1. Introduction

At the end of the twentieth century it appeared that we stood on the threshold of a new era in the legalization of international relations, one characterized by the exponentially increasing judicialization of dispute resolution.[1] States widely accepted the jurisdiction of international tribunals of various kinds to oversee their compliance with substantive legal commitments. Yet at the same time scholars lauded judicialization, some states began to retrench, withdrawing from their jurisdictional commitments with the more or less clearly stated intent to avoid procedures that would result in a finding of noncompliance. This essay considers whether a state violate the duty of good faith enshrined in  Article 26 of the Vienna Convention on the Law of Treaties (VCLT) if it unilaterally withdraws from or seeks to modify its jurisdictional obligations. Put differently, under what circumstances, if any, does the principle of good faith require imposing jurisdictional obligations on a state despite its objection?

I make two claims. First, the principle of good faith does not in general limit a state’s ability to withdraw from its jurisdictional obligations—by which I mean a legal obligation to participate in proceedings before an international tribunal the purpose of which is to assess the state’s compliance with its substantive obligations. The duty of good faith cannot be violated by actions foreseen in a treaty and the Vienna Convention on the Law of Treaties (VCLT) provides that withdrawal rights only exist when states intend to deviate from the default rule that they may not withdraw from a treaty. Put differently, withdrawal rights are in a sense always lex specialis.Those limits that do exist are creatures of contract. For example, states may and frequently do impose limits within a treaty on a state’s ability to refuse jurisdictional obligations. Even in the absence of limitations, the object and purpose test may limit a state’s ability to refuse jurisdictional obligations that are central to a treaty regime. This limitation, however, is itself based on the notion that a reservation may not modify the basic bargain struck in a treaty. It is therefore based on contractual notions of consent. Moreover, it only operates in those circumstances in which states use reservations to limit their jurisdictional obligations and in which jurisdictional obligations are central to a larger treaty regime.

Having demonstrated that states are, as a baseline matter, free to refuse jurisdictional obligations, I argue that a move towards implying greater jurisdictional obligations could have adverse unintended consequences for legal cooperation. First, jurisdictional obligations are costly. Imposing them over states’ objections would likely lead either to states refusing to join or remain within treaties entirely or to states watering down the treaty’s substantive commitments in order to reduce the costs of compliance. Second, making jurisdictional regimes mandatory would risk making tribunals’ more status-quo oriented in their interpretations of treaty commitments. Optional jurisdictional regimes that include only states particularly committed to cooperation have an often-overlooked benefit: by removing the states most likely to urge shallow interpretations of substantive obligations, they free tribunals to develop a jurisprudence that tends to deepen cooperation over time. Third, implying jurisdictional obligations risks crowding out non-judicial oversight mechanisms, such as the compliance and implementation committees commonly found in environmental regimes.

2. Does the duty of good faith limit the right of withdrawal?

The VCLT provides that ‘[every] treaty in force is binding upon the parties to it and must be performed by them in good faith’.[2] There are two ways in which this requirement of good faith might limit a state’s ability to withdraw from or modify its jurisdictional obligations. First, jurisdictional obligations are often contained in separate legal instruments, such as optional jurisdictional protocols. Such jurisdictional commitments and any associated withdrawal rights must be complied with in good faith just like any other legal obligation. The principle of good faith could thus directly limit states’ ability to withdraw from jurisdictional treaties. Second, the duty of good faith might be understood to link substantive and jurisdictional commitments when they are found in the same treaty. A state’s duty of good faith, one might argue, would thus prevent a state from withdrawing from or modifying jurisdictional commitments that are central to a larger treaty. I consider each of these two possible roles for the principle of good faith below.

a)  Does the principle of good faith compliance with jurisdictional commitments limit withdrawal rights?

