The Federalist Society

The Federalist Society Online Debate Series

Medellin v. Texas
Part I: Self-Execution
March 28, 2008

On March 25, 2008 the Supreme Court decided the Medellin v. Texas case. The Court ruled that neither Avena nor the President's Memorandum constitutes directly enforceable federal law that pre-empts state limitations on the filing of successive habeas petitions. Solicitor General for the State of Texas and attorney for the respondent Ted Cruz, Saint Louis University School of Law professor David Sloss, Georgetown University Law Center professor Nick Rosenkranz, and former Legal Adviser to the U.S. State Department and current partner at Sullivan & Cromwell Edwin Williamson discuss the Court's decision in Part I: Self-Execution. Read Part II: Presidential Power.


David Sloss


Medellin and the Decline of Textualism  

Justice Scalia has rightly warned us of the danger of constitutional interpretation that is unmoored from constitutional text.  Unfortunately, in Medellin v. Texas, the Supreme Court did not heed this warning. 

The Constitution states expressly that treaties ratified by the United States are supreme federal law.  In Medellin, though, the Court held that Article 94 of the U.N. Charter, a treaty ratified by the United States, is not supreme federal law.  This holding is directly contrary to the plain meaning of the Constitution.  Moreover, the Court reached this conclusion without even analyzing the constitutional text.   

Instead of analyzing the constitutional text, the Court analyzed the treaty text to support its conclusion.  This approach is fundamentally misguided.  To understand the nature of the international obligation created by Article 94, one must analyze the treaty text.  But to ascertain whether Article 94 has the status of federal law within our domestic constitutional system, one must analyze the text of the Constitution.  

The Supremacy Clause states, in relevant part: “[A]ll Treaties made . . . under the Authority of the United States, shall be the supreme Law of the Land.”  One could argue that the phrase “shall be” effectively grants the treaty makers the power to decide that one treaty shall become federal law automatically, by virtue of treaty ratification, but a different treaty shall not become federal until after Congress has enacted implementing legislation.  To address this argument, let us set aside two categories of treaties that are not at issue here: treaties for which implementing legislation is constitutionally required, and treaties for which Congress voluntarily decides to enact implementing legislation.  Thus, the question is this: in cases where implementing legislation is not constitutionally required, and Congress decides not to enact implementing legislation, do the treaty makers have the constitutional power to decide that a particular treaty shall not be federal law?  Such a power cannot be rooted in the Supremacy Clause.  Constitutional text specifying that all treaties shall be federal law cannot mean that the treaty makers have the power to decide that some treaties shall not be federal law. 

In sum, the Court holds in Medellin that the drafters of the U.N. Charter decided, at the time of treaty ratification, that Article 94 of the Charter shall not be federal law in the absence of implementing legislation.  But the Court fails to identify any constitutional provision that grants the treaty makers the power to decide that a particular treaty shall not be federal law, and the Court fails to acknowledge that the Supremacy Clause states expressly that all treaties shall be federal law.

Ted Cruz


Thanks for your provocative post.  Your observations seem more in the tradition of Justice Thomas than Justice Scalia, because in order for your interpretation to be correct, the Court would have to jettison centuries of precedent to get there (something Thomas, perhaps rightly, is often willing to do).  The distinction between treaties that are "self-executing" and "non-self-executing" goes back at least 200 years.  As the Chief Justice observed in the majority opinion:

"The interpretive approach employed by the Court today-resorting to the text-is hardly novel. In two early cases involving an 1819 land-grant treaty between Spain and the United States, Chief Justice Marshall found the language of the treaty dispositive. In Foster, after distinguishing between self-executing treaties (those 'equivalent to an act of the legislature') and non-self-executing treaties (those "the legislature must execute"), Chief Justice Marshall held that the 1819 treaty was non-self executing.  2 Pet., at 314. Four years later, the Supreme Court considered another claim under the same treaty, but concluded that the treaty was self-executing. See Percheman, 7 Pet., at 87. The reason was not because the treaty was sometimes self-executing and sometimes not, but because ‘The language of' the Spanish translation (brought to the Court's attention for the first time) indicated the parties' intent to ratify and confirm the land grant 'by force of the instrument itself.' Id., at 89."

You seem to be arguing that Article VI of the Constitution forecloses the possibility of non-self-executing treaties; i.e., that the Supremacy Clause mandates that "all treaties = federal law" and therefore that all treaties must, as a constitutional matter, be judicially enforceable.  That argument runs into two problems:  first, the longstanding precedent described above, and second, it's hard to state any persuasive logical reason--as a matter of first principles-- why that should be so.  I agree that treaties stand on the same footing as "laws."  But Congress passes all the time "laws" that are not enforceable in court, and it's up to Congress to determine (in the first instance) which is which.  Why would treaties be different?  And isn't the fact that pretty much every President and every Senate since the time of the Founding has understood that non-self-executing treaties might be desirable (and are permissible) a powerful refutation to your claim?

