I Ratification

Treaty law long accorded conclusive significance to treaty ratification by executives as an incident of the sovereign power of monarchs – passed along, naturally, to those with full powers to act on the sovereign’s behalf.Footnote ⁹ This was challenged by two developments. First, beginning by the mid-nineteenth century, a practice emerged in which states adopted constitutional provisions that required legislative approval of treaties. That tendency continues largely unabated today; the subjects encompassed by such provisions have often been expanded, along with the widening scope of treaty-making generally, and there has even been an uptick in supermajority requirements.Footnote ¹⁰ Second, agreements became more varied and often more informal, resulting in many that arguably did not require approval by traditional, treaty-oriented mechanisms.Footnote ¹¹ The first development meant that ratification was transformed from a formal and infrequent act by which sovereigns confirmed, as a matter of obligation, that their representatives had been authorized to reach agreement, into a legislative referendum on the underlying merits of the agreement. But the second development meant that ratification might sometimes be dispensed with altogether, and agreements concluded based on signature alone.Footnote ¹²

The significance for the law of treaties was hotly contested, including as to whether the violation of a state’s constitutional requirement of legislative approval affected whether a nonconforming treaty nonetheless bound the state under international law. Most agreed that domestic law regulated competence to consent to treaties and (often distinctly) governed competence to notify that consent to other states. But the consequences when consent had been notified by a state official competent to that role – notwithstanding the absence of valid consent under domestic law – were unsettled, and occasioned a diverse range of views that flourished from the interwar period through the drafting of the VCLT.Footnote ¹³

At one end of the spectrum, some suggested that international law treated notification by a competent notifying official as conclusive, either as a matter of dogma or based on a state’s responsibility to make good the acts of its officials. At the other extreme, some regarded constitutional law as being conclusive as to both the domestic and international effectiveness of consent. In the middle were those who regarded constitutional deficiencies as potentially bearing on international efficacy, but varying as to when.Footnote ¹⁴ The Harvard Research study, after extensive canvassing of publicists, state practice and jurisprudence, decided – almost arbitrarily – to recommend codifying that ‘[a] State is not bound by a treaty made on its behalf by an organ or authority not competent under its law to conclude the treaty’, but that a nonbound state might then be responsible for the reasonable reliance of others on its prior representations.Footnote ¹⁵ Under this approach, international law deferred to foreign relations law, but ameliorated any adverse effect on other states through a liability rule.

The matter was revisited by the International Law Commission (ILC) in drafting the VCLT. Initially, the ILC’s view was similar to that of Harvard Research: for a treaty to be binding on a state, it had to be adopted ‘in accordance with its constitutional law and practice through an organ competent for that purpose’.Footnote ¹⁶ Then it reversed course. The ILC’s next proposal stressed state autonomy and self-help: states could always, if they chose, negotiate ratification terms that accommodated domestic processes, seek collateral arrangements, caveat their signatures, qualify full powers or improve disclosures to other states during negotiations.Footnote ¹⁷ If, despite these options, states behaved internationally in ways that were incompatible with their domestic constitutions, the proposal would be less forgiving. The draft precluded a state from invalidating its consent by invoking the violation of ‘a provision of its internal law regarding competence to conclude treaties’ unless that violation was ‘manifest’.Footnote ¹⁸ This was later enhanced, in Article 46 of the VCLT, to define what ‘manifest’ meant – signifying a violation that would be ‘objectively evident’ to any state acting normally and in good faith – and to require that the violation also concern ‘a rule of its internal law of fundamental importance’.Footnote ¹⁹

The ILC took commentary and practice to lean, slightly, in favor of its resolution, and noted support from governments. But its conclusion also had a normative underpinning. In its view, recent treaty-making procedures had ‘done all that can be reasonably demanded … in the way of taking account of each other’s constitutional requirements’. Moreover, in most cases, states invoking the failure to abide by constitutional requirements had other motives for attempting to escape their obligations, and indulging them would threaten the security of treaty obligations.Footnote ²⁰

