Title Menu - The International Journal of Peace Studies.  
Return to Table of Contents Return to I J P S Home



Jacob Bercovitch and Su–Mi Lee



This paper examines some factors which may help to explain how mediators in international conflict structure their intervention and choose their behavior.  We argue that mediators may adopt a range of strategies which may be described as directive and non-directive strategies.  We develop some hypotheses that link the effectiveness of these strategies to three basic factors, nature of the issues, membership of a common bloc and perceptions of trust in a mediator.  We use a data set on international mediation events in the 1945-95 period to assess empirically how each of these affects mediation outcomes.  We find that in general directive strategies are more effective than non-directive ones, and that trust in a mediator is a particularly important dimension in explaining mediator strategy and success.



Mediation is one of the oldest and most common conflict resolution mechanisms in international conflicts. When it is applied correctly it can help to mange or settle a conflict.  Many questions about the process can be posed, but one of the most interesting  question relates to the need to understand when a mediation effort succeeds or fails to settle or resolve a conflict. To answer that question, we have to be aware of the relationship between strategies employed by a mediator, and mediation outcomes.  A number of attempts to explain the association between the choice of mediation strategies and mediation outcome have been made; few, however, have provided specific conditions where the effectiveness of certain mediation strategies becomes noticeable. This paper attempts to fill this gap.

In this paper, we will look at different strategies, and through an empirical examination of mediation attempts in international disputes between 1945 and 1995, we will argue that "directive" strategies are more effective than non-directive strategies. The association between directive strategies and successful outcomes is very strong in labor-management relations, as Hiltrop (1985) demonstrated. We suggest that it is as strong in international conflict.


Managing international conflicts has become a priority on the global agenda. The devastating consequences of conflict in an increasingly globalizing world order cannot be ignored. There are several peaceful ways to manage conflicts. These include avoidance, negotiation, mediation, arbitration, and adjudication.  Of these mediation offers many advantages.  It has been studied by scholars and students of political science, psychology, business management, and law as well as practitioners.  All have proposed various definitions of the process with very little consensus on any of these.

What are the characteristics of mediation?  First, mediation is a voluntary process (Bercovitch 1992; Moore 1996).  It takes place when disputants seek the assistance of third parties (Bercovitch 1992; Bercovitch and Houston 1993).  The right to accept or reject an offer of mediation or a mediation outcome rests entirely with the disputants (Ross 1993).  The fact that mediation is a voluntary process is directly related to its success or failure.  Without a high level of disputants’ willingness to concede, and motivation to engage in conflict management, a successful mediation outcome is unlikely to be achieved.
Secondly, the outcome of mediation is non-binding (Bercovitch 1992; Bercovitch and Houston 1993; Bercovitch 1997; Groom 1986; Moore 1996; Smith 1998; Touval and Zartman 1989; Wall and Lynn 1993).  Mediation's non-binding nature distinguishes it from other forms of external intervention such as arbitration and adjudication.  In mediation, third parties have no authority over disputants' compliance with a mediated outcome (Groom 1986).  Indeed, most disputants would not accept mediation in the first place if mediation bound them to an outcome.

These two characteristics mean that, for the most part, the outcomes of all mediation attempts depend entirely on the disputants' willingness to resolve their conflict and to abide by the mediator’s terms.  In other words, mediation cannot be successful if the disputants do not see any reason to resolve the conflict quickly, or if they refuse to adhere to the terms of an outcome. Disputants' motivation is a crucial factor, which affects many aspects of the process (see Rubin 1992). It is also amongst the important factors affecting the choice of a strategy.

Mediation is defined here as a pacific approach to conflict resolution in which impartial third parties help disputants resolve conflicts through a process of information and social influence, without using violence or invoking the authority of a legal system. The objective of disputants in inviting or accepting mediation is to reach compromise in a conflict, or at least to indicate willingness to do so (Richmond 1998). The third party in mediation may be an individual, organization, or country that is not a direct party to the conflict (Bercovitch and Houston 1993; Carnevale 1986; Touval and Zartman 1989; Wall and Lynn 1993).  For a mediation to be successful and for a compromise to be reached, an effective strategy must be employed by a mediator. But how can we determine which strategy is likely to be the most effective?

