MTI Report 06-07

 

 

SELECTIVE SCREENING OF RAIL PASSENGERS

 

 

February 2007

 

 

 

 

Brian Michael Jenkins

Bruce R. Butterworth

 

 

 

ACKNOWLEDGMENTS

This monograph is published under the auspices of the National Transportation Security Center at the Norman Y. Mineta Transportation Institute (MTI).

The authors wish to thank the following individuals whose help and guidance were essential: Stephan Parker of the National Academy of Science's Transportation Research Board (TRB); Kelly Leone of the Volpe Transportation Center in Cambridge, Massachusetts; Greg Hull of the American Public Transportation Association (APTA); Roshni Sherbondy, formerly of the Department of Homeland Security, Science and Technology Directorate (for her technical advice on security technologies); and Robert Hertan of APTA's Committee on Public Safety.

The authors are indebted to Janet DeLand for her valuable editing assistance. Thanks are offered also to MTI staff, including Research Director Trixie Johnson, Communications and Special Projects Director Leslee Hamilton, Research and Publications Assistant Sonya Cardenas, Webmaster Barney Murray, and Graphic Artist Shun Nelson. Editing and publication services were provided by Catherine Frazier and Irene Rush.

Table of Contents

Executive Summary 1

Organization of This Monograph 4

Key Judgments 5

Scope of the Study and Definition of Terms 7

The Current Terrorist Threat 9

The Objectives of Screening 15

Methods of Selecting and Screening Passengers 19

100 Percent Screening 19

Random Screening 20

Selective Screening 20

Technological Screening 23

Canine (K-9) Teams 23

Threat Information and Racial Profiling 24

Limitations on Passenger Screening 27

Current Practices 31

General Observations about Attitudes Toward Security
and Passenger Screening 31

The State of Policies and Planning for Screening 33

Resources 34

Current Passenger Search Procedures and Contingency
Plans 36

The Role of Screening after an Attack or Specific
Intelligence Indicating an Attack 38

The Role of Technology 39

Technologies for Screening in the
Public-Transit Environment 41

Technology Field Demonstrations to Date 43

The Value of Selective Screening Regimes in Different
Threat Situations 47

Characteristics of a Good Screening Program 53

Characteristics of an Effective Mass-Transit Screening
System 53

General Management Practices 54

Conclusions and Recommendations 59

Appendix: Security Survey for Transportation Organizations 61

Endnotes 65

Executive Summary 65

Scope of the Study and Definition of Terms 65

The Current Terrorist Threat 65

Methods of Selecting and Screening Passengers 66

Limitations on Passenger Screening 66

Current Practices 67

Technologies for Screening in the Public Transit
Environment 67

Characteristics of a Good Screening Program 67

Abbreviations and Acronyms 69

Bibliography 71

About the Authors 73

Peer Review 75

List of Tables

Post-9/11 Attacks on Rail and Bus Passengers 12

Executive Summary

Those charged with the security of public surface transportation systems in the United States face a dilemma. The terrorist threat is unquestionably real, yet terrorist attacks remain statistically remote events. Terrorists see public transportation as an attractive target, but there have been no major attacks on surface transportation in the United States in the past 10 years. Recent terrorist attacks on public transportation systems in Madrid (2004), London (2005), and Mumbai (2006), however, are seen by other terrorists as successes to be emulated, and similar attacks on the United States have been plotted. Were an attack on public transportation in the United States to occur tomorrow, no one could claim to be surprised.

Public or mass transit systems, particularly large urban systems that use heavy and commuter rail, are attractive terrorist targets for several reasons.

First, attacks on urban mass transit create economic harm and great disruption. More than half the workers in the central city areas of New York, New Jersey, and Connecticut use public transit. Nationally, 14 million people use public transit every weekday; another 25 million use it less frequently, but regularly.See "A Guide to Transportation's Role in Public Health Disasters," Surface Transportation Security, Vol. 10, Transportation Research Board, National Cooperative Highway Research Program Report # 525, pp. 60-61.

 Disruption of the transit system brings serious economic harm to local economies, more so if it causes a decrease in tourism.

Second, the level of security in surface transportation systems is relatively low. Public transit is designed and must operate as an open system with multiple points of entry and exit. Moreover, the volume of rail and transit passengers, which is significantly higher than in air travel, precludes the imposition of anything approaching the rigorous and costly security procedures now in effect at commercial airports. The delays would be unacceptable, and the net security benefits would be slender.

Third, crowds of anonymous strangers on mass transit create a relatively easy environment in which to attack and also to escape.

Fourth, the crowded, contained environments of rail cars, buses, and tunnels make both conventional explosives and nonconventional attacks more deadly.

The pattern of recent jihadist incidents also suggests public mass transit as a probable target. About one-fifth of all jihadist attacks contemplated or executed since 9/11 have targeted urban mass transportation. The average number of people killed per bomb in the successful attacks is relatively high, and the means for achieving these casualties are neither sophisticated nor expensive.

National terrorism alerts, specific threat information, or attacks on public transportation targets in the United States or elsewhere in the world may dictate that security be increased to prevent additional attacks, discourage copycats and pranksters, and reassure riders. In addition to ongoing efforts to improve the design of transportation facilities and increase surveillance, for example, by deploying closed-circuit television, local authorities and system operators should be prepared to rapidly increase security, even if only temporarily. This has generally been done by exhorting staff and riders to increase their vigilance and by deploying additional police and security personnel in the system. These additional security personnel may act as passive observers, or they may interact with riders, observing their behavior, engaging them in conversation, and selectively searching their carry-on luggage, backpacks, briefcases, purses, and parcels.

Various efforts have been made by federal agencies, nonprofit organizations, and transit operators to increase security.See For a current overview of efforts to increase the security of passengers in rail (including transit) transportation, see Cathleen A. Berrick, Passenger Rail Security: Enhanced Federal Coordination Needed to Prioritize and Guide Security Efforts, Testimony before the Committee on Commerce, Science, and Transportation, Washington, DC: U.S. Government Accountability Office, GAO-07-225T, January 18, 2007. The Department of Transportation (DOT) and the Federal Transit Administration (FTA) have issued useful training materials and guidelines for mass transit. The Department of Homeland Security (DHS) and its Transportation Security Administration (TSA) have also provided funds, guidance, publications, and resource centers. In 2004, TSA included mass transit in its security directives, mandating measures largely based on FTA and American Public Transportation Association (APTA) best practices,See Ibid., p. 14. including increased police officer presence, removing or hardening trash containers, video surveillance around platforms, encouraging passengers to look for suspicious behavior, and randomly inspecting passengers and packages, sometimes using canines. Many of these measures were already in place.See David Randall Peterman, Passenger Rail Security: Overview of Issues, Washington, DC: Congressional Research Service, Order Code RL32625, updated May 26, 2005, p. 2, available at www.fas.org/sgp/crs/homesec/RL32625.pdf. In December 2006, TSA issued a proposed rule that would require transit operators (and certain other transportation operators) to establish a 24-hour security coordinator, immediately report incidents and threats to TSA, and allow DHS/TSA personnel to test, inspect, and conduct other duties within their systems--in other words, TSA appears to be strengthening a federal regulatory system for transit security. See Berrick, op. cit., pp. 16-17.

