The Oxford Guide to Treaties Symposium: The General Law of Treaties and Its Limits

by Christian Tams

[Christian J. Tams holds the Chair of International Law at the University of Glasgow - School of Law]

Let me start off by saying that to participate in the Oxford Guide to Treaties has been a real privilege: it is a great book that combines theoretical reflection and practical insights. I am particular impressed by the list of treaty clauses included as Part VI. Unlike many other areas of international law, treaty law is a framework to give effect to the expressed will of parties. Of course, there are limits to what parties can agree to – jus cogens being the most important –, but they hardly ever act as real constraints on treaty-making. So any analysis of treaty law has to come to terms with the diversity of treaty practice. And as patterns of practice change over time, the topic itself is not static, but in a state of constant reform. That makes it so interesting.

At the same time, there is a risk – and it is reflected in Duncan’s post, even though I would describe it slightly differently. In many respects, the general law of treaties as laid down in the VCLT cannot ‘tame’, or regulate, treaty practice. Sometimes, this is simply pragmatic and realistic: a general treaty on treaties can only offer residual rules, just as a general document on responsibility will yield to special regimes (as per Art 55 of the ILC’s Articles on State Responsibility). The VCLT implements the lex specialis principle not in one general clause, but within the specific provisions, many of which contain caveats: Art 28 eg provides for non-retroactivity ‘[u]nless a different intention appears from the treaty or is otherwise established’; Art 54 as the opening provision of Part V Section 3 provides (unsurprisingly) that ‘The termination of a treaty or the withdrawal of a party may take place: (a) in conformity with the provisions of the treaty’. The phrase ‘unless the treaty otherwise provides’ is used 14 times throughout the VCLT. The general law of treaties is residual, and unlike in the case of State responsibility, it is a lex generalis that is very often derogated by special rules.

In other areas, the VCLT provides not residual rules, but no rules at all; and as a result, solutions are sought under other regimes of international law. Duncan rightly mentions remedies, which the VCLT treats rather in passing and which are sought and addressed under the law of dispute settlement, the law of responsibility and/or special treaty regimes. Geir Ulfstein’s reference to treaty bodies to me would also seem to be a case in point: a major element of the multilateral treaty process that is not regulated in any detail by the VCLT’s general rules, but by international institutional law. But questions of responsibility for treaty breaches are probably the biggest ‘regulatory gap’: pursuant to Art 73, the VCLT ‘shall not prejudge any question that may arise … from the international responsibility of a State’. As a consequence, the huge field of treaty breaches, which domestic legal systems often regulate as part of the law of treaties/contracts, is to the regime of international responsibility.

The combined effect of the application of the lex specialis principle and of the conscious decisions of drafters not to cover treaty law issues in their entirety is that quite often, answers to questions of ‘treaty law’ will have to be sought outside the VCLT. This does not affect its status as the key legal instrument in the field. But it means that it has to be read and applied (and presumably taught) in conjunction with other legal texts addressing treaty law issues. And it means that more than with other general legal texts (say, the ILC’s texts on State responsibility or on diplomatic protection; or the 2004 Convention on Jurisdictional Immunities of States and Their Property), the primary treaty law rules in their infinite variety will often determine outcomes. And because of this, Duncan should be congratulated on his decision – explained to contributors right from the start of the project – to complement the legal analysis by a long and detailed part on ‘Treaty Clauses’.

The Oxford Guide to Treaties Symposium: Evidence of “Secondary” Fragmentation

by Duncan Hollis

First of all, I need to say thank you to all the contributors to the current symposium on my book, The Oxford Guide to Treaties.  It’s quite common in academic circles to have symposia on “affairs of the day” (and, to be clear, those affairs often trigger very important issues like targeted killing, cyberwar, climate change, the EU fiscal crisis, etc.).  But, I think it’s equally important to step back from time to time and have conversations about the international legal system itself, of which treaty law and practice now forms a large part.  Thus, I’ve greatly appreciated the discussions over the last several days on whether reservations can be severed, the rule(s) of treaty interpretation, the increasing “publicness” of treaty functions, and the role of non-State actors in modern treaty-making.

In reading these posts, moreover, I was struck by how some of them suggest (albeit implicitly) a new way to think about the fragmentation of international law.  To date, we have tended to think about fragmentation in one of two ways.  First, we have the question of “normative” fragmentation, where two rules produced by different legal regimes conflict or compete with each other, meaning that a State has to choose to which rule to give priority.  We normally talk about this as the “trade and . . .” problem where WTO rules have been said to conflict with rules of international environmental law, international labour law, etc.  But, normative fragmentation is not limited to the trade context as witnessed by the question of what to do when UN Security Council resolutions on international peace and security conflict with EU Law in the Kadi case.  Second, fragmentation may also arise where the conflict is not between the rules but who applies them; that is, competition or conflict over which tribunal or court should be authorized to have the final say on which rules apply or what a particular rule means in a given situation.  The MoX case is a paradigmatic example of this inter-tribunal competition, with three different proceedings under three different normative regimes: an arbitral tribunal pursuant to the 1982 U.N. Convention on the Law of the Sea, dispute settlement under the Convention on the Protection of the Marine Environment of the North-East Atlantic, and proceedings before the European Court of Justice pursuant to the European Community and EURATOM treaties.

As I think about the law of treaties, however, I see the potential for a third type of fragmentation within international law, or what I’ll call “secondary” fragmentation.  My basic idea is that fragmentation is not limited to competition and conflict among primary rules, but can also occur with secondary rules.  Here, I’m employing H.L.A. Hart’s famous division of rules into primary and secondary categories.  Primary rules are rules of conduct — telling States and other subjects of international law what they are obligated to do (or not do).  Secondary rules, in contrast, are “rules on rules”, or rules that tell us how to form, interpret, amend, or extinguish primary rules.  When we talk about normative fragmentation, however, almost all the existing discussion has emphasized conflicts and competition among primary rules, e.g., should a WTO rule trump or defer to an international environmental rule?

