This article examines the impact of nongovernmental organization‐sponsored contact and communication on fostering peaceful solutions to ethnic conflict via case studies of the activities of the Project on Ethnic Relations (PER) in Romania, Macedonia, Montenegro, Kosovo, and Serbia. It explores five operational principles that guide PER activity: creating credible, neutral forums for dialogue; maintaining momentum; working within political realities; encouraging indigenous solutions from within existing processes; and acting with the backing of powerful states. These principles explain PER's success as a “weak mediator” of ethnic conflicts. According to this analysis, PER also exhibits organizational characteristics that contribute to success, including nonpartisanship, area expertise and extensive networks of local contacts, and an ability to secure the trust of local actors.

A significant indicator of the success of PER activities is the establishment by conflicting parties of institutionalized mechanisms for addressing their differences. Contrary to the view that electoral competition contributes to conflict, this study finds that the possibility of achieving an electoral advantage by participating cooperatively in conflict resolution activities creates incentives for local actors to recognize opportunities offered by PER activities and leads local actors to heed PER's advice. Finally, the article offers a cautionary observation. While PER's perceived influence with major international actors may contribute to its local successes, once a state actor with the power to impose a solution has committed itself to ending a conflict, its preferences outweigh any local interests in determining the outcome and renders the efforts of a “weak mediator” such as PER irrelevant.

This article examines the impact of nongovernmental organization (NGO)‐sponsored contact and communication on fostering peaceful solutions to ethnic conflict in the Balkans. The object of this study is the Project on Ethnic Relations (PER), a Princeton‐based organization. PER has been particularly active and has achieved its clearest successes in Romania, where it has helped foster peaceful approaches to resolving the conflict between the Romanian majority and the ethnic Hungarian minority in that country. It has also achieved significant successes in Macedonia and Montenegro. PER experience in these countries offers important lessons about the factors that contribute to success in such activities. PER has also been engaged since 1992 in efforts to alleviate conflicts involving Albanians and Serbs in Kosovo and Serbia, but these efforts have been far less successful.

A comparison of PER experience in Romania, Montenegro, and Macedonia with PER experience in Kosovo and Serbia suggests that the outcomes of such mediation efforts depend on the effect local actors perceive compromise and cooperation will have on their own concrete political interests. While PER experiences also suggest that NGOs may be able to affect local perceptions in ways conducive to agreement, they offer some important cautionary notes about the immense difficulties involved in nongovernmental third‐party conflict resolution efforts in the face of violent conflict, about the conditions that may be necessary for such efforts to succeed, and finally, about how powerful outside forces can limit NGO influence on local actors.

In this article, I will first summarize the basic PER approach to conflict resolution, which involves maintaining both neutrality and fairness as the organization engages both (all) sides of a conflict in dialogue aimed at encouraging the emergence of indigenous solutions. In the next section, I identify five general principles of PER engagement; each helps explain the degree of success achieved in each case. I then provide short case studies of PER activities in Romania, Montenegro, Macedonia, and Kosovo to illustrate the operation and consequences of these principles in practice. These case studies are based on both PER documentation of the meetings (cited in the notes) and my own analyses of the Kosovo, Montenegro, Macedonia, and Serbia meetings, in which I participated and for which, in many cases, I served as rapporteuer. Finally, in the conclusion, I suggest that the effectiveness of the PER approach depends not only on the perceptions of local actors, but on the preferences of “great powers,” powerful states external to the conflict but interested in its outcome. In the cases reviewed here, it is the U.S. and the European Union that play this role.

The PER engagement with ethnic conflict in formerly communist regions reflects the intellectual biography of the organization's founder, Allen Kassof. Kassof was already a distinguished scholar in the field of Soviet and East European studies when he became director of the International Research and Exchanges Board (IREX), which was the principal institutional sponsor of academic exchanges between the East and West during the Cold War period. From that position, he enjoyed unparalleled contact and communication with Soviet/Russian and East European scholars, and developed insight into social, cultural, and political developments in their home countries.

Kassof anticipated the explosion of ethnic conflicts that would come with the collapse of communism, and founded PER in an attempt to encourage their peaceful resolution. Presently, the PER mission is self‐consciously defined as reducing the probability of violent interethnic conflicts. Thus, PER also addresses such broader regional issues as political extremism and its causes, in an effort to help empower leaders to avert problems (PER 2002a). It should be noted that PER staff consider “prevention” too presumptuous a word to describe their activities. From an analytical perspective, however, PER activities have indeed included conflict prevention work, confidence‐building efforts, and postconflict peacemaking of the type usually encompassed by the term conflict resolution.

The modesty of PER staff reflects their organizational approach, which is to push those engaged in a conflict to develop their own solutions, rather than to impose a PER‐devised solution. PER can thus be characterized, according to the definition articulated by Dean G. Pruitt, as a “weak mediator.” According to Pruitt, “The weak mediator's main tools are in the realm of communication and formulation rather than manipulation. . . . They include the capacity to transmit and interpret messages, to bring realism to the parties' conceptions of each other, to reframe the issues, and to make suggestions for settlement” (Pruitt 2002: 51).

Consistent with its role as a weak mediator, PER does not produce policy recommendations or advocate particular solutions to the conflicts in which it is involved. The organization's leaders believe that maintaining neutrality is essential to preserving their ability to work with actors on both (all) sides of a conflict, and the term neutrality reflects PER's organizational commitment to remain impartial, or nonpartisan, in its approach to conflict. Kassof said, “we don't consider ourselves players” in the events (Kassof 2001). PER emphasizes fairness in the search for solutions. PER views nonpartisanship as a requisite condition for achieving fairness, a view that contrasts with the view of such scholars as Jacob Bercovitch, who argues that impartiality and fairness can be competing principles (Bercovitch 1996: 26).

The activities sponsored by PER are neither strictly “second track” (i.e., involving only nonofficial actors), nor strictly intergovernmental, although PER works with government officials, and support from governments is critical to the success of the organization. Kassof characterizes PER activity as “track one and one‐half”; that is, in his words, “diplomatic activity carried out by private persons dealing with decision makers” (Kassof 2001). Such an approach requires a dense network of contacts and extensive communication among the parties to the conflict. The ability to establish and maintain such a network is, in part, a legacy of the organization's origins in IREX, through which Kassof established professional contacts over many years with numerous scholars throughout the region, some of whom have since assumed significant public roles. But such a network would be impossible to maintain were it not for the extensive linguistic, cultural, and sociopolitical expertise of PER's very small operational staff.

The personal and professional network of Romanian contacts of PER's then executive director (now president) Livia Plaks' were, for example, crucial to the success of the organization's engagement there. PER also draws on the assistance of experts and consultants, both locally and internationally. (I served as an outside expert for PER, and have participated in many of the events reported here. Direct participant observation of PER activities is a major source for the analysis that follows.) PER's networks include contacts on all sides of the particular issue. The PER staff strives not only to remain well‐informed, but also to develop what might be called insider expertise, by tracking local political events over time, talking with local actors on an almost daily basis, being present in the area of concern through both permanent representation and frequent visits, and following local media accounts of events. Through these activities, PER seeks to gain practitioner and operational expertise as opposed to scholarly expertise.