Most treaties either do not have jurisdictional obligations or contain ‘arms-length’ jurisdictional obligations, meaning that the instrument creating jurisdictional obligations is different from the one containing substantive obligations.[3] For example, many jurisdictional obligations are contained in optional protocols. The Vienna Convention on Consular Relations contains an optional protocol permitting states to submit to the jurisdiction of the International Court of Justice (ICJ) for disputes arising under the Convention.[4] Other jurisdictional obligations are even further removed from the substantive ones. The jurisdiction of the International Center for Settlement of Investment Disputes (ICSID), for example, extends only to disputes that the parties have submitted to it through another instrument, often a bilateral investment treaty.[5]

In principle, one might think that the principle of good faith compliance with these jurisdictional treaties should limit states ability to withdraw when they are doing so for the purpose of avoiding an adverse determination as to state responsibility for conduct undertaken while party to the jurisdictional treaty. For example, in 2005 the United States withdrew from the Optional Protocol to the Vienna Convention on Consular Relations in part to avoid adverse decisions by the ICJ based on conduct like that at issue in the Avena case. One might claim that the United States should not be able to withdraw from the Optional Protocol, at least for claims based on actions that occurred while the United States was party to the Optional Protocol.

Withdrawal rights are, however, specifically bargained-for rights and thus the principle of good faith may not limit them. The default rule is that states may not withdraw from a treaty that is silent as to withdrawal. Withdrawal rights therefore only exist when states specifically intend to create them. Articles 54 and 56 of the VCLT make this clear. Article 54 provides that withdrawal may occur either pursuant to the rules for withdrawal agreed in a treaty or pursuant to the consent of all parties.[6] Article 56 says that if a treaty is silent as to withdrawal, no right of withdrawal exists unless a state can show either that ‘the parties intended to admit the possibility of denunciation or withdrawal’ or ‘the right of denunciation or withdrawal may be implied by the nature of treaty.’[7]

For its part, the duty of good faith ‘requires that a party to a treaty shall refrain from any acts calculated to prevent the due execution of the treaty or otherwise frustrate its objects.’[8] In principle, a state may violate the duty of good faith through an act that defeats the object and purpose of the treaty, even if the act itself is not expressly forbidden by the treaty.[9] However, as the ICJ held in the Nicaragua decision, ‘an act cannot be said to be one calculated to deprive a treaty of its object and purpose, or to impede its due performance [in violation of the duty of good faith], if the possibility of that act has been foreseen in the treaty itself.’[10] This holding is consistent with the ICJ’s approach to good faith in general. In noting that good faith is at the core of international law, the ICJ has declared that ‘[t]rust and confidence are inherent in international co-operation, in particular in an age when this co-operation in many fields is becoming increasingly essential.’[11] Yet trust, like its close cousin reputation,[12] is based on what states have actually promised to do. Absent a promise, there is no reasonable basis for trust. Perhaps for this reason the ICJ has also noted that the principle of good faith ‘is not in itself a source of obligations where none would otherwise exist.’[13]

It follows from this understanding of the duty of good faith that states do not breach the duty by withdrawing from a treaty in accordance with their treaty-based right to do so. Because the default rule is that withdrawal rights do not exist, a state may only lawfully withdraw from a treaty in circumstances in which states intended that they would have the right to do so. In order words, under  Articles 54 and 56 of the VCLT, a state may only withdraw from a treaty if states foresaw the possibility of lawful withdrawal. Yet the duty of good faith cannot be violated by actions foreseen in a treaty. The duty of good faith therefore does not limit withdrawal rights or the exercise thereof beyond what the treaty contemplates.

This is not to say, of course, that there are no limitations on the right of withdrawal where it exists. The right of withdrawal emerges from the parties’ consent and so too may limitations on that right. For example, the ICSID Convention makes clear that ‘[w]hen parties have given their consent [to jurisdiction], no party may withdraw its consent unilaterally.’[14] Somewhat more weakly, the UN Convention on the Law of the Sea (UNCLOS) provides that declarations by states limiting the application of otherwise mandatory dispute settlement procedures do not ‘in any way affect proceedings pending before a court or tribunal.’[15] Other treaties, such as the Marrakesh Agreement Establishing the World Trade Organization or, with exceptions, UNCLOS, make clear that accepting a treaty regime’s jurisdictional commitments are mandatory.[16]

