This month, Amtrak officials announced that they were stepping up security, including random searches of passenger carry-on bags, more officers patrolling platforms and trains, and bomb-sniffing dogs. That Amtrak would want to increase security should come as no surprise given that the railway system was the target of terrorist attacks in Madrid in 2004, London in 2005, and Mumbai in 2006. Indeed, one might even ask why it’s taken Amtrak so long to introduce increased security measures to protect against a similar fate in the United States. According to Amtrak president and CEO Alex Kummant, "These new procedures will strengthen Amtrak’s overall security, and they are vital in our efforts to deter, detect, and prevent a terrorist incident on the rail system."
But heightened security procedures do not necessarily translate into being safer and more secure. Will the benefits outweigh the new burdens?
Apparently, Amtrak’s decision to conduct random searches is based on the New York City subway random search program. Unfortunately, simple analysis shows that the New York City program is a flawed design that requires a relatively high manpower effort to yield a relatively low level of effectiveness. That, however, does not seem to concern Amtrak officials, who seem to like the New York City program because it has held up in court challenges.
One thing the Amtrak and New York City random search programs have in common is that they both allow people to refuse a search and walk away. In New York City, if that person was a would-be terrorist, he or she could simply try to gain access to the same subway station via a different entrance (if there is one), which may or may not be subject to search, or go to a different subway station knowing that the NYPD does not have enough manpower to conduct searches at all of New York’s 468 subway stations. On Amtrak, any passenger refusing a search is not allowed to board the train and has their ticket refunded. One question is how easy it would be for that person to simply buy another ticket and board a different train (presumably, a terrorist would be indifferent to which train they were on). It might be easy enough to identify that person if they bought another ticket from a station agent (which requires showing identification). But no ID is required to purchase a ticket from a self-serve kiosk (and a would-be terrorist might have multiple credit cards with different names for making such purchases).
Does refusing a search automatically make someone suspect, and does Amtrak then keep track of that person to ensure that they don’t board another train? Many Amtrak stations have multiple platforms (each with their own access) so it would be at least possible to board a different train, especially if searches are not being conducted on all platforms. (Of course, if the answer is "yes" to the above questions, that raises the larger issue whether refusing a search should automatically make one suspect and subject to tracking. Indeed, would refusing a search put you on some sort of terrorist watch list or a "no-ride" list similar to the no-fly list?)
And nothing would prevent that person from trying to get on another train on another day. Moreover, the random search program would not prevent a terrorist from detonating a bomb while standing in line (or even in the waiting area of a terminal, such as Grand Central in Manhattan or Union Station in Washington, D.C.). Just as it is reasonable to assume that terrorists would be indifferent to which train they were on, it is also reasonable to assume that the target of the attack does not have to be the train itself but simply a large enough crowd of people.
Also worth considering is that a terrorist attack might consist of several bombers (indeed, this seems likely given the other train bombings). Thus, even if a random search stopped one would-be bomber (obviously, a good thing), what are the odds that all the bombers would be stopped by the same random search?
As is the case with many other post-9/11 security measures, there is no real analysis to demonstrate the efficacy and cost-effectiveness of Amtrak’s random search program. Instead, there is simply assertion. According to Amtrak’s CEO Kummant, "This is just the correct step to take."
But if we’re worried about bombers, wouldn’t it make more sense to use the manpower and money for the random search program on more officers with bomb-sniffing dogs patrolling platforms and perhaps even trains, rather than randomly searching passengers in the hope that Amtrak security guesses right about which people to search?
Finally, there is this paradox to ponder. Administration officials often tout the fact that the United States has not (thankfully) been attacked since 9/11 as evidence that their Homeland Security policies and programs are correct. Indeed, last September, Department of Homeland Security Secretary Michael Chertoff told the House Committee on Homeland Security: "On Sept. 11, 2001, no one could have predicted the passage of six years without further attacks on our homeland. By any measure, this is a remarkable achievement. It is the result of our comprehensive efforts to secure our safety." And he asserted that we were "unequivocally safer." So, by Chertoff’s logic, if Amtrak has not been the target of a terrorist attack, and not being attacked is evidence that we are safer, doesn’t that suggest that Amtrak’s prior security measures (sporadic identification checks by train conductors) were working? Then why introduce more intrusive measures that aren’t demonstrably more effective?