Thanks to Matt for his very thoughtful comments. I agree with almost all of them, so will take this opportunity to amplify on some of the issues he raises.
First, Matt “wonder[s] whether administrative detention is so underdeveloped, or so expansive a concept, that it doesn’t make sense to think of it as a single model at all.” I agree with Matt that the parameters of lawful administrative detention are now underdeveloped. Human rights instruments generally require detaining states: (1) to prescribe in advance the permissible bases for detention and to follow their own laws; (2) to inform the detainee of the reasons for detention; (3) to afford him the opportunity for judicial review; and (4) not to detain arbitrarily. (The framework under the European Convention on Human Rights is more restrictive; instead of prohibiting arbitrary detentions, it prohibits detentions not falling into one of the specified categories.) Based on that formulation, international actors have long accepted that security-based administrative detention is sometimes lawful, but they have failed to provide much guidance on when and under what circumstances it is.
One of my arguments, then, is that the law in this area must be developed in order for it to fill the void for a sustainable detention regime in the fight against terrorism. Even if the law is developed, however, Matt is correct that it would—and, in my view, should—leave some room for state variance. Such variance is not new to human rights law and would not negate the existence of any international legal standard. But it would suggest that a state’s compliance with the international standard be assessed in light of the precise contours of the system it implements.
For instance, I argue that the procedural requirement for judicial review should be understood to mandate prompt, fair, and meaningful review; and that meaningfulness denotes at least three things. First, the reviewing body must have the authority to order the detainee’s release if it determines that detention is unauthorized. Second, the detainee must be equipped to participate in that process and to pursue his rights within it. (That almost certainly requires legal counsel or some other form of independent representation.) Third, the detainee must be informed of the factual basis for detention and be given a genuine opportunity to respond. Even if the international standard on judicial review is developed along those lines, however, states would have some discretion on how best to implement it within their domestic legal systems. In other words, the standard would continue to permit some variance, and whether a particular detention scheme satisfies the standard would be assessed on the facts. Yet, in my view, that variance would not by itself demonstrate a lack of cohesion in the concept of administrative detention.
Second, Matt implores us to focus on the substantive standard for detention. I couldn’t agree more. The standard of non-arbitrariness has been interpreted (outside the security context) to prohibit detentions unless reasonably necessary to satisfy a legitimate government interest. That standard is meaningless in the security context. All governments have a legitimate interest in protecting against serious threats to national security, and the determination that a threat renders detention necessary generally cannot be reviewed by international bodies. I suggest in my piece that security-based detention should be lawful (i.e., not arbitrary) only where the detainee himself poses a serious security threat, where detention is necessary to contain that threat, and where detention is calibrated to last no longer than necessary. Moreover, I argue that the security interests that may justify detention (e.g., disrupting ongoing terrorist activity or developing a more considered criminal case) give way, over time, to the liberty interests against detention. I thus suggest containing the duration of detention by requiring states to satisfy increasingly stringent evidentiary standards if they seek to hold suspects beyond incrementally set periods. After a certain period, all or almost all detainees should be released, deported, or criminally prosecuted. I underscore, however, that I intend for those suggestions only to begin the conversation. I very much appreciate recent efforts—including by Matt—to address in greater depth the appropriate substantive standard for detention.
Finally, Matt asks the difficult policy question of whether states should implement systems of administrative detention. I myself am torn on that question. Nevertheless, I do think it must be answered cognizant of the alternatives. What are the alternatives to a better regulated system of administrative detention? One may be expanding domestic criminal law in ways that contaminate the justice system. Another may be hamstringing state officials against preventing terrorist attacks. Still other, even less palatable alternatives are apparent in state practice: Several western democracies have “dealt with” non-battlefield terrorism suspects by participating in U.S. detentions based on the law of armed conflict; deporting terrorism suspects, despite the risk of home-country mistreatment; and detaining administratively without meaningful controls. If those are the alternatives, then a regulated system of administrative detention may well make sense.