Accession Rules for International Institutions

Abstract

Powerful states often accept unanimity voting on accession to international institutions, even though this enables weak states to blackmail powerful states into providing costly side payments. Whereas the literature attributes this choice mainly to efforts to bolster the legitimacy of international institutions, the authors demonstrate that the choice of unanimity also has a strategic component. The authors formally show that unanimous accession rules can profit powerful states by creating uncertainty as to the minimal level of reform that enables accession. If accession is valuable enough and the membership candidate is uncertain about the resolve of weak states, it plays safe by implementing ambitious reforms that improve the efficacy of the international institution. In this case, a legitimacy-efficacy trade-off does not exist: the unanimity rule enhances legitimacy while allowing powerful states to induce significant reforms by applicants to the benefit of current members.

Keywords

international institutions international organizations accession rules legitimacy institutional efficacy

If you have the guts, you can use the veto with great effect. You’ve got everybody by their balls.

Quote from an EU ambassador about bargaining in the European Council (Tallberg 2008, 695)

International institutions have become an integral part of global governance. Rather than acting unilaterally or forcing compliance by weak states, even powerful states such as the United States have voluntarily accepted and even promoted institutional constraints by delegating decision-making powers to international institutions, allowing weak states to influence policy outcomes through the formal decision-making process. The degree to which powerful states can use international institutions effectively to shape policies in their favor thus crucially depends on institutional design features, such as the distribution of votes and institutionalized voting rules.

A particularly consequential, but poorly understood, issue in the design of international institutions is the accession rule. The accession rule is of central importance because enlargement decisions will shape cooperation in the long run by shifting bargaining power and changing distributional considerations of current member states. For example, the European Union (EU) Eastern enlargement left a permanent imprint on the distribution of benefits and costs within the EU while China’s accession to the World Trade Organization (WTO) changed the core bargaining dynamics of the multilateral trade regime. One key problem in the design of such rules is that the nature of future accessions underlies great uncertainty, so states cannot easily tailor contingent accession rules for every conceivable situation. When the General Agreement on Tariffs and Trade (GATT) was negotiated in 1948, for example, Western democracies could hardly predict with any accuracy whether any given developing country would apply for membership in the near future.

Perhaps, the most common, though not universal, accession rule is unanimity voting, whereby every current member of an international institution can veto accessions. This rule is, for example, used by the EU, North Atlantic Treaty Organization (NATO), and the GATT/WTO. One obvious benefit of giving weak members some say over accession is to increase the legitimacy of the institution and alleviate concerns regarding institutionalized exploitation of the weak by the strong. By giving weak states some say over international cooperation, powerful states can enhance weak states’ willingness to participate voluntarily and they can avoid costly coercion. The need to secure voluntary participation provides a potential explanation for why powerful states often condone, and even embrace, unanimous or consensual decision rules for international institutions.¹

But the unanimity rule can come at a great cost. In December 1984, the Greek government opportunistically blackmailed billions of euros from powerful EU members by threatening to prevent the accession of Spain and Portugal (Nicholson and East 1987). This costly side payment was mainly funded by large member states that were already net contributors to the budget and firmly believed that the accession of Spain and Portugal was vital to prevent them from backtracking to autocracy. Side payments were too costly when Georgia threatened to veto Russia’s WTO accession in 2006.² Russia had negotiated for twelve years and concluded bilateral negotiations with all other member states, including the United States and the EU, so Georgia was alone blocking accession. Although the powerful states could have coerced Georgia to allow accession, they deemed this strategy too costly.³ By August 2008, Russia and Georgia were at war. After the war, Prime Minister Putin announced that Russia would abandon the reforms it had made during the accession negotiations since “we don’t see or feel advantages from the membership, if they exist at all, but we are carrying the burden.”⁴ Consequently, Russia reversed many of the essential economic reforms that had been implemented in preparation for WTO membership and even formed a customs union with Kazakhstan and Belarus.⁵

In the first case, a weak state used unanimity voting over accession to extract side payments from powerful member states. In the second case, a weak state prevented accession against the interests of powerful member states. Thus, if a powerful state leads the efforts to create an international institution, it may face a trade-off. On one hand, unanimity voting over accession may entice weak states to join the institution in the first place. On the other hand, unanimity voting may render the powerful state vulnerable to blackmail that would not be possible if accession rules reflected the actual distribution of power. Thus, from the perspective of the powerful state, the legitimacy benefits of unanimity voting are accompanied by a loss of efficacy: the unanimity rules may reduce the ability of the international institution to maximize the benefits that powerful states obtain from future accession rounds.⁶

We address this trade-off by formally investigating the strategic logic of unanimous accession rules. First, in line with the existing literature, our equilibrium analysis shows that in many circumstances, the trade-off between legitimacy and efficacy is indeed central to the bargain over the accession rule. If a powerful state believes that future accession candidates are of strategic importance, it will only allow unanimity rule if weak states refuse to join unless they can expect some say over future accession decisions. Second, we show that under certain (common) conditions, the powerful state benefits from unanimity voting over accession decisions and the trade-off between legitimacy and efficacy disappears. Specifically, unanimity makes accession dependent on the approval of all member states. The powerful state can therefore credibly commit to rejecting the membership application unless it obtains enough gains from accession to outweigh the cost of giving side payments to the other members in return for their approval. If the candidate is uncertain about the resolve of weak states, it plays safe by implementing ambitious reforms that improve the efficacy of the international institution. In other words, unanimity tends to increase the equilibrium level of reform accepted by the candidate states. To the degree that these reforms benefit the powerful state, the unanimity rule thus may produce a higher expected payoff than alternate voting schemes, such as weighted voting, for all members of the international institution.

The article therefore characterizes the consequences of unanimity voting over accession. While previous scholarship has correctly emphasized the legitimacy value of unanimity voting, we are not aware of any systematic study of the strategic logic of unanimity voting over accession decisions. We show that, contrary to common belief, unanimity voting may benefit powerful states by inducing candidates to implement extensive reforms that improve the effectiveness of future cooperation. On one hand, these benefits offer a rationalist insight into the design of accession rules: if powerful states anticipate future accession rounds, they may expect benefits from applying the unanimity rule. On the other hand, our characterization of the effects of unanimity voting applies even if the original decision over accession rules was motivated by other factors as well. Thus, the strategic logic applies to accession negotiations, and the resulting reforms, regardless of why the voting rules were originally designed in a given fashion.