Once we cross the bridge of admitting the possibility that treaties might be non-self-executing, the majority's argument concerning these particular treaties is compelling: (1) the text of the treaties does not suggest judicial enforcement, (2) the structure of the treaties suggests political and diplomatic enforcement, (3) the consistent position of the United States (owed great deference on such matters) has been--for over four decades--that these treaties are not judicially enforceable, and (4) every other country in the world that is a signatory to these treaties likewise deems them not judicially enforceable in their domestic courts.

Nick Rosenkranz


I agree with Ted, and will just add a few more thoughts on this point.  David argues that non-self-executing treaties are generally inconsistent with the Supremacy Clause, which provides that "all Treaties made, or which shall be made, under the Authority of the United States, shall be the supreme Law of the Land."  No party argued that position in Medellin, and no Justice endorsed it, because, as Ted points out, it was conclusively rejected by Chief Justice Marshall in Foster v. Neilson, 2 Pet. 253 (1829).  

And Chief Justice Marshall was right: non-self-executing treaties are entirely consistent with the Supremacy Clause.  What the Supremacy Clause means is that treaties are Supreme Law of the Land to the extent that they purport to be.  But if a treaty, by its own terms, purports only to create an international legal obligation to pass implementing legislation, then the treaty does exactly that and nothing more.  

Such non-self-executing treaties are perhaps analogous to non-binding congressional resolutions.  The Supremacy Clause provides that "the Laws of the United States ... shall be the Supreme Law of the Land."  Nevertheless, Congress regularly passes non-binding resolutions, which, by their terms, do not constitute domestic law, and no one thinks that these resolutions violate the Supremacy Clause.  The reason, of course, is that Laws of the United States made in pursuance of the Constitution are Supreme Law of the Land to the extent that they purport to be law.  Treaties work just the same way.

So the Court was correct to begin with the question whether the treaty at issue in Medellin is non-self-executing, and it was correct to conclude that it is.  As the Chief points out, Article 94(1) of the U.N. Charter provides only that "[e]ach Member of the United Nations undertakes to comply with the decision of the [ICJ] in any case to which it is a party."  This provision simply does not sound like a rule of domestic law; it sounds, as the Solicitor General explained, like "a commitment on the part of U.N. Members to take future action through their political branches to comply with an ICJ decision."  Brief for United States as Amicus Curiae in Medellin I, O.T. 2004, No. 04-5928, p. 34.  Since the underlying treaty is non-self-executing, the Court rightly held that an ICJ opinion could not trump inconsistent Texas law. 

David Sloss


I thank both Nick and Ted for their thoughtful replies to my post.  Let me clarify my position.  There are several variants of non-self-execution doctrine.  Some versions are consistent with the Supremacy Clause; some have roots dating back two hundred years.  In Medellin, the Court embraced a variant of the doctrine that originated in the 1960s, that has never previously been endorsed by the Supreme Court, and that is inconsistent with the text of the Supremacy Clause.

As of 1900, there were three different variants of the doctrine.  One variant holds that some treaties are constitutionally non-self-executing.  For example, a treaty obligating the U.S. to appropriate funds is constitutionally non-self-executing because the Constitution requires legislative action by both Houses of Congress to appropriate funds.  One could plausibly argue that such a treaty is not supreme federal law because the treaty makers, acting under Article II, lack the constitutional power to compel Congress to appropriate funds.  Regardless, no one claims that Article 94 of the U.N. Charter is non-self-executing in this sense.

A second variant holds that there are constitutional and/or prudential limits on the judiciary’s power to enforce treaties.  As Ted notes, there are a variety of limits on the judiciary’s power to enforce federal statutes.  Indeed, every federal statute is subject to at least some limitations on judicial enforcement – e.g., the Article III “case or controversy” requirement.  Those same limits apply to treaties.  However, the fact that judicial enforcement of statutes is subject to limitations does not mean that statutes are not federal law.  In Medellin, though, Justice Roberts asserts that Article 94 of the U.N. Charter is not federal law.  So, he clearly has a different variant of non-self-execution in mind.

A third variant, represented by Ware v. Hylton, 3 U.S. 199 (1796), distinguishes between “executory” and “executed” 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 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. 

Historical materials demonstrate that, in Foster v. Neilson, Justice Marshall endorsed the Ware v. Hylton version of the doctrine.  If one applies this version to Medellin, two things are clear.  First, Article 94 of the U.N. Charter is “executory.”  Second, the judiciary is the only branch of government capable of executing the international obligation at issue in Medellin, because the treaty obligates the U.S. to comply with the ICJ decision in Avena, the ICJ decision requires a judicial hearing for Mr. Medellin, and the judiciary is the only branch of government capable of providing a judicial hearing.

