Why many cooks if they can spoil the broth? The determinants of multiparty mediation

Abstract

Which factors drive the onset of multiparty mediation? Despite numerous studies explaining the onset of international mediation in conflicts, the literature offers little insight into when we are likely to see ‘multiparty mediation’, that is, mediation attempts that are conducted by a coalition of interveners. The latter are different from those interventions that only see one mediator, however: mediators forming a coalition are likely to be constrained by different domestic factors and may have different incentives than states pursuing unilateral approaches. Thus, ignoring these patterns can have serious consequences for our theoretical expectations and empirical inferences on mediation onset. In order to address this shortcoming, this article introduces concepts from the coalition-formation literature to the existent research on mediation, and examines the onset of multiparty mediation from a twofold perspective. First, there are supply-side determinants, that is, interveners’ incentives for multiparty mediation. Second, the author studies demand-side influences, that is, factors pertaining to the belligerents. By analyzing data for 1950–2000, the following research’s main contribution stems from the empirical test of the theoretical framework in a quantitative research design. The results point to a disconnect between the factors that promote single-party mediation onset and those that promote multiparty mediations.

Keywords

demand-side influences international mediation multiparty mediation supply-side incentives

Introduction

The 1978 Camp David Accords were induced by intense mediation¹ efforts of President Carter, in which the USA largely acted on its own and relied on little coordination with its allies. During the 1994 Bosnia conflict, however, the USA employed a strategy of ‘multiparty mediation’, that is, a mediation attempt that involves multiple interveners forming a coalition (Crocker, Hampson & Aall, 1999, 2001a,b), and established a mediating group comprising Russia, France, Britain, and Germany. The question that follows asks ‘why did the USA form a mediating coalition in 1994 but decide to pursue a unilateral approach for Camp David?’ More generally, under what conditions does multiparty mediation occur?

Despite significant work that examines the determinants of single-party mediation in interstate and civil disputes (see e.g. Greig, 2005; Greig & Regan, 2008), the literature offers little insight into when we are likely to see multiparty mediation. Generally, mediation is a ‘mode of negotiation in which a third party helps the parties find a solution which they cannot find by themselves’ (Touval & Zartman, 1985). Touval & Zartman (1985: 40) also emphasize that ‘third parties are only accepted as mediators if they are likely to produce an agreement or help the parties out of a predicament’. In other words, both the belligerents and potential mediators must have an interest in an intervention, that is, demand and supply-side incentives are a necessary requirement for (multiparty) mediation to occur (e.g. Crocker, Hampson & Aall, 1999, 2001a,b; Beardsley, 2008, 2010; Beardsley & Greig, 2009; Crescenzi et al., 2011; see also Keohane, 1990: 741ff). Greig (2005: 251) argues in this context that (a) the likelihood of mediation success, (b) the characteristics of the disputants, (c) the previous conflict management history of the disputants, and (d) the threats outside the rivalry faced by the disputants are key influences. Despite a compelling research design and important findings, Greig (2005) addresses all kinds of mediation efforts but does not distinguish between single- and multiparty mediation.

In the following, I contribute to the existing literature by making this distinction between single-party mediation and multiparty mediation. The previous work on mediation largely treats single- and multiparty mediation similarly, in the sense that it assumes that the same conditions influencing the onset of single mediations apply to the case of multiparty mediation. However, I argue that multiparty mediation is likely to be different from the former.² For example, Beber (2010) stresses that multiparty mediation allows each member of an intervening coalition to shoulder a smaller share of the fiscal burden and political risk associated with mediation. Also, ‘groups of mediators can pool resources, skills, and leverage in trying to sway disputants’ (Beber, 2010: 4). All these characteristics do not pertain to unilateral mediations. Similarly, mediators forming a coalition are likely to be constrained by different domestic factors and may have different incentives than states pursuing single-party mediations.

As notable exceptions, Crocker, Hampson & Aall (1999, 2001a,b) actually distinguish between single- and multiparty mediation. The scholars approach the latter in a threefold sense. First, there are simultaneous interventions of multiple mediators that do not necessarily coordinate their efforts. Second, conflicts may see sequential mediated interventions that involve more than one third party. Finally, there are composite actors such as coalitions of states (Crocker, Hampson & Aall, 1999: 10). In this study, I focus on the third ‘composite’ category where third parties actually form a coalition and seek to represent a single interest.³ However, Crocker, Hampson & Aall (1999, 2001a,b) focus on the effectiveness of these kinds of interventions, and it rather seems that they implicitly assume that the same conditions that affect mediation success also influence the likelihood of multiparty mediation onset in the first place.⁴

Against this background, Crocker, Hampson & Aall (1999, 2001a,b) are aware of the potential of multiparty mediation and clearly outline the benefits associated with it. As indicated above, multiparty mediation implies smaller shares of the fiscal burden and the political risk for each member of an intervening coalition. Multiparty mediators can further pool resources and skills and, thus, have in principle more leverage than a single mediator. Crocker, Hampson & Aall (1999: 230) stress here that multiparty mediation can create beneficial situations in which ‘the lead mediator gains the benefit of the partners’ insights, relationships, credibility, resources, diplomatic reach, and political balance. Broadening, when successful, isolates the spoilers and rejectionists common to most conflict situations’. A prerequisite of forming a mediating coalition is, however, that the members of this coalition largely share similar goals and are willing to work together (Crocker, Hampson & Aall, 1999: 22).⁵ Equally important, states in an intervening coalition have to maintain coherence, coordinate and sequence their initiatives, and have staying power over the process (Crocker, Hampson & Aall, 1999: 33, 38f).⁶

On the other hand, the title of their book Herding Cats implies that Crocker, Hampson & Aall (1999) are rather skeptical of multiparty mediation, and, thus, these scholars also emphasize the costs of it. First, a coalition of interveners may increase the complexity of the mediation process, leading to unanticipated coordination and collective action problems. Eventually, this can induce a serious deterioration of the overall conflict situation (Crocker, Hampson & Aall, 1999: 39f): ‘problems of handoff between one peacemaker and the next are all too frequently encountered when different mediators try to engage parties in negotiations over a prolonged period of time.… During this period, misunderstandings and conflicting interpretations about implementation are common’ (Crocker, Hampson & Aall, 2001a: 507). Furthermore, Crocker, Hampson & Aall (2001b: 53ff) emphasize that multiparty coalitions might lack a common vision or send mixed messages that induce a communication gridlock. Consequently, the UN Secretary-General Ban Ki-moon (2009: 6) highlighted that ‘multiple actors competing for a mediation role … reinforce fragmentation in the conflict and complicate resolution’. Nevertheless, multiparty mediation occurred in numerous instances, despite the fact that both decisionmakers and scholars question its potential for success. More specifically, 101 out of 330 issue disputes from 1945 to 2001 (i.e. about 31%) listed in the data I use below actually saw multiparty mediation. If it is perceived to be ineffective in the first place, however, why do actors frequently choose to mediate as part of a coalition?