In looking at the various posts on interpretation and Geir Ulfstein’s post on treaty functions, however, it seems there’s some evidence of a different kind of fragmentation emerging among the secondary rules of international law.  For example, Geir suggests at the end of his post that “Treaty law must be complemented by international institutional law”.  But treaty law and international institutional law are not required by any rule of international law to get along — it’s equally possible that the result produced by the law of treaties (say an interpretation of a treaty constituting an International Organization under VCLT Article 31) and international institutional law (say an interpretation of the same treaty employing the implied powers doctrine) could generate competing or conflicting results with respect to the same primary rule.  Catherine’s post makes this point more descriptively, noting how international institutional law has come to supplant the general law of treaties in the IO treaty context.  As with fragmentation among tribunals or primary rules, however, international law doesn’t tell us if this is the correct result.  Examples of secondary rules that are conciliatory to other secondary rules are relatively rare — although the VCLT does include a few examples with respect to IO treaties in Articles 5 and 20(3).  But, on the whole, the international legal order says little, if anything about whether one set of secondary rules should be accorded priority over another.

Moreover, I don’t think this competition over interpretative rules is an isolated case.  Although The Oxford Guide to Treaties does not explicitly flag this idea of secondary fragmentation specifically, there is evidence of it in several other chapters.  For example, although the VCLT’s rules on material breach purport to function differently than the law of state responsibility, Bruno Simma and Christian Tams’ chapter on remedies for treaty breach makes clear that these two sets of “secondary” rules are in competition with one another (and, moreover, that the law of state responsibility on countermeasures may be winning in the sense that it is those rules not the VCLT provisions on which States currently rely).  Malgosia Fitzmaurice’s chapter makes a similar point about tensions over exceptional circumstances where the law of treaties has doctrines — impossibility, and rebus sic stantibus — that may be threatened by the law of state responsibility’s doctrine of necessity.

The Oxford Guide to Treaties Symposium: Contractual and Institutional Elements in the Treaty Process

by Catherine Brölmann

[Catherine Brölmann is Associate Professor of Public International Law at the University of Amsterdam]

Particular features in the interpretation of constitutive treaties or secondary acts of international organizations reflect the special nature of the law of organizations, which brings both contractual and institutional features in the treaty process.

Following up on posts of Richard Gardiner, who brings up pertinent questions regarding treaty interpretation that are not addressed by the VCLT, and Geir Ulfstein, who points at the increasing publicness of international law ia through the creation of treaty bodies, this post briefly focuses on a phenomenon that is not covered by the Vienna Convention on the Law of Treaties and strictly speaking not by the law of treaties as such. This is the mixing of contractual and institutional elements in international treaty practice.

The Guide to Treaties addresses what is arguably a specific aspect of this combination of contractual and institutional elements, with a chapter on the interpretation of constitutive treaties and acts of international organizations. We find two trends in the interpretation of constitutive treaties or ‘law of the organization’ in general, that stand apart from the standard interpretive approaches in the Vienna Convention: a ‘teleological approach’ to the text, not very common in regular treaty interpretation but akin to traditions in national law of statutory interpretation in an organic, constitutional context; and particular importance attached to the ‘practice of the organization’ as opposed to the practice of the original treaty parties (such accompanied by a corresponding disinterest for the travaux préparatoires and the intention of the original parties).

The claim is that this ‘constitutional’ approach to the law of the organization can be traced to the notion of that law as special and ‘self-contained’ – not in the traditional sense of a particular area of substantive law or lex specialis, but rather as a semi-independent or internal legal order based on specific institutional rules. Constitutive treaties of organizations in fact have been attributed something of a constitutional character (cf the ICJ in 1996), binding member States to a set of coherent internal laws on the organization’s competences, functions and goals. The special character of the law of the international organization as as an internal legal order, semi-closed off from general international law, is also recognized by the Vienna Convention: think for example of the general reservation clause in Article 5, and the role for the competent IGO organ in the acceptance of a reservation provided for by Article 20(3).

The mix of contractual and institutional features which consequently marks the international organization as a legal environment is visible in more areas of treaty practice than just interpretation. The use of an institutional infrastructure for international processes dates back to the mid-nineteenth century, but recent decades show international organizations using in a particularly efficient and self-confident way their institutional mechanisms in order to streamline stages in a treaty-making or law-making process that were traditionally based on individual consent of the contracting parties – for example in the adoption of the text (UNGA), the initial signature subject to ‘ratification’ (ILO) or the inclusion of a treaty text in a binding resolution (UNSC).

This phenomenon falls outside the scope of the VCLT in a very fundamental way, as it moves in and out of the law of treaties framework altogether. In that respect it is different from new trends such as evolutive interpretation or ‘living instrument doctrine’ -, which may be unaddressed by the Vienna Convention, but which are considered to be within the law of treaties paradigm. On the other hand, in the case of interpretation of IGO law, a preliminary question arises as to the applicability of the Vienna Convention to begin with. Are UN Security Council resolutions a form of treaty law or do they amount to ‘international legislation’? Was the 1999 new ‘strategic concept’ of NATO a living instrument interpretation of the 1949 North Atlantic Treaty or a redefinition of powers and competences by an Organization making use of its compétence de la compétence to enact institutional innovation?

The Oxford Guide to Treaties Symposium: Enter Non-State Actors

by Peter Spiro

I’ll join the chorus of praise for my colleague Duncan’s book. It will clearly become the standard reference work in the area. As IL scholarship proliferates, there is a lot of smart money in handbook volumes such as this one. The Oxford Guide to Treaties is a one-stop source for the best thinking on the subject.