Moreover, limitations may be implied as well as express. The VCLT directs a treaty interpreter to examine subsequent agreements or state practice as a means of interpreting a treaty’s provisions.[17] Interpretations of a treaty text that are more limiting than the text suggests thus could become binding on the parties to the treaty. The parties would be bound in good faith to comply with limiting interpretations of the treaty, even if the limitations are not found in the text.[18] Similarly,  Articles 31, 32, and 56 each authorize treaty interpreters to resort to the context, travaux, and nature of the treaty, respectively, in interpreting the scope of withdrawal rights. In any given situation, these sources might suggest limitations on withdrawal rights not expressly found in the treaty.

Critically, however, such limiting interpretations remain grounded in the contractual nature of withdrawal rights. The principle of good faith may require compliance with limiting interpretations based on subsequent state agreements and practice, but those limitations are nevertheless derived from state conduct. The limitations are not implied through a principle that operates as a constraint on states’ freedom to create withdrawal rights. Put differently, the principle of good faith does not limit a state’s ability to withdraw from a jurisdictional treaty beyond the limits that states themselves have agreed to in creating withdrawal rights. Where states bargain for withdrawal rights, either expressly or implicitly, they have foreseen the circumstances under which withdrawal may occur. Having allowed for the possibility, states cannot subsequently complain that withdrawal violates the treaty.

b)  Does good faith compliance with substantive commitments limit the ability of states to withdraw from or modify their jurisdictional commitments?

Even if the principle of good faith does not limit the ability of states to withdraw from jurisdictional treaties, the principle might still limit the ability of states to withdraw from jurisdictional commitments that are contained in a larger treaty. For example, many human rights treaties contain reporting obligations requiring states to submit to the competence of human rights bodies. Similarly, many investment and free trade agreements contain provisions consenting to investor-state arbitration. In such situations, a treaty’s jurisdictional commitments may be so central to the overall purpose of the treaty that a state should not be permitted to evade only the jurisdictional obligations. In effect, one might argue that good faith efforts to comply with substantive legal commitments prohibit the withdrawal from or modification of jurisdictional obligations, at least in some circumstances.

To be clear, the question here is whether the principle of good faith limits the ability of states to evade only the jurisdiction obligations in a treaty. States attempts to withdraw from an entire treaty containing both substantive and jurisdictional provisions would be subject to the same analysis discussed above. Likewise, a state could lawfully withdraw from only the jurisdictional portions of a treaty if the treaty itself permitted such a limited withdrawal. That is, the VCLT provides that withdrawal rights only exist when they are bargained for. The possibility of withdrawal in accordance with bargained-for rights has been foreseen by the parties and therefore cannot run afoul of the good faith principle.

States have, however, occasionally tried to withdraw only from the jurisdictional obligations contained in a treaty, or try to withdraw from some portion of a tribunal’s competence, in situations where the treaty did not contemplate such a right. Such attempts at evasion can occur through a withdrawal from the entire treaty followed by reaccession to the treaty with a reservation to the jurisdictional obligation. Trinidad & Tobago and Guyana followed this latter course when they denounced the First Optional Protocol to the International Convention on Civil and Political Rights (ICCPR) and then reacceded with reservations purporting to deny the Human Rights Committee jurisdiction to review petitions from capital defendants.[19]

Although such action is really an attempt to withdraw from the tribunal’s jurisdiction, tribunals have not distinguished such reservations from reservations made when initialing joining a treaty. In both cases, tribunals have invalidated the jurisdiction-limiting reservations based on the rule that reservations may not violate the object and purpose of a treaty. For example, the HRC ruled that Trinidad & Tobago’s reservation to the ICCPR’s First Optional Protocol regarding capital applicants violated the object and purpose of that instrument.[20] Similarly, in General Comment 24, the HRC opined that

‘A State may not reserve the right not to present a report and have it considered by the Committee. The Committee’s role under the Covenant, whether under  Article 40 or under the Optional Protocols, necessarily entails interpreting the provisions of the Covenant and the development of a jurisprudence. Accordingly, a reservation that rejects the Committee’s competence to interpret the requirements of any provisions of the Covenant would also be contrary to the object and purpose of that treaty.’[21]

