Non-Self-Executing Treaties in the Draft Restatement of Foreign Relations Law

At its annual meeting on May 18–20, the American Law Institute (ALI) will consider portions of a draft Restatement (Fourth) of Foreign Relations Law. Unfortunately, the most recent draft of the Restatement on Treaties appears to endorse a variant of non-self-execution (NSE) doctrine that is unconstitutional. This essay distinguishes among five possible interpretations of NSE declarations. Under two of those interpretations, NSE declarations are clearly constitutional. Under one other interpretation, NSE declarations are probably constitutional. Under the last two interpretations, NSE declarations are probably unconstitutional. The April 2015 Discussion Draft on Treaties appears to endorse the most constitutionally problematic interpretation of NSE declarations.

Preliminary Observations

Judicial decisions applying the doctrine of self-executing treaties persistently conflate two distinct questions: 1) what is the content and scope of the international obligation created by the treaty? (the “international obligation” issue); and 2) which government actors in the United States have the domestic legal authority and/or the domestic legal duty to implement the treaty? (the “domestic application” issue). Given this distinction, one can divide self-execution doctrines into two categories. Under “indirect versions” of the doctrine, the treaty makers use their Article II power to make decisions about the content and scope of the international obligation and certain domestic consequences follow. Under “direct versions” of self-execution doctrine, the treaty makers use their Article II power to make decisions about domestic application of the treaty without modifying the international obligation.

Chief Justice Marshall’s opinion in Foster v. Neilson is open to several competing interpretations. However, one thing is crystal clear. Marshall was applying an indirect version of self-execution doctrine. He analyzed the treaty language to ascertain the content and scope of the international obligation contained in the treaty. He drew certain conclusions about domestic application based on his analysis of the international obligation. Nothing in Marshall’s opinion indicates that he believed that the treaty makers used their Article II power to control domestic application directly, without modifying the international obligation.

In contrast, when the treaty makers ratify a treaty subject to an NSE declaration, they are using their Article II power to control domestic application of the treaty directly, without modifying the international obligation. Like Marshall’s opinion in Foster, NSE declarations are subject to several competing interpretations. However, one thing is crystal clear. NSE declarations are based upon a direct version of NSE doctrine, not an indirect version. In every case where the United States has adopted an NSE declaration, the treaty makers have made clear that the declaration does not modify the content or scope of the international obligation contained in the treaty.

The recent Discussion Draft fails to distinguish between direct and indirect versions of self-execution doctrine. That failure creates serious problems. From a constitutional perspective, direct and indirect versions of NSE doctrine are fundamentally different. Indirect versions of the doctrine assume that Article II grants the treaty makers an affirmative power to shape the content and scope of international obligations. That assumption is unassailable. Direct versions of the doctrine assume that Article II grants the treaty makers an affirmative power to control domestic application of treaties directly, without modifying the international obligation. In my view, Article II does grant the treaty makers such an affirmative power. However, it is a limited power, not an unlimited power. As explained in more detail below, by failing to distinguish between direct and indirect versions of self-execution doctrine, the Discussion Draft implies that Article II grants the treaty makers an unlimited power to control domestic application of treaties directly. That proposition is untenable.

Five Interpretations of NSE Declarations

Some treaty provisions address matters that, absent a treaty, would be regulated in whole or in part by state law in the United States. Other treaty provisions address matters that, absent a treaty, would be regulated exclusively by federal law. I focus here on NSE declarations as they apply to treaties that intersect with areas of state regulatory authority.

To illustrate application of NSE declarations in that context, consider the following hypothetical case. Assume that the United States ratifies the Convention on the Rights of the Child. Article 37(a) states: “Neither capital punishment nor life imprisonment without possibility of release shall be imposed for offences committed by persons below eighteen years of age.” Assume, further, that the United States does not adopt a reservation to article 37, but it does adopt an NSE declaration. (These assumptions are admittedly unrealistic. Nevertheless, they are useful for clarifying the meaning and effect of NSE declarations.) What is the effect of that declaration? Consider a criminal defendant in Florida who is charged with murder. Defendant was 17-years-old when he committed the crime. The State seeks a life without parole (LWOP) sentence, which is authorized under state law. Defendant invokes the treaty as a defense. Article 37(a) clearly bars the LWOP sentence. Should the court apply the treaty to bar the sentence? Or does the NSE declaration preclude the court from applying the treaty? To address this question, I consider five different variants of an NSE declaration. All five variants share one feature in common: they do not alter the content or scope of the international obligation. Their effects are purely domestic.

