Three degrees of Success in International Mediation

The Elusive Nature of Success

Despite an impressive body of literature on the topic of international mediation, thus far the academic community has been unable to reach consensus regarding a comprehensive definition that would unequivocally delineate how to recognise success in international mediation. Even when scholars use somewhat similar definitions, their analyses seem to discuss essentially quite different aspects of success. Academic conceptualisations of success suffer from definitions that appear to be quite arbitrary, developed on a case-to-case basis, and not based upon a solid theoretical foundation. ³ This conceptual confusion regarding success in international mediation should not come as a surprise. On the one hand, international mediation processes are not uniform and, as such, it is quite difficult to establish a one-size-fits-all set of criteria with which to assess achievement of the various objectives of mediation. Parties’ objectives and goals tend to change throughout the process, so measuring mediation success in terms of parties’ goals might make it impossible to formulate a consistent conceptualisation of mediation success. ⁴ On the other hand (and more importantly), so far empirical studies have overwhelmingly treated mediation outcomes – both success and failure – as dependent variables. The focus has rarely been on how to measure success, but rather on which factors have the ability to affect the mediation outcome. In line with this approach, numerous studies have relied on analytical frameworks ⁵ in which mediation outcomes are represented as a function of large sets of factors related to the context (e.g. nature of the mediator, nature of the parties and nature of the dispute) and to the mediation process (mediator behaviour). While such approaches denote the complex interactions of a large number of factors, they still fail to provide a comprehensive set of tools with which to define and measure success. With all these analytical shortcomings in mind, how can we be certain that an outcome is a success? In other words, how do we evaluate success?

The bulk of existing studies – especially those that have explored correlation between a variety of factors and mediation outcomes – have treated success as a dichotomous phenomenon, usually linked to the existence of a signed agreement. In other words, mediation was successful only if parties reached a formal agreement. The fact that parties accepted mediation in the first place and issues related to agreement implementation were not part of their understanding of mediation success. ⁶ In International Mediation Greig and Diehl challenge this analytical oversimplification and suggest that the definition of success should correspond to different stages of the mediation process. ⁷ The first stage focuses on getting the disputing parties to accept mediation, which they refer to as ‘getting to the table’. ⁸ The second stage of mediation is related to the achievement of a formal agreement, regardless of its scope: whether it is a simple cease-fire or a comprehensive agreement it just needs to be a product of a mediation process. The final stage of mediation is the implementation phase. In this phase mediation success is directly linked to the durability of mediated settlements. According to Greig and Diehl, the stages are not identical and are only partly linked because ‘some of the same factors and processes that encourage actors to seek mediation also influence their willingness to come to an agreement’. ⁹ As such, each stage of mediation and consequent degree of success deserves a separate conceptualisation.

Getting the Parties to the Table

Making the parties amenable to mediation is not an easy task. In order for mediation to take place, parties need to perceive it as a reasonable alternative to their belligerent activities. Parties could use mediation as a convenient political cover for making unpopular decisions, such as making necessary concessions to the other side, which would be unimaginable without the presence of a third party. Nevertheless, accepting mediation also generates reputational costs (both domestic and international) for the conflicting parties. The parties may be apprehensive that their willingness to negotiate with ‘the enemy’ might be perceived as a sign of weakness and even labelled as treason by their constituencies. In light of such costs, Greig and Diehl argue that ‘the willingness of parties to negotiate with one another and accept mediation can develop as the costs of conflict between two sides mount (“pain”) and diplomatic approaches to settle the conflict grow more appealing than continued violence (“promise”)’. ¹⁰ The notions of ‘pain’ and ‘promise’ come from Zartman’s ripeness theory. ¹¹ ‘Pain’ refers to the perceptual condition of ‘mutually hurting stalemate’ (MHS), in which the parties are at an impasse and neither one is able to escalate the conflict to victory, making the continuation of the existing conflict both harmful and costly to each side. The second element – ‘the promise’ – is also perceptual and highlights the attractiveness of mediation as a ‘way out’ of the impasse. According to ripeness theory, both ‘pain’ and ‘promise’ are ‘necessary yet insufficient’ ¹² conditions for a successful peacemaking process. In order to ripen the conflict for resolution, mediators need to help parties perceive and understand how unbearable their stalemate is. Therefore, success at this stage also depends on the mediators’ willingness and ability to push the parties to commit to the peacemaking process and reach a mutually acceptable agreement.

