Executive Agreements

Executive Agreements

The Treaty Clause lays out one procedure for binding the nation to a foreign government. The President negotiates, and the agreement takes effect as a treaty only if two-thirds of the senators present consent.¹ The United States also makes international commitments by a second route, the executive agreement, which is a compact with a foreign government concluded without that procedure. The Constitution never mentions executive agreements by name, yet they are an established and heavily used instrument of American diplomacy. The trials in Reid v. Covert, discussed in the preceding section, rested partly on agreements of this kind with Britain and Japan.

Because an executive agreement skips the Senate's supermajority, it raises a question the treaty procedure answers automatically: where does the authority come from? The two-thirds requirement exists to keep a thin or temporary coalition from committing the nation at the highest level of domestic law. An agreement that bypasses it therefore needs some other source, and the candidates are a grant of power from Congress and the President's own constitutional powers. Jackson's three categories, introduced earlier in this chapter, supply the lens, because the strength of the President's position depends on how the agreement relates to the will of Congress. A second question follows once the agreement exists. A court asked to apply it must decide what rank it holds against state law and against federal statutes, an issue the Supremacy Clause settles expressly for treaties and leaves unaddressed for other agreements.

These questions are taken up in three steps below. The first sorts executive agreements by the source of authority behind them. The second examines a decision in which the Court tested that authority against a President's settlement of private claims. The third asks how much legal force a valid executive agreement carries once it is in place. The discussion begins with the kinds of executive agreements.

Types of Executive Agreements

An executive agreement is an international compact that the United States concludes without satisfying the Article II requirement of Senate advice and consent by two-thirds of the senators present.² The practice is old. Since the administration of George Washington the executive branch has entered into thousands of such compacts, and the authority to do so is treated as inherent in nationhood, an exception to the usual rule that the national government's powers are enumerated.³ Presidents turn to the form for practical reasons. An executive agreement can take effect despite Senate opposition that would defeat a treaty, and even where Senate support exists it can take effect at once, without the delay that ratification sometimes brings.⁴ Much routine international business therefore moves this way, because the form streamlines the process and avoids controversy.⁵ Practical limits also operate. A President whose agreement needs implementing legislation or an appropriation risks alienating Congress by leaving it out, and Congress can answer with a statute that overrides the agreement or requires renegotiation.⁶

Skipping the two-thirds vote removes the Senate from the process, so an executive agreement must draw its authority from some other source, and the standard classification sorts these agreements by what that source is.⁷ The types are these:

1. Congressionally authorized agreements rest on a prior statutory delegation or on later statutory implementation, and they are often called congressional-executive agreements.
2. Treaty-authorized agreements rest on the provisions of a preexisting treaty.
3. Sole-executive agreements rest on the President's own constitutional authority, such as the power to recognize foreign sovereigns or to command the armed forces.

The line between the first and third types turns on where the power to conclude the agreement comes from. When the President relies on authority that a statute confers, the agreement completes an arrangement that Congress set in motion by delegating, while an agreement resting on inherent Article II authority is a sole-executive one.⁸ The label matters because the validity of an agreement is measured, at least in theory, against the power behind it. If Congress uses its foreign commerce power to authorize a mutual trade agreement, the agreement's constitutionality is tested against the scope of that power, and if a treaty alone supports an agreement, its validity is derivative of the treaty's.⁹

Congressional authorization is the most secure footing. In Field v. Clark the Court accepted a grant of reciprocal tariff-making authority to the President, holding that it neither delegated legislative power invalidly nor invested the President with treaty-making power.¹⁰ A statute authorizing the President to set postal rates with foreign countries likewise permits an agreement for that purpose, whether or not the statute says how the rates are to be set.¹¹ The Senate and the House have already played their constitutional roles in passing the statute, so the agreement takes nothing from either. One caveat qualifies the picture. The federal government's authority to carry out a congressional-executive agreement stays within the scope of Congress's enumerated powers, whereas an Article II treaty can reach beyond them under Missouri v. Holland, so the executive-agreement route is not a perfect substitute for a treaty on every subject.¹²

Agreements authorized by an existing treaty follow the same logic. The President has an Article II duty to execute the laws and treaties of the United States, and the Senate has already consented to the treaty being carried out, so using an executive agreement to implement it leaves the Senate's role intact.¹³ Some treaties expressly state that they may be implemented through executive agreement, and the route may be appropriate even when a treaty is silent. The agreement's constitutional footing is only as sound as the treaty's, since its power is derivative.