More recently, in 2010 Pakistan entered a reservation to the ICCPR rejecting the HRC’s competence to receive and consider reports from States Parties.[22] Although the HRC did not explicitly invoke the ‘object and purpose’ test in its rejection of Pakistan’s reservation, the HRC noted that its ‘competence is of critical importance for the performance of the Committee’s monitoring function and essential to the raison d’être of the Covenant.’[23]

The object and purpose test can thus be used to limit a state’s ability to creatively use the right to withdraw from a treaty and the right to make reservations to a treaty to try to withdraw from the jurisdictional provisions of a larger treaty (or a portion of a tribunal’s jurisdiction created by a larger jurisdictional treaty). Any legal instrument embodies a basic bargain among the contracting states. The object and purpose test is best understood as a rule that reservations may not alter the essential terms of the bargain.[24] Thus, if oversight by a tribunal is part of the core bargain struck in a treaty, the object and purpose test is properly used to invalidate states’ efforts to limit a tribunal’s competence. In essence, jurisdictional obligations can become bound up with substantive obligations during a treaty’s creation if the parties to the treaty deem a tribunal’s competence to be fundamental to overall treaty scheme. The object and purpose test thus balances states’ freedom to design treaties to suit their needs while also recognizing that those treaties represent a package of concessions that states do not necessarily intend to be unilaterally modified after negotiations are completed. Although located in  Article 19 of the VCLT and distinct from the duty of good faith, the rule that reservations may not violate the object and purpose of a treaty is of course closely related to the duty of good of good faith’s prohibition on activities that violate the duty of good faith. In this sense, the duty of good faith can limit a state’s ability to evade jurisdictional obligations central to an overall treaty.

Although an effective way to limit states’ ability to evade supranational jurisdiction, the object and purpose test is also limited in its reach. First, it of course only applies to jurisdictional limitations attempted through reservations, either in conjunction with withdrawal and reaccession or on initialing joining a treaty. Because jurisdictional obligations are often in separate treaties, this limits the reach of the object and purpose test. Second, it applies only to situations in which states have not expressly made clear whether reservations are permitted to jurisdictional commitments. Third, states remain free to refuse to join or, where lawful, withdraw from treaties in their entirety. The object and purpose test thus does not provide a limitation on states’ ability to refuse to cooperate. It does not limit state sovereignty over both their legal commitments and how those commitments are supervised.[25] Instead, the object and purpose test is simply a doctrine that limits states’ power to unilaterally alter core terms of an interstate agreement.

3.  Should international law imply greater limitations on withdrawal rights?

As a descriptive matter, then, the law does not for the most part impose limitations on a state’s ability to withdraw from jurisdictional commitments beyond those limitations with a basis in the bargain struck by states themselves. One might argue that a mandatory link between substantive and jurisdictional obligations would deepen cooperation. On this view, oversight by a tribunal would increase compliance with substantive commitments. As long as those substantive commitments themselves remained the same, this increased compliance would be a boon to international cooperation.

The resistance to implying obligations beyond what states have bargained for is, however, good for international law’s ability to facilitate cooperation among states. Perhaps counterintuitively, imposing jurisdictional obligations through doctrines such as the duty of good faith would likely lead to shallower cooperation in at least three ways. First, states view jurisdictional commitments as costly. They would therefore likely offset greater implied jurisdictional commitments with weaker substantive commitments.[26] Alternatively, states might simply refuse to join a legal regime in the first place. Second, allowing states to opt out of jurisdictional obligations allows tribunals to more effectively develop a cooperation-minded jurisprudence. Third, a focus on judicial oversight mechanisms threatens to crowd out non-judicial oversight mechanisms, such as compliance and implementation committees authorized by Conferences of the Parties to multilateral environmental treaties, that are increasingly significant. I explain each of these arguments in greater detail below.

