RACIAL PROFILING AS A PREEMPTIVE SECURITY MEASURE AFTER SEPTEMBER 11: A SUGGESTED FRAMEWORK FOR ANALYSIS K. SHIEK PAL

RACIAL PROFILING AS A PREEMPTIVE SECURITY MEASURE AFTER SEPTEMBER 11: A SUGGESTED FRAMEWORK FOR ANALYSIS K. SHIEK PAL

Abstract

The events of September 11, 2001, marked the beginning of the United States’ renewed war on terror, and introduced racial and ethnic profiling as a tool in that struggle. This paper reviews the body of law and policy concerning profiling prior to and after September 11, and through statistical analysis the impact of preemptive profiling of individuals “of Middle Eastern appearance” to both society and those affected. The author draws attention to the collateral costs associated with broad population screening, and the disproportionate burden borne by the targeted group. Several courses of possible refinement and future research are developed.

Introduction September 11, 2001, marked a tragic day in the nation’s history and ended an era of American innocence and isolation. Americans were no longer absolutely secure in the belief that they were safe within their own borders, and similarly many Americans were no longer secure in the belief that race or ethnic origin would not deprive them of the same civil liberties afforded other Americans. The casualties of September 11 included a massive loss of innocent life, financial loses in the billions, and also the loss of the bedrock American value that race or eth- nicity should not be a factor in predicting criminal behavior. This development undermines an extensive body of law and public policy enacted over decades that delineated the appropriate consideration of race in such areas as jury selection, profiling, and internment.1

The question of racial and ethnic profiling after September 11 focuses on two new issues: the preemptive use of profiling as a prophylactic against terrorism and the identification of the so-called “of Middle Eastern appearance” (OMEA) group as the primary target of such profil- ing.2 The OMEA group is broadly defined to include any individuals whose physical charac- teristics could plausibly serve as a proxy for Arab, Middle Eastern, or Muslim affiliations.3 This proxy is theoretically used to identify threats against the United States emanating from extrem- ist Islamic terrorists and is premised on the idea that all prospective terrorists representing this threat will fall into this category.4 However, the use of preemptive profiling is a dangerous proposition because it carries inherent costs that must be evaluated against the potential bene- fits that would be derived from a foiled terrorist plot. These costs include the proportion of OMEA citizens falsely identified as threats and the associated harms of social stigma and fur- ther deprivation of civil liberties through detention, intensive interrogation, and invasive searches. This paper assesses these social costs as a means of evaluating the appropriate role of profiling under the current federal guidelines.

The terrorist attacks in New York and Washington on 11 September 2001 have resulted in both an intensification of law enforcement activities and the loss of some freedoms for all Americans, but especially those commonly dubbed “of Middle Eastern appearance.” The question of whether individual freedoms subserve the greater good, particularly in the United States, remains difficult. K. Shiek Pal presents a valuable framework for considering this question and for guiding further analysis. —Jules M. Delaune, editor

Kennedy School Review120

Preemption Preemption is controversial because it involves morally ambiguous questions of acceptable loss and calculated risk. However, in a global society increasingly susceptible to nontradition- al forms of devastating attacks oriented at amassing massive casualties, preemption offers an inevitable and effective method of protection. Categorically discounting preemption compro- mises a state’s ability to defend its citizens against enemies who do not adhere to similarly defined boundaries, and therefore preemption appears to be an inevitable future policy deci- sion. The distinction between preemption and prevention turns on immediacy—preemption employs various tools to terminate a perceived threat prior to its effectuation, whereas preven- tion takes a less aggressive and more protracted approach to create an environment where such threats are less likely. Preemption becomes necessary because prevention is not a viable option in situations where the harm is imminent and there is insufficient time to implement a more comprehensive response.5 The tools of preemption are more controversial than those of preven- tion because of the greater associated risks and costs. One such preemptive tool employed by authorities in the war against terrorism is racial profiling.

