Edited by Gerrit De Geest
Chapter 15: International treaties
[In: Volume 7, Francesco Parisi (ed) Production of Legal Rules]
This chapter concerns itself with legal agreements which are treaties. A treaty is a formal agreement between multiple states, recognizing specific rights and obligations laid out within the context of the treaty; it is meant to define the relationship between the parties. More specifically, this chapter examines the formation of and reservations to modern treaties which are governed by the Articles established in the Vienna Convention on the Law of Treaties.
Human history has seen a long line of treaties. The oldest known written treaty, the Kadesh Peace Treaty, was established about 16 years after the battle in 1274 BC between the Hittites and the Egyptians. Over the last three millennia, untold numbers of treaties were formulated to consolidate cooperation, to resolve conflicts, to seal promises, and to coerce compliance across countries. The need for international agreements has increased substantially in the last few decades, as the world grew more integrated and interactions between states became ubiquitous and increasingly unavoidable. Witness the over 500 major multilateral instruments deposited with the Secretary-General of the United Nations (as of 1 January 2009) and over 158,000 treaties and related subsequent actions recorded in the United Nations Treaty Series (December 1946–December 2006) (United Nations Treaty Collection, Overview section). As treaties multiplied, it was recognized that codification of an overarching set of principles and rules to govern the conclusion, application, and interpretation of treaties was needed. Subsequently, international effort began in 1949, and the Vienna Convention on the Law of Treaties (hereinafter the Vienna Convention on Treaties) was signed May 23, 1969 and entered into force in January 1980. This convention is generally accepted as the authoritative codification of treaty law by the international community.
Recently, law and economics scholars have studied different forces that lead to the formation of international agreements. In general, international agreements can be sorted into different categories. Chief among these categories are treaties, customary international law, and non-legal international agreements. A treaty is an international agreement that is binding under international law. p. 273Customary international law encompasses consistent practices regularly adopted by states due to a sense of legal obligation. The third category of instruments of international cooperation consists of non-legal agreements between states, such as administrative agreements, letters of intent, memoranda of understanding, joint declarations, and statements or declarations of principles. Even a handshake between heads of states can be construed as a non-legal agreement. We will delve briefly into how treaties distinguish themselves from customary international law and non-legal international agreements.
Section 2 tallies potential drivers for international cooperative behavior, such as the need for coordination and the existence of transnational externalities. Although instruments other than treaties can specify international agreements, the discussion of factors that bring about international agreements will be framed in terms of treaties, as treaties are the center of attention. While many factors that lead to treaties could also lead to customary international law, Section 3 discusses differences between treaties and customary international law and contemplates the strength of treaties over customary international law. Likewise, Section 4 compares non-legal international agreements and treaties. It suggests reasons why treaties may be the preferred instrument adopted by states to form agreements. Section 5 discusses the modalities of formation of treaties, analyzes the state’s incentive problem in deciding whether to be a founding negotiating state or an acceding state, and underscores the relationship between the acceding protocol of the treaty and the treaty content.
Next we turn to the issue of reservations to a treaty. Section 6 discusses the ability of an applicant state to express reservations while becoming a party of the treaty, and the legal effects of reservations and of objections to reservations, according to the procedures set forth by the Vienna Convention on Treaties. It shows different needs for reservations and how successfully the Vienna Convention fulfills those competing needs. Section 7 examines the benefits and costs of being a reserving state versus a non-reserving state, to better understand whether the Vienna Convention on Treaties is truly neutral toward both kinds of states. Section 8 summarizes research discussed in previous sections.
A. Formation of International Treaties
2. Factors Conducive to Treaties and Other International Cooperation
Traditional international legal scholars have highlighted morality and legality to explain the existence of treaties and customary international law. Goldsmith and Posner (1999, 2005) adopted a different avenue of research by using rational choice theory in general and game theory in particular to analyze why international treaties and customary international law arise. Other law and p. 274economics researchers extended and expanded the economic forces conducive to international cooperation. These factors, along with some non-economic factors, are discussed below. Before that, it is important to discuss the basic assumptions behind the rational choice theory of international cooperation and the characteristics of game-theoretical research.
The fundamental assumption behind the rational choice approach is the existence of a preference of the state. That a state is motivated by its own well-defined self-interest is both questionable and defensible, but it is taken here as a given. Further, it is assumed that a state only cares about its own interests and not the welfare of other nations. The preference of the state then provides a clear picture of the benefits and costs of international cooperation. Other practical constraints, such as the value of resources and technological know-how, determine the pecuniary portion of the costs of fulfilling an international obligation. Along this line, the cost of consenting to be bound by a treaty depends not only on the costs of compliance but also on the likelihood of those costs being realized (Hathaway, 2003). Lastly, the transaction costs, including negotiation costs, of international agreements need to be considered as well (Fon and Parisi, 2007). The transaction costs ought to be broad enough to include the monetary costs of reaching the agreements and the political costs of accepting and maintaining the agreements (Sykes, 2007). Thus, when states gain more than they lose through cooperation, they contemplate treaties. With benefits exceeding costs, nations are motivated to pursue their own interests while accounting for other states’ behavior and responses. Taking into account other nations’ reactions given one’s own action is the strategic framework that game theory employs and that many researchers use to understand international agreements.