Motivation

This section motivates our question of why powerful states agree to implement unanimous accession rules in many international institutions. We show that while the literature correctly recognizes the legitimacy benefits of unanimity rules, it neglects the strategic value of unanimous accession rules for powerful states. Specifically, we discuss (1) why giving up power over the accession decision is costly to powerful states, (2) why such costs are most likely to occur in international institutions dealing with distributional issues, and (3) why assumptions regarding exit options of weak states in the existing literature have led to an overemphasis of legitimacy requirements in institutional design.

The accession rule has important implications for the distribution of benefits and costs within an international institution. The effect of accession rules on members’ payoffs depends on the functions of the international institution in focus. Whereas many international institutions focus on a narrow range of relatively uncontroversial issues, such as research coordination or cultural exchange, other international institutions are distributional, as their decisions carry important allocative consequences. For example, the WTO can issue binding rulings on trade policy, while NATO requires contributions for collective security and peacekeeping. Accession of new states to such an institution can shift the balance of benefits and costs of member states and thereby affect the cooperation payoff to powerful states.⁷

The potential costs of giving up decision-making power over the accession of further member states are readily illustrated by the Mediterranean enlargement of the EU, already discussed in the introduction. Although Greece first strategically threatened to prevent accession, a few years later, it reversed course by colluding with Spain and Portugal to overhaul the system of income transfers between member states (Moravcsik 1991, 1998). Before the Mediterranean enlargement, redistribution had mainly benefited the powerful states, as France and Germany captured a large proportion of the EU budget through the Common Agricultural Policies. However, the Mediterranean coalition changed the ratio of funds allocated to the Common Agricultural Policies and to the Cohesion Policies dramatically to their favor. Along similar lines, in 2001, Mexico delayed China’s accession to the WTO by over a year although both the United States and the EU had already given their full support of Chinese membership by concluding their bilateral accession negotiations. Mexico tried to change China’s attitude toward Mexico’s request for a transitional period for the antidumping rules it had imposed against Chinese products that were all illegal under WTO law (Kraft 2007). Mexico’s bargaining tactics therefore delayed the opening of the Chinese markets to its main trading partners.

If powerful states may expect heavy losses from unanimity voting, as these cases illustrate, why do they accept unanimity voting? It is easy to see why majority voting is often not a credible alternative, as it may leave powerful states vulnerable to large coalitions of weak states. However, why not simply adopt a weighted majority rule, as was done in the Bretton Woods institutions? To explain this conundrum, scholars have argued that weak states need institutional safeguards against exploitation (Cooley and Spruyt 2009; Ikenberry 2000; Lake 2009; Rector 2009; Stone 2008, 2011; Zamora 1980). If a powerful state proposes international institutionalization without giving a weak state any say over future accessions, the weak state might be unwilling to participate in the institution. After all, future accessions may prove particularly harmful to members that are already weak, as powerful states can play weak states against each other for distributional gain. To create a notion of legitimacy for less powerful states, and thereby induce broad participation, “( . . . ) institutions have to be designed in such a way that all of their members benefit from participating, if not in every instance, at least in expectation” (Stone 2008, 593). “The most straightforward way to ensure that all states have a voice in decisions is to enforce a rule of unanimity” (Woods 1999, 50).⁸

Although legitimacy and safeguards are important to explain unanimous accession rules, an explanation that is purely based on legitimacy arguments has an important limitation: it depends on the assumption that weak states have credible exit options. Do developing countries really have credible alternatives to joining the multilateral trade regime and regional trade agreements? Can transitional democracies in Southern or Eastern Europe really afford to eschew European enlargement? Can the small neighbors of Russia really credibly defend their borders and interests against Russia without joining NATO? If powerful states understand that weak states are desperate to join an international institution, they might be tempted to abandon unanimity voting, especially in regards to such integral decisions as accession. Thus, there are good reasons to believe that, although important, legitimacy is only a part of the puzzle. In the following, we formally derive a strategic logic of unanimous accession rules that complements the logic of legitimacy.

Main Assumptions

Our theory is built on three main premises. First, the widening of an international institution is potentially beneficial for current members. An ideal membership candidate can productively participate in international cooperation, so current members prefer to admit it as a member. Benefits for current members are greatest if the membership candidate implements various political and economic reforms, defined as (partially) irreversible adjustments that increase the potential gains from cooperation to current members upon accession. For example, reforms may contain trade liberalization or privatization, as well as effective environmental regulations or a program against corruption. The exact nature of reform will depend on the characteristics of the issue area. This notion is widely accepted in the literature. As international cooperation theorists emphasize, the efficacy of an international institution depends on the ability of the member states to credibly commit to mutually profitable policies (Alesina, Angeloni, and Etro 2005; Downs, Rocke, and Barsoom 1998). Conversely, if the current members admit candidates that free ride or obstruct decision making, the value of cooperation declines. Accordingly, the EU only accepts new members if they meet the 1993 Copenhagen criteria that prescribe intensive democratization and transition toward a market economy, the WTO imposes trade liberalization requirements for accession, and NATO requires that new entrants improve military performance and institutionalize civilian oversight and control.

Second, accession is highly valuable to the membership candidate. This assumption builds on the notion that candidates self-select into the pool of applicant states because they expect to gain from membership. As Mattli and Plümper (2002) argue, for example, membership applications to international institutions depend on the value of membership to influential domestic groups in outsider states. We can therefore assume that states that applied for membership value membership more than states that did not apply for membership. Candidate states prefer being inside to being outside the international institution.

Third, membership candidates are uncertain about the probability that weak states are willing to accept their accession. Whereas candidates usually have information about the preferences of the powerful states in an international institutions, it is often hard to gather reliable information about the preferences of weak member states, particularly if the organization has many members. Such uncertainty can be thought of as the probability that a given weak state expects a very high cost from accession in the absence of extensive reforms. Thus, uncertainty pertains to the upper bound of the accession cost to weak states in the current membership. For example, the Eastern enlargement of NATO was highly contested even after the United States had successfully established a basic commitment to expansion. According to Schimmelfennig (1999), the current members had conflicting preferences, which induced great uncertainty among the candidates about the minimum required reforms for accession.