Ted Cruz


Alas, I think your argument attacks a straw man.  Nobody disputes that treaties CAN be self-executing.  If the President and the Senate so intend, a treaty--by virtue of the Supremacy Clause--can of course be made immediately enforceable in state and federal courts. 

But, what your argument seems to assert is that all treaties MUST be self-executing.  That's an altogether different matter. 

Stated differently, would you come to the same conclusion if these three treaties all said, in explicit all-cap bold letters, that these treaties cannot be judicially enforced?

And, if you then find yourself agreeing in that hypothetical that the President and the Senate can permissibly choose whether or not to make a particular treaty judicially enforceable, what is your precise basis for claiming that the particular President and Senate who signed and ratified these treaties intended for judgments of the ICJ under the treaties to be immediately enforceable in American courts of law?

David Sloss


You are the one who is attacking a straw man, not me.  I never asserted that all treaties must be self-executing.  In fact, I don’t think it’s possible to have an intelligent debate about this issue without first defining the term “not self-executing.”  Justice Roberts defined it to mean “not federal law.”  You want to change the definition to mean “not judicially enforceable.”  I suggest that we stick with Justice Roberts’ definition.  If Article 94 of the UN Charter is not federal law, then it’s obviously not judicially enforceable.  If it is federal law, we can address the question of judicial enforcement separately.

So, here are a few questions.  First, do you agree with Justice Roberts that Article 94 is not federal law?  I assume the answer is yes.  If so, how do you reconcile that view with the text of the Supremacy Clause?  Remember – the text says that ALL treaties “made under the authority of the United States” are federal law.  Let me suggest two possibilities: A) Article 94 is not “made under the authority of the United States,” or B) notwithstanding the text of the Supremacy Clause, SOME treaties made under the authority of the United States ARE NOT federal law.

Do you choose A or B?  Or is there a third option I’m overlooking?

Nick Rosenkranz


There is a third possibility, as I explained in my last post.  Under the Supremacy Clause, treaties--like Acts of Congress--are supreme law of the land, to the extent that they themselves purport to be law.  But if they do not purport to alter domestic law, then they do not.

Since no Justice considered, let alone endorsed, David’s argument, let me try to bring us back to the actual point of dispute on this issue. All the Justices agreed that a treaty can be non-self-executing, but they disagreed on how to tell if it is non-self-executing.   The Court begins with the proposition that “The interpretation of a treaty, like the interpretation of a statute, begins with its text.”  And the Court then proceeds to engage in careful and nuanced interpretation of the text of the treaty.  Likewise, even Justice Stevens, concurring in the judgment, agrees that the case turns on “the best reading of the words ‘undertakes to comply’” from the U.N. Charter.  And he agrees that those words are best read to “contemplate[] future action by the political branches.”  By contrast, the dissent insists that “the absence or presence of language in a treaty about a provision’s self-execution proves nothing at all.”  This is a very important methodological dispute, and in my view, the case is particularly heartening for its six votes in favor of the primacy of treaty text.  Do others agree?

Also, it is noteworthy that none of the Justices (expressly) resorted to a presumption to resolve this question.  The treaty is not explicitly self-executing or non-self-executing, and so the Court might have said that when there is ambiguity on this point, a presumption--one way or the other--should be used to resolve the question.  One might have thought (following a softer version of David’s argument) that the Supremacy Clause establishes a constitutional default rule or constitutional starting-point rule of self-execution.  On the other hand, one might have said that the revealed preference of the political branches, in modern times, for non-self-executing treaties justifies the opposite presumption, for non-self-execution.  The Court ostensibly did neither, interpreting the treaty text with no thumb on the scale in either direction.  And Justice Stevens explicitly noted the “[a]besn[ce] of a presumption one way or the other.”

I have written extensively about constitutional default rules and constitutional starting-point rules of statutory interpretation, see generally Federal Rules of Statutory Interpretation, 115 Harv. L. Rev. 2085, 2092-2102 (2002); that discussion is equally applicable to treaty interpretation.  But what do others think?  Is there a constitutional default rule or starting-point rule of self-execution or non-self-execution?  Or were the Justices right to read the treaties without a thumb on the scale?  

Or did they actually place a thumb on the scale, without saying so?  

Ted Cruz


I think Nick's exactly right.  But I also disagree with David's reading of the majority opinion.  Of course, all three treaties at issue (including Article 94 of the UN Charter) are "federal law," because all treaties are "federal law."  That wasn't the question before the Court.  The question was whether the treaties were "self-executing," by which the Court meant judicially enforceable in U.S. courts.  And that is a question, as Nick explains, that the Court quite properly said depends on the intent of the President and the Senate, as reflected in text, structure, and contemporaneous understanding of the treaties.