Taking nations as strategic players who would sign a treaty when the net private benefits from treaty participation are positive, the following considerations may influence the formation of treaties. Although many of these considerations may serve equally well to promote customary international law, the forces conducive to international cooperation will be framed and discussed in terms of treaties. The driving forces toward international cooperation include: coincidence of interest, coercion, cooperation, coordination, externalities, public goods, institutional designs, and morality-induced private gains.
Coincidence of Interest
When the interests of states align themselves and the same action is the best option available for each state regardless of the choices made by other states, states would find themselves adopting the same action, or cooperating, whether or not there is a treaty. Given that each nation would engage in the same action, there is no apparent reason for nations to form a treaty. However, it is possible that treaties may strengthen the cooperative relation between states even though p. 275there is a coincidence of interest. There may be underlying motives that inspire nations to come to an explicit legal agreement. One example is the Treaty of Moscow, signed in May 2002, in which Russia and the United States agreed to reduce their nuclear warheads to a certain level by 2012 (Goldsmith and Posner, 2005, p. 89). Although each nation finds the reduction of its nuclear arsenal to be its best option, without a treaty each country may be tempted to reduce its warheads more slowly to maintain its nuclear advantage, knowing that the other state will reduce its nuclear stockpile. A treaty increases the sense of security for both nations and the likelihood that the other party will do the right thing.
A powerful state or a group of powerful states with similar interests may force other states to engage in acts that they would not adopt absent pressure from the powerful, such as agreeing to a treaty (Goldsmith and Posner, 2005). Coercion exists when the powerful state’s net gain depends on the weaker state’s action and when the stronger state can punish the weaker state for a deviation that does not suit the powerful state. This is especially true if the cost of punishment incurred by the coercing state is relatively small and thus the threat of punishment is credible, and the punishment imposed on the coerced state is harsh. However, similar to coincidence of interests, coercion does not necessarily imply that a treaty is necessary, although a treaty may bring other advantages. When a treaty is formed and compliance with treaty obligations is practiced, the coerced state can have peace of mind, having fully understood all the demands from the coercing state, and is freed from other harassment or punishment. Meanwhile, when the rights and expectations are clearly specified and accepted, the coercing state can expect less resistance from the coerced and perhaps receive more respect in the international arena.
Sometimes, two states find themselves in a Prisoner’s Dilemma situation in which both gain if they cooperate, but individually each has the incentive not to cooperate. When this situation is repeated over time, game theorists have shown that parties will find it beneficial to cooperate repeatedly under some circumstances. Specifically, suppose that the following conditions hold: the parties clearly understand what counts as cooperation and as cheating; they care enough about the future (discount rates are sufficiently low); the situation repeats itself indefinitely and the states expect it never to end; and gains from defection are not too high relative to gains from cooperation. Then incentives to cheat in the short run can be overcome, and a strategy of cooperating indefinitely unless the other party defects will bring about indefinite cooperation. Goldsmith and Posner (1999, 2005) maintain that this situation provides solid ground for the states to enter into a bilateral treaty. Sandler (2008) extended the potential p. 276cooperative nature through indefinitely repeated situations beyond the Prisoner’s Dilemma, expanded the grounds that would foster dynamic cooperation, and provided a foundation for multilateral treaties.
The case of coordination is similar to the case of coincidence of interest in that the interests of the states converge. The difference is that, unlike coincidence of interest, under coordination each state’s best action depends on the action of the other state as well. Typically, with the convergence and interdependence of interests of the states, there is more than one collaborative solution and the problem is one of coordination. The use of treaty as an instrument of agreement will bring focus to the collaborative solution acceptable to all states involved. As Goldsmith and Posner (2005) point out, there may be difficulties coordinating multiple states, and thus bilateral treaties are more likely. Sandler (2008) extended the arguments to show that multilateral treaties are quite possible for states that are similar (homogeneous). However, a minimal number of collaborating states are often required for a treaty to have an impact (e.g. eliminating terrorist safe havens), and the probabilities of states’ adherence to the treaty terms are highly influential toward the success or failure of multilateral treaties. Specifically, the smaller the minimal number of collaborating nations that is required and the more likely it is states believe that other states will adhere to their treaty obligations, the likelier it is that coordination will prove successful and lead to a treaty.
In an increasingly globalized world, actions of one state can create transnational external effects beyond their borders. This engenders a need for states to cooperate in order to find ways to internalize these externalities. International agreements could enable states to adopt actions that bring the equilibrium with externalities closer to the Pareto frontier (Sykes, 2007). That is, if gains are to be had by some states without hurting other states’ welfare, international agreements induced by externalities could bring about these superior outcomes. Externalities provide an opportunity for states to cooperate and improve their welfare as long as transaction costs to reach international agreements are sufficiently low. As an example, when individual countries unilaterally impose tariffs on their imports, they alter the terms of trade to their advantage. These unilateral trade policies impose externalities on other states: the costs of these trade interventions are borne by exporting states. These externalities created by unilateral action of individual states then become engines of international cooperative trade policy. International agreements could diminish the terms of trade externalities across countries, while reducing tariffs and improving states’ welfare in the process.
p. 277Transnational Public Goods
In an international context, a public good is one where the consumption of a good by one state cannot diminish the consumption of the same good by another state (non-rivalry), and no state can be excluded from the benefit (or cost) of the good (non-excludability). The existence of a public good provides important incentives for states to form treaties (Sandler, 2008). While a transnational public good (or public bad) creates spillover benefits (or costs) for other countries, international treaties can help bring about joint efforts that produce benefits (or curtail costs and damages). The North Atlantic Treaty Organization, an alliance to provide common defense, is an international cooperative example of a transnational public good, and the Helsinki Protocol on sulfur emissions to control pollutants is an example of a transnational public bad.