Preemption through Racial Profiling The concept of racial profiling—attributing a predisposition for certain behavior to an individ- ual’s racial or ethnic heritage—arouses impassioned debate and moral outrage because it nec- essarily implicates a belief that race determines conduct in some discernable part.6 This notion fundamentally contradicts cherished bedrock values of equality in the American democratic system, and the Associated Press quoted President Bill Clinton’s description of profiling as “a morally indefensible, deeply corrosive practice.” However, in a world where any incremental advantage in winning a war against terrorism must be considered, there must be room to incor- porate the benefits of preemptive profiling if it sufficiently advances the cause of security. The controversy arises when entire classes of people are treated as suspects based merely and sole- ly on their racial or ethnic heritage. This type of race-based assumption harkens back to peri- ods of American history when discriminatory laws and policies resulted in a stratified citizen- ship wherein people were denied rights and liberties because of the color of their skin. Considering the costs borne by the targeted group on behalf of the overall society is important in evaluating the actual preemptive mechanism used to screen the group because of the legal, social, and philosophical principles that argue for equal treatment between races, with limited narrow exceptions.

Defining the Problem Proponents of profiling as a means of preempting terrorism point to the demographic common- alities of the September 11 hijackers, who were uniformly Muslim males of a certain age and primarily of Saudi origin.7 The basic argument of these proponents is that, given these com- monalities, it is ineffective to allocate scarce resources towards screening passengers outside this defined group, particularly when the process of randomized searches results in very young children or elderly women being selected for such scrutiny.8 By noting that a broadly dissemi- nated threat is emanating from a specific “identifiable” group (i.e., Muslim terrorists), those in favor of such profiling distinguish the measures they endorse from a mere assumption of crim- inality based on race.9 While it is true that the threat effectuated on September 11 can be attrib- uted to a group of men sharing some identifiable characteristics, it is equally true that there are also additional, less easily discernable factors that distinguish this particular subset of Muslim men from all those that would be captured under a more targeted screening process focusing on the general profile (young Muslim men of Middle Eastern appearance.) These additional

Kennedy School Review 121

factors include religious extremism, political ideology, and adherence to a socioreligious cul- tural war known as a jihad.10 Thus the fundamental problem with a preemptive profiling sys- tem using the OMEA standard is that this does not reflect these salient additional factors and therefore improperly captures far more innocent people than intended.

DOJ Guidelines on Racial Profiling The U.S. Department of Justice has issued guidelines on racial profiling that prevent the use of race in routine domestic investigative procedures but allow for particularized suspect descrip- tions incorporating racial characteristics.11 The guidelines state that “general enforcement responsibilities should be carried out without any regard to race or ethnicity,”12 and that “stereo- typing certain races as having a greater propensity to commit crimes is absolutely prohibited.”13

These provisions guard against the most common notions of racial profiling, such as discrim- inatory traffic stops or targeted searches of minorities. However, these guidelines do not com- pletely constrain the use of racial/ethnic profiling in combating terrorism insofar as there is substantial latitude for national security applications.14 Specifically, the guidelines use contra- dictory language and examples in delineating the scope of permissible racial profiling in the war on terror. Compare the following excerpts from the guidelines:

In investigating or preventing threats to national security or other cata- strophic events (including the performance of duties related to air trans- portation security), or in enforcing laws protecting the integrity of the nation’s borders, federal law enforcement officers may not consider race or ethnicity except to the extent permitted by the Constitution and laws of the United States.15

Given the incalculably high stakes involved in such investigations, federal law enforcement officers who are protecting national security or prevent- ing catastrophic events (as well as airport security screeners) may consid- er race, ethnicity, alienage, and other relevant factors.16

Because terrorist organizations might aim to engage in unexpected acts of catastrophic violence in any available part of the country (indeed, in mul- tiple places simultaneously, if possible), there can be no expectation that the information must be specific to a particular locale or even to a particu- lar identified scheme.17

This latter excerpt carves out an exceptionally broad set of circumstances wherein the precau- tions enunciated in the first except do not seem to apply. The elasticity in defining national security exceptions to the general prohibition against racial profiling creates an ambiguous environment without clear delineations of appropriate and inappropriate behavior. The danger identified by the Bush administration, and that which the administration sought to foreclose by issuing these guidelines, is that the unchecked use of racial profiling would completely desta- bilize the notions of transparent authority, accountability, and equality for many racial and eth- nic minorities.18 However, the lack of precise language in determining the extent to which such profiling may be applied in certain limited (but again, not clearly defined) situations results in an ongoing threat of racial profiling being used in an abusive manner against targeted groups.