Duncan is also to be congratulated for his forward-looking inclusion of a chapter by Kal Raustiala on NGOs and treaty-making. Kal’s entry represents the best kind of writing for the format. It gives the newcomer a parsimonious but informative survey of the subject. For those already engaged on the subject, it has an edge.

Kal argues that states have a self-interest in permitting NGO participation in treaty-making. NGOs provide information in both treaty negotiation and implementation. NGO inclusion pays domestic political benefits, helping to get NGOs on board during the negotiations who are then more likely to turn around and sell the deal back home. When it comes to delicate negotiations, states can have their cake and eat it too, by closing the door on NGOs in informal-informals and the like. Kal describes how NGOs have become more prominent in the treaty-making process. But in his view this rise

should not be viewed as necessarily antagonistic to State interests. Nor does it undermine the centrality of States in international law-making. The roles played by NGOs remain formally subject to State control. This control ranges from the accreditation process, which keeps out the overly radical or insufficiently organized (or just deeply disliked), to the use of informal negotiating venues and forums, which keeps out everyone. Yet this control is used sparingly and, I have argued, often strategically. The reality is that NGOs remain active in a wide range of treaty settings.

I agree that as a formal matter states are still in charge. In theory, they can reject NGO participation. In practice, however, no multilateral regime can succeed without a substantial NGO presence. That’s not just a matter of putting NGOs to work in the service of states. It’s about legitimizing global institutions. The results won’t always be to state liking, either: NGOs are influential in treaty-making, in a way that’s often consequential.

The treaty form works against NGO participation, at least for now. (There is an interesting chapter by Tom Grant on non-state treaty-makers, but his discussion focuses on other territorial creatures – subnational jurisdictions, external territories, and insurgent groups.) When NGOs get involved on the basis of something more like equality, it has made sense to take other institutional routes. Think the Kimberley Process or the Global Fund. There’s a selection error in focusing on treaties as a form of international agreement-making. As a legacy institution, treaties will reflect a lower level of non-state participation than more novel agreement forms. I understand that this is beyond Duncan’s brief. But it’s something to keep in mind as we think about the spectrum of international instruments.

As a bonus to Kal’s chapter, Duncan includes model clauses on the terms of NGO participation (see 673-75). I’m sure that’s something new in a treaty handbook. I expect that future editions of the book will be including more material relating to non-state actors as their power becomes more evident and as the treaty form adapts accordingly.

The Oxford Guide to Treaties Symposium: The Public Character of Treaty Law

by Geir Ulfstein

[Geir Ulfstein is Professor of Public and International Law at the University of Oslo]

Treaty law is increasingly acquiring a public character. One reason is that more and more treaties set up treaty bodies, i.e. organs that are neither formal international organizations nor international courts. Examples are the Conference of the parties (COPs) used in international environmental law, the Antarctic Treaty Consultative Meeting, or the supervisory organs established by human rights treaties, such as the Human Rights Committee.

These bodies may exercise public powers in the form of law-making, executive decisions or enforcement. Treaty law is essential for their creation. But treaty law must interact with international institutional law when it comes to determining the powers of the treaty bodies, as well as the legal status and effects of their decisions. With public powers comes also issues of the effectiveness and legitimacy of their activities.

The law-making powers of treaty bodies represent a move away from ad hoc consensus-based treaty-making. These bodies are permanent fora that may be tasked with law-making. Examples are COPs in environmental treaties empowered to adopt new protocols, such as the Kyoto Protocol, i.e. the adoption of formally new treaties. But some of these treaty bodies provide for their COPs to adopt or amend annexes to the treaties or protocols, subject to the non-acceptance of these decisions by individual states parties. By requiring action by states in order for them not to become committed rather than to become committed – opting out, instead of opting in – the efficiency of law-making is greatly enhanced.

Other examples are the ‘quasi-legislative’ powers of the UN Commission on the Limits of the Continental Shelf to adopt recommendations on the outer limit of the continental shelf that according the Law of the Sea Convention will, if implemented by the coastal state, be ‘final and binding’. Similarly, the ICC Assembly of States Parties has the power to adopt and amend ‘elements of crime’ which shall ‘assist the Court in the interpretation and application of Articles 6, 7 and 8 bis’ (genocide, crimes against humanity, war crimes and the crime of aggression), subject to the condition that such elements ‘shall be consistent’ with the Rome Statute.

In determining the powers of the treaty bodies the traditional canons of treaty interpretation apply, including the principle of effective interpretation. But it may also be a question of the applicability of international institutional law, especially the principle of implied powers. This principle has for example been invoked in relation to the powers of the Human Rights Committee to adopt General Comments, interim measures, as well as follow-up measures.

The legal status of the treaty bodies’ decisions may also raise difficult questions. For example, article 17 of the Kyoto Protocol enables the Meeting of the Parties to adopt ‘rules’ relating to the operation of the system for trading in emissions of greenhouse gases. It is not entirely clear whether such rules are legally binding. But even if the treaty bodies are not empowered to make binding decisions, such decisions are not necessarily without legal significance. Article 31(2)(b) of the Vienna Convention on the Law of Treaties requires that treaty interpretation takes into account ‘any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’. The practice of the treaty bodies should be seen as international institutional practice rather than the practice of states parties – especially if the treaty bodies are composed of independent experts. But responses by states practice may establish an interaction between such institutional practice and that of states parties.