A treaty can be framed to motivate nations to come to an agreement and to fulfill obligations without the need for enforcement. When designed properly, the treaty innovation can create the right incentive for states to commit to the treaty. By using principles that enhance self-enforcing behavior in which adherence to treaty terms is in the states’ interests, treaties become an important instrument for guaranteeing international cooperation (Sandler, 2008). Special innovation in the design of treaty rules can often convert reluctant states into willing participants. One example of special institutional innovation involves international joint peacekeeping, in which leadership has been borne by those countries closest to the conflict, as in the case of the United States for Haiti. Another example concerns international cooperative effort to fight diseases and pests, where a minimum threshold of effort must be reached to be effective. A properly designed treaty might result in rich countries and neighboring states of the problematic areas as the main contributors. In our third example, the International Monetary Fund (IMF) provides incentives for generosity; countries with higher cost shares (“quotas”) are rewarded with more votes in IMF decisions. Further, a treaty can also be designed to reallocate jurisdiction (Trachtman, 2008). For example, the European Union (EU) was established to trade jurisdiction from Member States of the EU to the EU Commission.
Morality, Conformity, and Other Factors
Rational choice theory can be augmented by morality, legality, justice, and conformity, the important promoters for international cooperation highlighted by traditional international legal scholars. Including morality- and conformityinduced private gains in the strategic framework strengthens the rational choice analysis of the formation of international agreements. For example, if citizens of a nation perceive that membership in a club of states who value clean air has many social, moral, and economic benefits, it would increase and fortify p. 278the state’s interest in joining an environmental treaty (Sandler, 2008). It is quite possible that the Montreal Protocol on Substances that Deplete the Ozone Layer grew from a few nations to most nations because of the increasing recognition of the importance of all these forces in countries around the world.
Besides the many forces already discussed, other factors can enhance the likelihood of treaty formation. One factor that can sway treaty participation and commitment is the manner in which individual states’ contributions determine the overall level of public good consumption. In the case of weakest-link public goods, the least stringent efforts to monitor a deadly disease such as bird flu can determine the safety of all at-risk nations (Sandler, 2008). In order to compensate for the potential lack of effort, perhaps due to financial constraints of some nations, the rich countries may decide to shore up the weakest link by providing income transfers or to simply pick up the missing contributions. Treaties then become the instrument to formalize the shoring-up process. Further, economies of scale to pursue common goals and network effects can also be important promoters of international cooperation (Fon and Parisi, 2007).
Using treaties as signaling devices (Martin, 2005) is the last driver that we shall mention. Choosing to sign a treaty over more informal means of agreement may signal the degree of commitment of states to the cause, both to their domestic constituents and worldwide, especially if the cost of non-compliance is large and severe. Before the Treaty of Moscow was made, President Bush proposed a handshake with President Putin to seal the deal on nuclear warheads reduction. Putin and the US Congress objected, insisting on having the treaty ratified. Thus, key parties in the two treaty signatories wanted to highlight the importance of the roles they play, as well as to strengthen and signal their strong commitment to the legalized treaty (Goldsmith and Posner, 2005). This last example brings out the issue of treaties versus other instruments such as non-legal international agreements and customary international law. To these comparisons we now turn.
3. Customary International Law and Treaties
Treaties and customary international law (henceforth called CIL) are both instruments of international cooperation. As characterized in Restatement (Third) of the Foreign Relations Law of the United States, 1987, §102(2), CIL “results from a general and consistent practice of states followed by them from a sense of legal obligation.” That is, CIL is conventionally characterized by a regular, widespread, and convergent state practice and a belief that the practice has taken on the role of a legal obligation. The latter requirement, the recognition and acceptance that adherence to the widespread practice is an international obligation, is known as opinio juris, the central concept in CIL.
Rather than the doctrine of opinio juris recognized under CIL, the doctrine of pacta sunt servanda holds under treaties. As specified in Article 26 of the p. 279Vienna Convention on Treaties, the pacta sunt servanda states that “Every treaty in force is binding upon the parties to it and must be performed by them in good faith.” Thus, every treaty party has an absolute obligation to comply with the treaty terms. Clearly, the pacta sunt servanda doctrine imposes a more forceful and definitive requirement on treaty signatories, whereas the opinio juris doctrine only provides guidance to a CIL practitioner state that believes the practice in question has reached the status of a legal obligation. No clear and solid guidelines determine when a common regular practice of states has been adopted widely enough to reach this legal status.
Why then should rational states follow CIL practices? Other than the sense and recognition of legal obligations, the economic drivers conducive to treaties enumerated in the previous section (coincidence of interest, coercion, cooperation, and coordination) could also lead to CIL (Goldsmith and Posner, 1999 and 2005; and Sykes, 2007). Convergence in state practice takes place because of the self interest of states, and self interest could give rise to continued adherence to CIL practices over time. Further, concern for reputation may be at work here (Guzman, 2002 and 2008). States may continue to cooperate on some issues to develop or maintain a good reputation for future negotiations; and reputational considerations may sustain regular state practices over time. Thus, not only do nations believe in their own legal obligations over some regular state practices, they feel that other nations should commit to their legal obligations as well. Defecting from CIL obligations will then lessen the reputation of a state.
