In the contemporary international legal system, each state decides how to give effect to its international legal obligations in its domestic law. Some states (those in the so-called “monist” camp) do so directly, while others (the “dualists”) require some affirmative act of “domestication.” Most fall somewhere between the two poles, taking different approaches to treaties (indeed, sometimes distinguishing between different kinds of treaties), customary international law, and general principles.
International law itself imposes no particular obligation in this regard, other than to preclude states from justifying non-compliance with their international obligations on the grounds of domestic law.Footnote 1 Given the diversity in national approaches, international agreements, and especially multilateral treaties, typically do not address the issue of domestic implementation.
For the United States, the U.S. Constitution provides what surely appears to be an unambiguous rule for the direct implementation of treaties. Article VI, Clause 2, states that:
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; and the Judges in every State shall be bound thereby, any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.
When this provision was adopted in the 1780s, the founding fathers were wrestling with an existential challenge posed by the new country's bifurcated (federal) structure to compliance with its international obligations—the refusal of states and their courts to give full effect to the peace treaty with King George. The intent of Article VI, Clause 2 could not have been clearer: the nation's treaty obligations were binding on the individual states and overrode any contrary state law.
Yet in its 2008 decision in Medellín v. Texas, the U.S. Supreme Court recognized that some treaties “do not by themselves function as binding federal law.”Footnote 2 While they constitute binding international law commitments, it said, such treaties (the so-called “non-self-executing” treaties) do not have “domestic effect” in the absence of implementing legislation.Footnote 3 “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.”Footnote 4 In consequence, the Court certainly seems to have meant that treaties in this category are not preemptive (“supreme”) law.Footnote 5
How is one to explain the yawning gap between the founding fathers’ clearly “monist” ideas about the role of treaties in our domestic legal system and the distinctly “dualist” concept reflected in the Medellín decision? That is the task Professor David Sloss, of Santa Clara University School of Law, set for himself in The Death of Treaty Supremacy, and he succeeds admirably, not only in leading his readers on a detailed explanatory journey from 1789 up through 2008 but also in challenging the legal community to consider the broader implications of this doctrinal evolution.
Simply stated, Sloss's thesis is that, over the years, through a process of “invisible constitutional transformation,” the courts, the Congress, the executive branch, legal practitioners, and scholars have together misinterpreted legal precedent and inverted fundamental constitutional principles so that, contrary to the evident intent of the founders, the mandatory “treaty supremacy rule” reflected in Article VI, Clause 2 has now become optional, leaving the constituent states free to disregard treaty obligations in the case of non-self-executing treaties. Not only is this change wrong-headed in Sloss's view, its “democratic legitimacy” is open to question.
The book has been widely applaudedFootnote 6 and earned the American Society of International Law's 2017 Certificate of Merit for a preeminent contribution to creative scholarship. It is necessary reading for all who study, practice or teach in the fields of international or foreign relations law or otherwise want or need to understand the role of treaties in the U.S. legal system.
I.
Sloss begins by explaining why the “treaty supremacy” rule was included in the Constitution at the outset. The very point of the Supremacy Clause was precisely to prevent state governments and courts from frustrating critical treaty obligations of the new nation. While the framers certainly understood the general importance of ensuring compliance with international law, their primary concern was to preserve the 1783 treaty ending the Revolutionary War. Under the Articles of Confederation, the central government lacked the power to compel state courts to provide remedies to British subjects as guaranteed under that treaty, and in response to American non-compliance the King had refused to remove his troops from U.S. territory. The issue (survival of the treaty) was unquestionably one of fundamental national security as well as national honor; resolving it was one of the main motivations for replacing the Articles of Confederation with the Constitution. The wording of Article VI, Clause 2 was intended to settle the issue in bold, underscored, with an exclamation point.
The doctrine of “treaty supremacy”—the principle that all treaties are hierarchically superior to state law and directly displace the latter to the extent of discrepancy—remained a “bedrock principle” respected and enforced by U.S. courts until the mid-twentieth century, when (Sloss argues persuasively) three principal developments combined to provoke a reinterpretation of the rule: (1) the advent of international human rights law, including the Universal Declaration of Human Rights and eventually the International Covenant on Civil and Political Rights, offering principles and mechanisms that might be used in domestic litigation to overcome racially discriminatory laws and practices; (2) the 1950 decision of a California appellate court in the Sei Fujii caseFootnote 7 invalidating a discriminatory state land law on the ground that it contravened the UN Charter; and (3) conservative opposition to the idea of “self-execution,” eventually taking the form of a proposed constitutional amendment (the “Bricker Amendment”), the aim of which was to prevent treaties from having direct domestic effect.