Sole-executive agreements depend on whether the subject falls within the power the President invokes. In United States v. Pink and United States v. Belmont the Court upheld the Litvinov Agreement, under which the United States recognized the Soviet Union and the Soviet Union assigned to the United States its interest in a Russian insurance company in New York, assets the government would use to pay American creditors' claims. The Court explained that, because the agreement was not a treaty, Senate approval was not required, and it held that states must comply with such agreements.¹⁴ The recognition power supplied the authority there. In the same way, an agreement with France to hold joint naval exercises in the Pacific rests on the commander-in-chief power, and an agreement designating the site of a United States embassy or consulate abroad rests on the power to receive ambassadors, which includes recognizing and dealing with foreign governments.¹⁵ In each situation the Constitution has assigned the power to the President, so choosing a form that avoids a Senate vote neither usurps the Senate's authority nor enlarges the executive's domain.

All three types are treated as complete alternatives to treaties, so an executive agreement can address any subject a treaty could, and the choice between them is essentially a political judgment for the executive branch.¹⁶ Major commitments have gone this way. In 1940 the Destroyer-Bases Agreement exchanged fifty American naval destroyers for ninety-nine-year leases on base sites in the Caribbean and Newfoundland, and it substantially expanded American involvement in World War II.¹⁷ The Supreme Court has never held an executive agreement invalid on the ground that it exceeded the national government's powers, and because the inherent foreign affairs power is so elastic, the limits that exist are largely theoretical.¹⁸ An agreement within the scope of the President's powers remains subject to the Constitution's guarantees of individual rights, which is the lesson of Reid v. Covert in the preceding section.¹⁹

The fact that every type is permitted does not settle the separation-of-powers question, because an agreement that skips the Senate can still look like a way of evading it. A treaty needs the votes of two-thirds of the senators present, and to the extent the Framers saw the Senate as a real check on excessive foreign entanglements, an executive agreement that bypasses it may raise serious concerns.²⁰ Senator Bricker's proposed constitutional amendment of the 1940s and 1950s, which would have curtailed executive agreements, never passed, but it reflected deep worry that the President could circumvent treaty ratification.²¹ The concern is graduated, and the three sources of authority fall along a sliding scale, much as Jackson's categories scale the President's power with the position Congress has taken.

The worry is weakest when the President uses an agreement to exercise a power the Constitution already gives the executive, as in the naval exercise and the embassy site, because nothing is taken from the Senate.²² It is also weak when the agreement implements a statute or treaty, since the Senate or Congress has already approved the measure being carried out.²³ The concern is strongest when the agreement does neither, and especially when it has a major domestic effect. In that setting the President may be taking over Congress's lawmaking role or the Senate's approving role, so the agreement may be invalid unless Congress has expressly or impliedly authorized its use.²⁴ An agreement with Mexico letting produce into the country despite state pesticide laws, made with no authorizing treaty or statute, is a case of this kind. It in effect regulates foreign commerce, which Article I assigns to Congress, so it is probably invalid and cannot displace the state rules. If the Senate had earlier ratified a treaty under which both countries agreed to encourage the exchange of agricultural produce, the same agreement would likely stand as an exercise of the President's duty to carry out treaties.²⁵ The Court's decision in Dames & Moore v. Regan fits the pattern. It upheld an agreement that displaced private claims against Iran in American courts, and it called Congress's implicit approval of the practice crucial to the decision, which suggests that without such approval the agreement would have failed.²⁶ The next section turns to that case.

Claims Settlement and Dames & Moore

Executive agreements usually draw their strength from a power the Constitution gives the President directly, or from a statute or treaty that stands behind them. Dames & Moore v. Regan tested an agreement whose most controversial feature had no express footing in any statute or treaty: it ended the right of American companies to press their lawsuits against Iran in United States courts. The decision is the leading answer to how far the President may go in settling the claims of American nationals against a foreign government, and the way the Court divided the problem into two separate questions is the part to carry onto an exam.