DHS/TSA has awarded grants for planning, training, equipment, and other security enhancements; has provided direct personnel (for example, federal air marshals) to enhance security; has trained and provided canine teams (53 teams in 13 mass transit systems);See Berrick, op. cit., p. 15. has inspected larger rail systems and provided law enforcement training for local authorities; has developed and tested equipment for detecting explosives and chemicals; and has conducted security assessments to define best practices and vulnerabilities.See "Securing Our Nation's Rail Systems," July 2006, available at www.dhs.gov/xprevprot/programs/editorial_0895.shtm, accessed December 5, 2006. Grant programs have been initiated to assess risks, strengthen security, and bolster both security and emergency response capabilities as part of a consequence-management effort.

The Transportation Research Board (TRB) has published a number of practical studies and guidance materials on subjects ranging from emergency planning to communications, robotics devices, and exercises and drills. APTA has a vigorous program of standards development and auditing and has developed useful best practices. The Mineta Transportation Institute (MTI) has produced several useful works on transportation security.1

By and large, these efforts have focused on risk assessment; the need to create additional detection and deterrence, primarily through passive means such as surveillance and observation; and consequence management. Calls for action in Congress focus on federal funding in general and have not considered whether any kind of passenger screening could be introduced in public transit, or under what circumstances. The issue is only now being addressed, although the DHS Science and Technology Directorate (DHS/S&T) and TSA demonstrations of technology to assist in screening have been useful.

Despite the fact that TSA security directives apparently touch on random screening of bags and passengers and eight transit agencies use behavioral observation to "screen" passengers,See Berrick, op. cit., p. 19. only two transit agencies have active programs of passenger screening that include inspection. Both of these programs involve partial, voluntary screening. However, interviews with transit officials suggest that most are keenly aware of the need for screening, and most believe that at some time in the future they will possibly, or even probably, have to implement a screening system that includes passenger inspections.

While screening is the most effective way to combine the essential elements of deterrence and defense, a host of factors--including resource limitations, legal constraints, and public acceptance--make transit authorities view this prospect with great trepidation. Anyone who has been through the development and implementation of airport screening knows that operators face a set of conflicting problems that can be managed but not fully solved.

In the case of mass transit, the only screening approach that is practical--screening less than 100 percent of all passengers--is the one the U.S. population resists the most. There is a general distrust of anything that seems intrusive and is not egalitarian, as well as dissatisfaction with any security solution that does not promise certainty of success. Until the traveling public understands security better and accepts a more realistic long-term level of risk, it will want passive security that promises full protection, equal treatment, and no delay--currently an impossible set of demands to meet.

Still, we strongly believe that maintaining the status quo is just as wrong as trying to implement a screening solution that is fundamentally incompatible with mass transit. We have, therefore, attempted to identify a screening approach that can be sustained for the long term, one that reduces risk while maintaining operational efficiency and public support.

Selective screening raises a number of questions regarding utility, effectiveness, application, workforce requirements, costs, public acceptance, even legality. This report addresses the following questions:

If 100 percent screening is not possible, do selective searches make sense?

If only some passengers are screened, where there is no specific intelligence, what should be the appropriate selection process?

What combinations of selection methods are appropriate under different conditions?

What role can current and future technology play in passenger screening?

What are the characteristics of a good screening program?

Organization of This Monograph

We review the current terrorist threat, which suggests a challenge to security that seems almost certain to persist because attacks on public or mass transportation, urban mass transit in particular, offer terrorists a good "return on investment"; and we examine the objectives and benefits of selective passenger screening and measures of its effectiveness.

The document is organized as follows:

The first chapter, See Scope of the Study and Definition of Terms defines the forms of public transportation considered in the report.

Next, See The Current Terrorist Threat examines past jihadist attacks, modes of operation, and trends, and identifies factors that are of particular relevance to public surface transportation generally and urban mass transit in particular. It assesses the value of urban mass transit as a jihadist target in absolute and relative terms.

specifies the goals screening should be designed (and not designed) to accomplish in the current threat environment.

See Methods of Selecting and Screening Passengers describes different methods of selecting and screening passengers.

See Limitations on Passenger Screening considers the structural, operational, legal, and public limitations on security screening, in particular, selective screening.

See Current Practices provides a snapshot of current security surveillance and screening practices and plans for future implementation to provide an appropriate level of security in the public transit environment.

See Technologies for Screening in the Public-Transit Environment briefly explores what technology can and cannot do now, and what it may be able to do in the future.

See The Value of Selective Screening Regimes in Different Threat Situations explores whether the screening of less than 100 percent of passengers provides value, the kinds of selection methods that are available, and the optimal mixes for different threat environments.

points out some basic mandates for maintaining effective and professional screening.

The final chapter, See Conclusions and Recommendations summarizes our answers to the questions addressed in the study and presents some key recommendations for the transit industry and the government.

This document concludes with an appendix that presents a security survey administered to transportation organizations; endnotes; abbreviations and acronyms; a bibliography; information about the monograph's authors; and an explanation of the report's peer review process.

Key Judgments

Our conclusions and recommendations are summarized below:

Terrorists see urban mass transportation as an attractive target--an easy means of inflicting civilian casualties, causing massive disruption, and provoking alarm. This threat will persist for years to come.

Screening 100 percent of urban mass transit passengers is not a realistic security option. The human resources required, added security costs, and delays would destroy urban mass transit.

Terrorism alerts, specific threat information, or attacks on surface transportation targets elsewhere may dictate that security measures be rapidly increased to prevent any further planned terrorist attacks, discourage copycats, and reassure riders. Selective screening offers a flexible response that can be implemented immediately and reduced as the threat decreases. Screening some passengers through selective searches is a viable security option.