a)  Costly Commitment

For the most part, states must still consent to their own legal obligations.[27] In deciding whether to join a legal regime, states balance the benefits and costs of doing so, joining only if they expect the benefits from joining to exceed the costs. Jurisdictional commitments create both costs and benefits for states. Jurisdictional commitments are costly for several reasons. Most importantly, losing a dispute is costly to states. If they refuse to comply with the tribunal’s ruling, they may face reciprocal sanctions or, at a minimum, reputational sanctions. Second, tribunals expose and publicize conduct that states might prefer remain hidden. Third, participating in dispute resolution can be administratively burdensome, especially for developing states. In terms of benefits, dispute resolution and the costs it creates should deter some marginal violations. Dispute resolution can also be a relatively more efficient vehicle for elaborating the meaning of vague substantive commitments as compared to future negotiations.

If these benefits routinely outweighed the costs, there would be no reason to debate whether jurisdictional obligations should be implied. States would include them in treaties without any need for prompting. The fact that states do not do so, or have retrenched in their willingness to do so, suggests that often the costs of dispute resolution outweigh the benefits.[28] Implying jurisdictional obligations where they are not freely accepted thus imposes costs on states.

Will states simply accept these costs? It seems unlikely. Instead, states will take one of two actions. First, states may refuse to join international treaties at all or may withdraw from them. This was Trinidad & Tobago’s response to the HRC’s determination that it could not use reservations to determine the scope of its jurisdictional obligations.[29] The second possibility is that states will adjust their substantive commitments by making them less onerous.[30] If one part of a treaty becomes more costly for states, they may reduce the overall costs created by the treaty by reducing costs in another area. Consequently, implying jurisdictional obligations may have unintended consequences. It risks what Laurence Helfer has called ‘overlegalization’, prompting a backlash that ultimately limits the ability of states to cooperate.

b)  Rulemaking by Tribunals

Tribunals perform at least two critical functions. First, tribunals perform a compliance function. By providing a (hopefully) neutral forum in which disputes can be aired, they are able to determine and publicize violations of treaties. The costs of having one’s violations publicized hopefully deter future violations. Second, and equally importantly, tribunals perform a rule-making function, developing a jurisprudence regarding the content of the substantive obligations contained in the treaties they oversee. Critically, this jurisprudence informs the commitments of not only those states subject to the tribunal’s jurisdiction, but all states subject to the underlying substantive obligations.[31]

States face a tradeoff between these two roles for tribunals. On the one hand, increasing participation in dispute settlement allows tribunals to perform their monitoring and compliance functions more effectively. On the other hand, increasing participation in dispute settlement may actually undermine tribunals’ ability to develop a cooperation-minded jurisprudence. To see this, consider that states that freely accept additional jurisdictional commitments are likely to be cooperation-minded. Those that avoid jurisdictional commitments are likely not to favor deep cooperation. Both groups have an equal interest in the tribunal’s rulemaking function, because both sets of states are bound by the underlying rules. But participating in the tribunal’s rulemaking requires accepting the costs that come with the tribunal’s compliance function. For states that are cooperation-minded, this cost is relatively low. Such states expect to be in compliance most of the time with their commitments. States that view their substantive commitments with a gimlet eye, however, bear considerably greater compliance costs from accepting jurisdictional obligations. They are therefore more likely to avoid such commitments if they can.

The result is that states that submit to optional jurisdictional obligations will tend to favor deeper cooperation and will therefore produce cases that allow tribunals to make cooperation-minded jurisprudence. Complainant states will bring cases based on facts that involve more forward leaning interpretations of substantive commitments. More importantly, because respondent states are cooperation-minded themselves, they will not be as hostile to tribunal decisions invoking broad interpretations of substantive commitments. Respondent states are both more likely to moderate the kinds of arguments they make to tribunals, as well as to be receptive to potentially adverse decisions that result in deeper substantive commitments.

Both of these behaviors by respondent states free tribunals, in turn, to be more cooperation-minded in their rulemaking role. Tribunals can be less concerned that their legitimacy or authority will be challenged by states appearing before them. Moreover, they can apply their prior rulings in a quasi-precedential fashion more easily, in part because they are less likely to see states that are hostile to their prior rulings before them.