The dispute began when the American embassy in Tehran was seized on November 4, 1979, and diplomatic personnel were taken hostage. On November 14, President Carter, acting under the International Emergency Economic Powers Act (IEEPA), declared a national emergency and blocked the removal or transfer of Iranian government property subject to United States jurisdiction.²⁹ A Treasury regulation followed, declaring that any attachment, judgment, lien, or other judicial process against such property was null and void unless licensed or authorized, and warning that any license could be amended, modified, or revoked at any time.³⁰ Dames & Moore, an American engineering firm, sued Iran, its Atomic Energy Organization, and several Iranian banks on December 19, 1979, claiming more than $3.4 million for services under a contract with the Atomic Energy Organization, and the district court issued orders attaching the defendants' property.³¹ Because the attachments came after the freeze, they depended on a license the government had reserved the right to withdraw, and that detail matters when the Court reaches the nullification.

The crisis ended through an executive agreement with Iran, commonly called the Algiers Accords. Iran released the hostages on January 20, 1981, under an agreement signed the day before, whose stated purpose was to terminate all litigation between each government and the nationals of the other and to settle the claims through binding arbitration before a new body, the Iran-United States Claims Tribunal.³² A claims tribunal is an international arbitral panel that hears claims in place of national courts. To carry out the agreement, President Carter issued orders on January 19 that revoked licenses for Iranian assets, nullified non-Iranian interests acquired after the freeze, and directed banks holding Iranian assets to transfer them to the Federal Reserve Bank of New York.³³ On February 24, President Reagan ratified those orders and suspended all claims that could be presented to the Tribunal, providing that they would have no legal effect in any pending American court action.³⁴ The suspension was conditional: it ended for any claim the Tribunal found outside its jurisdiction, and a claim was discharged only when the Tribunal paid an award or found that nothing was owed.³⁵

By then Dames & Moore had won summary judgment against Iran and the Atomic Energy Organization for the contract amount plus interest, and it was trying to collect through garnishment and execution in Washington state when the district court stayed execution.³⁶ The firm then sued the United States and the Secretary of the Treasury, arguing that the orders exceeded the President's statutory and constitutional powers.³⁷ The Supreme Court heard the case on an expedited schedule, before the court of appeals had ruled, because the Solicitor General warned that unless the government acted by July 19, 1981, Iran could consider the United States in breach of the agreement.³⁸

Two different executive actions were under attack, and the Court took each on its own footing. One was the nullification of the attachments and the transfer of the frozen assets out of the country. The other was the suspension of the pending claims. The division matters because the two actions sit at different points on the spectrum from Jackson's framework: the first had a statute behind it, while the second had to be justified without any statute that spoke to it.

The nullification was the easier half. IEEPA, the statute under which the President had declared the emergency and frozen the assets, supplied the authority to nullify attachments and to order the assets transferred out of the country, and the Court said so directly.³⁹ It added that this power does not depend on the President providing claimants a substitute forum, because it rests on the statute itself.⁴⁰ The licensing structure makes the result easy to follow. Congress had given the President power over blocked foreign property, the government had reserved the right to revoke any license, and an attachment obtained under such a license was always exposed to revocation. An act backed by a statute stands at the strong end of Jackson's scale, so a challenger had little room to argue against it.

The suspension of claims was harder, because the Court declined to conclude that IEEPA or the Hostage Act directly authorizes it.⁴¹ With no statute on point, the question became whether the President was acting alone or at least with the acceptance of Congress. The Court reasoned in steps. Congress cannot anticipate every action a President may need to take in a fast-moving dispute, so its failure to delegate a specific power does not imply disapproval, especially in foreign policy and national security. Legislation closely related to the question, if it shows an intent to give the President broad discretion, may invite measures taken on independent presidential responsibility. That inference holds most strongly where nothing in the legislation points the other way and where a history of congressional acquiescence exists.⁴²