Because PER's staff is small and its engagement intensive, the number of issues in which the organization may become engaged is by necessity small. But the organization's small size and intense involvement engender three organizational characteristics that have contributed to effectiveness in the cases reviewed here: persistence, patience, and presence. PER thus exhibits many of the characteristics identified by Bercovitch as requisites of successful mediation. It is “perceived as reasonable, acceptable, knowledgeable, and able to secure the trust and cooperation of the disputants” (Bercovitch 1996: 25).

PER efforts at mediation fall into two of the three categories of mediator activity identified by Saadia Touval and William Zartman, which are communication–facilitation, formulation, and manipulation (Bercovitch 1992: 16–18). PER engages predominantly in communication–facilitation strategies. The organization is by no means a “passive” mediator but does not engage in “manipulation” of participants in its activities. As a weak mediator, PER can neither promise nor withhold resources as a means of pressing parties to change their views. Rather, it assists disputants in understanding their conflict, in defining their own interests, and in identifying opportunities for transcending conflict and achieving cooperation. Each party to a PER‐mediated agreement gains resources and rewards from the cooperation itself or from international actors supporting PER activities. Kassof summarized the PER approach to conflict resolution in remarks to the Carnegie Corporation (Carnegie Commission on Preventing Deadly Conflict 1997: 53). These remarks can be reduced to five operational principles:

  • create credible, neutral forums for dialogue early,

  • maintain momentum,

  • work within political realities,

  • encourage indigenous solutions from within existing processes, and

  • act with the backing of powerful states.

Each of these operational principles is described below.

Creating Credible, Neutral Forums for Dialogue

PER undertakes a great deal of preparatory work on each side of the conflict. This involves identifying the relevant and requisite parties, persuading them to participate, devising an appropriate agenda for dialogue, and creating the appropriate conditions — substantive, procedural, and structural — that must be met for such dialogues to happen. Hence, PER events are conducted according to what are commonly defined as “Chatham House rules.” (See http://www.chathamhouse.org.uk.) Participants are free to discuss the meetings, including with representatives of the press, but only on a not‐for‐attribution basis. The meetings focus on issues that potential participants have indicated in advance discussions are important to them, and dialogues and meetings are conducted in protected environments and are designed to be conversational in style (to the extent possible, depending on the size of the meeting), not formal debates. The ability of PER to secure participation in such events is linked to the extensive networks of contacts and communication cited above. But it is also connected to the organization's conscious efforts to avoid becoming associated with only one issue or, especially, the interests of one group.

PER does not function as an advocate of minority rights. It recognizes that both majorities and minorities have rights and responsibilities, and that majority–minority dialogue is essential to finding peaceful solutions to conflicts that involve competing definitions of rights. Hence, PER sees its efforts to engage local governments in dialogue as an opportunity for meaningful exchanges with opponents at no political cost, but with the potential for advancing mutually acceptable solutions to their common problems. This perspective is reinforced by PER's efforts to address issues regionally when appropriate. When issues are regional in significance, it is important to engage all relevant actors in efforts to address them. Governments tend to be extremely sensitive to the question of status and will resist being “singled out” on an issue that transcends borders.

Maintaining Momentum

PER seeks to work directly with the parties to build on whatever has been achieved through dialogue, but this has not always been possible. In both Montenegro and Macedonia, initial PER efforts were overtaken by local events. In Macedonia, PER pursued an initiative on issues involving the Roma (the ethnic group often referred to in the past as “Gypsies”). But it was not simultaneously able to engage the political leadership in an effort to address issues surrounding the status of ethnic Albanians. This became possible only after the outbreak of violence in 2001 created stronger incentives on both sides to cooperate and for the U.S. and European actors to encourage dialogue as a means of addressing the grievances underlying the violence. Even when PER has been able to initiate dialogue and follow‐up has been successful, the time lag between meetings can be considerable, which can make maintaining momentum difficult, as in the case of PER's meetings on Albanians, Kosovo, and Serbia, which were interrupted first by war in 1999, and then by the outbreak of renewed violence in 2004.

Working within Political Realities

Choosing where and when to become involved amounts to an exercise in “creative opportunism.” PER's activities reflect its leaders' underlying sense of where the most important fault lines in the region are located. While PER cannot make major progress in places where the government is uninterested or even opposed, it nonetheless can make some progress. Even under the dictatorship of Slobodan Milosevic in Serbia, for example, PER worked in Vojvodina to encourage cooperation between the local ethnic minority parties and the democratic opposition parties from Belgrade who opposed Milosevic. According to the head of the Organization for Security and Cooperation in Europe (OSCE) mission to Yugoslavia at the time, these efforts facilitated the inclusion of the minority parties in the democratic coalition in Serbia, as well as the appointment of an ethnic Hungarian as deputy prime minister in the democratic government of Serbia, the first time an ethnic Hungarian had been appointed to such a high‐level position (Sannino 2001).

PER does not wait until a conflict is “ripe for settlement,” but it is hard‐nosed and realistic about the nature of conflict. When I asked him in an interview to explain what this meant, Allen Kassof said:

It is not the inability of individuals to recognize each other's humanity, to cooperate with one another, or even to develop deep friendships or love for one another that is at the heart of ethnic conflicts — ethnic conflicts are constructed conflicts; constructed by political and cultural elites acting in their own interests. This means that in some instances one can only hope to improve the situation, and one must not let one's hope for the best outcome deter one from seeking a better outcome. Change in leaderships may therefore represent real opportunities (Kassof 2001).

The nature of the postcommunist regimes in the countries in which PER has been active has varied from nationalist authoritarian to genuinely democratic. To the extent that the regimes involved in conflicts are democratic, the task of encouraging dialogue has been easier, not only because political leaders in democratic regimes tend to be more open, but also because international actors — the U.S. especially — are more interested in working with democratizing regimes than with nondemocratic regimes. In the course of PER's involvement in Kosovo, for example, leaders of democratic opposition parties in Serbia were more open to dialogue with Kosovar Albanian leaders than were officials of the Serbian government under Milosevic and his authoritarian ruling party. But because PER emphasized the interests of the parties themselves and invested time and energy in helping leaders on all sides to see their common interest in peacefully resolving the Kosovo conflict, it was possible for PER to arrange direct meetings between Kosovar Albanian leaders and representatives of both the Serbian government and its opposition.

Sustaining engagement in an issue over a long period of time can enhance opportunities for meaningful dialogue because “political realities” can become more favorable over time. After the shift in Serbia toward democracy, for example, some of the participants in earlier PER meetings became senior government officials and leading political figures in postwar Kosovo. In Romania, meetings involved mostly government and opposition‐party representatives. Former government opponents who had participated in the earlier meetings eventually joined the Romanian government where they adopted recommendations that had been developed in PER‐organized meetings as government policy.

In Serbia, however, the entry into government of past participants in PER processes has not necessarily resulted in their greater engagement with the project. PER activities had provided democratic political leaders in Serbia with an important channel for communication, including communication with international actors. But since they have taken power, the stakes involved in the resolution of ethnic and minority issues, especially with respect to Kosovo, appear to be too high even for democratic Serbian politicians to risk having the kinds of open discussion that characterize PER‐organized meetings. Not surprisingly, therefore, a new generation of opposition politicians from ethnic minority parties seem more interested in participating in PER activities in Serbia than do coalition party leaders or government officials (except for those officials responsible for minority affairs).