Mediation Strategies

Mediation strategy denotes an overall plan of mediators to resolve and manage conflicts (Bercovitch 1992). This definition is also supported by Kolb (1983:249), who notes that mediation strategy as “an overall plan, approach or method a mediator has for resolving a dispute. . .  [I]t is the way the mediator intends to manage the case, the parties, and the issue.”  consistent patterns emerge and are observable with respect to the overall strategy, or plan of action employed (kolb 1983). Differences in the implementation of various mediation strategies may be attributed to how a mediator chooses to handle the mediation process, and the specific context of the conflict. In essence, the practice and process of mediation revolve, to a large extent, around mediators’ choice of strategic behaviors.

Touval and Zartman (1985) identify three discrete categories of behavior, on an ascending level of third party involvement that describe the full range of mediation techniques. Their typology is particularly useful because it is part of a general framework of mediation.  Their three categories are the following: communication, formulation, and manipulation (Touval and Zartman 1985). This typology permits us to analyze and understand what mediators actually do when they get involved in a conflict, and how successful they may be by analyzing how different patterns of behavior lead to different outcomes. [This analysis can take the form of interviews, observations, or survey techniques.]

This taxonomy of mediator behavior was modified by Bercovitch and Houston (2000) where they present three categories of strategic behavior along a continuum ranging from low to high intervention.  These are communication-facilitation; procedural; and directive strategies (Bercovitch 1992; see also bercovitch and Houston 2000; Wall and Lynn 1993). These strategies are based on assumptions derived from Sheppard’s (1984) taxonomy of mediator behavior that focuses on the content (directive), process (communication-facilitation) and procedure (formulative) aspects of conflict management. Let us describe each of these strategies in turn.

Communication-Facilitation Strategies

These strategies describe mediator behavior at the low end of the intervention spectrum.  Here a mediator typically adopts a fairly passive role, channelling information to the parties and facilitating co-operation but exhibiting little control over the more formal process or substance of mediation. Tactics associated with this strategy include the following: making contact with the parties; gaining the trust and confidence of the parties; arranging for interactions between the parties; identifying issues and interests; clarifying the situation; avoiding taking sides; developing a rapport with the parties; supplying missing information; developing a framework for understanding; encouraging meaningful communication; offering positive evaluations; and allowing the interests of the parties to be discussed. When mediators adopt communication-facilitation strategies they play a role of “go-between,” such as passing messages from one disputant to the other, and providing disputants with unbiased information (Young 1967).  Communication-facilitation strategies are strongly supported by Burton, who claims that all disputes are products of misunderstanding, and that clear communication among disputants, through mediators, is the key to conflict resolution (Burton 1969).

Procedural-Formulative Strategies

These strategies enable a mediator to exert more formal control over the mediation process with respect to aspects of the environment of conflict management.  Here a mediator may control where mediation takes place, how often the parties meet, how the agenda is structured and information about progress is distributed.  Other aspects of this strategy include controlling constituency influences and media publicity, enhancing situational powers of weaker parties’, and chairing the communication process. Tactics associated with this strategy include choosing the site of meetings, controlling the pace and formality of meetings, controlling the physical environment, establishing protocols, suggesting procedures, highlighting common interests, reducing tensions, controlling timing, dealing with the simple issues first, structuring the agenda, keeping parties at the table, helping parties save face, and keeping the process focused on issues. procedural strategies are designed to create a favorable environment for conflict management.

Directive Strategies

These strategies are the most powerful form of intervention.  Here a mediator affects the content and substance of the bargaining process by providing incentives for the parties or issuing ultimatums.  Directive strategies deal with, and aim to change, the motivation and behavior of the parties in dispute.  The tactics associated with this strategy include changing the parties' expectations, taking responsibility for concessions, making substantive suggestions and proposals, making the parties aware of the costs of non-agreement, supplying and filtering information, suggesting concessions parties can make, helping the negotiators to undo a commitment, rewarding party concession, helping devise a framework for acceptable outcomes, changing perceptions, pressing the parties to show flexibility, promising resources or threatening withdrawal, and offering to verify compliance with agreement. Directive strategies represent the highest level of mediator involvement.