The history was the heart of the case. Claims by the nationals of one country against the government of another are a standing source of friction between the two states, so nations often resolve them by agreement, and the United States has settled such claims sometimes by treaty and, by longstanding practice, by executive agreement without the Senate's consent. In those agreements Presidents have renounced or extinguished claims in return for lump-sum payments or arbitration procedures, and since 1952 the President had entered at least ten binding settlements, including one of $80 million with the People's Republic of China.⁴³ Congress had responded to that practice with legislation. The International Claims Settlement Act of 1949 created a commission to distribute the funds from future settlements, and by building a procedure for implementing settlements it had not yet seen, Congress placed its stamp of approval on them. The Court called that implicit approval crucial to its decision.⁴⁴ Congress later amended the Act repeatedly to handle particular settlements, and the legislative history of IEEPA itself said that nothing in the statute was meant to impede the settlement of claims of American citizens against foreign countries.⁴⁵ The Court also drew on United States v. Pink, which treated the removal of obstacles to full recognition of a foreign government, such as the settlement of claims, as a modest implied power of the President.⁴⁶

Dames & Moore offered two arguments that Congress had withdrawn this power, and the Court's answers show what real congressional opposition would look like. The first was that restrictive sovereign immunity, adopted in 1952, removed the need for executive claims settlement, but at least ten settlements by executive agreement had followed that change, so acquiescence continued.⁴⁷ The second was that the Foreign Sovereign Immunities Act of 1976 took commercial suits against foreign states out of the executive's hands. The Court read that Act as removing one barrier to suit, sovereign immunity, and it noted that the Congress that passed it had considered and rejected proposals to limit executive agreements and had enacted IEEPA a year later with a statement protecting claims settlement.⁴⁸ A statute that merely touches the subject therefore does not amount to disapproval. Opposition has to be aimed at the power itself.

Taken together, the character of the legislation and the history of acceptance led the Court to conclude that the President was authorized to suspend the pending claims.⁴⁹ The Court was explicit that past practice does not by itself create power. What a long-continued practice does, when Congress knows of it and accepts it, is raise a presumption that the President acted with Congress's consent.⁵⁰ Two further facts buttressed the result. The suspension gave claimants an alternative forum, the Claims Tribunal, so they received something in return and the arrangement resembled the traditional settlements of the past.⁵¹ And Congress had not disapproved: it held hearings on the agreement but enacted no legislation and passed no resolution against it, and a Senate committee had called the Tribunal of vital importance to the United States.⁵²

The reason the Court leaned so heavily on Congress becomes clear when this agreement is compared with those in the preceding section. The agreement in Pink accompanied the recognition of the Soviet government, so it drew on the President's own recognition power and needed no delegation from Congress. No similar Article II power could be invoked for the agreement with Iran, and it did not implement a statute or a treaty, so congressional approval carried the weight.⁵³ The stakes for private parties explain why that approval mattered, since the accord displaced contract, tort, and other claims that companies held against Iran in American courts, a severe domestic effect for an agreement the Senate never approved.⁵⁴ The reasoning has critics. Treating long-running legislation and silence as approval sits uneasily with Youngstown, where Congress's rejection of a seizure power counted against the President, and the critics argue that the Court should have required more specific legislative approval for such far-reaching measures.⁵⁵

The Court closed by stressing how narrow its holding is. It did not decide that the President has plenary power to settle claims, even against foreign governments, because the sheer magnitude of such a power, set against the diversity and complexity of modern international trade, counsels against reading presidential authority more broadly than necessary.⁵⁶ The holding rests on two conditions stated together: the settlement was a necessary incident to resolving a major foreign policy dispute between the United States and another country, and Congress had acquiesced in the President's action.⁵⁷ The Court's own framing supports that reading, since it said it was resting on the narrowest possible ground and laid down no general guidelines for situations not before it.⁵⁸ The doctrine to hold is therefore a conditional one: the President, with at least the implicit approval of Congress, may settle claims of American nationals against foreign governments by executive agreement, including by suspending those claims in American courts.

Upholding the suspension did not close the courthouse to claimants entirely. Dames & Moore argued that suspending its claim took its property without just compensation, but both sides agreed the question was not ripe, because whether and how far the claims would go unpaid was still speculative.⁵⁹ The Court decided instead that the claimant would have a remedy in a suit against the United States in the Court of Claims under the Tucker Act, and it accepted the government's concession that the Court of Claims' treaty exception posed no jurisdictional obstacle.⁶⁰ The remedy matters because a taking must come with reasonable, certain, and adequate provision for obtaining compensation, so a company underpaid by the Tribunal can pursue compensation from the United States if the suspension amounted to a taking.⁶¹ On the attachments, the majority concluded that nullification did not itself effect a compensable taking, because the President's statutory power to prevent or condition attachments meant claimants held no protected interest in them.⁶² Justice Powell agreed with the suspension holding but would have left the takings claims, for both the attachments and the suspension, open for case-by-case resolution, and he wrote that the government must pay just compensation when it uses claims lawfully held by a few persons as bargaining chips in foreign policy.⁶³ The later discussion of Medellin v. Texas returns to the contrast between a President who builds on a long record of congressional acceptance and one who acts against Congress's implied will.