Variant One – The NSE declaration states: “This treaty shall not create a private cause of action in U.S. courts.” This declaration is based on the private right of action doctrine, which holds that Article II grants the treaty makers an affirmative power to decide directly, without modifying the international obligation, that a treaty will not create a private right of action. I assume that the declaration is constitutionally valid. In the hypothetical Florida case, the declaration would not preclude the defendant from invoking the treaty as a defense, nor would it preclude the court from applying the treaty. The court would presumably rule that the State may not impose an LWOP sentence because the treaty supersedes state law under the Supremacy Clause. The authors of the Discussion Draft apparently agree that the private right of action doctrine is constitutional. However, they contend that NSE declarations should be construed to “preclude all judicial enforcement,” rather than “merely denying a private right of action” (Discussion Draft, § 106, Reporters’ Note 4).

Variant Two – The NSE declaration states: “State and federal courts shall construe the treaty so as to avoid conflicts with state and local law, whenever reasonably possible.” This declaration is not based on any version of NSE doctrine. However, this variant merits consideration because Professor Bradley — one of the Reporters for the ALI Restatement on Treaties — has argued that NSE declarations are like statutory anti-preemption provisions. This declaration is similar to certain statutory anti-preemption provisions.

I assume that the anti-preemption declaration is constitutionally valid. (If one deleted the phrase “whenever reasonably possible,” the constitutional validity of the declaration would be more doubtful.) In the hypothetical Florida case, the court would presumably find that the declaration does not apply because it is not “reasonably possible” to construe the treaty in a manner that avoids conflict with the state law authorizing LWOP sentences for juvenile offenders. Therefore, the court would apply the treaty to bar the LWOP sentence. Nevertheless, in any conceivable real-world case the declaration would accomplish the objectives that advocates claim the NSE declarations are intended to accomplish because cases involving unavoidable conflicts between treaties and state law are routinely handled by adopting treaty reservations. The “anti-preemption” declaration would complement treaty reservations by addressing those cases where it is unclear whether the treaty conflicts with state law.

Variant Three – The NSE declaration states: “This treaty shall not supersede conflicting state laws unless Congress enacts legislation stipulating that the treaty supersedes conflicting state laws.” This declaration is based on the Fujii doctrine, which holds that the treaty supremacy rule codified in the Supremacy Clause is merely a default rule, not a mandatory rule. (The treaty supremacy rule says that treaties supersede conflicting state laws.) Accordingly, Article II grants the treaty makers an affirmative power to decide that a particular treaty will not supersede conflicting state laws. Variant Three differs from Variant Two because the declaration under Variant Two operates as interpretive guidance to courts, whereas the declaration under Variant Three modifies the ordinary operation of the treaty supremacy rule.

The constitutionality of the Fujii doctrine can be challenged on the grounds that it is in tension with the text of the Supremacy Clause. Even so, the Fujii doctrine is defensible on two distinct grounds. As an historical matter, the doctrine gained broad acceptance in the early 1950s as a way to mediate the tension between then-emerging human rights treaties and the traditional treaty supremacy rule, which had previously been understood as a mandatory rule.* History aside, there is a persuasive argument by analogy. The Fourteenth Amendment grants Congress an affirmative power to override the Eleventh Amendment. That Fourteenth Amendment power helps preserve the federal-state balance by ensuring that Eleventh Amendment immunity does not negate federal supremacy. By analogy, one could argue that Article II grants the treaty makers an affirmative power to override the treaty supremacy rule. That Article II power helps preserve the federal-state balance by ensuring that the combination of the Treaty Power and the Supremacy Clause does not negate state autonomy.

Assume that the Fujii version of the NSE declaration is valid, and that Congress has not enacted relevant legislation. In the hypothetical Florida case, the court would presumably impose an LWOP sentence because the declaration stipulates that the treaty does not supersede state law.