As stated previously, conflicting parties may accept mediation without any interest in reaching a peaceful settlement to their dispute. According to Greig and Diehl, when mediators face these ‘devious objectives’, ‘unless the third party can provide sufficient incentives to the two sides once they get to the bargaining table to make them more amenable to a settlement, the mediation effort will fail’. ¹³ Incentives may take the form of tangible carrots and sticks as well as different intangible inducements such as improved international reputation or enhanced relations with the third party. Mediation of international conflicts may be conducted by a variety of actors – states, international and regional organisations, NGOs – either individually or through an ad hoc multiparty effort and providing the necessary incentives is certainly a costly endeavour for all of them. Greig and Diehl highlight three general reasons why third parties are willing to get engaged in managing an international dispute: humanitarian concerns, national interests and third-party’s organisational predisposition. ¹⁴ According to their analysis, in the post-World War II period mediation efforts were ‘at least partially motivated by humanitarian concerns’, as slightly more than 13 per cent of all mediation efforts occurred in conflicts that witnessed a genocide. ¹⁵ They find that ‘many of the conflicts with the worst humanitarian emergencies are ignored by major-power mediators, leaving their mediations to non-major powers and non-state actors, if those conflicts are mediated at all’. ¹⁶

Addressing humanitarian concerns may not generate direct pay-offs for third parties. Therefore, it is plausible to presume that mediators are driven by their self-interests just as much as by humanitarian impulses. The list of self-interested reasons that motivate third parties to act as mediators is long. Mediation is a useful, cost-efficient foreign policy tool that can be used to advance state’s interests on the international level. Major powers may use it to expand their sphere of influence, at the same time reducing the influence of other rival states. Non-major power states may use mediation to contain and reduce negative externalities from neighbouring conflicts or to improve their reputation and credibility on the international level. The reputational incentive is also a strong motivator for international organisations, whose founding charters might describe peacemaking as their raison d’etre.

Since mediators might use mediation to promote their interests, it is not a surprise that conflicting parties are inclined to resist mediation by specific mediators. Mediation is a voluntary and non-binding process after all and, as such, third parties need to be perceived as acceptable by conflicting parties. In fact, Beardsley finds that:

When neighbouring and powerful states in the international system have strong interests in a crisis, the disputants actually resist mediation … this is understandable because many disputants prefer to limit meddling by external actors wishing to shape the regional and global orders in their favor. ¹⁷

Reflecting on this finding, Beardsley develops his core argument regarding the long-term effects of mediation (as will be discussed later in this article). He argues that highly intrusive mediation activities conducted by powerful mediators are likely to generate highly unstable results in the long run and, therefore, disputants tend to be hesitant about allowing ‘very eager third-parties’ to mediate their conflict. ¹⁸

Parties’ willingness to accept mediation and consequently to commit to the process is strongly related to mediator’s characteristics. Greig and Diehl note that ‘disputants desire a mediator that is perceived to be “even-handed,” that is willing to broker an agreement fairly’. ¹⁹ Although related, the concept of fairness should not be mistaken for mediator neutrality. Since international actors may have a vested interest in being involved in managing a conflict, their role will inevitably be affected by their bias. A biased mediator is useful insomuch as it is able to move the party to accept an agreement that it would never accept unilaterally. ²⁰ In order to produce such a move, mediators need to possess particular leverage that could create the necessary incentives aimed at altering parties’ conflictual behaviour. In one of his earlier works, Beardsley argued that weak mediators are more likely to be requested by disputants that want to use mediation insincerely, when they are actually not interested in reaching an agreement. ²¹ Parties that select weak mediators might be more inclined to stall the process to buy time for rearming or regrouping on the battlefield or to enable them to resist making concessions that, although necessary, might be quite unpopular at the domestic level.

