it is a type of crime that can be focused by world events or regional conflicts on particular nationalities or ethnic groups.  Unlike appearance or behavior, nationality and ethnicity are characteristics that are difficult to hide or change.  Conversely, because they are immutable characteristics, we have to be cautious that individuals with these characteristics are not converted into walking profiles subject to endless stops and searches.  This is why the combination of criteria and the training of agents for observational profiling is so essential.  It is to those issues that I would now like to turn.

III.

PROFILING CAN BE A CONSTITUTIONAL AND APPROPRIATE MEANS TO DETER TERRORISM AT UNITED STATES AIRPORTS.

          As stated earlier, I believe that passenger profiling would be constitutional - even with the inclusion of nationality or ethnicity - if crafted to fit existing standards under the Fourth and Fourteenth Amendments. There are four principle lines of cases that are most relevant to the issue of profiling of passengers at airports. First, there is a long line of cases in which profile stops that been specifically reviewed by the Supreme Court and lower courts.[8] The Supreme Court has never found racial profiling, let alone general profiling, to be unconstitutional as a threshold matter.  However, it is clear that racial profiling can be a violation of the Fourteenth Amendment's guarantee of equal protection as well as a violation of the Fourth Amendment's search and seizure protections in certain circumstances.  Race alone is not sufficient as a general matter to establish reasonable suspicion for stop.[9]  It seems likely, in my view, that a majority of the Court would accept race or nationality as a factor on a profile where the government could show a statistical justification or intelligence basis for the selection.

          Second, the Supreme Court has allowed law enforcement to stop individuals for questioning where their conduct or appearance creates a reasonable basis for inquiry. In Terry v. Ohio,[10] the Supreme Court found that an investigatory stop and patdown was constitutionally permissible when supported by "reasonable suspicion," a standard below that of the Fourth Amendment's probable cause standard.  This standard is tied to the totality of the circumstances and the use of an officer's reasonable judgment.[11] However, it is clear that officers cannot detain someone longer than reasonably necessary.[12] Accordingly, at airports, Terry stops can be used where a reasonable officer would view an individual as suspicious due to their conduct or appearance.  Any profile stops would be limited in time and character along the lines of a Terry stop.

          Third, there is a line of cases recognizing that the government must be given greater discretion and flexibility at airports as the functional equivalents of borders.[13]  Not only is the expectation of privacy less at an airport under the Katz v. United States[14] test, but airports are a primary site for profile stops by government agents working with Customs, the DEA, and the INS.  If a profile including nationality or ethnicity is to be upheld, it is most likely that it would be upheld at airports where the government's interests are at their apex.

Fourth, and finally, the Court has allowed stops to occur for the most minor violations - even when such reasons are merely a pretext.  In Whren v. United States, the Supreme Court ruled that the police can use pretextual stops.[15]  While the Court noted that the use of pretextual stops based on race could violate the Fourteenth Amendment, the Court stated that it would not delve into the subjective intent of an officer who can cite a valid pretext for a stop.[16] One interesting limitation on Whren is the Court's subsequent rejection of the use of pretextual roadblocks in City of Indianapolis v. Edmond.[17]  Ironically, profiling seeks to avoid the problem identified in Edmond.  Airport security does not want to use a pretext to stop every passenger but to remove an identifiable subgroup of travelers from the much larger population.  Pretextual stops are more relevant outside of the airports where simple violations of motor vehicles laws supply easy pretexts for law enforcement.[18]  Nevertheless, Whren would indicate that law enforcement can use minor violations as an alternative justification for investigatory stops at airports.

While there is much more that can be gleamed from these cases in terms of guidelines, these cases suggest that a profile on terrorism could be upheld, even with the inclusion of nationality or ethnic factors.  Such profiles would have to contain an array of factors beyond nationality or ethnic factors such as conduct, appearance, and independent intelligence factors.  The multiple criteria is particularly important for general profiling.  Few would seriously contest that greater importance can be given to nationality or ethnicity in a target profile where an individual or group is being specifically sought by the government. While we may call such an "information circular" a "description" of suspects as opposed to a "profile," it is often little more than an identification of likely suspect characteristics.  If a threat is viewed as imminent, there may be little for the government to use beyond nationality and a general profile of a likely terrorist's age and gender. Thus, if the government receives intelligence that a terrorist cell in Sudan is about to attack an airplane, it is particularly reasonable for the government to temporarily subject all Sudanese nationals and Sudanese flights to added scrutiny.  The greatest concerns arise in a general profile based on statistical analysis as opposed to a targeted profile based on specific intelligence. It is in these general profiles that we need to direct our greatest attention to define the proper use of such criteria and to impose some reliable form of monitoring.

IV.

CONGRESS SHOULD DEVELOP GUIDELINES AND MONITORING SYSTEMS FOR PASSENGER PROFILING AT UNITED STATES AIRPORTS.

It is clear that profiling at airports will occur informally regardless of whether it is adopted formally.  Cases like Terry allow police officers to use their reasonable judgment as to suspicious conduct or appearance.  An officer may describe this selection as based on a "hunch" supported by objective criteria, but it is often an informal profile.  Ultimately, a profile is the aggregation of the experience of hundreds or thousands of officers into a single list of common criteria.  Basic profiles of high-risk travelers (regardless of nationality) are obvious to any booking agent.  A passenger traveling without luggage, with a one-way ticket purchased the day of the flight will raise a series of "flags." Such profiles can be used either as part of a computerized system like the computerized passenger prescreening system (CAPPS) or as a secondary observational system by booking and security agents.  The key in both computerized and observational systems is obviously the use of true objective criteria.  It is obviously easier to scrutinize, control, and monitor a CAPPS system as opposed to stops triggered by visual observation.

