Exiting From War:
Cooperating Under Even the Most Difficult Circumstances

Once civil wars begin, they become almost impossible to end short of a decisive military victory. Nonetheless, almost half of all civil wars since 1940 experienced serious negotiations designed to find early solutions to the war, many of which drafter mutually-acceptable political settlements. The majority of these negotiations, however, failed and the combatants eventually returned to war. Despite what appeared to be strong incentives to stop fighting and despite what often appeared to be strong yearnings for peace, groups rarely agreed to settle for a compromise settlement that would allow them to either share political power or peacefully part company.

Until recently, most policymakers and scholars assumed that domestic adversaries almost always fought to the death because they had no desire to negotiate or because the underlying issues of the conflict were irreconcilable. But a detailed examination of the 53 civil wars that were fought between 1940 and 1992 reveals a surprisingly high rate of negotiation. ₁ The Chinese Communists and the Chinese Nationalists met three times between 1938 and 1949 to discuss political and military arrangements for a shared government during their war. Greek's Communists and Nationalists held formal talks in 1945 and eventually signed the Varkiza Agreement. The three rival factions in Laos met eight times between 1961 and 1972 and ultimately signed an agreement on February 20, 1973. The government of Uganda and the NRA guerrillas signed a peace accord and a powersharing agreement in December 1985. The Nigerian government and leaders of the Biafran movement held lengthy talks in May of 1968 and again between August and September of 1968. Four separate conferences of "National Reconciliation" were held between the national government in Chad and the guerrilla force in 1979 to try to end that war. The Angolan government and the UNITA rebels held a series of peace negotiations between 1991 and 1995. And this pattern continues today. The Serbs, Croats and Muslims met numerous times before they finally reached and signed the Dayton Accords. And extensive negotiations were held in Cambodia, Somalia and most recently Rwanda. Contrary to the image most people hold of the intransigence of civil war adversaries, almost half of them are publicly willing to consider compromise solutions and many draft formal peace agreements. Despite these promising attempts at peace, however, the outcome is almost always the same. The adversaries eventually walk away from the negotiating table and return to war, stopping only when one side eventually triumphs in battle.

This chapter explores this disturbing phenomenon by asking two questions. First, why do civil war adversaries consistently walk away from possible compromise settlements despite what-are often very sincere attempts to negotiate? And second, is there anything the international community can do to break this cycle of violence and help reconstruct broken states more peacefully?

Briefly, this chapter argues that domestic groups walk away from settlements because they cannot credibly commit to treaties that produce enormous uncertainty in the context of potentially dangerous disarmament. If adversaries truly wish to resolve their wars they must do two things: they must stop the fighting and they must fashion a political contract that reestablishes legitimate rule. This means, however, that groups must dismantle their separate militaries and integrate their individual assets at exactly the same time they submit to untested rules of a new political game. But once groups dispose of their partisan military forces they become unable to enforce their own agreement, protect themselves against surprise attack or credibly commit to honoring any subsequent terms. Thus, even under the very best conditions - conditions where all groups within a civil society actually want to peacefully rebuild their state - the desire for peace will clash with the realities of implementation, and groups will ultimately choose the safer, more certain option of

The fact that rebuilding a state is difficult, however, does not mean it is impossible. Many settlements would be acceptable and would succeed in ending a war if there was some certainty that the participants would be protected during the vulnerable implementation period and there was some certainty that these new political and military contracts would last even after the first ruling party was safely installed in government. Since domestic factions cannot retain their forces, outside intervention can serve this vital function by guaranteeing that groups will be protected and ensuring that treaties include important long-term strategies for self-enforcement. Once groups are reassured that they cannot be exploited by such a treaty, their promises gain credibility and they are far more likely to submit to the terms.

In what follows, I explore why so many "attempted negotiations" since 1940 have failed. These cases are especially intriguing because their high rate of failure indicates that even if all the other roadblocks to cooperation are removed and groups agree to negotiate, even these efforts will most likely fail. Bringing adversaries to the table, therefore, is not enough. If the international community truly wishes to resolve civil wars they must first identify what ultimately blocks success. Only if we can identify what is required to reach and implement treaties can we truly find a solution to these wars.

These cases are also important to study because they represent those cases where outside intervention has the greatest chance of success and can therefore do the greatest amount of good. If an important part of the problem is the inability of groups to credibly commit to an agreement that becomes much less attractive once implemented, then reassurance is precisely the type of help outside intervention can provide.

And finally, if it's true that promising negotiations collapse because groups can't commit to the high degree of vulnerability required by a settlement (and not because their conflict is hopelessly irreconcilable) then this fact has stunning implications for how we view the problem of civil war resolution. Suddenly, settlement is not as impossible or as difficult as some have argued. If this is really true, then outsiders should be able to identify in which instances a settlement would be acceptable and target their assistance accordingly. ₂

The following chapter begins by presenting the central puzzle: given the choice between negotiating a compromise settlement or fighting for hegemony, domestic groups almost always choose to fight. Section two presents four explanations for this phenomenon that focus on the problems of historical hatreds, conflicts of interest, greedy elites and security dilemmas as possible reasons for rejecting settlement. It then argues that although the first three explanations do identify important reasons why domestic adversaries might refuse to negotiate, they cannot explain the subset of cases where serious negotiations are attempted but still failed. Current security dilemma arguments, on the other hand, offer important insights into why negotiations might fail, but do not identify the crucial mechanism blocking implementation. They, therefore, also cannot account for variation in success and failure. Section three offers an alternative theory of civil war resolution based on the problem of fear and insecurity. It argues that domestic adversaries find it almost impossible to end their wars peacefully because they face an almost insurmountable dilemma: any place settlement will require them to relinquish independent assets and defenses which in turn makes the contract unenforceable. Section four then discusses the conditions necessary for successful cooperation and the important role outside intervention can play in the resolution of these wars.