In Mediation Dilemma, Beardsley contrasts mediation with and without leverage in order to understand their relative impact on the bargaining environment. On the one hand, mediation with leverage or ‘heavy mediation’, ‘tries to maximize the cost of nonagreement and thereby expand the set of mutually acceptable alternatives’. ²² According to this perspective, leverage can take the forms of positive inducements for reaching an agreement and negative inducements against belligerent activities. On the other hand, mediation without leverage or ‘light mediation’, ‘primarily involves enabling the actors to find an agreement within the existing set of alternatives that are mutually preferable to conflict’. ²³ In such circumstances, mediators are more focused on reducing parties’ bargaining uncertainties by facilitating information exchange, helping them formulate potential alternatives, and by providing the necessary political cover for unpopular decisions. In his differentiation of light and heavy mediation, Beardsley treats the concept of leverage as being essentially resource-based. This reductionist take leads him to overlook all other types of social power that could still be used to leverage the parties toward a mutually acceptable agreement. According to Carnevale, power in mediation – the ability or willingness to influence another party to achieve a goal – is a much broader concept that encompasses two distinct forms: resource-based (‘strategic’) and behavioural (‘tactical’). ²⁴ Resource-based strategic strength is what a party brings to the table, while tactical strength is what a party does at the table. So in addition to ‘reward’ and ‘coercive’ powers or ‘carrots and sticks’ (types that Beardsley equates with power), there are also other forms of resource-based power that a mediator might possess, including legitimacy, expertise, or prestige. Even ‘information power’ can be treated as a form of resource-based leverage because ‘the mediator may provide information that makes compliance with the mediator’s request seem rational’. ²⁵ Beardsley’s reductionist conceptualisation of power strongly conditions his main argument about the long-term effects of mediation, both with and without leverage.

Beardsley’s argument is directly related to the standard classification of mediation strategies – communication-facilitator, procedural-formulator, and directive-manipulator, ²⁶ the latter also referred to as ‘heavy’ or ‘mediation with muscle’ – based on mediator’s assertiveness and level of intrusiveness in the process. Quantitative studies, such as Beardsley’s, provide important insights regarding the correlation between such strategies and mediation outcomes. However, they do not tell us much about the complexities of potential causal mechanisms that exist throughout the mediation process. They do not account as much for various forms of ‘tactical strength’ ²⁷ that a mediator might use throughout the process, such as developing momentum between the parties and using ‘relational power’ (i.e. bias) with one of the parties to induce a change in its behaviour.

Svensson and Wallensteen recognise the importance of such forms of leverage, and consequently develop a completely new taxonomy of four mediation styles which allows them to observe the changes throughout various stages of the process, from the onset of mediation until an agreement has been signed. While not refuting the significance of the existing notions of mediation strategies – which in their taxonomy of styles are referred to as ‘methods’ of mediation; others being the ‘scope’ (inclusiveness or exclusiveness of the process), the ‘mode’ (processes’ level of transparency), and the ‘focus’ (short-term end of hostilities or a comprehensive peace that also upholds the principles of justice and human rights) – the authors also highlight three main ways in which mediation can move forward: mandates, resources and outcomes of mediation. ²⁸

Every mediation process starts with a mandate – an aspect of this topic that has been surprisingly overlooked by the academic community. Regardless of the fact of whether a mandate derived from the belligerents or from an international actor, ‘[the] mandate directs the way a third party enters a conflict, affects the mediation process, involves the international community, and finishes the assignment’. ²⁹ More importantly, mandates are the first indicator of a mediator’s objectives. Insights from Jan Eliasson’s experience show how practitioners and academics have somewhat different expectations in terms of what the purpose of mediation is. Svensson and Wallensteen note that ‘mediation does not always result in complete and durable peace agreements between the parties, [as] such an outcome may not even be the purpose of mediation’. ³⁰ In fact, Eliasson’s objectives generally included goals such as improvement and maintenance of communication channels between conflicting sides, alleviation of humanitarian crisis, and exploration of elements that could be used for a final agreement.

