Security Studies Program Seminar

  Power Sharing or Partition?
History's Lessons for Keeping the Peace in Bosnia

Alan J. Kuperman
Assistant Professor
LBJ School of Public Affairs
University of Texas, Austin

  April 12, 2006

Bosnia is not necessarily over, especially if we continue our current policy. Why is Bosnia even a question? Even though we are ten years past the Dayton accords, and there hasn't been civil war since 1995, three factors are important: 1) supreme political authority still in hands of unelected international official; 2) security still ensured by international peacekeepers, though the number have dropped; 3) Bosnia is still composed of two separate political entities, one a Serb republic and the other a Muslim-Croat federation, and they still disagree about whether Bosnia should remain two separate loosely linked entities, or whether they should be merged.

So Bosnia is still in an unsustainable limbo. Unsustainable because 1) very costly to the international community; 2) EU has this ongoing accession process with other Balkan states—but Bosnia can't join until we decide what Bosnia is; 3) we have impending final status negotiation for neighboring Kosovo, and if that final status is settled prior to Bosnia's, there could be uproar in Bosnia.

The fundamental question today is the same one since the early 1990s, involving why the war broke out, why it continued, how it was settled, why we have a continuing dispute, and why war might erupt again. The Muslims, who represented a near majority, wanted a unitary state. The Serbs, about 30% of population, have insisted that they want Serb political control over Serb territory, where Serbs are majority or large plurality.

Different attempts at a solution were made in the early 1990s, aiming to avert the outbreak of war. First was a proposal for Bosnia to remain in a rump Yugoslavia, after the secession of Croatia and Slovenia. Second was a proposal for ethnic cantonization, so that each ethnic group would have control over its own areas. Bosnia's Muslims initially agreed to each proposal, but then reneged, and instead joined with the republic's Croats to declare the independence of a unitary Bosnia, which was recognized by the international community. But this violated the Serbs' primary demand, for political control over their own areas, so they took up arms to achieve that goal by force. The Serbs make quick military progress, and the Muslims and Croats started fighting among themselves. In 1994, the US forged and facilitated the arming of a Croat-Muslim alliance, which reversed the Serb military gains in mid-1995. This enabled the signing in late 1995 of the Dayton peace accords, which was an ambiguous peace agreement. It was sold to the Serbs as an ethnic partition—Serbs have control over Serb areas. But the Muslims were told they were getting a unitary Bosnia, as they had always wanted. The accords did give strong to Bosnia's two entities, a “Serb Republic” and the Muslim-Croat federation. But the accords also required negotiations between the Serbs and the Muslim-Croat federation to explore the strengthening of central government institutions. This was the weak reed on which the Muslims could think there would be a unitary state. The international community has decided in recent years to push toward a unitary state—which the Serbs have always said they could never accept. The question is whether this is a good idea or not. Is siding with the Muslims a good thing or not?

There are two ways to answer this, and I will use both. One is to look at the specifics of this case. The other is to look more generally at the track record of power sharing after ethnic wars. There are two opposing theories in the literature on power sharing. One school says it never works, and the other says it works quite frequently, and ethnic wars are no harder to resolve than ideological civil wars. I wrote a Security Studies article on this. I examined Chaim Kauffman's claim that you can never put an ethnically fractured state back together again after a civil war—the Humpty Dumpty theory. During a war, ethnic identities harden, and people cannot let down their guard again. So the only solution is a de facto or de jure partition. Separate people into defensible mono-ethnic zones. By contrast, Mason and Fett say there are many instances of ethnic civil wars being resolved by power-sharing, and it is just as common as for ideological civil wars. So this hardening of identities had no impact.

How can these studies come to such different conclusions? The main answer is they use different universes of cases. So, I tried to look closer at every case since WW II of a civil war that was arguably resolved by power sharing, to determine if it really had been resolved by power sharing. This raises the question of what does it mean for an ethnic civil war to be resolved by power sharing? I argue there are 5 possible disqualifying conditions:

1. Regional autonomy. If a region was granted political and security autonomy within a defensible border, this is de facto partition.

2. Peace enforcement. There is an outside military force in the country. We cannot know for sure the role of power sharing in resolving the war if there are still peacekeepers. The test would come after their withdrawal.

3. Disguised victory. When one side is completely beating the other and the losing side is willing to compromise by “power sharing,” this is just a negotiated surrender if the weaker side is not really given any power.

4. Non-ethnic. There may be ethnic groups involved in a civil war, but the main schism is not along ethnic lines, but rather along political lines. If this were true, you would not expect the identity hardening that Kauffmann discusses.

5. Pause in fighting. If fighting resumes soon after an ostensible power sharing agreement, then it was just a ceasefire, not an actual resolution via power sharing. The standard test in the literature is whether peace lasts five years.

When you add all these conditions, there were six cases of true power sharing:

Lebanon, 1958
Sudan, 1972
Zimbabwe, 1979
Mozambique, 1992
South Africa, 1994
Guatemala, 1996

Kaufman's hyperbolic claims are not true. It's not impossible to resolve an ethnic civil war this way, but it's pretty darn tough. The percentage of ethnic civil wars resolved this way is lower than ideological civil wars.

For the five reasons above, we can't count many putative cases of power-sharing resolution of civil war, including in India, Spain, Nicaragua, Papua New Guinea, Lebanon, Georgia, Israel, Bosnia, Croatia, Mali, Philippines, Tajikistan, Bangladesh, and N. Ireland. In several of these cases, including N. Ireland and Lebanon, we have live experiments, as occupying troops are withdrawn, and we must wait to see if peace survives.