This conceptualisation specifies a clear distinction between various types of mediator behaviors and provides an extensive descriptive account of what exactly these behaviors entail.  It provides the basis for a logical and systematic explanation of mediation behavior that can be applied to the empirical analysis of mediation in international conflicts.  One is able to test whether a given profile fits a specific mediator role and how to enact it.  It also provides basis through which we can look at what influences and determines these behaviors.

The choice of strategy is rarely random, they all have their advantages and disadvantages. Directive strategies, for instance, allow mediators to control the process and the substance of a conflict, but this is achieved at the expense of the disputants’ freedom to control their own affairs.  When this strategy is used, disputants may be motivated to resolve a dispute as soon as possible before they cede further control to mediators (Muldoon 1996).  It is also possible that disputants may reject a mediator's proposal or even mediation itself when mediators put too much pressure on them (Touval and Zartman 1985).
The literature on mediator behavior suggests that in general directive strategies are more effective, especially in international militarized conflicts (Bercovitch and Houston 1993).  Bercovitch's (1986) analysis shows that directive strategies are the most effective in settling international conflicts.  In a later study, he found that communication strategies were most likely to be employed but less likely to lead to a successful outcome (Bercovitch 1991; Bercovitch, Anagnoson, and Wille 1991; Bercovitch and Houston 1993).   Can we go further and study the conditions where the use of directive strategies may lead to successful outcomes? This is what we hope to do below.


Developing Hypotheses


Disputant Willingness

The choice of a strategy is the product of many interacting factors.  Mediators choose their strategies based on their capabilities and their assessment of the context of the conflict. Many factors contribute to the success of mediation efforts. One of the crucial factors is the willingness of the parties to find a solution to their conflict. Since mediation is a voluntary process, disputants' motivations or willingness to reach to a settlement makes a great difference in mediation outcomes. The more the disputants are willing to settle a conflict, the more opportunities and resources are available for mediators and, therefore, the more likely it is that the mediation will be effective (Bercovitch 1997; Zartman and Touval 1996).  Rubin (1992) supports this observation by arguing that disputants’ motivations are a pre-requisite for effective international mediation. 

Where disputants have low willingness to settle their conflict, a strategy that is more likely to motivate disputants to concede is the key to a successful mediation  (Moore 1996; Smith 1998). Directive strategies can create this willingness and motivation by persuading and pressuring the disputants to agree to an outcome. Thus, we hypothesize that the lower the trust, willingness, and motivation of disputants, the higher the need for mediators to deploy directive strategies.


The idea that directive mediation strategies are the most effective under some conditions, raises an interesting question regarding the level of trust the disputants have in mediators. In mediation, it is obviously important for mediators to attain the disputing parties’ trust. Muldoon (1996) suggests that both parties must trust a mediator employing directive strategies.  Otherwise, the parties will merely resent what they perceive as an intrusion. The more a mediator gains the parties’ trust, the more resources will be available to him/her to use directive strategies.

Where non-directive strategies are employed, a mediator may gain the parties’ trust through the judicious use of information and strict adherence to the norms of impartiality.  However, trust may not be as crucial in this context as it is where directive strategies, with their emphasis on proposals, suggestions and the utilization of carrots and sticks, are employed.  Mediators who utilize such strategies retain authority over the mediation process, but little authority over an outcome.   In either case a proposal provided by a mediator who is not trusted by the disputants is hardly likely to be supported or accepted by the parties. 

Relationship between the disputants

Relationships between disputants influence mediator strategies and parties’ motivations to reach an agreement.  Wall and Lynn (1993) claim that mediation is likely to be more successful where disputing parties have close relationships. In international relations, close relationships may be represented by membership of international regimes.  International regimes are defined as “sets of implicit or explicit principles, norms, rules, and decision-making procedures around which actors’ expectations converge in a given area of international relations” (Krasner 1983:2). Through facilitating communication among countries, international regimes can affect countries’ interests and expectations, and consequently their behaviors (Kydd and Snidal 1993:123; Keohane 1984; Keohane 1993).  Some international regimes do not have institutional form, others are so evolved as to be embedded in international or regional organizations.
Using the notion of international regimes to signify close political or economic association and linking it to directive strategies leads us to offer the following hypothesis:

Hypothesis 1.  When disputants do not belong to the same international regime, directive mediation strategies are more effective than non-directive strategies.
Disputants’ level of trust in a mediator is thus expected to influence the choice of mediation behavior and presumably its effectiveness.   Directive strategies are more likely to be used when there is a high level of trust in a mediator.