Encouraging Indigenous Solutions

PER does not develop its own policy recommendations for solving a conflict. Instead, the organization seeks to encourage authentic, indigenous solutions from within existing processes or institutions, and, if necessary, to help the actors create such processes and institutions where they are absent. Meetings organized by PER thus provide extensive opportunities for informal contact among participants, so that the sides can develop trust, respect, and understanding of each other's underlying interests. Because this is seldom achieved in a single meeting, PER remains engaged with an issue for many years. PER thus shares many of the same organizational and operational characteristics attributed to the Community Sant'Egidio by Cameron Hume (Hume 1994) and to the Conflict Management Group‐Norwegian Refugee Council mediation effort in the Caucasus by Susan Allen Nan (Nan 2004).

Like all third‐party mediators, PER encourages parties to a conflict to redefine their self‐interests in ways conducive to cooperation. PER advised leaders of the Hungarian minority in Romania, for example, that they should avoid general demands for “autonomy” and urged them to focus on more specific demands. “Autonomy” represented a probable dead end, while specific demands offered opportunities for compromise and, eventually, cooperation. PER took the same approach in its work with the Albanian leadership in Montenegro. Kassof reports that PER has “cajoled, scolded, instructed, educated, and motivated local leaders as part of getting them to change their own sense/definition of self‐interest” (Kassof 2001). Clearly, PER could not do this without first earning the participants' trust.

For PER, the most important indicator of success is the parties' creation of mutually agreed methods, processes, or even formal institutions for dealing with the sources of conflict. This has been achieved in Romania, which is why the Romanian case can be considered a clear‐cut success. The project in Macedonia, which has followed somewhat the model of the Romanian project, was moving in a positive direction. The recent change of government will undoubtedly require PER to re‐establish trust and cooperation between parties whose roles are now reversed. In Montenegro, PER facilitated the adoption of new legislation on minority rights but key elements of the legislation have been called into doubt by a recent ruling of the constitutional court. In the case of Kosovo and Serbia, a PER‐brokered agreement between the two sides to convene a series of government‐to‐government discussions on finding a mutually agreed solution was immediately negated by strong domestic opposition.

Backing of Powerful States

To achieve success, PER must work with the backing of powerful states and multilateral organizations, which indicates to local actors that their actions will have international consequences. Securing the participation of representatives of powerful states and significant international organizations is one way to signal to local actors that PER has this kind of support. Officials from the U.S. and European Union participated in the many PER sessions that I attended. Their involvement encouraged key local leaders to participate and to articulate their local grievances and demands and also gave them opportunities to hear international responses to their concerns firsthand.

The participation of these representatives, however, can be a double‐edged sword. International actors intent on controlling local developments can suppress participants' initiatives. More fundamentally, international actors are often highly partisan in their relationships with conflicting parties. Neither PER nor any other NGO that pursues an impartial, inclusive strategy can operate effectively in the face of partisan opposition from such powers. A project to address the issue of Serbs and Serbia paralleling the project on Albanians, for example, was hampered by the opposition of the U.S. and other powerful outside actors to the inclusion of some Serbian parties. Of course, there is little room for an NGO to operate when great powers are determined to resolve conflict directly. PER efforts to address the Kosovo issue were rendered moot by the U.S. effort to resolve the Kosovo status question initiated in early 2005.

PER engagement in Romania began with a conference organized with the Romanian Academy of Sciences at the suggestion of leaders of the Hungarian minority in Romania. The conference brought together minorities in Romania and some foreign experts in June 1991 (PER 1991). There had been violence between Hungarians and Romanians in the city of Tirgu Mures in 1990. The official response had been one‐sided in favor of ethnic Romanians, contributing to a sense of moral outrage among minorities and fears of further violence.

Several circumstances made it possible for PER to act effectively in Romania. PER staff members had extensive contacts and maintained strong friendships on both sides of the ethnic divide in Romania. With a local staff person in Romania, PER was also able to establish a partnership with the Romanian academy. But leaders of Hungarian and Romanian groups, as well as government leaders, were nonetheless reluctant to meet. PER reported later that “it took almost a year to persuade Romanian officials and leaders of the ethnic Hungarian community just to gather around the same table, so deep was the mistrust” (PER 2001a). Indeed, PER resorted to inviting the actors to “lunch” as a means of initiating a first meeting. In a June 2001 interview, Livia Plaks, PER's executive director, described relations at the first meeting as “very frosty, hostile” (Plaks 2001). The PER report for that meeting acknowledges that the participants held

fundamentally different interpretations of the nature of the problem and of the various measures that should be taken to address it. Agreement on the nature of the primary problem could not be reached. . . . In fact, it turned out that there were several, often very different, views of what constituted primary problems (PER 1991: 3).

“It was easy to see at that meeting,” Plaks reported, “that these groups had not been talking to one another, and the meeting stimulated interest among Hungarians and Romanians in getting together to talk about their common problems” (Plaks 2001). A follow‐up meeting on the Black Sea coast in June 1992 provided additional opportunity for frank discussions between the two sides. Plaks (2001) described them as “substantive and argumentative, including the emotional articulation of grievances.” But while there was real engagement in dialogue, there were no constructive outcomes. PER calls this “putting the pieces of the puzzle on the table.” This corresponds to what Harold Saunders describes as the first two “stages” of a dialogue process, which are creating an opportunity for dialogue and allowing the “venting” of grievances (Saunders 2003: 86). This represents an important element in improving mutual understanding between conflicting parties.

The June 1992 meeting was followed by meetings in Romania, Switzerland, and the U.S. Moderates on both the Hungarian and the Romanian side were criticized by extremists in their own parties. Plaks writes that “the moderates on the Hungarian side were almost destroyed politically in the process and it is only due to their great courage in pushing the agenda through that they succeeded” (Plaks 2003). Patrice McMahon's study of the role of transnational actors in managing ethnic conflict in Romania confirms that PER was viewed by local actors, along with the North Atlantic Treaty Organization (NATO), the OSCE High Commissioner for National Minorities, and the European Union as “making important contributions to the management of ethnic relations.” McMahon reports:

According to Gabor Adam, a program officer at the Open Society Institute in Cluj, PER may have been the first international group to get government officials and minority group leaders to meet together. In his opinion, PER helped to create a culture of discussion in Romania (McMahon 2005: 22).

PER helped increase the international visibility of Romanian elites willing to engage in dialogue by arranging meetings for them with policymakers in Washington. Facilitating such contacts served dual purposes. A U.S. embassy officer in Romania reported to McMahon “that PER provided the United States with crucial information about what was happening in Romania and through its connections to the U.S. government was able to influence Romanian and Hungarian intellectuals” (McMahon 2005: 24).

The discussions organized by PER led to the development of a number of concepts, mechanisms, and agreements for managing interethnic relations in Romania. These included:

  • the establishment of a Council for National Minorities composed of minority members of parliament, which acts as an advisory body to the government;

  • adoption of bilingual signage;

  • education reforms at both the secondary and postsecondary levels, including revision of textbooks and introduction of textbooks in minority languages; and

  • a major reform of the civilian police in Romania.

PER helped achieve the last item through collaboration with the Southern Police Institute of the department of justice administration of the University of Louisville in Kentucky in a program of reform that extended from 1994 to 1998 (PER 1999; Wilson and Walsh 1995). A department of prevention was created within the police inspectorate of the Ministry of Interior devoted to monitoring and improving police and minority relations.