The inherent inconsistency in the terms and conditions presented in these guidelines poses a problem for law enforcement agents, civil rights activists, legal theorists, and targeted minori- ties because it allows virtually unregulated profiling to occur under the exception for national security. The weakness of the guidelines and the glaring loopholes contribute to the opaque nature of the discussion of preemptive profiling. Evaluating the social costs of profiling and

Kennedy School Review122

reconsidering the guidelines through the lens of those bearing these costs allow for a more comprehensive analysis of what measures are needed to effectively implement any form of racial consideration.

The Costs of Racial Profiling . . . and the Costs of Not Racially Profiling In evaluating the costs and benefits of preemptive racial profiling in a security context, a com- prehensive understanding of each of the potential outcomes for every person identified as being part of the suspect class and subjected to profiling is essential. An analysis of the probabilities of each outcome given certain factors provides a basis to evaluate whether the costs outweigh the benefits. But as discussed below, there is ample room for subjective weighting of the vari- ous outcomes, and therefore a clear quantifiable answer may not be feasible. The process by which the weighting takes place may ultimately be more instructive.

In considering the probabilities of how preemptive racial profiling will affect both securi- ty and civil rights, two factors must be incorporated into the analysis—the accuracy of the screen used in the profiling process, and the actual correlation between race and criminality (i.e., the percentage of people in the profiled group who are actually terrorists).19 Each factor has two possible outcomes, positive and negative. Assume arguendo an extremely exaggerat- ed hypothetical case wherein one out of every ten thousand Middle Eastern Arabs is actually a terrorist (0.0001 percent) and that an extraordinarily effective screening mechanism is devel- oped (incorporating racial profiling) that is 98 percent accurate;20 multiplying out the possibil- ities produces four potential outcomes for each person subjected to the screen, as displayed in Table 1.

These results indicate that 98 percent of actual terrorists will be identified as such, and that only 2 percent of innocent Arab passengers will be falsely identified as terrorists. The respec- tive probabilities of falling into each category are displayed in Table 2.

The correlation numbers used are extremely exaggerated given the overall global popula- tion of people meeting the basic OMEA definition and the relatively small subgroup of terror- ist extremists. But as demonstrated in Table 2, these numbers still produce an extremely high number of false positives—ten thousand false positives for every false negative.

The Trade-Off in Errors—Designing a Better System The purpose of this type of analysis is to measure the accuracy of a screening system as depict- ed by the relative proportions of the type I and type II errors generated by it. The practical com- parison is between false positives (type I errors) and true positives, i.e., how many innocent people are caught up in the additional security measures for each true terrorist actually appre- hended. But the comparison that best illuminates the possibilities for improving the system is between the false positives and false negatives (type II errors). Generally, American society prefers false negatives to false positives.21 But this preference cannot be sustained in the con- text of terrorism because of the devastating consequences of events such as September 11. Therefore finding the appropriate balance between the two types of errors is essential.

Given the criminal disposition of the terrorist, there are no associated harms inherent in the true positive category—the threat is successfully identified and controlled. However, there are substantial associated harms with the false negatives—those terrorists that slip through the sys- tem and ultimately engage in criminal behavior with massive casualties. Thus an appropriate comparison might be between the two groups with the greatest associated harms—the harms borne by the false positives in order to minimize the potential harms caused by the false nega- tives. Thus the crux of the deliberation is the balance between false positives and false nega-

Kennedy School Review 123

tives, and how each of these outcomes is weighed. The remainder of this paper focuses on how to consider and derive this balance.

The Associated Costs and Effects—What Matters More?

One false negative is not worth the same amount as one false positive because the false nega- tive left unto itself will lead to massive casualties, whereas the false positive may lead to a weakened state of support or security among the wider racial/ethnic pool from which the false positive is drawn but will not produce greater casualties in and of itself. In evaluating false pos- itives, a distinction must be made between casualties and other harm, so that the erroneous assumption that false positives create no harm is not perpetuated. Rather the harm to false pos- itives must be carefully considered and balanced as a cost against the benefits to be derived from minimizing the false negatives.

False positives (innocent travelers who are identified as risks) can face a range of conse- quences, depending on the severity of the misidentification, contextual security concerns, and length of time required to rectify the false identification. However, even the most onerous and severe of these outcomes cannot rationally be equated to the loss of life resulting from a false negative. At the same time, these costs can be extremely painful and burdensome for those who have to bear them and cannot be minimized.