We hypothesize this as follows: Hypothesis 1-1.  When disputants do not belong to the same international regime but have high level of trust in a mediator, directive mediation strategies are more effective than non-directive mediation strategies.


Issues in conflict also determine the effectiveness and consequences of mediation.  In attempting to explain mediators' roles in the success or failure of mediations, we group issues into tangible issues and intangible issues.  This categorization makes it possible to explore a relationship between the types of issues and the mediation strategies. Where conflict issues are tangible, disputants need only a minimal involvement of mediators.  Since these issues are somewhat divisible or lend themselves to some form of measurement, the injection of information by a mediator may be sufficient to mange such conflicts. However, in dealing with intangible issues where parties can not even reach an agreement on the nature of their issues, let alone know how to assess or measure these, mediators may need to adopt a more active role if they are to make any difference to the conflict.

Examining labor mediations in the United Kingdom (UK), Hiltrop (1985) suggests that the effectiveness of mediation strategies or techniques varies in different types of issues.  His research certainly displays that an association between the effectiveness of mediation strategies and types of issues exists. Non-directive strategies appear to be effective in salary disputes, whereas they appear to be ineffective in non-salary intangible issues - e.g. union recognition (Hiltrop 1985). Hiltrop (1985:91) justifies the results by saying that "when matters of principles [intangible issues] are involved, more pressure is required to induce movement from the parties' stated positions than when more 'compromisable' issues such as pay are involved."  In other words, when dealing with intangible issues over which disputants are reluctant to compromise, directive strategies are needed to draw the disputants to a negotiation table and create the basis of a settlement. The second set of hypotheses links directive mediation strategies to the nature of the issues in disputes. Thus, we hypothesize  the following:

Hypothesis 2. Directive mediation strategies are more effective than non-directive strategies in resolving intangible issues.

Hypothesis 2-1.  When the issues involved are intangible and disputants have high level of trust in a mediator, directive mediation strategies are more effective than non-directive mediation strategies.

Finally, all hypotheses are incorporated in Hypothesis 3.

Hypothesis 3.  When disputants have a high trust in a mediator and do not belong to the same international regime, and the issues disputed are intangible, directive strategies are more effective than non-directive strategies.

Data Introduction

This research is designed to examine the effectiveness of directive strategies under specific conditions, namely, when disputants are not in the same regime and where the issues are intangible.  In addition, we also want to look at how disputants' trust in a mediator impacts on the effectiveness of directive strategies.

To study these relationships we make use of a data set on international mediation events from 1945-95 generated by Bercovitch and his associates over the last fifteen years.  The data set encompasses numerous cases of mediation and negotiation, each of which is coded in terms of many independent variables.  Each mediation event is coded in terms of its behaviors, consequences, issues and other contextual conditions.  The objective is to give as accurate a picture as possible of conflict management events.  In addition to information about conflict management events the data set contains information on parties in conflict, their characteristics, and their socio-political dimensions.  There is also information on the nature of each conflict, the issues in conflict, duration, fatalities and other key components. The data set has been compiled over many years, and is constantly updated (for a fuller description, see Bercovitch and Houston 2000).

Cases and Units of Analysis

Bercovitch's current data defines an international conflict as "an organized and continuous militarized conflict or a demonstration of intention to use military force involving at least one state" (Bercovitch and Houston 2000: 184). A total of 295 conflicts in the 1946-1995 period met these criteria. Of these 295 conflicts, 171 experienced some form of mediation. In total we have 1,583 cases of discrete mediation attempts. These cases were coded by scanning systematically Keesing's Archives, the New York Times, the New York Times Index, the Times (London), and Reuters. Of the 1,583 mediation attempts, 189 mediation attempts were rejected by either or both disputants. Sixty-eight variables were analyzed for each of 1,583 mediation attempts.  Among them, six variables are selected to serve this research; these are, mediation attempt outcome, the mediators' previous relationship with the disputants, mediation strategies, the disputants' alignment, the type of conflicts, and the type of issues at a conflict.  Each variable will be explained in detail in the following sections.