In the end, the appellate decision in Sei Fujii was reversed by the California Supreme Court, which invalidated the law in question on the basis of the Fourteenth Amendment to the U.S. Constitution rather than the UN Charter but, in so doing, held that treaties only supersede inconsistent local laws when they are “self-executing.”Footnote 8 A revision of the Bricker Amendment that would have elevated that principle to constitutional status narrowly failed of adoption but was effectively incorporated by the American Law Institute's (ALI) Restatement (Second), Foreign Relations Law of the United States (1965) (p. 284).Footnote 9 And courts since then have repeatedly embraced and applied the rule.
Sloss analyzes with care the reasons why the California Supreme Court chose to rule as it did, how and why its approach was incorporated into the Restatement (Second), Foreign Relations Law, and how Congress and the courts have since become comfortable with this version of “non-self-execution,” all the way up to the 2008 Medellín decision. It is, in his view, a tale of “invisible” constitutional transformation through a combination of domestic advocacy, political maneuvering, legislative efforts, and of course judicial interpretation, rather than by formal constitutional amendment.
II.
In untangling the threads of this story, Sloss articulates a number of doctrinal distinctions that help to clarify how this transformation actually occurred. Most importantly he distinguishes (1) “treaty supremacy” as an aspect of federalism in its original Article VI, Clause 2 sense from (2) “self-execution” initially understood as a function of separation of powers at the federal level. Where the first speaks to the ability of the central government to bind the states through the treaty power, the second (in its original conception) concerned only whether a treaty could be enforced directly by the executive branch without legislative implementation.Footnote 10 In his view, it is the conflation of these two doctrines (or more accurately the subordination of treaty supremacy to the idea of self-execution) that accounts for the current situation (p. 254).Footnote 11
Sloss also parses the different (and confusing) ways in which the “self-execution” doctrine has been interpreted and used by the courts and others over time.Footnote 12 The predominant (and most constrained) version is what he labels the Fujii (or “optional” treaty supremacy) doctrine—that the president and the Senate, as the constitutional treaty-makers, have the power to designate a given treaty as self-executing or not.
Some courts have posited that the distinction is not in fact for decision by the political branches but turns instead on the “intent of the treaty makers.” Sloss traces the origins of this “intent” theory to a 1926 article in this Journal by Edwin Dickinson, who argued that the issue is essentially one of treaty interpretation.Footnote 13 As Sloss points out, such a “one step” analysis is flawed for a number of reasons, including that treaties rarely address the issue textually, and thus a focus on the drafters’ intent leads the analysis astray. The proper approach, he contends, involves two distinct if related steps: first, determining whether the treaty is “executory or executed” (that is, what it requires of the parties), and then deciding who in the federal government can or should carry out that obligation (a separation of powers question).
In other contexts, “self-execution” has been used to denote that the treaty provides no private right of action in U.S. courts (i.e., may not be invoked by plaintiffs),Footnote 14 or more broadly that it provides no private right of enforcement by any litigant (i.e., cannot be invoked in defense or as an interpretive tool). These are not unique concepts; as law students are surprised to learn, domestic statutes are sometimes cabined by similar rules. More troubling, perhaps, is when the term is used to suggest that a non-self-executing treaty may be law but cannot be judicially enforced—suggesting that ensuring compliance falls to the political branches.
Perhaps the most radical approach is the one some think the U.S. Supreme Court took in Medellín—that in saying that non-self-executing treaties (like International Court of Justice (ICJ) judgments) do not themselves “give rise to domestically enforceable federal law,”Footnote 15 it meant that they are not the Law of the Land. In that decision, of course, what the Court held was that neither the ICJ's Avena judgment nor the president's memorandum seeking to compel Texas's compliance with it had automatic domestic legal effect. As to the latter point, it said, the president could not “unilaterally convert” a non-self-executing treaty into a self-executing one, since that responsibility falls to Congress—meaning by legislative action.Footnote 16 The necessary implication is said to be that the distinction is between “law” and “not law.”Footnote 17
Variants of the “optional treaty supremacy” rule have been embraced by sequential versions of the ALI's Restatement, Foreign Relations Law. The 1965 Restatement (Second), for example, stated that only a “self-executing” treaty is “effective as domestic law of the United States” and “supersedes inconsistent provisions of earlier acts of Congress or of the law of the several states of the United States.”Footnote 18 As Sloss observes, the distinction was said to turn on whether the treaty itself “manifests an intention that it shall become effective as domestic law of the United States at the time it becomes binding on the United States,” for which he contends there has never been strong authority (p. 276).