CIL is and will remain an important part of international law. Many fundamental principles of international law are governed by CIL (for example, territorial sovereignty, diplomatic immunity, and sovereign equality). The strength of CIL is that it does not require large transaction costs for negotiation, administration, enforcement, and institutional innovations. Behavioral regularities that have become CIL can be sustained by the states’ sense of legal obligations and the desire to maintain their reputations, without incurring monitoring and enforcement expenses. Even in areas where treaties are abundant, CIL may provide interpretive presumptions, extend treaty norms beyond a group of signatories, and induce efforts to expand treaty regimes (Goldsmith and Posner, 2005). Generally, states may avoid an obligation by objecting at the early stage of an emerging practice, but once they have demonstrated their agreement through conforming behavior, deviation is no longer possible. Not surprisingly, CIL works better and is more useful when states have homogeneous preferences and similar constraints (Fon and Parisi, 2009b).
While CIL offers considerable strength for international cooperation, this strength can turn into weakness under different circumstances; treaties then become the more desirable instrument for international agreements. As intimated earlier, it is generally difficult to decide whether a certain regular practice of states has been consistent and should be considered a legal obligation. Further, p. 280a sense of legitimacy, conformity, and justice may be fleeting. States can and often do change their views on the content of CIL, even in a short period of time. The uncertainty of the status of CIL then makes the use of treaties, where every obligation is spelled out, more attractive.
Goldsmith and Posner (2005) presented various case studies to document how CIL rules are often violated when it is in the interest of a state to deviate. A salient example is provided by the frequency with which rogue states violate the rules of diplomatic immunity. Because of the short time horizon and high discount rate for future welfare, rulers of rogue states value short-term gains much more than long-terms benefits. Thus, we observed the Iran hostage crisis, the 1967 attack on the British embassy by followers of the Chinese Cultural Revolution, and the 1958 Iraqi military coup that burned the British embassy. Goldsmith and Posner also pointed out that states tend to limit CIL to practices that are acceptable to them from their normative perspective, while besmirching CIL practices incompatible with their normative or cultural beliefs.
Various other reasons make treaties the more appropriate instrument for international agreements. Sykes (2007) reminded us that although reputation considerations may help maintain CIL practices, they may not be practical in many settings. Current events, domestic or international, may make the benefit of maintaining a reputation for compliance insignificant, especially when the CIL practice lacks wide support for being considered a legal obligation. Complicated problems encompassing a broad array of issues which require much negotiation, monitoring, and enforcement among many nations also tip the scale to make treaties more valuable. Externalities are yet another factor that can be handled more easily under treaty negotiation than under CIL. Likewise, Sandler (2008) reasoned that treaties can deal with the public goods cooperation problem better because institutional designs can be adopted to take care of special public goods production problems. Finally, treaties can provide a better infrastructure, well-recognized signatories, and clearly defined obligations.
All these pros and cons of employing treaties or CIL as instruments of international agreements make both instruments indispensable. Both treaties and CIL, and also non-legal international pledges, reside in the muddy grounds of international diplomacy. It is interesting to note that in spite of the importance of the Vienna Convention on Treaties and the fact that the United States is a signatory to many treaties, as of 2010, the United States never ratified the Vienna Convention. In responding negatively to the question whether the United States is a party to the Vienna Convention, the Treaty Affairs, Frequently Asked Questions section in the US Department of State website states, “The United States considers many of the provisions of the Vienna Convention on the Law of Treaties to constitute customary international law on the law of treaties.” The ground is very muddy indeed.
4.4 Non-Legal International Agreements and Treaties
p. 281Other than treaties and CIL, states often use non-legal agreements as instruments for international cooperation. Although some scholars apply the term “soft law” to treaties with weak obligations, “soft law” often represents non-legal agreements between states. Likewise, other scholars prefer the term “pledges” to mean non-legal agreements. In this section, pledges and non-legal agreements mean the same thing. Sometimes, it is difficult to decipher whether an unfamiliar international agreement is a treaty or a pledge. For example, the 2001 Convention on International Interests in Mobile Equipment is a treaty, while the 1988 Basle Capital Accord, the agreement to promote international convergence of capital measurement and capital standards, is considered a pledge.
Why do states use non-legal international agreements? Two prominent explanations – flexibility and domestic law and politics issues – have been advanced in the literature (Guzman, 2005). States use non-legal agreements instead of treaties because non-legal agreements are less binding on states and they give states more flexibility. If necessary, states can more easily deviate from the terms, renege, or renegotiate the terms of agreements. The domestic processes for approving international agreements matter quite a bit. For example, in the United States, a treaty requires a two-thirds super-majority vote in the Senate but no consideration from the House, a congressional-executive agreement requires a simple majority vote in both the Senate and the House and approval by the President, and a sole executive agreement only requires consent from the President. Since a small minority of senators can prevent a super-majority consent, treaties in the United States are politically vulnerable, more cumbersome, and often embed less reliable legal commitments than the alternatives. Although seemingly less dignified than treaties, congressional-executive agreements and sole executive agreements have been on the rise in the United States. Thus, Hathaway (2008) advocates replacing most treaties by congressional-executive agreements in the United States.