The fostering of institutional development within governments has been a major achievement for PER, and PER‐inspired approaches in Romania have been influential beyond that country's borders. The police initiative in Romania was repeated successfully in Hungary (PER 2000a; Wilson and Walsh 1995). The training program was also adopted in Slovakia, which has its own ethnic Hungarian minority (PER 2001b). This represents a dramatic example of the positive regional spillover potential of “best practices” for managing interethnic conflict adopted in any single state.

PER extended its efforts in Romania by encouraging moderate parties in both the government and opposition to continue to deal with interethnic relations. McMahon suggests that, prior to the 1996 election, “PER was able to get Romanian politicians who otherwise would not have talked to one another to sit down together” (McMahon 2005: 24). The November 1996 elections resulted in formation of a coalition government that included an ethnic Hungarian party for the first time. This government established a ministerial Department for the Protection of National Minorities, which included a National Office of Romani Affairs.

But an anti‐Hungarian backlash in the media and some sectors of Romanian society made it more difficult to adopt reforms in education, local administration, and use of the Hungarian language. To help resolve these issues, PER convened a seminar in Romania in November 1997, attended by representatives of all the coalition parties, Romanian academics, the head of the Department for Protection of National Minorities, and civic and human rights organizations in Romania. This meeting gave Romanian elites the opportunity to identify viable compromise formulations on these sensitive issues (PER 1998). PER continued to work in Romania at the local level with leaders of the Hungarian and Roma minorities, separately, to help them define their demands more effectively; with media and opinion makers on the representation of minorities in the press; and with educators regarding the revision of textbooks.

In anticipation of the November 2000 parliamentary election in Romania, PER convened a meeting of representatives of six major parties, including the governing parties and the two major non‐nationalist opposition parties, in February 2000 to discuss how the anti‐Hungarian extremism that had characterized electoral campaigns in 1992 and 1996 might be avoided in the upcoming campaign (PER 2000b). The meeting, according to a PER report, was characterized by “sober and reasoned discussion” of the ethnic issue. At this meeting, “PER proposed that an agreement might be drawn up by the parties . . . concerning their conduct during the upcoming elections. The objective would be to avoid nationalist rhetoric in the campaigns.” A “senior politician” drafted such a statement, and “several days after the conclusion of the meeting, the participants unanimously agreed to sign and publish the statement” (PER 2000b: 15–17).

The 2000 election ended the coalition government, sending the Hungarians into opposition. But the degree to which PER‐sponsored communications had bridged the ethnic divide in Romania was made clear after the election when the new governing party committed itself to preserving past gains in interethnic relations. To facilitate further progress, PER organized a meeting of governing and opposition party leaders and representatives of the Romanian president's office in February 2001 to discuss ethnic issues. At this meeting, “the debates between Romanians and Hungarians (and the Roma) . . . were civilized, self‐analytical, and modest, even when there were disagreements. Above all, each side was well‐informed and aware of the needs and interests of the other, and was at least willing to take them into account and to discuss them if not always to act on them” (PER 2001a: 2). In 2001, a law was passed that gave greater power to local governmental units and called for the use of minority languages and bilingual signage in areas where members of a minority group constitute more than 20 percent of the local population — its passage was a major achievement for Romania's Hungarians.

This increasingly cooperative behavior toward ethnic Hungarians has been undertaken by politicians who had previously opposed cooperation. Livia Plaks reported that these opponents asked ethnic Hungarian leaders “if we give you this, will you stop asking for more?” (Plaks 2001). Thus, what might be seen from outside the process as a “change of heart” or the emergence of “good will” can also be seen as a revised calculation of political interest on the part of participants in these discussions.

Although participants in the Romanian dialogues credit PER with facilitating this change, it cannot be attributed to PER efforts alone. The Romanian parties formed a coalition with Hungarians primarily to further their own interests and gain more power. Romania needed to come to some form of accommodation with the country's ethnic Hungarian minority to be granted membership in the European Union and NATO. As a result, while the successor government did not include the Hungarians, it did develop a formal “understanding” with the major Hungarian party that its interests would be accommodated. The PER effort to facilitate communication across the ethnic divide in Romania encouraged and enabled both Romanian and ethnic Hungarian political leaders to see their shared interests and provided a process by which they could work toward agreement.

In the former Yugoslavian republic of Montenegro, PER has engaged in successful preventive action. PER began work there in 1998 after the Montenegrin leadership broke with Serbian leader Slobodan Milosevic and signaled its interest in moving toward more inclusive governance. PER dispatched a program officer, Alex Grigor'ev (now director for the western Balkans), to consult with government, opposition, and community leaders. Grigor'ev's discussions helped establish good relations and trust between PER and local leaders.

In December 1999 PER organized a meeting of Montenegrin officials, Serbian democratic opposition leaders, officials from other Balkan states (including senior officials from Albania — the first such visit to Montenegro in more than a decade), and U.S. and other international representatives (PER 2000c). It seemed increasingly likely that Milosevic would use force against the independent‐minded Montenegrin leadership, but the PER meeting gave democratically inclined Serbian opposition figures an opportunity to reassure their Montenegrin colleagues that they opposed the use of force against them. It also gave international actors an opportunity to reassure Montenegrin leaders that they were pressing Milosevic not to use force.

The meeting facilitated a discussion of interethnic relations inside Montenegro before they became conflictual; an issue of particular concern was the growing demands of the ethnic Albanian minority in Montenegro for greater recognition in the wake of the war in Kosovo. Much of the discussion involved the airing of minority grievances and government efforts to address them. But, by clarifying for participants the constraints on action imposed by international actors, including other states in the Balkans, the meeting encouraged the participants to identify what lay within “the realm of the possible” with respect to improving the status of ethnic minorities in Montenegro. PER was asked by the Montenegrin president and the leader of the largest ethnic Albanian party to follow up this meeting with a dialogue focused specifically on interethnic relations.

In October 2001, after two years of follow‐up efforts, PER brokered an agreement between leaders of Montenegro's Albanian minority, leaders of the republic's ruling coalition parties, and the parliamentary opposition. At a roundtable held in the Montenegrin city of Ulcinj, participants agreed to set the more difficult political issues aside and focus instead on the specific and practical needs of the Albanian community. The participants ultimately reached a consensus on six steps to improve the situation of the Albanians in Montenegro. These included:

  • establishing a maternity hospital in Ulcinj;

  • giving separate municipal status to Tuzi, a region of the capital city of Podgorica with a predominantly Albanian population;

  • establishing an Albanian‐language program at the University of Montenegro;

  • opening a border‐crossing with Albania near Ulcinj, which has a predominantly Albanian population; and

  • consulting with local authorities to appoint ethnic Albanians as the chief of police and head judge in Ulcinj.

In effect, by helping create these initiatives, PER had moved to what Saunders classifies as stages three and four of a dialogue process: disciplined examination of specific problems and identification of options for dealing with them, and designing steps to change troublesome relationships (Saunders 2003: 86).