Systematically addressing when we are likely to see multiparty mediation in conflicts not only helps policymakers and scholars of conflict resolution better understand this phenomenon but is also likely to have implications on areas that are generally beyond common mediation studies. First, the main theoretical contribution of my article is the introduction of concepts from the literature on coalition formations (e.g. Tago, 2005) to the mediation research. As emphasized, third parties actually form a coalition and seek to represent a single interest in the case of multiparty mediation. The research on the formation of military coalitions might facilitate our understanding here. Second and from a methodological perspective, the work on mediation in general faces selection problems. Disputes that see multiparty mediation are unlikely to be a random set, since mediators strategically select themselves into those efforts. This problem of selection bias, which may cause wrong estimates on mediation onset and, in turn, effectiveness if not controlled for, becomes even more serious for multiparty mediations, since there is more than one mediator intervening, and actors of a mediating coalition have to cooperate and coordinate their interactions. This study seeks to further shed light on this issue by outlining the crucial factors that lead to the formation of mediating coalitions.

In what follows, I elaborate my theoretical framework that builds upon prior research on mediation and introduces concepts from coalition formations. Given that both the belligerents and (potential) mediators must have an interest in an intervention (e.g. Beardsley, 2008, 2010; Crescenzi et al., 2011), I examine the onset of multiparty mediation from a twofold perspective. First, there are supply-side determinants, that is, interveners’ incentives. The research concentrates here on mediators’ economies, the accountability of their executives, capabilities, and possible negative externalities. Second, I study demand-side influences, that is, factors pertaining to the belligerents, where I focus on the intensity of a dispute, the regime type of the warring parties and their capabilities, and belligerents’ attempts to balance a potential bias of interests in a mediating coalition. Afterwards, I describe the research design and empirically test the theory with a quantitative research design by analyzing data from the Issues Correlates of War (ICOW) project for 1950–2000 (Hensel & Mitchell, 2007; Hensel et al., 2008). The analysis supports a large share of the hypotheses and points to a disconnect between the factors that promote single-party mediation onset and those that promote multiparty mediations. The final section concludes and stresses the implications of my research.

Why many cooks if they can spoil the broth? The influences of multiparty mediation

Although both policymakers and the scholarly literature are rather skeptical of the effectiveness of multiparty mediations, the existing theoretical arguments focus on a power aspect for explaining this phenomenon (Crocker, Hampson & Aall, 1999, 2001a,b): a coalition of interveners should in principle be more likely than a single third party to establish leverage over the antagonists. This, in turn, raises the likelihood of a stable peace agreement (see Crocker, Hampson & Aall, 1999: 22, 2001a,b). Touval & Zartman (1985) emphasize accordingly that ‘third parties are only accepted as mediators if they are likely to produce an agreement … and for this they usually need leverage’. From the perspective of both the antagonists and the prospective mediator(s), the basic reasoning behind choosing a mediating coalition over a single mediator could then be that the former strategy especially addresses the more difficult cases as the prospects of success may seem higher. Belligerents seek third-party assistance if it is too difficult to terminate fighting by themselves. A single intervener often does not have the necessary amount of leverage for effectively solving conflicts peacefully, but the combined leverage of a mediating coalition can create obligations and expectations that help enforce compliance more effectively.⁷ This should increase the chances of multiparty mediation by definition.

Table I.

Mediation effectiveness and size of mediating coalitions

	Size of mediating coalitions
	1	2	3	4	5	6	Sum
No agreement	175	37	8	7	5	22	254
	(46.17)	(61.67)	(36.36)	(38.84)	(31.25)	(61.11)	(47.74)
Agreement	39	5	2	2	9	2	59
	(10.29)	(8.33)	(9.09)	(10.53)	(56.25)	(5.56)	(11.09)
Compliance/end of claim	165	18	12	10	2	12	219
	(43.54)	(30.00)	(54.55)	(52.63)	(12.50)	(33.33)	(41.17)
Sum	379	60	22	19	16	36	532
	(100)	(100)	(100)	(100)	(100)	(100)	(100)
χ²	44.72	p = 0.000

Table entries are counts, with percentages out of column totals in parentheses.

For evaluating this assumption, I briefly examine the relationship between effectiveness and the size of mediating coalitions.⁸ Although descriptive and arguably selective, Table I depicts a strong relationship between both factors (χ ²= 44.72; p = 0.00) and that mediation effectiveness does indeed increase by coalition size. Consequently, while single-party mediations are fully effective in only 43.54% of the observations, the effectiveness rate of a mediating coalition of three actors increases by about 11 percentage points. Note, however, that effectiveness seems to decrease again when more than four mediating parties are involved. While the agreement rate of multiparty mediations comprising five actors is still notable (56.25%), this agreement rate and the share of fully effective interventions (5.56% and 33.33%, respectively) substantially drops when six mediators form a coalition. Based upon these results, some constellations of multiparty mediation do indeed induce higher compliance rates than single-party interventions, and I will rely at least to some extent on this general assumption for the development of the following theoretical arguments.

Supply-side incentives

As in the case of single-party mediations, interveners need a crucial interest in a conflict per se in order to make mediation occur. However, due to the special circumstances of multiparty mediations, third parties also face the additional cost of forming a coalition and coordinating with other interveners. In this context, the literature on coalitions (e.g. Tago, 2005: 588f) largely concentrates on domestic factors and emphasizes that coalition formations can be explained by (a) a higher degree of legitimacy in the use of force and (b) burden sharing among members of a coalition. In turn, these benefits may outweigh the costs of multiparty mediation (see Beber, 2010: 4; Crocker, Hampson & Aall, 1999, 2001a,b). More precisely, I argue that the economies of mediators, their accountability, and their capabilities are likely to create signals for the domestic audiences and influence the degree of legitimacy and burden sharing. This may then determine whether a mediator forms a coalition of states or, instead, pursues a single-party mediation. Also, I incorporate the impact of possible negative externalities into the cost–benefit analysis of potential multiparty mediators.

First, multiparty mediation may be perceived as burden sharing by the domestic audience as in the case of any other kind of coalition formation (Bennett, Lepgold & Unger, 1994). Trying to settle a conflict peacefully causes costs for the intervener(s), for example expenses for travel, meetings, or related security arrangements (Beber, 2010: 4). Unlike in the case of single-party mediations, however, multiparty mediation signals to the domestic audience that the costs of intervention are being shared and, hence, should be lower than in the former case. Simultaneously, if the costs of a mediation attempt are shared in the sense that more than one party intervenes, this also increases the legitimacy of an intervention. Tago (2005: 588) argues in this regard that the economic situation is one condition that may lead states to form military coalitions. If a country’s economic situation is weak, its government may be more willing to consider multiparty mediation instead of a unilateral approach: other states participate in sharing the costs due to the outside funding sources, and a mediation attempt per se is unlikely to pose a ‘major threat’ toward worsening a nation’s economy in the case of poor conditions anyway. Ultimately, although a country then participates in mediation, it is not perceived as a costly choice by the domestic audience and also increases the legitimacy of the intervention. Therefore, the first hypothesis on supply-side incentives is formulated as:

Supply-side hypothesis 1: The worse the economic situation of a mediator, the higher the likelihood that mediation will be attempted by a coalition of interveners.