The Model

In the model, a powerful state and a weak state form an international institution, anticipating that a candidate will submit a membership application.⁹ The powerful state is dominant and it begins the game by offering an accession rule to the weak state. The accession rule is either unilateral (such as weighted voting) or bilateral (such as unanimity voting). Under the unilateral accession rule, the powerful state unilaterally decides over enlargement. Under the bilateral accession rule, the weak state can formally block accession. We assume that the weak state can reject the offer, so that no international institution is formed. This goes back to the notion that powerful states want to maximize the legitimacy of newly formed international institutions.¹⁰ In this case, the payoff to each state is, without loss of generality, normalized to zero. If the international institution is formed, both countries immediately obtain payoffs $V_{p o w e r f u l}$ , $V_{w e a k} > 0$ . These payoffs can be thought of as the value of cooperation prior to any future accessions, as compared to unilateral outside options. All model notation is summarized in Table 1.

Table 1.

Model Notation

Symbol	Interpretation
V _powerful, V _weak	Instantaneous payoffs from forming an international institution
ρ	Probability of “strategic times” (powerful state uninterested in reform)
θ	Reform by candidate
c(θ)	Cost of reform to candidate
λ(θ)	Equilibrium probability of successful accession given reform
B	Accession benefits to candidate
T	Side payment from powerful to weak state
Y	Accession payoff to the powerful state in “strategic times”
θ − 1	Accession payoff to the powerful state in “ordinary times”
θ − X	Accession payoff to the weak state (X is subject to uncertainty)
$\overset{ˉ}{X}$	Highest possible accession cost to weak state

We use three simplifying assumptions to ease modeling. First, the powerful state can set the agenda for negotiations. This assumption is quite realistic, because we assume that only the powerful state is capable of leading cooperation efforts. It is not necessary for the results, however, as the key bargaining dynamics would apply for many other bargaining protocols, such as Rubinstein repeated offers or the Nash Bargaining Solution. Second, the distribution of gains, future accessions notwithstanding, is exogenous. This assumption is not necessary for the results, as one may imagine that the payoffs $V_{p o w e r f u l}, V_{w e a k}$ include all side payments and issue linkages that are commonly used in international negotiations. Third, the accession rule is not malleable over time, so that the international institution cannot easily adapt it to changing circumstances. This assumption is empirically realistic, as voting rights usually change slowly or not at all over time. Most importantly, we are not aware of any case in which the international institution changed accession rules from unanimity to majority voting.¹¹

The decision rule is chosen under uncertainty about the state of the world $ω$ . It is strategic times $α$ with probability $ρ$ and ordinary times $β$ with the complementary probability $1 - ρ$ . The state of the world is simultaneously revealed to all players only after the powerful state has chosen the decision rule. In strategic times, the powerful state has an overriding strategic interest in accession. In ordinary times, the powerful state prefers accession if and only if the member candidate has undergone political and economic reforms. This distinction captures the notion that if a country is strategically important, it need not reform.

After the state of the world has been revealed, the candidate selects a level of reform $θ \in [0, \infty]$ .¹² We assume the candidate has no intrinsic interest in reform. The cost of reform is $c (θ)$ , where c is an increasing and strictly convex function for which the Inada conditions hold. This assumption is innocuous. If the candidate benefits from reform, one can say that the candidate has already chosen the intrinsically ideal level of reform normalized to zero. The expected payoff to the membership candidate is

u_{c a n d} = λ (θ) \cdot B - c (θ),

where $λ = λ (θ)$ is the probability of successful accession given reform $θ$ and $B > 0$ is the net payoff from accession relative to status quo.

After the candidate has chosen a level of reform, it automatically submits a membership application. The powerful state reviews the membership application and decides whether to support it or not. If the powerful state does not support membership, there is no accession and the game ends. Under the unilateral rule, the game also ends if the powerful state supports accession, so that the candidate joins the international institution. Under the bilateral rule, the powerful state offers a side payment $t \in (- \infty, \infty)$ to the weak state in exchange for supporting the membership application. For generality, the side payment can be negative, so that the powerful state could demand concessions from the weak state. The weak state accepts or rejects the offer. If it accepts, accession takes place and the weak state receives the side payment. If it rejects, accession does not take place and the weak state does not receive the side payment. In strategic times, the payoff from accession to the powerful state is Y, where Y is so high that the powerful state prefers accession even in the absence of any reform and is willing to pay a side payment to secure accession. In ordinary times, the payoff from accession to the powerful state is $θ - 1$ , so that the powerful state prefers to maximize reform but is willing to support accession as long as $θ \geq 1$ .

Regardless of the state of the world, the payoff from accession to the weak state is $θ - X$ , so X is the lowest level of reform that is acceptable to the weak state. Substantively, X is interpreted as the cost of accession without reform to the weak state. It could not only follow from economic competition but also from political animosities or hostile public opinion. It need not directly relate to reform, as long as sufficient reforms $θ$ can potentially outweigh the cost $X$ . Thus, the weak state is only willing to support accession such that the reform benefits exceed the costs, $θ \geq X$ .

We consider two possible information structures. First, under complete information, the value of $X$ is known to all players ex ante. Second, under incomplete information, the value of $X$ is drawn from a commonly known uniform probability distribution on $[0, \overset{ˉ}{X}]$ , where $\overset{ˉ}{X} \geq 1$ .¹³ It is only revealed to the powerful and weak state. Importantly, the notion of uncertainty pertains to the upper bound of the accession cost to the weak state. As $\overset{ˉ}{X}$ increases, uncertainty increases in the sense that the accession costs may be even higher than previously, while the lower bound zero remains unchanged. We model uncertainty in this specific sense so as to focus attention on the consequences of incomplete information regarding the willingness of the weak state to reject an accession application.¹⁴

If there is a side payment t, it is subtracted from the payoff of the powerful state and added to the payoff of the weak state. For tractability, we omit all inefficiencies associated with side payments, but all qualitative results hold even if there is a transaction cost $z \cdot t$ that the powerful state must incur for a side payment. Let $R = 1$ if the bilateral rule is used and $R = 0$ if the unilateral rule is used. Combining these elements, the payoff to the powerful state is zero without accession and otherwise

u_{p o w e r f u l} = \{\begin{matrix} θ - 1 - R \cdot t & | ω = β \\ Y - R \cdot t & | ω = α \end{matrix} .