By early 2006 the agreement had been fully implemented. The most difficult task was establishment of an Albanian‐language university faculty in the republic. In practice it proved extremely difficult to find a location for the school acceptable to all parties. To help break this impasse, in October 2003 PER organized a meeting of key leaders from both the Albanian community and the Montenegrin government and brokered a compromise: the Albanian‐language school would be established in Podgorica, with a liaison office in Ulcinj. At least 50 percent of instruction in the department would be in Albanian, and the new faculty would become the basis for Albanian studies at the University of Montenegro, a program that may eventually include studies in Albanian literature, culture, and history. This compromise was officially endorsed in February 2004 with assistance from PER, and the first students attended classes in Podgorica in September.

This agreement represented an important step forward in interethnic relations in Montenegro because establishing Albanian‐language study at the university level has been highly important to the Albanian community; but it has also been a difficult issue on which to create consensus. The agreement showed what can be achieved when ethnic group leaders move from single‐minded insistence on broader principles to practical negotiations of concrete problems, a change in their focus that PER seeks to facilitate.

In Montenegro, working together to address practical issues helped establish a rapport that facilitated cooperation on other issues. For example, PER helped facilitate development of consensus in Montenegro on minority policy and legislation (PER 2005a). A key factor in the Montenegrin case was the full participation of the Albanian population in Montenegrin institutions. Their electoral weight created important incentives for Albanian leaders to participate in politics and for their Montenegrin governmental partners to cooperate with them. However, the constitutional court recently struck down some important elements of the minority policy legislation. It remains to be seen whether, having achieved its goal of independence from Serbia, the Montenegrin government will continue to cooperate with its ethnic minorities.

In the wake of internal violence followed by a negotiated general agreement brokered by representatives of the European Union and the U.S. (the Ohrid Framework Agreement), and at the request of two Macedonian deputy prime ministers, PER has been bringing Macedonian political leaders together periodically since May 2003 in a project that has come to be known as the “Mavrovo process” (named after the resort town in which most of these meetings take place). These discussions have taken place, for the most part, in an environment “free from the external pressures and influences of constituents and media” (Bercovitch 1996: 29). This freedom has facilitated the emergence of interparty cooperation across the Albanian–Macedonian ethnic divide based on good faith, trust, mutual respect, and, most importantly, the fulfillment of specific agreements. I observed firsthand the emergence of cross‐ethnic cooperation, and the adoption of agreements articulating agreed norms of political behavior when I attended these meetings, and Macedonian and Albanian participants in the Mavrovo process have reported positively on their progress to other regional meetings organized by PER to address other Balkan conflicts.

One product of the Mavrovo meetings has been progress in establishing an Albanian‐language university in Tetovo, which was achieved without international participation. As in Montenegro, the discussions in Macedonia suggest that when agendas can be made more specific, more progress is possible. PER therefore attempts to formulate agendas that encourage engagement in specifics — a process which I have also witnessed. In Macedonia, initiatives are already under way to establish mixed police units; to use the Albanian language in the media, parliament, and public documents; and to achieve proportional representation of ethnic Albanians in public institutions, although none of these issues, as further discussions in Mavrovo in December 2004 and June 2005 made clear, has yet been resolved (PER 2005b, 2005c). Agreements in principle have not yet led to agreements on specifics.

The participants in these meetings identified a number of issues and problems and created a consensus among government and opposition parties that must be addressed. The discussions, in which I participated, were open, candid, and at times quite vigorous. More importantly, however, they were at all times constructive. All participants declared their interest in, and willingness to work toward, mutual progress. These discussions have thus set the stage for a government effort to develop and implement concrete policies. The discussions in December 2004 also made it clear that most of the parties were prepared to enter into an “informal” or “gentlemen's” agreement that would establish an agreed code of conduct for political parties and the media in advance of upcoming local elections. On the basis of these discussions and PER experience with the same issue in Romania, a follow‐up meeting was organized in Skopje in February 2005 at which a code of electoral conduct was adopted for the upcoming local elections. Such an agreement, to which all parties have so far abided, cannot take the place of institutional and procedural changes, but it has helped stabilize Macedonian politics and has moved the Mavrovo process to what Saunders has described as “stage three.”

Two long‐running, controversial issues continue to divide participants of the June 2005 discussions: the display of national (ethnic) symbols (mainly the issue of flags) and the status of national languages. The desire to stay on track toward Macedonian accession to the European Union, however, seems to have motivated most of them to work toward compromise solutions on these and other outstanding issues during a PER‐sponsored meeting in December 2005 (PER 2006). Interventions by a European Union participant, and by the Macedonian minister responsible for accession, exerted significant influence over the discussion.

Undoubtedly the earlier outbreak of violence and intervention of international actors to end the violence and broker the Ohrid Framework Agreement created strong incentives for interethnic cooperation in Macedonia. The inclusiveness and privacy of the Mavrovo process, as well as the focus of the PER moderator on drawing attention to outstanding issues and creating opportunities for all participants to express themselves on these issues, provided an important and constructive context in which Macedonian politicians could explore opportunities for cooperation. Because the meetings have been conducted away from the capital, participants have had many opportunities for intensive interaction outside the organized discussion sessions. For smaller ethnic minority party leaders in particular, the Mavrovo process provided a unique opportunity for detailed, private discussion with senior government officials. Mavrovo also served as the only venue in which the second‐largest ethnic Albanian party, the Democratic Party of Albanians, took part in discussions with other parliamentary parties during its almost year‐long boycott of the Macedonian parliament.

The organized discussion sessions allowed leaders of all parties to hear and evaluate and, if they wished, to respond to the differing views of their peers. The substantive commonalities and differences that emerged in these sessions underscore the importance of other interests — social, economic, and political — in shaping positions on issues often otherwise characterized simply as “ethnic.” The Mavrovo process thus provides support for the view that appeals to identity by elites engaged in political competition represent, to some extent, a rational strategy of mobilization and that even elites engaged in such strategic appeals may be able to deal pragmatically with concrete interests. After two years of Mavrovo meetings, discussions had shifted from debate across ethnic lines to policy debates between government and opposition, each of them multiethnic. A new government elected in July 2006, consisting of former opposition parties, has signaled its support for continuing the Mavrovo process.

PER's first meeting involving ethnic issues in the Kosovo region of Yugoslavia took place with ethnic Albanian leaders in the city of Subotica in the region of Vojvodina, Yugoslavia (Serbia), in 1992. A meeting on interethnic relations in Serbia organized by PER in New York in September 1993 involved only a few local actors in a preliminary, exploratory discussion (PER 1993). This meeting took place outside the region during the wars that raged inside the region in Croatia and Bosnia‐Herzegovina and involved almost as many outside experts as local actors. It is unsurprising, therefore, that local participants tended only to restate positions they had already articulated publicly, rather than engage in a serious effort to find areas of substantive agreement. But the fact that the meeting even took place demonstrated the strength of PER's commitment to foster peace and made it easier for the organization to work inside the region with a local partner organization to enlist wider participation in a follow‐up effort. This suggests that persistence over a significant period of time following even a modest success engenders a certain degree of respect among local actors. This respect is reflected in the quality of local participation in follow‐up events.

After almost two years of effort, PER convened a meeting in Belgrade in June 1995 that brought together representatives of the Kosovar Albanian political leadership, Muslim leaders from the Sandzak region of Serbia, Hungarian leaders from Vojvodina, representatives of Serbian government and opposition parties, and others to discuss “democratic processes and ethnic relations in Yugoslavia.” Representatives of the OSCE and the European Union also attended.