The majority opinion concludes that (1) none of the treaties, by their terms, created binding federal law, enforceable in U.S. courts, and (2) neither the Presidential Memorandum nor the ICJ's Avena decision were "federal law" at all (and, a fortiori, were not binding and enforceable in U.S. courts).

Repeatedly, when the Court talks about the treaties, it uses phrases such as "binding federal law" or "enforceable in U.S. courts".  See, e.g., slip op. at 8 ("But not all international law obligations automatically constitute binding federal law enforceable in United States courts." (emphasis added)); id., at 9 n.2 ("[A] 'non-self-executing' treaty does not by itself give rise to domestically enforceable federal law." (emphasis added)).

Just like statutes can constitute "federal law" and, by their terms, not be judicially enforceable, so too can treaties. 

David Sloss


I’ll address Nick first.

You are using the term “self-executing” to muddle the distinction among three very different issues.  First issue – does the treaty contemplate future action?  Here, we agree that it is essential to look to the text of the treaty to see what it says.  This is precisely what Marshall did in Foster

Second issue – which actors, within the domestic legal system, are responsible for taking the action required by the treaty?  It does not make any sense to look to the text of the treaty to answer this question, because treaty negotiators almost never address this question.  The firmly entrenched international legal rule is that each state must decide for itself how to implement the treaty within its own domestic legal system – that is why treaty negotiators almost never address this issue in the text of the treaty.  In Foster, Marshall most emphatically did NOT look to the text of the treaty to answer this question.  Occasionally, the President and Senate address this question by adopting a unilateral condition that is attached to the instrument of ratification.  With a couple of caveats that need not concern us here, I agree that such unilateral conditions are controlling as a matter of federal law.  However, the President and Senate did not adopt any such condition when the U.S. ratified the U.N. Charter.  Thus, the text of the U.N. Charter and the U.S. instrument of ratification don’t say anything at all about which actors in the U.S. legal system are responsible for domestic implementation.  We must look elsewhere to answer this question.

Third issue – does the treaty have the status of federal law within the U.S. legal system?  As Ted apparently concedes, the Framers of the Constitution answered this question when they wrote that ALL treaties “made under the authority of the United States” are supreme federal law.  Nick wants to rewrite the Supremacy Clause to say, “unless the treaty makers decide otherwise.”  But that is not what the Constitution says.  Since the Framers did not grant the treaty makers the power to “decide otherwise,” one must look to the text of the Constitution (and associated constitutional rules), not the text of the treaty, to determine whether a treaty is federal law.

Three other quick points:

First, contrary to Ted’s claim, Roberts really did say that the U.N. Charter is not federal law.  See Roberts slip op. at 24 (“the particular treaty obligations on which Medellin relies do not of their own force create domestic law”).  This was not mere dicta.  His further conclusions – that the courts can’t enforce Article 94, and that the President can’t execute Article 94 – are based on the premise that Article 94 is not domestic law.

Second, the nub of the problem is really this.  The Restatement (Second) of Foreign Relations Law, published in 1965, seemed to endorse a radical new version of non-self-execution doctrine that had no basis in Supreme Court precedent, and virtually no basis in lower court precedent.  See Sloss, 36 U.C. Davis L. Rev. 1 (2002).  The Restatement seems to endorse the methodology that Nick favors – examine the text of the treaty to decide whether the treaty is federal law.  Since that time, numerous lower courts have adopted this approach and many commentators have endorsed this approach.  Medellin is the first case in which the Supreme Court has endorsed this approach.  I continue to believe that this approach is fundamentally misguided.  The question whether a treaty is federal law is a question of constitutional law.  It makes no sense to examine the text of a treaty to answer a question of constitutional law.

Last point – Nick says that treaties, “like Acts of Congress,” are law if they purport to be law.  But Acts of Congress are not like this at all.  If an Act of Congress is approved by majority vote in both Houses, and signed by the President, it’s law because the Constitution says so.  If Congress inserted a clause into such a law stipulating “this Act shall not become law, even after it’s approved by both Houses, and signed by the President,” I have no doubt that the courts would hold that such a clause is unconstitutional.  Whether it’s severable is a separate matter.  Congress can, of course, do lots of other things – such as specify an effective date for the law, and establish pre-conditions that must be satisfied before the law takes effect.  But Congress can’t just say, ipso facto, that it’s not law in the circumstances where the Constitution says it is law.  The same is true for treaties.

David Sloss



I’d like to respond to Ted’s comments on judicial enforcement.  I will assume, for the sake of argument, that his interpretation of the Court’s opinion is correct – Article 94 of the UN Charter is federal law, but it’s not judicially enforceable.


