The efficient level of torture

Cite this Article
Todd Henderson, The efficient level of torture, Truth on the Market (March 05, 2010),

Every morning on my 1-mile drive to work, I pass two signs expressing outrage about torture – one is a hand-made yard sign, the other an ominous black banner hanging from a church window: “torture is wrong.” (Yes, punctuation by e.e. cummings it seems.) Is it? I’m not sure.

The optimal amount of torture is certainly not zero. Only a zealot would claim otherwise. The simple law-school hypothetical of the ticking time bomb shows the absurdity of the claim: if a nuclear bomb were known to be about to go off in Chicago, and if waterboarding were known to be an effective method of extracting information, and if there were no other way of getting the information (three big “ifs,” I admit, but don’t fight the hypothetical), everyone making the decision at the time would torture. Everyone would prefer to scare someone (not even seriously hurt them) to save millions. So we might quibble with the “ifs,” but this isn’t denying torture as a matter of first principles, rather based on details that are quite contestable. For instance, several high-level intelligence figures here and in other countries claim demonstrable success using these techniques. The debate is described here.

Although not zero, the optimal amount of torture may be small, even extremely small. Torture is thus like killing. Everyone agrees the optimal amount of killing other humans is not zero. I can kill an intruder who enters my house and threatens me; the police can kill under limited circumstances; the state (or at least some states) can kill heinous criminals; and the federal government can kill pretty much at will, sometimes massively and indiscriminately (e.g., the firebombing of Dresden), to protect us from perceived threats. While individuals may differ about the wisdom, efficacy, or legality of some of these, only the most idiosyncratic of us would consistently reject any killing of any kind, especially when the connection with human flourishing is established.

To this point about large-scale violence being perpetrated in our name, I find it odd there are no signs on my way to work reading: “Predator attacks are wrong,” or “cluster bombs are wrong,” or any of the other, far more lethal things we do are wrong.  One might say that torture is worse than killing because the latter is often done in “the heat of battle” or when there are no alternative choices because of an imminent threat. There are two problems with this line of reasoning. First, there are millions and millions of graves filled with the bodies of those killed in cold deliberation and where the threats were hardly obvious. Was the murder of tens of thousands of German civilians necessary to impede the Reich’s war production efforts? Perhaps. But, perhaps not. Isn’t it odd to say that we can’t slap a terrorist in the hope of intimidating him into confession but we can kill that same person (plus his entire family) if we use a Predator drone to attack his house in Waziristan? Ah, you might argue, if he is free, he is a bigger threat than when in custody. This is the second problem with this argument. Lawyers call this fallacy the act omission distinction. Is the act of setting a bomb (when free in Waziristan) really more threatening than the omission of telling us where a bomb is set (when held in Bagram Air Base)?

A counter argument to this analogy between killing and torture is that the greater does not always include the lesser – the fact that we can kill does not mean we can do things less than killing. Take the example of animals. Killing animals is socially acceptable – about 9 billion were slaughtered last year in the US for our enjoyment. But torturing animals is generally not socially acceptable. Professional football player Michael Vick served nearly two years in federal prison for his involvement in dog fighting. Why is this? Why doesn’t the right to kill include the right to torture?

To get at this, it is important to remember we do torture animals. We use animals not only when we kill them “humanely,” whatever that means, but we also use them when they are subjected to treatment that is akin to torture. A college friend of mine experiments on pigs with the hope of developing more efficacious heart surgery techniques – the pigs are made sick, subjected to numerous painful surgeries, and then killed. Another friend does research on the brains of chimps – again, these chimps are not happy about this. Why is this form of torture OK (I understand to some it is not), and yet the use of dogs for entertainment (put aside dog racing, if you can) is not?  It must have something to do with the end results. My friends are trying to save human lives and they are acting as compassionately as they can in dealing with the animals on which they experiment. Michael Vick, on the other hand, wanted a cheap thrill and acted like an animal himself. In other words, we punish not because of the impact on the animal, but because of a cost-benefit analysis of the impact and a view about what the conduct tells us about the perpetrator. If your goal is one that is important and the costs are minimized, torture is OK.

This distinction lets us draw a sensible line in the current debate about torture. What rogue soldiers did at Abu Grahib looks like it was not part of a plan, not well designed, not managed, not intended to resolve an imminent threat, and more revealing about the nature of the people doing the torture. In short, the perpetrators look more like Michael Vick than my medical researcher friends. The case of the waterboarding of high-level al Qaeda operatives, in contrast, looks more like the latter. Sure it makes us uncomfortable to think about our government doing this, but doesn’t it make us uncomfortable to think about the pigs and the chimps, not to mention the children incinerated in Dresden? Our top intelligence officials believe “coercive interrogation methods,” gave us “deeper understanding of the al Qaeda organization that was attacking this country.” Why don’t we believe them?

Perhaps we do. My hope is that the compromise we’ve reached is to publicly condemn torture but to privately signal that in extreme cases we will forgive those who do it. This is probably the best legal regime to handle something like this. The banning of torture is good PR, but more importantly, it puts the onus on potential torturers to make sure that when they do torture, it is for the extreme, ticking-time-bomb case. In other words, the legal uncertainty about torture means we will have less of it. If we believe the optimal amount of torture is less than the amount before the ban, this should be expected to lower it. But, the use of pardons after torture is revealed should provide sufficient protection for those who are certain under the circumstances that torture is the right decision. Of course, for this to work well – to generate the efficient level of torture, if you will – the public needs to have the debate in a sensible way. All the hand wringing and political posturing is not helping, since it likely gives our intelligence officials less comfort about whether a pardon would be forthcoming, even in extreme cases. Better we face the cold, hard realities of the world with a pragmatic view, than to simply condemn something because it gives us the willies. Killing animals or children gives me the willies too, but sometimes, tragically, it must be done.