But shouldn't the six successes give some hope for Bosnia? This is where you need to do some process tracing and see what the cases really tell you.

Lebanon doesn't give us great hope—not very many casualties compared to Bosnia, so reconciliation was easier. In Sudan, war stopped for 11 years, but then war restarted, and it was even bloodier than the first. This power-sharing agreement fell apart, and another 2 million people died. In Zimbabwe , there was a decolonization war between whites and blacks. This was a true power-sharing agreement in 1979. In the intervening years, a new civil war broke out between the two main black tribes and parties. Blacks have taken more power and started marginalizing the whites, with raids on their farms. Whites are fleeing the country—again, hardly power sharing. In Mozambique, the war included ethnic groups, but the main divide was not ethnic but political. It was a war between a communist, Soviet-leaning group and a right-wing group sponsored by South Africa. Again, we would not expect hardening of ethnic identities as in Bosnia. In South Africa, there was a long, low-level conflict between ANC and the white apartheid regime. This has been settled peacefully, and there is power sharing. On the other hand, the death toll was never very high—about 8,000 over 2 decades. Also, Nelson Mandela was able to lead the blacks to suppress their demands and see what their long-term interest was. If there were a Bosnian Nelson Mandela, things might be different. In Guatemala , this was an ethnic conflict, with high levels of conflict. I think this is a true success. True resolution of ethnic civil war with power sharing. The government accepted one of the ethnic leaders to be in charge of implementing the agreement. So I would argue that there is one case of real successful power-sharing after an ethnic civil war in the post-WW II period.

So has the international community looked at this weak track record of power sharing in regards to Bosnia? No. The community has insisted that the Dayton accords be revised for true power sharing in Bosnia. They have done this through incentives by the EU, US, and NATO, and through unilateral moves by the appointed international authority in Bosnia. Muslims have welcomed this revision of the Dayton accords, but the Serbs are resisting it. The international community has had to impose centralization. The high representative also took away veto power of ethnic groups and took other legal steps against Serb autonomy. High representative also dismissed Serb political and police officials.

Isn't this a good thing? Isn't the high representative just getting rid of spoilers? Unfortunately, in some cases, this attempt to insist on moderation has backfired, with hardliners coming in to replace those who resign in protest. But the carrot of possible EU accession has coerced the Serbs grudgingly to make concessions toward centralization, including of military and police forces.

But I don't think the problem is over. Because all the Serbs have done is sign a piece of paper. Who knows if they will implement it. A lot of the police agreement is ambiguous. The tough part is a provision that says police jurisdiction will not follow entity lines. This means that Muslims and Croats with guns with be authorized to go into Serb areas and arrest Serbs. Some Serbs are saying this will mean war, which is chillingly reminiscent of what they said before the outbreak of war last time. They might be bluffing, but there are significant quantities of light and heavy weapons all over the country not under international monitoring, according to the International Crisis Group. There are the means and motivation to restart this conflict. The only additional thing we need is Croats and Muslims with guns starting to arrest Serbs in the Serb Republic.

I don't mean to be alarmist. I don't think war is inevitable. And I don't ignore the down sides of partition, including the problem of contagion. If the Bosnian Serbs get their own state, the Albanians in Macedonia, the Albanians in Kosovo, and other ethnic concentrations in the former Yugoslavia will demand the same right. There is danger of a domino effect here, which is why I don't support a hard partition of Bosnia. But there are clearly some major risks in the status quo. Nick Burns has been pushing to give the Muslims the unitary state they have been demanding, and take away what the Serbs want.

Five Options are on the table:

1. Coerced unification. This is what we're currently doing, and I've discussed the downsides. Backfires by prompting Serb nationalism and possibly renewal of war.

2. Partition de jure, Kauffmann solution, which I think has dangerous contagion effects.

3. Adjust borders. Hold an international conference for all of the Balkans. It has been proposed every few years for hundreds of years in this region, yet it has never prevented war. This is impossible.

4. Freeze the status quo (limbo). This is not terrible, but I don't think it is sustainable, especially with Kosovo final status coming up, which is going to raise questions in Bosnia.

5. Regional autonomy. It is close to de facto partition, without the prize of independence. It is a compromise solution. It is closer to the partition side, but avoids rewarding militants with independence. Indeed, this is the option that could have averted war in the first place – giving the Serbs the one thing they have wanted, strong autonomy – and also was the promise in the Dayton accords that managed to end the war. I think this could also serve as model for Kosovo. Serbia doesn't want Kosovo to get independence, and Russia could prevent this with a veto in the UN Security Council. Whereas if we have strong regional autonomy that solves the Bosnia problem, we can insist on the same deal to the Kosovo Albanians.

If this is the sweet spot, who could be against it? First, the US, on grounds that it rewards ethnic cleansing because the contiguous Serb territory was created through war. If we restore autonomy to it, the Serbs will have gained for their actions. But I don't think this is such a grave concern. The Serbs did not gain any net territory through war, and they would not be rewarded with independence. Second, the EU opposes my plan, because EU accession requires certain reforms at the state level based on central institutions. Strong regional autonomy would interfere with this. They say they need a single state for reasons of uniformity. But if the EU had to deal with two entities in Bosnia, it surely could. In my view, the concerns of the US and EU pale in comparison to the goal of consolidating a still fragile peace in Bosnia .

[For further detail, see the forthcoming chapter: Alan J. Kuperman, “Power-Sharing or Partition? History's Lessons for Keeping the Peace in Bosnia,” in Michael Innes, ed., Bosnian Security After Dayton: New Perspectives ( New York: Routledge, 2006).]

Rapporteur: Caitlin Talmadge

back to Wednesday Seminar Series, Spring 2006