The second reason why third parties might form a coalition in the light of burden sharing and higher legitimacy is the accountability of a regime (see Tago, 2005: 589). Basically, citizens can express their will on the performance of a government more effectively in more accountable, that is, democratic countries and, in particular, in the realm of national elections. Bueno de Mesquita et al. (2004) argue that the size of a leader’s winning coalition and selectorate influences accountability.⁹ Also, the higher the leaders’ level of accountability, the more likely it is that they have an extra incentive to make policies that are favorable to the domestic audience. This incentive is driven by a leader’s primary goal of retaining office. Mediation is certainly not an exception here. More precisely, in order to be perceived as favorable to the domestic audience, governments seek cost-saving and legitimate actions. Consequently, less risky and less burdensome alternatives are more likely options for a government. Hence, an executive that is confronted with the decision to mediate is more likely to do this as part of a coalition, if it has higher accountability in terms of institutional characteristics and elections:

Supply-side hypothesis 2: The higher the accountability of a mediator, the higher the likelihood that mediation will be attempted by a coalition of interveners.

The third factor on the supply side pertains to the capabilities of an intervener. Capabilities are essentially leverage, that is, the more capabilities, the higher the leverage of a country. This leverage can be used to gain influence over fighting parties. I derive two conclusions from that logic. On one hand, the less powerful a mediator is, the more likely that it will need additional capabilities from other states in order to mediate a conflict effectively. This should increase the probability that this country will form a coalition. Consider the example of the Intergovernmental Authority on Development’s (IGAD) mediation efforts during the war between the Sudanese government and the Sudan People’s Liberation Movement (Beber, 2010). Initial mediation attempts failed in the mid-1990s primarily because IGAD lacked substantial capabilities (El-Affendi, 2001). The situation changed for the 2005 Naivasha Agreement, however, where IGAD greatly benefited from the diplomatic weight, logistical support, and technical expertise of the USA in a multiparty mediation approach.

On the other hand, more powerful mediators should also seek to limit the size of a coalition in order to avoid the additional costs of multiparty mediation, which would evolve due to the necessary coordinative efforts with coalition members. Although additional partners might imply higher legitimacy, states generally value more their sovereignty and independent decision-making powers. Gent (2007: 1095) similarly emphasizes that ‘an intervening power may not want other states to participate, since this may prevent it from imposing its most preferred policy’. Ultimately, the likelihood of multiparty mediation should decrease with the capabilities of a mediator:

Supply-side hypothesis 3: The higher the capabilities of a mediator, the lower the likelihood that mediation will be attempted by a coalition of interveners.

Finally, countries seeking to mediate may be driven by strategic and security-related interests. Gleditsch, Salehyan & Schultz (2008), for example, demonstrate that civil wars increase the likelihood of interstate wars in/with the neighborhood of a conflict state. These ‘negative externalities’ can affect the decision-making calculus of geographically close states (Gleditsch, Salehyan & Schultz, 2008: 486f).¹⁰ More specifically, conflict countries may initiate military action against neighboring states in order to undertake cross-border counter-insurgency actions or to engage in retaliatory attacks against interventionist neighbors in the hope of coercing those into withdrawing support (Schultz, 2010). This, in turn, increases the likelihood of fighting between the conflict country and its neighboring states. As a result, the latter have strong incentives not only to prevent the spread of conflict, but also to address the source of this potential negative externality, that is, the hostility as such. In other words, conflicts raise the likelihood that surrounding states have a valid security interest in this dispute. Ultimately, this should increase their incentives to mediate:

Supply-side hypothesis 4: The closer the geographical proximity between a mediator and either belligerent, the higher the likelihood that mediation will be attempted by a coalition of interveners.

Demand-side influences

As indicated above, also the belligerents must have incentives in order to agree on (multiparty) mediation. For this, the benefits of multiparty mediation have to outweigh the costs associated with it. Hence, compared to and unlike in the case of single-party mediation, recall that Crocker, Hampson & Aall (1999, 2001a,b) stress the power aspect of multiparty interventions, that is, a coalition of mediators should have more leverage than a unilateral intervention. Accordingly, mediating coalitions may be more effective than single-party mediations as mirrored by Table I. Consequently, from the antagonists’ perspective, the basic rationale for demanding a mediating coalition instead of a single mediator is precisely the power/leverage advantage of multiparty mediation, since the prospects of success may seem higher then. And this may also outweigh the belligerents’ additional costs when seeking to pursue multiparty mediation, that is, higher coordination costs, higher heterogeneity of interests due to the involvement of more actors, or potentially higher uncertainty.

Against this background, the first argument pertaining to the demand side centers on the salience and duration of a dispute. Bercovitch & Gartner (2006), Bercovitch, Anagnoson & Wille (1991), and Quinn et al. (2006) contend that high-intensity disputes are more likely to see third-party intervention. These tough cases pose a more severe threat to both the belligerents and to the international environment – including prospective mediators.¹¹ The argument can be extended to the duration of a conflict (see Greig, 2005: 251). The longer a dispute lasts, the higher the likelihood that the belligerents’ positions are fixed, their attitudes are hardened, and the conflict cycle has escalated (see Bercovitch & Langley, 1993: 676). As a result, fighting parties may be more willing to seek a peaceful settlement of their dispute via a third party as the antagonists are less likely to solve their issues on their own. Owing to the assumption that multiparty mediation is more likely to be more effective than single-party mediations, the chances for multiparty mediation should also increase with salience and duration. This logic is mirrored by Crocker, Hampson & Aall (2001b: 62f) who state that ‘the long history of the various official and nonofficial mediations in the Middle East, the multi-track effort in Northern Ireland, and the almost overwhelming multiple-institution effort in Bosnia provide numerous … examples of multiparty efforts to intervene when violence is at its highest or the conflict appears most intractable’. I therefore suggest:

Demand-side hypothesis 1: The higher the salience of a dispute, the higher the likelihood that mediation will be attempted by a coalition of interveners.

Demand-side hypothesis 2: The longer the duration of a dispute, the higher the likelihood that mediation will be attempted by a coalition of interveners.

Second, it is likely that the regime type of belligerents influences whether conflicts see multiparty mediation or not. Drawing upon the democratic peace literature (e.g. Oneal & Russett, 1999), democracies have a common structural background and a common political culture, which gives these kinds of regimes more inclusivity and a higher level of transparency, and promotes a common understanding that induces shared norms. One of the most important norms of democracies is the peaceful resolution of conflicts. This should make democratic belligerents more likely to agree to mediation (e.g. Dixon, 1994; Raymond, 1994; Bercovitch & Houston, 2000). Other studies come to the opposite conclusion, however, that is, democracies are more likely to use bilateral negotiations for settling issues (e.g. Mitchell, 2002; Mitchell, Kadera & Crescenzi, 2008; Ellis, Mitchell & Prins, 2010). The basic argument here states that democracies are able to strike effective bargains without third-party influence on their own and more successfully than non-democracies (Ellis, Mitchell & Prins, 2010: 376). Regardless of these opposing arguments, however, all those scholars do not distinguish between single-party mediation and multiparty mediation. Since I lack an a priori reason to expect a somewhat different impact in the case of multiparty mediation, I seek to test:

Demand-side hypothesis 3: The likelihood of multiparty mediation should be equal to the likelihood of single-party mediations when the belligerents’ level of democracy increases.