The payoff to the weak state is zero without accession and otherwise

u_{w e a k} = θ - X + R \cdot t .

Equilibrium Analysis

Since our game includes incomplete information, our solution concept is the perfect Bayesian equilibrium.¹⁵ We solve the game as follows. First, we find the expected payoffs for the two states under the two rules. Second, we solve for the optimal accept–reject rule of the weak state. Finally, we find the optimal offer for the powerful state. The game has a unique equilibrium.

Unilateral Rule

Under the unilateral rule, the powerful state accepts a membership application if and only if the payoff exceeds zero. In strategic times, the powerful state always accepts to obtain a payoff of Y. If the state of the world is ordinary, the powerful state accepts if and only if $θ - 1 \geq 0$ . The membership candidate conditions reform on the state of the world. If it is strategic, the candidate understands that it can accede without any reform at all, so it selects $θ = 0$ . If it is ordinary, the candidate must choose $θ = 1$ to gain accession. Thus, the candidate reforms $θ = 1$ if and only if the net value from accession exceeds the cost of reform, $B \geq c (1)$ .

In equilibrium, the expected payoff to the powerful state is simply $ρ \cdot Y$ . In strategic times, the powerful state captures all of the surplus because the weak state cannot intervene. In ordinary times, the membership candidate chooses the minimal acceptable reform, leaving the powerful state exactly indifferent between accession and the outside option. Consistent with previous research, the unilateral rule is optimal for the powerful state in strategic times and ineffective in ordinary times.

The Bilateral Rule

Under the bilateral rule, the powerful state only secures accession if it can profitably offer a side payment to the weak state. If the state of the world is strategic, the value of accession to the powerful state is $Y$ . The value to the weak state is $θ - X$ , so the minimal side payment that the powerful state must offer is $t = X - θ$ .¹⁶ The total payoff to the powerful state is therefore $Y - X + θ$ . With $Y$ high enough, this payoff is strictly positive, so the powerful state does secure accession, but at a higher cost than under the unilateral rule. In ordinary times, the value to the powerful state is $θ - 1$ and the value to the weak state is $θ - X$ . The side payment that the powerful state must offer is $t = X - θ$ , so the value of accession to the powerful state is $2 θ - 1 - X$ . This payoff is negative only if $θ$ is low enough, so accession succeeds if and only if $θ \geq \frac{1 + X}{2}$ .

Reform under complete information

Under complete information, the candidate knows the preferences of the powerful and the weak state. In strategic times, the payoff to the powerful state is $Y - X + θ$ . With $Y$ large enough, the candidate selects $θ = 0$ . The equilibrium payoff to the powerful state is $Y - X$ and the equilibrium payoff to the weak state is $X$ . In ordinary times, the value of accession to the powerful state is $2 θ - 1 - X$ , as shown in the previous paragraph, so the candidate selects $θ = \frac{1 + X}{2}$ if $B - c (\frac{1 + X}{2}) \geq 0$ and $θ = 0$ otherwise.

Reform under incomplete information

Under incomplete information, the candidate does not know what the true preferences of the weak state are. By increasing reform, the candidate can increase the probability of accession, so the candidate’s choice of reform must satisfy

λ^{'} (θ) \cdot B = c^{'} (θ)

unless $λ (θ) = 0$ . Accession is possible if and only if the joint surplus between the powerful and the weak state is not negative, so that $2 θ - 1 \geq X$ . Accession succeeds if and only if $X \in [0, 2 θ - 1]$ . The probability that $X \in [0, 2 θ - 1]$ is in turn given by

λ (θ) = \{\begin{matrix} 0 & | θ < \frac{1}{2} \\ \frac{2 θ - 1}{\overset{ˉ}{X}} & | \frac{1}{2} \leq θ < \frac{1 + \overset{ˉ}{X}}{2} \\ 1 & | \frac{1 + \overset{ˉ}{X}}{2} \leq θ \end{matrix} .

This accession probability is graphically represented in Figure 1.

Figure 1.

The probability of successful accession $λ$ as a function of reform $θ$

If $c (\frac{1}{2})$ is high enough, it is optimal for the candidate not to reform at all. Otherwise, the candidate chooses the level of reform $θ$ optimally in $[\frac{1}{2}, \frac{1 + \overset{ˉ}{X}}{2}]$ . We can write equation 4 as

2 \frac{B}{\overset{ˉ}{X}} \geq c^{'} (θ),

where the inequality is strict if and only if $θ = \frac{1 + \overset{ˉ}{X}}{2}$ . The left side represents the benefit of increasing the probability of accession by $\frac{2}{X}$ , while the right side represents the cost. For the candidate, the optimal reform is thus increasing in the benefits of accession $B$ . On the other hand, the optimal reform increases with $\overset{ˉ}{X}$ if and only if $\overset{ˉ}{X}$ is so low that accession is certain. Figure 2 illustrates this relationship.

Figure 2.

The length of each thick line measures equilibrium reform for the corresponding parameter values. If the benefits of accession are high $B$ , an increase in uncertainty about the necessary level of reform, ${\overline{X}}^{*} > \overline{X}$ , results in higher equilibrium reform. If the benefits of accession are low $B^{*}$ , this increase results in lower equilibrium reform.

Bargaining

For the weak state, the accept–reject rule should maximize the expected payoff from the game. Since the bilateral rule guarantees at least a payoff $V_{w e a k} > 0$ , the weak state accepts it. The unilateral rule produces a zero payoff in ordinary times and a payoff $- X$ in strategic times. The expected payoff from the international institution is $- ρ \cdot \frac{\overset{ˉ}{X}}{2} + V_{w e a k}$ , so the weak state accepts if and only if $V_{w e a k} \geq ρ \frac{\overset{ˉ}{X}}{2}$ . The powerful state secures a payoff $ρ \cdot Y + V_{p o w e r f u l}$ under the unilateral rule. Unless the candidate selects zero reform, the payoff under the bilateral rule is $ρ \cdot (Y - X + θ) + (1 - ρ) E (m a x {0, 2 θ - 1 - X}) + V_{p o w e r f u l}$ . The powerful state selects the accession rule, anticipating the weak state’s accept–reject decision, to maximize its expected payoff from the game.

