Criminal Cues

TSA turns to behavior detection to spot trouble at airports.

As her name might suggest, Maureen O'Sullivan is from New York, and she likes cops.

And yet, the University of San Francisco psychology professor will be the first to tell you that police officers are not very good judges of whether a person is lying, at least compared to Secret Service agents.

O'Sullivan, who is finishing a four-year study on the kinds of people who can accurately tell lies from truth, is one of the many researchers in a field called behavior detection. It involves reading people's nonverbal cues to make informed guesses about their mental or emotional state. The research she and others have been performing has caught the attention of Transportation Security Administration chief Kip Hawley, who is looking to increase the use of behavior detection as a tool for airport security.

In 2003, TSA launched its Screening of Passengers by Observation Technique program, or SPOT, which trains officers to scan airport crowds, looking for body language that could signal malevolent intent. TSA says it will expand the program to the highest risk airports by the end of this year.

"We want to be able to address threats that weren't yesterday's threat," Hawley said at a Jan. 17 hearing before the Senate Commerce Committee. "And finding somebody with hostile intent is a critical component of that."

Here's how the program works: If a trained SPOT officer sees cues that raise suspicions, he approaches the passenger to initiate a conversation, which usually explains the behavior the officer observed. A passenger's brow could be furrowed because he is thinking about being dumped by his girlfriend, for example. Many times, the passenger does not even realize he was targeted by the program. But if the passenger continues to exhibit suspicious behavior, the TSA officer could impose a more extensive checkpoint search or call in local police for interrogation. Israeli security personnel have long used this sort of targeted approach. TSA officials will not say what is on the behavior checklist, for fear of helping criminals learn how to beat the system. But scientists in the field, some of whom have advised TSA in developing the system, say officers look for body language and micro-expressions-tiny but observable, involuntary muscle movements in the face-that are significantly different from those of the typical airport passenger.

There is no "Pinocchio sign," they are all quick to point out; any physical cue correlated with lying also could come from completely innocent sources. A woman whose behavior suggests she is concealing an emotion might be on her way to a funeral. But reading nonverbal cues allows screeners, who see 700 million passengers a year, to narrow the group they are searching for malevolent actors-focusing on one section of the haystack, where the needle is overwhelmingly likely to be.

"You have a certain number of people you can give additional screening to, and the question is what is the criteria you're going to use to figure that out," says Mark Frank, a State University of New York at Buffalo communications professor who has advised TSA and is doing research funded by the Homeland Security Department in this area. "You kind of do it randomly now. . . . If you can apply a little science to this . . . to make those decisions a little better, the costs to that are minimal, and yet you may improve your overall hit rate on somebody with malfeasant intent."

Civil liberty groups have protested, suggesting that any system based on the subjective judgments of officers is ripe for abuse and ethnic profiling. The national coordinator of the American Civil Liberties Union's campaign against racial profiling is suing Boston police after an encounter with a behavior detection officer at Logan International Airport.

Scientists say the method is based on behavior that is not distinct to any particular ethnic group, and the system works well as long as officers understand the limitations. "People are not arrested," says Frank. "All it does is prompt a conversation." Focusing on identifying hostile people over detecting weapons would be a change for TSA. Two such programs-to check passenger records against terrorist watch lists and to pre-screen passengers who want to be expedited through the checkpoint-have fallen far behind schedule, due to privacy concerns and other problems. And the SPOT program is tiny compared to the billions of dollars the agency has spent on metal detectors, explosive-sniffing "puffers," imaging technology and other systems to detect hidden weapons. Technology firms have begun to offer solutions for this kind of hostile intent detection, in Israel and elsewhere. Paul Ekman, a pro bono adviser to TSA and psychology professor at the University of California, San Francisco, says camera systems that scan crowds and identify nonverbal cues for SPOT officers to follow up on will be ready within a year. TSA pilot-tested a system in 2006 that asks passengers or employees simple questions and analyzes their hand pulse, sweat and other

biometric readings for signs of deception. Israeli manufacturer Suspect Detection Systems Ltd. says the system, which is installed at Ben-Gurion International Airport in Tel Aviv and ready to be deployed elsewhere, is being tested by TSA at more locations. TSA consultant Frank says he is not aware of any testing that shows vocal pitch-based technology with a success rate above 60 percent. The communications professor is quick to point out the many myths about lie detection-many think, for example, that averting one's eyes means someone is probably lying, but research does not support that theory. He cautions that more scientific research is needed. "People tend to think it's technology, therefore it's science, and that's not always the case," Frank says.

The SPOT program began at about a dozen airports, but TSA won't say how many now employ SPOT teams, or how many officers have trained. Ekman, who has been studying behavior detection for 40 years and has been advising TSA on its program, says he wishes the agency had more SPOT officers, more training for those officers and more research into the effectiveness of that training. "It's not enough," he says. "But it's a step in the right direction. And the best I know from the couple of meetings I had with the TSA director is if people gave him the budget, he would do these other things."

In its fiscal 2008 budget request, TSA asked for $60 million to add 2,000 document checkers-a function now performed by airline employees-who would be trained in behavior detection. And DHS and other agencies are funding behavior detection research, although the agency and researchers will not discuss it publicly.

"I think the behavior observation-a lot of the person-related screening-as opposed to [weapons] screening is really where we've got to go," Hawley said at the Commerce Committee hearing, "because we can't just keep taking away things from people based on we think they could use it as a weapon."