The VCLT thus wound up contributing to an important divide. Treaty law ceded to each state ‘the determination of the organs and procedures by which its will to conclude treaties is formed’, concerning itself ‘exclusively with the external manifestations of this will on the international plane’ – meaning that the international-law determination of whether a state’s agent was competent to commit the state largely superseded, for international purposes, any other constitutional delicts. In principle, as the ILC explained, each government ‘had the necessary means of controlling the acts of its representative and of giving effect to any constitutional requirements’, and perhaps making any failings ‘the clear responsibility of the Government of the State concerned’ was more feasible. Any other approach, it suggested, ‘would certainly be regarded as an inadmissible interference’ in another state’s internal affairs, particularly those that followed more dualistic approaches to the relationship between international and domestic law.Footnote ²¹

This was not the only possible international rule, even by the ILC’s lights. And every approach the ILC considered, not just the one it ultimately adopted, allowed states to work around the default by developing treaty practices that respected both constitutional requirements and the conclusion of treaties.Footnote ²² Unsurprisingly, though, the default it adopted was urged by governments: in practical terms, the executive representatives of governments, who were also the rule’s beneficiaries. These representatives would retain primary authority over any workarounds that treaty negotiations might yield. More basically, the VCLT rule established an international-law constraint on foreign relations law. A state’s representatives, rather than the state as a whole, would be entrusted (absent extraordinary circumstances) with exclusive responsibility for determining whether a state had complied with its constitutional rules, irrespective of whether those rules vested them with such responsibility or purposefully divided domestic treaty-making authority. Even in the event of manifest violations, it would ordinarily fall to those same representatives to invoke any transgression – one made through themselves or other executive-branch agents – unless the domestic legislature or some third party developed standing to do so.

There was never any illusion that this would permit the vindication of constitutional norms, even those expressed by laws regarding competence to enter into treaties. There had already been well-known instances in which Luxembourg and Argentina had acceded to the League of Nations despite violations of provisions in their constitutions requiring parliamentary approval – though there was room for disagreement as to whether these violations had been ignored as merely domestic matters or whether tacit parliamentary approvals might be surmised from their funding of League participation.Footnote ²³ The ILC itself emphasized the exceptional nature of manifest violations, and even that other states had never acquiesced in attempts to invoke such violations.Footnote ²⁴

Subsequent assessments, and subsequent practice, only vindicate its assessment. The governments of major powers have rebuffed criticisms that they ignored the obligation to secure legislative assent, without any apparent effect on their international obligations.Footnote ²⁵ Most prominently, the International Court of Justice rejected Nigeria’s contention that a bilateral declaration with Cameroon was nonbinding because the Nigerian constitution at the time required approval of its Supreme Military Council; while the provision was of ‘fundamental importance’ under Article 46, it was not sufficiently ‘manifest’, including in part because heads of state could be presumed competent.Footnote ²⁶ A later judgment was to much the same effect, and made clear that – even though the Somali Parliament had rejected an agreement, making relatively prominent its potential authority under domestic law – that mattered little when the Prime Minister of Somalia failed to question its validity in later interactions on the international plane.Footnote ²⁷

It is possible that the rise in legislative authority at the national level might shape international expectations and make deviations sufficiently apparent to register as manifest, though the trend has yet to have any such effect; perhaps the nascent project on comparative foreign relations law will even assist by illuminating pertinent commonalities and differences.Footnote ²⁸ It is unlikely, in any case, that the basic rule will be reconsidered. If anything, parallel developments suggest similar biases toward the creation of international obligations and against means by which they might be subverted. The unilateral acts doctrine, per the ILC’s Guiding Principles of 2006, likewise indicates that while ‘[a] unilateral declaration binds the State internationally only if it is made by an authority vested with the power to do so’, necessarily ‘[b]y virtue of their functions, heads of State, heads of Government and ministers for foreign affairs are competent to formulate such declarations’.Footnote ²⁹

The emerging unilateral acts doctrine seems to be of a piece with the full powers doctrine for treaties and, more broadly, the executive capacity assumed for traditional sources of international law, insofar as it also encourages the stability of international commitments.Footnote ³⁰ Still, this assignment of responsibility – like treaties, to a limited class of officials, without regard to what a state’s foreign relations law provides – lacks the same compelling need to find an agreeable means of concluding a principal source of international law, and lacks any explicit exception for manifest violations. Interestingly, the Guiding Principles also inhibit the ‘arbitrary’ revocation of such commitments, with potential parallels for treaties.Footnote ³¹ Some governments were discontent with this exercise, but it seems to have been due primarily to the difficulty of establishing general principles – and, perhaps, the risk that they would be regarded as binding their states unintentionally – and not the risk of making international law by means that were inconsistent with domestic law.Footnote ³² The basic calculation is unsurprising. A sufficiently deliberate capacity would appeal to the political actors that may make use of it, executive officials, who might have less regard to the loss in authority suffered by others. If consulted, national legislatures might be more inclined to regard the emerging doctrine as creating a new rival to treaty-making by participative means.

