I just finished reading and absorbing Medellin today. I mentioned the case several times in my Constitutional Law I class in the fall, and students seemed quite intrigued by the interesting fact pattern and issues presented by the case.

Which leads me to the following question: Can readers think of a good place to teach Medellin in an introductory Constitutional Law class? At GW, our required, introductory Constitutional Law focuses on structural issues (federalism and separation of powers). That means that one large part of the class is spent on the Commerce Clause/Spending Clause/Tenth Amendment, and the other large part of the class tends to focus on Youngstown/war on terror cases/Chadha/executive privliege.

Where can you fit Medellin in? As part of the Youngstown discussion? Can you make the self-executing issue part of a Constitutional Law course? Have others thought of teaching Medellin in a Constitutional Law course, or have they perhaps taught Sanchez-Llamas in their classes before?

[William Dodge is a law professor at UC Hastings]

I’ve not seen anyone comment yet on what I thought was one of the more notable aspects of Chief Justice Roberts’s opinion, its application of Justice Jackson’s Youngstown analysis. The question is how to read congressional silence. Although I am greatly oversimplifying, Jackson seemed to read Congress’s failure to authorize what the President did in Youngstown as implicit disapproval (category 3). In Dames & Moore, by contrast, Justice Rehnquist read Congress’s silence as approval (category 1). Roberts’s opinion in Medellin seems to follow Jackson rather than Rehnquist and to treat silence as implicit disapproval (category 3).

As chance would have it, Rehnquist clerked for Jackson the term that Youngstown was decided and Roberts clerked for Rehnquist clerked the term that Dames & Moore was decided. I’m sure there is a great law review article to be written here (though I am not the one to write it).

I wanted to jump in with a quick response to Marty’s awesome post about what is, without question, the most surprising (and for me, delightful) part of the Medellin opinion: Justice Stevens’ concurrence.

I love this concurrence, especially because I am (to put it mildly) rarely fond of Justice Stevens’ forays into foreign relations law. This is, after all, the author of both Rasul and Hamdan. But give the Court’s senior justice his due: Justice Stevens seems to be genuinely constrained by his view of the law, and not his very clear policy preferences (to rule for Medellin). This may be true of the other Justices, but it is much harder to tell.

So here’s my take on his thinking: Article 94 does not, in his view, require the United States to comply with the ICJ ruling. All it has to do is “undertake to comply,” which in his reading, is not a mandatory obligation. So the Supremacy Clause is imposing a non-mandatory obligation on the U.S. here and neither authorizes the President’s action nor requires Texas to do anything. All Texas is obligated to do is whatever the U.S. is obligated to do, which is to “undertake to comply,” which is non-mandatory.

To put it another way, the “obligation” here for Texas is one of upholding honor and integrity with respect to conduct with foreign nations. Justice Stevens is recognizing here something that I have argued in a number of law review articles (at greatest length here): state governments have often believed they had the primary duty to manage certain aspects of foreign relations that intersected their domain. And the federal government has often agreed that the states have this task, and has left such questions to the states. (Recall that in both the Breard and LaGrand cases, President Clinton basically took that position with respect to those earlier ICJ rulings).

This is why Stevens is asking, pleading, for Texas to come to its senses and give Mr. Medellin his hearing. He is recognizing that, in the U.S. system, the states often are the only governmental entities empowered to fulfill certain treaty international law obligations (although they have no constitutional duty to do so). It is a bit a strange result, but it is actually (in my reading of historical practice) hardly unprecedented.

In the beginning of his concurrence in Medellin, Justice Stevens reads Article 94 not to require the Texas state courts to take steps to ensure that the U.S. complys with the ICJ judgment. I disagree with his interpretation of “undertakes to comply,” but he’s almost convinced me that it’s a close question, at least with respect to whether the treaty (plus the Supremacy Clause) imposes an obligation on the state court to entertain a habeas petition that state law would otherwise foreclose.

In any event, and more to the point, Justice Stevens then includes (page 4-5) a remarkable paragraph that in my view gets right to the heart of the matter, which is not whether the ICJ judgment is “enforceable in court,” but instead whether the State of Texas as a whole has a constitutional obligation to see to it that the ICJ judgment is honored:

Under the express terms of the Supremacy Clause, the United States’ obligation to “undertak[e] to comply” with the ICJ’s decision falls on each of the States as well as the Federal Government. One consequence of our form of government is that sometimes States must shoulder the primary responsibility for protecting the honor and integrity of the Nation. Texas’ duty in this respect is all the greater since it was Texas that—by failing to provide consular notice in accordance with the Vienna Convention—ensnared the United States in the current controversy. Having already put the Nation in breach of one treaty, it is now up to Texas to prevent the breach of another.

OK, so Texas (as well as the federal government) has a constitutional obligation to prevent the U.S. from breaching Article 94. I agree. And if Texas were to execute one of the 51 defendants before they were afforded “review and reconsideration” of their sentences by someone, Texas would then cause the U.S. to breach Article 94, which Texas may not do.

Having read this paragraph, one might expect Justice Stevens to then conclude that, even if there is no way (absent federal statute) to judicially enforce this obligation against Texas, nevertheless Texas is constitutionally required to do the right thing.

But for some reason, Justice Stevens pulls his punches at the last minute: Instead of writing that “The Court’s judgment, which I join, does not change the fact that the State of Texas is required to take appropriate action to prevent a breach,” Stevens writes that “The Court’s judgment, which I join, does not foreclose further appropriate action by the State of Texas.”

As though Texas has a choice in the matter.

Can anyone reconcile this closing sentence with Stevens’s earlier paragraph (quoted above), which (correctly, in my view) speaks of Texas’s constitutional “duty” and “obligation”?

As many have noted, the Supreme Court’s decision yesterday in Medellin v. Texas raises serious questions about the binding nature of United States’s treaty commitments. The Court holds as follows: “Because none of these treaty sources creates binding federal law in the absence of implementing legislation, and because it is uncontested that no such legislation exists, we conclude that the Avena judgment is not automatically binding domestic law.” (Slip Op. at 10). In other words, the Optional Protocol to the Vienna Convention on Consular Relations may be binding as a matter of international law, but it is not binding domestic law because it is not self-executing and there is no federal legislation that puts the obligation into effect.

