The September elections in Bosnia highlighted what was until then an implicit aspect of the current peace: it is more likely to move Bosnia toward the ethnic states for which the war was fought than to reestablish the multiethnic Bosnia that once was. Indeed, as the Dayton process unfolds, it becomes clearer that the peace agreement signed in November 1995 after three and a half years of war was something historically familiar: a so-called peace accord that is in reality a partition agreement with an exit clause for outside powers.

At the same time, while key aspects of the document, such as the creation of two "entities" with virtually separate legislatures, administrations, and armies, tend toward partition, the pact attempts to get around some of the more hostile legacies of partition through a common economic space and arms control, and it creates structures that could reverse the partition process by returning refugees and rebuilding civil society. So far, these structures have been dormant, and the holding of national elections in a still highly uncertain peace marks the tilt toward partition. As was widely predicted, the Bosnians gave their ethnic leaders new mandates, and Bosnia took another step toward partition. However, the postponement of the municipal elections due to irregularities in voter registration means the international community is not yet in a position to accept partition as the democratically expressed will of the people.

The Bosnian war and the Dayton peace agreement have reignited a debate on whether partition is an effective solution to ethnic conflict. Although Bosnia is the starting point, the arguments in this debate have broad resonance at a time in which the rapid spread of ethnic and communal wars east and south of Bosnia is of increasing concern to the international community. Defenders of partition make an argument that runs as follows. When an ethnic war is far advanced, partition is probably the most humane form of intervention because it attempts to achieve through negotiation what would otherwise be achieved through fighting; it circumvents the conflict and saves lives. It might even save a country from disappearing altogether because an impartial intervenor will attempt to secure the rights of each contending ethnic group, whereas in war the stronger groups might oust the weaker ones. In fact, its advocates say, the ideal strategy for resolving an ethnic conflict is to intervene and take partition to its logical conclusion by dividing a country along its communal battle lines and helping make the resulting territories ethnically homogeneous through organized population transfers. This will ensure that partition is more than a temporary means of containing conflict. Less thorough partitions, however, can still be a lasting means of containment.-1

Partition, however, has its own sordid history, not arising as a means of realizing national self-determination, but imposed as a way for outside powers to unshoulder colonies or divide up spheres of influence -- a strategy of divide and quit. Although described as the lesser of two evils, the partitions in Cyprus, India, Palestine, and Ireland, rather than separating irreconcilable ethnic groups, fomented further violence and forced mass migration. Even where partition enabled outside powers to leave, as in India, it also led to a disastrous war. Often thought of as a provisional solution, it has been unable to contain the fragmentation it triggers among dispersed or overlapping ethnic groups that are not confined by neat geographic boundaries, and it gives birth to weak civil institutions demanding supervision. Similar conditions ensure that the partition of Bosnia, which from the start should have been reintegrated, will also amount only to a policy of divide and be forced to stay. The Dayton accords should not evoke memories of Munich, but rather of Cyprus.


The argument for ethnic partition is not new, but its terms changed considerably over this century before settling upon the current rationale of the lesser of two evils. Before World War I, most partitions were effected for the needs of empire, to strengthen rule or simplify administration. After 1918, however, colonial empires were increasingly challenged, and subsequent partitions took place as part of a devolution of authority or a Cold War policy of spheres of influence. There were two distinct rationales for the partitions resulting from the fall of colonial empires: Wilsonian national self- determination, applied to Poland and Romania, and the British colonial policy of identifying irreconcilable nationhoods, applied in Ireland, India, and, as a delayed response, Cyprus and Palestine. Though both rationales took ethnic identity as an important determinant of political rights, Wilsonian policy supported ethnic self- determination as freedom from colonial rule, while the British reluctantly espoused partition as a lesser evil than constant civil war.