The goal of any security measure is risk reduction, not the prevention of all attacks. Selective searches can contribute to deterrence, oblige terrorists to take greater risks, complicate their planning, force them to use smaller amounts of explosives, and, better yet, divert them to less lucrative venues, thereby reducing casualties.

Existing technology and the use of canine teams can facilitate security inspections, but we are still years away from full technological solutions; security will continue to be labor intensive.

The introduction of any passenger-screening program, especially one that involves selection, runs against Americans' preference for security that is passive and egalitarian. Screening programs must be carefully planned and closely managed to reduce the inevitable allegations of discrimination or profiling based upon race or ethnicity.

Legal challenges to selective screening should be anticipated. A recent court decision upholds selective searches, but only if they incorporate certain features.

A good selection process must be planned in advance; must be based on clear policies and procedures; must combine random selection, behavioral profiling, and specific or general threat information; must maximize unpredictability and deterrence; must allow for program expansion, redeployment, and reduction; should use existing technology; and must maximize interaction with riders, but not in a way that is perceived as harassment.

Vigorous public information programs that outline risk-reduction goals must accompany the introduction of any new security measure that directly engages riders, to allay potential public concerns.

Scope of the Study and Definition of Terms

This study specifically addresses the application of passenger-screening methods in public transportation, or mass transit, systems, particularly in major metropolitan areas. We begin by defining the terms used in this report.

Public transportation, or mass transit , is service operated by transit agencies or systems and provided on fixed routes, that is, with vehicles stopping to pick up passengers on a fixed schedule along a specific route and delivering them to specific locations. This study focuses primarily on urban mass transportation, because urban systems have the greatest density of passengers and therefore the highest potential for casualties in the event of an attack. We focus on heavy, light, and commuter rail service rather than buses; however, this is not to suggest that terrorists do not consider bus service to be a valuable terrorist target--experience in Israel and London shows that terrorists do target buses. See In London, the bomber (Hasib Hussain) apparently detonated aboard a bus either because of delays in the subway on the way to his target train or because his device malfunctioned and required a new battery. The bus appears to have been a secondary target. See Report of the Official Account of the Bombings in London on 7th July 2005, House of Commons Report HC 1087, May 11, 2006, pp. 5-6. The following types of service are considered:

Heavy rail (metro, subway, rapid transit, and rapid rail) : electric railways characterized by high speed and rapid acceleration, with passenger rail cars operating singly or in multicar trains on fixed rails; separate rights-of-way from which all other vehicular and foot traffic is excluded; sophisticated signaling; and high platform loading. (Examples: Washington Metropolitan Area Transit Authority [WMATA] and the Metropolitan Transit Administration's New York City Transit [MTA-NYCT])

Light rail (streetcar, tramway, and trolley) : lightweight passenger rail cars operating singly (or in short, usually two-car, trains) on fixed rails in rights-of-way that are often not separated from other traffic, typically powered by an overhead electric line via a trolley or a pantograph. (Examples: the Santa Clara County [California] Valley Transportation Authority light rail system and the San Diego, California, trolley system)

Commuter rail (metropolitan rail, regional rail, or suburban rail) : electric-or diesel-propelled railways for short-distance, urban passenger-train service between a central city and adjacent suburbs, characterized by multitrip tickets, specific station-to-station fares, railroad employment practices, and usually only one or two stations in the central business district. (Examples: Virginia Railway Express [VRE] in the Washington, DC, metropolitan area and the Long Island Railroad [LIRR] in the New York area)

Passenger ferry service : service provided by vehicle ferries that have at least one deck for vehicles and additional decks for passengers, or passenger-only ferries that have only passenger decks, sometimes with space for bicycles. Although immediate blast effects on ferries may be less lethal than blast effects in a heavy rail system, ferries carry large numbers of passengers and provide poor escape routes. They therefore represent valuable targets. (Examples: vehicle ferries in the Seattle, Washington, area that are more than 460 feet long, each accommodating 2,500 passengers and 218 vehicles, and the 310-foot, passenger-only Staten Island ferries in New York, each of which can accommodate 6,000 people)

Passenger screening is the questioning or physical examination of passengers or members of the public by employees of the transit agency in an area under the control of the transit agency. Physical examination can include passengers' clothing; articles they carry; their identification; and bags or packages they carry. We assume that no cargo is carried on the transit system separately. Searches may be done by hand or may use a technology that can detect the bulk or trace particles of an explosive; the presence of firearms or other dangerous weapons; or the presence of chemical, biological, radiological, or nuclear (CBRN) agents.

The Current Terrorist Threat

Although the individual passenger's risk of a terrorist attack is statistically minute, the threat to public transportation systems, especially urban rail transit, is growing. This is evidenced by the deadly attacks in Madrid (in 2004), London (in 2005), and Mumbai (in 2006). The greatest current threat comes from jihadists inspired by al Qaeda ideology.

In response to the 9/11 attacks on the World Trade Center and the Pentagon, the United States and its allies concentrated on degrading the operational capabilities of the al Qaeda terrorist organization that carried out the attacks. Al Qaeda's Taliban protectors in Afghanistan were toppled, its training camps were dispersed, its key operational planners were apprehended, and its top leadership has been kept on the run. As a consequence of unprecedented focus and cooperation among intelligence services and law enforcement agencies worldwide, many intended terrorist operations have been uncovered and thwarted.

Nevertheless, al Qaeda's leaders continue to communicate and inspire angry young men to violence. Jihadist extremists, certainly motivated by al Qaeda but not exclusively controlled by them, radicalize, recruit, indoctrinate, and instruct eager volunteers who plan, prepare, and launch terrorist operations. In the five years following 9/11, jihadists carried out more than 30 large-scale terrorist attacks, not counting any that occurred in Afghanistan, Iraq, or Israel, or in the insurgency zones of Kashmir or Russia.

Moreover, at least that many attacks have been thwarted, indicating that the jihadist enterprise is operating at a significantly higher level than it did before 9/11, although it is pursuing mostly pre-9/11 scenarios. Roughly one-fifth of the attacks, including some of the deadliest events, have been aimed at public surface transportation. The current fighting in Iraq, the escalating insurgency in Afghanistan, the "jihadization" of the long-running insurgency in Kashmir, new jihadist footholds in Somalia and elsewhere in Africa, the apparent appeal of jihadism to Muslim immigrants, sons of immigrants, and converts in European cities, as well as terrorist plots uncovered in the United States and Canada, all suggest that the jihadist global terrorist campaign will persist for years to come.