This decision just serves to prove a point that I make in my article, Treaties’ End, which was the subject of a forum on Opinio Juris last week. In the article, I argue that treaties can be more difficult to enforce in U.S. courts than congressional-executive agreements. That is in part because some treaties are not self-executing. (I put to one side, for now, the question of whether the Court was right that the treaty obligations at issue in Medellin are or are not self-executing.) In such cases, two problems can emerge. First, a non-self-executing treaty could impose an international obligation on the United States that would be unenforceable as a matter of domestic law—because the necessary implementing legislation has not been passed—leaving the country in violation of its international obligations. (That is essentially what the Medellin Court decided had happened in that case.) To avoid this problem, the Senate usually postpones its advice and consent to a non self-executing treaty until implementing legislation can be enacted concurrently. Alternatively, it might give its advice and consent to the ratification of a treaty contingent upon the subsequent enactment of implementing legislation. That did not happen here—either because everyone involved assumed the treaty obligations were, in fact, self-executing or because the nature of the treaty obligations were not expected to give rise to individual claims under the treaty that could only be satisfied if there were federal implementing legislation (if the latter were true, that would represent a serious lack of foresight that should have been addressed once the individual claims were in fact filed).

As I argue in my article, congressional-executive agreements make it possible to avoid these dilemmas. Congressional-executive agreements are created by means of legislation passed by both houses of Congress. Unlike non-self-executing treaties, congressional-executive agreements provide one-stop shopping: the same act that provides the authority to accede to the international agreement can also make the necessary statutory changes to implement the obligation incurred. Had the Vienna Convention and Optional Protocol been entered as congressional-executive agreements, it is extremely unlikely that the Court would have held in Medellin that the treaty obligations were unenforceable in U.S. courts as a matter of domestic law.

[Mark Weisburd is the Martha M. Brandis Professor of Law at UNC Law School.]

I find it difficult to read Medellin as institutionalizing a presumption against self-execution. If that had been Roberts’s intent, the form of his argument should have been, “We presume non self-execution, is there anything to overcome the presumption?” Instead, he analyzed the text, ratification hearings, and practice of other treaty parties to conclude that Art. 94 was not intended to create obligations for domestic courts. The conclusion seems reasonable to me - 94(2)’s according the Security Council discretion to refuse to enforce an ICJ judgment is hard to reconcile with a domestic judicial duty to enforce those same judgments - but it certainly isn’t reached with the aid of presumption.

David Sloss’s post makes an important point regarding the branch of the federal government with the repsonsibility to execute particular treaties. As he pointed out to me in a colloquy some time ago, whatever the status of some generic ICJ judgment, this particular judgment specifically requires action by American judges and it is impossible to carry out the international obligation admittedly created by the judgment without judicial action. The problem I see is that, if Art. 94 in general does not require domestic judicial implementation, and if the Senate consented to American submission to the ICJ only on the understanding that there was no requirement of domestic judicial enforcement, what happens when a particular ICJ judgment is meaningless without such enforcement? I find the Senate’s understanding crucial. Necessarily, it seems to me, the “treaties” to which the Supremacy Clause refers are those to which the Senate understood itself to be consenting. That is, a treaty for purposes of American judicial treatment imposes only those obligations which the Senate saw the treaty as creating. So - to address David’s argument - if the treaty to which the Senate thought it consented never requires judicial enforcement of ICJ judgments, then that’s the treaty which is the supreme law of the land. It may well be reasonable to argue that, at least on these facts, that not the best reading of Art. 94, but, if I’m right, the only issue is determining the Senate’s understanding of the treaty, not determining whether that understanding necessarily makes sense. Indeed, when the Court holds that the Senate’s understanding of Art. 94 as creating no domestic legal effects disables the president from seeking to implement Avena, it seems to put just that degree of weight on the Senate’s understanding.

Three other quick points. First, I think Ernie Young is exactly right that upholding the effect of Bush’s memo here would have had immense consequences. The administration’s argument was that the president can negate state law in order to carry out international legal obligations not otherwise binding in the US. Given the breadth of at least some readings of customary international law these days, it’s hard to imagine a subject as to which the President could not, effectively, legislate by decree if that argument had prevailed. I would add that the Court’s take on the consequences of the Senate’s understanding would seem to put to rest the controversy during the Reagan administration regarding the President’s authority to “reinterpret” treaties, according them a meaning different from that the Senate thought they had when it consented to ratification. Finally, regarding Paul Stephan’s point about the Court’s examining other states’ readings of Art. 94, I would note that this follows straight from the Vienna Convention on the Law of Treaties. Indeed, one of the reasons the ICJ was wrong in LaGrand and Avena was that it ignored other states’ readings of the Consular Convention.

[Edward Swaine is an Associate Professor of Law at the George Washington University Law School. ]

The quick scorecard on Medellin is pretty simple: Texas wins, the ICJ loses, and the President loses. I have lots of reactions to what the opinions say about the ICJ, non-self-execution, and even comparative law, but let’s just focus on this bottom line.

The first two claims have already been spun. As to Texas, Peter makes the point that freedom may have a price. (And Justice Stevens thought that Texas might step up and take one for the team; I suppose hope springs eternal.) As to the ICJ, Chief Justice Roberts suggests that a sparer approach to self-execution is indispensable for U.S. treatymaking, since any other approach might “hobble” U.S. willingness to enter into agreements by causing too much anxiety about what courts would do. If this reasoning is accepted – and there’s something to it, though I am not persuaded – it probably applies even more forcefully to dispute-resolution mechanisms, so this decision could be celebrated as a shot in the arm for U.S. willingness to go before the ICJ! (Not that it will particularly improve the reception once we get there, but you can’t have it all.)

So what about the third result — the President’s loss? I myself have argued that President should be understood to have the authority to implement even some non-self-executing treaty obligations via the Take Care Clause, which is a constrained kind of power — it authorizes only to the extent it binds. The Court gives that the back of the hand. It also rejects what the executive branch was arguing, which depended on the power to make sole executive agreements, and should be understood as cutting back on attempts to extrapolate from that line of precedent. As many have pointed out, however, the circumstances giving rise to this assertion of presidential power are pretty unusual: Not only is the President trying to embrace an ICJ decision that the United States lost, but he is doing so while continuing to insist that the ICJ was wrong. Not too appealing a pitch, and not too likely to come up in the near future, even if the Court has single-handedly saved dispute resolution.

So is there a countervailing upside for the President, like there is for the ICJ? Something much more substantial, to my reckoning – if not exactly to my liking. The near presumption against self-execution, the reliance on domestic political branches to confirm that presumption, and the deference to executive branch treaty interpretation, among other things, all force the conclusion that it will be yet harder in the future to invoke treaty obligations in court contrary to executive branch interests. Even a marginal change along this line of authority is quite important to presidential authority, since it applies in many more circumstances, and in many cases of keener concern to the President, than will the case’s holdings with regard to ICJ decisions. Geneva Conventions, anyone?