The Dependent Variable: Mediation Outcomes

Mediation outcome, as the dependent variable, refers to the extent to which mediation has made an observable difference to the conflict or the parties’ behaviors.  It should be noted that this variable does not represent the overall outcome of an international conflict; it refers to the outcome of each mediation attempt.  Mediation outcomes are categorized as follows: no management, mediation offered only, unsuccessful, cease-fire, partial settlement, or full settlement. These we dichotomize into two categories, successful outcomes and unsuccessful outcomes.

Here we define "unsuccessful" as efforts by a mediator which produced no change at all in the level of conflict or the parties’ behaviors.  We group cease-fire, partial settlement, and full settlement and describe these as exemplars of a successful outcome.  If mediation was offered, but rejected, the mediation attempts are not included in the research reported here.  There is no attempt here to delineate outcomes in terms of how satisfied the parties may be with specific mediation efforts.  This is a line of research that can not be advanced using our particular methodology.   

Mediators’ Previous Relationship with Disputants

This variable measures disputants' trust in a mediator.  The level of disputants' trust for a mediator can be measured by looking at the previous relationships between a given mediator and the parties.  Bercovitch's data (see Bercovitch and Houston 1993) distinguishes previous relationships between mediators and disputants along five dimensions - no previous relationship, different bloc, same bloc as one party, same bloc as both parties, and mixed relationship.  Here we consider the first three and the last dimension to be indicative of low level of disputants’ trust, while the fourth dimension to be indicative of a high level of trust. The logic for this is quite simple. Mediators who belong to the same organization or alliance are bound to have mutual interests or common principles, and a degree of trust in each other.


Mediation Strategies

Three groups of mediation strategies are introduced in Bercovitch's data (see Bercovitch 1989).  He constructs his taxonomy of mediation strategies according to the level of mediator involvement in the mediation process. The most passive activities of mediators are communication/facilitation strategies such as provision of information.  Mediators utilizing communication/facilitation have little control over the substance of mediation.  Such mediators aim to promote communication between disputants by delivering messages from one to the other. The second group, procedural strategies give more power to a mediator, although they do not allow him/her to get involved in the substance of mediation.  Procedural strategies are designed to create a favorable environment where the disputants can resolve their conflict.  Such strategies include the arrangement of a meeting at a certain place and time, the disclosure of the mediation progress, and the overall supervision of communication. Lastly, the most active mediation strategies are directive or manipulation strategies.  Of all mediation strategies, none gives the mediator more power than this one.  In this strategy mediators are allowed to become engaged in the substance of mediation.  They can influence the disputants on making a decision about settlement, and they can pressure disputants to agree to a proposal they draft.  Directive strategies enable mediators to lead a mediation process by allowing them to set the agenda.

Disputants’ Alignment

In this study, disputants’ relationship is believed to affect the choice and effectiveness of directive mediation strategies.  Disputants in the same regime or alliance are more likely to settle a conflict because they may be concerned about the overall damage to their group or association. Bercovitch’s data (1989) groups disputants’ alignment into seven distinct types: members of opposing blocs, members of the same bloc, bloc member vs. unaligned, both unaligned, different regional organization, same regional organization, and regional organization vs. unaligned. This research assumes that two states are members of the same regime if they are members of the same bloc or regional organization.  This is not meant to suggest that states without such a tie necessarily have interests or resources in opposition, merely that they may have less willingness toward conflict settlement.

Types of Issues

Bercovitch (1989) categorizes issues in conflict in terms of; territory, ideology, security, independence, resources, and ethnicity.  Although it would be very interesting to investigate the relationship between individual types of issues and mediation outcomes, this research divides instead the above issues into tangible issues and intangible issues.  As explained in the previous section, tangible issues are divisible and measurable.  Here, territory and resources are classified as tangible issues, while interdependence, security, ideology and ethnicity are treated as intangible issues.