After the last attempt to ratify a partition -- Cyprus after the Turkish invasion in 1974 -- the notion that partition was an effective solution to ethnic conflict fell into disuse for a quarter-century. Paradoxically, its revival followed hard on the heels of German reunification and the potential integration of Europe that it heralded. In the first phase of the revival of partition theory, Wilsonian self- determination was invoked more often than the lesser-evil argument. Indeed, the prevailing feeling was that the end of the Cold War -- and the relatively peaceful dissolution of the Soviet Union -- meant that separations could be negotiated. In the early 1990s the most frequently cited example of a peaceful negotiated division was Czechoslovakia's "velvet divorce." When asked on The News Hour with Jim Lehrer in November 1995 whether the Dayton agreement was a partition, Assistant Secretary of State Richard C. Holbrooke said he preferred the example of Czechoslovakia's voluntary dissolution. But fewer people now refer to the Czech split. That the Czech Republic and Slovakia were relatively homogeneous and that dissolution of the federation did not require an alteration of internal borders or a substantial displacement of people make the comparison with Bosnia untenable. A comparison between Bosnia and the partitions of Ireland, India, and Cyprus, or the incomplete partition of Palestine, would be better, because each involved ethnically mixed and dispersed populations and each was held to be a pragmatic recognition of irreconcilable ethnic identities.

It is worth examining these partitions' relevance to Bosnia in more detail. All relied heavily on the lesser-evil argument, but in at least two of them the decision for partition was prompted not by a desire for peace and self-determination, but because the colonial power, Britain, wanted to withdraw. The recognition of irreconcilable nationhoods followed as a consequence -- it would be easier to withdraw quickly if the aims of the ethnic leaders were fulfilled by territorial grants. Looking back on the 1947 partition of India in 1961, former civil servant Penderel Moon summed up as "divide and quit," in a book of the same name, the British policy of pushing partition through without establishing the boundaries of new states or planning for the wars that might ensue; it was the post-World War II imperative of quitting that drove the decision to divide, he said. It was arguably the post-World War I imperative of quitting the Irish conflict that led the British to espouse a partition of Ireland.

That both divisions were driven by considerations extraneous to the needs and desires of the people displaced does not necessarily mean that partition was not a solution to their conflicts. However, as in India and Ireland, partition has more often been a backdrop to war than its culmination in peace; although it may originate in a situation of conflict, its effect has been to stimulate further and even new conflict. Indeed, India's experience raises the question of whether a peaceful transition to partition is possible. India's political leadership agreed to partition the country before the spread of large-scale conflict; the 1947 partition agreement between the Indian National Congress and the Muslim League was intended partly to prevent the spread of communal riots from Bengal in eastern India to northwestern India, which was also to be divided. But the riots that followed in 1947-48 left more than a million people dead in six months and displaced upwards of 15 million.

Moreover, partition arises in high-level negotiations long before it becomes evident on the ground. The British partition of Ireland in 1921 was a late addition to negotiations for home rule during the 1919-21 Anglo-Irish war for independence, but partition had been on the drawing board since 1912, when it was suggested by a group of conservative and liberal members of parliament that Protestant-majority counties be excluded from the proposed Irish Home Rule Bill. Calls for partition were renewed in 1914, 1916, and 1919; the offer of a double partition of Ireland and Ulster based on religion led to the spread of conflict between English and Irish across the south, west, and north of Ireland, escalating to guerrilla warfare when Catholic rebels formed the Irish Republican Army in 1919. Nor did the war end in 1921 when Britain negotiated a treaty with Sinn Fein, the political arm of the IRA, offering dominion status to southern Ireland in return for a separate Ulster under British administration. The decision to accept partition led to a split in Sinn Fein, and internecine conflict was added to communal conflict, ending two years later with the defeat of the faction led by Eamon De Valera. It took almost four years of war to achieve the partition of Ireland, and those four years were themselves a culminating phase in a movement toward partition that had begun ten years earlier.

Significantly, the British rejected the partition option in Palestine in the same years that they espoused it in India. The two reasons they gave were infeasibility and the risk of a military conflict that would involve an expanded British presence. Although partition had been proposed in 1937 by the Peel Commission, which concluded that cooperation between Jews and Arabs in a Palestinian state was impossible, and had been the subject of debate in Britain throughout the 1930s, in 1946 the British members of the Anglo-American Committee of Inquiry argued that because ethnic groups were so dispersed, partition would entail massive forced population transfers, and that the territories created -- a tiny Arab state, a Jewish state in two parts, and three blocs under continuing British administration -- would be infeasible. Moreover, they said, moves toward partition could cause a war. In 1947 the British referred the dispute to the United Nations. The Security Council opted for partition, with a special U.N. regime for Jerusalem and a continuing economic union for the whole of Palestine. The plan required Britain to undertake a substantial role in its implementation, but after the Ministry of Defence forecast that Britain's military presence would have to be reinforced in the wars that would follow, Britain announced that it would withdraw in May 1948. In April the Jewish Agency, which represented the Jewish community under the British mandate, announced that it would declare a Jewish state when the British withdrew. War broke out, resulting in a kind of skewed partition by which one new state was created but not the other. Subsequently there have been three Arab-Israeli wars, and the issue of territorial feasibility continues to dog the peace process.