Third, antagonists that approximate military parity, that is, neither side has a clear military advantage, are generally more likely to accept mediation (Mason & Fett, 1996: 550; Greig, 2005: 258).¹² Therefore, power discrepancies or asymmetric conflicts basically decrease the incentives for the most powerful antagonist to agree to a peaceful settlement as it stands a better chance of winning a conflict militarily. If we subscribe to my assumption from above (Table I), though, an asymmetric conflict that is characterized by a high degree of belligerents’ military heterogeneity should signal (to potential mediators and to themselves) that the conflict in question might be a difficult one to solve. This stems from the fact that the stronger military capabilities of one side may lead to a further escalation of the dispute. Owing to this potentially higher intensity, even the existence of either belligerent (or, additionally, of actors outside the fighting dyad) may be threatened. Therefore, I argue that asymmetric conflicts should be more likely to see multiparty mediation as the belligerents (especially the weaker disputant) and the outside parties¹³ may push for a multiparty mediation that principally has more leverage than a unilateral approach:

Demand-side hypothesis 4: The higher the power discrepancy within a dyad of belligerents, the higher the likelihood that mediation will be attempted by a coalition of interveners.

Finally, when belligerents ask for mediation, they seek to address states that could favor their interests (Crocker, Hampson & Aall, 1999, 2001a,b) or are impartial (Crescenzi et al., 2011; Gent & Shannon, 2010).¹⁴ An antagonist seeks to avoid a (mediator or a) mediating coalition that is biased toward its dyadic counterpart, because this would increase its costs, owing to a higher risk that a settlement outcome might favor its opponent. Hence, fighting parties will seek to establish a balance between mediators (Gent & Shannon, 2010: 20): the larger the bias of one mediator to a belligerent, the higher the likelihood that the opponent will ask for an additional state that represents its interest. Accordingly, the chances of multiparty mediation increase. For example, over the course of the conflict between Israel and the Palestinian Liberation Organization (PLO), the former frequently asked the USA to mediate. However, the PLO feared that the USA would merely mirror Israeli interests and, hence, also demanded that the USSR should mediate. As a result, the USA and the USSR formed mediating coalitions on various occasions, for example during the 1991 peace conference in Madrid (Choukri-Fishere, 2008). The last demand-side hypothesis, thus, argues:

Demand-side hypothesis 5: The stronger the bias of a mediator toward either belligerent, the higher the likelihood that mediation will be attempted by a coalition of interveners.

Research design

Data

I employ data from the ICOW project (Hensel & Mitchell, 2007; Hensel et al., 2008), which covers territorial claims in the Western Hemisphere and Western Europe, river claims in the previous two regions plus the Middle East, and maritime claims in the Western Hemisphere and Europe. Claims are identified according to explicit evidence of contention involving official representatives of two or more nation states over the issue type in question. The advantages of the ICOW data are twofold. First, we can compare cases across various issues of contention. Second, the data’s cases are not limited to situations that become militarized at some point, which addresses the problem of selection bias to some extent. Owing to data constraints, the temporal domain of my study is from 1950 to 2000, while the data are essentially cross-sectional. However, some temporal dependence may persist, since interventions are unlikely to be independent from previous conflict management efforts over a claim between a specific dyad. To address this problem, I cluster the standard errors on mediating coalitions¹⁵ and consider a duration variable that is specified below.

For capturing the concept of mediation, I rely on Touval & Zartman (1985). Hence, after dropping cases that experienced military interventions or bilateral conflict management, the final data include information about attempts to manage or settle each ICOW claim through peaceful third-party mediations.¹⁶ The unit of analysis comprises the ‘individual mediator’ of the ICOW data, that is, I use the set of all interveners of all ICOW disputes in 1950–2000, which have (at least) the potential to form multiparty coalitions. More specifically, I disaggregated each settlement along any third party involved and, in turn, use each single actor therein as my unit of analysis.¹⁷ This approach has three key advantages. First, I do not create an artificial overdispersion of zeros on the dependent variable, which would be the result if I considered, for example, all states in the international system. Second, it is unlikely that I exclude any potential multiparty mediator, since I consider every country that intervened so far in an ICOW claim. Also, I do not employ any sort of exclusion criteria. For example, my data include Norway as a potential (multiparty) mediator. However, a coding rule that is based on, for example, geographical proximity between the belligerents and a potential mediator would exclude Norway for Latin American and African conflicts, despite the fact that Norway has mediated in Colombia and Sudan. Finally, my approach allows me to study the characteristics of mediating states directly, that is, key features that largely pertain to the supply side of multiparty mediation.

Core variables

The dependent variable is a binary item indicating whether multiparty mediation occurred or not. As emphasized, multiparty mediation refers to instances of composite efforts only, where third parties actually form a coalition and seek to represent a single interest.¹⁸ The values of 0 comprise conflicts that saw instances of single-party mediation, but leave out conflicts without any interventionist effort. I use logistic regressions for testing my hypotheses.

Second, the supply side of multiparty mediation covers five items. A mediator’s economic situation is operationalized by GDP per capita, which is measured in 1948 US dollar value (Gleditsch, 2002). With regard to mediators’ capability, Crocker, Hampson & Aall (2001b: 59) claim that it may consist of material means to induce the parties in a dispute to change their behavior.¹⁹ I take the CINC variable from the Correlates of War data (Singer, 1988). The data on a country’s accountability stem from Bueno de Mesquita et al.’s (2004) ‘W/S’ measure, that is, the ratio of the size of a country’s winning coalition (W ) to its selectorate (S). Further, the data on election dates are taken from Goemans’s extension of the Archigos data (see Goemans, Gleditsch & Chiozza, 2009). This is a dummy variable indicating if the year of a mediation attempt sees an election. The last item on the supply side, geographical proximity, is operationalized via the capital-to-capital distance between a belligerent and a mediator. I use a weakest-link specification where the smallest distance between any antagonist and an intervener affects the chances of (multiparty) mediation.

Third, five variables refer to the demand side. The intensity/importance of an issue claim is operationalized by salience, which measures the characteristics of an issue claim. Following Hensel & Mitchell (2007), the sample’s variable ranges in [2; 12], with higher values indicating greater salience. Duration addresses the second demand-side hypothesis and any persisting temporal dependencies. As emphasized, multiparty attempts should be more likely to occur in longstanding conflicts. I incorporate a variable that measures the time elapsed from the start of an issue claim until (multiparty) mediation occurred. I also consider the belligerents’ regime types. The data are taken from Polity IV’s polity2 item (Marshall & Jaggers, 2002), relying on a weakest-link specification, where the least democratic country in a belligerent dyad influences the likelihood of (multiparty) mediation onset. For testing the fourth hypothesis on the demand side, I include the natural log of the ratio of the stronger country’s capability to that of the weaker country (Singer, Bremer & Stuckey, 1972). Finally, I follow Gent & Shannon (2010) and use S-scores of alliance portfolio similarities to measure bias (Signorino & Ritter, 1999). This S-score measure is particularly appropriate, since it reflects agreement between states over security matters (Gent & Shannon, 2010: 19). Bias is coded as the absolute value of the difference between the S-score of the third party and the target, and the S-score of the third party and the challenger in a dispute.

Control variables

I also have to control for other variables that may influence (multiparty) mediation onset, although they are not explicitly addressed by my theory. Following Greig (2005), I consider three variables. First, the number of previous mediation attempts may influence the prospects for future efforts. This is likely to apply to both single- and multiparty mediations. I use a count measure. Second, although I consider the intensity of a dispute, the salience and duration items might not fully capture influences such as fatalities over the course of a dispute. Therefore, I include a dummy item indicating whether a militarized interstate dispute is going on during a mediation attempt; and, finally, I consider whether a dispute broke out due to territorial claims, since these are likely to cause the most intense conflicts (Hensel et al., 2008).