According to the authors, ‘bringing the parties to the table is an achievement in itself. The mediator then has to keep them there.’ ³¹ In order to keep parties committed to the process, mediators may employ not only tangible (and intangible) incentives, but also particular tactical moves to preserve the momentum. For instance, Eliasson used to ‘wear down’ the parties by keeping them at the table for many hours. ³² At the same time, a mediator may examine how credible a party’s threat to walk away from the negotiations is and emphasise the benefits of a negotiated deal. Mediators may also help the parties reframe the issues and persuade them to believe that they are close to achieving an agreement. ³³ Nevertheless, since mediation is an inherently voluntary and non-binding process, ‘a mediator cannot force the parties to end a conflict when one or other does not want to end it‘ ³⁴ Therefore, according to Svensson and Wallensteen, in order to determine how best to measure outcomes and success, it is not enough to ‘narrowly focus’ on cease-fires and peace agreements, but ‘it may be necessary to incorporate more nuanced understandings of the operational objectives of international mediators, and particularly, those of the parties to the conflict’. ³⁵ The present discrepancy between practitioners’ experiences and academic knowledge of the mediation process is evidently a result of different perspectives and expectations of what a mediator is supposed (or even able) to achieve throughout the process. A more careful categorisation of mediator’s objectives – which may be related to their direct mandates – could be a suitable starting point from which to better understand how to operationalise the first degree of success in mediation: ‘getting the parties to the table’ and making them commit to the process.

Reaching an Agreement and Making it ‘Stick’

Reaching an agreement is nothing short of a true accomplishment for any mediator. For Greig and Diehl, while ‘the assistance of a willing mediator and the consent of the conflicting sides to sit down and talk is an important initial step toward peace, it is by no means a guarantee that a settlement can be reached and peace established’. ³⁶ Their scepticism toward the relevance of the first degree of success is based upon the presence of ‘devious objectives’ the parties might have when entering the mediation process. Therefore, for them ‘the ultimate value of an agreement is manifest only if the settlement is actually implemented and endures for a significant period of time’. ³⁷ In other words, the true challenge of any mediation activity is to prevent the conflict from relapsing into violence. Mediators may use agreements to create specific provisions that deter defection and reduce uncertainties between belligerents. According to Beardsley, the formalised nature of the agreement is relevant both theoretically and practically. They are easily identifiable and are able to ‘better indicate a successful bargaining outcome because of the effort needed to negotiate the specific terms, define what compliance looks like, and decide on any enforcement provisions’ ³⁸

Beardsley finds that mediated conflicts are more likely to achieve formal agreements than those that are not mediated. ³⁹ Depending on their primary scope and ability to address the conflict’s underlying issues, agreements may take different forms. The least comprehensive form is cease-fire agreements. As their primary aim is to stop the violence, they represent the first step in creating the momentum required to produce a more comprehensive settlement. However, while they are the simplest form of agreement to achieve, they are also the easiest to break. Due to the fragility of cease-fire agreements, Greig and Diehl argue that it is quite difficult to label their achievement as successful, ⁴⁰ especially when they are not followed by more far-reaching settlements. Partial or robust agreements contain provisions that reflect a greater degree of convergence between the parties. With them, the parties indicate at least a formal willingness to alter their belligerent behaviour and agree on specific measures that could solidify peaceful relations.

In order to achieve these agreements, mediators often face the challenges of inducing concessions from the parties. While the presence of a mediator increases the likelihood of concessions, Beardsley also finds that despite the useful political cover they might offer, mediators still fail to induce major concessions from belligerents. According to him, this finding supports the notion that ‘mediators can interfere with the natural trajectory of settlement and promote outcomes that are not self-enforcing’. ⁴¹ Therefore, the true ‘mediation dilemma’ is not whether mediation should be conducted, but how involved the mediators should be throughout the process. ⁴² Beardsley finds that mediated agreements are more stable in the short run – which in his analysis are treated both as a period of two years and five years – than agreements that are reached without the presence of a mediator. ⁴³ However, mediation may produce negative effects in the long term for three reasons: the artificiality of external leverage, the obfuscation of the future bargaining environment, and parties’ insincere motives. ⁴⁴ For Beardsley, this is especially true when the process is conducted by a powerful mediator, as ‘the increased use of third-party leverage will lead to ever-stronger short-term outcomes – more formal agreements, concessions, and lulls in hostilities – and ever-weaker long-term stability’. ⁴⁵ Accordingly, ‘heavy mediation’ produces better pay-offs in the short run, but generates long-term instability, while less intrusive strategies (facilitation and formulation) are less risky in the long run, but are still less effective in the short term. ⁴⁶ Although extremely insightful, Beardsley’s quantitative analysis is unquestionably conditioned by three empirical oversimplifications. On the one hand, there is the aforementioned reductionist categorisation of leverage (or power) that includes only resource-based forms of social power. On the other hand, quantitative analyses rely on data sets that treat mediation strategies as uniform behaviour and, as such, do not account for the possibility that a mediator actually employs two (or even all) mediation strategies at the same time. Consequently, while we might gain important insights as to the correlation between a dominant strategy and an outcome, we are still short of any understanding about the actual causal mechanism by which a specific form of behaviour yields a particular outcome. Lastly, and perhaps most importantly, in the ten-year period – which is used as the time span in which the long-term effects of mediation are measured – the context may change so drastically so as to induce the (formerly) conflicting parties to develop new disagreements that have nothing to do with the prior mediation effort. As such, it is reasonable to assume that linking the mediation efforts with these new conflictual dynamics even between the same parties might be overly simplistic.