To analyze this claim, it’s helpful to distinguish between “wholesale analysis” and “retail analysis” of judicial enforcement.  When courts analyze the judicial enforcement of statutory claims, they almost always engage in retail analysis.  In other words, they divide the question of judicial enforcement into several component parts: Does the plaintiff have standing? Is the claim ripe? Is the claim moot? Does the court have jurisdiction? Does the statute at issue create individual rights? Is the plaintiff a member of the class of people that the statute was intended to protect? What kind of relief is the person seeking?  The list goes on.  I am not aware of any case where a court declared that a statute was not judicially enforceable by any person at any time in any forum in any situation.  (If you know of such a case, please tell me.)  In short, courts apply retail analysis, not wholesale analysis.  Moreover, it bears emphasis that many (but not all) of the questions listed above are NOT questions of statutory interpretation.


Until about 1970, courts generally applied the same type of retail analysis to treaty claims.  Since 1970, treaty jurisprudence has been split.  In treaty cases involving claims between private parties, courts still apply retail analysis.  In a study to be published later this year, a co-author and I show that in a randomly selected set of cases, courts held treaties to be non-self-executing in fewer than one percent of the cases involving claims between private parties.  However, since 1970, courts have often applied wholesale analysis, rather than retail analysis, in treaty cases that pit private parties against government actors.  When courts apply wholesale analysis, they ask whether the treaty is “judicially enforceable,” or whether it is “self-executing,” without dividing that question into its many component parts.  Note that this approach merges together treaty interpretation questions and other questions into a single inquiry.  This is precisely the type of analysis that law professors train law students NOT to do.


This gives rise to two questions: 1) is there any justification for applying wholesale analysis in the treaty context, given that courts apply retail analysis in the statutory context? and 2) is there any justification for applying the wholesale approach in treaty cases involving government actors, given that courts apply a retail approach in treaty cases involving private parties?  My answers are “no” and “no.”


At present, I’ll just say a few things about question one.  First, it rarely if ever advances the cause of cogent legal analysis to lump together a whole set of discrete questions into a single inquiry.  Second, when applied to treaties, the wholesale approach is even more problematic than it is in the statutory context -- as I explained in my previous post, treaty negotiators almost never address questions about the proper mechanism for domestic application of treaties because they view these issues as questions of domestic law that are not appropriate to address in a treaty.  Hence, when courts examine the text of a treaty in an attempt to determine how the treaty should be applied domestically, they will almost always find (if they’re intellectually honest) that the text does not answer the question.


To avoid possible confusion, let me add one further point.  If one divides the self-execution inquiry into its component parts, there are some parts of the inquiry that are properly viewed as treaty interpretation questions.  I think we can all agree that courts must engage in careful textual analysis to answer treaty interpretation questions.  It’s a mistake, though, to think that every question related to judicial enforcement of treaties involves treaty interpretation.  When courts apply wholesale analysis, they mistakenly look to the text of the treaty to answer questions that were never the subject of negotiations among the parties.

David Sloss


In my last post, I argued that courts should approach self-execution by applying retail analysis.  Below, I explain the idea of retail analysis by applying it to Medellin.  This analysis assumes that Article 94 of the U.N.

Charter is supreme federal law.

1.  What does Article 94 obligate the U.S. to do?  Answer -- it obligates the U.S. to comply with ICJ decisions.  This is clear from the treaty text.

2.  Does Article 94 itself accomplish compliance with ICJ decisions?  Answer -- no, further action is needed to ensure compliance.  This is also clear from the treaty text.

3.  Does the UN Charter specify who, within the US legal system, is responsible for ensuring compliance?  Answer -- no, the treaty text is silent on this point.

4.  Did the President and Senate, at the time of ratification, decide who would be responsible for ensuring compliance with ICJ decisions?  Answer -- no, they recognized that they could not predict what the ICJ might decide in future cases, and that steps to achieve compliance would have to be taken on a case-by-case basis.

To illustrate point 4, consider three different remedies that the ICJ might have ordered in Avena.

First, the ICJ might have ordered the US to pay compensation to Mexico.

This would clearly have required legislative action, because congressional action is necessary to appropriate funds.

Second, the ICJ did order the US to issue a formal apology to Mexico.  This clearly requires executive action, because the President controls diplomatic communications with other countries.

Third, the ICJ did order the US to provide judicial hearings for Mexican nationals, including Medellin.  Neither Congress nor the executive branch has the capacity to comply with this decision, because they are not courts.

The judiciary is the only branch of government capable of implementing this decision -- it's the only branch that can provide a judicial hearing for Medellin.  Therefore, Article 94 requires judicial action in these circumstances.

