An article on the front page of Saturday's NY Times examined models for President Obama's plan to hold a small number of terrorism suspects for "prolonged detention." These include: quarantine of people with infectious diseases; pre-trial detention of criminal suspects; and preventive detention of people who are mentally ill and dangerous or sexually violent predators. However, as I was quoted saying in the article: "We have these limited exceptions to the principle that we only hold people after conviction . . . but they are narrow exceptions, and we don’t want to expand them because they make us uncomfortable.” And why do they make us uncomfortable? Because they violate a presumption of liberty, the core notion that people should enjoy the most basic freedom---freedom from confinement---absent some very good reason.
We could go further and say that the more the basis for confinement looks like a fear of criminality, the more uncomfortable we are (or should be) about confinement in the absence of proof of a past crime. Thus, quarantine is probably the least problematic form of detention without proof of guilt precisely because it is conceptually so distant from criminality. Quarantine could in principle be abused by the state, but proof of ebola or some other terrible disease is unlikely to be used as a short-cut around proving guilt beyond a reasonable doubt. The other grounds for detention without proof of guilt are harder because the harm we fear looks a lot like crime.
Thus, it is important to ask just how the people who will be eligible for prolonged detention differ from ordinary criminals as to whom the government wants to take a short-cut. Is it the nature of the acts we worry they will commit that justifies holding people for "prolonged" periods even without a conviction by a civilian or military court? If so, would someone like Timothy McVeigh have been eligible for prolonged detention in the event that the government determined he could not be tried and convicted?
Under the President's proposal, the answer is pretty clearly "no," but not because McVeigh posed a lesser threat than the terrorism suspects now at Gitmo. (If you think he did pose a lesser threat, imagine a nuclear-armed McVeigh). McVeigh would be treated differently, I think, because the best model for prolonged detention is the detention of prisoners of war during a very long military conflict, and the Gitmo detainees look a lot more like POWs than McVeigh does.
The Bush Administration didn't want to give Gitmo detainees POW status because that would have precluded interrogating them. However, even if they had been given or were now given POW status, we would still have a puzzle: Because they do not fight for any state that can surrender or sign an armistice, how will we know when hostilities are over? The Bush theory was that we would never know when the war was over, and therefore we could hold them indefinitely. As I understand the Obama plan, it is different in one important particular: There will be periodic (perhaps annual?) review of whether individual detainees need to remain detained. Presumably, after the initial combatant status determination has been made, such review will focus on whether there is sufficient reason to believe that the detainee, if released, will continue to pose a threat to the United States.
That does indeed seem like a better approach than indefinite detention simply on the President's say-so, but it's not clear that it will lead to a very different result. Consider that persons involuntarily confined as mentally ill and dangerous are entitled to periodic review of their condition, at which the government continues to bear the burden of proof by clear and convincing evidence that they are ill and pose a threat. The key testimony at such hearings typically comes from the state psychiatrist, and it is highly unusual for a judge to release someone who, in the judgment of the state psychiatrist, needs to remain confined. How likely is it that the review system planned by the Obama Administration will lead to a different pattern?
Posted by Mike Dorf