In 1987, the Restatement (Third) took a slightly different approach, that an international agreement is “non-self-executing” (and will not be given effect as law in the absence of necessary implementation) if it “manifests an intention that it shall not become effective as domestic law without the enactment of implementing legislation” or “if the Senate in giving consent to a treaty, or Congress by resolution, requires implementing legislation” or “if implementing legislation is constitutionally required.”Footnote 19 Reporters’ Comment (h) to that section clarified that “the intention of the United States determines whether an agreement is to be self-executing in the United States or should await implementation by legislation or appropriate executive or administrative action.”
The most recent formulation, in Section 310 of the Restatement (Fourth), The Foreign Relations Law of the United States (2018), accepts that “[a] treaty provision is directly enforceable in courts in the United States only if it is self-executing” and that
[c]ourts will evaluate whether the text and context of the provision, along with other treaty materials, are consistent with an understanding by the U.S. treaty makers that the provision would be directly enforceable in courts in the United States. … If the Senate's resolution of advice and consent specifies that a treaty provision is self-executing or non-self-executing, courts will defer to this specification.Footnote 20
It thus embraces the Sei Fujii approach as well as the not-judicially-enforceable interpretation.Footnote 21 The Reporters’ Notes observe, however, that “[t]he Supreme Court has not fully explained the relationship between the concept of non-self-executing treaties and the Supremacy Clause” and consequently “there is no clear reason at present to conclude that non-self-executing provisions are, as a general matter, less than supreme law.”Footnote 22
III.
In significant part, Sloss places the blame for the subordination of the treaty supremacy doctrine on resistance, in the immediate post-World War II era, to the use of international human rights norms to force changes in domestic law regarding racial discrimination and equal protection. There is no doubt, of course, that then-existing domestic law was inconsistent with the emergent international norms (which were in fact strongly supported by the federal government), that segregationists and other conservatives feared the use of those norms as a tool to permit the federal government to accomplish what it could not otherwise achieve under existing interpretations of constitutional law, and that they found justification for those fears in the appellate decision in Sei Fujii, which held in effect that, by ratifying the UN Charter, the United States had abrogated discriminatory state laws.
At the same time, it paints with too broad a brush to single out national arrogance (“American exceptionalism”) to the exclusion of substantive disagreements about the content of the international norms or the mechanics of how and by whom domestic law gets changed.Footnote 23 In the context of the times, with the recent victory over fascism fresh in the public mind and the Cold War confrontation with communism growing, the thought that the new United Nations might actually dictate domestic laws was certainly provocative for many, even those dedicated to ending segregation. Surely not all opposed to automatic acceptance and direct enforcement of international norms were then (or are now) racist or anti-democratic,Footnote 24 and an uncritical acceptance of the de facto superiority of international norms (“reverse exceptionalism”) is as ill-considered as an unthinking rejection of those norms.
In any event, it is hardly “exceptionalist” to have concerns about “legislating” directly through the treaty power. One can hear such concerns even today, in contexts far removed from issues of non-discrimination and equal treatment, as recent treaty ratification debates in the Senate have demonstrated.Footnote 25 Indeed, versions of the Bricker Amendment are still introduced in the House and Senate from time to time.
Sloss rightly acknowledges that the optional “non-self-executing” doctrine adopted by the California Supreme Court actually helped to mitigate the tension between human rights and states’ rights and to defeat the Bricker Amendment (describing it as “a compromise between the normative preferences of competing groups”) (p. 322). Substantively, he embraces the doctrine as a defensible interpretation of the Constitution, while rejecting some of its more recent transformations, including the “no private right of action” and “non-judicially enforceable” interpretations (p. 255).
He is also correct to point out that substantively the gradual emergence and eventual acceptance of the “optional treaty supremacy” doctrine was a clear modification of what the framers intended. He contends that “[i]nvisible constitutional transformation is difficult to reconcile with principles of democratic legitimacy because ‘the people’ cannot easily mobilize to resist constitutional constructions that are shielded from public view” (p. 329). True enough, although the transformation in question was not really “invisible” or “shielded” in any meaningful sense, since as he acknowledges, the debate in fact played out loudly and publicly in the Congress, the courts, the press and other fora such as the American Law Institute.
Whether it lacked the requisite “political transparency” or was less than fully consistent with principles of democratic legitimacy is more debatable. If the point is that a decision on how to interpret Article VI, Clause 2 should have been taken by “the people,” perhaps by popular plebiscite or formal amendment of the Constitution, one can wonder how such a process might have worked and whether the outcome would actually have been at the time (or would likely be today) in favor of returning to an originalist “direct incorporation” rule.