The jihadists have adapted to today's hostile operating environment. Their activities are now increasingly decentralized. Radicalization and indoctrination occur locally. Operations are planned and prepared on local initiative, with little evidence of external material assistance. This limits the resources available to the terrorists, but it puts their operations under the radar, reducing the international communications, money transfers, frontier crossings, and other transactions that national intelligence services monitor.

Jihadist planners anticipate a conflict that will last centuries, but local jihadists plan single attacks that must succeed operationally, not long campaigns. This is particularly true of suicide attacks, where terrorists get only one chance to demonstrate their conviction, their courage, and their prowess as warriors. To succeed is to be worthy of God's support and, at the same time, to demonstrate that they have it. Failure suggests its absence and cannot be risked. This pushes the terrorists toward soft targets and reliable tactics. Their primary measure of success has become almost exclusively body count. Virtually any crowded place will do.

This quest for high body counts and soft targets that ensure tactical success makes public transportation an especially attractive target. Public transportation systems offer terrorists easy access and, for those who are not suicidal, ease of escape. Passengers are crowds of strangers, guaranteeing the attackers anonymity; the contained environments of train coaches and buses, stations, and tunnels enhance the effectiveness of both conventional explosives and unconventional weapons, thereby guaranteeing high body counts. Finally, attacks on public transportation systems cause great alarm and disruption, which is the primary objective of terrorism.

MTI's chronology of nearly a thousand attacks on surface transportation over the past four decades shows that terrorists who attack transportation, including those who came before the jihadists, are out to kill. Two-thirds of the attacks were clearly intended to cause fatalities, and 37 percent succeeded (compared with between 20 and 25 percent of terrorist attacks on all categories of targets). Multiple fatalities resulted from 75 percent of the fatal attacks; 28 percent involved 10 or more fatalities.See Brian Taylor et al., Designing and Operating Safe and Secure Transit Systems: Assessing Current Practices in the United States and Abroad, MTI Report 04-05, San José, CA: Mineta Transportation Institute, November 2005.Almost every attack in the past several years has been intended to kill, and the current jihadist terrorist campaign underscores these trends (see See Post-9/11 Attacks on Rail and Bus Passengers).

One way local jihadist terrorists with limited resources and expertise can achieve high body counts is by assembling hundreds of pounds of explosives in a vehicle and detonating it in a public place. The spectacular bombings in Bali, Jakarta, Karachi, Riyadh, and Istanbul are examples of such operations. Fifteen of the jihadists' attacks in the five years since 9/11 have involved vehicle-borne improvised explosive devices (VBIEDs). On average, these attacks killed 24 persons per bomb, with a median number of fatalities in the high twenties.

However, terrorists can easily achieve high body counts with small explosive devices (10 to 20 pounds) by detonating them on trains and subways. These operations are much cheaper than attacks in other public places--no vehicle is required and fewer explosives are needed. Operational security is also easier to maintain, and the devices are easier to construct. Finally, for terrorists who are not suicidal, escape is likely to be easy.

The "return on investment" in public-transit attacks is almost as good as that in large VBIED attacks. An analysis of the terrorist bombings of trains and subways in Stavropol (December 2003), Moscow (February 2004), Madrid (March 2003), Russia (August 2004), London (July 2005), and Mumbai (March 2003 and July 2006) shows that the 26 bombs that detonated delivered an average return of 20 fatalities each, with a median of 19--almost as good as the rate achieved by VBIEDs, for much less investment and with a much lower risk of failure.

Although the sample populations are small and the analysis is crude, these figures demonstrate a powerful operational rationale for attacks on public surface transportation, one that jihadist terrorist planners are surely aware of. They see the attacks on trains and subways in Madrid (where 191 persons were killed), London (52 killed), and Mumbai (207 killed) as successes to be emulated or surpassed.

In addition to the successful attacks, terrorists have plotted to release cyanide on New York's subways, to bomb a subway station in New York, to spread ricin on London's Heathrow Express, to attack London's subways with bombs and chemicals, to bomb train stations in Melbourne or Sydney, to blow up a commuter train in Milan, and to breach underwater subway tunnels in New York. They attempted a second bombing of London's subways and a bus in July 2005, and in August 2006, they left suitcase bombs that failed to explode on a train in Germany. Transportation targets are clearly in the jihadists' playbook.

It can happen here in the United States. In 1997, there was a plot by Islamic extremists to carry out a suicide bombing in New York's subways; in 2003, a plot to release cyanide in New York's subways; in 2004, a plot to bomb a New York subway station; and in 2006, a plot to attack trains in tunnels under the Hudson River.

Post-9/11 Attacks on Rail and Bus Passengers

Date

Location

Description

December 2001

Singapore

Authorities arrested terrorists involved in planning attacks on various targets, including the city's metro system.

January 2003

United States

Authorities learned of a plot to release hydrogen cyanide in multiple locations on New York's subways.

March 14, 2003

India

A bomb on a Mumbai commuter train killed 10 persons.

December 5, 2003

Russia

Terrorists detonated a bomb on a train in Stavropol, killing 42 persons.

February 6, 2004

Russia

Terrorists detonated a bomb in Moscow's metro, killing 49 persons.

March 11, 2004

Spain

Ten bombs exploded on the Madrid commuter transit system, killing 191 persons and wounding approximately 1,900. A jihadist terrorist group claimed responsibility.

August 27, 2004

United States

New York police thwarted a plot by Islamic extremists to bomb the subway station in midtown Manhattan.

August 31, 2004

Russia

Terrorists detonated a bomb in Moscow's metro, killing 10 persons.

April 17, 2005

United Kingdom

Authorities revealed that al Qaeda-trained terrorists planned to place ricin on the handrails and in the lavatories of the Heathrow Express.

July 7, 2005

United Kingdom

Suicide bombers set off three bombs on trains in London's underground tube and one on a London bus, killing 52 persons and wounding more than 700. The attackers claimed association with al Qaeda.

July 21, 2005

United Kingdom

In an apparent effort to duplicate the July 7 attack, jihadists set off three bombs in London's tube and one on a London bus. The devices failed to detonate.

August 2005

United Kingdom

British authorities reported thwarting a terrorist plot begun in 2004 to release deadly chemicals at the Houses of Parliament or in the London tube.

November 2005

Australia

Authorities arrested terrorists planning to attack a number of targets, including train stations in Melbourne or Sydney.

March 7, 2006

India

Terrorists detonated two bombs at a train station in Varanasi, killing at least 15 persons.

April 2006

Italy

Authorities discovered a terrorist plot to blow up a commuter train in Milan.

