For a long time, I’ve had a problem with the “ticking time bomb” hypothetical used by torture apologists. Basically, it is a way to argue against a categorical ban on torture. The argument goes something like this:
Imagine you have a terrorist in your custody who has knowledge of a nuclear bomb in the center of a major American city. He won’t tell you where the bomb is or how to defuse it. So do you torture him to get the information or just let a million people be killed?
The hypothetical has always bugged me because it isn’t a trivial matter to argue against. At the same time, it is clearly a ridiculous hypothetical. And it is evil because the people are not using it to argue that torture is acceptable in extreme cases, but in pretty much all cases. The logic is: if you are willing to torture one person to definitely save a million people, you aren’t completely against torture; therefore, it’s just a question of where you draw the line; thus you liberals want to draw it way out in a case that would never exist, and we want to use it to maybe gain actionable intelligence (even though we can’t point to a single piece of useful intelligence we got through torture).
Two months ago, Conor Friedersdorf provided what I think is an absolutely fabulous counter response to the hypothetical, Torture, Ticking Time Bombs, and Waterboarding Americans:
The thought experiment is no less absurd when applied to “ticking time bombs” that can only be stopped by torturing.
This gets to the very heart of what I think I’ve always known about the “ticking time bomb” hypothetical: it has nothing to do with whether we should torture or not. As Friedersdorf’s example shows, torture is irrelevant to the hypothetical. You could put anything in there. Perhaps a better example would be killing your children. Would you be willing to kill your children to stop a terrorist from murdering a million people? Maybe you would, but that tells us nothing at all about the morality of killing our children.
I highly recommend reading Friedersdorf’s whole article. It isn’t long, but it deals with a couple of other issues that are worth taking note of. One is his discussion of the hypothetical as something straight out of a comic book. As he puts it, “[U]nless Hollywood screenwriters start engaging in murderous acts of performance art, no actual terror plot is ever going to involve a time bomb, a code to defuse it, a collaborator in custody who has that code, FBI or CIA agents who know it, and a waterboarding table on hand.” It’s just sad that the popularity of 24 has made a lot of people think that (1) torture works and (2) such situations are not only possible but common.
 I am not saying that torture does not in many cases provide information. I know, for example, that anyone could torture me and I would provide them with any information they wanted. There are two fundamental problems with this meaning that torture “works.” First, it is almost certain that anyone torturing me would be looking for information that I don’t actually have. So I would make up stuff just to have the torturing stop. Thus, there is a big problem of sifting the good information from the bad. Second, there is a huge opportunity cost with torturing me. In general, rapport building has been found to be far more effective than torturing. I know it would be with me. (I still don’t have any secret information anyone would be interested in, but what are you gonna do?) So the cost of torturing is greater than the gain. Thus, torture does not “work.”