Torture

This op-ed piece, published anonymously in the WaPo, has attracted a lot of attention.

I was prepared to like it better than I actually did when I read it. Some points:

1. I am disappointed that the military is stone-walling this guy's book. It would, of course, not be the first time the military has shot itself, metaphorically, in the foot. It seems to me though, that as a general rule open and honest debate are good things, even in terms of military strategy. Normally, that debate occurs outside the public view (largely due to lack of public interest) but torture policy has aspects beyond just efficacy that should be included in decisions about policy.

2. I think the op-ed goes off the rails a bit in interpreting the claims of individuals being interrogated that it was American torture that led them to join Al Qaeda. Could it be that they are saying this in the hope that it will reduce their chance of being tortured? I am sure that our torture policies had some effect on recruitment but I think the op-ed author reduces his credibility by not explicitly noting that those making these statements have a strategic incentive to do so. This leads, of course, to real issues with the broad claims that are made about loss of life due to US torture policy in Iraq.

3. The authors epistemic stance is basically that his one data point trumps however many data points advocates of torture could bring forward. One problem with this is the same one that plagues evaluation of apparently successful, small-scale employment and training programs: you don't know if it is the technique or the person. If it is the technique, you can scale it up. If it is the person, you cannot. Maybe this fellow is very good at his style of interrogation, and at getting his underlings to do it successfully, but maybe others who tried to emulate his style would not be. As such, a more compelling end to the op-ed would have been a call for the systematic collection and analysis of evidence on the efficacy of alternative interrogation techniques rather than the complete elimination of one techinque based on what is essentially an anecdote.

I liked this interivew piece in the Atlantic a couple of years ago much better - it has, of course, also has the advantage of length, which allows more nuance. Also, my liking it better is not particularly related to the somewhat different policy conclusions of the two pieces. I just think the Atlantic piece is a lot more thoughtful. My policy reviews here remain largely unformed, with the exception that I think the evidentiary bars for policies of either occasional or routine torture are very high given the normative costs, much higher than it appears we have reached at present.