Treaty bodies may also have important functions in supervision and enforcement of treaty obligations – complementing treaty law and the law of state responsibility. Examples are the supervisory bodies in human rights, international environmental law and disarmament law. But, again, it may be difficult to determine the legal status of enforcement decisions. This is illustrated by article 18 of the Kyoto Protocol establishing that mechanisms ‘entailing binding consequences shall be adopted by means of an amendment to this Protocol’. It is not obvious what kind of measures should be regarded as entailing ‘binding consequences’, but it would be difficult to accept e.g. deduction of emissions quotas at a penalty rate, as adopted by the MOP as part of its Marrakesh Accords, without a treaty basis in the form of an amendment.

It may be concluded that treaty law is necessary, but not sufficient to understand the legal aspects of treaty bodies and their activities. Treaty law must be complemented by international institutional law. This branch of international law is characterized by its open-ended nature, and has sufficient flexibility to be adapted to the specific functions of treaty bodies.

The Oxford Guide to Treaties Symposium: Comparing International and US Approaches to Interpretation

by Jean Galbraith

[Jean Galbraith is Assistant Professor at Rutgers-Camden School of Law]

Congratulations to Duncan Hollis and the contributors to The Oxford Guide to Treaties [OGT].  This is a magnificent volume — one that fully lives up to its aim of “explor[ing] treaty questions from theoretical, doctrinal, and practical perspectives.”  For an edited volume, it is a remarkably coherent treatise.  Personal views of the individual authors do emerge, but they rarely dominate the doctrinal presentation.  Where the law is unclear, this is acknowledged by individual authors or sometimes by the very structure of the book itself.

Nowhere is this more clear than with regard to treaty interpretation.  By my count, OGT offers no fewer than four approaches to treaty interpretation.  Three are laid out in Part IV, which is explicitly devoted to treaty interpretation.  Richard Gardiner explores the foundational VCLT rules; Catherine Brölmann focuses on the interpretation of constitutive treaties creating IOs; and Başak Çali addresses the interpretation of human rights treaties.  The fourth approach to treaty interpretation is found in David Sloss’s contribution on the domestic application of treaties.  Professor Sloss points out that some domestic courts take what he terms a “nationalist approach” and emphasize domestic rather than international principles of treaty interpretation (such as a canon of deference to the positions of their executive branches).  His lead example for this is the United States, and he explicitly cites the Supreme Court’s Medellin decision from 2008.

Here, I want to suggest that international and U.S. domestic approaches to treaty interpretation have largely moved in opposite directions over the last thirty years.  The Restatement (Third) of Foreign Relations Law (1987) emphasized that in the “United States tradition, the primary object of interpretation is to ‘ascertain the meaning intended by the parties’ rather than focus simply on the text” (§ 325, reporter’s note 4).  It specifically noted the special nature of IO charters, saying that the interpretation of such treaties should be subject to Chief Justice Marshall’s observation that “‘we must never forget that it is a constitution that we are expounding’” (§ 325, cmt. d).

Today, as Professor Brölmann indicates, international law accepts a strong teleological emphasis for the interpretation of at least some categories of treaties (such as IO charters).  Yet in the United States, this prospect has faded in favor of a more strictly textualist approach.  Contra to the Restatement, the Supreme Court in Medellin showed a marked preference for text as opposed to other possible sources of meaning.  It also showed no interest in a broader teleological approach in interpreting the U.N. Charter and the Statute of the ICJ.  Where the Restatement emphasized that IO charters are constitutional, Medellin treated them like ordinary treaties.  One can overemphasize this shift, which undoubtedly has its roots in broader changes in how U.S. courts approach statutory and constitutional interpretation, but it is nonetheless a real one.

One interesting question for treaty interpretation going forward (particularly for those of us interested in treaty interpretation by the United States) is to what extent this shift in U.S. judicial approaches has and will affect interpretative positions taken by the U.S. executive branch.  In some sense, this is the flip of Professor Sloss’s point about U.S. courts deferring to the executive branch.  U.S. courts may indeed defer to the executive branch, but the executive branch in turn may interpret treaties in line with the interpretive principles set out by U.S. courts.  This in turn might affect the positions taken by the United States as an actor within IOs.

This is just one example of the issues and questions that reading the OGT brought to my mind.  I wish I had had it available to me in my past work, and I know I will be turning to it often in my future work.  It is a truly rich resource for practitioners and scholars alike.

The Oxford Guide to Treaties Symposium: Starting a Conversation on Interpretation

by Richard Gardiner

For the past 15 years courts, tribunals, practicing lawyers and academics concerned with treaty interpretation have been paying increasing attention to the three articles on the topic in the 1969 Vienna Convention on the Law of Treaties. Because the International Law Commission as architect of these provisions confined their drafts to what they saw as general principles, stated laconically and in places elliptically, there has been some scope for dispelling misunderstandings.

Among the points made by the ILC were that the general rule is the whole of article 31 VCLT; that subsequent agreement and subsequent practice are authentic interpretation (the latter showing particularly that the approach is not a purely textual or literal exercise); that ordinary meaning, as part of the general rule, does not indicate some mythical single meaning but typically one selected by reference to context and to object and purpose; that the reference to relevant rules of international law was, at least in part, included to take account of intertemporal considerations; and that consideration of the circumstances of conclusion and of preparatory work was not to be unduly inhibited, the stated limitations applying only to their use as a determining factor.

However, the ILC never saw these provisions as anything amounting to a complete set of formulae for interpretation.  Thus investigation of the subject has moved on to consider, among other things, whether there are particular approaches to be taken in special subject areas such as human rights or international trade, how to deal with time factors, whether particular considerations arise if international organisations are involved, whether there is a useful potential crossover from the  originalist/constructionist debate in constitutional interpretation, and whether an evolutionary method of interpretation forms a distinct approach.

The Guide takes up some of these issues but much of its consideration of the topic is set in the context of the VCLT provisions.  Perhaps now is a good opportunity to take stock of new lines of investigation.