But substantial false positives will necessarily accompany each true positive and any false negatives. Recalibrating the screening process to reduce the false negatives correspondingly increases the number of false positives. Finding the appropriate and ideal balance between the two extremes is the challenge in devising such a system. Further complicating this calibration is the political impossibility of acknowledging the necessity of incurring any false positives or false negatives to each respective constituency, and therefore this is a calculation that must be sheltered from public scrutiny because the reality of each outcome is socially repugnant.

But the overwhelming numbers of false positives bear the costs of such a system, therefore these costs need to be carefully evaluated. The false positives here are the innocent travelers who are eyed with suspicion, or perhaps removed from a flight because their ethnic appearance causes distress among other “non-OMEA” passengers, such as reported by the IndUS newswire. In a more extreme case, a false positive can be placed on a no-fly list or detained for a lengthy period due to mistaken identity, irregularities in immigration status, or outstanding questions about social connections or affiliations.22 For the false positives, the intensity of their experience is compounded by the lack of distinction between true and false positives in the eyes of those witnessing the screening process. These are just some of the costs of a preemp- tive system of racial profiling—others are less tangible and include the effects on the psyche, security, and comfort of American citizens who fall, rightly or wrongly, into the OMEA pool. These are the costs that need to be examined and weighed.

A false negative (an actual terrorist that is not identified as such) carries the most substan- tial harm because a consummated terrorist act could result in the loss of hundreds or thousands of innocent lives. The impossibility of truly quantifying the impact of such losses makes poten- tial false negatives devastating to society and makes the overall tolerance for such false nega- tives nearly zero, irrespective of the actual number of casualties for a given incident. The pur- pose in considering, let alone implementing, a system of preemptive racial profiling is precise- ly to address the heightened intolerance for such casualties. In the aftermath of September 11, there was popular rhetoric arguing that in a time of war, citizens should be willing to sacrifice their own civil liberties for the sake of national security, but interestingly, this rhetoric was not being generated by the same segments of society that would have to bear the requisite costs or make those sacrifices.23

Therefore the balancing must turn on properly valuing the false positives. Given the pre-

Kennedy School Review124

mium placed on saving innocent lives, there is a legitimate question as to whether there are any limits on what an acceptable cost would be for a mechanism to prevent such losses. However, the system is not fail-proof, and therefore it seems logical that there should be some cap on the cost for a system that is less than 100 percent effective. It is even more so when the costs are borne disproportionately by one constituency although the benefits are shared by all. Thus, the key consideration in evaluating preemptive profiling is how to weigh the false positives because that is where the social costs of the program will be borne.

The effects of a false positive identification carry associated costs that have prolonged effects. These costs range from the pervasive awareness of a being in an inferior tier of citizen- ship, to a loss of faith in the security and identity of an America where equality is not predicat- ed on race, and a lingering bitterness that corrupts the integration of different races and ethnic- ities into a unified society. This type of long-term effect magnifies the weight that needs to be attributed to the costs of false positives. An apt comparison might be to death row prisoners who are acquitted before execution—they still bear the costs of time served, emotional trauma, social stigma, and resentment against the system. These costs are not alleviated by the eventu- al acquittal.

Further complicating the current situation is the fact that the compromised civil rights are not limited to one isolated individual, but rather an entire class of people broadly (and vague- ly) defined to be “of Middle Eastern appearance,” which in practice encompasses not only those of Arab descent, but also South Asians, Mediterranean peoples, some Latinos, and gen- erally “brown” people.24 The fundamental problem with the OMEA description is that the par- ticularized threat against America comes not from a narrow ethnic group, but rather an extreme subset of one of the world’s major religions—one that has a sizable following in this country.25

The OMEA description is not well designed to serve in a future prophylactic capacity but rather is premised entirely upon some selected characteristics of the group of terrorists who success- fully attacked America on 11 September 2001.26 The danger of relying on this description is that the source of the threat is broader than that narrow group, and as previously noted, the description itself is vague and overbroad. Thus, the OMEA description is simultaneously over- broad and too narrow by failing to capture Muslim extremists from other ethnicities but includ- ing non-Muslim, non-Arab, non-extremist people, as well as the vast majority of peaceful, law- abiding Arab Muslims.

