The Breakfast Table

The Proper Error Rate Is Never Zero

Dear Dahlia, Cliff, and Walter:

I think Dahlia and I have hit on the hardest and most important policy question in the design of a detention system for dangerous terrorists: how the system should strike the balance between the erroneous detention of the innocent and the erroneous release of dangerous terrorists. The bad guys do not wear uniforms or dog tags, so the chance of mistakes is higher than in a traditional war and so, too, is the cost of those mistakes, which is potentially indefinite detention. On the other hand, the average terrorist today is much more dangerous than the average soldier in World War II or the average criminal. So mistakes in either direction are really bad.

In analyzing Justice Scalia’s claim that Americans would die because of the Boumediene decision, I said that “higher standards of judicial review designed to minimize false positives in military detentions will likely produce false negatives that mean more Americans will be killed than would otherwise be the case.” My point was that we shouldn’t flinch from recognizing (as I probably incorrectly took you to be doing) that higher standards for detention that better protect innocent detainees will at some point likely result in the deaths of the innocent attacked by terrorists erroneously released. 

You respond that it is a “fundamental legal principle” that we “protect the rights of ‘false positives’—i.e., the innocent.” I agree. But this “fundamental” principle is not an absolute one. We might be able to design a procedural system that would guarantee no mistaken detentions (I doubt it), but that system would come at the very high cost of allowing many mistaken releases of dangerous terrorists. Every system of detention faces this quandary. The criminal law system’s “beyond a reasonable doubt” standard is commonly said to be premised on the idea that it is better for 10 people to go free than for one innocent to be convicted. This means that the criminal system tolerates the conviction of innocents (as many DNA cases have shown), and we build many layers of legal doctrine to privilege the finality of those convictions over the ability of innocent convicts to vindicate themselves. We do that because at some level, we recognize that the proper error rate, even for criminal justice, is not zero.

If our criminal law system is not and is not designed to be perfect, then we should not expect our military detention system to be, either. This is why I disagree with you that recognizing the possibility of mistaken detentions in this system constitutes a “willingness to sacrifice both those nondangerous prisoners and those fundamental principles that frustrate liberals most.” Even liberals, I assume, would not want to design a system that ensured no mistaken detentions, for that system would be more demanding than our criminal law system and would involve the release of many very dangerous people. Am I right about this?

If I am right, then the hard question is where and how to draw the line: What kind of risks in both directions are we willing to assume, and which procedural system best gets us to that point? I think we both agree that the system in place before Boumediene was not up to the task, though I suspect we disagree on how to strike the balance going forward. One point I was trying to make in my last post is that while courts undoubtedly have a role to play in monitoring and legitimizing the military detention system, the political branches are best positioned to decide the hard liberty and security trade-offs that the system tries to guarantee. They are best positioned because they have more information about the risks of terrorism, weighed against the risk of being too aggressive against terrorism, and because they face the electorate if they get that trade-off wrong. Whether the political branches—in particular Congress—have the political desire to exercise their duties and assume these responsibilities is another question altogether.

This might be my last post as tomorrow’s decisions will move us on to terrain beyond my expertise. Thanks for letting me join in!

Jack