In many ways Cyprus offers the most striking parallels to Bosnia, and its history again raises the question of whether a peaceful transition to partition is possible. Although the British proposed the partition of the island in a divide-and-rule move in 1956, they subsequently rejected the plan on the same grounds as in Palestine -- infeasibility and the risk of conflict. The British-brokered constitution of 1960 that made Cyprus independent was an attempt to avert division of the island between ethnic Turks and ethnic Greeks, but the idea that ethnic politics could be contained by providing for ethnic representation at every level proved a failure. The constitutional creation of separate municipalities and a distribution between the two ethnic groups in the presidency, legislature, civil service, police, and army added communal (that is, interreligious) conflict to internecine conflict. In 1963 the "Green Line," the first partition boundary to be drawn, divided Greek and Turkish Cypriots in Nicosia, the capital. Ethnic conflict only intensified, and a Turkish Cypriot declaration of support for partition followed in 1964. Although U.N. troops arrived that year, tensions escalated, with a counter- declaration of unification by Greece and Cyprus in 1966, a military coup in Greece, renewed conflict in Cyprus, a Turkish Cypriot announcement in 1967 of a Provisional Administration, increasing Greek support for the radical Greek underground in Cyprus, and finally a Turkish invasion in 1974 that reinforced the de facto partition of the island. Thus it took 14 years to establish what continues to be a shaky partition of Cyprus.


How successful have these partitions been at reducing conflict and permitting outside powers to end their involvement? It is not clear that the partitions of Ireland and Cyprus can be said to have worked, even in the lesser-evil sense. Although the former was a move to divide and quit -- in which all sides accepted division as the price of self-determination -- the British are embroiled in a military operation in Northern Ireland that continues 70 years later. The troop presence curtailed the toll that communal conflict might otherwise have taken; indeed, it could be argued that it contained the Irish conflict and kept deaths to a minimum. But it also brought the conflict to the heart of Britain as the IRA mounted terrorist attacks in London to increase pressure for a British withdrawal, and it could just as well be argued that from the British point of view independence would have been a more effective way to contain the conflict because it would have thrown the onus of peace onto the Irish; moreover, it might have encouraged regional compromises rather than a prolonged stalemate.

The partition of Cyprus can only be described as a partition by default that the U.N. presence inadvertently aided. The conflict following independence in 1960 was compounded by the fact that Turkey, Greece, and Britain were appointed protecting powers by the constitution. The formal structure this gave to a wider engagement in the conflict drew both the Greek and Turkish armies in and permitted international acceptance of Turkey's invasion in 1974 and what was until then a de facto partition. While casualties have been restricted since then, the division of Cyprus is little more than a long standoff that remains volatile and continues to require the presence of U.N. troops. Nor can the conflict be confined to Cyprus. Over the 20 years since partition, its short fuse is evident. A violent demonstration by Cypriots in August 1996 resulted in Greece and Turkey threatening war. The costs of containment, therefore, include permanent vigilance on the part of NATO and the Atlantic allies.

In many ways, despite the violence and displacements it produced, India's was the most successful ethnic partition, both because it allowed the British to quit and because the conflicts that ensued were by and large contained. But this had less to do with the wisdom of ethnic separation than with other factors, among them the subcontinent's distance from Europe. Unlike Ireland, Cyprus, and Bosnia, the Indian subcontinent is so large that a dozen or more new states could have been created. The deployment of the ethnic two-nation theory, however, which holds that Hindus and Muslims could not live together, had a paradoxical effect -- the new state created, Pakistan, was divided into two parts by roughly 2,000 miles of Indian territory. The subsequent separation of those parts points up the inadequacy of the principle of ethnic separation for effecting stable territories. In the late 1960s, resentment at West Pakistani political and economic dominance led to a regional Bengali movement for independence, a war between the two parts in which India intervened in support of the Bengalis, and the birth of Bangladesh in 1971.