Other important considerations in the choice between treaties and non-legal agreements, such as form and substance, may elevate treaties to be the preferred instrument of agreement (Raustiala, 2005). Two features of agreements relate to form: legality and structure. Legality refers to the choice between the accord being legally binding or non-binding, and structure concerns the specific provisions for monitoring and sanctioning non-compliance. The other important consideration of agreements relates to substance, which addresses the issues of substantive provisions and their deviations from the status quo. These features are not unrelated and there are trade-offs among them. The choice between treaties and pledges depends on the trade-off between form and substance. For example, if compliance with pledges is highly problematic, then legality needs to be the salient feature of the agreement and treaties are preferred. When states are uncertain of the distribution of future cooperative benefits, they tend to p. 282prefer substance over form, and pledges are the preferred instrument of their agreements. Similarly, when states do not understand the underlying problems well, they cannot estimate compliance costs for the legal obligations (Abbott and Snidal, 2000). In these cases, pledges are preferred, allowing states to take more risk under uncertainty.
As in the case of treaties versus CIL, treaties are not always superior to pledges. However, a treaty generally embeds a higher level of commitment and increases the credibility of commitment relative to pledges. Treaties may be employed to reveal information about a state’s strong and lasting commitment to the agreement. Treaties should be formed if states want to stress their intent to have binding agreements, thereby inducing more impact on other states’ behavior. Formal treaties are subject to the default rules of the Vienna Convention on Treaties, reducing uncertainty on interpretations of agreement terms. Lastly, treaties are more likely than non-legal international agreements to enjoy higher levels of domestic support and to withstand shocks in political arenas, and are more durable following changes in administrations (Guzman, 2005; Goldsmith and Posner, 2005; and Sykes, 2007).
5. Modalities of Treaty Formation
When a group of states considers it beneficial to form a treaty and start the process of drafting the treaty, besides agreeing on the treaty content, the states must choose from among the alternative modalities with which they can organize their treaty. Specifically, the negotiating states – those states which take part in drawing up and adopting the treaty text – need to decide whether and in what manner membership of the treaty can be expanded. Article 15 of the Vienna Convention on the Law of Treaties states that, “The consent of a State to be bound by a treaty is expressed by accession when: (a) the treaty provides that such consent may be expressed by that State by means of accession; (b) it is otherwise established that the negotiating States were agreed that such consent may be expressed by that State by means of accession; or (c) all the parties have subsequently agreed that such consent may be expressed by that State by means of accession.” This Article implicitly requires prior or subsequent consent by the signatory states for an applicant state’s accession to an existing treaty, and explicitly lists the three modalities that govern the manner in which a state may subsequently join the treaty agreement.
Fon and Parisi (2007) denote the three approaches specified in the Vienna Convention on Treaties as open, semi-open, and closed treaties, and analyze the different structures and incentive issues confronting states. In open treaties, the negotiating states agree to leave the treaty open for accession. Semi-open treaties accept acceding states through approval by a majority of the existing member states. Closed treaties require unanimous consent of parties to the p. 283treaty – states which have consented to be bound by the treaty and for which the treaty is in force.
Open treaties do not require affirmative action by the original negotiating states. Any state is welcome to express its consent to be bound by the treaty by accession. By formulating the conditions and prerequisites for accession, the negotiating states avoid the necessity to formally renegotiate the treaty, while permitting expansion to states with similar preferences that are more likely to accede to the treaty. Open accession clauses are common in multilateral treaties, particularly those intending to promote cooperation and foster dispute resolution across states. The Vienna Convention on Diplomatic Relations (1961) is an example.
Closed treaties disallow automatic accession; acceptance of a new member to a closed treaty requires unanimous approval by the current party states. A closed treaty typically focuses on narrow issues and provides exclusivity. For example, in an effort to limit the number of states that hold special privileges with regard to nuclear weapons, the Treaty on the Non-Proliferation of Nuclear Weapons, signed in 1968, extended the special privileges only to states that manufactured nuclear explosive weapons prior to 1967. Most bilateral treaties, especially bilateral investment treaties, are closed since they concern a specific relationship between two states.
The intermediate case of semi-open treaties consists of treaties in which acceptance of a new member requires majority approval of the signatory states. Semi-open treaties leave greater political discretion to current member states. At the same time, expansion of treaty participation means that the treaty content can be modified by a majority of the member states, brought about by accession of new states. The revised treaty content may differ in important ways from the treaty content with which the original negotiating states agreed to be bound. This creates a risk for the treaty’s negotiating states. The 1974 Agreement on an International Energy Program, which promotes secure acquisition of oil, is an example of a semi-open treaty. In order to maintain a balance of power among the original negotiating states, the treaty requires an acceding state to gain approval of a majority of the Governing Board, which is composed of representatives from the founding states.
States can become part of an international treaty in two ways: as a founding negotiating member of a treaty or by acceding to an existing treaty. In light of the possibility of joining an existing treaty through accession, a state should consider the benefits and costs of being a founding member versus being an acceding state. A founding member of the treaty can strongly influence treaty content and set rules to control accession. But the negotiating and drafting costs can be substantial and have to be paid upfront, while benefits will be delayed. If being an acceding member of the treaty is possible, the costs of acceding to an existing treaty are typically lower, costs of compliance with treaty terms are p. 284less risky, and benefits are more immediate. However, while an acceding state can make reservations to certain treaty obligations in the process of ratification, thereby altering the effect of the treaty through its interaction with states who accept the reservations, its impact on treaty content is limited. The next two sections will turn to this issue.
