Suicide Bombed

In my last post at Volokh, I promised to extend my economic theory of suicide to the case of suicide bombers. Over a month later, I’m finally getting around to it.

The story so far: I model the suicidal person as engaged in a search for suicide opportunities, akin to a job search or product search. The agent ranks the possible methods in order of desirability (e.g., guns are preferred to pills, pills are preferred to drowning). But some methods are more readily available than others, and waiting for a better method to come along means enduring a bit more of life’s misery. The model counter-intuitively predicts that decreased availability of a preferred method could increase the frequency of suicide, because the agent becomes more willing to employ the less-desirable but more-readily-available methods. (The job-search analogy: A worker is more likely to take a lousy job offer now if he’s unlikely to get a good job offer in the future.)

As I originally conceived the model, all the utility values are negative: life sucks, shooting yourself sucks, etc. The agent just tries to minimize the sum of suckiness. But actually, all the math goes through just as well with positive values. Imagine a would-be terrorist. He doesn’t necessarily hate his life – it has a net positive value – but he also anticipates great utility from the killing of infidels, enough so that he might be willing to sacrifice his life. The more dead infidels, the greater the utility. This provides a natural ranking of his suicide opportunities: a planeful of victims is preferred to a busful, a busful to a carful. But some opportunities are more readily available than others. A 9/11 takes months of preparation (at least), and only a few hijackers. Opportunities to bomb buses arise more often.

The conclusion? If increased security diminishes opportunities for “major” terrorist acts, the result could be increased frequency of “small” terrorist acts. Would-be terrorists, faced with fewer chances to commit big atrocities, become more likely to commit small ones. The total number of terrorist acts or victims could rise or fall. Example: The U.S. had (maybe still has) lousy airport security. We got 9/11. Meanwhile, Israel has had its airport security tight as a drum for decades. They get countless truck bombings, car bombings, restaurant bombings, etc.

I know, I know… that’s purely anecdotal evidence, and I wouldn’t claim otherwise. The U.S. and Israel differ in numerous important ways other than their airport security, including their size and their distance from the terrorist breeding grounds. It would be difficult, maybe impossible, to do a true controlled study. So there’s no proof here, but I think the comparison is suggestive.

Does the theory indicate that increasing airline security is a bad idea? Not necessarily, since the total effect is ambiguous. But at a minimum, we shouldn’t simply assume that increasing security in any given area will have reduce the incidence terrorism. If I worked for Homeland Security, I would anticipate an increase in the number of small-scale terrorist attacks that don’t involve airplanes.