There is much one could say about Medellín, but I want to focus on the meta-question of what this decision portends for the future of international courts and tribunals. While the domestic effect of ICJ decisions is now cast into serious doubt (at least in terms of direct enforcement), I think there is far more reason to be hopeful than some are suggesting.

First, the Court emphasized that the effect to be given to international courts and tribunals depends first and foremost on whether there is a federal mandate to respect such decisions. It emphasized that such a federal mandate might be found in a self-executing treaty or a congressional statute. The Court said it agreed “as a general matter [that] an agreement to abide by the result of an international adjudication can be a treaty obligation,” but found that “the particular treaty obligations on which Medellín relies do not of their own force create domestic law.” (p. 24). The Court also agreed that a statute could have the same effect. “The judgments of a number of international tribunals enjoy a different status because of implementing legislation.” (p. 25).

Second, the Court fully embraced the principle that domestic effect should be given to decisions of international courts and tribunals if that is what federal law requires. As I have written elsewhere, this domestic effect falls along a continuum of deference. The Court cited with approval the “full faith and credit” approach of 22 U.S.C. 1650a, which treats ICSID decisions exactly the same as domestic court decisions. (p. 25). It also cited with approval an “arbitration model” under the New York Convention that accords great deference to international arbitral decisions pursuant to the Federal Arbitration Act. (p. 26). The decisions of the Iran-United States Claims Tribunal are the best example of an international tribunal that falls within this sort of approach. Although somewhat less clear, the Court also appears to accept a “foreign judgment” model, provided the international tribunal is rendering monetary awards (rather than injunctive relief) and provided the international decision does not contravene domestic law. (p. 26). Mass claims tribunals such as the UNCC are possible candidates for such a foreign judgment model. (It is also worth noting that the citation in footnote 1 to the La Abra case involving the U.S.-Mexico Claims Commission–one of the few Supreme Court decisions utilizing a foreign judgment model for an international tribunal decision–may suggest that if a foreign judgment model is to be employed, again the treaty (or implementing legislation) must mandate that approach.)

Third, the Court effectively relegated ICJ decisions to the same status as the decisions of the WTO Appellate Body. Direct recognition of WTO decisions is precluded by implementing federal legislation (19 U.S.C. 3512(c)). Under this implementing legislation, the political branches must decide what domestic effect to give to WTO decisions. Apparently the same now applies to ICJ decisions. ICJ decisions may be given domestic effect, but the mechanism is through the political branches. The President tried to do that, but failed in his choice of mechanism. Obviously if it so desired, Congress could achieve what the President’s Memorandum did not. That frequently happens with WTO decisions, with Congress amending the law to bring the United States into conformity with our international obligations as interpreted by WTO Appellate Body decisions.

Fourth, the Court did not address the issue of indirect recognition of decisions of international courts and tribunals. On this score nothing has changed. Charming Betsy remains vibrant and there is every reason to think that domestic courts in construing statutes will continue to rely on decisions of international courts and tribunals (including the ICJ) to interpret international law. The same goes for using international decisions as persuasive authority to understand the content of international law in matters such as ATS claims or boundary disputes.

We are witnessing the end of the era of “respectful consideration” and the birth of disaggregated deference. That is, the degree of deference domestic courts should accord to decisions of international courts depends on what federal law (i.e., self-executing treaties or implementing legislation) requires. That mandate may be more or less than “respectful consideration.” In the absence of such a federal mandate, international tribunal decisions will not have direct effect, but they will continue to enjoy indirect recognition as tools of interpretation.

I wanted to react quickly to two of the other items posted earlier today. First, Paul Stephan makes the point that Justice Breyer’s dissent is “very problematic.” I agree. Breyer’s opinion does not make a helpful contribution to untying the “Gordian knot” of non-self-execution doctrine.

Second, Curt Bradley claims that the Court’s decision effectively reserves “to Congress the determinations of whether and how to comply with the ICJ decision.” With due respect for Curt, whose views are usually very insightful, I think this assessment is wrong. Even if the Court held that courts are obligated to enforce the Avena decision, Congress would retain the power to decide “whether and how to comply with the ICJ decision,” because Congress could enact a statute to override the Court’s decision. The real issue here concerns default rules — what should the courts do in the absence of congressional action? The Court effectively endorses a default presumption against judicial enforcement (although the Court does a very poor job of explaining when that default presumption applies). In contrast, the Framers endorsed a default presumption in favor of judicial enforcement, as evidenced by the constitutional text specifying that “the judges in every State shall be bound” by treaties. Thus, the Court’s decision does not empower Congress. To the contrary, it effectively amends the Supremacy Clause by instructing state courts not to enforce treaties — or not to enforce some ill-defined category of treaties — in the absence of Congressional action.

Eric Posner has this post up on Medellin on Slate’s new law blog Convictions, which wraps up with the following:

But this is at least partly false. The U.S. is in fact clearly in greater compliance with the VCCR today than it was in, say, 1998, when the Supreme Court first addressed the consular notification rights issue in the Breard case. In 1998 there were probably only a handful of local law enforcement officials, even high up the chain, who had even heard of the VCCR. A decade later, it’s only one level below Miranda. Even law enforcement officers in Texas probably try to toe the line, if only because they now understand the kind of headaches that come with noncompliance. It’s an astonishing success story, the entrepreneurial use of the VCCR. It could only have happened against a backdrop in which IL was generally coming to count for more.

No doubt today’s decision is a loss for the transnationalist camp, as was the US withdrawal from the Optional Protocol, and there may have been some overreaching/wishful thinking along the way. But equating Medellin with necessarily diminished US compliance is to take an overly court-centric view of the international legal dynamic. Although it would help to enlist them more fully as enforcement agents, international law does not need the federal courts to achieve meaningful efficacy. Other countries with substantial interests in the VCCR can keep up the heat through other channels, in ways that will work to further improve aggregate U.S. compliance over the long run.

In Medellin, the Court held “that neither Avena nor the President’s Memorandum constitutes directly enforceable federal law . . . .” This comment focuses on the effect of the Avena judgment itself, and disregards the President’s Memorandum. The majority was undoubtedly correct to hold that Avena is not “directly enforceable federal law.” In fact, Avena is not federal law at all. The Constitution is federal law. Statutes are federal law. Treaties are federal law. But decisions of the ICJ are not federal law.