July 2006

United States

Authorities uncovered a plot to blow up subway tunnels beneath the Hudson River in New York.

July 11, 2006

India

Terrorists planted seven bombs on trains in Mumbai's metro system, killing 209 people and wounding more than 700. A jihadist extremist group claimed responsibility.

July 31-August 1, 2006

Germany

An unexploded suitcase bomb was discovered on a double-decker commuter train in Dortmund. A second unexploded bomb was discovered at a railway station in Koblenz. Police subsequently arrested two Lebanese men suspected of involvement.

 

The Objectives of Screening

Defining the objective of screening would seem simple: To prevent any kind of attack. By that definition, however, almost all defensive measures will fail. Nevertheless, screening in urban mass transit, while it will fall short of absolute prevention, can achieve useful and realistic goals. It may deter an attack from being planned altogether. It may divert attacks to targets of lesser value. It can complicate operational planning. It can provide law enforcement with greater opportunities for interception. Finally, it may force attackers to alter their plans at the last moment, reducing their chances of success or the lethality of their attack.

To deter an attack altogether or divert it to another target is no small accomplishment. Ultimately, the terrorist is driven to venues or devices that will produce fewer casualties and will have less economic impact.

Complicating the terrorists' operational planning is a reasonable goal. If conspirators have to lengthen their planning cycle and conduct more surveillance, there will be more opportunities for authorities to intercept them.

Not all terrorist operations are suicide, or "martyrdom," operations. The conspirators in the Madrid and Mumbai bombings hoped to escape. It is easier to recruit terrorists who believe they have a reasonable chance of escaping than to recruit people who are committed to dying in the process. There were 13 bombers in Madrid and only 4 in London. Fewer bombs meant fewer casualties--191 compared with 52.

In other words, the effectiveness of screening should not be measured only by the weapons, explosives, or attackers actually stopped. Finding weapons, bombs, prohibited items, or suspicious passengers is important, but it cannot be the only objective of screening or the sole criterion of effectiveness.

U.S. aviation history provides several examples of attacks redefined and deferred by security measures. Unfortunately, the details of examples remain classified. An unclassified example comes from the January-March 1995 attempt of Middle Eastern terrorists to place bombs on as many as 11 wide-bodied aircraft bound for the United States from Asian airports. Ramzi Ahmed Yousef, the organizer of that plot, who also masterminded the first attempt to bring down the World Trade Center towers in New York City, had researched developments in explosives-detection technology. He had determined--erroneously--that U.S. airliners might at that time be using effective technologies in their screening of passengers and cabin baggage that were relevant to certain components of the device he was constructing. Apparently because of this fortunate misunderstanding, Yousef decided to use a homemade alternative. The terrorists' insufficient care in making the extremely unstable homemade devices caused a fire in their Manila apartment, which led to the discovery and disruption of the plot.

Failing to find anything does not mean the system is not working. Equally important, missing something also does not mean the system is not working, particularly if the screening function is determined, in the aftermath of an attack, to have reduced the attack's lethality.

If risk reduction, rather than risk elimination, is the objective, it is also true that where 100 percent passenger screening is not possible--as is currently the case in mass transit--programs involving partial and voluntary screening can still have value, particularly if they are managed properly to maximize their deterrent and detection benefits.

This can be demonstrated by two examples. First, in the voluntary screening regimes used by transit operators in New York and New Jersey, passengers can choose not to submit to screening. They are not allowed to board at the station where the screening was refused, but they can leave and board at another one. While at first glance this may seem nonsensical, it is not if one assumes--as we have every reason to do--that trained officers, both uniformed and plainclothes, are observing all passengers entering the screening process and particularly those electing to refuse it. Thus, voluntary screening systems have considerable value. Second, if the screening location is randomly shifted from one station to another, and if the selection of passengers is difficult to predict, deterrence is significantly increased. Unless the decision to determine where screening operations will take place or how passengers are selected is known or can be determined by observation, this complicates potential attackers' operational planning.

A screening regime that will be in place for a relatively long period of time will also have to meet other, secondary objectives that are directly related to the goal of risk reduction. The regime must be kept at a high level of professionalism and effectiveness, which is difficult in the gritty real world of managing both normal operational personnel and security personnel. Being explicit about what the screening is and is not designed to do, having operational procedures that are followed, and being interactive and courteous with passengers are vitally important from the standpoint of withstanding legal challenges and minimizing negative public reaction. To the extent that either current technology or canine units are used to augment the search of selectees or their packages, understanding the complexity of the interface between operator and technology and maintaining quality control and resolution protocols are also key. We explore many of these features in beginning on .

It is important to acknowledge that situations of increased threat often are not identified by specific intelligence. In a perfect world, the various sources of intelligence would be able to track all terrorist operatives. If security intervention were needed at all, it would be covert and invisible to the passenger, because the identities of the attackers would be well known. In reality, information about most threats, even under the best of circumstances, is fragmented, incomplete, and uncertain as to either location, timing, type, or number of attacks. More worrisome, the homegrown nature of the London and Madrid conspiracies, along with the relatively short period of time from conception to execution of the attacks, suggests that the chances of timely intelligence alerts are not great. In short, those responsible for security measures--screening among them--may get no warning.

Methods of Selecting and Screening Passengers

100 Percent Screening

Although the U.S. air passenger transportation system has been criticized for having security gaps involving cargo and ground-service personnel such as catering company employees, aviation security authorities have put in place a system in which every passenger, employee, or object that enters the aircraft, the sterile area, or the Secure Identification Display Area (SIDA) on the tarmac is physically screened or has had relevant paperwork or data examined. Before the first security screening was implemented in the 1960s, commercial aviation was as vulnerable to security threats as mass transit is now.

However, commercial aviation was inherently more restricted. Before the inevitable call for similar screening of transit systems is made, it is important to clarify how the aviation system differs from the more complex public-transit systems.

By design, commercial aircraft have few entry points. This is clearly not true for trains used in urban mass transit. Furthermore, because people have been killed on the tarmac by vehicles, or by walking into propellers, or being ingested into jet turbines, procedures to keep passengers away from commercial aircraft have always been in place. By contrast, mass transit is designed so that the public can be within feet of train or bus doors. The physical configuration of mass transit is entirely different from that of commercial aviation.