In regions of multiple ethnicities -- where, for example, the same individual might have loyalties to one community defined by its religion and another by its language -- attempts to make one ethnic identity dominant can trigger further fragmentation and conflict. The temporary success of the Indian People's Party in whipping up Hindu nationalism during the destruction of the Babri mosque in Ayodhya and in the riots that followed in the winter of 1992-93 ultimately led to the party's isolation and failure to form a government after the 1996 elections. The case of Kashmir is more poignant. Since 1947, India and Pakistan have been embroiled in a conflict that has twice flared into war, over what has been described, in a phrase dear to politicians on both sides, as "the unfinished business of partition": Kashmir. On ethnic grounds it can be argued that the conflict has continued because India retained the Muslim- majority Kashmir Valley, which should have gone to Pakistan. But following ethnic dividing lines could well entail a further three-way partition of the state -- the valley, Buddhist Ladakh, and multiethnic Jammu -- which would not only set the stage for intensified conflict and ethnic cleansing, as much of Jammu lies between Pakistan and the valley, but would also dissolve Kashmir.


Bosnia-Herzegovina, like these other partitioned territories, has problems of dispersed populations and continuing fragmentation, which the Dayton agreement shows little promise of resolving. It is difficult to contain a conflict when partition is still in progress. Thus NATO and Implementation Force (IFOR) officials under its auspices increasingly worry that hasty implementation of civilian aspects of the agreement such as elections might renew the conflict; instead of reversing partition or facilitating its peaceful execution, the prelude to the elections brought renewed low-level conflict in August. To this extent the Bosnian elections are to the Dayton process what last year's Israeli elections were to the Palestinian peace process: they prove that partition is still incomplete on the ground. Nearly half the pre-1992 population of Bosnia is still living as refugees outside Bosnia. Ethnically homogeneous territories can be created only if the refugees are refused repatriation to the towns and villages from which they were driven, which puts pressure on their host countries not to enforce repatriation. The refugees have become a key constituency that is used both to further and to challenge the consolidation of such ethnic territories, as in the Serb efforts to force refugees to vote from towns other than their prewar residences, and the subsequent threat of Bosnian Muslims to boycott the elections.

Both IFOR and the risibly named Office of the High Representative, which oversees the civilian implementation of Dayton under former Swedish Prime Minister Carl Bildt, make no secret of a concern the elections highlighted: the third partition that always hovered in Dayton's wings, between the Herzegovinian Croats and Bosnian Muslims now living together in the Bosnian Federation. Since the 1994 agreement signed in Washington that established the federation -- which was intended to limit Bosnia's partition but produced a constitution remarkably similar to the 1960 Cyprus constitution -- Herzegovinian Croats have asked: If a two-way partition is acceptable, why not a three-way one? Unsurprisingly, the answer, that Croatia has already been offered a far more substantial quid pro quo than Serbia and that a tiny and probably landlocked Muslim Bosnia would perpetuate Muslims' resentment, does not satisfy Herzegovinians. But Muslim resentment should at least give partition revivalists pause. Indeed, their argument that the United States should jettison the federation in favor of a tripartite partition begs both the resentment issue and a related matter Bosnian nationalists have raised: that they be given Serbia's predominantly Muslim Sandjak province as territorial compensation. Moreover, their argument ignores the problematic relationship between Croatian and Herzegovinian Croats, whose distrust of each other rivals that between Croats and Muslims.