The Oxford Guide to Treaties Symposium: Can Severability Be Made Less Severe?

by Edward Swaine

[Ed Swaine is Professor of Law at George Washington University Law School]

I’m pleased that the subject of reservations, which is near and dear to my heart, is attracting the attention of such esteemed commentators. The illuminating comments by Professor Stewart and Dean Koh, which I had the chance to read while preparing this post, focused in part on the severability solution, according to which impermissible reservations can at least sometimes be disregarded so as to leave a reserving state as party to the entire agreement.  Their comments, which query the compatibility of this approach with state consent and with securing Senate consent to future agreements, may be misperceived as resulting from parochial concerns – particularly in light of objections to the U.S. reservation to the Convention on Certain Conventional Weapons, in which some states likewise asserted the capacity to disregard the U.S. reservation and hold it to the “entirety” of the Convention.

I think the objection to the solution goes deeper, for reasons I could only touch on in my chapter.  ILC Guideline 4.5.3 would create a presumption that impermissible reservations are severable unless it is shown that the reserving state viewed the reservation as a sine qua non of its consent (as opposed to, say, a half-hearted ask or an attempt to yank another state’s chain).  Unless the matter rises to the legitimate attention of a treaty monitoring body or an international tribunal, this puts two inquiries to states trying to evaluate another state’s obligations – first, whether the reservation is impermissible because it violates the treaty’s object and purpose; and second, the psychological conditions of another state’s consent – which must also be anticipated by the reserving state in its turn.

This does not seem very workable, and the reason is in part because of the (ultimate) indulgence of the consent principle.  One might defend the first inquiry as potentially objective, though states are bound to differ on a treaty’s core values; the VCLT wisely pretermits the inquiry’s relevance to the extent it opens all reservations to objection on any grounds. The second inquiry, though, is not only subjective, but foreign to the responding state.  Still, it would have been open to the ILC to defend this presumption of severability as an information-forcing default rule – one that makes the reserving state disclose, via the reservation text or otherwise, its intent to insist on the reservation as a condition of adherence.

It is at that point, though, that the Guidelines seem to run out.  The opposability construct of the VCLT, to which the Guidelines are somewhat hostile, works after a fashion only because the call-and-response of reservations and objections establishes a shared understanding through what look like unilateral vetoes: first a veto by the reserving state of the agreed treaty content, then a veto by any objecting state to the reservation’s effect against it (and, unfortunately, either to the reserved-to provision or the reserving state’s consent, which is not an appealing choice).  Guideline 4.5.3 instead supposes that states will mutually intuit an offense against a treaty’s object and purpose; then develop a shared view of the reserving state’s “reservation price” based on the totality of circumstances relating to its intentions; and last but not least, as a belated concession to the consent principle and to its relative expertise, permits the reserving state to “express at any time its intention not to be bound by the treaty without the benefit of the reservation.”

The ILC is to be commended for taking seriously the problem of reservations that are inimical to a treaty’s mission and for addressing the limitations of the objections scheme. And it is notable that it did so while (at least nominally) resisting the position that each state can decide for itself whether a reserving state remains a party to the entire agreement, having recognized the importance of a state’s consent to the initial set of terms it assumed. But permitting a state to end its obligations at any time, with potentially retroactive effect, on the pretext that its reservation is impermissible – well, that converts a problem of uncited violations into one in which a team can blow the whistle on itself or, in any event, pick up the ball and go home when it chooses.  The Guidelines probably anticipate more refereeing by treaty monitoring bodies, but that will elicit its own objections. In the meantime, it seems likely that the proposed compromise on severability will gain adherents only slowly, if at all, and in the meantime make the players even more uncertain of the rules.

The Oxford Guide to Treaties Symposium: Validity of Reservations Revisited

by Marko Milanovic

[Marko Milanovic is Lecturer at the University of Nottingham School of Law]

I am grateful to Duncan for inviting me to contribute to this conversation, inspired by his important new book. Let me continue where David Stewart and Harold Koh left off, namely with the issue of the validity of reservations and the innovations in that regard in the ILC’s freshly minted Guide to Practice. I would disagee with Professor Koh that the approach ultimately adopted by the ILC ‘makes little sense and smacks of unfairness’. On the contrary – the ILC has not only brought much needed clarity to the Vienna Convention regime, but has succeeded in reconciling seemingly irreconcilable doctrinal and ideological positions, in a compromise that I hope most governments will in the end find palatable. Let me try to explain why.

When one reads the (rather paltry) Articles 19-22 VCLT, particularly in light of the ICJ’s Reservations to the Genocide Convention opinion, one cannot avoid the impression that the process of determining whether a reservation was invalid as being contrary to the object and purpose of a treaty was meant to be more or less inter-subjective: each state should determine for itself whether a given reservation was compatible with the treaty’s object and purpose, and if it was not it should make an objection to that effect. But such an inter-subjective approach looks remarkably unappealing from the perspective of major multilateral normative treaties, particularly in the human rights context. The rights of individuals, so the reasoning among many human rights lawyers went, should not depend on the existence of objections, vel non, by third states, especially when reciprocity of state obligations has little place in the human rights context and when for a variety of reasons states routinely fail to object to reservations even when there are perfectly good reasons to do so. While objections to reservations would be probative, they could not be dispositive. It would indeed primarily be upon courts or treaty bodies to determine whether a reservation is compatible with the object and purpose of the human rights treaty, while the consequence of invalidity would normally not only be the nullity of the reservation, but also its severability, so that the reserving state would remain bound by the human rights treaty without the benefit of its reservation.