The Discussion Draft appears to reject the Fujii doctrine. The Draft is not entirely unambiguous on this point, but it says that treaties are “the supreme Law of the Land,” (Discussion Draft, § 101, cmt. a) that they are supreme over State and local law,” (Discussion Draft, § 104) and that they “displace contrary State and local law” (Discussion Draft, § 104, cmt. A). None of the quoted passages distinguishes between self-executing and non-self-executing treaties. Thus, these statements imply that even NSE treaties supersede conflicting state law. Under the Fujii doctrine, though, NSE treaties do not supersede conflicting state law, because the treaty makers exercised their Article II power to override the default rule in the Supremacy Clause. In my view, the Fujii doctrine provides a more constitutionally defensible interpretation of NSE declarations than either of the variants discussed below.

Variant Four – The NSE declaration states: “This treaty supersedes conflicting state laws. However, state and federal courts shall not apply the treaty to override state law unless Congress enacts legislation authorizing judicial enforcement.” This declaration is based on the no judicial enforcement (NJE) doctrine. The NJE doctrine holds that Article II grants the treaty makers an affirmative power to order state and federal courts to refrain from applying a treaty that has the status of supreme federal law, even in cases where the treaty is otherwise applicable. The doctrine assumes that such an order is binding on state and federal courts.

In my view, the NJE version of the NSE declaration is unconstitutional. The assumption that Article II grants the treaty makers an affirmative power to order courts not to apply supreme federal law is inconsistent with core constitutional principles. Consider the hypothetical Florida case. Under the NJE doctrine, Article 37 of the treaty supersedes conflicting state law, rendering the LWOP sentence illegal as a matter of federal law. Nevertheless, the judge is permitted to impose an LWOP sentence because the NSE declaration precludes courts from applying the treaty to invalidate the state law authorizing an LWOP sentence. That outcome cannot be correct. Just this term, in Armstrong v. Exceptional Child Center, Inc., a case implicating the correct interpretation of the Supremacy Clause, the Supreme Court stated: “To say that the Supremacy Clause does not confer a right of action is not to diminish the significant role that courts play in assuring the supremacy of federal law. For once a case or controversy properly comes before a court, judges are bound by federal law. Thus, a court may not convict a criminal defendant of violating a state law that federal law prohibits.” Similarly, a court may not impose a sentence on a criminal defendant that federal law prohibits. Under the NJE version of NSE doctrine, though, a court may impose a sentence on a criminal defendant that “federal law” prohibits, because an NSE treaty is “federal law.” This is not true under the Fujii doctrine, because under the Fujii doctrine an NSE treaty is not supreme federal law. Thus, the Fujii doctrine avoids the central constitutional difficulty inherent in the NJE doctrine.

The Discussion Draft appears to endorse the NJE doctrine. Section 104 says that all treaties (presumably including NSE treaties) are supreme federal law. Section 106 says that NSE treaties are not “directly enforceable by the judiciary.” Reporters’ Note 4 to Section 106 suggests that NSE declarations are binding on the judiciary, and that they bar direct judicial enforcement by the judiciary. Thus, the Discussion Draft appears to endorse an unconstitutional version of NSE doctrine.

In his published writing, Professor Bradley has defended the validity of the NJE doctrine. His argument can be summarized briefly as follows. Premise One: Article II grants the treaty makers an affirmative power to control domestic application of treaties without modifying the content or scope of the international obligation. (I accept this premise.) Premise Two: the Supremacy Clause does not obligate courts to apply all treaties in all cases; some limits on judicial enforcement are entirely consistent with the Supremacy Clause. (I accept this premise also.) Conclusion: Article II grants the treaty makers an affirmative power to bar judicial enforcement of a treaty by all courts in all circumstances. (I reject this conclusion.)