This was the first public meeting between representatives of the Kosovar Albanian political leadership and the Serbian ruling parties in four years. The frank exchanges between Kosovar and Serbian leaders reflected the degree to which fighting in the region had made it difficult to achieve moderation. Nonetheless, PER was able to facilitate an agreement between the Serbian ruling party and the Kosovar Albanian Democratic League of Kosovo (LDK) to continue discussions in the fall of 1995 “aimed at facilitating the beginnings of political negotiations on concrete issues, beginning with the problem of schooling and the official use of the Albanian language in Kosovo” (PER 1995: 22). This PER initiative may have contributed to the later success of the Italian NGO, Community Sant' Egidio, in negotiating a school agreement between the Serbian and LDK leaderships, which was signed by Milosevic and Kosovar Albanian leader Ibrahim Rugova in September 1996.

Following up on its earlier efforts and taking advantage of the end of fighting in Bosnia — which elevated the international significance of the Kosovo issue — PER convened a meeting of Kosovar Albanian and Serbian opposition party leaders in New York in April 1997 (originally planned for 1996 but postponed because of pro‐democracy demonstrations in Belgrade) devoted to peaceful accommodation in Kosovo (PER 1997). Unlike the earlier New York meeting, holding the meeting in the U.S. this time encouraged both participation and a positive outcome. The U.S. was now perceived in the region as the actor that would decide the outcome of conflicts there, and the meeting was organized with the obvious, although still unofficial, support of the U.S. government.

The U.S. government impact on the meeting, however, was not entirely positive. The state department sent senior observers, which lent a degree of official support to the meeting, but Secretary of State Madeleine Albright refused to receive delegates from Milosevic's ruling Socialist Party, while extending this welcome to heads of the Serbian opposition parties. As a result, the Socialist Party did not send a delegation to the talks, leaving representation of the government to its junior coalition party. Serbian democratic opposition parties, Kosovar Albanian opposition parties, and a government party of only limited influence were represented, but not the government itself. Moreover, none of the participants in the meeting was in a position to address the demands of the others. This blunted the effect of the meeting on Serb–Kosovar relations, which by this time were growing more tense in the wake of increasing violence in Kosovo and increasing conflict between the government and democratic opposition in Serbia. The discussion in New York, in which I participated, was characterized by a clash of seemingly irreconcilable political positions taken by the Kosovar and Serbian participants in their opening remarks. At some points, the discussions became highly emotional. Efforts to focus discussion on specific issues, rather than finding general principles on which some agreement might be reached, also ran into difficulty.

PER helped push these discussions toward more productive ground when Kassof intervened to state that the historical arguments made by each side were “right” and that neither side would be likely to persuade the other, nor outside actors, to support its position. This approach was picked up by one of the Kosovar Albanian participants, who offered a list of modest principles for developing frameworks for negotiating future agreements, rather than developing principles to address substantive issues in dispute. Although PER usually encourages parties to develop their own solutions, the participants in New York were unwilling or unable to do so until one of the outside expert participants began to “shuttle” back and forth between Albanian and Serbian participants and caucus with them separately during a break late in the meetings to help each side achieve some coherence and to facilitate discovery by the two sides of mutually acceptable language in which to frame an agreed set of principles. These principles became the object of extended discussion and, eventually, led to agreement on a set of “jointly agreed positions” (PER 1997: 12).

The New York meeting highlighted the deep differences between Kosovar Albanian and Serbian views on Kosovo and among actors on each side. These differences made it difficult to achieve agreement even on general principles. The Kosovar situation contrasted with what was happening in Romania where Hungarian leaders decided to participate in Romanian politics, even though they were initially divided over whether they should pursue that strategy or a secessionist one. They formed a party, competed in Romanian elections, won seats in parliament, and became a significant factor in Romanian politics. The Hungarian community was thus clearly committed to achieving equality within the framework of the existing state. The Kosovar Albanian leadership, in contrast, opted to pursue independence. The modest agreement achieved in New York therefore proved to be temporary, and the conflict inside Kosovo rapidly escalated soon thereafter. In February 1998, after a year of preparatory discussions, PER persuaded the Serbian ruling parties to join the talks. But the outbreak of violence in the Kosovo region of Drenica the following month made it clear that it was too late, and PER was unable to make any progress on its Kosovo agenda until after the war.

Following the war in Kosovo, it was no longer possible to address the Kosovo issue independently from its Balkan context. The status of Kosovo had become part of a larger “Albanian question” affecting not only Serbia, but Macedonia and Montenegro, as well as Albania itself. Moreover, resolution of the Kosovo issue held important implications for developments in Bosnia, as well as other minority issues in the region, such as the status of the Roma. Thus, the number of groups and actors interested in Kosovo multiplied, making it more difficult to maintain the “track one and one‐half” approach that PER had pursued in the period before the war. Additional senior government officials and international actors were now more interested in participating in any serious effort to resolve the conflict in Kosovo. The war also created a significant break in the process, and in some ways rendered it moot — many of the actors had changed, with powerful new actors emerging on both sides. So, in April 2000 when PER convened the first meeting of Serbian and Kosovar Albanian leaders since the war, they structured it differently than preceding meetings.

That meeting, which took place in Budapest, had a regional focus. It was much larger in size, with participation of leaders from more countries, and significant representation of international actors (PER 2000d). In light of the changed circumstances, it was unsurprising that discussion returned to the “articulation of grievances and demands” that had characterized the meeting in New York in 1997. Participants focused on justifying their own past actions and advancing their own interpretations of the current situation. This behavior was encouraged by the presence of significant international representation in the meetings, and the presence of more official participants from more countries in the region. This, in effect, transformed the process from the “private diplomacy” or “track one and one‐half” approach of earlier meetings to something closer to multilateral diplomatic negotiation.

The participation of international actors had contradictory consequences for this meeting and later PER meetings dealing with the Kosovo issue. The articulation and confrontation of conflicting realities is, of course, a necessary part of the process of conflict resolution, but the PER meetings in which I participated suggest that it is more likely to lead to progress if it takes place outside the view of international actors. The presence of significant international actors makes it difficult for participants to change their positions, thus discouraging open dialogue and exploration of the issues among the conflict parties themselves. The conflict parties focused their efforts on influencing international actors rather than on having a constructive dialogue with each other. At the same time, however, the participation of international representatives and senior governmental and diplomatic officials from several states in the region had an important positive effect: it highlighted the strong interest of international actors in avoiding violence, resolving the issue peacefully, and maintaining regional stability. European participants delivered a clear message that the region's states would be unable to join the European Union until their local conflicts were settled. The participation of international actors thus helped convince local actors that they would have to negotiate and made them more willing to participate in a follow‐up meeting organized by PER in Athens in December 2000, only eight months later.

The Athens meeting, for which I served as rapporteur, was a similarly large one, again involving representatives of many governments and international organizations (PER 2001c). The dramatic change in Yugoslav government and politics that took place in October 2000 with the ouster of Milosevic and the corresponding change in Serbian government and politics certain to result from imminent elections, increased the perception among regional actors that meaningful dialogue and significant change was now possible. These circumstances increased the perceived importance of this meeting. Some local participants again chose to make defensive and demanding speeches and to direct their comments toward the international community rather than to their peers. Indeed, one participant, representing a minority political party participating in the governing coalition in his state, unexpectedly presented a comprehensive set of demands for radical change that were met with astonishment and surprise by a representative of that government. The response of most other participants made it clear that they considered such demands inappropriate and put PER leadership in a position to advise this actor to change his approach if he wished to achieve any of his more practical goals. (This episode echoed PER experience in Romania, reported above.)