The Court erred, however, by concluding that Article 94 of the U.N. Charter is not federal law. See Roberts, slip op. at 10 (stating that the U.N. Charter does not create “binding federal law in the absence of implementing legislation”); id. at 24 (“the particular treaty obligations on which Medellin relies do not of their own force create domestic law”); id. at 31 (“A non-self-executing treaty, by definition, is one that was ratified with the understanding that it is not to have domestic effect of its own force.”) The Chief Justice, unfortunately, confused two entirely separate questions: whether Article 94 of the U.N. Charter is federal law, and how the treaty obligation is to be executed.

Article 94(1) of the Charter stipulates: “Each Member of the United Nations undertakes to comply with the decision of the International Court of Justice in any case to which it is a party.” Chief Justice Roberts tried to answer the question whether Article 94 is federal law by analyzing the text of the treaty. This is like trying to answer a question about Venezuelan law by looking in the U.S. Code. The question whether the U.N. Charter is federal law is a question about U.S. constitutional law. Accordingly, the answer is to be found in the text of the Constitution, not in the text of the treaty. The Constitution states that “all Treaties made . . . under the Authority of the United States, shall be the supreme Law of the Land.” Since the U.N. Charter was made under the authority of the United States, it is the supreme Law of the Land: i.e., it is federal law. By deciding that the U.N. Charter is not federal law, the Court has effectively rewritten the text of the Supremacy Clause to say that treaties are the Law of the Land unless we, the Supreme Court, decide otherwise.

Given that Article 94 is federal law, the next question is how to execute the U.S. treaty obligation. As noted above, Article 94 obligates the U.S. to comply with the ICJ decision “in any case to which it is a party.” There is no dispute that the U.S. is obligated to comply with the ICJ decision in Avena because the U.S. was a party in Avena. At the risk of over-simplifying, one can say that Avena obligates the U.S. to provide a judicial hearing for Medellin for the purpose of deciding whether he was prejudiced by the violation of his rights under Article 36 of the Vienna Convention on Consular Relations (VCCR). So, in the present case, the question of how to implement the U.S. obligation under Article 94 becomes a question of how to implement the U.S. obligation to provide a judicial hearing for Medellin.

In this regard, it is helpful to recall Justice Iredell’s opinion in Ware v. Hylton, 3 U.S. 199 (1796). In Ware, Justice Iredell distinguished between executed and executory treaty provisions. Treaty provisions are “executed” if “from the nature of them, they require no further act to be done.” Id. at 272. In contrast, executory treaty provisions require some further action by the U.S. government. Justice Iredell divided executory treaty provisions into three groups: legislative, executive, and judicial. See id. at 272-73. Whether an executory treaty provision requires legislative, executive, or judicial action depends on the nature of the international obligation, and the capacity of the respective branch of government to implement that obligation.

Chief Justice Marshall’s analysis in Foster v. Neilson, 27 U.S. 253 (1829), was entirely consistent with Iredell’s analysis in Ware. Marshall thought that Article 8 of the 1819 treaty with Spain was executory because the specific treaty language – “shall be ratified and confirmed” – required further government action. (It bears emphasis that Marshall was drawing a distinction between executory and executed treaty provisions, a distinction that depended on whether the treaty required further government action. See David Sloss, Non-Self-Executing Treaties: Exposing a Constitutional Fallacy, 36 U.C. Davis L. Rev. 1, 19-24 (2002)). The specific government action required by the treaty, in Marshall’s view, involved the transfer of real property from one private party to another private party. Legislative action was necessary because the treaty obligated the U.S. to convey title to real property, and the legislature was the only branch of government competent to execute that obligation. Foster neither states nor implies that legislative action is always necessary to execute an executory treaty provision. Thus, the Court in Medellin erred by construing Foster to mean that a non-self-executing treaty always requires legislative implementation. See Roberts slip op., at 30. As Justice Iredell explained in Ware, some executory treaty provisions require legislative action, but others require executive or judicial action, depending on the nature of the international obligation.

The application of this framework in Medellin is very straightforward. As noted above, the U.S. obligation under Avena and Article 94 of the U.N. Charter is to provide a judicial hearing for Medellin. There is only one branch of government capable of executing that obligation: the judicial branch. As Justice Breyer noted in his dissent, the obligation could be implemented either by the federal judiciary or the Texas state courts, but there are a variety of factors that weigh in favor of state court implementation. Regardless, the correct application of Foster and Ware to the facts of Medellin leads inexorably to the conclusion that Article 94 is an executory treaty provision that requires judicial execution because the judicial branch is the only branch competent to execute the U.S. obligation to provide a judicial hearing for Medellin. This does not mean that every ICJ decision is directly enforceable in U.S. courts. As Justice Iredell explained in Ware, it depends on the nature of the obligation that flows from the particular ICJ decision.

The fundamental flaw in the Court’s analysis in Medellin stems from its failure to distinguish between two very different questions: 1) is Article 94 of the U.N. Charter federal law?; and 2) what is the appropriate mechanism to execute U.S. treaty obligations under Article 94? The Court conflated these two questions by combining them into a single question: whether Article 94 is self-executing. This muddled analytical approach is symptomatic of a broader trend in U.S. jurisprudence that can be traced, in part, to the rise of legal realism a century ago. Justice Holmes thought that a so-called “law” is not really “law” if it can’t be enforced. Henry Hart argued persuasively that effective application of Holmes’ insight necessarily requires a two-step analysis: 1) is the relevant instrument a “law”?; and 2) what is the best way to enforce that law? Under Hart’s approach, the assumption is that all laws must be enforced in some way because the very nature of “law” is that it must be enforced.

Unfortunately numerous courts and commentators have twisted Holmes’ idea to produce the opposite result. They think that courts should simply bypass step one, proceed directly to step two, and ask whether the relevant law explicitly requires judicial enforcement. Under this approach, if the law does not explicitly require judicial enforcement, courts should refuse to enforce it. Whereas Holmes believed that the idea of an “unenforceable law” is a contradiction in terms, modern realists have perverted Holmes’ key insight and produced a wide range of judicial decisions that effectively render valid laws unenforceable. Medellin is the latest in this misguided series of decisions. In Justice Roberts’ perverted version of Holmesian realism, Article 94 of the U.N. Charter is not domestically enforceable (without legislative action) because it does not specify a domestic enforcement mechanism. Since Article 94 is not domestically enforceable, it is not federal law – even though the Constitution states unambiguously that it is federal law!!! Justice Holmes is rolling over in his grave.