Legal Force of Executive Agreements

Once an executive agreement is validly made, the remaining question is how much existing law it can displace. The answer depends on what the agreement collides with. Against state law the rule is settled and favors the agreement. Against earlier federal statutes and treaties, part of the answer is settled and part has never been resolved by the Supreme Court. This section marks the line between those situations, because a student who treats the contested view as settled has misstated the law.

State law comes first. The Supremacy Clause makes the Constitution, federal laws made in pursuance of it, and treaties the supreme law of the land, and it does not mention other international agreements.⁶⁵ Even so, valid executive agreements preempt conflicting state law, because the preemption doctrine treats valid federal law of every kind, including treaties and executive agreements, as displacing inconsistent state law.⁶⁶

The foundation cases are United States v. Belmont and United States v. Pink, which both arose from the Litvinov Assignment. When President Roosevelt recognized the Soviet Union in 1933, an exchange of diplomatic correspondence with Maxim Litvinov assigned to the United States all Soviet claims against Americans who held funds of Russian companies seized after the Revolution. In Belmont the United States sued to recover funds that a Russian corporation had deposited with a New York banker, and the lower courts dismissed the suit because enforcing the Soviet confiscation would offend New York's public policy. The Supreme Court sustained the agreement and held that it took precedence over the conflicting state policy, because recognition, the establishment of diplomatic relations, and the assignment were parts of one transaction within the President's competence, and the Senate's participation was not required.⁶⁷ Pink later confirmed that compacts of this kind have a dignity similar to treaties under the Supremacy Clause.⁶⁸

The reason for the rule lies in the structure of foreign relations. The Constitution gives the national government complete power over dealings with other nations and gives the states none, so a state has no authority of its own to chart a competing course. When the President commits the nation to a foreign government, the promise comes from the nation as a whole, and a foreign government that had to wait and see whether each state would honor it could not rely on it. A state rule that undercut the agreement would leave the President unable to deliver what was promised. Belmont reasoned along these lines when it found it of no consequence that Article VI mentions only treaties, since complete power over international affairs lies in the national government.⁶⁹

The omission from the Supremacy Clause has a plausible explanation that also marks a limit. The Framers may have expected agreements short of treaties to stay within diplomatic and overseas matters, with anything of significant domestic impact taking treaty form. If so, an agreement that conflicts with state law stands on firmest ground when it can be defended as an exercise of the President's own Article II powers or as the product of authority Congress delegated, which fits the refusal in Dames & Moore to recognize a blanket power to use executive agreements in place of treaties.⁷⁰ Critics press the textual point harder. They argue that only treaties are given the force of domestic law, and that Belmont and Pink are better read as cases in which the President's independent power to recognize the Soviet Union, not the agreement as such, displaced the state rule.⁷¹ Either reading yields the same outcome on the facts of those cases.

In American Insurance Association v. Garamendi the Court applied the rule to an agreement that said nothing about displacing state law. After German courts lifted a postwar moratorium on Holocaust-era claims, class actions were filed in American courts against companies that had done business in Germany during the Nazi era, and the German government and German insurers protested. In 2001 President Clinton signed an executive agreement under which Germany would create a foundation, funded by the German government and German companies, to compensate people who suffered at the hands of German companies, including holders of insurance claims. The agreement stopped short of barring suits, but the United States promised to tell any court that American policy interests favored dismissal on any valid ground and to use its best efforts to have state and local governments treat the foundation as the exclusive mechanism.⁷² California had meanwhile enacted a statute requiring every insurer doing business in the state to disclose details of European policies in force between 1920 and 1945, on pain of losing its license and facing prosecution for false statements.⁷³