The benefit of a formal profiling system is that we can control the criteria and better monitor their use.  One of the greatest corruptions of profiling is the use of "objective" criteria that cover every possible conduct or behavior as well as its inverse.  Thus, past controversial profiles have shown such inversive criteria as avoiding eye contact as well as making eye contact.[19]  This can create the false appearance of scientific objectivity when the cited criteria merely rationalize a stop based on bias.[20]  Accordingly, the United States Customs Service changed its 1999 Personal Search Handbook to eliminate over forty (often conflicting criteria) with six criteria relevant to customs violations.[21]

Terrorism involving airplanes is a relatively narrow category of crime that avoids the past problems associated with generalized and race-based theories of criminality.  Even critics of past racial profiling have accepted that some limited nationality or racial factoring is permissible when there is greater specificity as to the crime and threat.  Thus, academics have accepted as valid profiles that focus on limited categories of criminality, distinctive patterns of ethnic or racial offenders, specific locations, and patterns of conduct.[22]  Many of these criteria mirror successful profiles used in cases like United States v. Sokolow.[23]

Obviously, many of the most obvious profile criteria do not involve nationality or ethnicity. Richard Reid is an example of the obvious value of a general profile in the isolation of particular travelers.  Reid was identified by a profile at the Charles de Gaulle airport. Security officials with ICTS, a subcontractor at the American Airlines counter, identified Reid because of (1) lack of a verifiable address; (2) a one-way ticket; (3) no clear travel plan; (4) no luggage; and (5) use of cash to buy his ticket.[24]  This led to a stop that led to Reid missing his first flight - and quite possibly a successful attempt at terrorism.  The failure of the French authorities to completely search Reid, including his shoes, was a case in which a profile was successfully used but undermined by poor security procedures.[25]  On December 11, 2001, the FAA issued an "information circular" that specifically stressed that shoes should be inspected for possible weapons or devices.[26] 

As noted earlier, the value of nationality or ethnicity is that they are characteristics that are more permanent.  While future terrorists are likely to take care to learn from Reid's experience and follow the profile of a tourist rather than a terrorist, nationality or ethnicity are more difficult to disguise.  The danger is that these criteria would become the only criteria or outcome-determinative criteria.  The greatest danger is that nationality or ethnicity would be the sole criteria used by an agent, who will simply add "nervous in appearance" as the other reported criteria. Moreover, care must be taken to avoid the simple adoption of any criteria that have a greater than average yield.  For example, all of the terrorists on September 11^th were Muslim and the terrorist threat against the United States is closely connected to Islamic fanaticism.  However, use of Muslim faith as a criterion would be problematic on a number of levels. First, there are over 1.6 billion Muslims in the world and only a small number of such individuals are terrorists.[27]  With one out of four humans a Muslim, the population group is too high to yield a meaningful statistical value to a profile.  Second, the social costs of such a criterion heavily outweigh its value.  Questioning individuals on religion creates a chilling effect, particularly with a religious group that is already confronting growing prejudice.  Third, there is no easy way on travel documents to easily isolate faith and only some Muslims wear clothing that indicates their faith.  This creates the added danger of misidentification by security attempting to use the criterion.  Fourth, this is a criterion that any terrorist could easily hide or deny.  As a result, law-abiding Muslims would occupy security officials and the criteria would serve as a distraction for security.  This does not mean that an individual who has been stopped on legitimate criteria cannot be legitimately questioned about their religious views as part of a later interrogation.  However, such questioning should only occur after an initial stop has revealed additional basis for suspicion.

          Congress can help implement a system of monitoring to identify abuses or statistically unsound profiling. [28]  This system should include easily available reporting mechanisms at airports for abuses or complaints, mandatory reporting of complaints, expedited review of complaints, active oversight from Congress, and the assignment of supervisory personnel.  Perhaps the most important protection is the continual review of criteria used for airport investigatory stops.  Given the shifting intelligence on threats, a passenger profile is likely to evolve more than a drug or customs profile.  It is important for the government to be monitored to keep profile criteria updated and relevant to the current threat.  Moreover, the greatest danger to citizens comes not in the brief investigatory stop but in their treatment after the stop.  Congress is essential in guaranteeing that stops are limited in time and conducted in a respectful and non-threatening fashion by law enforcement.  Finally, a bi-annual report on the number and characteristics of passengers stopped should be mandated.  Unlike the current system where profiling likely occurs on an informal basis, this reporting would give us our first comprehensive tracking system for such stops.  While this report would have to be classified to some extent (including the current profile criteria), much of it could be made public to assure citizens that "flying while Arab" is not a new addition to our national lexicon.

V.

CONCLUSION

It is the continual burden of a free and democratic society that we must balance our needs as individuals with the interests of our nation in times of peril.  It is far easier to address such questions when only the security of the nation is a concern.  We, however, chose a more difficult path when we formed this representative democracy.  Profiling is and should be a controversial step in the war against terrorism.  We should implement such a system with the greatest caution and reluctance.  In the best of times, I would oppose profiling as a security device. However, in my view, these are difficult times that require difficult choices.  Complaining about the need for added security is like complaining about the weather.  The risks at airports are real and present.  We must respond to those risks.  I believe that we can do so in a measured and constitutional fashion.  I admit that I will not likely shoulder the burden of such precautions at airports.  We owe it to our fellow citizens who will be inconvenienced to demonstrate that these stops are based on articulated and well-founded criteria.  Our commitment to their civil rights will be measured by the resources that we commit to the monitoring and investigation of any abuses in this system.

There is obviously much on this subject that I have not addressed in the limited time and space available.  I hope, however, that this testimony will give the Subcommittee a useful outline of the issues to help grapple with this difficult question.  To that end, I am available to address any later questions that may arise on this subject.

I would be happy to answer any questions that the Subcommittee might have at this time.

_______________