A Biometric Screening System

When people travel internationally, they usually move through defined channels, or portals. They may seek to acquire a passport. They may apply for a visa. They stop at ticket counters, gates, and exit controls at airports and seaports. Upon arrival, they pass through inspection points. They may transit to another gate to get on an airplane. Once inside the country, they may seek another form of identification and try to enter a government or private facility. They may seek to change immigration status in order to remain.

Each of these checkpoints or portals is a screening-a chance to establish that people are who they say they are and are seeking access for their stated purpose, to intercept identifiable suspects, and to take effective action.

The job of protection is shared among these many defined checkpoints. By taking advantage of them all, we need not depend on any one point in the system to do the whole job. The challenge is to see the common problem across agencies and functions and develop a conceptual framework-an architecture-for an effective screening system.³⁵

Throughout government, and indeed in private enterprise, agencies and firms at these portals confront recurring judgments that balance security, efficiency, and civil liberties. These problems should be addressed systemically, not in an ad hoc, fragmented way. For example:

What information is an individual required to present and in what form? A fundamental problem, now beginning to be addressed, is the lack of standardized information in "feeder" documents used in identifying individuals. Biometric identifiers that measure unique physical characteristics, such as facial features, fingerprints, or iris scans, and reduce them to digitized, numerical statements called algorithms, are just beginning to be used. Travel history, however, is still recorded in passports with entry-exit stamps called cachets, which al Qaeda has trained its operatives to forge and use to conceal their terrorist activities.

How will the individual and the information be checked? There are many databases just in the United States-for terrorist, criminal, and immigration history, as well as financial information, for instance. Each is set up for different purposes and stores different kinds of data, under varying rules of access. Nor is access always guaranteed. Acquiring information held by foreign governments may require painstaking negotiations, and records that are not yet digitized are difficult to search or analyze. The development of terrorist indicators has hardly begun, and behavioral cues remain important.

Who will screen individuals, and what will they be trained to do? A wide range of border, immigration, and law enforcement officials encounter visitors and immigrants and they are given little training in terrorist travel intelligence. Fraudulent travel documents, for instance, are usually returned to travelers who are denied entry without further examination for terrorist trademarks, investigation as to their source, or legal process.

What are the consequences of finding a suspicious indicator, and who will take action? One risk is that responses may be ineffective or produce no further information. Four of the 9/11 attackers were pulled into secondary border inspection, but then admitted. More than half of the 19 hijackers were flagged by the Federal Aviation Administration's profiling system when they arrived for their flights, but the consequence was that bags, not people, were checked. Competing risks include "false positives," or the danger that rules may be applied with insufficient training or judgment. Overreactions can impose high costs too-on individuals, our economy, and our beliefs about justice.

• A special note on the importance of trusting subjective judgment: One potential hijacker was turned back by an immigration inspector as he tried to enter the United States. The inspector relied on intuitive experience to ask questions more than he relied on any objective factor that could be detected by "scores" or a machine. Good people who have worked in such jobs for a long time understand this phenomenon well. Other evidence we obtained confirmed the importance of letting experienced gate agents or security screeners ask questions and use their judgment. This is not an invitation to arbitrary exclusions. But any effective system has to grant some scope, perhaps in a little extra inspection or one more check, to the instincts and discretion of well trained human beings.

Recommendation: The U.S. border security system should be integrated into a larger network of screening points that includes our transportation system and access to vital facilities, such as nuclear reactors. The President should direct the Department of Homeland Security to lead the effort to design a comprehensive screening system, addressing common problems and setting common standards with systemwide goals in mind. Extending those standards among other governments could dramatically strengthen America and the world's collective ability to intercept individuals who pose catastrophic threats.

We advocate a system for screening, not categorical profiling. A screening system looks for particular, identifiable suspects or indicators of risk. It does not involve guesswork about who might be dangerous. It requires frontline border officials who have the tools and resources to establish that people are who they say they are, intercept identifiable suspects, and disrupt terrorist operations.