Many governments were less than pleased with what they saw as a power-grab by human rights bodies and a usurpation of their sovereign prerogatives. The ILC, being the bastion of international law orthodoxy, was no more pleased, nor was Alain Pellet as its Special Rapporteur. How could international law survive as a coherent, unified system if more of its branches followed the human rights example and asserted that because they were special they needed special rules, rather than the outdated Vienna framework. If that was true for human rights, why would it not be true for trade, the environment, or whatever other topic people became strongly devoted to. Fragmentation had to be resisted, and Pellet and the ILC could not, would not accept the ‘human rights are special’ argument. No true international lawyer, even a gentle, human rights-loving one, could accept its basic ideological premise.

But, during the (very) lengthy process of working on the Guide, the ‘generalists’ and the ‘human rightists’ actually talked to one another, including at a series of meetings organized in Geneva between the ILC and human rights treaty bodies. Rather than harden, their respective positions evolved. While from the generalist perspective the specialty claim could never be accepted, there was still room for compromise. Perhaps it was the general regime itself that could be so interpreted – or adjusted – to accommodate the concerns of the other side, and this time not just for the benefit of human rights. And so we have now the Guide, in which Pellet so very cleverly succeeded in reconciling positions that before seemed irreconcilable. He and the ILC did so by making a series of crucial conceptual moves.

First, according to the Guide, Article 19 VCLT should be regarded as laying down objective criteria for the validity of reservations, as inter-subjectivity is a recipe for chaos. Secondly, Articles 20-23 VCLT only deal with those reservations which are objectively valid under Article 19; they do not mention nor pertain to reservations which are in fact invalid. Thirdly, while states may object to reservations that they consider invalid, this is merely persuasive evidence of invalidity. In fact, objections only have real legal effect if they are made against reservations which are objectively valid; the objecting state may object for any reason whatsoever, simply because it does not want to accept the modified treaty bargain that the reserving state is offering. Fourthly, while the VCLT does not say what are the consequences of an invalid reservation, the only sensible option is to accept that such a reservation is null and void. Fifthly, however, saying that an invalid reservation is a nullity does not resolve the issue of the reserving state’s status as a party to the treaty. That will depend on the intention of the reserving state, which has a choice – either stay on as a party to the treaty without the benefit of the invalid reservation, or say that it no longer considers itself bound by the treaty, while acting under a rebuttable presumption that the reserving state intends to remain a party. (See gudelines 4.3 and 4.5).

Whether this is really the Vienna regime, ‘Vienna-plus,’ or something else entirely will, I imagine, be the object of some debate. But what seems to be beyond debate is that the Guide’s approach to the invalidity of reservation accommodates most of the human rights-inspired critique of Vienna without giving any ground to the idea of specialty. This is a general regime applying to all treaties, but it still moves from the inter-subjective approach in which state objections are the only thing that matters, it treats invalid reservations as a nullity, and it allows them to be severed. Yet they can only be severed if the reserving state does not actively oppose its continued status as a party to the treaty, with the presumption of severability operating on the unstated assumption that passivity is frequently the politically most likely reaction by states, and that the retention of the state within the treaty regime but without the reservations would be a probable outcome.

The Guide thus achieves a compromise that avoids fragmenting international law, accommodates concerns arising from the context of normative multilateral treaties, including human rights ones, yet still gives ultimate say on the matter to the principle of state consent. For instance, all the United States would need to do to avoid the severability of any of its reservations to the ICCPR, on the assumption that some of them in fact are objectively invalid, would be to say (as it has done so several times already) that it considers them an integral component of its consent to the entire treaty. That does not seem to be too onerous a burden, and is indeed a far cry from the approach articulated by the Human Rights Committee in its General Comment No. 24.

There is, in short, much to be commended in this anti-fragmentationist yes still progressive compromise, but it of course remains to be seen whether both governments and human rights institutions will appreciate it for what it is. I certainly hope they do.

The Oxford Guide to Treaties Symposium: What Happens if a Treaty Reservation is Invalid?

by Harold Hongju Koh

[Harold Hongju Koh is the Legal Adviser, U.S. Department of State.]

Professor Duncan Hollis’ magisterial new book, The Oxford Guide to Treaties, collects an enormously useful amount of up-to-the-minute scholarship on myriad pressing questions of international treaty law. Its publication comes at a particularly opportune moment, as the International Law Commission’s (ILC’s) Guide to Practice on Reservations to Treaties was finalized by the Commission in 2011, and is currently on the calendar for action next year by the U.N.’s Sixth Committee and General Assembly.

The chapter on Reservations to Treaties, by Professor Ed Swaine, a former Counselor on International Law in the Office of the Legal Adviser, addresses one of the thorniest issues in the already complicated area of reservations – namely, the effect of objections to so-called “invalid” reservations.  For good reason, Professor Swaine does not attempt to resolve this contentious issue definitively in his chapter, as it remains subject to differing and contested views among States.

But the penultimate draft of the ILC Guide provided that “[w]hen an invalid reservation has been formulated, the reserving State or the reserving international organization is considered a contracting State or a contracting organization or, as the case may be, a party to the treaty without the benefit of the reservation, unless a contrary intention of the said State or organization can be identified.” (emphasis added).  Simply put, under this approach a State that has made a reservation that is later deemed “invalid” could nevertheless be bound by the treaty without the benefit of the reservation it took pains to assert.  For good reason, this proposal drew a raft of vigorous comments from States at the annual Sixth Committee discussion in 2010.

The final draft of the ILC Guide has largely maintained the same approach, although happily giving somewhat more deference to the intent of the reserving State. According to current section 4.5.3, “[t]he status of the author of an invalid reservation in relation to a treaty depends on the intention expressed by the reserving State or international organization on whether it intends to be bound by the treaty without the benefit of the reservation or whether it considers that it is not bound by the treaty.” (emphasis added) But the presumption is clear: “Unless the author of an invalid reservation has expressed a contrary intention or such an intention is otherwise established, it is considered a contracting State … without the benefit of the reservation.”