II Withdrawal

International law’s attitude toward the domestic law regarding withdrawal from treaties is less concrete – but as a matter of inference, at least as indifferent. The VCLT allows a state to withdraw by consent or in conformity with a treaty’s terms, or if the parties so intended or treaty’s nature so implies.Footnote ³³ Additional provisions address notice and subsequent withdrawal for cause.Footnote ³⁴

The ILC, in drafting the VCLT, noted that withdrawal serves as a safeguard when a state’s consent to a treaty violates its constitutional principles.Footnote ³⁵ Accordingly, ‘a defect in [the state’s] consent to be bound’ – presumably, such as would satisfy Article 46 – is one of the grounds that may be invoked and notified to other states parties.Footnote ³⁶ But nothing addresses whether there might be any ‘manifest’ or other constitutional violation related to the act of withdrawal itself – that is, addressing whether a state’s invocation of some basis for withdrawal (or comparable act) might itself be impeached on the basis of a constitutional violation.Footnote ³⁷ The VCLT incorporates notice periods, at least as defaults, presumably for the benefit of other states parties,Footnote ³⁸ and allows the notifying state to withdraw such notice or instrument before it takes effect – including, one expects, on the basis of a constitutional violation.Footnote ³⁹ But the practical significance is elusive. The timing in which a state can change its mind is a serious constraint.Footnote ⁴⁰ In any event, the decision appears confided to the discretion of executive-branch officials, exactly as one might expect.Footnote ⁴¹ The result, as Hannah Woolaver has observed, is that ‘in contrast to international law powers to join treaties, the authority of the executive to withdraw the state from treaties in Article 67 of the VCLT is, prima facie, absolute in international law, unlimited by any checks that may exist in domestic law’.Footnote ⁴²

This withdrawal regime now faces challenges reminiscent of those once confronted by ratification. States are slowly retreating from the practice by which the executive is entitled to withdraw unilaterally from treaties, instead adopting new provisions that require parliamentary approval.Footnote ⁴³ During the same period, multilateral and even bilateral treaty-making has seen a pronounced slowdown,Footnote ⁴⁴ augmented (at least in the short term) by withdrawals and similar attempts at disengagement.Footnote ⁴⁵ The result is that formal constitutional revision has been accompanied by high-profile instances in which executive-led attempts at withdrawal, like the United Kingdom’s initial attempts to withdraw from the European Union and South Africa’s (initial?) attempt to withdraw from the International Criminal Court, were slowed by judicial decisions that recognized for the first time parliamentary rights to participate in exit from certain types of treaties.Footnote ⁴⁶

Despite the evolution of foreign relations law, there is no evident movement to develop international law. As with ratification, international law effectively regards executive determinations, in notifying a state’s will, as exhausting treaty law’s deference to domestic law, leaving any potential legislative role to be enforced solely through a state’s own legal processes. This has clear appeal, both internationally and domestically, in instances where national requirements are as yet uncertain. But where national requirements are well defined, it remains troublingly clear that foreign relations law is not self-enforcing. For a number of actual or potential withdrawals – including those by the United States, a frequent flyer of sorts – justiciability principles are likely to limit the capacity of courts.Footnote ⁴⁷ Even where courts are willing and ostensibly able, any domestic principle will be difficult to apply prior to a unilateral executive notice taking effect as a matter of international law.