This analysis shows that, while the weak state unambiguously prefers the bilateral rule, the powerful state’s preferences are indeterminate. On one hand, the unilateral rule allows full discretion in strategic times. On the other hand, the bilateral rule may increase reform. In the following section, we fully characterize the decision.

Theoretical Results

The formal analysis has implications for the influence of unanimous accession rules on the efficacy of international institutions after enlargement, and therefore the strategic logic of the design of accession rules. We begin by presenting the results on the design of accession rules. We find that unanimity voting introduces uncertainty about the required level of reform which induces candidate states to choose higher levels of reforms than under certainty. Since all current member states will gain from higher levels of reform, this provides a new strategic rationale for unanimity voting.

Before we begin, note that it is difficult to assess or test empirically which factors (i.e., legitimacy or expected future gains) states considered during the bargaining process. Public statements have their limitations because the literature shows that states often use conflicting rhetoric during secret meetings and public announcements (Stasavage 2004). While we cannot provide direct empirical evidence that states considered the effect of unanimous accession rules on the applicant’s reform efforts, our theoretical results imply that if states design international institutions rationally, they would consider this effect (Koremenos, Lipson, and Snidal 2001). In other words, if we can show that uncertainty introduced by the unanimity rule increases the level of reform that the candidate state is willing to implement—a relationship that is actually testable—then this provides some indirect evidence that current members (including the powerful states) at the very least gain from the unanimity rule. Thus, we are able to provide a new strategic rationale for unanimous accession rules.

After characterizing the equilibrium design of accession rules, we present the theoretical propositions about the influence of uncertainty on the expected level of reform, provide some empirical illustration using the EU Eastern enlargement as a case, and discuss how this relationship could be tested on a more general level.

The Design of Accession Rules

How does unanimity voting over accession influence the payoff to the powerful state? We find that under complete information, the unilateral rule is optimal for the powerful state unless the weak state must be bribed to join the international institution. However, the bilateral rule is optimal for the powerful state under incomplete information if the probability $ρ$ of strategic accession is low enough, regardless of the weak state’s preferences. The probability of a bilateral rule increases with the benefits of reform to the candidate, but the effect of increasing uncertainty about the preferences of the weak state is not monotonic.

Consider complete information and recall that the net cost of using the bilateral rule in strategic times is $X$ . The candidate never reforms because it understands that the powerful state will offer a side payment to the weak state. In ordinary times, the candidate selects the level of reform that leaves the powerful state exactly indifferent to accession, regardless of the decision rule, so the decision rule has no effect on the payoff. The net cost of using the bilateral rule under complete information is $ρ \cdot X$ . Thus, the powerful state only selects the bilateral rule if $V_{w e a k} < - ρ \frac{\overset{ˉ}{X}}{2}$ .

Proposition 1. Under complete information, the powerful state offers the unilateral rule to the weak state unless $V_{w e a k} < ρ \frac{\overset{ˉ}{X}}{2}$ .

Proof. In the main text.

If the powerful state can endogenously create constraints, why does the Schelling conjecture fail? Since the membership candidate decides on the level of reform and side payments are possible, constraints have no effect on the distribution of surplus. The membership candidate understands that as long as accession is not strictly harmful to the international institution, accession succeeds. But if the weak state fears that the value of cooperation is low relative to the outside option, the powerful state may have to choose the bilateral rule, as $V_{w e a k} < - ρ X$ . This finding shows the importance of the weak state’s outside option, as we have already argued. Considerations of legitimacy and mutual gain are indeed central, but only if the weak state can credibly threaten to reject international cooperation otherwise.

What about incomplete information? The expected cost of using the bilateral rule in strategic times is now $\frac{\overset{ˉ}{X}}{2}$ —the expected value of the weak state’s preferences. In ordinary times, the powerful state can always say “no” to reap zero surplus. Unless the equilibrium level of reform is zero, the probability that the weak state also benefits from accession is positive. Thus, it is possible that the joint surplus from accession, $2 θ - 1 - X$ , is positive. The powerful state can leave the weak state exactly indifferent to accession to reap this surplus, so the bilateral rule increases the payoff to the powerful state.

Proposition 2. If the probability of strategic times $ρ$ is low enough or $V_{w e a k} < ρ \frac{\overset{ˉ}{X}}{2}$ , the powerful state offers the bilateral rule to the weak state.

Proof. If $V_{w e a k} < ρ \frac{\overset{ˉ}{X}}{2}$ , the weak state rejects the bilateral rule. The marginal effect of selecting the bilateral rule on the powerful state’s payoff is $- ρ \cdot \frac{\overset{ˉ}{X}}{2} + (1 - ρ) \cdot E (m a x {0, 2 θ - 1 - X})$ . This expression cannot be negative if $ρ$ is low enough.

Under incomplete information, the membership candidate is not sure how much it should reform to enable accession. If it fails to secure accession, the powerful state obtains the payoff from the “outside option,” perhaps interpreted as a delay in accession. However, the uninformed membership candidate could choose excessive reform. Consequently, the powerful state produces a strictly positive joint surplus that it then captures by compensating the weak state for saying “yes.” If the probability of ordinary times is high enough, the net cost of using the bilateral rule for strategic accession is rather irrelevant, so the bilateral rule is optimal for the powerful state.

This proposition shows that under plausible conditions, there is no legitimacy-efficacy trade-off. A powerful state understands that, even though unanimity voting is costly in strategic times, as the weak state can demand compensation for accession, the unanimous decision will also induce the candidate to choose higher levels of reform. To the degree that this benefits both the powerful and the weak state, there is no distributional conflict. This finding provides a powerful and general explanation for unanimity voting in international institutions.