Nevertheless Beardsley also explains a potential mechanism by which a settlement may endure over time. Using insights from a number of illustrative cases, such as Kissinger’s and Carter’s involvement in the Middle East, Carter’s engagement in the Korean dispute, and Roosevelt’s mediation of the Russo-Japanese war, Beardsley concludes that ‘leverage itself is thus not bad for a self-enforcing peace as long as the impact of that leverage does not wane and leave the actors with time inconsistency problems’. ⁴⁷ In fact, the third parties may employ a number of tactics in order to address the ‘credible commitment problem’. ⁴⁸ In order to encourage a settlement to endure over time, mediators may prescribe detailed sets of guarantees as to the implementation of the agreement that help reduce the level of uncertainty between the disputing sides, such as setting up demilitarised zones, drafting disarmament provisions and/or employing cease-fire and election-monitoring missions. ⁴⁹ More importantly, as Beardsley points out, mediators may commit themselves to stay longer in the zone and thus contribute to the agreements’ longevity. Nevertheless, the durability of settlements should not only be naively associated with a mediator’s will and skill. Conflict intensity and issues at stake are often the primary reasons why some conflicts relapse into violence. ⁵⁰ Although mediators might be called in to help improve communication between the parties and assist them in reframing the issues, thereby ripening the conflict for resolution through the creation of various inducements, their impact is highly dependent upon the parties’ willingness to compromise.

Conclusion

International mediation is a goal-oriented process. While formal agreements are certainly a useful, albeit abstract, measure of mediation success, exclusive reliance on them remains an oversimplification of the complex nature of mediation. Just as mediators come in different forms, their goals inevitably differ in scope and nature. They may be mandated to reduce the level of hostilities and to bring the parties to the negotiating table, or to produce a robust settlement that deals with all the underlying issues of the conflict. Taking into account the varying nature of a mediator’s objectives (i.e. mandate) is the first step in deciding upon how to go about measuring or defining mediation success. At the same time, mediation should be understood as a process that is characterised by a sequence of events. As each stage is aimed at achieving a certain outcome, mediation success is best understood as a sequential phenomenon as well. This article identified three distinct, yet highly interrelated degrees of success: getting the parties to the table, reaching an agreement, and making the agreement endure over time. Each one of these stages is affected by a number of factors (such as parties’ and mediators’ characteristics, interests and resources) and almost all of them tend to be influential across all three stages.

The three books here under review are not only noteworthy for their general contributions to the growing field of international mediation, they are also highly valuable in understanding the particular the mechanisms by which the aforementioned degrees of success can be achieved. The identification of three general degrees of success and factors that affect them is just the starting point from which any analysis should begin. Instead of resorting to arbitrary, oversimplified, and case-to-case conceptualisations of mediation success, future studies need to accept that mediation success is as complex as the mediation process itself. Accordingly, within each one of the mediation stages, there could be potential sub-stages that also follow a cause-and-effect (or even a chronological) pattern. While keeping a keen eye on the mediators’ actual goals (or mandates), the juxtaposition of various sub-degrees of mediation success in the form of a causal chain would unquestionably reduce the elusiveness of a definition of success as an analytical concept.