HILJ Symposium: Consent to the Use of Force, Jus Cogens, and Manifest Violations of Domestic Law

This post is part of the Harvard International Law Journal Volume 54(1) symposium. Other posts from this series can be found in the related posts below.

In my previous response to Ashley Deeks’ article, “Consent to the Use of Force and the Supremacy of International Law,” I examined some of the practical, doctrinal, and systemic implications associated with Deeks’ challenge to international law supremacy. As I noted there, I do think the problem of unreconciled consent requires attention, if not a solution, in the use of force context. I would prefer that solution to come from domestic law. Nonetheless, to the extent international law is asked to fix this problem, I’d like to explore the context in which it would have to do so, and suggest an alternative solution to the problem that avoids giving domestic law supremacy over host State consent.

Deeks suggests her duty to inquire (and the invalidity of any subsequent unreconciled consensual agreements) could arise via state practice or a modification to VCLT Article 46. I think both paths are problematic if State consent takes a treaty form (in contrast, if it’s a political commitment, I think a total or partial override of that commitment in favor of domestic law is much easier). In the treaty context, State practice favoring a duty to inquire runs up against VCLT Article 42(1):

1. The validity of a treaty or of the consent of a State to be bound by a treaty may be impeached only through the application of the present Convention.

This isn’t to say customary international law cannot override States’ treaty obligations under the VCLT (or the VCLT’s status as customary international law) but it’s not as simple an analysis as if States were creating a duty to inquire on a clean slate. The VCLT purports to be an “exclusive” list of grounds for invalidating State consent, which cuts against finding new or additional grounds for invalidity even in the use of force context.

What about amending the VCLT itself? Obviously, that’s possible, but I don’t think Deeks’ proposal aligns well with the VCLT’s existing criteria for invalidity. With one exception – discussed below – the VCLT bases the validity of State consent entirely on whether that consent was informed and freely given, not how that consent aligns with substantive rules or obligations. As Jan Klabbers puts it in his chapter in The Oxford Guide to Treaties:

In essence, most of these (Articles 46 through 52) deal with defects in the consent of a State to be bound, underlining the point made above that consent is practically speaking the most relevant general validity requirement when it comes to treaties. Such defects can have three broad sources: they can be based on improper procedure or authorization; they can be the result of something misleading; or they can be the result of coercion. In all cases, the underlying rationale is that the consent to be bound has been affected: but for the affecting circumstance, the State concerned would not (or would most likely not) have expressed its consent to be bound. Treaties are typically conceptualized as the result of a consensus ad idem; if so, a premium is placed on the reality of this consensus, and if the consensus ad idem is affected by fraud, coercion, or misrepresentation, it follows that there is not really a consensus ad idem and thus, in a meaningful way, no treaty.

In its present form, therefore, the VCLT might take care of some cases of unreconciled consent, i.e., where Heads of State or Government consent to a use of force, rendition, or drone deployment when such consent was, in the words of Article 46, “expressed in violation of a provision of its internal law regarding competence to conclude treaties” that was also “manifest and concerned a rule of its internal law of fundamental importance.”

The VCLT might also support invalidity where consent was given by someone (e.g., Italy’s military intelligence chief) who had no authority to commit the State. Note, however, claims of a lack of authority are further limited by VCLT Article 47:

If the authority of a representative to express the consent of a State to be bound by a particular treaty has been made subject to a specific restriction, his omission to observe that restriction may not be invoked as invalidating the consent expressed by him unless the restriction was notified to the other negotiating States prior to his expressing such consent.

But what the VCLT does not address — and what concerns Deeks — is consent in violation of substantive domestic law protections, not domestic procedural provisions on treaty-making. And when it comes to substantive constraints on State consent to international agreements, the VCLT has only one substantive constraint — jus cogens. VCLT Article 53 prohibits consent inconsistent with these “pre-emptory norms” of international law. Moreover, it does so in much the same way as Deeks envisions her duty to inquire would operate. For example, States cannot rely on, or excuse, acts in violation of jus cogens norms (such as prohibitions on slavery, genocide and, torture) even where they consented to such acts.

Given such similarities, to what extent could jus cogens solve the problem of unreconciled consent if not completely, then at least in the most egregious cases? For example, consent to a State rendering an individual for torture could be invalidated under Article 53 without having to ask whether or not the acts violated the host State’s domestic law. Moreover, to the extent, use of force rules are themselves jus cogens, that might provide a path to clarify and regulate the scope of host State consent to use of force as well. This may not resolve all cases of unreconciled consent – I don’t think it responds to Mexican consent to U.S. drone deployment for fighting drug traffickers. But it does seem like there are already various international law doctrines that could be brought to bear on the problems of unreconciled consent and human rights violations without necessarily turning to domestic law for answers.

Unfortunately, I suspect that neither my jus cogens suggestion nor Deeks’ proposal to adopt a duty to inquire will draw much support from State practice. For all the attention treaty law doctrine gives to establishing the boundaries for valid treaties, it’s remarkable how rarely those boundaries are actually invoked. Jus cogens is widely discussed in academic circles, but I am unaware of a single instance in which a State’s consent to an agreement was invalidated on the grounds that its subject-matter breached jus cogens. There have been more claims of invalidity for failure to comply with internal treaty-making procedures (see, e.g., Cameroon v. Nigeria: Equatorial Guinea intervening), but Courts have tended to dismiss them, leaving the bar for a VCLT Article 46 claim high, if not beyond reach. The one exception lies outside the treaty context in the Kadi case. When the ECJ considered an actual treaty in France v Commission it recognized that, even if the European Commission’s conclusion was ultra vires as a matter of European law, international law still required EC compliance. Thus, whatever normative appeal might attach to improving the implementation of the VCLT’s existing validity criteria or adopting new ones altogether, States and international tribunals appear to have little enthusiasm for such a project. The predictability and certainty associated with consent and pacta sunt servanda remain, for now, nearly inviolable.