Uncertainty, Reform, and Accession

In addition to several straightforward comparative statics, such as the positive effect of accession benefits $B$ on reform $θ$ , the formal model illuminated the central role of the level of uncertainty over current member states’ preferences. In this section, we summarize these implications. Since they pertain to actual reform decisions and accession terms that are directly observable, as opposed to implicit negotiation preferences over accession rules, we also offer illustrative examples from the EU Eastern enlargement. These examples are useful because they show that the central dynamics of our model were relevant for the interactions between established EU members and various accession candidates in Eastern Europe. In addition, we will discuss how a more general test could be conducted.

Uncertainty about necessary reforms

What is the effect of incomplete information about the necessary level of reforms? The simplest way to evaluate this effect is to allow the highest possible cutoff for accession preference $\overset{ˉ}{X}$ increase. This parameter captures an exogenous increase in the probability that the weak state is really willing to prevent accession for a given level of reform. Interestingly, we find a nonmonotonic relationship.

Proposition 3. Consider an international institution that uses unanimity voting. Suppose the highest conceivable accession cost to weak current members ( $\overset{ˉ}{X}$ ) increases. If the benefits of accession to a candidate ( $B$ ) are low enough, reform ( $θ$ ), and the probability of accession ( $λ$ ) decrease. If the benefits of accession to a candidate ( $B$ ) are high enough, equilibrium reform ( $θ$ ) increases while the probability of accession ( $λ$ ) remains constant.

Proof. For low enough $B$ , the equilibrium probability of accession $λ (θ)$ is less than one. Use equation 6 to verify that equilibrium reform $θ$ is locally decreasing in $\overset{ˉ}{X}$ . For high enough $B$ , the equilibrium probability of accession $λ (θ)$ is one. Use equation 6 to verify that equilibrium reform $θ$ is locally increasing in $\overset{ˉ}{X}$ .

If the probability of accession is uncertain to begin with, increases in $\overset{ˉ}{X}$ reduce both reform and the probability of accession. As the marginal effect of reform on the probability of accession decreases, the cost–benefit ratio of reform worsens. Notably, this hurts the powerful state, because it wants to maximize the equilibrium reform and the probability of accession. But if the membership candidate cannot afford failure, it responds to increased uncertainty by increasing the equilibrium level of reform.

The EU Eastern enlargement negotiations nicely illustrate how uncertainty about the necessary level of reform-induced membership candidates to implement radical reforms, which benefited the current member states. Uncertainty arose mainly from the introduction of the Copenhagen Criteria in 1993 which requires membership candidates to provide stability of institutions guaranteeing democracy, the rule of law, human rights and respect for and protection of minorities; a functioning market economy and the capacity to cope with competitive pressure and market forces within the Union; and acceptance of the acquis communautaire, the total body of EU law.

First, the scope and nature of the first two conditions were abstract and interpreted differently by the member states (Maniokas 2000). EU officials claimed “to know what was acceptable or unacceptable when they saw it” (Jacoby 2004, 7). Although the Commission led the negotiations, all member states had to agree that a membership candidate was ready to begin accession negotiations. Such consensus was difficult to achieve given disagreement and ambivalence among current member states. Second, even if accession negotiations were opened, the implementation of the acquis communautaire required that all member states agree on all thirty-one chapters of the common acquis.¹⁷ The candidates had to ensure that the domestic reforms they promised to the EU would be deep enough to satisfy each individual fifteen current member states. Disapproval during the negotiations about a chapter typically delayed the accession process tremendously which provided big incentives for candidates to satisfy all EU members in the first round of negotiations.¹⁸ In this situation, it did not help that Germany and France publicly announced that they would not be willing to unilaterally pay for the full burden of enlargement (by, e.g., providing side payments necessary to coax the poor member states) but rather demanded reform-willingness from the application states. Consequently, membership candidates understood that ambitious reform secured accession with high probability while partial reform could lead to failure. Knowing that sufficient reform willingness was a key to accession, many membership candidates accepted demanding “shock therapy” reforms early in the enlargement process. For example, although Germany supported Poland’s accession to the EU, the current recipients of large agricultural subsidies expressed grave concern about the effect of Polish accession without major reforms in her agricultural sector.¹⁹ Poland therefore implemented radical reforms even though they were domestically highly contentious (Wilkin 2001).²⁰

For a general test, it is important to find appropriate measures for uncertainty about the required level of reform. Several empirical factors could predict uncertainty about the preferences of weak states. First, as in the case discussed earlier, uncertainty could vary across international institutions. International institutions in which accession regulations give much room for interpretation and power to individual member states should induce much higher uncertainty. Second, uncertainty could vary within international institutions. Candidate states should have less information about democratic countries with frequent political turmoil and public opinion hostile to accession. Notably, this could explain why the Southern European countries were able to extract concessions from the EU in accession negotiations. Third, candidate states should have less information about weak states that do not frequently interact with membership candidates. Whereas China probably has a good understanding of political realities in the United States, it probably has limited information about the true preferences of the Mexican government. Similarly, Eastern European countries more frequently interact with Austria than Portugal. Fourth, uncertainty could vary across candidate countries. Poor membership candidates face higher uncertainty because they tend to cause distributional conflict for at least some current member states. Finally, uncertainty could vary with characteristics of the enlargement process itself. The EU Eastern enlargement shows that a high number of applicants can further aggravate uncertainty.

Uncertainty about accession

We have so far assumed that accession depends only on reform in ordinary times. However, accession can also be uncertain regardless of reform. To capture this possibility, we modify the model so that in ordinary times, accession fails with probability $γ$ , regardless of the level of reform $θ$ . Then the total probability of accession is simply $(1 - γ) \cdot λ (θ)$ . For example, negative public opinion could prevent even the accession of an ideal membership candidate. This form of uncertainty unambiguously reduces equilibrium reform.

Proposition 4. Consider an international institution that uses unanimity voting. If the probability ( $γ$ ) that accession fails for reasons unrelated to reform ( $θ$ ) increases, reform ( $θ$ ) decreases.

Proof. Replace $λ (θ)$ with $(1 - γ) \cdot λ (θ)$ and let $γ$ increase. In equation 6, the left side decreases, while the right side remains constant for a given $θ$ , so equilibrium reform must decrease.

