Monday, November 22, 2010

The Cost of Terrorism and Airline Security

Lost in much of the debate on airline security and invasive searches has been a realistic examination of the worth of the costs and benefits of these invasive searches. Given that the last few attempted terrorist attacks, such as the al-Qaeda's Shoe Bomber, or Umar Abdul Mutallab, the Underwear Bomber were stopped in the air rather than by airport security, it's unclear whether or not security has stopped any attacks that Corrections has heard of. Let us leave that aside.

Let's say that airport security could stop all attacks, and without it two Boeing 737's would explode in the mid-air with 175 passengers every year, with no survivors. The average flyer is 42 years of age, and they would lose around of their expected 77.9 year life, or 33.9 years. This is a total 12,565 life-years lost. There are 809,611,003 airline passengers per year, and according to the wait time figures Corrections found, each individual spends about 5-6 minutes in line (between 1 and 25). This is a waste of approximately 7,697 life-years to 9,236 life-years each year.

If this tradeoff is about correct, then a pure utilitarian back-of-the-envelope calculation makes it look as though the time tradeoff from waiting in line alone is about the same magnitude as the cost of death from two Boeing 737's blowing up every year.

For skeptics, the annual budget is $8.1 billion a year. Taking the typical government $6.9 million per life saved, (a statistical measure of the value of human life), it appears as though we are spending the value of 1,173 people each year on the TSA. That is, if the TSA isn't saving 1,173 people per year, the money may be better spent elsewhere.

From these back of the envelope calculations, it appears that everyone would benefit from the government abolishing the TSA and allowing many people to die each year. Further, this assumes the TSA actually has some nonzero impact.

Corrections is interested in comments. Obviously these are back-of-the-envelope calculations and assume away some issues. What first-order losses or gains are we missing?


  1. In addition the TSA operations impose a number of costs. The irradiation of large numbers of Americans - with scanners known to put our more radiation than originally claimed - poses some cancer risk especially for frequent flyers. TSA delays increment travel time - introducing an unpredictable extra obstacle to air travel. More fundamentally though TSA is conditioning Americans to yield their privacy rights, to submit to being groped routinely even though the gropee knows they pose no threat and the TSA gropers are going through the motions. This intrusion is far more negatively impactful against the privacy rights of Americans and their rights not to suffer unreasonable searches and seizures than the much celebrated Guantanamo base sequestering of large numbers of foreign nations apprehended on the battlefield making war against Americans. TSA activities are part of a long term erosion of our liberties that will cause them to disappear with a whimper rather than a bang. If a parent will submit to having their baby or small child's genitalia groped for what the parent knows is not reason, by a government official wearing a uniform, just what limits to inappropriate government intrusion are left? To this point all the attempted Islamofascist plane attacks have been conducted by Muslim males, most of them with recent travel to Yemen, the Sudan, Saudi Arabia or Pakistan - what possible purpose does it serve to grope blue haired grandmothers or children. An intelligent approach to screening focusing more attention on the actual threat group including the question approach the Israelis have found so effective.

  2. Corrections largely agrees with your thoughts on take on the efficiency of statistical discrimination.

    The claim by people who don't like statistical discrimination is that Islamic terrorists will just recruit non-Islamic-looking individuals. It is an open empirical question whether or not this would help Islamic terrorists or Americans more. On the one hand, Americans, both those who would be singled out for profile and those who would not, would likely have faster travel times.

    The pareto improvement Corrections is trying to illustrate is as follows. Say an innocent individual in Muslim garb could make a choice between clearing the line in front of them and being invasively examined or waiting and having a positive (perhaps high) probability of being invasively examined, they may choose the latter. Everyone cleared from the line would obviously also benefit. If this is the case, all Americans benefit.

    Do Terrorists benefit? It depends on the cost and difficulty of recruiting non-Islamic looking individuals. Corrections posits this cost is very high, as the FBI has a vast array of white men willing to go undercover. Corrections posits that "random" recruiting of whites for Islamic terrorism will turn up an FBI informant or agent 80% of the time.

    Even if race isn't used, age crossed with race is likely rather safe to use. As you say, "blue haired grandmothers" and perhaps children, are unlikely to be carrying bombs or be available for carrying bombs.

    It's unclear whether or not we have actually undergone an erosion of rights as interpreted as the ability to freely choose between different activities. This will perhaps be the basis of a discussion in a future post.