I’ll venture a cautious prediction beyond the four corners of the opinion: Medellin will not be executed. I don’t think it will result from anything that happens in Washington; Congress just won’t have adequate incentive to help out a brutal murderer in the name of international legal obligation. The Administration might try to push for implementing legislation sufficient to satisfy the Medellin majority. It’s possible that the Government of Mexico itself will make some approaches on the Hill as well as at Foggy Bottom (where diplomatic hands have now been more tightly bound), thus further bringing Congress directly into the arena of international relations on yet another front. But I just don’t see Congress picking up that ball.

The more intriguing possibility is that Mexico starts waving a big stick around Austin with talk of what I’ve called targeted retaliation. If domestic Mexican constituencies (and for that matter, Mexican citizens resident in the US, who can now vote in Mexican presidential elections) are mobilized, the GOM will have to take some action beyond useless diplomatic demarches (nor does the Security Council route look very promising, even from a PR perspective). So why not start talking boycott? I doubt it could be undertaken as an official move, consistent with the international trade regime, but I don’t think it’s at all implausible that Mexican elites (governmental and non) could start quietly talking about redirected trade and investments. Texas had more than $40 billion in annual exports to Mexico as of 2003. That’s serious money and a lot of jobs. (Why buy a Dell when you can go with Gateway?) Even if a very small percentage made its way to California or other states instead, that would have to make Medellin’s execution look a little less like a top priority. And of course we have the Torres case in Oklahoma by way of precedent (which, in Janet Levit’s account, included “implicit threats of economic retaliation”).

So we might be at the end of the federal judicial road, as Mark Movsesian suggests, but that will hardly be the end of the story here. By further gumming up the treaty enforcement process at the national level, the Court pushes decisionmaking downstairs. But that isn’t necessarily a bad place to advance the rule of international law.

As lead counsel on the scholars’ amicus brief in support of Texas, I am not entirely unbiased here. But when one can get scholars with as diverse views of executive power as John Yoo and Erwin Chemerinsky to sign on to a brief arguing that the President has gone too far, it shouldn’t be entirely surprising to find that the Court agrees. Here are some early thoughts on the opinions:

1. This opinion certainly gives aid and comfort to those who have argued for a general presumption that treaties are not self-executing, although it might be a stretch to say it holds as much. The Chief’s majority opinion does strongly reject the dissenters’ opposite presumption, and that is important in itself. But keep in mind that the Chief carefully distinguishes between the different ways that treaties may or may not be self-executing. The Vienna Convention is plainly self-executing in that it binds the Houston police to give warnings without further implementing legislation, and it may be self-executing in the sense that individuals can assert violations on their own in court (the Court doesn’t decide). What the Court rejects is that the ICJ’s judgments under the treaty are self-executing in the sense of being directly enforceable in domestic courts. But Medellin was on unusually weak ground here to argue otherwise, given that the Executive had taken the position that such judgments are not self-executing, and both the Executive and the Court (in Sanchez-Llamas) were on record that the judgment to be enforced was incorrect on the merits. These things are going to have to be fought out treaty by treaty, which is probably the right result.

2. The presidential power holding, although it takes a back seat to the self-execution holding in the majority opinion, may be more sweeping in at least one sense. The Court holds pretty categorically that the President lacks power unilaterally to execute a treaty that is otherwise non-self-executing. In fact, the Court says that a determination that the treaty is non-self-executing means that Congress has implicitly disapproved actions to execute the treaty, placing presidential actions to execute it in Category 3, not 2, under Youngstown. Given the broad and amorphous nature of many of the non-self-executing treaties to which we are parties—think of some of the more open-ended trade or human rights instruments—a contrary holding would have been a broad grant of power to the President indeed.

3. The majority also takes what seems to be a major bite out of the sole executive agreement cases like Garamendi, Dames & Moore, and (looking further back) Pink and Belmont. Chief Justice Roberts says that these cases “involve a narrow set of circumstances” concerning the settlement of claims against foreign nations. It will be harder, in future, to cite Garamendi and Dames & Moore for open-ended presidential authority to create binding federal law by sole executive agreements without congressional action.

4. The internationalism of Justice Breyer’s dissent is really quite striking, as is the extent to which this case replicates the usual left-right split on the Court. (Justice Stevens concurs in the result, but his heart seems to be with the other liberals in dissent.) I think that’s unfortunate. The legal question dividing the Court in Medellin concerned the domestic effect of international law, and the allocation of authority between domestic and supranational courts. That should be a left-right issue only on the most cynical view of international law, which is that it provides a vehicle to achieve more liberal results on issues like the death penalty than the domestic political consensus would otherwise stand for. But even if we take that view, the truth is that both liberals and conservatives have things to gain and things to fear from increasing or decreasing the influence of international law and institutions in the domestic legal system. Free market conservatives may approve (and liberals disapprove) of decisions by supranational trade tribunals rejecting local environmental or labor laws, for instance. Reasonable people can differ about the extent to which we should open up the domestic legal system to international law and courts, but they should not differ on the traditional left-right grounds.

[Mark Movsesian is the Frederick A. Whitney Professor of Contract Law at St. John's University School of Law.]

Thanks to Chris for inviting me to say a few quick words about today’s very significant decision.

Hooray for Dualism: Just as he did two years ago in Sanchez-Llamas, CJ Roberts endorsed a dualist approach to the judgments of international tribunals. Dualism teaches that international and domestic regimes are entirely independent; in the absence of a domestic act of incorporation, international judgments cannot serve as rules of decision in domestic courts. Because there was no valid act of incorporation in this case – the Court believed that the Bush Memorandum did not qualify – Avena did not bind American courts.

This approach is entirely sensible. As a formal matter, neither the Optional Protocol nor the UN Charter can fairly be read to provide for direct effect for ICJ judgments. The functional arguments for dualism are as strong as well. By reserving the final decision on international judgments to domestically accountable actors, dualism promotes legitimacy and avoids unnecessary friction between national and international bodies. Moreover, dualism actually increases the likelihood that nations will create international tribunals. If international judgments automatically bound domestic courts, without any further act on the part of domestic authorities, nations would be much more wary of signing up to international regimes.

The Foreign Judgments Analogy: The Court correctly dismissed the idea that enforcing the ICJ judgment would be effectively the same as enforcing a foreign court judgment or an international commercial arbitration award. While it’s true that domestic courts routinely enforce such rulings, foreign court judgments and ICA awards typically deal with commercial disputes that do not impinge on domestic public policy. Here, by contrast, the ICJ ordered the retrial of scores of defendants who had been convicted of serious crimes and indirectly questioned American policy on the death penalty. These are hardly the sort of issues that come up in the foreign-judgments and ICA contexts.