Our society has historically rejected state conduct that broadly targets an entire class of people based on their appearance, but here that appearance seems to have a stronger predictive value (although still miniscule) than in other contexts.27 Similar arguments and considerations have previously been made in support of racial profiling in other security contexts, for exam- ple, in drug trafficking or urban street crime.28 Courts have generally disallowed or severely limited the use of such tactics because of the lack of evidence of a reliable predictive value based on race or ethnicity, as weighed against the substantial intrusion upon civil liberties embodied by these types of policing measures.29 However, the use of racial profiling to prevent terrorism may be distinguishable from these other cases because of the ethnopolitical and reli- gious foundations of the threat against America.30 As previously discussed, the more nuanced argument considers the use of race or ethnicity in the terrorism context as an example of “pre- emptive suspect description” based on an articulated threat rather than traditional profiling. However, the DOJ guidelines on racial profiling outline a slightly more stringent standard for this approach by noting that there must be an explicit, temporally limited threat in order to look for a suspect.31

But even within the hypothetical situation outlined in the guidelines, there is sufficient ambiguity to complicate the use of profiling in a controlled and limited manner. For example, there are no restrictions on how specific information must be or how imminent a threat must be before profiling can be appropriately used. The guidelines cite an example of a credible

Kennedy School Review 125

threat against an airline from a terrorist organization in “the next week” and use this threat to justify an undefined period within which the strict prohibitions on racial profiling can be bent or suspended.32 This is a critical point to understand the difficulties of instituting such a pro- gram because the ambiguity of the terms combined with the elasticity of the loophole create a situation where civil rights can be substantially diverted for an extended period without prop- er oversight or accountability—the only requirement is a “credible threat” against national security, and in the contemporary global war on terror, such threats are frequent.

Given the historical burdens of racial prejudice and discrimination in this country, there should be a clearly compelling state reason to employ profiling tactics, and those tactics should be narrowly constructed so as to minimize the costs borne by the profiled group in terms of the compromise that members of that group must make in their civil liberties. The DOJ guidelines meet the compelling reason by noting that a credible threat is required, but by failing to pre- cisely outline the minimal thresholds, the guidelines are not narrowly tailored enough to be acceptable. There are no limits within the guidelines that would restrict the use of profiling if even a minimal argument can be mounted, as evidenced in the previous example. This results in undue costs being borne by the target group, which in this case would be the OMEA group— and more specifically the false positives within the OMEA group.

Balancing the Costs of Racial Profiling Part of these costs on OMEA passengers come in the insecurity of knowing that there are no clear parameters for the circumstances under which profiling can be instituted. This under- mines the ability of OMEA citizens to be secure in their faith in the protections of the system because the decisions as to when the conditions are appropriate for profiling are somewhat arbitrary. This insecurity is a substantial cost because it is only borne by the targeted group, as opposed to all citizens broadly or those engaging in illicit conduct. The costs of profiling also bear a linear relationship to the effectiveness of the profiling system, insofar as the most min- imal forms of profiling may not make a discernible difference to the overall effectiveness of the security process, whereas greater degrees of profiling may yield substantial results. Thus the challenge is to determine the equilibrium point between the costs and the effectiveness of profiling.

Balancing more onerous forms of profiling against the benefits of apprehending a terrorist implicate the mathematical calculations of the likelihood of successfully achieving a true pos- itive. Here the probability analysis indicates a far greater likelihood of yielding false positives than true positives, which means that in sheer numbers there will be far more innocent OMEA members subjugated to the higher scrutiny (and consequently greater harm) than there will be actual terrorists apprehended. However, it also seems probable that these measures will be much more valuable in preventing or impeding terrorists, and offer greater payouts from a util- itarian perspective. Thus in order to implement a screening system that is even moderately effective, it seems clear that the costs borne by the OMEA group will be at least at the inter- mediate level discussed above, if not greater. Any system that requires a lower level of intru- sion will not be effective.

The burden placed on all OMEA members will need to be balanced against the benefits of preventing even a single terrorist activity. Because of the difficulty in quantifying the value of a life as compared to varying degrees of harm associated with the practical costs of racial pro- filing, there is not a clear or easy answer as to whether the potential benefits justify these costs. Theoretically, there can be another level of analysis that attempts to quantify not just the value of a saved life against the infringement of civil rights, but seeks to determine how many lives must be saved in order to impose the highest level of scrutiny. It may be difficult to argue that saving one solitary life justifies the costs borne by an entire group of citizens, but it is equally

Kennedy School Review126

difficult to enunciate a rationale by which to measure how many lives are necessary to balance that equation.