In their study, Fon and Parisi unveil an important relationship between the formation of the treaty and its substantive content. When a treaty is left open and accession of other states is expected, treaty content is set not only according to existing interests of the negotiating states, but also on the basis of their expectation of treaty enlargement. Hence treaty content may differ from that which would be chosen under other treaty forms or if no expansion was expected. Lastly, it is observed that homogeneous states are most likely to form an international treaty to cooperate; they surmise that the treaty would likely be open or semi-open. Any state willing to join the open treaty and accept the treaty content as specified is likely to be similar to the founding negotiating states. Likewise, acceptance by a majority of signatory states of the accession of a new state to a semi-open treaty means that similar interests in all these states will be served. In the heterogeneous case with strong and weak states, the treaty is likely to be closed, since the more powerful states would refuse to give away advantageous stakes from the negotiation.
The default rule dictates that a treaty is closed for accession, unless specified otherwise (Bishop, 1971, p. 119). Since 1966, the International Law Commission has unsuccessfully advocated changing the default rule to make all multilateral treaties open for accession unless otherwise stated. The prevailing default rule favoring closed-form treaties may be justified. States face incentive problems when confronted with treaty participation. Since treaty negotiation is costly, leaving all multilateral treaties open for accession by default could undermine states’ incentives to invest in the initial negotiation and drafting of treaty agreements. Further, treaties with open accession clauses cannot ensure that the rights and obligations of the negotiating states remain undisturbed (Fon and Parisi, 2007; Stark, 1989, pp. 458–9).
B. Treaty Reservations
Under the CIL of sovereign equality, no state can be forced to consent to be bound by any specific treaty term, let alone the entire treaty. When widespread participation in a treaty is considered a high priority, there is a need to reconcile minor disagreements. In order to promote greater international cooperation, meet the need for greater flexibility brought about by the increasing number of multilateral treaties, and accommodate differences in opinions and interests concerning a treaty, the Vienna Convention on Treaties allows an applicant state p. 285to make reservations against some treaty provisions when signing, ratifying, or acceding to the treaty, unless specified otherwise. In particular, a state that expresses its consent to be bound by the treaty in general is free to make reservations on selected provisions of the treaty, as long as the reservations are not prohibited by the treaty and are not incompatible with the object and purpose of the treaty (Article 19).
If the provisions for which a reservation is made are deemed essential for the application of the entire treaty to uphold the object and purpose of the treaty, Article 20(2) requires unanimous acceptance of the reservation by all parties to the treaty. Otherwise, individual treaty participants decide whether the reservation is acceptable. Hence, while a state can propose a reservation to a treaty, the reservation must meet with the approval of another non-reserving state for the relevant provisions to apply between the two states. If a non-reserving state accepts, or at least raises no objection to, the reserving state’s reservation, the relevant treaty provisions are modified for both of them under Article 21. However, the reservation does not modify the relevant treaty provisions for other parties to the treaty inter se. If a non-reserving state objects to the reservation within 12 months, then the provisions to which the reservation relates do not apply between the two states.
Thus, in order to promote international cooperation, to ease frictions arising from some treaty provisions, and to advance consensual relations between parties to a treaty, reservations are used to inject greater flexibility into treaties. This is necessary; otherwise, a state in favor of the treaty content in general but finding some specific treaty terms unacceptable will have no course of action other than not participating in the treaty. The Vienna Convention on Treaties provides an intermediate ground that allows states to include reservations along with their acceptance of other treaty obligations. However, if the unilateral reservations from states are not constrained, the essence and cohesiveness of multilateral treaties may be corroded and the net benefits of treaty participation for signatory states may be reduced. To this end, a reservation must be acceptable to the non-reserving state before the provisions can be legally modified. A treaty modified by the reservation enters into force only between the reserving state and those states that do not object to the reservation. Conversely, states that object to the reservation and the reserving state are not bound by the reserving provisions in their mutual relations. Further, other parties to the treaty remain bound to the original treaty provisions in their relations with one another.
International law lacks an overarching enforcer of the rules. To infuse multilateral treaties with flexibility, reservations are allowed when states agree to accept the treaty obligations. To ensure integrity of treaties, reservations must be accepted, explicitly or tacitly, before the relevant treaty provisions come into force between the reserving and non-reserving states. The Vienna Convention on Treaties achieves these goals through the principle of reciprocal p. 286reservations. If a state introduces a reservation, the treaty relation between the state and a non-reserving state is modified according to the scope of the reservation: the exception or limitation claimed by the reserving state applies equally to both states. This effectively introduces an identical reservation against the reserving state.
Parisi and Ševcenko (2003) investigate the tension between treaty flexibility and integrity and illustrate that the principle of reciprocity, or the matching-reservation constraint, may ease the tension. A state seeking to exempt itself from a treaty obligation must permit other nations to escape the same burden. Thus, when states are homogeneous with symmetric incentives, they have no reason to attach merely strategic reservations to the treaty during negotiations. States know that the matching-reservations mechanism will make their sought-after advantage automatically available to others. A strategic reservation matched by others may occasion mutual losses for all states involved. States then refrain from introducing strategic unilateral reservations to maximize their expected returns from the treaty relationship; optimal treaty obligations are likely included in the original treaty agreement. That is, in settings where states have symmetric incentives, the matching-reservations constraint set forth by Article 21 leads to socially optimal levels of treaty ratification and participation.