Empirical results

Table II displays the results of my analysis. Model 1 incorporates my core variables solely, while Model 2 also includes the control items. In order to unveil the implied magnitude of my substantial findings, the second and fourth columns in Table II report predicted probabilities while holding all other variables at their mean value (or mode for dichotomous variables).

My results are generally robust and do not depend on model specifications such as dropping or including specific variables. Further, the controls actually contribute to the model fit, but this is largely driven by the impact of ongoing MID, which is positive and significant. If an issue conflict comprises a continuing MID, prospective mediators perceive this as a high-intensity signal. This increases the incentives for these actors to put more effort into their actions and, ultimately, raises the chances of multiparty mediation. The remaining control variables fail to reach any significant level. Hence, their (fairly substantial) predicted probabilities should be interpreted with caution.

The findings for the other variables support a large share of the hypotheses and clearly point to a disconnect between the predictors of single-party and multiparty mediation. Starting with the variables on the supply side, only accountability and the variable on economic performance show consistent results that match my theoretical expectations. As Table II reveals, the worse the state of a country’s economy, the higher are the incentives for this actor to pursue a strategy of burden sharing and higher legitimacy. Ultimately, this increases the chances for multiparty mediation: the likelihood that a conflict sees a mediating coalition instead of a unilateral intervention decreases on average by 78 percentage points when moving toward the maximum value of GDP per capita. The ‘W/S’ item has also the expected effect: the higher the accountability of a mediator, the more likely it is that this country will mediate as part of a coalition.

With regard to capability and election, the findings remain inconclusive. Although both variables have the expected sign, they turn out to be insignificant. The result on the election dummy can be interpreted fairly easy, however. My data comprise all kinds of countries, including autocratic mediators. Consequently, although it is unlikely to observe an election in those regimes, election is coded as 0 here, which, in turn, is likely to drive the results of this variable into insignificance.²⁰ Finally with regard to the supply side, geographical proximity has a positive and consistently significant effect. Apparently, my argument does not hold, as I obtain evidence against the claim that the smaller the distance between a potential multiparty mediator and the closest antagonist, the more likely it is that the former will perceive this as a direct and substantial threat. The predicted probability of multiparty mediation increases by about 39.5 percentage points when approaching the maximum of geographical proximity.

Table II.

The onset of multiparty mediation, 1950–2000

	Model 1 (Estimates)	Model 1 (Pred. probs.)	Model 2 (Estimates)	Model 2 (Pred. probs.)
Supply-side influences
GDP per capita	−0.0001	0.92→0.16	−0.0001	0.91→0.11
	(0.0000)***		(0.0000)***
Capability	−2.546	0.77→0.60	−1.511	0.70→0.60
	(2.650)		(2.888)
Accountability	1.219	0.52→0.78	1.173	0.45→0.73
	(0.636)*		(0.630)*
Election	0.313	0.75→0.80	0.259	0.69→0.74
	(0.340)		(0.352)
Geographical proximity	0.0005	0.63→0.97	0.0006	0.53→0.98
	(0.0002)***		(0.0002)***
Demand-side influences
Salience	0.190	0.53→0.88	0.223	0.41→0.87
	(0.109)*		(0.113)**
Duration	−0.010	0.81→0.41	−0.011	0.77→0.29
	(0.005)*		(0.007)*
Belligerent's democracy	0.063	0.58→0.83	0.091	0.42→0.82
	(0.027)**		(0.029)***
Capability ratio	0.296	0.54→0.93	0.327	0.44→0.92
	(0.091)***		(0.102)***
Bias	−2.955	0.83→0.45	−2.403	0.76→0.44
	(1.276)**		(1.310)*
Controls
Previous mediation attempts			−0.039	0.73→0.24
			(0.028)
Ongoing MID			1.403	0.69→0.90
			(0.762)*
Territorial issue			0.749	0.69→0.82
			(0.593)
Constant	−0.445		−0.925
	(0.895)		(0.949)
N	334		334
Log pseudo likelihood	−140.048		−133.532
Wald χ²	75.35***		75.43***
Pseudo R²	0.336		0.367

Robust standard errors in parentheses; columns 3 and 5 show change in predicted probabilities when moving from minimum to maximum values of each specific variable while holding all other variables at their means (or modes for binary variables).

*significant at 10%; **significant at 5%; ***significant at 1% (two-tailed).

Turning to the demand side now, consistent with the result of ongoing MID, salience has a positive and significant impact on the likelihood of multiparty mediation. The predicted probability of multiparty mediation increases by 40.5 percentage points on average. Consequently, the higher the intensity of an issue claim, the more likely it is that the belligerents and, simultaneously, prospective mediators realize that single-party efforts might not suffice for settling the conflict. This partly mirrors Table I and is in line with my core assumption, since the chances of multiparty interventions increase as expected.

Further, I argued that the longer the duration of a dispute, the higher the likelihood of multiparty mediation. The findings go against my expectation, though, as duration has a significantly negative impact on the likelihood of multiparty mediation. The probability of multiparty efforts decreases by about 44 percentage points. Hence, Touval & Zartman (1985) actually seem to be right in the case of multiparty mediation: third parties are more likely to be reluctant to intervene – and to form a mediating coalition in my case – in longstanding conflicts that have witnessed a history of mutual attack and atrocity, because the chances of mediation success could be lower in those disputes. This seems to hold despite the potentially higher leverage of multiparty mediation.

Coming to belligerents’ democracy, this variable turns out to be positive and significant in both model estimations. More substantially, the more democratic the least democratic antagonist, the higher is the likelihood of multiparty mediation. This finding does not mirror my argumentation, as I expected to see no difference between single- and multiparty interventions. Having said that, this emphasizes again a disconnect between multiparty and single-party mediation: several studies come to the conclusion that democracies are more likely to use bilateral negotiations for settling issues instead of mediation (e.g. Mitchell, 2002; Mitchell, Kadera & Crescenzi, 2008; Ellis, Mitchell & Prins, 2010). This research, however, does not distinguish between different forms of mediation and, thus, implicitly assumes that multiparty interventions should depict the same patterns as unilateral approaches. Yet, my results demonstrate that there is a significant difference between the two.

Moreover, when studying the capability of belligerents, the literature on single-party mediations suggests that there is a negative impact. Contrary to those expectations, I obtain a positive and highly significant effect. On average, the likelihood of multiparty mediations increases by about 43.5 percentage points if we move from the minimum value of capability ratio to its maximum. In other words, if belligerents are characterized by a high degree of military heterogeneity, this signals to these actors as well as to potential multiparty mediators that the conflict in question may be difficult to solve. This makes it less likely that either actor perceives single-party interventions to be sufficient for solving a conflict, and the chances for multiparty mediation increase.