On its face, this approach makes little sense and smacks of unfairness.  As the United States has consistently noted, it is difficult to square this approach with the bedrock principle of treaty law that States are bound only to those obligations they affirmatively consent to undertake.  Reservations, which allow States to condition the scope of their consent, are a well-established and critical feature of international treaty law. If a State’s condition to consent is later deemed invalid because it is incompatible with a treaty’s object and purpose, it makes little sense to pretend after the fact that the reserving nation de facto gave the broader, unconditioned consent that it expressly chose to withhold in the first place.

The Vienna Convention on the Law of Treaties (VCLT) rules on reservations are admittedly opaque.  But acknowledging that States are only bound to their treaty commitments with their genuine consent offers a far more workable framework for addressing reservations that are later characterized as invalid by another State.  For example, when the U.S. objected to Pakistan’s reservations to the Convention Against Torture (CAT) and International Covenant on Civil and Political Rights (ICCPR), we took the position that the totality of those reservations was incompatible with the object and purpose of the treaties.  Yet we nevertheless chose to maintain –consistent with VCLT Art. 21.3–treaty relations with Pakistan, except to the extent of their reservations. In short, severing an invalid reservation from a treaty is not like severing an unconstitutional clause from a piece of legislation. A fair-minded observer still must determine the scope of the reserving nation’s genuine consent to the treaty’s provisions, and make a good faith effort to respect both the party’s actual consent and its effort to carve out exceptions to that consent through particular reservations.

The ILC’s Guide to Reservations is sure to foster ongoing debate among international lawyers. It is the work of what the Statute of the ICJ, Article 38 calls, “the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of [international] law.”  But as that debate, and others, unfold, it will be a blessing to have Professor Hollis’ Oxford treaty volume, whose chapters –also authored by distinguished publicists with their own legal views–consistently give us a thoughtful, useful starting point –and counterpoint –for thinking through  this and many other knotty treaty questions.

The Oxford Guide to Treaties Symposium: Treaty Reservations and ‘Objections-to-Reservations’

by David Stewart

[David P. Stewart is a Visiting Professor of Law at Georgetown University Law Center]

Duncan Hollis deserves hearty congratulations on the publication of the Oxford Guide to Treaties.  There’s no doubt that it will quickly become the essential reference for lawyers and other treaty specialists in foreign ministries and international organizations everywhere, to say nothing of judges, professors and private practitioners.  Its 25 substantive chapters cover the full range of issues raised by contemporary treaty practice, both bilateral and multilateral, so it’s definitely not a volume you’ll try to read at one sitting.   But international lawyers will certainly want to keep it handy.   Many thanks to Duncan and his co-authors for giving us one of those “can’t-do-without” books.

For various reasons, I was drawn in particular to Ed Swaine’s chapter on treaty reservations and was delighted to find a thorough and thoughtful analysis which combines practice and theory and contrasts the rules of the Vienna Convention on the Law of Treaties with the International Law Commission’s 2011 Guide to Practice.   He traces the evolution of the reservations system as multilateral treaties proliferated over the past century and takes up the knottiest of issues – for example, how reservations differ from other unilateral statements (such as interpretive statements or “understandings”), whether some types of reservations are void ab initio or only when objected to by other States, and what legal effect such “objections” actually have — and who gets to decide.

As Swaine rightly notes, the “objections-to-reservations” rules in VCLT articles 20 and 21 leave much to be desired, at least with respect to clarity and certainty.  An objecting State can, at its option, decline to accept treaty relations with the reserving state; otherwise, “the provisions to which the reservation relates do not apply as between the two States to the extent of the reservation.”  (Art. 21(3)).

The ILC’s Guide reiterates the first part of this approach (noting, in para. 2.6.6, that a state or international organization “may oppose the entry into force of the treaty as between itself and the author of the reservation”) but then introduces a further complication by distinguishing between “valid” and “impermissible” reservations and by introducing the concept of “establishing” a reservation.  In summary, an “established” reservation is one which is authorized by the treaty or otherwise permissible, properly formulated and communicated, and accepted by the other party (para. 4).  By definition, then, established reservations are objection-proof (or, put otherwise, objections prevent establishment.) But an objection to a reservation which has not been “established,” even if that reservation is “valid,” has the legal effect of precluding the reservation “from having its intended effects” but does not preclude the entry into force of the treaty as between the reserving and objecting states, unless the objecting state expresses such an intention (para. 4.3, 4.3.1 and 4.3.5).  Whether this additional level of detail adds to the clarity of the rules seems questionable.

These issues get even more difficult when the objection is to only part of a reservation (a subject not explicitly addressed by the Convention or the ILC Guide).   But perhaps the most troubling recent development in recent “treaty reservation practice” has been the emergence of a so-called “severability” rule – according to which some reservations are considered invalid or impermissible and therefore “severable,” meaning that the reserving state is bound to the treaty in question without the benefit of its asserted reservation.  The origins of this approach, as Swaine notes, lie in the European Court of Human Rights (in particular its 1988 decision in the Belilios case); the UN Human Rights Committee adopted a similar approach in its General Comment 24 (1994) formulated in the context of the International Covenant on Civil and Political Rights.  Although the Covenant contains no prohibitions on reservations, the Committee adopted a very restrictive view on the subject, opining that no reservations could be taken to provisions that represent customary international law (and a fortiori those having the character of peremptory norms), or that make rights non-derogable, or that provide “supportive guarantees,” or that interfere with its own “monitoring role” or its competence to interpret the requirements of the Covenant.  The General Comment drew immediate and sharp protests from the United States, the United Kingdom and France for a variety of reasons, among them the suggestion that the Committee could itself declare a given reservation invalid and “severable” and consider the reserving State bound to the Covenant without benefit of the reservation in question.