Others will object that this argument leaves the judiciary in charge of sensitive foreign policy decisions.  This objection is unfounded.  When the Senate approved the UN Charter, and the President ratified the treaty, they decided that the United States WILL COMPLY with ICJ decisions.  Congress is free to reverse that decision by enacting legislation to preclude compliance with a particular ICJ decision.  Unless Congress acts, however, the treaty makers' decision TO COMPLY with ICJ decisions is an authoritative expression of the will of the political branches, which is binding on state and federal courts.  This does not mean that every ICJ decision is judicially enforceable: the judiciary cannot pay money to Mexico, or issue an apology to Mexico.  However, since the obligation to comply with ICJ decisions is binding federal law, and since the ICJ ordered the US to provide a judicial hearing for Medellin, and since the judiciary is the only branch of government capable of complying with that decision, the judiciary is obligated to provide a hearing.  This is what the Court should have said in Medellin.

Nick Rosenkranz



You accept that Congress can specify an effective date for a statute, but what if the effective date is a million years from now?  You accept that Congress can establish preconditions before a law takes effect, but what if the precondition is that hell freezes over?  Are such statutes supreme law of the land?  Sure, yes, in a formal sense.  But they have no legal effect.  They are not merely unenforceable in court.  They are not enforceable by the President.  They are not enforceable by anyone.  They do not change the rights of individuals or the powers of government.  In short, they have no effect whatsoever on the state of current, operative U.S. law.  Why?  Because they do not purport to.


Less fancifully, it is well accepted that Congress can limit the domestic legal effect of its acts, by specifying that they do not pre-empt state law or even trump prior inconsistent federal law.  See generally Curtis A. Bradley & Jack L. Goldsmith, Treaties, Human Rights, and Conditional Consent, 149 U. Pa. L. Rev. 399, 446-47 (2000), and sources cited therein.  Plus, of course, Congress can and does enact statutes that have no enforceable legal effect all the time, because they are--by their terms--merely precatory or aspirational.  See, e.g., 36 U.S.C. 108 ("The President is requested to issue each year a proclamation ... designating September 17 through September 23 as Constitution Week.") (emphasis added).  For countless other examples, see 36 U.S.C. 101 et. seq.  These Acts of Congress are, formally, supreme law of the land, but they have no legal effect.  They do not change the rights or duties of anyone.  Why?  Because they do not purport to.  Yet no one has ever thought that any of this is inconsistent with the Supremacy Clause.


Non-self-executing treaties are just the same.  Nothing in the Supremacy Clause prevents a treaty (or a statute) from limiting or disclaiming domestic legal effect. Chief Justice Marshall explained this in 1829, the Restatement (Second) of Foreign Relations Law restated it in 1965, and all nine of the Justices accepted it in Medellin.  The hard question is just whether any given treaty in fact disclaims domestic legal effect.  And, as six of the Justices make clear, the answer to this question is to be found in the text of the treaty.  

David Sloss



Let’s be clear about what is at stake here.  The Framers of the Constitution included treaties in the text of the Supremacy Clause to address a very specific problem.  Under the Articles of Confederation, judges in state courts were refusing to enforce Article IV of the Peace Treaty with Great Britain.  By doing so, they placed the United States in violation of its treaty obligation.


The national leaders were appalled.  So, they wrote in the Supremacy Clause that treaties are the “Law of the Land” and that “the Judges in every State shall be bound thereby.”  The purpose of this language is crystal clear – the Framers wanted to ensure that judges in state courts would not render decisions that place the U.S. in violation of its treaty obligations.


In Medellin, Texas state judges rendered a decision that places the United States in violation of its treaty obligations.  The U.N. Charter obligates the U.S. to comply with ICJ decisions.  Texas courts refused to comply with the ICJ decision in Avena.  Their refusal to comply is a violation of the UN Charter.  This is precisely the result that the Framers thought they had precluded when they drafted the Supremacy Clause.


The Framers recognized that the United States, as a sovereign state, has the power to violate its treaty obligations.  But they did not want to distribute that power to every judge in every state.  Their view was clear – if the U.S. is going to violate its treaty obligations, the decision should be made by national political authorities, not state judges. 


Do you really think it’s a good idea to have a constitutional system in which every judge in every state has the power to render decisions that place the United States in violation of its treaty obligations?  Why would you want such a system?  That is not the system the Framers created.

Nick Rosenkranz



I am glad to see that you have shifted from textual arguments to historical and pragmatic ones, since, as we discussed above, nothing in the text of the Supremacy Clause prevents a treaty (or a statute) from limiting or disclaiming domestic legal effect.


You are right that, under the Articles of Confederation, the States often placed the United States in breach of its treaty obligations.  You are right, too, that the Framers considered this a serious problem.  And, finally, you are quite correct that the Framers solved this problem with the Supremacy Clause, which empowers the President (with the Advice and Consent of the Senate) to put us in immediate and automatic compliance with a new treaty, by making it self-executing.