Finally, I expected to see a positive impact of bias on multiparty mediation. However, we obtain a highly significant negative influence that is lowered by about 35 percentage points when moving from the minimum to the maximum value of this variable. Thus, the higher the bias of a mediator toward either antagonist, the less likely it is that we see multiparty mediation. Essentially, this may provide one of the first empirical proofs that ‘counter-balancing’ (see e.g. Gent & Shannon, 2010) does generally not occur for mediation. Consequently, the case illustrated above for the USA and the USSR seems to be an exception rather than the rule. On the contrary, if a mediator is biased toward either side, it seems that the non-favored fighting party does not have many incentives to argue for an additional mediator but perceives this intervening actor – also in the shadow of prospective mediation success – as sufficient to help settle a conflict. Arguably, this is an ad hoc explanation, but it mirrors Kydd (2003) who claims that a biased mediator is more likely to be effective. If this applies, it would virtually preclude the need for multiparty mediation. In other words, if biased mediators are trusted by one side and if they can get the belligerents to the negotiation table, fewer actors will need to intervene.

Robustness

In order to ensure the robustness of my findings, I changed a variety of model specifications and ran the estimates again. First and as indicated above, I consider all kinds of voluntary third-party interventions, which includes both ‘limited forms of mediation’ and ‘heavy mediations’ such as arbitration and adjunction (Rauchhaus, 2006: 224). In order to rule out the possibility that my results are driven by the latter category, I estimated a model that only considers unbinding third-party interventions. As shown in Model 3, however, my findings virtually stay the same, but duration and ongoing MID only approach statistical significance.

Second, the somewhat surprising findings on geographical distance and bias could partially be explained by including ‘multilateral negotiations’ and ‘peace conferences’ in my analysis. In more detail, these types of interventions comprise, for example, cases like the Antarctica negotiations where a lot of different (and geographically distant as well as biased) states were involved. Although ‘peace conferences’ are dropped from my original analysis already because of missing values, I excluded ‘multilateral conferences’ and estimated a model with bias and geographical proximity. Table III demonstrates, however, that this does not influence my results.

Table III.

The onset of multiparty mediation, 1950–2000: robustness

	Model 3 (Unbinding)	Model 4 (No multi. conf.)	Model 5 (Outcome eq.)	Model 5 (Selection eq.)
Supply-side influences
GDP per capita	−0.0001		−0.0001
	(0.000)***		(0.000)***
Capability	−1.738		−1.531
	(2.997)		(1.606)
Accountability	1.270		0.800
	(0.635)**		(0.304)***
Election	0.331		0.154
	(0.368)		(0.199)
Geographical proximity	0.001	0.001	0.0002
	(0.000)***	(0.000)***	(0.000)***
Demand-side influences
Salience	0.237		0.110	0.018
	(0.117)**		(0.039)***	(0.023)
Duration	−0.010		−0.004	−0.001
	(0.007)		(0.002)	(0.001)
Belligerent's democracy	0.072		0.029	−0.016
	(0.030)**		(0.013)**	(0.007)**
Capability ratio	0.357		0.158	0.152
	(0.101)***		(0.048)***	(0.022)***
Bias	−2.758	−3.689	−1.576
	(1.298)**	(1.860)**	(0.613)**
Controls
Previous mediation attempts	−0.040			−0.008
	(0.030)			(0.004)*
Ongoing MID	1.114			0.632
	(0.793)			(0.231)***
Territorial issue	0.630			−0.073
	(0.602)			(0.126)
Constant	−1.096	−0.418		−0.783
	(0.972)	(0.316)		(0.180)***
N (censored)	321	159	913	(592)
Log pseudo likelihood	−127.794	−101.690	−676.89
Wald χ²	77.45***	8.67**	197.65***
ρ			−0.445

Robust standard errors in parentheses.

*significant at 10%; **significant at 5%; ***significant at 1% (two-tailed).

Third, a more intuitive design for testing which type of third-party management is selected might be a two-stage process, that is, a first stage where the actors decide between bilateral negotiations and any third-party intervention, and a second stage where we distinguish between single- and multiparty mediations (see Heckman, 1979).²¹ For this estimation, I use a Heckman-type probit model with sample selection. The unit of analysis stays the same as in Table II, although I now also include issue claims that do not see any third-party involvement as observations in my data.²² Model 5 summarizes my findings for this and, as demonstrated, the results hardly change. Note, however, that belligerents’ democracy has a negative sign in the first stage, which mirrors, for example, Mitchell (2002), Mitchell, Kadera & Crescenzi (2008), and Ellis, Mitchell & Prins (2010) who find that democratic belligerents are more likely to use bilateral negotiations.

Conclusion

What factors influence the onset of multiparty mediation? This article has sought to increase our understanding of this phenomenon by a quantitative analysis. Based upon theoretical arguments from the previous mediation and coalition-formation literatures, this study not only identified the scale of impact of various items on multiparty mediation, but also emphasized that we have to crucially differentiate between single- and multiparty efforts. Generally, this conclusion may have a significant impact on how both policymakers and academics should treat multiparty mediation in the future.

Table IV.

Summary results

Supply-side influences
GDP per capita	yes
Capability	no
Accountability	yes
Election	no
Geographical proximity	no
Demand-side influences
Salience	yes
Duration	no
Belligerent’s democracy	no
Capability ratio	yes
Bias	no

Support for specific hypothesis is signified by ‘yes’; no support for specific hypothesis is signified by ‘no’.

First, my study emphasizes the benefits (advantages) and costs (disadvantages) of multiparty mediations for policymakers. On one hand, a coalition of mediators is likely to increase the legitimacy of an intervention and to improve the burden sharing of associated costs. As a result, if state leaders have an interest in mediating but might lack legitimacy or the economic resources for doing so, then the formation of a coalition may appear to be a more suitable strategy. On the other hand, there are also disadvantages associated with multiparty mediation. States may face the more difficult cases here, or they might signal to other actors that their economy is weak or that they highly value the accountability of their regime type. Depending on a specific bargaining environment, for example trade negotiations, this may signal a certain degree of ‘weakness’.

Coming to the scholarly implications, both my theory and the empirical results point to a disconnect between those variables that drive single-party mediation and those that influence the onset of mediating coalitions. I argued for this disconnect because of my theoretical framework based upon a core assumption (Table I), and the results of those items are summarized in Table IV. The belligerents’ capabilities and their level of democracy, for instance, are obvious examples. Unlike much previous work, therefore, future research has to take into account those differences in order to avoid misleading inferences.

Finally, mediation is not a random set – regardless of whether we look at single- or multiparty mediation efforts. I tried to address this issue thoroughly in the robustness section, and my research might also help to further mitigate the problem of selection bias, since I identified significant differences between unilateral approaches and the factors that drive multiparty mediations. This point in particular might be useful to future studies that explicitly examine the effectiveness of (multiparty) mediation (see Böhmelt, 2011).

Footnotes

Replication data

The dataset, codebook, and do-files for the empirical analysis in this article can be found at .

Acknowledgements

I am indebted to Kyle Beardsley, Holger Schmidt, and the participants of the Folke Bernadotte Academy’s Conflict Prevention Workshop (New York, 21–22 January 2011) for valuable suggestions on an earlier draft. A version of this article has also been presented at the Annual Meeting of the European Peace Science Society in Amsterdam (28–30 June 2010) and the Annual Meeting of the American Political Science Association, Washington DC (2–5 September 2010). Finally, I thank JPR’s Associate Editor Mats Hammarström and the anonymous reviewers for remarkably detailed comments that enabled me to improve the manuscript.

1

Throughout this article, I use the terms ‘mediation’ and ‘intervention’ interchangeably.

