Do Congressional-Executive Agreements Establish More Reliable Commitments Than Treaties?

As we explained in our first post, we agree with the broad “interchangeability” thesis as a matter of constitutional law, and we are, in general, quite sympathetic to Oona Hathaway’s comprehensive case on behalf of a presumption in favor of the use of congressional-executive agreements (CEAs) rather than treaties . . . principally because of the more robust democratic nature of such statutes. Professor Hathaway, however, argues further that CEAs are preferable to traditional treaties because they “create more reliable international commitments.” Although we think this is occasionally true–in circumstances we described in our prior post–we remain doubtful that the “more reliable commitments” argument is sound as an argument for a more comprehensive abandonment of treaties in favor of CEAs.

Oona’s case for the greater reliability of CEAs rests principally on two claims: (i) that commitments made through CEAs will be easier to enforce (presumably against the Executive branch) in domestic courts; and (ii) that it will be more difficult for the President to withdraw the United States from CEAs. In our view, both prongs of this argument are doubtful.

1. Professor Hathaway first claims (see page 181 of her draft) that “[i]nternational law truly binds only when there is a way to enforce a state’s obligation under international law in domestic courts,” and that CEAs are preferable in this respect because they “allow[] for a one-stage rather than multi-stage process to create [this] enforceable legal commitment.”

For starters, we question the premise of this argument (and we suspect Oona does not really mean to embrace such an unqualified proposition). Even where there is no means of enforcing a treaty or CEA in court, that agreement obviously does, at least in a formal legal sense, “truly bind” the United States: Not only does it impose an international law obligation, but the Supremacy Clause (and, for the President, the Take Care Clause) requires the political branches of the United States (and, where relevant, the States) to comply with the treaty or statutory agreement as a matter of domestic constitutional law. Take, for example, Common Article 3 of the Geneva Conventions. Congress’s recent diminution of judicial review over claims involving Geneva violations, and its dilution of the War Crimes Act (in sections 5 and 6 of the Military Commissions Act, respectively), have significantly (although perhaps not entirely) foreclosed judicial enforcement of Common Article 3 against U.S. government actors. Nevertheless, the United States, and the President, remain obligated, under both international and domestic law, not to engage in cruel treatment and torture with respect to detainees, by virtue of the prohibition in Common Article 3—and notably, no one in either the legislature or even the Bush Administration has suggested otherwise.

Perhaps, however, Oona has in mind a less formal concept of “bindingness.” She implies that the absence of judicial enforcement renders our legal commitments, as a practical matter, of little or no worth in the eyes of other nations. But that would be an overstatement. Indeed, state parties are often quite skeptical of the fairness with which domestic courts resolve international law disputes. In the United States, for example, skepticism about foreign court interpretations of international law trace back to the period of the Napoleonic Wars and the decisions of British Admiralty courts affecting U.S. neutral shipping. The Sabbatino decision rested on just this concern, applying it, self-consciously, even to U.S. courts. In any case, international law has flourished in many countries even without any form of judicial enforcement. Both here and abroad, domestic courts have rarely played a dominant role in enforcing international law obligations. (The sui generis case of the EU is an exception, of course, but an exception made possible, among other things, only because of the shadow cast by the ECJ’s compulsory jurisdiction.) In this nation, as in many others, whether as a result of various kinds of justiciability doctrines or because many international law disputes do not yield plaintiffs willing to bring their disputes to domestic courts, international law is most often enforced by the political branches without involvement by the judiciary. Therefore, although we agree that judicial enforcement is a matter of importance in some contexts, we think it is a mistake to suggest, as a general proposition, that the “truly bind[ing] character of international law” depends on its availability in all cases.

But happily conceding that judicial enforcement will sometimes be important in enhancing the reliability of U.S. treaty commitments, we nevertheless question Oona’s claim that it is easier to provide for such enforcement by using CEAs as opposed to treaties. The whole question of judicial enforcement of treaty obligations in the United States involves complexities and uncertainties that we will not try to resolve here. It is clear, however, that for many, if not most treaty commitments, the President and Senate—depending upon, inter alia, the treaty language, existing statutory provisions, and the context in which a treaty claim is raised (e.g., as the basis for a cause of action or as a defense in a habeas or criminal proceeding)—have constitutional authority, without the involvement of the House, to provide for judicial enforceability. And, indeed, throughout U.S. history courts have frequently enforced treaty commitments even in the absence of legislation.

In the relatively small number of instances where this is not the case, Oona’s claim is potentially more persuasive. As she observes, in the case of a CEA the necessary implementing legislation can be included in the same enactment by which Congress authorizes or approves the agreement itself—thereby necessitating only a one-step, rather than a two-step, process. Of course, that this can be does not mean that it will be done, and recent experience with, for example, human rights treaties, suggests that the problem is not one of institutional misalignment but instead of political will. Even had these agreements been submitted as CEAs, it is doubtful that Congress would have specifically provided for judicial enforcement—as, indeed, it has not, at least as to private enforcement, in the case of NAFTA and the WTO. Be that as it may, however, the important point is that the added burden of the two-step process only exists in some cases and, therefore, cannot be a strong ground for an argument for the presumptive superiority of the CEA over the treaty. The CEA is superior in this respect, at most, in those unusual cases where the President and Senate are constitutionally prevented from providing for judicial enforceability.