The Athens meeting was also characterized, however, by the emergence of a genuine effort on both the Kosovar Albanian and Serbian sides to begin a process of engagement, a development that can be attributed to a number of factors. Perhaps the most important was the fact that the change in Yugoslav politics brought new actors to the table. Several members of the Serbian delegation proved to be extraordinarily open and honest in their approach to sensitive questions, and thereby elicited equally open and honest responses from some Kosovar representatives. The PER moderator helped to focus the meeting on this emerging change, rather than allowing this development to be overwhelmed by the many different issues that other participants had been raising up to this point. With the support of PER, and as the result of discussions outside the formal sessions, the Kosovar and Serbian participants moved toward agreement on the desirability of convening a direct dialogue between them.

A PER participant noted that a Serbian delegate and a Kosovar delegate each made constructive comments in the morning discussion and approached each one separately during a break, suggesting that these comments represented an opportunity. He then suggested to PER president Kassof, who was serving as moderator, that he begin the next round of discussion by recalling these remarks and asking whether these delegates wished to pursue them, which enabled these delegates to engage in a direct and constructive exchange. This illustrates how PER facilitates agreement, and the organization's general approach to mediation: identify emergent opportunities and encourage them. It also appeared to move the process closer to “stage three” of the Saunders classification.

In response to this exchange, some participants argued that, with impending elections in Serbia that would likely further advance democratization efforts there, democratic elections should be required in Kosovo in order to establish a symmetrical dialogue. The idea of a direct dialogue, however, and especially the idea of holding elections in Kosovo, met with stiff opposition from some international participants, who seemed to fear losing control over the process of defining Kosovo's status. This turn of events illustrated one of the tensions inherent in the participation of national and international representatives in PER‐sponsored processes: PER seeks to encourage local initiative but international actors may be reluctant to cede control over local events.

In an effort to encourage dialogue, PER organized a meeting in Pristina, Kosovo in April 2001, which brought together leaders of all of Kosovo's ethnic communities and the Kosovar Albanian leadership for the first time to discuss interethnic relations in Kosovo. In October 2001, PER organized a meeting between Serbian Deputy Prime Minister Nebojsa Covic and the leaders of the three largest Kosovar Albanian political parties to discuss participation by Kosovo's ethnic Serbian minority in the elections scheduled to take place in Kosovo in November 2001. This meeting only demonstrated “how difficult it is to establish any communication between Belgrade and Pristina” (PER 2002b). The dialogue called for at the Athens meeting in December 2000 was put on hold until after elections in Kosovo, and three months of negotiations among the three largest Kosovar Albanian parties in the newly elected parliament produced a democratically legitimated leadership that could enter into symmetrical dialogue with Serbian/Yugoslav representatives.

With a government in place in Kosovo by March 2002, PER convened another meeting devoted to “Albanians and their neighbors” in Lucerne, Switzerland, in November (PER 2002c). At that meeting, Covic apologized publicly to Kosovo Prime Minister Bajram Rexhepi for statements he had made a year earlier about Mr. Rexhepi, attributing them to inaccurate information and expressing his sincere regrets. The PER meeting in Lucerne was also the occasion for the first direct government‐to‐government talks between Pristina and Belgrade since the end of the war, talks that would have been impossible without Covic's apology and Rexhepi's acceptance of it. Both men showed the kind of goodwill and interest in finding a solution to their common problem that had made it possible for PER to function so effectively in Romania. PER's role as intermediary between the two sides in the process of preparing the meeting and the immediate follow‐on discussions was also essential. It is unlikely that PER could have played this role were it not perceived to be an “honest broker” with a long‐term commitment to seeing the process of reconciliation through to a successful conclusion. Indeed, the Serbian Deputy Prime Minister immediately requested that PER preside over any government‐to‐government meetings that might result from the November 2002 agreement. Rexhepi, however, was widely criticized when he returned to Pristina, and nothing came of this agreement.

The next meeting devoted to “Albanians and their neighbors” was not convened until May 2004, again in Lucerne (PER 2004). This delay was the result of the outbreak of renewed violence in March 2004. Similar in format to the earlier meetings, the May 2004 discussions underscored the growing contrast between positive developments in relations between Albanians and their neighbors in Montenegro and Macedonia on the one hand, and relations between Albanians and Serbs in Kosovo on the other. While circumstances in Montenegro and Macedonia differed in important ways from those in Kosovo and Serbia (the most important of which was the full participation of Albanian political parties in the democratic electoral politics of Montenegro and Macedonia), the discussions in Lucerne made clear the need for more intensive discussions among Kosovo Albanian and Serb leaders, among Serbian leaders, and, eventually, between Kosovar and Serbian leaders. If positive developments were to be achieved from such discussions, however, PER experience in Montenegro and Macedonia made it clear they would have to be focused on specifics.

PER attempted to pursue a strategy of focusing on specifics by convening further meetings with Serb and Albanian leaders in Pristina in June 2004, with senior officials from the region in Bucharest in November 2004 (PER 2005d), and with Kosovo Serb leaders in Pristina in April 2005 (PER 2005e). The April 2005 meeting produced an agreement by the Serbs to end their boycott of Kosovo institutions. Follow‐up was left in the hands of U.S. officials, however, who failed to achieve implementation. Later discussions proved even less productive than earlier efforts, as political developments in Serbia and Kosovo — and especially the U.S. drive to initiate resolution of the status issue that began in June 2005 — led to a hardening of positions on all sides. The positions expressed during discussions convened in Lucerne in July 2005 were simply irreconcilable, leading some participants to conclude that dialogue on this issue was no longer possible, and to raise the prospect of an imposed solution (PER 2005f). Direct talks between the Kosovo Albanians and the Serbian leadership, carried out in Vienna in the spring and summer of 2006 as part of the U.S. initiative to resolve the status question and mediated by U.N. Special Envoy Martti Ahtisaari failed to reconcile the two sides, making an imposed solution all but inevitable.

The PER experience in the Balkans suggests that NGO mediation can play a constructive role in resolving ethnic conflicts. PER's ability to help resolve ethnic conflict arises from:

  • its leaders' expert knowledge and extensive networks of contacts;

  • its organizational persistence, patience, and presence in the region of conflict;

  • its perceived influence with major international actors; and

  • its carefully nonpartisan approach to conflicts.

Paradoxically, success can also be attributed to PER's powerlessness, that is, its inability to challenge the power of governments or the claims of their opponents. This has meant that participation in PER activities carries few political risks or costs for local actors. The effectiveness of PER processes is also a product of the advance preparation of meetings through discussion with prospective participants, the flexible format of discussions that allows the moderator to pursue opportunities and abandon dead ends, and, especially, the provision of considerable opportunities for informal contact and conversation during the course of meetings.

