A new word entered the American vocabulary a few years ago: profiling. People became aware of it as a technique used by law enforcement to catch criminals, especially serial killers, by looking at the ways in which their crimes were committed. Specialists then reason backward to figure out what kind of a person committed the crimes, hopefully enabling them to predict the criminal’s future behavior and narrow the circle of potential suspects. Soon law enforcement “profilers”—as they came to be called—were appearing in popular movies and TV shows: Clarice Starling in The Silence of the Lambs, Fox Mulder on The X-Files, Frank Black on Millennium, and Sam Waters on Profiler.
In the racially sensitive environment of the 1990s, a parallel term also made its way into the American vocabulary: racial profiling. A position paper on the American Civil Liberties Union’s website explains it this way: “‘Racial profiling’ occurs when the police target someone for investigation on the basis of that person’s race, national origin, or ethnicity. Examples of profiling are the use of race to determine which drivers to stop for minor traffic violations (‘driving while black’) and the use of race to determine which motorists or pedestrians to search for contraband” (“Is Jim Crow Justice Alive and Well in America Today?” www.aclu.org).
Based on definitions and examples of this sort—as well as a desire to stamp out the kind of race-based injustice to which minorities, especially African-Americans, often have been subject—the media elite concluded that racial profiling is immoral and must be prohibited by law. To this a few voices responded that the targeting of individuals for investigation solely on the basis of race, national origin, or ethnicity is indeed a bad thing and should be illegal; but this does not mean that such considerations are irrelevant to criminal investigations. There are situations and crimes for which these factors can prove useful and even essential.
In the politically correct environment of the last several years, these voices were contending for the losing side of the argument, in the court of media opinion if not law enforcement practice.
Then September 11 happened.
The Change
Suddenly America was under attack by individuals who fit a very clear profile: They were young men, Muslims, foreign nationals, and of Middle Eastern extraction. And in the shock of the terrorist attacks, a lot of people’s views on racial profiling started to change.
Some, of course, did not. I recall, in the days after the attack, Sam Donaldson and Cokie Roberts declaring on ABC’s This Week that the recent events must not lead to racial or ethnic profiling of such individuals.
Others—even at the same end of the political spectrum—were more adaptive to the reality of the situation. Driving to work one morning, I was listening to the local National Public Radio station and a local (very liberal) artist was confessing that, at least when it came to getting on a plane, his views on ethnic profiling had “changed one hundred eighty degrees” in the last few days, “just like everybody else’s.”
The Problem
On the one hand, it was now obvious that a particular cross section of the public—young, male, Muslim Middle Easterners—needed additional scrutiny to prevent further attacks. On the other hand, it was also clear that innocent people must not be persecuted. How could these two objectives be harmonized?
The answers on a practical level—what specific steps should or should not be taken—are still being worked out and will be under development for a long time. Inevitably, mistakes will be made on one side or the other. It is still possible, though, to provide a sketch of the basic moral principles that are relevant to the profiling question. A good place to start is by recognizing the scope of the profiling phenomenon.
The Basis of Profiling
Profiling is a new name for a process that is as old as mankind. It isn’t just a technique used by law enforcement officers. It’s something that every person does every day.
The reason humans profile is because they have incomplete information. There are too many questions we encounter in the course of our lives for which we lack specific information: Is this a book I will like? Should I watch this movie? Will that dog start barking at me? Will that lion try to eat me? Will this person be a good employee? Is the person in the plane seat next to me a physical threat?
We can’t know whether we will enjoy a book or a movie until we’ve read or watched it. We can’t peer inside the heads of animals to see what they’re going to do. Neither can we do that for people—especially people we’ve only just encountered. Yet there is a practical need to answer these questions, because they impact our course of behavior and because guessing wrongly could even, in some circumstances, cost us our lives.
Torn between the lack of specific information and the practical need to answer these questions, humans do the best they can: They profile. Not having the specificinformation they need to answer a question, they take a look at what they know on the subject in general and then try to estimate the specific answer to the question.
For example, when trying to decide the question “Will I like this book?” I normally have nowhere near enough time to read it in order to find out. So I pose smaller, more easily answerable questions: Do I like other books by the same author? Do I like the topic or genre of the book? Is the publisher a good one? How many pages does it have? What does the description on the back of the book say about it? Who is endorsing it? By answering these small questions, I try to build up a portrait of the book—a profile—that will enable me to make an educated guess about my main question: “Will I like it?”
Because it is universal in the human condition to encounter questions one needs to answer but for which one lacks specific information, profiling is a universal human behavior. Reason—and therefore natural law—dictates that when we lack the specific information we need, we engage in the profiling process. We must review what we do know and look for indicators to point us in the direction of what we don’t know. Over time, experience will show what indicators are relevant and how strongly relevant they are, and our profiling process will become more sophisticated.
For example, I may enjoy the writings of a particular author of fiction, a fact that creates in my mind a strong correlation between the fact that he is the author of a book and the idea that I will like it. However, over time I may discover that, while the author is great at fiction, his nonfiction is less enjoyable. As I incorporate this fact into my profiling system, I learn to predict with greater precision which of his books I am likely to enjoy.
The idea of not profiling is inconsistent with human nature. Indeed, even if the Fall of Man had never occurred, humans would still profile, as they would still need the answers to specific questions (“Is this serpent lying to me?”) in the pursuit of which they might pose other, more easily answerable questions (“Does that fruit look good?”).