2

From a more general perspective, : 330ff) stresses crucial differences between uni- and multilateral approaches and argues that the latter might be more efficient than the former.

3

This treatment is in line with : 740) who identifies ‘a consistent strategy’ as a core characteristic of any multilateral approach. Also note that I focus only on states’ multiparty mediation because of the limited availability of data. The operationalization of multiparty mediation is discussed below.

4

The previous research on single-party mediations has shown, however, that this is unlikely to be the case (Kleiboer, 1996; Greig, 2005).

5

See for a counter-argument, where he contends that states are more likely to join military interventions the less similar their preferences with intervening actors.

6

Again, ‘sharing similar goals’, the ‘willingness to work together’, or ‘maintaining coherence’ are factors that solely pertain to multiparty mediation and, hence, there are real and important differences between multiparty and single-party interventions.

7

Martin (1992) also emphasizes a leverage aspect when studying the effectiveness of multilaterally enforced sanctions. Mansfield & Reinhardt (2003: 838) highlight the same for preferential trade agreements.

8

I employ two variables for this table. The first, mediation effectiveness, relies on ICOW’s settlement attempt outcome items, where I aggregate the different categories. The second, size of mediating coalition, simply disaggregates mediation attempts by size.

9

The selectorate is the set of people who have the ability to choose a country’s leader, while the winning coalition signifies the portion of the selectorate that keeps a leader in power.

10

In a similar, albeit different manner, Gleditsch, Salehyan & Schultz (2008: 487) highlight that conflicts can create new tensions because of spillovers from the fighting (see also Sandler, 2004: 201). For example, conflicts may induce mass refugee migration that poses severe challenges to neighboring states, since refugee flows can be an economic burden and cause political instability.

11

Note that negative externalities may be caused by the salience of a dispute. Hence, they should be considered as determinants on their own, which I argued above. Admittedly, however, the mechanisms of salience and negative externalities are similar, although the latter pertain to the supply side.

12

See also Hansen, Mitchell & Nemeth (2008), who find that states are more likely to turn to international organizations as mediators in situations of power parity.

13

Note that especially outside parties, that is, potential mediators will try to ‘enforce’ a multilateral intervention under those circumstances, because it is in their interest to prevent negative externalities. Arguably, therefore, this factor pertains to some extent to the supply side as well.

14

The bias literature is more comprehensive on the mediation effectiveness issue. Kydd (2003) argues that a mediator must be biased toward a belligerent if it is to communicate with it honestly. Smith & Stam (2003) contend that biased mediators are not effective, and Rauchhaus (2006) obtains evidence that although biased mediators may be effective, impartial mediators are even more so. Finally, discovers that if a mediator is to help resolve a conflict caused by mutual mistrust, it must be unbiased.

15

Also, not all mediating coalitions (or single-party mediations) are unique. Thus, it may be that the very same coalition intervenes in different conflicts (or even the same) at different points in time. Due to this reason, clustered standard errors control for intra-group correlations as well.

16

This basically treats mediation in the broadest possible way by including all kinds of voluntary third-party interventions, that is, good offices, inquiries, conciliations, ‘classical’ mediations, arbitrations, adjunctions, multilateral negotiations, and international peace conferences. A potential problem with that approach, however, is that it considers both ‘limited forms of mediation’ and ‘heavy mediations’ such as arbitration and adjunction, which employ more forceful means (Rauchhaus, 2006: 224). I will return to this issue in the robustness section.

17

This rule aligns with Bercovitch & Schneider (2000), for example.

18

ICOW does not explicitly use coordination as a coding rule for determining if multiparty mediation occurred. Thus, I rely on the following rationale for identifying multiparty mediations of the composite category: I only code those interventions where (a) more than one actor intervened, and (b) a single and unique conflict management strategy was employed. Because of the latter requirement, especially, I infer that the actors of a settlement attempt to pursue a common strategy and that they seek to represent a single interest.

19

An example illustrates this. The initial mediation efforts of the Contact Group mainly failed due to Bosnian Serb unwillingness to agree to the terms and the imposed embargo on the Drina River. The military operation ‘Deliberate Force’, which bombarded the Bosnian Serb military, was then intended to bring the Bosnian Serbs back to the negotiation table.

20

More specifically, I believe that the result is likely to stem from the coding of this item, that is, even countries that never (or are unlikely to) see an election at all (such as most autocracies) are coded as 0. However, treating those cases as ‘missing’ might be more appropriate since – and unlike those countries that have elections – there is not even a chance that the value of election will change to 1.

21

I thank the anonymous reviewers for this suggestion.

22

An unreported robustness check employs a specific ICOW settlement attempt as the unit of analysis. Although this excludes the ability to code individual mediator variables, the findings for the remaining items basically remain the same.

References

Ban

Ki-moon

(2009) Report of the Secretary-General on enhancing mediation and its support activities. United Nations Security Council Document S/2009/189.

Beardsley

Kyle

(2008) Agreement without peace? International mediation and time inconsistency problems. American Journal of Political Science 52(4): 723–740.

Beardsley

Kyle

(2010) Pain, pressure, and political cover: Explaining mediation incidence. Journal of Peace Research 47(4): 395–406.

Beardsley

Kyle

Greig

Michael

(2009) Disaggregating the incentives of conflict management: An introduction. International Interactions 35(3): 243–248.

Beber

Bernd

(2010) The (non-) efficacy of multiparty mediation in wars since 1990. Unpublished manuscript.

Bennett

Andrew

Lepgold

Joseph

Unger

Danny

(1994) Burden sharing in the Persian Gulf war. International Organization 48(1): 39–75.

Bercovitch

Jacob

Gartner

Scott

(2006) Is there method in the madness of mediation? Some lessons for mediators from quantitative studies of mediation. International Interactions 32(4): 329–354.

Bercovitch

Jacob

Houston

Allison

(2000) Why do they do it like this? An analysis of the factors influencing mediation behavior in international conflicts. Journal of Conflict Resolution 44(2): 170–202.

Bercovitch

Jacob

Langley

Jeffrey

(1993) The nature of the dispute and the effectiveness of international mediation. Journal of Conflict Resolution 37(4): 670–691.

10.

Bercovitch

Jacob

Schneider

Gerald

(2000) Who mediates? The political economy of international conflict management. Journal of Peace Research 37(2): 145–165.

11.

Bercovitch

Jacob

Theodore Anagnoson

Wille

Donnette L

(1991) Some conceptual issues and empirical trends in the study of successful mediation in international relations. Journal of Peace Research 28(1): 7–17.

12.

Böhmelt

Tobias

(2011) Disaggregating mediations: The impact of multiparty mediation. British Journal of Political Science 41(4): 859–881.

13.

Bruce

Bueno de Mesquita

Smith

Alastair

Siverson

Randolph

Morrow

James

(2004) The Logic of Political Survival. Cambridge, MA: MIT Press.

14.

Choukri-Fishere

Ezzedine

(2008) Against Conventional Wisdom: Mediating the Arab–Israeli Conflict. Olso: Center for Humanitarian Dialogue. Available at: http://www.hdcentre.org/files/Ezzedine%20Choukri-Fishere%20Middle%20East%20WEB.pdf

15.

Crescenzi

Mark

Kadera

Kelly

Mitchell

Sara McLaughlin

Thyne

Clayton

(2011) A supply side theory of mediation. International Studies Quarterly 55(4): 1069–1094.

