Is Racial Profiling Racist? It Depends on Whether It Is Based On Known Facts or Statistical Generalizations

If a rapist was described by his victims as a white man, no one would consider it racist if the police limited their search for suspects to white men, a racial group.

If the arsonist of a ski resort was known to be a member of Earth First!, no one would consider it un-American if the police limited their search for suspects to members of Earth First!, an ideological group.

If the murderer of a nightclub owner was known to be a member of the Mafia, no one would consider it unfair if the police limited their search for suspects to members of the Mafia, a criminal group.

But if thousands of Americans are mass murdered in an act of war by a group of Muslim terrorists from the Middle East, police and security personnel are accused of being racist, un-American, and unfair if they limit their search for suspects to Muslims from the Middle East.

Why?

Incredibly, because of a semantic error. The public is confusing the above method of criminal profiling, which is uncontroversial because it relies on known facts about specific criminals, with a different method of criminal profiling, which is very controversial because it relies on statistical generalities concerning criminal activity as a whole.

The statistical approach to criminal profiling is usually referred to as “racial profiling.” Racial profiling uses statistical generalizations that correlate various ethnic groups with criminal activity. The classic example of racial profiling is when a traffic cop stops a black motorist for questioning, based on a statistically higher crime rate among blacks than among whites. How does this example of racial profiling differ from the other examples of criminal profiling described above?

What is common to the other examples is that certain concrete facts about particular criminals are known: in the case of the rapist, his race; in the case of the Environmentalist, his ideology; in the case of the Mafioso, his organization. Because these specific facts about the suspects are known, their presence becomes a necessary condition for being considered a suspect of the crimes in question.

By contrast, in the statistical approach, no concrete facts about particular suspects are known, because no particular crime is being investigated. While, statistically, blacks as a group may commit crime more frequently than whites as a group, because people have free will being black is obviously not a necessary condition for being a criminal. In this sense, racial profiling based on statistical correlation is not dissimilar to racist generalizations, and so it is not surprising that it raises similar objections.

However, these objections do not apply to the method required to locate Bin Ladin’s terrorists within the United States, because we know several specific facts about these terrorists. As in the example of the rapist, we know the terrorist’s physical characteristics: they are all from the Middle East. As in the example of the Environmentalist, we know the terrorist’s ideology: they are all Muslims. As in the example of the Mafioso, we what criminal group the terrorists belong to: they are all members of Al Queda. These are all facts, not statistical generalizations, and taking cognizance of these facts in order to locate the terrorists is not an instance of racial profiling or of racism. The issue is whether we will use these facts to find the killers, or whether our law enforcement officials will idly wait while thousands of enemy soldiers on our own soil plan their next attack on a disarmed civilian population. At stake are the lives of thousands—or millions—of our citizens, our economy, and, ultimately, the survival of the United States itself.