If additional reform has little effect on the probability of accession, the membership candidate hesitates to pay the cost. Again, the EU Eastern enlargement provides a nice illustration of the causal mechanism because the EU experimented with two different approaches to the enlargement process, both of which where designed to increase reform efforts of the applicant states. In 1997, the Council of Luxembourg tried to intensify the candidates’ reform efforts by beginning accession negotiations only with the Luxembourg Group (Cyprus, Hungary, Poland, Estonia, the Czech Republic, and Slovenia). The Helsinki Group (Romania, Slovakia, Latvia, Lithuania, and Bulgaria) had to wait. The idea was to differentiate between countries based on their readiness to fulfill the criteria and to increase reform efforts by those who were falling behind.

This strategy of differentiation slowed rather than accelerated reform efforts for countries outside the Luxembourg Group, as the differentiation approach raised fears that even full reform could be insufficient (Maniokas 2000; Glenn 2004; Mattli and Plümper 2004). The Copenhagen criteria stated that membership was only possible as long as the EU could absorb new members, and the limits were unclear. Additionally, whereas the EU set a timetable for the Luxembourg Group, it refused to announce any specific dates for the commencement of negotiations for the Helsinki Group until the end of 1999.

The negative effects of differentiation led the 1999 Council of Helsinki to abandon this strategy. As Poul Skytte Christoffersen, a key EU negotiator said, “Leaving somebody behind—awaiting another ‘wave’ in an uncertain future-could take away the pressure for reform and modernisation in the countries left on the shore” (Christoffersen 2007, 32). The EU decided to begin accession negotiations with the Helsinki Group in 2000.²¹ The “second wave” was given a chance to accede, exactly as the “first wave” had a chance to accede previously. The approach was dubbed “Regatta” because it gave the remaining countries an equal opportunity without a corresponding commitment to simultaneous accession. This decision spurred the reform process, especially in countries that had just started negotiations. Lithuania successfully closed negotiations on over twenty-eight chapters after only two years of negotiations, passing former frontrunners of the first wave (Plümper, Schneider, and Troeger 2006). To increase the chances of early accession, candidates accepted shorter transition periods for the implementation of the acquis and longer transition periods for current members.

For a general test, it is essential to distinguish between different types of uncertainty. Mattli and Plümper (2004) show that when the EU delayed the accession of several Eastern European countries, these countries reduced the pace of reform. Our analysis implies that while uncertainty over acceptable accession increases reform in eager membership candidates, an increase in the probability that accession fails, regardless of reform should reduce equilibrium reform. Indeed, our model subsumes the partial analysis that Mattli and Plümper (2004) conduct, as we endogenize the decision to approve accession. In the case of Eastern enlargement, it can therefore explain why reforms increased dramatically in some of the second-tier countries after accession negotiations were commenced. In addition, it can explain why Russia decided to abandon many of the reforms that were implemented during accession negotiations with the WTO. Georgia’s veto had dramatically decreased the likelihood that Russia could join the WTO anytime soon without an intervention by the United States or the EU.

Conclusion

This article examined the strategic logic of unanimous accession rules in international institutions. Weak states can use egalitarian rules to blackmail powerful states, so it appears as though there is a genuine trade-off between legitimacy and efficacy. We demonstrated that if unanimity voting endogenously creates uncertainty about the requisite level of reform and the membership candidate cannot afford to stay outside, all current members of the international institution gain from unanimity voting. Our findings shed light on questions at the heart of decision making in international institutions. Previous research has not produced detailed predictions regarding the specific decision rules that give powerful states an aura of legitimacy at the lowest cost. Instead of accepting the received notion of an unavoidable trade-off, we have shown that unanimous accession rules can often secure voluntary participation by weak states while actually benefiting powerful states.

The broader implications of our findings are potentially interesting. First, Downs, Rocke, and Barsoom (1998) argue that sequential enlargements, as opposed to inclusive enlargement, mitigate the broader-deeper trade-off that exists in international cooperation. Our findings indicate that inclusive enlargements can also mitigate the broader-deeper trade-off because competition over accession among entrants prompts reforms that enable deeper cooperation between current and new members. Second, while previous research has emphasized that uncertainty can cause bargaining failure (Fearon 1995), our equilibrium does not carry the burden of inefficiency. Uncertainty surrounding the readiness of an international institution for enlargement benefits current members at the expense of the membership candidate, so it is not costly for powerful states to strategically create uncertainty as to the probability of accession without reform. Third, our findings show how powerful states can use international institutions to redistribute the gains from cooperation. Gruber (2000) shows that powerful states can sometimes worsen the status quo for weak states to secure “voluntary participation” on expedient terms. We extend this argument by showing that through voting rules, international institutions can achieve redistribution even if the status quo cannot be influenced.

The central role of uncertainty for the efficacy of international institutions bears some interesting policy implications. If powerful states prefer to expand the membership of an international institution, but only after applicants undergo reforms, they should increase uncertainty about accession without full reform while decreasing uncertainty about accession given full reform. In this respect, the proposal by the French President, Jacques Chirac, in 2005, to require that accession to EU require domestic approval appears troubling. Subjecting accession to the vicissitudes of domestic politics could close the door to many aspiring membership candidates and thereby slow down ongoing reform efforts in transition countries dramatically.

Mathematical Appendix on Generalizing Uncertainty

In the main text, we assumed that the powerful state’s preferences are revealed to the accession candidate. In this appendix, we generalize the model as follows. First, let $F^{u n i}$ be the cumulative distribution of the powerful state’s minimal acceptable reform in ordinary times. Second, let $F^{b i}$ be the joint distribution of the zero surplus cutoff. The only assumption is that each distribution is continuously differentiable on the positive side of the real line. Let $f^{r}$ denote the density of $F^{r}$ for $r \in {u n i, b i}$ . We also assume that the accession candidate is uncertain as to the state of the world (ordinary or strategic). Thus, $ρ$ is the probability that the accession candidate ascribes to the strategic state of the world.

Under the unilateral rule, the candidate maximizes

(1 - ρ) B F^{u n i} (θ) - c (θ) .