Both aviation and mass transit depend on reliable and frequent departures and arrivals, but the difference in passenger volume is staggering. According to APTA, New York's Penn Station alone handles the same number of passengers during each morning's peak hours as Chicago's O'Hare International Airport handles in 60 hours.See American Public Transit Association (APTA) Teleconference, "Transit Security Priorities Post-9/11," October 2006. The highest peak passenger load is in New York, on the Lexington Avenue Express and Local trains. These two routes alone are used by 63,200 passengers an hour traveling in the main rush-hour direction; the highest rate for any 15-minute period in that hour is slightly over 15,000 passengers.See Transit Capacity and Quality of Service Manual-2nd Edition, Transit Cooperative Research Program (TCRP) Report # 100, Washington, DC: Transportation Research Board, 2003, pp. 2-21. By contrast, an average of 210,000 passengers per day pass through O'Hare,See "Chicago's Airports Forecast Record Summer Travel," Chicago Airport System News Release, City of Chicago, Department of Aviation, May 25, 2006. only 3.32 times the number the two train lines experience per hour in just one direction.

In short, the throughput demands at screening points at airports are already challenging, but the challenge would be hundreds of times greater at major ground transit stations, especially at peak hours. A lengthy screening process could generate crowds that themselves could become prime targets for suicide bombers seeking mass casualties.See At some ferry terminals, the throughput demands are less, the access points are fewer, and the departures are less frequent, so while problems of staffing, costs, equipment, screening protocols, and passenger delay and frustration would remain, the design of a screening system is less inherently problematic.

It is, therefore, essential to understand the methods of performing selective searches, that is, screening in which not all passengers are searched.

Random Screening

In random selective screening, passengers or luggage to be screened are selected by random methodology. Such screening can probably best be explained in the negative: passengers are not selected on the basis of behavior, clothing, or attitude; on the basis of apparent nationality, race, or gender; or on the basis of identification or personal documentation. Neither is the selection based on patterns such as time of boarding, origin, or destination, or anything about the travel itself. The selection is mathematically random and, therefore, unpredictable.

There are many ways of generating random selections. The simplest would be to select every tenth or twentieth passenger, but this has the obvious disadvantage of allowing anyone monitoring to figure out the sequence unless it is changed frequently enough to avoid useful prediction. A handheld computer can randomly generate a number between 0 and 999 for each passenger, and those whose number is less than, say, 30 can be selected to ensure that a given percentage of passengers are screened. Various information carried in the reservation system or on the ticket also could be used.

Unpredictability and verifiable independence of the selection from anything about the passenger and his or her travel are the key features of random screening. These features can be important in sustaining legal approval and countering charges of racial profiling.

Selective Screening

Selective screening is the inverse of random screening. In selective screening, certain factors are considered to be indicative of a possible attacker and, therefore, increase the likelihood of a search. Jocelyn Waite provides a good description of selective screening, sometimes referred to as profiling, in her excellent article "The Case for Searches on Public Transportation":

Profiling can distinguish by behavior (e.g., buying a one-way airline ticket); a combination of behavior and appearance (e.g., wearing a large loose overcoat in weather not calling for overcoats); or appearance alone (apparent racial or ethnic identity). Singling out particular types of clothing (that is, clothing that could conceal weapons or explosives) or particular sizes of packages (those that could conceal weapons or explosives) could be deemed profiling and could be used to target security screening. However, the term "profiling" most commonly calls to mind racial or ethnic profiling, which has given rise to considerable controversy. It should be noted that the use of racial classifications are [sic] not per se unconstitutional, but [they] are subject to strict scrutiny to justify them.See Jocelyn Waite, "The Case for Searches on Public Transportation," TRB Legal Research Digest, No. 22, October 2005 (published under TCRP Project J-5, "Legal Aspects of Transit and Intermodal Transportation Programs"), pp. 18-19.

Behavior and Appearance

Behavior and appearance can be related and are often examined together. Many countries have developed sophisticated and continuously evolving behavioral profiles and associated training regimes that help observant personnel detect certain behaviors. The behaviors can include attitudes such as hesitancy; body language such as avoiding eye contact or being in a trancelike state; behavior in crowds such as trying not to call attention to oneself; and reacting to police or uniformed personnel by avoiding eye contact with them or avoiding where they are. When combined with certain clothing--bulky jackets that could hide a suicide belt--or bags of a certain size, they can provide trained officers with reason to select an individual for search.See Terrorist Activity Recognition and Reaction for Transit Employees, National Transit Institute, Federal Transit Administration (Washington, DC: U.S. Department of Transportation, n.d.).

Officers or employees with long experience in a particular transit setting often develop a sixth sense that something is wrong or out of place. This enables them, for example, to quickly distinguish between a mentally deranged person or someone anxious not to be late for a connection or meeting and someone who might be placing or about to explode a device. Trained officers can also question such people and interpret whether their responses warrant suspicion. Nevertheless, identifying potentially dangerous individuals on the basis of behavior, clothing, or baggage is quite difficult.

Baggage Type

Certain kinds of clothing or bags can conceal weapons or explosives. Where specific threat information about a weapon or device is available, passengers could be selected for screening on the basis of the size or type of bag or parcel they are carrying.

Gender and Ethnicity

While selection solely on the basis of gender or ethnicity is prohibited and can be ineffective as a security measure, case law indicates that it is permissible to consider certain factors.See Waite, op.cit., p. 19. The U.S. Supreme Court upheld the use of racial profiling by U.S. Border Patrol agents making stops along the U.S.-Mexico Border. Authorities who have reliable intelligence that attackers are most likely to be males appearing to be from Southwest Asia or North Africa would be foolish not to focus on individuals fitting this description for as long as the intelligence information remains valid.

Ticket, Method of Purchase, and Identification

For public surface transportation, screening based on the method of ticket purchase or purchaser identification is currently mostly theoretical. There is no requirement to show identification to use mass transit. A large and growing percentage of tickets are purchased without any interface with a station agent. Many people purchase tokens and magnetically encoded tickets through machines, and larger transit systems now use smart cards, which are purchased for multiple trips (on both rail and bus) and require interface only with a card reader in the entrance lane. While some smart cards are purchased online with credit-card numbers, addresses, or other personal information, this will not become the dominant means of purchasing tickets for a long time, and will never be the only means. In addition, there will be significant legal constraints on the use of information entered in order to purchase a smart card. While such information could be required or used in the future, this seems remote and costly at the moment.See An increase in the use of smart cards and contactless smart cards will facilitate a faster flow of passengers than is currently the case, making screening lines even more infeasible.