Critically, a tribunal’s jurisprudence nevertheless informs the obligations of all states bound by the underlying substantive commitments. Of course, tribunal rulings are not formally binding on states not subject to the tribunal’s jurisdiction. But tribunal decisions are for the most part not binding even on states that are subject to the tribunal’s jurisdiction. Indeed, at most tribunal decisions are binding on the parties to the dispute with respect to the facts of the dispute – and even then, not always. This observation has led Andrew Guzman and I to note that most international decisions, what we call international common law, are in fact soft law—nonbinding rules that nevertheless affect the expectations of parties.[32]

Despite the soft law nature of tribunal decisions, no one would say that they are irrelevant. The ICJ’s advisory opinion on Legality of the Threat or Use of Nuclear Weapons, for example, is widely cited as authoritative, despite the fact that as an advisory opinion it has no binding legal force. Where rule-making is concerned, tribunal decisions have their primary effect because they shape the expectations of states as to what counts as compliance with the underlying substantive obligations. States may not agree with tribunal decisions—they may not even be subject to the jurisdiction of the tribunal—but the decision’s existence forces states to respond to its holding. It burdens, in effect, future legal arguments about what the substantive rules requires.[33] Moreover, all states bound by the underlying substantive rules will have to react to the tribunal’s decisions when such decisions (or the arguments and interpretations contained therein) are invoked by other states.

Mandatory or implied jurisdiction would upset this balance. Because all states would appear before a tribunal and be potentially subject to the tribunal’s application of precedent in a subsequent dispute, the overall behavior of states before the tribunal would change. The tribunal would see more arguments favoring shallow interpretations of substantive commitments. The tribunal would also likely be warier of issuing decisions that states properly before it would simply ignore. Such challenges to the tribunal’s authority might undermine it going forward. Tribunals might also be inclined to issue much narrower, fact-bound rulings, as a way to avoid challenges to its legitimacy.

c) Non-judicial oversight mechanisms

The third risk of implying jurisdictional obligations is that it may crowd out non-judicial forms of dispute resolution and enforcement. These more administrative or sometimes legislative mechanisms are especially common in international environmental law. For example, the Meeting of the Parties (MOP) to the Montreal Protocol has established non-compliance procedures that are additional to dispute resolution procedures contained in Article 11 of the Vienna Convention for the Protection of the Ozone Layer.[34] Article 11 is essentially an optional jurisdictional obligation, permitting parties to accept, inter alia, the jurisdiction of the ICJ to resolve disputes.[35] The MOP created a much more detailed, and very different, enforcement procedure. States may express ‘reservations regarding another Party’s implementation of its obligations under the Protocol’ by notifying the Protocol’s secretariat and supplying corroborating information.[36] The other Party is then given the opportunity to respond.[37] Parties may also self-report compliance problems.[38] All of these reports are then forwarded to an Implementation Committee, which considers the reports and makes recommendations to the full MOP.[39] The MOP then decides on the appropriate course of action, choosing from the Indicative List of Measures that Might be Taken by Meeting of the Parties in Respect of Non-Compliance with the Protocol.

Similar, although not identical, enforcement regimes exist for the Kyoto Protocol, the Convention on the International Trade in Endangered Species, the Basel Convention on the Control of Transboundary Movements of Hazardous Wastes and their Disposal, among others. These enforcement regimes have a number of characteristics in common. First, they are non-judicial in the sense that they are administered directly by the Parties to the treaty, rather than by judges acting as individuals. Second, they are often adopted through decisions of the Parties that are properly termed soft law. Although compliance decisions by parties can result in penalties, such as the suspension of the right to trade pollution permits, neither the instruments that establish the review committees nor the enforcement decisions themselves are formally adopted as binding instruments or amendments to the governing treaty. Third, these enforcement mechanisms often have a much more robust set of remedies available to them than do ordinary tribunals. In addition to the ability to suspend a party’s privileges under the governing treaty, administrative enforcement mechanisms may also have the ability to take a more facilitative approach, offering financial or technical assistance to help a state boost its compliance capabilities. Fourth, because they are administered directly by the parties, resolutions to compliance disputes are more easily negotiated than in tribunals. In some instances, the ability to negotiate a resolution through an administrative decision making process can avoid the tensions associated with adversarial proceedings and can also allow compliance decisions to be tailored more readily to the circumstances, capabilities, and politics of a given circumstance. This tailoring, in turn, may increase the likelihood that a state will comply with the enforcement proceedings’ outcome.