A bit more clearly, Sloss reflects an evident preference to have left the issue entirely to the courts, in particular the U.S. Supreme Court. He correctly points out that the story of our Constitution has largely been one of judicial adaptation and reinterpretation in light of changed circumstances but the Supreme Court “played almost no role in establishing the contours of the de facto Bricker Amendment” (p. 323). More broadly, he notes, “the history of the de facto Bricker Amendment suggests that constitutional change outside the courts is not always consistent with principles of democratic legitimacy” (p. 13).
Yet much of the change (and not a little of the doctrinal confusion) in this area has in fact occurred through judicial decision-making (beginning with Ware v. Hilton Footnote 26 and Foster v. Neilson Footnote 27), and many point in particular to the Medellín decision as a distinct failure of the Supreme Court to clarify the issue. Regarding constitutional transformation in general, one can enthusiastically embrace the importance of an independent judiciary as protection against the political pressures of the legislative process, without necessarily accepting that the latter is less legitimate democraticallyFootnote 28 or that the courts are more accountable to the public.Footnote 29 Neither is fidelity to the intent of the founders by itself the only appropriate measure of democratic legitimacy; if it were, a number of constitutional amendments would fail the test.
IV.
In debating the merits of the “optional treaty supremacy” rule, account also needs to be taken of the dramatically changed nature of treaties today. At the time of the founding, most treaties involved straightforward questions of external bilateral relations like war and peace, boundaries and trade. These were clearly matters for which national governments were accountable and unable to tolerate non-compliance by their subordinate components (like states, counties or provinces). Today, treaties (especially multilateral treaties) deal increasingly with “internal” matters (such as criminal law, family law, commercial law, tax, intellectual property, judicial jurisdiction, etc., as well as individual rights) on which substantial domestic law already exists. The point of such treaties is frequently to achieve some measure of international conformity. In consequence, compliance will typically require changes to domestic law.
To illustrate, it is one thing to conclude a bilateral peace treaty that requires one party to provide mechanisms to consider the claims of the other party's nationals for compensation for wartime damage and loss, and a very different thing to accept legally binding obligations to ensure that in domestic law and practice the education of all children is directed to the development of the “full potential” of their personality, talents, and mental and physical abilities, provides every child a legally enforceable right to a “standard of living adequate for the child's physical, mental, spiritual, moral and social development,” and guarantees “the right of the child to rest and leisure, to engage in play and recreational activities appropriate to the age of the child and to participate freely in cultural life and the arts.”Footnote 30 Laudable and important as those goals may be, the question (for present purposes) is not whether they are worthy of embrace and implementation as moral or policy matters but whether incorporating them directly into federal law by means of the treaty power is the best mechanism for giving them domestic effect.Footnote 31
Unquestionably, there is today much greater acceptance in the United States (as compared to the Sei Fujii era) of the validity of international standard-setting and the utility of complying with internationally agreed rules including by incorporating or enacting them into U.S. law. Even so, the U.S. Senate remains unenthusiastic about such important instruments as the Convention on the Elimination of All Forms of Discrimination Against Women, the Convention on the Rights of Persons with Disabilities, or the UN Convention on the Rights of the Child (to say nothing of the UN Law of the Sea Convention or the Rome Statute).
One reason is substantive. Such multilateral treaties are necessarily negotiated on a global scale, in international organizations whose member states come from every part of the world and have markedly different legal systems and widely diverse approaches to the role of the courts. Few have federal structures. In consequence, the final formulation of the treaty is unavoidably the product of compromise and often worded so that the obligations it imposes simply cannot be incorporated directly into domestic law (at least in the United States). Careful legislative drafting is thus required to ensure that the treaty obligations can in fact be properly implemented in domestic law.Footnote 32
Another reason is that those obligations (especially in the human rights area) often implicate matters falling outside the scope of constitutionally delegated federal responsibility or occasionally, given our notions of limited government, beyond the reach of government regulation as a whole (the above excerpts from the Rights of the Child Convention provide some illustrations). By virtue of Article VI, Clause 2, duly ratified treaties unavoidably “federalize” the matters they cover; whether and when they can in fact augment the federal government's limited powers as delegated under the Constitution, or permit the federal government to act in areas otherwise preserved to the states under the Tenth Amendment, has been a recurring issue even before and after Missouri v. Holland,Footnote 33 and controversies about it persist.Footnote 34
There is also a question about how much the “self-execution or not” issue actually matters. In point of fact, the vast majority of treaties ratified by the United States in recent years (whether bilateral or multilateral) are not functionally self-executing. While many (mostly bilateral treaties) have been ratified on a “self-executing” basis, they have overwhelmingly been implemented legislatively; this is true in particular for bilateral tax, extradition and mutual legal assistance treaties, which are given effect in federal law through preexisting legislative structures. The record on self-executing multilateral treaties is mixed; some (such as the 1999 Montreal ConventionFootnote 35) work well but domestic compliance with others (such as the 1963 Vienna Convention on Consular Relations and the UN Convention on the International Sale of GoodsFootnote 36) has proven problematic.Footnote 37 The difficulties the courts have had in interpreting and applying these two treaties in particularFootnote 38 raise serious questions about the assumption that compliance is guaranteed by leaving the issue entirely to the judiciary. Recent U.S. treaty practice demonstrates that neither the executive or legislative branch has voiced a clear preference for self-execution.