But the key word here is "empowers."  The Supremacy Clause empowers the federal government to ensure immediate and automatic compliance with new treaty obligations; it does not require the federal government to do so.  As you know, treaty-makers may instead, if they wish, rely on Congress to execute a non-self-executing treaty--and take the risk that the House of Representatives will refuse to do so, or that a later Congress will put us in breach.  (After all, as you say, "the United States, as a sovereign state, has the power to violate its treaty obligations.")  Likewise, the federal government may, if it wishes, rely on the States to execute a non-self-executing treaty, and take the risk that they will, perhaps, decline to do so.  Even Justice Stevens, who is generally no federalism hawk, acknowledges this possibility: "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 .... [I]t is up to Texas to prevent the breach of [the U.N. Charter]."  Medellin (Stevens, J., concurring in the judgment).  


Again, relying on the States to execute a treaty obligation is merely a constitutional option, which the federal government can exercise or not as it sees fit, depending on the treaties that it chooses to make.  But I have suggested at least one reason why, at least sometimes, exercising this option might be a good idea: "[I]n light of Congress's power to 'encourage' states with conditional spending, see South Dakota v. Dole, 483 U.S. 203 (1987), this option might prove quite effective in some circumstances, as well as flexible and respectful of federalism .... [I]f a non-self-executing treaty potentially could be executed with a variety of legislative measures, then states may be left to experiment with the different possible measures--while encouraged by conditional spending to stay within the parameters of the treaty."  Executing the Treaty Power, 118 Harv. L. Rev. 1867, 1919 n.238 (2005).   

David Sloss



It’s important to note that there is a firmly established rule of international law that treaty obligations are binding on all government actors, including legislative, executive and judicial branches, and including state and local officials in federal systems.  This is merely a default rule; states can opt out of this rule, but they must do so explicitly.  The U.N. Charter does not include any language indicating that the parties collectively opted out of this rule, and the U.S. instrument of ratification does not include any language indicating that the U.S. unilaterally opted out of this rule (which it could do by adopting a reservation).  So, as a matter of international law, the Charter obligation to comply with ICJ decisions is binding on state judges.


Your position, though, is that the Charter is not binding on Texas state judges.  I see only two ways to defend this position: 1) the U.S. did something at the time of ratification to alter the international meaning of the Charter; or 2) there’s a constitutional default rule in the U.S. that trumps the international default rule.  As noted above, the U.S. did not adopt a reservation when it ratified the Charter to alter the international meaning of the Charter.  So, unless I’m missing something, your position boils down to this – the U.S. has a constitutional default rule that treaties are not binding on state judges, and that rule trumps the international default rule that treaties are binding on state judges.  The consequence of this position is that treaties ARE NOT binding on state judges unless the treaty makers state explicitly, either in the text of the treaty or at the time of ratification, that A PARTICULAR TREATY IS binding on state judges.


There are four big problems with this proposed constitutional default rule: it’s contrary to the text of the Supremacy Clause; it’s contrary to the original understanding of the Founders; it’s contrary to dozens, perhaps hundreds, of cases decided over the past two hundred years; and it’s a bad idea as a matter of policy because it increases the likelihood that the U.S. will violate its treaty obligations, even without any affirmative decision by the federal political branches to do so.


Of course, you have five Justices on your side.  As a practical matter, that may outweigh all the other factors.

Nick Rosenkranz



Now you have shifted to arguing from international law, but in making the shift, you have conflated international law and domestic law.  Everyone agrees that the U.S. has an international legal obligation to "undertake to comply" with the ICJ opinion.  And that obligation may be binding, as a matter of international law, on state officials as well as federal officials.  But because the underlying treaty is non-self-executing, it does not, of its own force, create federal law requiring Texas to undertake to comply.  This is why Justice Stevens could write that "it is up to Texas to prevent the breach of [the U.N. Charter]," yet still decide the case for Texas.  The ICJ opinion creates an international legal obligation but no enforceable domestic federal law.


There are six Justices on my side, by the way, not five.

David Sloss


You accuse me of conflating international and domestic law.  But I was simply adopting your premise for the sake of argument.  You said that the question whether the treaty creates binding federal law is a question of treaty interpretation.  (I said this was a question of domestic constitutional law.)  Based on your premise, I argued that, if one approaches the issue as a question of treaty interpretation, it leads to the conclusion that the treaty is binding on state officers (and therefore self-executing, under the terms you defined).  Now you are trying to say that the treaty has one meaning for purposes of international law (binding on state officers) but a different meaning for purposes of domestic law (not binding on state officers).  We can ascertain the international meaning by engaging in treaty interpretation.  But if the treaty has a different meaning domestically, how, in your view, do we ascertain that meaning?  I agree that the treaty can have a different meaning domestically if the U.S. adopts conditions at the time of ratification to give it a different meaning.  But if the U.S. does not adopt any such conditions – as in the case of the U.N. Charter – how is one supposed to ascertain the domestic meaning of the treaty, if not by interpreting the international legal instrument?