Other Factors

In addition to the method of selection, other factors, such as the following, may determine the deployment and intensity of searches:

Time and route of travel. Selective screening can focus on routes and stations that are the most central to the system and through which the greatest number of people travel. Both the Madrid and London bombings took place during the rush hour, and although the terrorists initially entered at other stations, the attacks focused on central stations and trains at times of peak capacity. Locations where evacuation or access by emergency personnel are the most difficult--including trains between stations--could also be factors in deploying searches. Countering an attack that might involve chemical, biological, radiological, and nuclear (CBRN) weapons would entail different considerations, including a focus on places where air circulation could reach the greatest number of people.

Proximity to a target with high symbolic value. Although the primary objective of jihadist attacks is mass casualties, a secondary objective is to destroy or damage assets of symbolic value. Stations that are close to national landmarks, particularly those viewed by jihadists as symbolic of Western power or decadence, may require additional security.

Technological Screening

The experience in aviation suggests that where 100 percent screening is not feasible, technology can help determine whom to screen (or to screen more intensively) if it is properly implemented. Explosives can be detected either by their bulk composition or by the presence of particles attached to a surface or released into the air--so-called explosive traces. Relatively few bomb makers are skilled enough to not leave such traces on their clothing, documents, persons, or baggage. The relatively large amounts of explosives needed to constitute a serious threat in the transit environment compared to the amounts needed for attacks on commercial aviation also make the use of bulk detection devices highly feasible.

There is no technological method for screening all mass transit passengers as is done in airports, and the development and deployment of technologies suitable for screening passengers in a rail environment for explosives at standoff distances of 10 meters or more is many years off. However, using technology to search those selected more effectively by some other means holds out some intriguing possibilities, if only because the amounts of explosives used in the London and Madrid bombs were 10 to 20 times the amounts that can bring down a pressurized airliner and are thus relatively easy to detect, even in the "dirtier," more cluttered, fast-moving environment of mass transit.

Canine (K-9) Teams

Well-trained canine teams are available now and could be considered an effective "technology." Subject to proper quality controls and training and used for short periods of time, a well-functioning and alert canine team can be effective at detecting particles and vapors of explosives. The dogs have the unique ability to follow the vapor to its source and therefore hold considerable promise for examining suspicious or unaccompanied bags and for searching areas after an alert. If trained to focus on people in a transit environment rather than objects, they can also help screen passengers. However, "the potential benefits of K-9 deployment will not be achieved in the transportation environment unless management actively attempts to understand what will be gained from the K-9 unit and how its performance can be measured."See K-9 Units in Public Transportation: A Guide for Decision Makers, Report 86: Public Transportation Security, Vol. 2, Washington, DC: TRB Transit Cooperative Research Program, June 2002, p. vii. This guide did not consider canine units for routine screening of passengers. Public and handler response to canines is generally positive and can mask the significant challenges of ensuring solid initial and recurrent training and acceptable detection, particularly since most search dogs cannot function accurately for more than 30 minutes at a time. This factor suggests that the only canine approach to 100 percent screening would be to rotate large numbers of search dogs trained specifically to focus on moving people in a transit environment, which could entail many complications and costs.

Threat Information and Racial Profiling

No practice or concept has been more controversial than racial profiling, nor have allegations of any other practice rightfully brought more screening systems to their knees. Racial profiling can be defined as selecting certain passengers for screening based solely on their race, presumed national origin, ethnicity, or religion, or the use of a selection method that effectively guarantees that only people with the same physical characteristics are screened. This practice strikes at the heart of civil rights protections, and although race may be considered when there is specific supporting threat information, transportation operators must be extremely careful here.

Racial profiling unrelated to specific threat information is unacceptable and can be counterproductive. Unless there is reliable information that only certain populations, defined in national or racial terms, are the likely attackers, selections can divert attention away from passengers who may be exhibiting traits or behaviors that are far more directly linked to the actual threat. As the jihadist enterprise acquires more converts, certain things become less certain. Converts to the cause who are not from Southwest Asia, North Africa, or the Middle East have been used for some missions specifically because they do not fit stereotypical racial profiles.

The most common denominator among recent attackers is that they tend to be young men who have become immersed in and converted to a jihadist worldview. If today's terrorism is primarily motivated by ideological conviction, the diversity of attackers is likely to increase, not decrease.

Terrorist groups can often find people to carry out attacks who do not fit known profiles or who may be disguised. According to Bruce Hoffman of the RAND Corporation, "Sometimes the bombers disguise themselves: male shaheed (Arabic for `martyrs') have worn green Israeli Defense Forces (IDF) fatigues; have dressed as haredim (ultra-Orthodox Jews), complete with yarmulkes and tzitzit , the fringes that devout Jews display as part of their everyday clothing; or have donned long-haired wigs in an effort to look like hip Israelis rather than threatening Arabs. A few women have tried to camouflage bombs by strapping them to their stomachs to fake pregnancy.See Bruce Hoffman, "The Logic of Suicide Terrorism," Atlantic Monthly, June 2003, p. 2. The bottom line is that racial or ethnic profiling not tied to threat information is not only illegal and unsustainable, it can be tragically ineffective as well.

At the opposite extreme is the situation in which specific threat information is learned about an impending attack, not in time to interdict the individuals or the terrorist cell, but specific enough to know what to look for. In this case, screening would be time-limited and could include any number of factors, including the number of attackers, their age, their gender, their clothing, their appearance, their national origin, what they may be carrying, and when and where they might be traveling. Factors of race can legally be included in such searches because they are specifically tied to threat information.

Limitations on Passenger Screening

In addition to the physical and operational limitations on passenger screening discussed in the previous chapter, there are other equally important constraints.

The first is resources. Whether technology is available or not, screening is workforce intensive. Transit security chiefs interviewed in this study could not provide estimates of how many work-hours would be required to maintain a screening function, in part because no practical screening concept has been defined by many transit agencies. However, most transit agencies indicated that they were not capable of sustaining any kind of screening function for more than a few weeks without augmentation from other authorities.

The other two major limitations are legal authority and public support. Passenger screening will confront a difficult political environment. Advocates of civil liberties will challenge screening on the grounds of guarding individual rights against warrantless and unreasonable searches, seeing themselves as part of the checks and balances needed to preserve democracy. Legal challenges are more likely if there is little public and passenger support for screening.