The effectiveness of PER's efforts have depended greatly, however, on the level of interest of the conflicting parties in resolving their common problem. The cases of Romania, Montenegro, and Macedonia all suggest that competitive electoral politics may create opportunities and incentives for cross‐ethnic cooperation in the quest for power. The PER experience in these countries suggests that an NGO can help local actors to perceive such opportunities created by events occurring “away from the table,” encourage parties to exploit them, and create circumstances in which parties can establish cooperation. The trust established by PER with local actors allows PER to push them toward more pragmatic positions that make compromise possible, but still serve their interests. But neither PER nor any other NGO can create such interests; they can only help parties to see them. The existence of competitive electoral politics, therefore, seems to be a necessary, but not sufficient, condition for the successful application of the PER model of NGO mediation of ethnic conflict.

The contrast between PER experience in Romania, Macedonia, and Montenegro with its far less successful experiences in Kosovo and Serbia suggests the extent to which the belief of local actors that concrete political gains can be achieved through compromise and cooperation constitutes a defining limit on the ability of any NGO to mediate conflict. In Romania, Montenegro, and Macedonia, political actors shared a common interest in securing electoral power. In Romania, the cooperation of some Romanian parties with the Hungarian minority offered them political advantages over their domestic opponents and leverage in relations with neighboring Hungary, NATO, and the European Union. In Montenegro, the electoral support of the Albanian minority was crucial to the legitimacy of the president and the government. In Macedonia, while cooperation with an Albanian party was not required to secure governmental power, it was required to secure domestic peace, and to legitimize the state domestically and internationally. As in Romania, interethnic cooperation in Macedonia has brought dividends with respect to relations with critical neighbors (Albania and Kosovo), and the European Union. Interethnic peace is a critical factor with respect to Macedonia's accession to the European Union. Over time, cooperation across the ethnic divide in Macedonia has also become a crucial political asset for each government. The political advantage to be gained through cooperation creates additional incentives for local actors to recognize the opportunities created by PER activities and to heed the advice offered by PER when it identifies opportunities for mutual gain.

In Serbia and Kosovo, in contrast, Kosovar Albanians and the Serbs (in Serbia and, for the most part, in Kosovo) do not perceive any common interests. Unlike the other states of the region, here the question of state borders, and whether coexistence in a single state is possible, has not been resolved. The views of the conflicting parties seem to be irreconcilable. The “stakes” of the Kosovo conflict are therefore much higher than those of the conflicts in Romania, Montenegro, and Macedonia. The very characteristics that contribute to the effectiveness of PER in these other states — patience, neutrality, and especially powerlessness — make it difficult for PER to affect the perceptions and calculations of Albanian and Serb political actors. And, most importantly, the initiative taken by the U.S. to resolve the future status of Kosovo appears to have eliminated any incentive for the two sides to compromise. Once a state actor with the power to impose a solution has committed itself to ending a conflict, its preferences outweigh any local interests in determining the outcome and can render irrelevant the efforts of a “weak mediator” such as PER. Any organization committed to facilitating the ability of the groups to find their own compromise solution to the conflict is, under such circumstances, forced to the sidelines.

Bercovitch
,
J.
1992
.
The structure and diversity of mediation in international relations
. In
Mediation in international relations
, edited by
J.
Bercovitch
and
J. Z.
Rubin
.
New York
:
St. Martin's Press
.
Bercovitch
,
J.
1996
.
Introduction: Thinking about mediation
. In
Resolving international conflicts
, edited by
J.
Bercovitch
.
Boulder, CO
:
Lynne Reinner
.
Carnegie Commission on Preventing Deadly Conflict
.
1997
.
Preventing deadly conflict
. New York:
Carnegie Corporation
.
Hume
,
C.
1994
.
Ending Mozambique's war
. Washington, DC:
USIP Press
.
Kassof
,
A.
2001
.
Interview with the author
. June 7, Princeton, NJ.
McMahon
,
P.
2005
.
Managing ethnicity: The OSCE and trans‐national networks in Romania
.
Problems of Post-Communism
1
:
15
27
.
Nan
,
S. A.
2004
.
Track one and a half diplomacy: Searching for political agreement in the Caucasus
. In
NGOs at the table
, edited by
C.
Church
and
M
.
Fitzduff
. Lanham, MD:
Rowman & Littlefield
.
Plaks
,
L.
2001
.
Interview with the author
. June 7, Princeton, NJ.
Plaks
,
L.
2003
.
Personal Communication
. January 15 (email).
Project on Ethnic Relations
.
1991
.
Romanian‐American symposium on inter‐ethnic relations
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
1993
.
Interethnic relations in Serbia/Yugoslavia: Alternatives for the future
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
1995
.
Democratic processes and ethnic relations in Yugoslavia
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
1997
.
The New York roundtable: Toward peaceful accommodation in Kosovo
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
1998
.
Schools, language, and interethnic relations in Romania: The debate continues
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
1999
.
Building Romanian democracy: The police and ethnic minorities
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2000a
.
Toward community policing: The police and ethnic minorities in Hungary
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2000b
.
The year 2000 elections in Romania: Interethnic relations and European integration
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2000c
.
Montenegro on the brink: Avoiding another Yugoslav war
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2000d
.
Albanians and their neighbors: Unfinished business
. Princeton, NJ:
Project on. Ethnic Relations
.
Project on Ethnic Relations
. 
2001a
.
Political will: Romania's path to ethnic accommodation
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2001b
.
Bulletin
. 16. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2001c
.
Albanians as majorities and minorities: A regional dialogue
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2002a
.
Political extremism and interethnic relations in the new millennium
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2002b
.
Yugoslavia at the crossroads
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2002c
.
Albanians and their neighbors: Is the status quo acceptable?
Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2004
.
Albanians and their neighbors: Moving toward real communication
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2005a
.
Developing a minority policy in Montenegro
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2005b
.
Macedonia: The next stage
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2005c
.
Macedonia: On the road to Brussels
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2005d
.
Kosovo 2005: Assuring security for the neighborhood
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2005e
.
Interethnic relations in Kosovo: Toward implementation of standards
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2005f
.
Kosovo and the region prepare for change: Relations, responsible governance, and regional security
. Princeton, NJ:
Project on Ethnic Relations
.
Project on Ethnic Relations
. 
2006
.
Macedonia: Agenda 2006
. Princeton, NJ:
Project on Ethnic Relations
.
Pruitt
,
D. G.
2002
.
Mediator behavior and success in mediation
. In
Studies in international mediation
, edited by
J.
Bercovitch
. Houndmills and New York:
Palgrave Macmillan
.
Royal Institute of International Affairs. N.D.
The Chatham House rule
. http://www.chathamhouse.org.uk/index.php?id=14.
Sannino
,
S.
2001
.
Personal Communication to Allen Kassof
. September 6 (letter).
Saunders
,
H. H.
2003
.
Sustained dialogue in managing intractable conflict
.
Negotiation Journal
19
:
85
95
.
Wilson
,
D. G.
and
W. F.
Walsh
.
1995
.
Reflections on the transfer of knowledge to support democratic policing in Hungary and Romania. Policing in emerging democracies: Workshop articles and highlights
. National Institute of Justice Research Report. Washington, DC, National Institute of Justice, U.S. Department of Justice and Bureau of International Narcotics and Law Enforcement Affairs, U.S. Department of State.
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0) license, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. For a full description of the license, please visit https://creativecommons.org/licenses/by/4.0/legalcode.