A simple illustration demonstrates why the Committee’s approach was of such concern to the United States.   Article 20 of the Covenant provides that “any propaganda for war shall be prohibited by law” and that “any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law.”  For obvious First Amendment reasons, the United States could not comply with the full import of that article, and thus conditioned its adherence to the Covenant on a carefully crafted reservation stating that “article 20 does not authorize or require legislation or other action by the United States that would restrict the right of free speech and association protected by the Constitution and laws of the United States.”  But the Committee took direct aim at this reservation, declaring that a state may not reserve the right to permit the advocacy of national, racial or religious hatred, because that prohibition was a matter of customary international law (if not a peremptory norm).

It is possible that the Committee failed to understand the real import of its statement, since it would leave denunciation of the Covenant as the only real option.  Had General Comment been issued before the United States undertook to adhere to the Covenant, ratification would have been far less likely.  Had the severability doctrine been understood to apply, it is hard to see how any responsible legal or political adviser could have recommended U.S. ratification in the first place.  The example simply illustrates that restrictive reservation rules may well work against broad adherence to some types of multilateral treaties, as Swaine’s chapter notes.

The ILC Guide takes a more nuanced approach than General Comment 24.  It states clearly that the fact a treaty provision reflects a rule of customary international law “does not in itself constitute an obstacle to the formulation of a reservation” but that no reservation can exclude or modify the effect of a treaty contrary to a peremptory norm.  (paras. 3.1.5.3, 4.4.2).  Nor can reservations be taken to non-derogable provisions unless “compatible with the essential rights and obligations arising out of that treaty.” (para. 3.1.5.4).  A somewhat similar tack is taken to reservations concerning dispute settlement provisions or monitoring (para. 3.1.5.7).  On the question of severability, the Guide states that unless it has expressed a contrary intention, a state is bound to the treaty without the benefit of an invalid reservation.

In an ideal world, “clean” ratifications are much to be desired.  But as multilateral treaties become more detailed and seek to address more issues on which various domestic laws and practices differ widely, it grows commensurately more difficult to achieve agreement on texts which all states can accept and implement.   Perfection in drafting is a virtue attributable to few if any treaties, especially those produced in multilateral fora.  Within limits, therefore, reservations continue to serve a useful function, permitting states to adhere to imperfect treaty texts by providing a means of accommodating differences.  In practice, the severability rule may turn out to be a significant disincentive to broad treaty adherence – particularly when decisions about the validity of reservations no longer rest on state consent.

The Oxford Guide to Treaties: An Opinio Juris Symposium

by Duncan Hollis

OGT CoverI’m extraordinarily pleased to be able to announce that today marks the start of the Opinio Juris symposium on my recently-edited volume, The Oxford Guide to Treaties (you can buy your copy here and there’s even a discount for Opinio Juris readers!).

The Oxford Guide provides a current and comprehensive guide to treaty law and practice. It does this in two parts.  First, it presents 25 chapters written by the world’s leading treaty-experts, exploring the world of treaties in five areas: (i) what a treaty is and who can make them; (ii) how a treaty is made (including the treaty-making process, signature, provisional application, deposit, registration, and reservations); (iii) how treaties are applied (including their territorial reach, third party rights and obligations, amendments, domestic application, succession, treaty bodies and conflicts); (iv) the rules on treaty interpretation generally and with respect to treaties on human rights and international organizations; and (v) how to avoid or exit a treaty commitment (including questions of validity, remedies for breach, exceptional circumstances, and termination). Second, the book pairs these explanations of existing rules and practice with examples of how modern treaties are drafted. Thus, the last section of the book includes 350 treaty excerpts on 23 treaty topics ranging from how to deal with multiple language treaty texts to the use of simplified amendment procedures (for those looking for a longer introduction to the project, see here).

Since the book is consciously treatise-like in its coverage, this symposium has opted for a slightly different format than the norm.  In lieu of comments on the book’s thesis, over the next few days we will use The Oxford Guide’s coverage as a launching pad for a discussion of some of the most pressing treaty questions confronting international lawyers.  The current schedule is (roughly) as follows:

(1) Today will focus on a discussion of reservations and other unilateral statements, with particular attention to the International Law Commission’s Guide to Practice on Reservations to Treaties

(2) Tomorrow, we will turn to treaty interpretation, particularly the phenomenon of evolutionary or “dynamic” interpretation; and, after a weekend respite,

(3) Monday, we will discuss the variety of functions treaties perform, such as their increasing “publicness”, the role of non-state actors in modern-treaty making, as well as any final comments that participants care to make.

In terms of participants, I’m pleased to have a truly distinguished group of experts participating in this on-line symposium.  Several of them are returning to the fold in the sense that they already contributed their time and expertise to The Oxford Guide itself, including Ed Swaine (who wrote the chapter on Reservations); Geir Ulfstein (who wrote on treaty bodies and regimes); Richard Gardiner (who wrote on the Vienna Rules on treaty interpretation); Catherine Brölmann (who wrote on interpreting constitutive treaties of International Organizations); Başak Çalı (who wrote on human rights treaty interpretation); and Christian Tams (who co-authored with Bruno Simma the chapter on remedies for treaty breaches)

In addition, I’m honored to have a group of very distinguished outside experts lend their voices to the conversation. I’m particularly pleased (and grateful) to have Harold Koh, the Legal Adviser to the U.S. Department of State, offer some thoughts on treaty reservations.  I’d also like to welcome three other commentators — Jean Galbraith, Marko Milanovic and David Stewart — and thank them for making the time to participate in these discussions.  I’m hopeful that one or more of my fellow Opinio Juris contributors may weigh in from time to time as well.

Altogether, we’ve got a set of really interesting topics and a great bench of experts to discuss them. I, for one, am really looking forward to the conversation.