A recent Gallup poll indicates that 60 percent of Americans want Arabs to undergo more intensive screening at airports. The Federal Motor Carrier Administration, which inspects trucks carrying hazardous materials, has announced that it "will be looking closely at the drivers, and if the person looks to be of Arab descent that would be enough" for stepped-up scrutiny. Those are just two instances of ethnic profiling now urged on us in the aftermath of September's horrific events.
But can ethnic or racial profiling ever be justified? After all, national security is a legitimate function of the federal government. Even hardcore civil libertarians concede that it would be foolish to treat civil liberties as inviolable when the lives of innocent thousands are at stake. So, what is to be done?
The Constitution guarantees all persons, including non-citizens, due process and equal protection of the laws. Yet those rights are not absolute. The Supreme Court has insisted that the government pass a rigorous two-part test if it intends to discriminate on the basis of race or national origin. First, government must show that it has a "compelling interest" in employing its discriminatory scheme. Surely, protection against the kind of terror that we experienced on September 11 would qualify as compelling. But second, government may not discriminate unless it adopts means that are "least restrictive" when compared against alternative approaches to accomplish the same ends. That second principle will ultimately control disputes over ethnic profiling.
Where, then, should we draw the line? That's a tough question to answer. But there is an analytical framework that can be applied in determining whether ethnic profiling should be condoned. To some, that framework will seem too quantitative, even though it deals with factors that are difficult, perhaps impossible, to quantify. The goal is not to reduce the factors to a cost-benefit calculus. Rather, the goal is to frame the issues so that both practical and moral considerations can be assessed in a structured, systematic manner.
First, the easy part: At one extreme, we should reject uncompromising views of national security. Ethnic profiling cannot be defended simply by asserting that some people will be more secure. When Rep. John Cooksey (R-La.) vented his spleen - "If I see someone [who] comes in that's got a diaper on his head and a fan belt wrapped around the diaper on his head, that guy needs to be pulled over" - rational people condemned such drivel. At the other extreme, civil liberties cannot be allowed to trump national security despite overwhelming evidence that ethnic profiling could, on balance, safeguard individual rights. Imagine, for the sake of argument, that 90 percent of New Zealanders were terrorists and 90 percent of terrorists were New Zealanders. Only a fool would forbid New Zealand nationality to be added to a composite profile of suspected terrorists.
Unhappily, we are left with a vast number of tough cases that require more nuanced analysis. Toward that end, I propose the following standard, which combines two tests centering on individual rights and a third test that is explicitly utilitarian. Ethnicity may be included as one factor in a multi-factor profile if, and only if, all three of these criteria are met:
First, the addition of an ethnic factor must significantly improve the effectiveness of the profile in ferreting out the guilty. Thus, if one in a hundred truck drivers who transport hazardous materials is likely to be a terrorist, and one in a hundred Arabic truck drivers who transport hazardous materials is likely to be a terrorist, the government obviously may not initiate stepped-up scrutiny because "the person looks to be of Arab descent." People who are similarly situated must be treated the same. The key, here, is evidence, not guesswork, that the profiled class is indeed different.
Second, there must be reasonable suspicion to believe that a meaningful portion of the profiled ethnic class is guilty. Otherwise, the profile will be unduly over-inclusive - vesting the sins of the guilty on the innocent. In the truck driver context, it may be that Arabs are 10 times as likely to be terrorists as non-Arabs. But if only one tenth of one percent of hazardous materials truck drivers of Arabic descent are terrorists, the addition of ethnicity to the profile cannot be justified without violating the rights of 999 out of every 1,000 persons. That seems clearly excessive to me, although I might reconsider if the government could find a way to compensate those innocent persons whose rights are abridged.
Third, the benefit of including ethnicity must exceed its cost. On the benefit side, two dynamics enter into the equation: First, the chance that addition of an ethnic factor will prevent a terrorist incident that would not otherwise have been prevented. Second, a measure of the likely harm if the incident were to have occurred. The product of those two terms tells us the value of profiling ethnicity. Again, evidence rather than conjecture must control. For example, the Federal Bureau of Investigation issued an advisory warning that future terrorist attacks might come in the form of hazardous materials trucks used as rolling bombs. If substantiated, that report must certainly weigh heavily in assessing the possible damage.
On the cost side of profiling are potentially grave impositions on innocent persons, triggered by their membership in a targeted ethnic group. Naturally, a vital consideration is the scope of the imposition. We should not object if police use ethnic profiling simply to limit their investigations. Even the questioning of profiled suspects raises few concerns if the suspects are free not to answer and free to leave. But subpoenas, custodial interrogations, and extended detentions are another matter. And, of course, the number of profiled suspects is of paramount importance. It's one thing for government to stop a dozen Arab truckers of hazardous materials. It's quite different to detain all Arab non-citizens unless and until it can be proven that they pose no threat.
No doubt that framework raises more questions than it answers. The devil is in the details. Still, in light of recent events, all three branches of government could be asked to tradeoff precious civil liberties against the prospect of grievous losses to innocent civilians from terrorist acts. That tradeoff cannot be based on seat-of-the-pants speculation, or knee-jerk invocation of theories that refuse to yield despite the exigencies of the moment. We face a new and unprecedented evil, which we must defeat without abandoning the liberties that set us apart from every other country in the world. That difficult task calls for logic, not emotion; for evidence, not rumor; and for a structured approach that weighs the competing interests rationally and morally.