Data Analysis

Relations between disputants and Directive Mediation Strategies

Hypothesis 1 anticipated that directive strategies would be more effective than non-directive mediation strategies when the parties belong to different blocs or regimes. Parties in different regimes or blocs may not be enthusiastic about resolving a conflict as early as they can. They may need cajoling and inducement from a mediator to settle their conflict.
Table 1 shows that directive strategies are more effective than non-directive strategies when disputants belong to a different bloc. Clearly, more pressure has to be used on such parties than on members of the same bloc.

Different Regimes




Directive  Strategies

43.3% (84)

56.7% (110)

100% (194)

Non-Directive Strategies

34.6% (103)

65.4% (195)

100% (298)





Table 1:  Different Regimes and Directive Mediation Strategies          (Chi-Square P-value = 0.051)

Relations between disputants, Directive Mediation Strategies,
and Disputants’ High Level of Trust in Mediators

Scholars emphasize the importance of trust in a mediator employing directive strategies. The level of trust in mediators has a great effect on the success or failure of mediation especially when directive strategies are employed (Muldoon 1996, 154). Table 2 examines the relationship between trust and mediation strategies.

Different Regime/ High Trust




Directive  Strategies

75.0% (33)

25.0% (11)

100% (44)

Non-Directive Strategies

50.9% (30)

49.1% (29)

100% (59)





Table 2:  Different Regime, Highly Trusted, and Directive Strategies       (Chi-Square P-value = 0.013)

We see that when mediators can be said to have the disputants’ trust, three-fourths of all directive strategies produce some measure of success. Whereas just over a half of non-directive strategies appear to produce any success.  The importance of disputants’ trust in mediators cannot be overemphasized in any discussion on mediators’ choice of behavior. 

The Types of Issues

Here we examine varying types of issues as an independent variable to assess how they affect mediation outcome. This kind of research was conducted by Hiltrop (1985) in labor-management disputes. He found a clear association between intangible conflict issues and unsuccessful conflict management experience. Bercovitch and Houston (1993) in their examination of international conflicts, find that conflicts involving issues of principle are very difficult to deal with, manage or mediate.

Intangible/Tangible Issues and Directive Mediation Strategies

Hypothesis 2 investigates which mediation strategies are more effective when dealing with intangible issues.  The results of this analysis on the effectiveness of directive strategies are not significant statistically, and in any event, the difference in success rates between the two strategies is minimal.  Table 3 does not support Hypothesis 2 that directive mediation strategies are more effective than non-directive strategies in resolving intangible issues.  It seems that this pattern holds true in labor-management conflicts, but is not present in international militarized conflicts.      

Intangible Issues*




Directive  Strategies

40.4% (159)

59.6% (235)

100% (394)

Non-Directive Strategies

39.0% (289)

61.0% (453)

100% (742)





Tangible Issues**




Directive Strategies

44.4% (67)

55.6% (84)

100% (151)

Non-Directive Strategies

31.8% (94)

68.2% (202)

100% (296)





Table 3: Intangible/Tangible Issues and Directive Strategies    (Chi-Square P-value = 0.644* / 0.009**)

Table 3 does reveal that mediation using directive in tangible conflicts are more likely to be successful than those utilizing non-directive strategies. When conflicts are over tangible issues mediators can assess the issues, suggest compromises and drive for mutual concessions. Issue tangibility gives mediators additional resources, competence and power.

Intangible/Tangible Issues, Directive Mediation Strategies,
and Disputants’ High Level of Trust in Mediators

When we combine level of trust and tangibility of issues, we can see that directive strategies in the context of tangible issues produce some success 67.5% of the time.  Non-directive strategies, in the same kind of context, produce some success in only 40.5 % of all efforts. Bearing in mind that the success rate of directive strategies, regardless of the level of trust, was 44.4% in all tangible issues, the 67.5% success rate may well reveal the vital importance of trust in mediators. Disputants’ trust in a mediator is a crucial variable that may considerably increase the chance of mediation success.