The End of the Road: For the last decade, the American international law academy has been pressing the Court to be more receptive to ICJ judgments. Ten years ago, in Breard, the Court held that a preliminary ICJ ruling merited only “respectful consideration.” Two years ago, in Sanchez-Llamas, the Court declined to treat a final ICJ judgment as binding precedent. And now the Court has ruled that ICJ judgments are not enforceable. For better or worse (on balance, I think it’s for better), it’s clear after this morning that the campaign to change the Court’s mind has failed. New justices might see things differently, of course — though it’s significant that even Stevens voted with the majority today. For now, though, it seems time to move on to other projects.

Our own Duncan Hollis was cited by Justice Breyer in his Medellin dissent. Justice Breyer cites to Duncan’s essay A Comparative Approach to Treaty Law and Practice, which is from the book National Treaty Law and Practice, co-edited by Duncan, Merritt R. Blakeslee, and L. Benjamin Ederington. The first cite to Duncan’s essay appears on page 11 of the dissent. Other authors from Duncan’s book are cited as well.

Congratulations, Duncan!

A first read through the Medellín opinions leads to tentative observations, subject to revision:

• Chief Justice Roberts’ opinion for the Court is modest and fairly careful. He does not articulate a presumption against self-enforcement, or offer a general interpretive template. The analysis of the Optional Protocol and the UN Charter is specific to those two instruments. As my prior briefs and published work indicate, I find this part of the opinion completely persuasive. I take issue with the glib assumption that a commitment to comply with an international tribunal’s decision implies an automatic assignment to the judiciary of the authority to ensure that the commitment is honored.

• Although the opinion is limited in the sense that it does not offer a general rule for inferring self-executing from treaties, its dicta states strong views (it might be too strong to say it disposes of) concerning several controversies that the academic community has taken seriously. (a) The Court understands self-execution to refer to all forms of domestic enforcement, not just to the existence of a private right of action. Its definition of self-execution in footnote 2 may clarify our discussing going forward, even if some may quarrel with the definition used. (b) Reservations, declarations and understandings that limit or foreclose self-execution of a treaty that might otherwise have domestic effect seem acceptable to the Court. The Sosa Court also hinted as much. (c) And the idea of domestic enforcement of the awards of international tribunals does not seem to cause any great concerns, at least in the abstract. This will disappoint some who have suggested that domestication of such awards might present problems under Article III or other constitutional provisions.

• As a teacher of comparative law, I was delighted to see the Court’s reliance of the evidence of other country’s enforcement of ICJ decisions. I missed seeing a discussion of the recent decision of the German Constitutional Court regarding the Vienna Convention, although it may be too recent, too complex, and too tangential to make any of the briefs. The basic point that domestic implementation of international obligations has a comparative component and that an appreciation of foreign practice enriches our understanding of our own.

• As I was serving in the Executive Branch at the time of the drafting of the U.S. amicus brief and the oral argument, I am disappointed by the last part of the Court’s opinion. I would have thought that there was more to the US’s argument that the Optional Protocol, the UN Charter, and 22 U.S.C. § 287 can be read as assigning to the President the discretion to implement ICJ decisions through changes in domestic law. This argument, to be sure, is neither clear nor ineluctable. Still, I came away feeling that the Chief Justice was a bit like the person who, having a hammer, sees everything as a nail. That is to say, the opinion works so hard to clarify and establish what it means to say that a treaty is not self-executing that it rushes past a plausible and even useful refinement, namely that the treaty makers in advance might specify a nonlegislative mechanism for deriving valid domestic law from an otherwise non-self-executing treaty. To accept this argument, one would have to see Dames & Moore , Belmont and Pink not simply as cases recognizing a limited Presidential power that inheres in Article II, but also an expression of the expectations of the legislative branches when authorizing the President to enter into dispute resolution with foreign states. One might still argue that the treaty makers or Congress have to do more than simply sign on to dispute resolution to give the Executive the authority to choose to implement an international award or not. But here the Court’s opinion struck me as less careful or persuasive than what went before.

• If I had had any doubts about the persuasiveness of the majority’s discussion of the non-self-executing issue, Justice Breyer’s dissent would have put them to rest. The Chief Justice was remarkably restrained in his deflection of the dissent’s very problematic claims and proposals.

• This will not end all Vienna Convention litigation. We still have to decide what, if anything, Section 1983 adds: The Circuits are split. So the gift to which Julian refers will keep on giving for at least a little longer.

There is a way in which the Medellin decision fits very nicely with our discussion last week about congressional-executive agreements. Like Oona’s article, the decision in Medellin is very pro-Congress. The Court’s finding of non-self-execution means that it is reserving to Congress the determinations of whether and how to comply with the ICJ decision. Similarly, the Court’s presidential power holding means that the President must work with Congress if he or she wishes to convert non-self-executing international law into U.S. law. Finally, the Court’s reliance on Justice Jackson’s framework from Youngstown, particularly category 3 of that framework, envisions a significant role for Congress, even in foreign affairs. All of this is to be applauded, in my view.

First, my thanks to Kevin, Peggy, and the OJ crew for a chance to post my own initial reaction to Medellín.

Leaving the international law to the international law scholars, and the Court’s odd parsing of the VCCR’s ratification history to those who are more familiar with it, my own interest in the Chief Justice’s majority opinion in Medellín is in his rather bold attempt to clear up decades of uncertainty over just what, for constitutional purposes, a “non-self-executing treaty” actually is. (I assume, for the sake of argument, that the VCCR’s optional protocol is one such treaty.) Footnote 2 of the majority opinion provides the Court’s new definition:

What we mean by “self-executing” is that the treaty has automatic domestic effect upon ratification. Conversely, a “non-self-executing” treaty does not by itself give rise to domestically enforceable federal law. Whether such a treaty has domestic effect depends upon implementing legislation passed by Congress.

Whereas forests have been felled on the question of whether a non-self-executing treaty merely fails to provide a cause of action, or also fails to provide judicially enforceable rights, I take footnote 2 as resolving that longstanding debate—and doing so in favor of the latter, broader, view. Indeed, in the paragraph leading up to footnote 2, the point is made even more explicitly—that non-self-executing treaties “do not by themselves function as binding federal law.”

If so, this is an extremely important development, and one that seems thoroughly at odds with the plain text of the Supremacy Clause (to wit, “This Constitution, and the Laws of the United States which shall be made in Pursuance thereof; and all Treaties made, or which shall be made, under the Authority of the United States, shall be the supreme Law of the Land . . . .”). Is the Chief Justice’s view really that non-self-executing treaties are not “treaties” under the Supremacy Clause? If so, then Medellín overrules a litany of earlier cases that declined to draw such a distinction for purposes of the Supremacy Clause, and sub silentio at that…

I had always thought the more compelling argument was that non-self-executing treaties don’t create private rights of action. On that view, they still create positive law, per the Supremacy Clause, but law that could only be privately enforced through otherwise available causes of action.