NATO and IFOR have also pointed out an even more crucial concern: a partition dependent on the awkward boundary line between the Bosnian Federation and Republika Srpska, the Bosnian Serb entity, can last only so long as a large international force is there to enforce it. It is not Mostar in the federation but Banja Luka in the Serb entity that may bring the simmering partition war to a head. One look at the map of the Republika Srpska shows why. Like Pakistan after the partition of India, the Serb entity is divided into two parts, connected only by the narrow Posavina corridor, in which the disputed town of Brcko is key; Serb attempts to rig local elections there were a major factor in the postponement of municipal elections in Bosnia last year. Additionally, the two main parts of the Serb entity lean in opposite directions, Banja Luka toward Zagreb and the eastern strip toward Belgrade. Normalization would again pull Banja Luka toward Zagreb economically and diminish its links to the east. That might mean a further division of the Republika Srpska, rather like the division of Pakistan that created Bangladesh, in which the Serb republic would be reduced to a strip of eastern Bosnia. Banja Luka, therefore, must be forced to look eastward, and is -- with tacit U.S. support, if Richard Holbrooke's recent suggestions that the Bosnian Serbs make Banja Luka their capital are anything to go by. But Banja Luka's isolation amid federation territory can be maintained only if Serb leaders keep the city in a state of anarchy and mafia rule, like Mostar. Banja Luka's location, however, makes this task much more difficult because Mostar, which remains integrated with Croatia despite being incorporated into the Bosnian Federation, is close to the Croatian border. It is debatable whether anything short of a fortified wall will keep Banja Luka isolated.

Thus, while elections may well be a step toward ratifying partition politically, efforts to consolidate a partition will not only perpetuate conflict but will eventually show that Bosnia can be successfully divided into two only if the Republika Srpska is further partitioned, with the western part reintegrated into Bosnia and the eastern part joining Serbia. Paradoxically, this may be the only way of ensuring a stable partition under the Dayton agreement, as the reintegration of Banja Luka will help keep the Croats in the federation. But if a multiethnic Bosnia can be re-created in one part, then why not in the whole, especially since a further partition will renew conflict around Gorazde in eastern Bosnia?


As pressure mounts to accept the September elections as a mandate for partition, more emphasis is being placed on the reintegration option of the Dayton agreement, which assumes that economic interests and the provisions for a common economic space will erode the partition lines by making them irrelevant. It is being argued that the partial partition the Dayton agreement partly accepts is only a means of buying time for Bosnia to undergo this process. Historically, however, the failure to inject substantial and timely aid has only hardened ethnic divisions.

Partition has rarely been seen as anything other than a temporary solution to a crisis, which can be reversed as the crisis recedes. However, ethnic partitions have never been reversed; their implementation has inexorably driven communities further apart. Sinn Fein's acquiescence to the partition of Ireland was on the condition that there be a referendum on unification; the referendum did not take place, and now that negotiations on the status of Northern Ireland have been revived, Sinn Fein faces the ironic possibility that Ireland may no longer want unification.

Ethnic partition can often hamper the development of postwar economies. Although economic cooperation could improve South Asia's economies enormously, the ongoing conflict between India and Pakistan over Kashmir has impeded attempts to build it. The Dayton agreement's hope that economic interests will militate against ethnic boundaries was also voiced in Ireland and Palestine. Irish nationalists and the U.N. mediators in Palestine both hoped that mutual dependence, geographic proximity, and the benefits of shared infrastructure would gradually dissipate the aftermath of ethnic partition. Indeed, the U.N. plan for the partition of Palestine was based explicitly on the premise that economic union would compensate for the difficulties of the proposed territories. Instead, partition's legacies thwarted economic union and kept both Ireland and what was left of Palestine in poverty.

If the lessons of these examples are noteworthy, it may be because Bosnia will constitute a turning point in partition theory. The fact that NATO is preparing for an extended presence indicates that the alliance recognizes the unlikely success of a divide and quit approach in this situation. Though divide and quit was a motive in Britain's support for partition in Ireland, Palestine, and India, it got Britain out quickly only in India, and that was because South Asia is distant from Britain. From the sequence of events in Bosnia, it is clear that European and American leaders, and the rest of the international community, were prepared to accept partition if it would curtail Western intervention in the conflict and limit Western involvement in the region. But as the partition process unfolds, it is being recognized that divide and quit might turn into divide and be forced to stay. Unlike Somalia or Rwanda, Bosnia is a high-profile intervention because the Balkans have played an important and generally unwelcome part in European security. So far, the West has not been able to walk away from this war, and each halfhearted intervention, however delusory, has led to more rather than less involvement. As the realization takes hold that a Bosnian partition may mean an indefinitely prolonged commitment to a chronically volatile region, investment in reintegration may be discovered to be an easier route to withdrawal.’

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Radha Kumar, a Warren Weaver Fellow at the Rockefeller Foundation, is the author of the forthcoming book Divide and Fall: Bosnia and the Annals of Partition.
  • More By Radha Kumar