A recent leaked report on the TSA revealed glaring failures by the agency. But screenings aren’t worth the cost even in the best of circumstances.
Airline security is a miserable business. Getting through a checkpoint is much like being in a fire drill in the middle of a traffic pileup at the busiest intersection in the city. And now, thanks to a leaked report on the Transportation Security Administration, we have a hard number to assign to our feelings of frustration: Airline security is 95 percent ineffective.
ABC News reports that TSA agents failed to detect a threat in 67 out of 70 recent trials. Posing as passengers, U.S. Department of Homeland Security “Red Teams” were able to carry weapons and fake explosives (“simulant” threats) through security checkpoints without any trouble. Former TSA Administrator James Loy, who led the agency in its first year, called the leaked results an “abominable failure.”
Homeland Security has since announced a series of steps to overhaul the agency, including the resignation of the TSA’s acting director. But the nation should be looking more critically—and more analytically—at the costs and benefits of even attempting to provide total airport security. When the TSA isn’t doing its job, then airport security is a complete waste of money. But by the strictest measure, airport security screenings are not worth the cost even in the best of circumstances.
Back in 2008, researchers from the Centre for Infrastructure Performance and Reliability at the University of Newcastle Australia tried to calculate the “annual cost per life saved” of various then-new security measures implemented in the U.S. after the attacks of September 11, 2001. The paper is a bit dated now, and it depends on a number of generous assumptions. But the fundamental inquiry is still sound, and even more relevant now in light of the revelations about the TSA: How much does totalistic American risk-mitigation for flight really cost us?
The researchers found that reinforced cockpit doors come at an annual cost per life saved of $800,000—a great investment. (According to the researchers, annual costs per life saved of $1–10 million pass the cost–benefit threshold for most U.S. federal agencies.) On the other hand, another security feature designed to prevent a 9/11–style attack, the Federal Air Marshal Service, has an annual cost per life saved of $180 million.
Using the same formula* as the researchers (and some similar assumptions), I estimated the same costs for airport pre-boarding security. TSA checkpoints have an annual cost per life saved of $667,000,000—two-thirds of one billion dollars.
Here’s how this equation works. One critical figure is counterfactual: the number of lives saved annually by post–9/11 security measures. We have no way of knowing what this number really is. Terrorists haven’t hijacked or bombed any American flights since 9/11. We have no basis for comparison, either. No one was ever killed by a hijacking within the U.S. before 9/11.
So the researchers—Mark G. Stewart and J. Mueller—came up with a number: 300. Here’s how they arrived at this figure:
As the threat to the U.S. homeland is from Al-Qaeda, it would be reasonable to consider the period of a heightened threat from Al-Qaeda to be a suitable time period for the estimation of fatality rate—this is a 10-year period, 1992–2001. Accordingly, we will assume that, in the absence of enhanced security measures, there would be a 9/11 replication every 10 years in the United States. That is, the annual fatality rate before enhanced security measures is 300 per year.
Risk reduction is also difficult to calculate. Stewart and Mueller explain the assumptions that led to their calculation of risk (R) mitigated by pre-boarding security measures (efforts that may go beyond checkpoint security).
The extra and more vigilant intelligence, immigration and passport control, airport screening, and other pre-boarding security measures implemented since 9/11 as arrayed in the TSA’s 14 layers of security should result in an increased likelihood of detection and apprehension of terrorists. Increased public awareness is also of significant benefit to aviation security. Added to this are the much-enhanced preventative policing and investigatory efforts that have caught potential terrorists including, in the U.K. in 2006, some planning to blow up airliners.
Combined, we suggest, these measure by themselves reduce the risk of a replication of 9/11 by at least 50%, and this is likely to be a lower bound value. There has been no successful hijacking anywhere in the world since 9/11 and very few attempts at blowing up airliners—and none of these in the United States. In consequence, we suspect, R(pre-boarding security) is likely to be much greater than 50%. Nonetheless, for the present analysis assume R(preboarding security)=50%.
To calculate R(security checkpoints), I used R(pre-boarding security) x Pr(checkpoint efficiency) = 50.00 x 0.05 = 2.5 percent. If TSA agents caught every threat, then checkpoint efficiency (Pr) would be 1.0. Since they failed to catch 95 percent of threats in recent tests, the rate falls to 0.05.
Those are the soft variables it takes to estimate the annual cost per life saved. The final variable is a hard number: The TSA spends $5.3 billion on aviation security annually, about 70 percent of its budget. Let’s say $5 billion, since presumably not all of those funds are directly tied to checkpoint security.
Here’s the equation that Stewart and Mueller use:
Plugging in the variables for security checkpoints comes out to an annual cost per life saved of $667,000,000.
Now, it’s only reasonable to toggle the variables. My assumptions might not be correct, after all. Let’s say everyone at TSA performed admirably. If agents caught every single explosive or weapon at every airport checkpoint, then the annual cost per life saved would still be $33 million: still excessive, at least in terms of federal tolerances. Cut down the cost of checkpoint security from an estimated $5 billion to $1 billion—and assume also perfect checkpoint performance—and the annual cost per life saved is $7 million.
Finally, within the appropriate cost–benefit range! Just not close to reality by any measure involved.
There’s no good reason to think that TSA security is going to get cheaper or more efficient any time soon. As Shirley Ybarra explains in a 2013 report for the Reason Foundation, the agency suffers from a fundamental conflict of interest. “All other aspects of airport security—access control, perimeter control, lobby control, etc.—are the responsibility of the airport, under TSA’s regulatory supervision. But for screening, TSA regulates itself,” Ybarra writes. “Arm’s-length regulation is a basic good-government principle; self-regulation is inherently problematic.”
The Heritage Foundation would like to solve the problem by expanding trusted-traveler programs and engaging private airport screeners. Maybe TSA checkpoint alternatives could bring the costs down a ways. Better technology could also potentially save money. There are a range of possibilities worth considering. In another study, Stewart and Mueller estimate that expanding the Federal Flight Deck Officer program and cutting the Federal Air Marshal Service would “potentially [save] hundreds of millions of dollars per year with consequences for security that are, at most, negligible.”
No matter what, though, eliminating the risk presented by every single passenger on every single flight is always going to be expensive. Especially when the threats that the TSA does detect—take a look at the agency’s Instagram account for a broad sample—are nowhere near a 9/11 in scale. Factor in the loss of productivity associated with security-checkpoint bottlenecks and flight delays, and the cost of total airport security grows even higher. It’s especially expensive when it doesn’t work at all.
To be sure, a 9/11 attack every decade would not be an acceptable price to pay for getting to leave my shoes on at the airport. It’s hardly certain that this is the price we’d actually pay, though. It’s time to consider the real costs and the real risks of flying—and what we’re getting for what we’re spending.
*Correction: This post has been updated to correct a mathematical error: $667 million is two-thirds of $1 billion, not one third. The author regrets the error.