Intangible Issues/ High Trust*




Directive  Strategies

59.7% (40)

40.3% (27)

100% (67)

Non-Directive Strategies

45.4% (39)

54.6% (47)

100% (86)





Tangible Issues/ High Trust**




Directive  Strategies

67.5% (27)

32.5% (13)

100% (40)

Non-Directive Strategies

40.5% (30)

59.5% (44)

100% (74)





Table 4: Intangible/Tangible Issues, Highly Trusted, and Directive Strategies
       (Chi-Square P-value = 0.078*/0.006**)

Disputants’ High Level of Trust in Mediator,
relations between disputants, and Intangible/Tangible Issues

The results in Table 5 do not lend much support to our hypothesis that directive strategies, in the context of intangible conflicts and parties who belong in the same regime, will be successful.  The results suggest that directive strategies work best in conflicts over tangible issues, where the parties are not in the same regime, but when they have trust in their mediator. This finding is statistically significant.

Different Regimes/
Intangible Issues/ High Trust*




Directive Strategies

68.7% (11)

31.3% (5)

100% (16)

Non-Directive Strategies

72.7% (8)

27.3% (3)

100% (11)





Different Regimes/
Tangible Issues/ High Trust**




Directive Strategies



100% (28)

Non-Directive Strategies



100% (48)





Table 5: Different Regime, Intangible/Tangible Issues, and Highly Trusted
       (Chi-Square P-value = 0.824*/0.005**)

The success rate of directive strategies as shown in Table 5 is quite remarkable.  In 78.6% of cases, directive strategies turned out to be successful in achieving some success where issues in conflict were tangible, disputants were from different blocs or regimes but had a high trust in a mediator. 


This paper examined the conditions that may affect directive mediation strategies.   The choice of a strategy by a mediator is often a complex process where many dimensions have an impact. The choice of a strategy, in any area of human behavior, is the product of perceptions, expectations and some contextual conditions. We looked in particular at three factors that might impact on the choice and effectiveness of a strategy: nature of the issues in conflict, memberships of same regime or bloc, and the degree of trust in a mediator.
We have tried here to go beyond descriptions of specific mediators in specific conflicts and examine the overall structure of mediation behavior.  We organized our study in terms of a conceptual framework which identified types of mediation strategies, stipulated variables, and examined their relationship using a data set of international mediation events. This kind of approach demonstrated that factors such as trust and the nature of the conflict exercise a strong influence on mediation outcomes. These factors, more so than mediator identity, are more directly associated with mediation success or failure than other factors.

In general we find that directive strategies are much more effective than non-directive strategies, this is especially so in the context of tangible issues, parties from different blocs, and trust in a mediator, are far more effective than non-directive strategies. Much more work needs to be done, both theoretically and empirically, to understand what influences mediators’ choice of strategies, and how to delineate the crucial factors that have an impact on the process and outcome of mediation. Hopefully, we can build on this work and develop a better understanding of how types of mediation strategies are matched with different kinds of conflicts. Research on this important topic has been mainly notable for its brevity.



Bercovitch, Jacob.  1997.  “Mediation in International Conflict.” In I. William Zartman and J. Lewis Rasmussen, eds., pp. 125-53, Peacemaking in International Conflict.  Washington, D.C.: United States Institution of Peace Press.

Bercovitch, Jacob.  1992.  “The Structures and Diversity of Mediation in International Relations.” In J. Bercovitch and J. Z. Rubin, eds., pp. 6-29, Mediation in International Relations: Multiple Approaches to Conflict Managements.  New York: St. Martin’s Press.

Bercovitch, Jacob.  1991.  “International Mediation and Dispute Settlement: Evaluating the Conditions for Successful Mediation.”  Negotiation Journal, Vol. 7, No. 1, pp.17-30.

Bercovitch, Jacob. 1989.  International Dispute Mediation: A Comparative Empirical Analysis.” In Kenneth Kressel and Dean Pruitt, eds., pp. 53-67, Mediation Research. San Francisco: Jossey Bass.

Bercovitch, Jacob.  1986.  “International Mediation: A Study of the Incidence, Strategies, and Conditions of Successful Outcomes.”  Cooperation and Conflict, Vol. 21, pp.155-68.

Bercovitch, Jacob, Theodore J. Anagnoson, and Donnette Wille.  1991.  “Some Conceptual Issues and Empirical Trends in the Study of Successful Mediation in International Relations.”  Journal of Peace Research, Vol. 28, No. 1, pp. 7-17.

