For longer than I care to admit, I’ve been working on an economic theory of suicide. (If you actually read the paper, I encourage you to skip the empirical section, which still needs lots of work.) Other economists have tackled the issue of suicide, though the field is still dominated by psychologists and sociologists.
Oddly, no economic papers that I know of directly address the question of how suicidal persons choose their suicide methods (though some have addressed it indirectly, by modeling suicidal persons as choosing their probability of death). Yet the choice-of-methods question lies at the center of public policy debates. For instance, would restricting access to guns reduce the number of suicides? Gun control advocates usually say, “yes, obviously,” while opponents say, “no, because people who want to kill themselves will just find another way.” Who’s right? I say neither.
The innovation of my approach is to treat the suicidal person as engaging in a search for suicide opportunities, akin to a job-seeker’s search for job opportunities. Suppose a job seeker gets a lousy job offer. Should she take it now, or should she turn it down and continue the search? Naturally, the answer depends on how likely it is that she’ll get a better job later. The lower the chances of a good offer later, the more likely she is to settle for the lousy offer now.
Similarly, a suicidal person can be characterized as engaging in a kind of passive search for opportunities to die. Suppose he has the chance to kill himself by a non-preferred method – say, overdosing on pills. But he would prefer to kill himself with a gun (over 70% of male suicides use guns). Whether he’s willing to use the pills now depends, among other things, on his chances of getting the opportunity to use a gun later. The lower the chances of a gun later, the more likely he is to use the pills now.
Counter-intuitively, the argument indicates that a policy decreasing the availability of a suicide method (like guns) could actually increase the suicide rate. But to be fair, the model’s predicted effect is ambiguous – the suicide rate could go up, go down, or stay the same. On the one hand, suicides increase as people substitute into less preferable but more readily available methods. On the other hand, suicides decrease because they will have fewer opportunities to use their more preferred methods. Which effect predominates? A priori, we can’t say. It depends on the strength of the subject’s preference for one method over another, the magnitude of the drop in availability in the preferred method, the availability of less preferable methods, and so on.
Past studies have tried to find a statistical connection between availability of guns and suicide rates, with little success. Some show a connection between guns and gun suicides, but few if any show a connection between guns and total suicides. What’s going on? The natural response is the old “people who want to kill themselves will find a way” theory. But that theory is implausible, because plenty of evidence shows that suicidal persons have strong preferences over suicide methods. The notion that people will casually switch from one method to another relies on the assumption that they regard different methods as perfect substitutes. I don’t buy it.
My theory provides an alternative explanation for the absence of a statistical connection between guns and suicide – an explanation that depends upon, instead of denying, people’s preferences about methods. People do prefer some methods over others, but they will expand the set of methods they’re willing to use at the margin in response to changes in perceived availability. As a result, any change in availability creates offsetting effects that will at least partially cancel out.
In the future, I hope to blog about the applications of my suicide model to terrorism. But since my blogging stint here at the VC is almost over, you’ll just have to visit Agoraphilia to get that part of the story. I want to thank Eugene for the invitation to blog here – it’s been great fun, and the feedback has been excellent.
Comments are closed.