With this as background, we are in a position to consider how profiling is impacted by considerations of race, ethnicity, national origin, and religion.
Racial Profiling Revisited
If by racial profiling one means “allowing considerations of race, ethnicity, national origin, or religion to be determiners of how we regard individuals,” then it is a bad thing. No question: It’s immoral. But if one means “allowing such factors to assume their proper weight as indicators of a person’s beliefs and actions,” matters are different.
Because profiling is a universal human behavior—one designed into us by God to help us cope with lack of specific information—it would be not only irrational but also impossible to refrain from profiling. Further, if experience shows that a particular indicator tends to correlate with a particular result, then it would be irrational not to take note of that in our profiling. Racial, ethnic, national, and religious indicators are not exceptions to that rule.
Indeed, knowledge of these indicators is often essential for getting along. To take a benign example, many (though not all) Jews do not eat pork. Knowing that, I am less likely to offer a piece of bacon to a Jewish person. That’s politeness. It’s also using a religious indicator in a profiling situation. If it later turns out that this particular Jewish person has no problem eating pork (e.g., if he’s a Jewish Christian), I’ll offer him bacon, but until then, out of courtesy, I’ll stick with what the more generic profile tells me.
Such profiling considerations are also relevant in evaluating what kind of threat a person may pose. Ku Klux Klansmen and Neo-Nazis tend to be white. Black Panthers tend to be black. IRA members tend to be Irish. Al-Qaeda members tend to be Middle Eastern and Muslim. These correlations are real.
That does not mean that these correlations are sufficient. Just because someone is white doesn’t make him a Neo-Nazi. And just because someone is Arab or a Muslim doesn’t make him an al-Qa’eda member. Racial, national, ethnic, and religious factors are relevant to the question of whether a particular individual is likely a member of a particular organization, and it’s foolish to pretend that isn’t the case. But it’s equally foolish to pretend that these factors—taken in isolation—establish an individual’s views, organizational memberships, or likely courses of behavior.
It is only in conjunction with multiple factors that a meaningful threat assessment of an individual can be made. If a guy is white and has a swastika and a shaved head, then he may be a Neo-Nazi. If someone is Arab and young and male and seeking to rent a crop duster, then he may be an al-Qa’eda terrorist.
Then again, maybe not. The first guy actually might be a Buddhist, for both the swastika and shaven heads are Buddhist as well as Neo-Nazi symbols. The latter guy may be farmer who wants to increase his crop yield.
The point is that, as we profile those we encounter, we must strive to neither overestimate nor underestimate the strength of particular indicators. As we get more details about an individual, our profile of him can become more sophisticated and more reliable as a predictor of his views, associations, and likely actions. Making it more reliable means giving each fact we know about him its proper weight.
Failing to do this results in one of two forms of prejudice. In deciding whether an individual is a threat, to give a particular factor greater weight than experience shows it to have is to be biased against persons with that characteristic and constitutes a refusal to learn from experience that individuals with that characteristic are not the kind of threat that they are initially supposed to be. Conversely, to give a particular factor less weight than experience shows it to have is to be biased for a person with that characteristic and is just as much a refusal to learn from experience.
The Way Ahead
The correct course of action, then, is not to flee from reason by pretending that particular factors have no bearing on risk assessment. Neither is it to pretend that they are conclusive when they are not. Instead, we must learn to properly weight particular indicators.
That’s a learning process. It takes time to find the right mix, and it is inevitable that mistakes will be made in the process.
The process is complicated by the fact that the terrorists have tried to blend into American society. They don’t go around armed, dressed, and acting the way they did in training camps in Afghanistan or Somalia. They have tried to eliminate any distinguishing features that might peg them as potential terrorists. In other words, they have tried to change their profile by generalizing it. But they cannot eliminate it: They will still appear to be young, male, and Middle Eastern.
The consequence is that in the war on terrorism a larger pool of individuals will be scrutinized because more people fit the profile. As a result, innocent people will be subjected to scrutiny, suspicion, and inconvenience that they don’t deserve. That’s bad. But let us realize where culpability for this lies: Is it with law enforcement officials who are doing the best they can to prevent civilian deaths by using the only profile they have? Or is it with the terrorists who have deliberately changed their profiles—muddied them—so that a larger number of people will have to be looked at?
In an editorial for the Wall Street Journal (Oct. 19, 2001), Peggy Noonan put the point well. If it turned out that the “terrorists were all middle-aged American blond women who tend to dress in blue jeans and T-shirts and like to go by Catholic churches and light candles, I would be deeply upset not only because the terrorists had done what they’d done. I would also be upset because they were just like me! I fit their profile! I look like them! I act like them! Everywhere I went people would notice me and give me hard looks and watch what I was doing. I would feel terrible about this. But you know what else I’d do? I’d suck it up. I’d understand. I wouldn’t like it, but I’d get it, and I’d accept it.”
That’s part of the unpleasant reality we now must live with. A cross has been placed on all of us: on the American people, on our leaders, our military, our law enforcement officers, and on the innocent people who fit the terrorists’ muddied profile.
In the wake of September 11, what is called for on the part of those responsible for national defense is a willingness to learn, to refine the terrorists’ new profile as much as possible so it is more accurate and so that it burdens fewer innocent people. What is called for on the part of those who suffer from profiling mistakes is patience as American society goes through a learning process. What is called for on the part of all is vigilance— and charity and prayer, so that mere vigilance does not become vigilantism or vengeance.