Nick Rosenkranz



I don't have much more to add about this, except to echo and applaud the Chief Justice's masterful opinion, so I think that I will rest on my prior posts and let David have the last word.

Edwin Williamson


First, thanks to Ted, Nick and David for your very thoughtful posts.  I think you have thoroughly covered the non-self-execution issue, and I would like to move the discussion to some related issues.  Before doing so, but without, I hope, extending the n-s-e discussion, I would like to make a couple of observations:

I agree pretty much straight down the line with Nick, including his description of the opinion.  Even if one does not agree with everything the Chief said, I think we will all agree that he has provided us with a bright, clear opinion that should go a long way in clearing up much of the muddle in this area.

On the domestic v. international law issue, I think the US is the only (US) person bound as a matter of international law.  Other US actors are not bound, as a matter of international law, but they may cause the US to violate its obligations as a matter of international law.  To prevent this, the USG needs to take action to make sure that other actors do not do this.  Article VI permits this, but simply ratifying a treaty in accordance with one of the two Constitutionally available processes is not enough.  The treaty must, by its terms, be self-executing or there must be separate legislation.

The self-execution question is not a question of enforcement of a treaty -- it is a question whether the treaty is intended to be law.  If it is so intended, then it is the supreme law of the land under Article VI.  

Second, based on our pre-decision discussions, the most obvious next issue would be Presidential power.  In fact, the majority opinion is so clear on the issue of non-self-executing that this will be a very short discussion.  What the majority opinion very helpfully cleared up (at least for me, because this is where I really got it wrong) is that notwithstanding the President's powers and responsibility for conducting foreign policy, taking care to see that the laws are enforced, etc., unless a treaty is self-executing or Congress otherwise makes it federal law, the ratification of a treaty does not give the President authority to do anything that he otherwise is not empowered to do.  A couple of comments --

Pink, Belmont and Dames&Moore appeared to me to permit an expansion of Presidential power when accompanied by an international agreement, even without having to go through the treaty-making process. The opinion does a masterful job of limiting those cases.

Implicit in the opinion, however, is the notion that Congressional and Presidential power can be expanded under the guise of treaty-making -- pursuant to a self-executing treaty, the President could "establish binding rules of decision that preempt contrary state law" (see page 35).   Cf. Missouri v. Holland?   Of course, this expansion is subject to Constitutional limitations.  E.g., a treaty obligation to bar (or require) religious observances would not be enforceable, but what about a treaty obligation that required the US to bar anti-Canadian bias in state court proceedings?  (Loewen, a Canadian funeral company, brought an arbitration proceeding against the USG under NAFTA c. 11 after losing a race- and xenophobic- charged state court case in Mississippi.  In order to avoid liability under NAFTA for state court proceedings that did not afford Canadians "fair and equitable treatment", the USG arguably would need to adopt legislating regulating the conduct of MS state court proceedings.)  I think an argument can be made that the Convention on the Rights of the Child outlaws abortion.  Roe v. Wade would bar the enforcement of that portion of the Convention as a matter of domestic law, but were Roe not on the books, could the USG adopt a ban on abortion under the authority of Convention (I assume that such a ban is not within the President's or Congress' authority otherwise)?  (The US has signed but not ratified the Convention.

Third, notwithstanding the clarity of the Chief's opinion, I worry that it may make the process of making treaties federal law may get overly complicated and become awkward politically.  Are the following real issues or am I seeing ghosts?

The majority opinion says that a treaty becomes federal law in one of two ways: Senate consent to a self-executing treaty or legislative implementation.  Where a treaty is submitted to the Senate for its consent, is there a suggestion that the self-executing intent could not be manifested outside the body of the treaty?   Normally, the Executive branch sends the treaty up, with its explanation, including whether the treaty is self-executing, what additional authorities are needed, etc.  Could the federal-law implementing language be included in the Executive branch message?  Could the Senate do this in its resolution?  Could the Senate make a treaty more self-executing than the President indicated?  Less?

I assume that, however the Senate/Congress makes the treaty federal law, it could delegate to the President full authority as to enforcement (e.g., adopt regulations, impose penalties for violation, etc.).  Agreed?

Does the majority opinion's requirement that the Senate/Congress be a participant in a clear implementation of a treaty into federal law undercut the President's power to terminate a treaty?  Could the Senate/Congress authorize the President not to enforce the treaty -- e.g., not to enforce an ICJ decision?  Could the President reserve this authority?  Does he need to?

Read Part II: Presidential Power.

The Federalist Society