On the other hand, if an innocent life is lost in a terrorist act, it is extremely difficult to explain to the victim’s family that there were additional screening measures which could have been taken and that might have reduced the likelihood of that act by 0.0001 percent, but that those measures were not taken because the costs imposed upon the OMEA class were not deemed to be proportionate. In that type of scenario, a plausible argument can be made that no cost is too great and no imposition unwarranted if it furthers the objectives of national securi- ty and contributes to the safety of innocent civilians. It is impossible to articulate to the fami- ly of a victim the rationale that the life of their loved one was not valuable enough to outweigh the civil rights of another group. At that point, lives and rights become two separate and unre- lated issues, and the notion of balancing them in an equation seems devoid of human emotion.

Ultimately, the political decision determining which side of this balance should be given preference will depend entirely on which constituency the authorities would prefer to answer to—the OMEA population or those who lose loved ones in a terrorist act that could perhaps have been made more difficult by imposing these measures. This decision necessarily requires consideration of earlier historical lessons, but also turns significantly on the nature of a new modern threat that perhaps undermines the traditional analysis of balancing rights and lives because the perpetrators themselves devalue lives (including their own) to such a great degree. Perhaps the degree of danger we face as a society, and the level of commitment on the part of those who threaten us, requires addressing our fundamental values as a nation in order to defend ourselves. But if so, then the question of how far we are willing to deviate from our val- ues, and for how long, must be answered in the interests of preserving the continuity and integrity of our most cherished freedoms.

Conclusion There is no right answer as to whether the potential payoffs of racial profiling justify its costs, but the key to making that determination is posing the right questions such that an informed decision can be made. Here, those questions turn on the measurement and balancing of the costs and benefits borne by each of the affected parties. This analysis indicates that even the most highly developed profiling system will still result in massive amounts of false positives for each true positive. But the false positives in this case may be an acceptable cost given the enormous benefit of identifying the true positives and preventing false negatives. An alterna- tive approach is to broaden the screening process by incorporating other nonracial factors into the process such that race is only one measure. These more informative measures can include travel patterns, professional or familial connections, criminal records, financial history, and affiliations with other groups/individuals. Race can be an effective element in such a portfolio of criteria, and the other factors can reduce the false positives. In fact, race can even be used as an exclusionary factor to draw conclusions about individual passengers whose overall pas- senger profile may otherwise arouse suspicion. For example, computerized data indicating fre- quent travel between Pakistan and the United States is a trigger for further scrutiny, but can be partially mitigated if racial data supports family connections that would explain regular visits. This additional information can then help distinguish those cases that require further investiga- tion from benign ones.

The costs of racial profiling realized upon the false positives cannot be diminished or min- imized in pursuit of a most critical goal, namely eliminating the false negatives. The key is striking a balance on a macro level that ensures the overall integrity of the system for all par- ties. There will always be individual cases of false positives, but as long as the system works on the group level, that might be an acceptable and necessary consequence. But creating a sys- tem that fails to incorporate some measures to defend the rights of the OMEA group is inap- propriate because it does not consider the burden of the false positives and is therefore not bal-

Kennedy School Review 127

anced. To make this system work, at the very least the OMEA definition must be redrawn more precisely, and race should only be one of several contributing factors—not a dangerously deter- minative one. The tremendous costs associated with racial profiling and the disproportionate burden of these costs on the false positives appears to outweigh the incremental security ben- efits. Therefore, notwithstanding the fact that there could be some marginal improvement in security, ultimately the social costs of a preemptive profiling system are not sustainable, and the system cannot be recommended without the aforementioned adjustments.

Endnotes

1. Randall Kennedy, Race, Crime & the Law (New York: Pantheon Books, 1997), 3–28, 174–175.

2. John Derbyshire, “At First Glance: Racial Profiling, Burning Hotter,” National Review Online, October 5, 2001, http://www.nationalreview.com/derbyshire/derbyshire100501.shtml.

3. Ibid.

4. Ibid.

5. A related argument has been raised in the context of using torture under similar circumstances. Alan M. Dershowitz, “Tortured Reasoning,” in Torture: A Collection, ed. Sanford Levinson (New York: Oxford University Press, 2004), 258–259. (Dershowitz discusses the hypothetical case of a suspect tor- tured to reveal the location of a “ticking bomb” that would otherwise harm hundreds of civilians.)

6. Department of Justice (DOJ), “Fact Sheet on Racial Profiling,” http://www.usdoj.gov/opa/pr/2003/June/racial_profiling_fact_sheet.pdf.