Under the bilateral rule, the candidate maximizes

(1 - ρ) B F^{b i} (θ) - c (θ) .

Thus, the candidate selects $θ^{*} = 0$ or such that

(1 - ρ) B f^{r} (θ) \leq c^{'} (θ),

where $r \in {u n i, b i}$ and the inequality is strict if and only if $F^{u n i} (θ^{*}) = 1$ .

It is straightforward to verify that if $B$ is high enough, the candidate selects $θ^{*}$ such that $F^{r} (θ^{*}) = 1$ . Thus, if $f^{b i}$ is low enough everywhere, $θ^{*}$ approaches infinity. In these circumstances, the bilateral rule maximizes the expectation of $u_{p o w e r f u l}$ .

Similarly, verify that if $ρ$ is high enough, $θ^{*} \to 0$ . Now, the candidate does not reform at all, regardless of the decision rule. Thus, the expectation of $u_{p o w e r f u l}$ is maximized by the unilateral rule if and only if the weak state accepts it.

Footnotes

Acknowledgments

The authors thank Terry Chapman, Mareike Kleine, Tonya Putnam, Branislav Slantchev, two anonymous reviewers, and a seminar audience at IMT Lucca for helpful comments.

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Alesina

Alberto

Angeloni

Ignazio

Etro

Federico

. 2005. “International Unions.” American Economic Review 95 (3): 602–15.

Buchanan

Allen

Keohane

Robert O.

. 2004. “The Preventive Use of Force: A Cos-mopolitan Institutional Proposal.” Ethics and International Affairs 18 (1): 1–22.

Christoffersen

Poul Skytte

. 2007. “The Preparation of the Fifth Enlargement.” In The Accession Story. The EU from Fifteen to Twenty-Five Countries, edited by Vassiliou

George

, 24–33. Oxford, UK: Oxford University Press.

Cooley

Alexander

Spruyt

Hendrik

. 2009. Contracting States: Sovereign Transfers in International Politics. Princeton, NJ: Princeton University Press.

Downs

George W.

Rocke

David M.

Barsoom

Peter N.

. 1998. “Managing the Evolution of Multilateralism.” International Organization 52 (2): 397–419.

Drezner

Daniel W.

2007. All Politics is Global: Explaining International Regulatory Regimes. Princeton, NJ: Princeton University Press.

Fearon

James D.

1995. “Rationalist Explanations of War.” International Organization 49 (3): 379–414.

Glenn

John K.

2004. “From Nation-States to Member States: Accession Negotiations as an Instrument of Europeanization.” Comparative European Politics 2:3–28.

Gruber

Lloyd

. 2000. Ruling the World: Power Politics and the Rise of Supranational Institutions. Princeton, NJ: Princeton University Press.

10.

Ikenberry

G. John.

. 2000. After Victory. Princeton: Princeton Universtiy Press.

11.

Jacoby

Wade

. 2004. The Enlargement of the EU and NATO: Ordering from the Menu in Central Europe. New York: Cambridge University Press.

12.

Koremenos

Barbara

Lipson

Charles

Snidal

Duncan

. 2001. “The Rational Design of International Institutions.” International Organization 55 (4): 761–99.

13.

Kraft

Christian

. 2007. Joining the WTO: The Impact of Trade, Competition and Redistributive Conflicts on China’s Accession to the World Trade Organization. Frankfurt, Germany: Peter Lang.

14.

Lake

David A.

2009. Hierarchy in International Relations. Ithaca, NY: Cornell University Press.

15.

Maniokas

Klaudijus

. 2000. “Methodology of the EU Enlargement: A Critical Appraisal.” Lithuanian Foreign Policy Review 5: 1–29.

16.

Mattli

Walter

Plümper

Thomas

. 2002. “The Demand-Side Politics of EU Enlargement: Democracy and the Application for EU Membership.” Journal of European Public Policy 9 (4): 550–74.

17.

Mattli

Walter

Plümper

Thomas

. 2004. “The Internal Value of External Options: How the EU shapes the Scope of Regulatory Reforms in Transition Countries.” European Union Politics 5 (3): 307–30.

18.

Moravcsik

Andrew

. 1991. “Negotiating the Single European Act: National Interests and Conventional Statecraft in the European Community.” International Organization 45 (1): 19–57.

19.

Moravcsik, Andrew . 1998. The Choice for Europe: Social Purposes and State Power from Messina to Maastricht. Ithaca: Cornell University Press.

20.

Nicholson

Frances

East

Roger

. 1987. From the Six to the Twelve: The Enlargement of the European Communities. Keesing's International Studies. London, UK: Longman.

21.

Plümper

Thomas

Schneider

Christina J.

Troeger

Vera E.

. 2006. “The Politics of EU Eastern Enlargement: Evidence from a Heckman Selection Model.” British Journal of Political Science 36 (1): 17–38.

22.

Rector

Chad

. 2009. Federations: The Political Dynamics of Cooperation. Ithaca, NY: Cornell University Press.

23.

Schimmelfennig

Frank

. 1999. “NATO’s Enlargement to the East: An Analysis of Collective Decision-making.” EAPC-NATO Individual Fellowship Report 1998-2000.

24.

Stasavage

David

. 2004. “Open-Door or Closed-Door? Transparency in Domestic and International Bargaining.” International Organization 58 (3): 667–703.

25.

Stone

Randall W

. 2008. “The Scope of IMF Conditionality.” International Organization 62 (4): 589–620.

26.

Stone

Randall W

. 2011. Controlling Institutions: International Organizations and the Global Economy. Cambridge, NY: Cambridge University Press.

27.

Tallberg

Jonas

. 2008. “Bargaining Power in the European Council.” Journal of Common Market Studies 46 (3): 685–708.

28.

Wilkin

Jerzy

. 2001. Poland into the New Millenium, 231–45. Northampton: Edward Elgar Publishing Chapter Polish Agriculture, Rural Areas and EU Enlargement.

29.

Woods

Ngaire

. 1999. “Good Governance in International Organizations.” Global Governance 5 (1): 39–61.

30.

Zamora

Stephen

. 1980. “Voting in International Economic Organizations.” American Journal of International Law 74 (3): 566–608.