The readiest solution would be to infer an analogous exemption for manifest violations in the withdrawal context to match that for ratification.Footnote ⁴⁸ However, that is difficult as a matter of treaty interpretation. Article 46 is plainly limited to ‘provisions of internal law regarding competence to conclude treaties’; that heads of government and comparable officials are elsewhere given authority to perform ‘any other act with respect to a treaty’ offers no textual basis for concluding that one of those acts, withdrawal, must also be constrained in the event of manifest violations.Footnote ⁴⁹ Indications in the travaux préparatoires are mixed or negative.Footnote ⁵⁰ That state practice did not support any such principle as a matter of customary international law – as Woolaver, a strong but judicious advocate for this interpretation, acknowledgesFootnote ⁵¹ – further undermines any basis for attributing such intention to the states agreeing to the VCLT, and the lack of supportive state practice since suggests that it would be unreasonable to find any grounding in subsequent treaty practice.

To be sure, teleological or policy arguments for comparable treatment are appealing, including that the security of treaty obligations – one of the objectives of Article 46, including in limiting its exception to manifest violations only – is arguably hindered by the lack of any exception for constitutional violations regarding withdrawal, since that permits too-easy disengagement.Footnote ⁵² Necessarily, too, the backstops are different: while withdrawal serves as a safeguard for uncorrected constitutional violations concerning ratification, re-ratification is needed to cure withdrawal in violation of domestic law, and that tends to be more cumbersome as a matter of domestic and international procedure. Finally, it may be possible to muster additional arguments for cross-applying Article 46 and its manifest violation standard. For example, one might argue that states appearing to have withdrawn might revoke a prior notification on the ground that the relevant instrument did not ‘take effect’ under Article 68 if it manifestly violated a constitutional requirement of legislative participation.Footnote ⁵³ No matter the precise basis, however, any such exemption would be doctrinally tenuous and narrow in its potential application.Footnote ⁵⁴ It remains extremely unlikely that this would change so long as the manifest violation standard serves as a catchall for redressing other, potentially unknown, constitutional violations as well.

What else might be engineered? If treaty law is to be progressively developed – or, with great difficulty, interpreted on the basis of Article 46 or otherwise – states and treaty depositories might adopt the assumption that a withdrawal must satisfy the same separation-of-powers procedures by which the state approved consent, as an international-law embodiment of the acte contraire doctrine. That is quite vulnerable as a general principle, given the actual discrepancy in many states between domestic laws regarding legislative approvals of ratification and withdrawal.Footnote ⁵⁵ But the severity of any such rule may also be tempered. A state might overcome any such assumption by explaining its contrary (or still indeterminate) domestic principle. And the consequence – if a discrepancy was observed and remained unexplained, but withdrawal could still be approved in reasonable time – might not be to render the initial attempt invalid, thereby sustaining the full-fledged obligation, but instead to establish in the interim an obligation simply to abide by the treaty’s object and purpose.Footnote ⁵⁶ Still less dramatically, governments declaring withdrawal might be required by depositories or other states simply to disclose their authority and its compatibility with constitutional requirements – which might at least have the effect of making sufficiently ‘manifest’ any violations of a requirement to secure legislative approval.

Other solutions involve bearing possible constitutional restraints in mind while negotiating withdrawal clauses in particular treaties. The easiest lift is to restrict any immediate effect of a notice or instrument of withdrawal – for example, by restricting when withdrawal may first be effectuated, as is typical for International Labor Organization agreements, and evidenced in the Berne Convention and the Paris AgreementFootnote ⁵⁷ – so as to permit legislative or judicial intervention in the case of a potentially unlawful, unilateral executive measure. More particularly, a treaty’s withdrawal clause might require the disclosure of authority mentioned above, following the precedent of clauses that obligate withdrawing states to accompany their notices with explanations.Footnote ⁵⁸

III Conclusion

Foreign relations lawyers looking beyond their jurisdictions will naturally focus on any directly comparable national systems, but they should also consider the impact of international law on their discipline. First, international law not only provides opportunities for domestic actors that may be in tension with domestic principles, but also establishes rules that constrain foreign relations law. Second, while those rules of international law intend to respect ratification developments, they also guarantee that (in most circumstances) only constitutional rules assigning authority to the executive to communicate state decisions regarding ratification or withdrawal will be respected. As a result, these international rules have not adequately taken into account developments in legislative authority. Third, none of this is inevitable. In developing a comparative approach to foreign relations law, experts should not overlook the degree to which the effectiveness of domestic norms require attention to internationally established conditions.