Published by The Lawfare Institute
in Cooperation With
Is it legal for the U.S. government to hold an American citizen in military detention in Iraq if that person was captured in the Syrian combat zone and was a fighter for the Islamic State?
That’s the fundamental legal issue at the heart of Doe v. Mattis, a habeas corpus case pending before Judge Tanya Chutkan in D.C. federal district court. It’s the sort of case that used to draw headlines; not so much anymore, it seems. Yet it’s a terribly important case, and worth your attention.
I previously summarized the emerging factual dispute in this case. Basically, the government contends that John Doe was a vigorous advocate for the Islamic State, went to great pains to get to Syria to join it, and served it in various capacities (including bearing arms). For his part, Doe apparently will contend that he was a freelance journalist who went into Syria to report on what was really happening inside the Islamic State, who was coerced into serving it, and who did his best to escape when the chances came. For the moment, though, Doe is not pressing his factual claims. He is focused initially on convincing the court that there is no legal authority to detain him even if the government’s account is true.
By the time the legal dispute is fully briefed, there will be five relevant documents filed by the parties. Four are available now.
It starts with the habeas petition itself, and the government’s return. Then there is Doe’s response to the legal positions the government asserted in the return. Next, the government filed its reply.
There will be one more filing on March 14, when Doe submits his final response. I will post something about that when it happens and will flag any significant new arguments. That said, we have more than enough clarity at this point for me to sum up the essential issues.
I am not aiming in this post to precisely catalogue the back-and-forth in the briefing. Instead, I’m trying to distill the most important disputes, place them in a logical order, and give my preliminary views on which arguments are most persuasive. No doubt both parties would feel I’ve left out some key aspects of their respective briefs.
Okay, on to the summary of the issues.
1. Is the case ripe?
The government concedes, as it must, that Judge Chutkan has jurisdiction to hear Doe’s petition for a writ of habeas corpus. But it makes a threshold argument that she should hold off from exercising that jurisdiction for the time being—basically, asserting that the case is not actually ripe.
The underlying theory appears to be that the government is still unsure what path it ultimately wants to take with Doe. Maybe they will decide to just hold him in military detention, yes, but maybe instead they will cut a deal to transfer him to Saudi Arabia or Iraq, or perhaps they will bring him to America for prosecution. (Note: Military commissions do not have statutory jurisdiction to prosecute U.S. citizens.) On this view, the court should stay out of it until the government makes up its mind.
This is a bad argument. True, Justice Anthony Kennedy in Boumediene v. Bush observed (quite properly) that a habeas court confronted with a military detainee captured overseas should not intervene “the moment” jurisdiction attaches. That is an important principle that I very much embrace. But it obviously is meant to be limited. And though there is ample room to disagree about precisely when this initial period should end—Steve Vladeck and I disagreed on precisely this point during the early weeks of the detention last fall, across many episodes of the National Security Law Podcast—it is preposterous to claim that the period can extend more than half a year.
With the passage of so much time, this is no longer even a close call.
So, on to the merits.
2. What is the essence of Doe’s legal objection?
Doe’s claim that the government lacks authority to detain him rests on a simple pair of propositions. He must prevail on both:
First, he asserts that an American citizen cannot lawfully be detained without criminal trial unless there is a clear statutory grant of authority by Congress for that purpose.
Second, he denies that there is a clear statutory authorization relevant to his particular circumstances.
Let’s unpack those claims in order.
3. Is there really a rule against detaining a U.S. citizen without statutory approval?
There is indeed. Doe points to two separate sources for such a rule.
First, Doe says this rule derives from the Constitution. Specifically, he says it is compelled by the Fifth Amendment due process clause. Citing Hamdi, Salerno, and Ex parte Endo, Doe argues that due process doctrine requires Congress to approve non-criminal detention of U.S. citizens through a clear statement in statute. (This line of argument resonates with the more general proposition that the due process clause embodies the Magna Carta principle that detention must always rest on the actual “law of the land” rather than on executive discretion.)
Is this argument persuasive? It is clear that the due process clause is at least concerned with clarity. Endo, for example, repeatedly emphasizes the importance of clarity when considering a claim of implied power to detain a citizen. In that case, the court concluded that in order to avoid constitutional difficulties, “any such implied power must be narrowly confined.” But note that the court did not articulate a bright-line rule outright forbidding any implied detention authority. It made clear such implications are disfavored, and strongly so. That’s a bit different, and it’s why Doe’s alternative argument matters.
Doe’s alternative argument is that Congress has passed a statute directly imposing a requirement that all non-criminal detention of a U.S. citizen be justified by a statute. In 1971, Congress enacted the Non-Detention Act (18 U.S.C. §4001). It states that: “No citizen shall be imprisoned or otherwise detained by the United States except pursuant to an Act of Congress.”
So far so good for Doe: The Non-Detention Act plainly requires a relevant statute to justify non-criminal detention of citizens. But note that the text leaves open an important interpretive question: just how clear must the statutory source of the detention authority be?