Implied jurisdictional obligations could threaten the viability of these non-judicial enforcement techniques. By making judicial oversight optional but non-judicial, party administered oversight mandatory in regimes like the Montreal Protocol, states have established enforcement mechanisms that balance the costs and benefits of different modes of dispute resolution in a way that attracts participation and does not dilute substantive commitments. By increasing the costs of oversight, implying jurisdictional obligations could eliminate states’ willingness to also implement these alternative mechanisms. Moreover, substituting a more adversarial judicial form of oversight would risk the possibility that states as a whole would become less engaged with compliance issues. The existence of implementation committees allows states that may lack the capacity to comply to pressure wealthier states to provide assistance. Judicializing disputes puts the focus squarely on the state allegedly violating its obligations, relieving developed states from the diplomatic and political pressure to assist developing states in their compliance efforts.

4. Conclusion

International relations have ebbed and flowed in the extent to which they have been legalized. The recent increase in soft law, rather than binding treaties, as a vehicle to create substantive commitments is consistent with the same trend observable in regards to jurisdictional commitments. These ebbs and flows are a response to changes in political dynamics, both domestic and international, and the relative power of states. But the increasing hesitation of states to submit to judicial oversight of their legal obligations is not per se a violation of international law. Moreover, a move to impose jurisdictional obligations not grounded in state consent could backfire by deterring states from joining or remaining within treaty regimes, producing a less-cooperation oriented jurisprudence, and undermining the increasingly large number of non-judicial oversight mechanisms.

[1] For but a small slice of the literature examining the judicialization phenomenon, see C P Romano, ‘The Proliferation of International Judicial Bodies: The Pieces of the Puzzle’ (1999) 31 NYU J Intl Law and Politics 709; E Benvenisti and G W Downs, ‘Prospects for the Increased Independence of International Tribunals’ (2011) 7 German LJ 1057; K J Alter and L R Helfer, ‘Legitimacy and Lawmaking: A Tale of Three International Courts’ (2013) 14 Theoretical Inquiries in Law 479.

[2] Vienna Convention on the Law of Treaties (adopted 23 May 1969, entered into force 1 January 1980) 1155 UNTS331, art 26 [hereinafter ‘VCLT’].

[3] For example, Barbara Koremenos’s research indicates that roughly one half of international agreements contain no dispute resolution clause at all. B Koremenos, ‘If Only Half of International Agreements Have Dispute Resolution Provisions, Which Half Needs Explaining?’ (2007) 36 J of Legal Studies 189.

[4] Optional Protocol to the Vienna Convention on Consular Relations concerning the Compulsory Settlement of Disputes (adopted 24 April 1963, entered into force 19 March 1967) 596 UNTS 487.

[5] Convention on the Settlement of Investment Disputes between States and Nationals of Other States (adopted 18 March 1965, entered into force 14 October 1966) 575 UNTS 259, art 25(1) [hereinafter ‘ICSID Convention’].

[6] VCLT (n 2) art 54.

[7] VCLT (n 2) art 56.

[8] O Dorr and K Schmalenbach (eds), Vienna Convention on the Law of Treaties (Springer 2011) 431, 445-46. Sir Humphrey Waldock’s draft article on pacta sunt servanda actually contained this original provision. Although the International Law Commission removed this language, it believed it ‘clearly implicit in the obligation to perform the treaty in good faith.’ Ibid.

[9] ibid 446.

[10] ICJ Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States) [1986] ICJ Rep 14, para 272.