Where this distinction would make an enormous difference is where a statute provides a cause of action for the enforcement of “treaties,” without specifying whether the treaty must be self-executing or not. An obvious example, of course, is the federal habeas statute, which provides both jurisdiction and a cause of action for claims by a detainee that “he is in custody in violation of the Constitution or laws or treaties of the United States.” (emphasis added).

I’ve argued previously that non-self-executing treaties should nevertheless be enforceable under the habeas statute, given that the Habeas Corpus Act of 1867 deliberately modeled this provision on the Supremacy Clause, and given that non-self-executing treaties, in my view, are still “treaties” under that constitutional provision.

I’m perfectly happy to accept that I might be wrong; it would hardly be the first time. But would it have been so hard for the Medellín majority to provide more than a cursory explanation for why? Given the significance of the implicit suggestion here–that non-self-executing treaties simply aren’t “treaties” under the Supremacy Clause–at least some analysis might have been helpful…

My former boss and Medellin’s counsel Donald Donovan (of Debevoise & Plimpton LLP) sends out this reaction to the Medelllin decision.

Donald Francis Donovan of Debevoise & Plimpton LLP, New York, counsel to petitioner Jose Ernesto Medellín, in response to the March 25, 2008 decision of the United States Supreme Court in MEDELLIN v. TEXAS:

As the early analysis starts to pour in on the Supreme Court’s Medellin opinion, I did a quick scan of the headlines. As I have argued here and here, the case is, at bottom, about the persistence of the death penalty in the U.S. and the efforts of close allies and neighbors of the U.S. to do something about it. Check out the early headlines, which reflect the centrality of the case to the death penalty and immigration debates:

US High Court Denies Hearing to Death Row Mexicans (Bloomberg)
Court rules against Bush, Mexican on Death Row (Reuters)
Supreme Court overrules Bush, OKs Texas execution (CNN)
Bush Overruled in Death Penalty Dispute (ABC)

Of course, there are the inevitable Texas-themed headlines:

Court doesn’t mess with Texas; denies new trial (Baltimore Sun)

As we read and watch the news coverage of the case, it will be interesting to count the uses of the word “Mexican,” “immigrant,” “murder,” and “death penalty.” And “World Court,” which sounds so much more ominous than “International Court of Justice.”

The Supreme Court’s Medellin decision today brings to an end a fascinating decade-long series of interactions between the U.S. Supreme Court, the International Court of Justice, and various state governments. Beginning in 1998, the Supreme Court has now weighed in four times on the ICJ’s various interpretations of the Vienna Convention on Consular Relations, the UN Charter, and the ICJ Statute (once in Breard, twice in Medellin, once in Sanchez Llamas). But although I could wish for yet more litigations, I think this is the last one, and it has been (from a legal academic standpoint) a wonderful ride.

The Court’s decision today may be the most important of the four decisions, since it tries to clarify a number of questions about the self-executing treaties and relationship of international judgments and state law, and the President’s power (or lack thereof) to carry out such international judgments. As a whole, Chief Justice Roberts’s decision is clear and (mostly) convincing. And it rightly rejects the more aggressive claims of groups like the ICJ Experts and other international lawyers that filed amicus briefs.

Here are the key holdings, as I see them:

1) Self-Execution

The key portion of the majority’s opinion is its analysis of the key treaty provisions (the Optional Protocol to the Vienna Convention on Consular Relations and Article 94 of the U.N. Charter) to conclude that these treaty provisions are not meant to be self-executing. The line between self-execution and non-self-execution has always bedeviled courts and commentators, but the Court here doesn’t seem deeply troubled. All you have to do is carefully analyze the text of the treaty to determine the intent of the treaty-makers, and perhaps consider some external sources such as the executive’s interpretations of the treaty and other states’ practice under the treaty.

2) The Enforceability of International Court Judgments

The question of whether an international court judgment is enforceable directly in US courts is entirely a question that turns on the particular treaty or statute or executive agreement in question. There is no presumption in favor of enforcing international court judgments. On the other hand, as the Court makes clear, there is no reason that Congress or the treaty-makers could decide to give international court judgments direct enforceability. They just haven’t done so here.

3) The President’s Limited Domestic Foreign Relations Power

Surprisingly, given the general media focus and interest in this case, the President’s attempt to enforce the ICJ judgments through a “Memorandum” does not occupy the Court too much (nor the dissent). The logic is again all about self-execution. If the treaty is not self-executing, then it is not federal law, and therefore it gives the President no further authority. Following Youngstown, therefore, we are at best in category two, where there is no express congressional authority. The President’s general foreign affairs power, recognized by the Court in Dames & Moore and, most recently, in Garamendi, is limited too executive agreements involving civil claims by U.S. citizens against foreign states. (Why this doesn’t also extend to claims by Mexican citizens against U.S. states, or the US in general, is not addressed other than that there is not longstanding practice in such cases).

My General Take:

I am on board for most of the Court’s analysis, which seems fairly sensible and reasonable. It is not overreaching, since it makes clear that there are indeed treaties that are self-executing, and international court judgments that could be self-executing (just not these ones).

The most important part of the Court’s opinion deals with self-execution, since its analysis there is the key the rest of the decision. And I don’t think it creates a “presumption” against self-execution, even against self-executing international court judgments (even though it perhaps ought to). But that is a subject of deep complexity, which I hope others tackle in more depth today.

Where I part from the Court is its rather brief dismissal of its own precedents in Dames & Moore and Garamendi, which I read to recognize that the President could preempt state law claims by virtue of sole executive agreements or a general foreign relations power. This power, it seems to me, seems to fit pretty well here since we have a Presidential attempt to settle a claim by a foreign government by preempting inconsistent state court judgments. But the Court is unimpressed and suggests this would be too different since those cases involved civil claims by U.S. citizens against foreign governments, whereas this involves interference with a state’s police powers.

My instinct has always been that somewhere, somehow, someone in the federal government has the power to vindicate the ICJ judgment led me astray. Absent legislation from Congress, an ICJ judgment is basically meaningless as a matter of domestic law.

There is one legal entity, of course, that has the power to give effect the ICJ judgment as well: the State of Texas. It is interesting that Justice Stevens’ surprising concurrence rested in the end on a plea to Texas to come to its senses and give Medellin a hearing. Good luck! Still, Justice Stevens recognizes that, in effect, we are going to have to rely on state governments to carry out ICJ judgments, absent Congressional action. The States, I’ve argued in prior work, are becoming substantially important foreign policy players. This decision will only enhance this role.