That is a critical question in this case, for the statutory authorities on which the government relies to detain John Doe (discussed below) do not expressly state that citizens may be detained. Instead, they are statutes that expressly authorize the use of military force, without being particular about what sorts of force might be used and without saying anything directly about the view of Congress regarding whether citizens are included within their scope.
Doe argues that the Non-Detention Act should be read strictly so as to exclude reliance on implied grants of detention authority. The government counters that a more flexible rule is needed at least for war-related detention. Let’s turn to that argument and look at it in detail.
4. Should the Non-Detention Act be read to allow implied grants of authority to detain in the context of war?
The idea that the Non-Detention Act should be read to allow for implied grants of detention authority in war-related cases faces a significant obstacle: The Non-Detention Act was created in no small part to foreclose a recurrence of Japanese internment. That being the case, how can the government argue for a flexible reading here?
That’s a strong argument at first blush. But there is a considerable difference between imprisoning a vast body of American citizens based on nothing more than national origin as a proxy for possible disloyalty and detaining combatants for the enemy who were captured in a warzone. That the Non-Detention Act plainly meant to foreclose the former does not mean that it also forecloses the latter, where an enemy combatant turns out to have U.S. citizenship. (Remember: Doe at this stage is conceding for the sake of argument that the government’s account of his actions and intentions are correct; this analysis therefore treats him as a willing Islamic State fighter captured in Syria.)
One way to operationalize this distinction is to note that the law of armed conflict does not authorize or permit racial roundups like the World War II internment of Japanese-Americans, but it does authorize or at least permit detention of captured combatants for the duration of hostilities. Indeed, this appears to be the theory of the plurality in the Supreme Court’s 2004 Hamdi decision, which dealt with this exact Non-Detention Act question in a closely analogous factual situation.
Yaser Hamdi’s situation was much like that of Doe: Both were born in the U.S. of Saudi parents, had dual U.S.-Saudi citizenship, lived their adult lives abroad, were captured in an overseas combat setting by U.S.-allied local forces, and then turned over to the U.S. for overseas military custody. (In Hamdi’s case, he was said to be a Taliban fighter, captured by Northern Alliance forces in Afghanistan.)
Faced with a Non-Detention Act claim, the plurality concluded that the 2001 authorization for the use of military force—though it said nothing expressly about detention of anyone, let alone American citizens—provided sufficient statutory authorization under the Non-Detention Act because the AUMF implicitly included authority to detain combatants in accordance with the law of armed conflict. If Hamdi was indeed a Taliban fighter, then, he was detainable despite his citizenship.
So what follows from this? The government wins, if the Hamdi plurality’s approach controls, and if the court concludes that the Islamic State today is within the scope of the AUMF in the same sense as was the Afghan Taliban in 2004.
Let’s turn now to that latter question: Is the Islamic State within the scope of a relevant AUMF?
5. A reminder that this is a familiar debate
Lawfare readers know that this question has been a source of controversy for many years, ever since the Obama administration advanced the claim in order to explain why it did not need a fresh statutory authorization in order to fight the Islamic State in Iraq and Syria. I will provide only a thumbnail sketch of that debate here.
The Islamic State once was the organization known as al-Qaeda in Iraq (AQI). While the U.S. military battled AQI in Iraq prior to 2009, there was relatively little doubt that AQI came within the scope of either the 2001 AUMF applicable to al-Qaeda and the Taliban, the 2002 AUMF specific to Iraq, or both. AQI, after all, seemed the quintessential example of an al-Qaeda “associated force” engaged in hostilities against the United States.
Here’s where it gets tricky: The U.S. left Iraq for a time, ending its combat role there. And AQI in the interim broke with al-Qaeda’s senior leadership after a disagreement over control of the Nusra Front in Syria. AQI rebranded as the Islamic State and metastasized rapidly and horrifyingly. The United States eventually reentered the picture to resume combat against it.
Against that backdrop, the relevant question is whether the two AUMFs continued to apply all along and to this day, or whether instead their relevance terminated on or after either the American withdrawal or the Islamic State’s break with al-Qaeda.
Doe notes that many observers (myself among them) publicly expressed great skepticism when the Obama administration first advanced its theory that the 2001 and 2002 AUMFs remained relevant and applicable to the Islamic State. I was indeed skeptical at the time. The more I have learned about the continuity of AQI with the Islamic State, however, the less skeptical I have become. On reflection, moreover, I’m troubled by the notion that the intramural squabble within the al-Qaeda network that led to the Islamic State break should have the immediate legal effect of exempting Islamic State from the scope of the AUMFs (assuming that Islamic State as AQI otherwise would have remained within scope). These considerations, combined with the principle that courts should afford a measure of (non-binding) deference to the executive branch when it makes a well-considered determination of factual and policy judgment in such a significant matter of armed conflict and foreign affairs, lead me to think that the court should accept the government’s position.
6. Must the court defer to the executive branch on this question, even if it otherwise would not agree that the AUMFs apply to the Islamic State?
Let’s assume Judge Chutkan would not agree with the government, all other things being equal. Might she be duty-bound to give some degree of deference to the government’s interpretation of the statutes, perhaps even a binding degree of deference?