The constitutional and precedented authority for conducting passenger screening has been examined in Jocelyn Waite's excellent article, which summarizes the legal constraints any transit authority must consider when instituting the kind of administrative search regime that generally has been held to be constitutional. She emphasizes the need for such a regime to be carefully thought out and controlled:

An administrative search must be conducted as part of a general regulatory scheme (although a regulation per se may not be necessary) that furthers an administrative purpose, rather than furthering criminal investigation. A warrant is not required where it would frustrate the governmental purpose behind the search. The search derives its legitimacy from governmental authorization, not consent to search. The court will balance these factors: the need to search, which should promote a substantial government interest (e.g., search for explosives and weapons prior to boarding aircraft), against the invasion that the search entails. The scope of the search is limited to furthering the administrative need. At least some courts will invalidate such a search notwithstanding a legitimate government interest if there is also an impermissible motive. The search must have a nondiscretionary application. Both notice and methodology will affect the assessment of the intrusiveness of the search.See Waite, op. cit., p. 25.

Waite continues:

Although the law on transit security searches is still emerging, transit agencies may look to established case law on suspicionless searches for guidance. Legal authority on checkpoints, special needs, and general administrative searches will provide useful insight into the requirements for constitutional screening policies. Cases on airport security and sensitive area entry screening should prove particularly relevant.

For passengers, security measures constitute a tradeoff between risk and inconvenience, delay, and intrusiveness. Immediately following a terrorist attack on a similar transportation system or upon receipt of credibly communicated intelligence information of an impending attack, the objectives of reduced risk and increased safety will outweigh inconvenience. As information about the threat diminishes or if the attack is geographically distant, public support for screening, particularly screening that is viewed as intrusive or unprofessional, will diminish.

Other factors may also contribute to the decline of public support for security searches. Seven months after 9/11, Harvard University surveyed law students (including some minority students) about their reactions to different kinds of aviation screening situations.See W. Kip Viscusi and Richard J. Zeckhauser, Sacrificing Civil Liberties to Reduce Terrorist Risks, Cambridge, MA: John M. Olin Center for Law, Economics, and Business, Discussion Paper No 401, p. 32. The survey found that support for selective screening that targets a particular group for more intensive screening rather than affecting all passengers increases considerably with the level of delay imposed on all passengers. In other words, the prospect of delays for all passengers increase support for selective searching. When a delay of 10 minutes in the system was proposed, 44.7 percent favored targeting screening on a particular group; when the delay was increased to an hour, the percentage increased to 73.9 percent. The study also found that majority populations (whites) are more willing to accept profiling focused on them if it results in less wait for others, whereas minority populations showed less support for screening focused on them and were relatively insensitive to the decrease in delays for others.

Perhaps the most interesting conclusion of the Harvard study was that the sensitivity to perceived terrorist risk varies considerably among individuals and is not rationally related to specific probabilities that risk will be reduced by a given percentage. The only screening regime that garnered strong support was the 100 percent regime:

Whether the terrorism risk is reduced to 50% of its current level, [to] 1 in 1 million per flight, or 1 in 10 million per flight, or zero, is not significantly different. Indeed, a regression analysis [of the delays that the respondents were willing to accept] does not differ significantly across three scenarios in which risk is reduced but not to zero. Doing something about terrorism risks that is incomplete but beneficial has a fairly similar attractiveness across these three options. However, a policy that would completely eliminate the risk is much more attractive and commands a much higher willingness-to-pay value.See Ibid., p. 25. (Emphasis added)

In short, inconvenience and delay are fully accepted only when the security system causing them promises absolute safety. Since no security system can guarantee absolute safety--and screening in transit most definitely cannot--this is a problem. The U.S. public expects egalitarian treatment of all, and some population groups--in particular, ethnic minorities with a history justifying suspicion of governmental action--become deeply resentful of selective security procedures. To garner support for such procedures, a carefully thought out campaign of public and passenger education, along with outreach to various communities that takes fully into account the unique aspects of each locality, is essential.

There is a well-known phenomenon in the U.S. governmental system regarding security: The higher the official making a statement about security systems, the greater the temptation to promise 100 percent safety. This is a fatal error, as politically tempting as it may be. Without divulging information that could undermine detection and deterrence, the public information arms of transit authorities should speak about increases in security, not the absence of risk, about work in progress, not a solution. It is more important to maintain credibility for the long term--and it will likely be a long term--than to gain short-term support. Such claims also unnecessarily invite, and then increase, the negative consequences of the inevitable testing of security system effectiveness by the media.

Current Practices

To assess the current state of security in general and screening procedures or plans in particular, MTI's National Transportation Security Center sent a questionnaire See The questionnaire is reproduced in the appendix of this report. to eight transit operators chosen because they are representative of transit systems and active in the field of security. Most of the respondents were candid, although one or two appeared to be trying to put the best face on the difficult situation they were responsible for managing.

Because of time and budgetary constraints, site visits were not made to verify the information provided. Nevertheless, the responses provide an informative outline of security surveillance and screening practices in place now and plans for future implementations; current resources and constraints; and prevailing concepts about the challenges of providing an appropriate level of security in a public-transit environment. The responses are generally consistent with the Government Accountability Office's (GAO's) January 18, 2007, report on passenger rail security.See Berrick, op. cit.

General Observations about Attitudes Toward Security and Passenger Screening

All respondents implicitly or explicitly consider U.S. public transit to be a possible or likely terrorist target; some seem to believe an attack is a matter of when rather than if . There seems to be a general recognition that requirements for system openness create vulnerabilities that terrorists could exploit to achieve mass casualties and spectacular results, using relatively unsophisticated techniques.

Nearly all respondents view passenger screening--which a few prefer to define as "baggage screening" because of sensitivities to the physical searches of people and to profiling generally--as either necessary now or likely to be needed in response to future attacks on transit systems or credible threats. All view any scheme of 100 percent passenger screening as completely incompatible with the speed and openness needed by transit systems to meet basic operational requirements. Nevertheless, one transit system does plan to search all bags if threat conditions elevate to the Red level.

All the respondents consider the primary limiting factors to be a combination of legal liability, perceived resistance by the public or by general management anticipating public reaction, and availability of human resources (law enforcement officers). The combinations of these factors differed among respondents, revealing two geographical divides: distance from the site of New York's World Trade Center, and distance from the East Coast:

For New Jersey and New York, which already have conducted and continue to conduct passenger inspection or screening, the primary limitations are resources, especially now that the New York City screening system has been upheld by the courts; the legality of Boston's system was affirmed during the 2004 Democratic National Convention.

For two East Coast and two West Coast transit agencies, public and general management reactions weigh more heavily. Recent court rulings have provided more certainty on the legality of a search regime in a public-transit environment.

Two West Coast agencies are taking many good proactive meas