Bercovitch, Jacob and Allison Houston.  2000.  “Why Do They Do It Like This?”  Journal of Conflict Resolution, Vol. 44, No. 2, pp. 170-202.

Bercovitch, Jacob, and Allison Houston.  1993.  “Influence of Mediator Characteristics and Behavior on the Success of Mediation in International Relations.”  International Journal of Conflict Management, Vol. 4, No. 4, pp.297-321.

Burton, John W.  1969.  Conflict and Communication.  London: MacMillan.

Carnevale, Peter J.  1986.  “Strategic Choice in Mediation.” Negotiation Journal, Vol. 2, pp. 41-56.

Groom, John A. R.  1986.  “Problem Solving in International Relations,” In Edward E. Azar and John W. Burton, eds., pp. 85-91, International Conflict Resolution.  Boulder, CO: Lynne Rienner Publishers.

Hiltrop, Jean Marie.  1985.  “Mediator Behavior and the Settlement of Collective Bargaining Disputes in Britain.”  Journal of Social Issues, Vol. 41, No. 2, pp.83-99.

Keohane, Robert.  1993.  “The Analysis of International Regimes: Toward a European-American Research Programme,” In Volker Rittberger, ed., pp. 23-45, Regime Theory and International Relations, New York: Oxford University Press.

Keohane, Robert.  1984.  After Hegemony.  Princeton, NJ: Princeton University Press.

Kolb, Deborah. 1983. “Strategy and Tactics of Mediation.” Human Relations, Vol. 36, No. 3, pp. 247–68.

Kydd, Andrew and Duncan Snidal.  1993.  “Progress in Game-Theoretical Analysis of International Regimes,” In Volker Rittberger, ed., pp. 112-35, Regime Theory and International Relations.  New York: Oxford University Press.

Krasner, Stephen D.  1983.  “Structural Causes and Regime Consequences: Regimes as Intervening Variables.” In Stephen Krasner, ed., pp. 1-22, International Regimes.  Ithaca: Cornell University Press.

Moore, Christopher W. 1996.  The Mediation Process: Practical Strategies for Resolving Conflict.  San Francisco: Jossey-Bass Publishers.

Muldoon, Brian.  1996.  The Heart of Conflict.  G.P. Putnam’s Sons, NY.

Richmond, Oliver.  1998.  “Devious Objectives and the Disputants’ View of International Mediation: A Theoretical Framework.”  Journal of Peace Research, Vol. 35, No. 6, pp. 707-722.

Ross, Marc Howard.  1993.  The Management of Conflict.  New Haven: Yale University Press.

Rubin, Jeffrey Z.  1992.  “International Mediation in Context,” In Jacob Bercovitch and Jeffrey Z. Rubin, eds., pp. 249-72, Mediation in International Relations.  New York: St. Martin’s Press.

Sheppard, William.  1984. “Third Party Conflict Intervention: A Procedural Framework.” Research in Organization Behavior, Vol. 6, pp. 141-90.

Smith, Richard P.  1998.  “Quantitative Methods in Peace Research.”  Journal of Peace Research, Vol. 35, No. 4, pp.419-27.

Touval, Saadia, and I. William Zartman.  1989.  “Mediation in International Conflicts,” In Kenneth Kressel and Dean G. Pruitt, eds., pp. 115-37, Mediation Research.  San Francisco, CA: Jossey-Bass.

Touval, Saadia and I. William Zartman.  1985.  “Mediation in Theory,” In Saadia Touval and I. William Zartman, eds., pp. 7-17, International Mediation in Theory and Practice.  Boulder, CO: Westview Press.

Wall, James A. and Ann Lynn.  1993.  “Mediation: A Current Review.”  Journal of Conflict Resolution, Vol. 37, No.1, pp. 160-94.

Young, Oran.  1967.  The Intermediaries: Third Parties in International Crises.  Princeton, NJ: Princeton University Press.

Zartman, I. William and Saadia Touval.  1996.  “International Mediation in the Post-Cold War Era.” In Ehster Crocker, Fen Hampson and Pamela Aall, eds., pp. 445-61, Managing Global Chaos.  Washington, D.C.: United States Institute of Peace Press, 1996.