In national security cases, one certainly expects the government to invoke deference principles in relation to both factual judgments and legislative interpretations—usually citing a constellation of familiar precedents while sounding themes of both comparative institutional competence and comparative institutional legitimacy. (If you really would enjoy a deep dive into such things, you will perhaps enjoy this paper.) And we certainly get that here.
The government suggests that interpretations of the organizational scope of the AUMFs—mixed questions of law and fact, perhaps—might well constitute a political question that courts should not attempt to resolve. This is a bit hard to square with some of the GTMO habeas litigation, however, for there have certainly been cases in that vein that wrestled with whether groups (such as Hezb-i-Islami Gulbuddin) should be read to fall within the scope of the 2001 AUMF.
Things get more complicated if one recasts these arguments in terms not of the political question doctrine as a trump card but rather as a simple argument for a degree of deference to be afforded the executive branch’s well-considered estimation of the Islamic State’s status in relation to al-Qaeda over time. (That’s not really how the government has framed its argument, mind you, but it’s how the argument should have been framed and needs to be addressed by the court (assuming the court rejects the bid for binding deference under the political question doctrine)). In short, the real deference issue is not whether the court should just accept whatever the executive branch says, but rather whether the court should afford some additional weight to the executive’s assessment (which may or may not be enough to dispose of the issue).
(At this point, I want to give a shout-out to whomever at the Justice Department came up with this passage: “that al Qaida and ISIL now manifest themselves as two heads of a modern day Leranaean Hydra should not undermine the Executive’s ability to battle both heads to defeat the beast.” Feb. 28 filing at p.10).
7. Has it been a mistake all along to read the 2001 AUMF to include “associated forces”?
Doe has a fallback argument should he fail in his bid to have the court conclude that the Islamic State is outside the scope of the AUMFs: Perhaps it has been a mistake all along for the executive branch to claim that the 2001 AUMF encompasses any “associated forces” of al-Qaeda and the Taliban? If so, then the question of succession from AQI to the Islamic State becomes moot; AQI would not have properly been within the scope of the 2001 AUMF to begin with.
The problem with this argument is that Congress, the courts, and the executive branch have all rejected it. Both the Bush and Obama administrations interpreted the 2001 AUMF this way, it was accepted during the GTMO habeas litigation, and then in the National Defense Authorization Act for Fiscal 2012, Congress expressly codified it.
If Doe were a noncitizen, then it seems to me this argument could go nowhere at this point. But, Doe is a citizen. That adds an important wrinkle, for the 2012 NDAA also went out of its way to say that it takes no position on the more-general question of whether AUMF detention authority extends to citizens. One might therefore argue that it is not proper to cite to any aspect of the 2012 NDAA (such as its general codification of the associated forces category) in any case involving a citizen.
That’s an interesting argument, but it is not obvious that the right way to read the NDAA’s agnosticism about citizen detention is to therefore also treat all other parts of the NDAA as irrelevant any time a citizen detainee case arises. It seems more logical to me to keep those questions separate, reading the NDAA as always controlling as to the list of organizations within the scope of detention authority while still being neutral as to whether a citizen member of any such group is subject to detention authority.
8. Has Congress ratified the Obama-Trump extension of the AUMFs to the Islamic State?
Another question to watch in this case is whether any doubts about the AUMFs reaching the Islamic State have been settled through congressional ratification, in light of years of intense congressional involvement in support of the ongoing war effort there.
As an initial matter, there is a significant question about the relevance of ratification-through-funding. Not everyone accepts the potential relevance of war funding to such questions (and for whatever it is worth, the War Powers Resolution does try to preclude such arguments). Having said that, let’s assume it might have at least some relevance in some cases. Is this a good case for this sort of argument?
On one hand, the AUMF-scope question has been very visible for many years, and throughout that time, Congress has provided sustained support for the war effort. On the other hand, there have been a few fruitless efforts to pass a new AUMF that would settle this issue by expressly naming the Islamic State. The existence of those efforts at first might seem like some evidence that Congress does not agree with the executive that the existing AUMFs encompass the Islamic State. Both White House and congressional proponents of these bills have tended to describe them as clarifying existing authority rather than providing an otherwise-absent authority, however. Bearing that in mind, the ratification argument here is a bit muddled but reasonably strong (for those who are open to ratification-by-spending at all).
9. One final thing: What if the court concludes that the Non-Detention Act’s requirement of statutory authorization just isn’t satisfied here?
In that case, the analysis must go on to consider a final and weighty question: Does Article II of the Constitution preclude application of the Non-Detention Act to an enemy combatant captured in the context of armed conflict? That is, does the commander-in-chief authority override the statutory rule? The government has raised this possibility, urging the court to construe the Non-Detention Act in a way that will avoid the need to confront such a serious issue. Needless to say, that argument is intensely controversial, harkening back to similar arguments that arose during the early post-9/11 period (in relation to matters such as interrogation and surveillance). It has not been thoroughly developed in the briefing to this point, but should the court decide the Non-Detention Act has been violated here it will not be possible to avoid it entirely.
I will post again when things develop significantly further on these issues.