Friday, August 12, 2016

Dr. Daniel Pipes: A Call for Intelligent Profiling [by Frederick Schauer]

http://www.danielpipes.org/1385/a-call-for-intelligent-profiling-by-frederick-schauer


In mid-November, Yusuf Suleman Motala, a Muslim leader in the United Kingdom said to be highly regarded and have a vast following, was at Heathrow Airport on his way to the lesser pilgrimage in Mecca. But British officers stopped him and Mr. Motala reports they asked him questions about the Islamic religion, the instruction at schools under his guidance, and his association with "jihadi groups." The resulting delay caused him to cancel his pilgrimage.

The Muslim Council of Britain responded with "outrage and shock" and demanded that such "profiling of Muslims " not recur.
Is this demand reasonable? What, in an effort to ferret out the enemy, is the proper place of profiling? For that matter, what is profiling?

In a just-published book titled Profiles, Probabilities, and Stereotypes (Harvard University Press), Frederick Schauer, Frank Stanton Professor of the First Amendment at Harvard, offers a thoughtful analysis of these questions. Mr. Schauer starts by offering terminology to discuss this subject.
  • generalization is the process by which, "on the basis of a characteristic of some members of a class, we reach conclusions or make decisions about the entire class."
  • Spurious factors are those with no statistical import in reaching generalizations (i.e., the role of gender in predicting intelligence); nonspurious factors do have statistical importance (i.e., gender in predicting physical strength).
  • Prejudices are views based on spurious beliefs about a group.
  • Profiling is "a process of generalization, seeking to narrow the list of possible suspects by identifying an area of interaction among numerous generalizations."
The origins of profiling lie in the early 1950s, when the New York City police, hoping to identify the perpetrator who had set off more than 30 bombs, turned to a psychiatrist named James Brussel for help. Mr. Brussel reviewed the evidence and concluded that the bomber was a middle-aged Catholic of Eastern European extraction who once worked for the Consolidated Edison Company, lived in Connecticut, probably lived with his siblings, and had a serious heart condition. These and other details proved so eerily accurate to describe George Metesky, the science of profiling was born.
Profiling enjoyed high repute until it came out that police forces had simplified Mr. Brussel's elaborate construct and crudely focused on a single factor — race. This reductionism smacked of prejudice, and it had two harmful effects: race as a factor in profiles became taboo and profiling more generally was discredited.

Still, Mr. Schauer notes, profiling remains a routine and completely accepted way of doing business for many U.S. government agencies — so long as race is not involved. For example, a profiling program called the Computer-Assisted Passenger Pre-Screening System (CAPPS) determines which passengers' baggage receives special scrutiny; it looks at such factors as gender, age, purchase with cash or credit card, time of check-in, type of luggage, and demeanor.

Beyond this, "Tax inspectors use profiling to target certain taxpayers for intense audits, customs officials use profiling to determine which arriving passengers warrant close scrutiny, occupational safety and health officials use profiling to decide which businesses to inspect, and police detectives continue to narrow the focus of their inquiries by constructing profiles of likely suspects."

What, however, if race turns out to be nonspurious? What, Mr. Schauer asks, if terrorists who bomb, hijack, or otherwise assault American airliners are "disproportionately younger Muslim men of Middle Eastern background?"

An effective profile, he emphasizes, must consist of dozens of elements, so including a Middle Eastern background means adding it to a long list of other features already in CAPPS, not making it the only or the decisive factor.

Mr. Schauer acknowledges that there is a "strong argument" in favor of including Middle Eastern background as a factor, yet remains ambivalent. He recognizes the high stakes involved and that many observers find it "obvious beyond question" that a Middle Eastern background must be taken into consideration; still, he wonders if it might be possible to avoid this by having passengers arrive a half-hour earlier at the airport, providing time enough to increase scrutiny of other profiling elements.

Mr. Schauer is one of the most sophisticated analysts of profiling and his endorsement of this practice, though qualified, carries real weight. It's time for governments also to explain the subtleties of profiling, justify its intelligent use, and proceed to take into account every nonspurious factor.