[11] ICJ Nuclear Tests (New Zealand v. France) [1974] ICJ Rep 457, para 49.

[12] See A T Guzman, How International Law Works: A Rational Choice Theory (OUP 2008) 71-117.

[13] ICJ Border and Transborder Armed Actions (Nicaragua v. Honduras) [1988] ICJ Rep 69, para 94.

[14] ICSID Convention (n 5) art 25(1).

[15] United Nations Convention on the Law of the Sea (adopted 10 December 1982, entered into force 16 November 1994) 1833 UNTS 3, art 298(5) [hereinafter ‘UNCLOS’].

[16] Marrakesh Agreement Establishing the World Trade Organization (adopted 15 April 1994, entered into force1 January 1995) 1867 UNTS 154, art II(2); UNCLOS (n 15) art 309.

[17] VCLT (n 2) art 31(3).

[18] T L Meyer and A T Guzman, ‘Customary International Law in the 21st Century’ in R A Miller and R M Bratspies (eds), Progress in International Law (Brill 2008) 213 (discussing implied limitations on North Korea’s right to withdraw from the Nuclear Nonproliferation Treaty).

[19] L R Helfer, ‘Overlegalizing Human Rights: International Relations Theory and the Commonwealth Carribean Backlash Against Human Rights’ (2002) 102 Columbia LR 1831, 1881 [hereinafter ‘Overlegalizing’].

[20] ibid.

[21] UNCHR, General Comment 24 (52) on Issues Relating to Reservations Made Upon Ratification or Accession to the Covenant or the Optional Protocols Thereto, or in Relation to Declarations Under Article 41 of the Covenant [1996] UN GAOR, Hum Rts Comm, 50th Sess, Supp No 40, UN Doc A/50/40, 119, para 11.

[22] Statement on the reservation of Pakistan to Article 40, Human Rights Committee, session 101st, 30 March 2011, available at <http://www2.ohchr.org/english/bodies/hrc/ docs/PakistanStatement.doc>.

[23] ibid.

[24] E T Swaine, ‘Reserving’ (2006) 31 Yale J. Intl L. 307, 309-312.

[25] Observations of the United States of America on General Comment 24, in J P Gardner (ed.), Human Rights as General Norms and a State’s Right to Opt Out: Reservations And Objections to Human Rights Conventions (BIICL 1997).

[26] A T Guzman, ‘The Cost of Credibility: Explaining Resistance to Interstate Dispute Resolution Mechanisms’ (2002) 31 J. Legal Studies 303 [hereinafter ‘Cost of Credibility’]; K Raustiala, ‘Form and Substance in International Agreements’ (2005) 99 AJIL 581.

[27] For a discussion of exceptions to this norm, seeL R Helfer, ‘Nonconsensual International Lawmaking’ [2008] University of Illinois LR 71.

[28] The costs may outweigh the benefits in part because sanctions are usually negative-sum in international law. That is, in many contexts the state found liable under international law suffers penalties, if only reputational, while the aggrieved party is not made whole in the form of compensatory damages. See Guzman, Cost of Credibility (n 24).

[29] Helfer, Overlegalizing (n 19) 1881-82.

[30] Raustiala (n 26).

[31] A T Guzman and T L Meyer, ‘International Soft Law’ (2010) 2 J of Legal Analysis 171.

[32] ibid. 202-03.

[33] H G Cohen, ‘International Law’s Erie Moment,’ (2013) 34 Michigan J Intl L 249, 276-279.

[34] Report of the Tenth Meeting of the Parties to the Montreal Protocol on Substances that Deplete the Ozone Layer [1998] UNEP/OzL.Pro.10/9, Annex II [hereinafter ‘MOP Report’].

[35] Montreal Protocol on Substances that Deplete the Ozone Layer (adopted 16 September 1987, entered into force 1 January 1989) 1522 UNTS 3, art 11(3).

[36] MOP Report (n. 32) Annex II(1).

[37] ibid, Annex II(2).

[38] ibid, Annex II(4).

[39] ibid, Annex II(9).