Professor Bradley’s conclusion does not follow from his premises because the preceding argument conflates the critical distinction between direct and indirect versions of NSE doctrine. Premise One applies to direct versions of NSE doctrine. Premise Two, though, is based on an indirect version of NSE doctrine: certain types of treaty provisions are not judicially enforceable because of the nature of the international obligation codified in the treaty. Imagine a hypothetical treaty provision obligating the United States to “make best efforts to eliminate poverty.” Even without an NSE declaration, it is difficult to imagine any case in which a state or federal court would enforce this treaty provision, because the content of the international obligation is such that domestic application of the treaty is beyond the scope of judicial competence. The fact that some treaties create international obligations whose application is beyond the scope of judicial competence does not mean that Article II grants the treaty makers an affirmative power to bar judicial enforcement of treaty provisions that are within the scope of judicial competence. As the Supreme Court’s recent decision in Armstrong makes clear, courts have a constitutional duty to apply a treaty that has the status of “federal law” whenever a “case or controversy properly comes before a court,” if the treaty creates a rule of federal law that courts are competent to apply.

Variant Five – The NSE declaration states: “This treaty supersedes conflicting state laws. However, no private party may invoke the treaty as the basis for a claim or defense in any state or federal court unless Congress enacts legislation authorizing private judicial enforcement.” This declaration is based on the no private enforcement (NPE) doctrine. The NPE doctrine holds that Article II grants the treaty makers an affirmative power to decide that private parties may not invoke a particular treaty as the basis for a claim or defense, absent implementing legislation. The doctrine assumes that such a decision is binding on state and federal courts. The NPE doctrine differs from the NJE doctrine because it does not preclude the federal government from suing to enforce a treaty. The NPE doctrine differs from the private right of action doctrine because it precludes individuals from invoking treaties defensively.

The Discussion Draft neither endorses nor repudiates the NPE doctrine. Based on the text of the draft Restatement, it appears that the Reporters have not considered the NPE doctrine as a variant distinct from the NJE doctrine.

Consider application of the NPE doctrine to the hypothetical Florida case. Under the NPE doctrine, Article 37 of the treaty supersedes conflicting state law, rendering the LWOP sentence illegal as a matter of federal law. Nevertheless, the judge is permitted to impose an LWOP sentence because the NSE declaration precludes the defendant from invoking Article 37 as a defense, unless Congress enacts implementing legislation. Thus, the NPE doctrine is arguably unconstitutional for the same reasons that the NJE doctrine is unconstitutional: it allows a court to impose an illegal sentence on a criminal defendant.

On the other hand, two distinct arguments support the validity of the NPE doctrine. First, statutory provisions in legislation implementing NAFTA, the WTO, and other free-trade agreements tacitly assume that Article I grants Congress an affirmative power to decide that private parties may not invoke a particular international agreement as a basis for a claim or defense. If Congress has such a power under Article I, then one can make a persuasive argument that Article II grants the treaty makers a comparable power. Second, in the Armstrong case discussed above, Justice Scalia wrote that judges are bound by federal law “once a case or controversy properly comes before a court.” Under the NPE doctrine, if the defendant in the hypothetical Florida case invokes Article 37 as a defense, one could argue that the treaty issue is not properly before the court, because the NSE declaration precludes the defendant from invoking Article 37, even defensively.

Despite these arguments, I believe that the NPE doctrine violates the Due Process Clause because, in cases like the hypothetical Florida case, the NSE declaration would deprive the defendant of his constitutionally protected “opportunity to be heard” on his treaty-based defense. In a prior article, I argued that the relevant statutory provisions in legislation implementing free-trade agreements are unconstitutional for the same reason. Most scholars and government officials have assumed that such provisions are constitutional, but no one has presented an analytically rigorous defense of that assumption.

Conclusion

The ALI should repudiate the NJE version of NSE doctrine because the NJE doctrine tacitly assumes that Article II grants the treaty makers an affirmative power to control domestic application of treaties that is not bounded by the ordinary constitutional rules that constrain other acts of federal lawmaking. Specifically, the doctrine assumes that Article II grants the treaty makers an affirmative power to order courts to disregard supreme federal law, even in cases where a case or controversy is properly before the court. The Discussion Draft appears to endorse the NJE doctrine. The Reporters should re-think their approach to NSE doctrine as they continue their important work on the Restatement.

__________________

* See, e.g., Preliminary Draft 5 of the Restatement (Second) of Foreign Relations Law, sec. 3.04, cmt. c (Jan. 1959) (“The provisions of Art. VI, § 2 of the Constitution . . . are in effect, permissive rather than mandatory.”). 

About the Author(s)

David Sloss

Professor at the Santa Clara University of Law