Thanks to Marty for his pointer on the decision and his instant analysis (which despite being instant, is also still quite interesting). Throughout the day today and into tomorrow, Opinio Juris will post thoughts and comments on the Medellin decision from leading commentators and scholars, in addition to (of course, our substantial “in-blog” expertise. Stay tuned!

Six to three. The decision is here. My very quick and preliminary reaction, after having read only a bit of the opinion, is that the presidential power question is not the most important aspect of the opinion. That would be, instead, the Court’s interpretation of Article 94 of the U.N. Charter as merely imposing a future obligation on the U.S. federal political branches to do something to comply with its requirement — and not to impose any independent obligation on the United States, including Texas, to actually take steps to comply with an ICJ judgment. This strikes me as an implausible interpretation, and as potentially very troubling for construction of treaty obligations going forward.

The article reads that the U.S. “undertakes to comply with with the decision of the [ICJ] in any case to which it is a party.”

The Court reads this obligation not to actually require the United States and its component parts to, uh, actually comply with an ICJ decision. Indeed, it apparently permits Texas (part of the U.S., last time I checked) to intentionally refuse to comply with such a decision.

What will this sort of treaty interpretation portend for, say, article 16 of the CAT, which provides that “each State Party shall undertake to prevent in any territory under its jurisdiction other acts of cruel, inhuman or degrading treatment or punishment”?” Apparently, that no longer means we are forbidden from intentionally inflicting such treatment on detainees — or so the Chief Justice reasons.

One last note regarding the discussion of Treaties’ End…

On behalf of all of us at Opinio Juris, I would like to thank Oona Hathaway, David Bowker, Curtis Bradley, David Golove, Marty Lederman, and Catherine Powell for joining us for the discussion of Oona’s article. As is evidenced from the discussion, each of our guests put a great deal of time and effort into analyzing the issues presented by Oona’s thought-provoking article. Speaking personally, I learned a great deal.

We hope you found it interesting, as well, and we look forward to similar online symposia in the coming weeks and months.

Let me begin my final post with a heartfelt thank you to Chris Borgen, Opinio Juris, and all the participants in the forum for a remarkable conversation about my article, Treaties’ End.

The conversation has touched on a wide range of issues that deserve much deeper treatment than I can give them here. I will just briefly mention several of the issues that have been raised over the course of the past week. This post is not meant to end the conversation but instead to invite continued debate and discussion.

First, the conversation has raised questions about the meaning of the term “lawmaking” in the context of international law. The subtitle of the article is, “The Past, Present, and Future of International Lawmaking in the United States.” Duncan Hollis asks whether the term is really correctly applied to international treaties. Duncan is right that the use of the term in this context is not standard. As I explain in the article, the terminology is quite deliberate–and part of my effort to reframe the debate. I refer to treaties and congressional-executive agreements “as ‘international lawmaking’ to emphasize the dependence of international law on individual countries’ decisions to commit to it. International law may be negotiated by states in New York or Geneva or Montreal, but it is not made at the negotiating table. It is made by countries when they agree as a matter of law to a binding international commitment. For it is the act of consent by each country that transforms an international agreement from a piece of paper devoid of any legal force into law that binds.”

Second, the conversation has repeatedly touched on issues of federalism. I argue in my article that the Article II Treaty Power does not face the same limits as Congress’s delegated powers under Article I. The Treaty Power is instead subject to limits of its own–a treaty must “have the consent of a foreign nation” and must be genuine, that is the parties must have a mutual interest in the subject matter of the agreement. Curt Bradley raises the debate over Missouri v. Holland, which he has addressed in depth in his scholarly work. He is concerned that lack of federalism constraints on the Treaty Power could lead to an expansion of federal lawmaking authority that lies in tension with democratic legitimacy. Cathy Powell also discusses this issue in the context of the debate over ratification of the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW). She disagrees with Curt, arguing that “federalism has been frequently used as a red herring in the context of ratification debates over human rights treaties”—and noting that the debate over the CEDAW was no exception. I respond to many of Curt’s points in a separate post and will not repeat that here. As I note there, much of our disagreement stems not as much from our different views of the Treaty Power as from our different views of Congress’s Article I powers, which Curt says do not reach some human rights agreements.

Third, our discussion has raised questions about the political feasibility of my proposal to conclude nearly all international agreements as congressional-executive agreements, thus brining Article II treaties (almost) to an end. Why, Duncan Hollis, Marty Lederman, and David Golove ask in various posts, would the Executive Branch want to “clarify the admittedly obscure landscape involving treaties and CEAs.” Would it really be “easier” to get the support of 51 Senators plus 218 Representatives than to get the support of 67 Senators? Why would the President ever go along with a plan to create stronger international commitments that are more difficult for the President to unilaterally undo? Peter Spiro adds, “it’s not clear that the Senate will go quietly into the Hathaway night.” I have already answered some of these questions. I acknowledge in the article that it would not necessarily be “easier” to gain a majority vote in both houses than a supermajority vote in the Senate (particularly, as I discuss in some depth, in light of the filibuster). Even so, I think the voting structure for congressional-executive agreements is more likely to lead to policies that reflect the preferences of the broader electorate than is the voting process for Article II treaties because it does not hand a veto to such a small and extreme minority. And I urge those considering the feasibility of the plan to keep in mind that presidents not only wish to maintain flexibility for themselves, but they might also want to create real constraints for their successors. That desire might lead presidents to endorse strong international commitments. The Senate, too, may not provide the opposition some expect: It has been surprisingly quiescent in the face of a grand expansion in the use of congressional-executive agreements, even though that expansion has diminished its unique power under the Treaty Clause.

Fourth, the conversation has raised questions about whether congressional-executive agreements or treaties lead to more reliable international commitments. In my article, I claim, contrary to the claims of nearly all the literature to date, that congressional-executive agreements have the potential to create stronger international commitments than do Article II treaties. David Golove and Marty Lederman take issue with this claim, arguing that treaties and congressional-executive agreements create equally strong international commitments. In an earlier post, I responded that even admitting that the two instruments are equal in this regard represents a significant move away from the current consensus that treaties are the stronger of the two. I go one step further, however, and argue that congressional-executive agreements are frequently easier to enforce and can be more difficult to undo than treaties–features than in my view make the congressional-executive agreement a more attractive mechanism for creating strong international commitments.

Fifth, the conversation has turned to the meaning of democratic legitimacy and its relationship to popular sove