When states face asymmetric incentives, however, recent game-theoretic models raise doubt about the effectiveness of matching constraints, such as those set forth by Article 21 (Fon and Parisi, 2003). Matching constraints positively influence cooperation behavior, but with heterogeneous parties they cannot generate a global optimum unless additional benefits to cooperation are very high. Thus, when asymmetric states have different benefits and compliance costs from the treaty, the dominant ratification strategy is to attach unilateral reservations that preserve national interests in spite of the matching reservations. Rules introduced by the Vienna Convention cannot discourage all reservations; erosion of the original treaty content may be inevitable, unless negotiating states preclude reservations in the treaty (Fon and Parisi, 2009a).
Fon and Parisi (2008) argue that due to the reciprocity principle, reservations under the Vienna Convention on Treaties create a bias when states are not homogeneous. Specifically, consider a state that is relatively high-cost and low-benefit in terms of issues covered in the treaty. If such a state wishes to avoid application or to reduce the content of a treaty, it can do so via unilateral reservations. But no similar opportunity is afforded to states that desire to add provisions or to increase treaty obligations. Thus, given the opportunity to introduce reservations, Article 21 tilts the balance in favor of high-cost, low-benefit states that can take advantage of the reciprocity mechanism. States with a comparative disadvantage in treaty implementation have a systematic advantage under the Vienna Convention on Treaties. This bias, introduced by p. 287the reciprocity principle in treaty reservations, has potential distributive effects among different groups of states.
7. Reserving versus Non-Reserving States
The Vienna Convention on Treaties seemingly endows equal rights to reserving and non-reserving states. If one state is given the right to formulate a reservation, then the other party ought to have the right to object to the reservation. Article 21 ensures such a balance. Does this mean that the reservation paradigm under the Vienna Convention on Treaties is truly impartial in its treatment toward reserving and non-reserving states?
There is some disagreement on the answer to the question, although it is widely considered that reserving states are favored. For example, Greig (1994) criticizes the Vienna Convention on Treaties for providing an unacceptable advantage to reserving states. Klabbers (2000) advances that whether or not a state accepts the other state’s reservation, the reserving state generally gets what it wants. Fon and Parisi (2008) bring out the fact that the Vienna Convention on Treaties favors states with comparative disadvantages in treaty implementation that would turn into reserving states. Meanwhile, Swaine (2006) disagrees that the Vienna Convention on Treaties disfavors the non-reserving states. Helfer (2006) extends Swaine’s thesis, providing more reasons why non-reserving states may not be disadvantaged after all. The researchers stress that, while there are gains to making reservations, a reserving state must bear some costs as well. Further, benefits can be derived by non-reserving states. Consequently, the Vienna Convention on Treaties may plausibly serve the interests of non-reserving states more than reserving states.
It is recognized that the Vienna Convention on Treaties carries various ambiguities. For example, a reservation can be made only if it is not incompatible with the treaty’s object and purpose. But it is not clear how, when, and in what way the incompatibility kicks in. If a non-reserving state is uncertain whether the reservation is considered compatible with the treaty’s object and purpose, it may fail to raise its objection within the required 12 month period. In general, any limbo resulting from uncertainty over the reservation process benefits the reserving state. To better understand the controversy, it is also important to consider the incentive problem facing a reserving state. After all, it is the pursuit of the state’s interest that inspires the state to consider participating in a treaty while making reservations that incur unacceptable costs. As discussed in the last section, reserving states also have strategic benefits (Parisi and Ševcenko, 2003; Fon and Parisi, 2008; Fon and Parisi, 2009a).
Swaine (2006) highlights the informational value generated from negotiating, concluding, and administering treaties. In fact, many treaties are oriented toward information production, especially if more worthy and demanding goals are beyond reach. Human rights treaties are a good example. In the international p. 288arena, states often hide their motives. But information is useful because it may help influence future behavior, build consensuses, and form other treaties. Sometimes information on other states is very difficult to generate; states do not wish to disclose embarrassing information about themselves, create disagreements, however minor, with other states, or risk retaliation. That is to say, making a reservation has an informational cost. Making a reservation to a treaty signals that the state is willing to cooperate, but only on its own terms and according to its own schedule. Helfer (2006) further points out that the informational signal is stronger the earlier the state makes the reservation, although reputation costs to early reserving states can be substantial. But information goes both ways. Reserving states may learn something from the reactions of other states. Further, non-reserving states may withhold their objections to early reservations for fear that doing so may cause other states not to ratify or to participate in the treaty. These are among the important benefits and costs that a potential reserving state must juggle before making a decision.
It is universally agreed that the international reservations regime contributes to broader participation: reservations lead to more states agreeing to be bound by a treaty. Treaty participants, including non-reserving states, may gain from broader participation whether they accept or object to specific reservations. With more participants to the treaty, each party state may acquire positive externalities and other benefits, such as lower compliance costs occasioned by the additional participation of the reserving states. Besides the benefits of breadth (more participation), the international reservations regime can also increase the depth of treaty participation. Given the opportunity to make reservations and to be exempt from certain provisions, treaty terms may be more demanding and therefore generate more benefits for participating states, even after some reservations are made and accepted. These benefits are available to all treaty participants, reserving and non-reserving states alike.