16.

Crocker

Chester A

Hampson

Fen O

Aall

Pamela

, eds (1999) Herding Cats: Multiparty Mediation in a Complex World. Washington, DC: United States Institute of Peace Press.

17.

Crocker

Chester A

Hampson

Fen O

Aall

Pamela

(2001a) Is more better? The pros and cons of multiparty mediation. In: Crocker

Chester A

Hampson

Fen O

Aall

Pamela

(eds) Turbulent Peace: The Challenges of Managing International Conflict. Washington, DC: United States Institute of Peace Press, 497–513.

18.

Crocker

Chester A

Hampson

Fen O

Aall

Pamela

(2001b) A crowded stage: Liabilities and benefits of multiparty mediation. International Studies Perspectives 2 (1): 51–67.

19.

Dixon

William

(1994) Democracy and the peaceful settlement of international conflict. American Political Science Review 88(1): 14–32.

20.

El-Affendi

Abdelwahab

(2001) The impasse in the IGAD peace process for Sudan: The limits of regional peacemaking? African Affairs 100 (401): 581–599.

21.

Ellis

Glynn

Mitchell

Sara McLaughlin

Prins

Brandon

(2010) How democracies keep peace: Contextual factors that influence conflict management strategies. Foreign Policy Analysis 6(4): 373–398.

22.

Gent

Stephen

(2007) Strange bedfellows: The strategic dynamics of major power military interventions. Journal of Politics 69(4): 1089–1102.

23.

Gent

Stephen

Shannon

Megan

(2010) The effectiveness of international arbitration and adjudication: Getting into a bind. Journal of Politics 72(2): 366–380.

24.

Gleditsch

Kristian Skrede

(2002) Expanded trade and GDP data. Journal of Conflict Resolution 46(5): 712–724.

25.

Gleditsch

Kristian Skrede

Salehyan

Idean

Schultz

Kenneth

(2008) Fighting at home, fighting abroad: How civil wars lead to international disputes. Journal of Conflict Resolution 52(4): 479–506.

26.

Goemans

Henk

Gleditsch

Kristian Skrede

Chiozza

Giacomo

(2009) Introducing Archigos: A dataset of political leaders. Journal of Peace Research 46(2): 269–283.

27.

Greig

Michael

(2005) Stepping into the fray: When do mediators mediate? American Journal of Political Science 49(2): 249–266.

28.

Greig

Michael

Regan

Patrick

(2008) When do they say yes? An analysis of the willingness to offer and accept mediation in civil wars. International Studies Quarterly 52(4): 759–781.

29.

Hansen

Holly

Mitchell

Sara McLaughlin

Nemeth

Stephen

(2008) IO mediation and interstate conflicts: Moving beyond the global versus regional dichotomy. Journal of Conflict Resolution 52(2): 295–325.

30.

Heckman

James

(1979) Sample selection bias as a specification error. Econometrica 47(1): 153–161.

31.

Hensel

Paul

Mitchell

Sara McLaughlin

(2007) Issues Correlates of War Project: User manual for ICOW data, Version 1.1. Available at: http://www.icow.org

32.

Hensel

Paul

Mitchell

Sara McLaughlin

Sowers

Thomas

Thyne

Clayton

(2008) Bones of contention: Comparing territorial, maritime, and river issues. Journal of Conflict Resolution 52(1): 117–143.

33.

Keohane

Robert

(1990) Multilateralism: An agenda for research. International Journal 45 (4): 731–764.

34.

Kleiboer

Marieke

(1996) Understanding success and failure of international mediation. Journal of Conflict Resolution 40(2): 360–389.

35.

Kydd

Andrew

(2003) Which side are you on? Bias, credibility, and mediation. American Journal of Political Science 47(4): 597–611.

36.

Kydd

Andrew

(2006) When can mediators build trust? American Political Science Review 100(3): 449–462.

37.

Mansfield

Edward

Reinhardt

Eric

(2003) Multilateral determinants of regionalism: The effects of GATT/WTO on the formation of preferential trade agreements. International Organization 57(4): 829–862.

38.

Marshall

Monty

Jaggers

Keith

(2002) Polity IV project: Political regime characteristics and transitions, 1800–2002. Dataset user’s manual. College Park, MD: University of Maryland.

39.

Martin

Lisa

(1992) Coercive Cooperation: Explaining Multilateral Economic Sanctions. Princeton, NJ: Princeton University Press.

40.

Mason

David

Fett

Patrick

(1996) How civil wars end: A rational choice approach. Journal of Conflict Resolution 40(4): 546–568.

41.

Mitchell

Sara McLaughlin

(2002) A Kantian system? Democracy and third-party conflict resolution. American Journal of Political Science 46(4): 749–759.

42.

Mitchell

Sara McLaughlin

Kadera

Kelly

Crescenzi

Mark

(2008) Practicing democratic community norms: Third party conflict management and successful settlements. In: Bercovitch

Jacob

Gartner

Scott

(eds) Empirical Studies in International Mediation. New York: Routledge, 243–264.

43.

Oneal

John

Russett

Bruce

(1999) The Kantian peace: The pacific benefits of democracy, interdependence, and international organizations, 1885–1992. World Politics 52(1): 1–37.

44.

Pahre

Robert

(1994) Multilateral cooperation in iterated prisoner’s dilemma. Journal of Conflict Resolution 38(2): 326–352.

45.

Quinn

David

Wilkenfeld

Jonathan

Smarick

Kathleen

Asal

Victor

(2006) Power play: Mediation in symmetric and asymmetric international crises. International Interactions 32(4): 442–470.

46.

Raymond

Gregory

(1994) Democracies, disputes, and third-party intermediaries. Journal of Conflict Resolution 38(1): 24–42.

47.

Rauchhaus

Robert

(2006) Asymmetric information, mediation, and conflict management. World Politics 58(2): 207–241.

48.

Sandler

Todd

(2004) Global Collective Action. Cambridge: Cambridge University Press.

49.

Schultz

Kenneth

(2010) The enforcement problem in coercive bargaining: Interstate conflict over rebel support in civil wars. International Organization 64(2): 281–312.

50.

Signorino

Curtis

Ritter

Jeffrey

(1999) Tau-b or not Tau-b: Measuring the similarity of foreign policy positions. International Studies Quarterly 43(1): 115–144.

51.

Singer

David

(1988) Reconstructing the Correlates of War dataset on material capabilities of states, 1816–1985. International Interactions 14(2): 115–132.

52.

Singer

David

Bremer

Stuart

Stuckey

John

(1972) Capability distribution, uncertainty, and major power war, 1820–1965. In: Russett

Bruce

(ed.) Peace, War, and Numbers. Beverly Hills, CA: Sage, 19–48.

53.

Smith

Alastair

Stam

Allan

(2003) Mediation and peacekeeping in a random walk model of civil and interstate war. International Studies Review 5(4): 115–135.

54.

Tago

Atsushi

(2005) Determinants of multilateralism in US use of force: State of economy, election cycle, and divided government. Journal of Peace Research 42(5): 585–604.

55.

Touval

Saadia

Zartman

William

(1985) Introduction: Mediation in theory. In: Touval

Saadia

Zartman

William

(eds) International Mediation in Theory and Practice. Boulder, CO: Westview, 7–17.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB