One of the problems with discussing torture as an interrogation technique is that it is almost impossible to get any sort of objective analysis of its efficacy. Torture is usually done in secret, and even declassified reports don't necessarily give the whole picture. Attempts at scientific study are therefore done second-hand using an incomplete data set. Whenever I see a purported study of torture, I rhetorically ask, how many people did they torture? Experts, most of whom don't actually qualify as experts on torture, tend to see things from their own biased perspective, and often ignore contrary evidence. Obviously we can't torture people just to study the results. Or can we?
Since waterboarding is usually physically harmless, it is legal to perform if someone agrees to it and signs the necessary releases. Several journalists have allowed themselves to be waterboarded, including Christopher Hitchens. Here's my proposal for a simple test of the efficacy of waterboarding at extracting accurate information.
Assemble a volunteer group of test subjects. Each subject would be paid for participating. A time limit would be set such as one week, or whatever was medically safe, as would the number of waterboarding sessions per day. Before each session one team of researchers would supply the subject with an easy to remember password. The interrogation team would have a password protected laptop for which they did not know the password, which would be known only by the subject. (It would be changed each time for each subject). The interrogation team would use waterboarding in an attempt to extract the correct password, and the subject would attempt to resist.
Here's the twist. To encourage resistance the subjects would be paid an additional fee every time they successfully held out. The person resisting the longest without giving up the correct password would receive a huge prize, say ten million dollars. If more than one person made it to the time limit they would split the prize. A reward of that magnitude would be the closest possible simulation of a life & death need to protect information.
The results of such a test would be meaningful only for waterboarding, and only for its effectiveness in obtaining simple, easily confirmed information. But they'd still be a more objective measure of a torture-technique's effectiveness than anything else that's out there.
It might also make a good reality television show, say "Torture Island."