Non-reserving states can gain from information: the informational costs from reservations incurred by a reserving state also represent informational benefits to non-reserving states. Since the act of reserving reveals valuable information about a state’s reputation and its propensity to comply, non-reserving states gain knowledge about the risks and benefits of contracting with reserving states. Non-reserving states can build on this information, reserve their judgment as to whether and how the reservations should be objected, and try to find ways to shift risks in their favor.
The number of reservations attached to international treaties is considered relatively low despite the rules governing reservations set forth in the Vienna Convention on Treaties. For example, Gamble (1980) found no reservations in 85 percent of all multilateral treaties that allowed reservations; only 61 treaties had more than three reservations. Given the competing benefits and costs of reserving, perhaps this is not as surprising as it may first sound.
p. 289In their pursuit of self-interest, states often need to interact with other states. International law and economics scholars reveal many factors that may secure the formation of treaties and other international agreements. Among the important factors are coincidences of interest, coercion, repeated cooperation over time, coordination to align states’ actions, external effects induced by one state’s action on other states, the transnational public goods problem which may require joint effort to produce benefits or to curtail costs affecting multiple states, and the treaty design, which if framed properly can create the right incentives for states to cooperate.
In the international arena, while agreements can function as treaties, customary international law, and other informal agreements, treaties are sometimes preferred over other instruments. Relative to customary international law, which are behavioral regularities that states consider they have a legal obligation to adhere to, treaties are more desirable when, for example, there is uncertainty over the legality of the states’ practices, where lengthy negotiations are involved, externality problems are prevalent, or a more organized infrastructure for interstate agreements is required. Relative to informal international agreements, sometimes referred to as pledges, treaties are superior when, for example, there is little uncertainty relating to the benefits and compliance costs of the agreement, when states want to signal their strong commitment and increase their credibility, or domestic constituents need to be placated through more formal agreements.
According to the Vienna Convention on the Law of Treaties, there are three modalities in which treaties can be formed. Negotiating states may agree to leave the treaty open for accession, acceptance of a new party to the treaty is made possible through approval by a majority of existing member states, or acceptance of a new party requires the unanimous consent of the parties to the treaty. Each protocol has its own strength; in general, the form of the treaty and the substantive content are related. For example, if unanimous approval of accession is required from existing parties, treaty obligations may be made more stringent, making it more manageable to keep treaty obligations intact to achieve the object and goal of the treaty.
To facilitate broad participation in international treaties, the Vienna Convention on Treaties extends the possibility of reservation to treaty provisions to a state during its signing of or accession to the treaty. While the acceding state may propose a reservation to the treaty, the Vienna Convention dictates that a non-reserving state must accept the reservation, or else the two states are not bound by the provisions related to the reservation. Thus, the reservations regime under the Vienna Convention attempts to balance the need to respect the conditional consent of reserving states and to honor the unqualified consent of p. 290the treaty by the non-reserving states; it bridges the gap between the differing desires of heterogeneous states and the need for consensus.
Unfortunately, much ambiguity remains in the legal reservations regime under the Vienna Convention. It is arguable that the uncertainties induced by the ambiguities, the possible strategic moves, and the freedom to make reservations mean that the Vienna Convention favors the reserving states. But the information content generated by the reservation process may be invaluable to non-reserving states as well. Much controversy remains concerning which parties the Vienna Convention favors.
Even with millennia of bargaining experience to establish interstate agreements, states still often find it difficult to come to an agreement. It is not surprising that many states attempt to bargain around the Vienna Convention on Treaties. For example, an unforeseen increase in compliance costs may inspire a state to denounce the treaty and to re-accede the same treaty with reservations. As Helfer (2008) notes, grounding treaties on formal consent has many advantages. But consent has significant costs as well, including the inability to ensure that all states affected by the transnational problem join treaties that endeavor to alleviate those problems. In light of the many problems addressed, additional analyses of treaties are warranted and much needed.
p. 292Other References
Basle Committee on Banking Supervision, International Convergence of Capital Measurement and Capital Standards (Basle Capital Accord), available at: <http://www.bis.org/publ/bcbs04a.htm>
Convention on International Interests in Mobile Equipment, International Institute for the Unification of Private Law, available at: <http://www.icao.int/icao/en/leb/capetown-conv.pdf>.
Draft Articles on the Law of Treaties, International Law Commission, available at: <http://untreaty.un.org/ilc/texts/instruments/english/commentaries/1_1_1966.pdf>.
Restatement (Third) of the Foreign Relations Law of the United States (1987), American Law Institute.
Treaty on the Non-Proliferation of Nuclear Weapons, available at: <http://www.iaea.org/Publications/Documents/Infcircs/Others/infcirc140.pdf>.
United Nations Treaty Collection, Overview, available at: <http://treaties.un.org/Home.aspx>.
US Department of State, Frequently Asked Questions on Vienna Convention on the Law of Treaties, available at: <http://www.state.gov/s/l/treaty/faqs/70139.htm>.
Vienna Convention on Diplomatic Relations (1961), available at: <http://untreaty.un.org/ilc/texts/instruments/english/conventions/9_1_1961.pdf>.
Vienna Convention on the Law of Treaties, available at: <<http://untreaty.un.org/ilc/texts/instruments/english/conventions/1_1_1969.pdf>.