2. Professor Hathaway’s second and more central claim is that it is easier to prevent executive withdrawal from CEAs than from treaties. We are not persuaded, however, that there is a meaningful difference in the two models when it comes to restricting withdrawal. Oona concedes, at least arguendo, that the President has unilateral authority to withdraw the United States both from treaties and from congressional-executive agreements, in accord with the terms of the agreements themselves and background rules of international law. She likewise concedes (see note 284) that both the Senate in giving consent to a treaty, and Congress in adopting a joint resolution approving an international agreement, can provide that their consent/approval is conditioned on restricting the President’s withdrawal power, such as by requiring that withdrawal be effected only by congressional resolution presented to the President. So where is the difference?

Oona suggests that the difference might lie, not in how the President can be limited before a withdrawal, but instead in what happens after a withdrawal: she argues that whereas when the President terminates a self-executing treaty, that treaty’s status as domestic law under the Supremacy Clause immediately ceases, legislation implementing a congressional-executive agreement can, by contrast, survive the President’s termination of the agreement.

This theory is intriguing, but ultimately unconvincing. As an historical matter, Congress in fact has not generally enacted the substantive provisions of CEAs into domestic law. More often, its joint resolution merely authorizes the President to conclude a particular agreement, or approves an agreement he has already negotiated. Such authorization itself is no moment, however, once the underlying agreement is undone and, thus, it has virtually no effect in most cases of withdrawal.

More importantly, even legislation implementing a pact will ordinarily be dependent on the ongoing validity of the agreement under international law. In the absence of clear congressional language to the contrary, there is simply no reason to think that the legislation will outlast the international obligation, not because the President has repealed or violated the legislation (something he may not constitutionally do), but because the legislation no longer serves the purpose for which it was enacted: Once the international agreement is denounced, the implementing legislation will become defunct of its own accord. (Notably, this is equally true of legislation implementing a so-called non-self-executing treaty.) Does Oona believe, for example, that the provisions of the NAFTA implementing legislation dealing with dispute resolution panels (or, for that matter, any other portion of the implementing legislation) would survive a presidential termination of the agreement?

It is here that Oona makes her most striking suggestion—that CEA legislation might not only authorize an international agreement, but also commit the United States to abide by the terms of the agreement, as a matter of substantive domestic law, even where the agreement itself is not in effect: that the enactment might “specif[y] a course of action by the United States,” such as protecting particular human rights or observing particular environmental standards” (page 198 & n.291), and that this substantive obligation would not be dependent on the continuing existence of the international agreement.

This does not appear to be an argument in favor of using CEAs, as such—it is, instead, an argument for the enactment of substantive statutes that are equivalent to agreements the President has negotiated, only without the (potential) expiration date. Notably, Congress could enact such a substantive law regardless of whether the parallel agreement is a treaty subject to Senate ratification, or a CEA. But more to the point, we very much doubt that either of the political branches would ever consider such a practice in either case. If they were, however, so motivated, the far more sensible approach would be simply to require, as Oona concedes they may, that the President may terminate an agreement only with a supporting congressional resolution presented to the President.

Finally, even if Congress were to prescribe that its substantive legislation—whether for a treaty or a congressional-executive agreement—shall survive a presidential termination of the underlying agreement, foreign states are not apt to view the continuing validity of the legislation as evidence that the United States is a more reliable treaty partner. The whole point of making treaties, as opposed to engaging in parallel acts of legislation, is to create a binding international law obligation on both parties to enact and retain the relevant legislation. If the President formally withdraws from a treaty commitment, the former treaty partner will view the United States as unreliable, in Oona’s sense, regardless of whether Congress in the meantime continues ex gratia to keep in place legislation that is consistent with the (former) treaty obligations.

3. One final point, about the feasibility of Professor Hathaway’s proposal: Let’s say Oona were right and we are wrong–that CEAs can, in fact, more effectively constrain the Executive branch by (i) providing judicial oversight and enforcement of international obligations and/or (ii) by limiting the circumstances under which the President can withdraw the U.S. from agreements, either formally or in effect. The especially odd thing about such arguments is that Oona is addressing them to the Executive branch. Her article is a plea to the Executive branch that it should, as a matter of practice, choose to use the mechanism for concluding international agreements that, in her view, would most constrain the President’s discretion in future implementation of, and withdrawal from, such agreements. Why is there any reason to think that the Executive branch would embrace such a self-restricting proposal? To the extent the Executive branch does prefer CEAs, the more plausible hypothesis is that it does so because it finds it easier to obtain consent to agreements it has negotiated by going to Congress than by submitting them to the Senate where the minority veto still looms large.

Oona further concedes that the Senate’s cooperation would be necessary to the realization of her proposal. Yet, she seems to underestimate the degree of resistance that is likely to come from that quarter. There is much history, going back to the late 1960s—the National Commitments Resolution, various failed legislative proposals to deal with executive agreements, the Senate’s failure to give consent to the Vienna Treaty on the Law of Treaties, and so on—suggesting that the Senate and the House do not see eye to eye on this problem and that the Senate’s acquiescence in the CEA mechanism therefore has its limits as a practical political matter. However normatively attractive the idea of phasing out the treaty may be, Professor Hathaway has not yet fully explained how that is likely to come about.