7. Michael A. Smerconish, Flying Blind, (Philadelphia: Running Press, 2004), 14–18.

8. Ibid., 51–55.

9. Ibid., 58–60.

10. Ibid., 38.

11. DOJ, “Fact Sheet on Racial Profiling,” 6. There is an open question as to the extent to which these guidelines are binding upon law enforcement officers or can actually be enforced. For the purposes of this analysis, it is assumed arguendo that these guidelines will be implemented and enforced as written. The jurisprudential basis for the guidelines is enunciated in Brown v. Oneonta, 221 F.3d 329, 337–338 (2d Cir. 2000). In Oneonta, a woman was assaulted in her home and only managed to see the forearm of her assailant. In her description to the police, she noted that the assailant was a Black male with a cut on his forearm. The police then canvassed the town for all Black males, and given that the town was pre- dominantly White, the canvass resulted in a majority of the town’s Black population being questioned. The court held that the use of race in this manner was not illegal because it was based on an eyewitness account of the actual perpetrator and was particularized to the suspect (through the description of the forearm cut) and because the search was geographically and temporally limited in scope. The court specifically noted that even though the police canvass resulted in a large number of Black men being questioned, the racial commonality of these men was directly and relevantly associated with the victim’s eyewitness account. Most importantly, this use of race does not infer any type of generalized tendencies or greater proclivities for criminal behavior to Black men, but rather uses race only as a means of identi- fying one specific suspect. The manner in which the information was obtained, and the subsequent can- vass conducted, did not incorporate any broad generalizations of criminality upon Black men but rather was oriented to collecting the most salient information required to make a timely arrest.

12. DOJ, “Fact Sheet on Racial Profiling,” 2.

13. Ibid., 3.

Kennedy School Review128

14. Ibid., 5-6.

15. Ibid., 5.

16. Ibid., 5.

17. Ibid., 5.

18. Ibid., 1 (comments from Attorney General John Ashcroft, 28 February 2002).

19. There is another complicating factor in this particular case that has been excluded from the analysis for illustrative purposes. This factor is the lack of clarity inherent in the OMEA designation. Theoretically, the profile relied upon to screen out potential Islamic terrorists is Arab or Middle Eastern males, between the ages of eighteen and fifty, who are Muslim. But using visual assessments will not always capture religious affiliation and will be over-inclusive by capturing other ethnicities that are fre- quently mistaken for Arab, such as South Asians and other Mediterranean peoples. Moreover, examining basic passport data will still lead to errors in identification along religious lines and possibly even ethnic lines. These inherent problems with this particular profile make it particularly problematic in terms of generating excessive false positives, but for the purposes of illustration, the discussion of analysis above assumes an accurate ethnic identification and then conducts the rest of the analysis from that assumption.

20. Note that for illustrative purposes, extreme numbers were deliberately used in these hypothetical cal- culations. The realistic numbers would be far more conservative, and accordingly the results of this analysis would be much more dramatic. The hypothetical numbers portray the best-case scenario—in actuality the percentage of actual terrorists is much lower and screening mechanisms are far less reliable. Thus the number of false positives would be much greater.

21. Paul Rosenzweig, “Civil Liberty and the Response to Terrorism,” Duquesne Law Review 42 (2004): 663, 680 (citing Winship, 397 U.S. 357, 372 (1970)).

22. American Civil Liberties Union of Northern California (ACLUNC), “Caught In the Backlash: Arshad Chowdhury,” http://www.aclunc.org/911/backlash/chowdhury.html.

23. Compare Peggy Noonan, “Profiles Encouraged,” Wall Street Journal, 19 October 2001, http://www.peggynoonan.com/article.php?article=72 (editorial advocating that citizens be proactive when seeing “mid-eastern looking” men acting in “suspicious” manners) with Raywat Deonandian, “Profile This!” http://podium.deonandian.com/profiling.html.

24. Derbyshire.

25. Ibid.

26. Ibid.

27. Mathias Risse and Richard Zeckhauser, “Racial Profiling” Philosophy & Public Affairs 32(2004): 132.

28. Kennedy, 13.

29. Ibid.,147.

30. Smerconish, 14–19.

31. DOJ, “Fact Sheet on Racial Profiling,” 4.

32. Ibid., 6.

Kennedy School Review 129


Comments are closed.