V.
The Supreme Court's decision in Medellín was primarily about whether the ICJ's judgment in Avena had (or could be given) direct domestic effect, and the critical treaties at issue (the UN Charter and the Optional Protocol to the Vienna Convention on Consular Relations) were undeniably not self-executing. While there is room therefore to argue that the decision did not establish a definitive presumption against treaty self-execution, there can be little question that in rejecting the president's argument that he could unilaterally make the Avena judgment enforceable against Texas or convert a non-self-executing treaty into a self-executing one, the Court saw the difference as involving a lawmaking function that must necessarily involve the Congress. On that point, one has to think that the founders would agree.
In this reviewer's opinion, Sloss misses the point in criticizing the Medellín Court for holding that Texas did not violate Article 94 of the UN Charter because that provision is not federal law (pp. 311, 313). The real problem was that Texas (like many other U.S. states) failed to comply with, and thereby put the United States in violation of, the “consular notification and access” obligations in Article 36 of the Consular Convention. That treaty is self-executing, yet the executive branch was unable to compel Texas to comply with its requirements or the ICJ's Avena judgment.Footnote 39 Most observers agree that U.S. compliance with Article 36 is self-evidently deficient. The situation continues and cries out for a solution, but so far, the states have been resistant, the courts have declined to enforce Article 36 (largely on grounds of the “procedural default” rule), and the Congress has shown little interest in addressing the issue. As a result, the efficacy of self-executing treaties is open to serious question, at least in some circumstances.
Sloss is emphatically correct to stress the importance of fidelity to our treaty obligations and the need to prevent states from “engag[ing] in conduct that would trigger a violation of U.S. treaty obligations” (p. 119). Yet it appears that to some extent he has misconceived how the “non-self-executing” doctrine is actually employed. It is not an “opt out” clause that automatically frees the states to disregard the nation's treaty undertakings or allows them to drive “a wedge between the nation's international obligations and the domestic implementation of those obligations” (p. 123).
The United States does not ratify treaties with which it knows it will be unable to comply. When the executive branch determines that compliance with the treaty in question would require a change in relevant domestic law (federal or state) and that self-execution is (for one of a number of reasons) not the appropriate method for accomplishing that change, then it will propose implementing legislation and (if the Senate gives its advice and consent to ratification) defer deposit of the instrument of ratification until that legislation has taken effect in accord with normal constitutional procedures (and sometimes that can mean a substantial waitFootnote 40). The point is precisely to avoid the situation that Sloss is worried about.Footnote 41
In pursing that goal, U.S. practice is more nuanced and flexible than many appreciate.Footnote 42 In some cases, the United States considers that the treaty can be ratified on a non-self-executing basis because domestic law and practice already comply with the relevant treaty requirements.Footnote 43 Treaties that do require changes in domestic law can be self-executing (in whole or in part) but very often are implemented by federal legislation, which can be very detailed or comparatively simple.Footnote 44 In a few cases, the text of the treaty itself has been incorporated directly into U.S. law by federal statute.Footnote 45 In others, the requisite implementing legislation has been adopted at the state (rather than federal) level or even by a combination of methods.Footnote 46 In still others, state and federal courts have been given concurrent jurisdiction over actions arising under the treaty.Footnote 47
Professor Sloss usefully highlights the complexity of treaty implementation issues, which continue to challenge those in the U.S. legal community concerned with ensuring fidelity to our international obligations. The “self-executing/non-self-executing” distinction remains confusing and contentious, yet it is a mistake to cast it as creating a stark binary choice (whether the treaty in question is binding/not binding, judicially enforceable/not enforceable, or law/not law). He is right to reject a constitutional interpretation that permits or encourages U.S. ratification of treaties under conditions that allow the states to frustrate national compliance. If, however, the only other option is to incorporate all treaties directly as preemptive federal law, we will then be disabled, legally and politically, from joining important multilateral regimes at the international level.