The Court held the California statute preempted even though the agreement contained no preemption clause, finding a sufficiently clear conflict in the intent of the agreements and in statements by high-level executive officials, because the state law compromised "the very capacity of the President to speak for the Nation with one voice in dealing with other governments."⁷⁴ The conflict lay in method, since the state and federal schemes shared the goal of compensating Holocaust victims and nothing in the state law made it physically impossible to honor the agreement.⁷⁵ The President had deliberately chosen voluntary cooperation with European governments and companies to preserve amicable relations, and California's disclosure regime, backed by license suspension, undercut that choice. In the Court's phrase, California sought "to use an iron fist where the President has consistently chosen kid gloves."⁷⁶ Two further facts pointed the same way: the subject was not a traditional state matter, and the conduct that gave rise to the state law did not occur within the state.⁷⁷ The agreement also fell squarely within the President's Article II power to conduct foreign relations.⁷⁸ The decision extended the foreign affairs preemptive power beyond treaties to executive branch acts that are not treaties.⁷⁹

The first version shows why touching foreign affairs is not enough by itself. A state law that merely brushes against foreign affairs does not automatically fall, although that connection is hardly irrelevant.⁸⁰ Preemption turns on conflict with the policy the agreement embodies, so the question is whether the state rule interferes with what the President chose to do.

Implied preemption is the part that drew resistance. Four Justices dissented in Garamendi, uneasy that the President can displace state law through an executive agreement when the preemptive effect is only inferred by courts. They would limit preemption in this setting to cases where the agreement's express terms call for it, because otherwise federal judges become "the expositors of the Nation's foreign policy" in situations where the President has not taken a clear stand.⁸¹ That position was the dissent's, and the majority's approach remains the governing one: a valid agreement preempts when it expressly says so, and also when a state law clearly interferes with the foreign policy the agreement embodies.

Against earlier federal law the picture changes, and the first task is to separate the settled part from the open part. For an agreement that Congress authorized, whether by prior delegation or by later implementing statute, or that a valid treaty authorized, the later-in-time rule that governs conflicts between statutes and treaties also governs conflicts with earlier statutes and treaties, so a congressionally authorized agreement adopted after a treaty overrides the treaty's conflicting terms.⁸³ The reason follows from the source of the agreement's authority. Such an agreement borrows the force of the process that authorized it, the instruments that those processes produce stand on equal rank, and the most recent expression of the national will therefore controls once courts have tried and failed to read the two so that both can stand.

The harder case is the agreement that rests on the President's authority alone. When such an agreement is self-executing it clearly preempts contrary state law, but its power over preexisting federal law is disputed, and the Supreme Court has never addressed the dispute.⁸⁴ One view holds that a purely presidential agreement cannot alter preexisting federal law without congressional implementation, a position associated with the Fourth Circuit's decision in United States v. Guy W. Capps, Inc., affirmed on other grounds.⁸⁵ The reasoning draws on how each instrument is made. A statute carries the approval of both Houses and the President, and a treaty carries the consent of two-thirds of the senators present, so an agreement made by one official has the thinnest pedigree and the weakest claim to displace either. In the terms of Jackson's framework from earlier in this chapter, an agreement that contradicts an existing statute is a third-category act, so the President prevails only if the power at stake belongs to the executive exclusively.

The competing view reasons from the President's independent powers. If the President concludes an agreement under an independent executive power, such as the power to recognize foreign nations or the commander-in-chief power, the agreement is valid federal law entitled to the same dignity as a statute or treaty, so the only question is whether it falls within the executive power, and if it does, the later-in-time rule applies to it as to any other valid federal law.⁸⁶ The strongest form of this view says that when the President exercises an exclusive inherent power, the agreement trumps earlier inconsistent congressional enactments, and it points to Zivotofsky v. Kerry, where the President prevailed against an Act of Congress because the recognition power belongs to the President alone.⁸⁷ The disagreement is therefore over whether the President's own power, standing alone, is enough to give an agreement the rank of a statute. Some read Medellin, which refused to let a presidential memorandum give domestic force to a non-self-executing treaty, as a hint that the Court may narrow the conventional conception of sole-executive agreements, though the Court has not said so.⁸⁸

The two hypothetical versions mark where the contest actually lies. A stem that pits an agreement against a state statute calls for the settled preemption rule. A stem that pits a purely presidential agreement against an earlier federal statute calls for both views and the observation that the Court has not chosen between them, since an answer that announces either view as the law overstates what the cases decide.