1. The Puzzle

Once civil wars begin, fighting factions are rarely able to end their war without one side first destroying the other. Although discouraging, this pattern of resolution obscures a more hopeful reality: almost half of all the civil wars between 1940 and 1992 initiated formal negotiations to try to end the conflict. ₃ Unfortunately, only half of these negotiations (20% of all civil wars) actually ended in a successful peace settlement ₄ Instead, civil war adversaries consistently abandoned promising talks in favor of decisive battlefield struggles. ₅ Unlike the majority of interstate wars which end in some form of negotiated settlement, the vast majority of civil wars end with the extermination, expulsion or capitulation of the losing side. Despite what are often promising negotiations and despite what often appears like a sincere desire for settlement, groups consistently spiral back to war.

Until now, this high rate of failure was not viewed as particularly puzzling. Policymakers and academics generally assumed that groups fighting civil wars hated each other, had opposing aims, and had no desire to work together. And even if groups would consider sharing a state, the stakes in civil wars were often difficult to divide, and groups had strong reasons to fear settlement. Besides, civil wars usually indicate a failure to negotiate in peacetime; once violence begins, the brutality of war only makes collaboration less likely. In short, there are many reasons fighting factions would refuse a settlement. Heated emotions, predatory practices, irreconcilable differences, and fear of exploitation all play a role in war and make reconciliation difficult.

Domestic adversaries, however, also have strong incentives to negotiate a settlement without first fighting what is usually a devastating domestic war. Civil wars tend to be long, averaging 35 months. ₆ They tend to be bloody, averaging approximately ? in battle deaths. And they tend to create enormous destruction to the homeland. Groups contemplating civil war must consider the possibility of relentless bombing and shelling of their neighborhoods, a devastated landscape, roads and farmlands fled with land mines, the cutoff of basic necessities like running water and electricity, rampant cholera and malnutrition, not to mention a potentially high number of military and civilian deaths. These are not wars that can be fought on foreign soil leaving houses and families protected at home. These are wars that can consume all that is valuable in a state.(Examples: Angola, Liberia, the American South?) In many cases it is unclear why combatants would continue to fight even after their country has been decimated by war, even after any prize has been exhausted.

The fact that almost half of all civil war opponents meet to negotiate peace terms indicates that they too are aware of these incentives to compromise. Clearly, groups are getting together to try to rebuild their states peacefully. But they are also consistently walking away from any possible settlements. What is it about these settlements that, in the end, becomes so unappealing?

2. Current Explanations

To date, no study or general theory has addressed the specific question of why domestic adversaries consistently walk away from negotiations, especially those negotiations where the underlying political issues have been resolved. And no theory has attempted to link the more narrow problem of civil war resolution with the larger problem of cooperative state-building under conditions of anarchy. At least five of the current theories on civil wars, however, can be applied to this question of intransigence: historical hatred arguments that assert that groups fight because they hate each other and have no desire to cooperate; conflict of interest arguments that claim that groups fighting for control of a single state inevitably encounter conflicts of interest that are difficult to reconcile; greedy elite arguments that point out that stubborn, threatened, or self-interested leaders often have little to lose by continuing to fight; and security dilemma arguments  that highlight how fear and uncertainty during war can ultimately sabotage cooperation efforts and perpetuate violence.

The Historical Hatreds Approach

The historical or ancient hatreds approach so often discussed in newspaper accounts of Bosnia, Somalia and Rwanda views conflict as the natural outcome of different ethnic, religious, or tribal divisions within a state that are suddenly reactivated. ₇ According to this view, competing identity groups often have no desire to share a common territory or a common government and they certainly have no desire to live under an opponent's rule. The recent violence in Yugoslavia, Georgia, Tajikistan, Nagorno-Karabahk, and Chechnya is often cited as an excellent example of unleashed hatreds at work. Once Soviet control of internal affairs ceased in 1991, diverse groups in the "states" were suddenly free to pursue their own agendas, seek revenge for past grievances, or simply rid themselves of despised enemies. In short, Serbs and Croats, Hutus and Tutsis, Blacks and Whites carry with them histories of past battles, memories of past injustices, and dreams of independent states. They do not want to cooperate. Only if these contentious cleavages can be suppressed through rigid domestic laws, outside restraint or physical separation will peace be possible. ₈

Proponents of the historical hatreds approach, however, argue that two types of negotiated settlements are possible under the right conditions. Adversaries will settle their war off the battlefield if a balance of power can be created so that no one group can win an easy victory. This balance would not only make fighting a less attractive option, but it would also enable groups to deter future aggression and thus make agreement more inviting. The only other type of settlement that could overcome these enmities would be one based on the partition of the territory into separate homogeneous regions. As Fred Ilcle argues,

If an effective balance-of-power cannot be created among the competing groups because one side is clearly stronger than the others and outside states are not willing to fortify the weaker side(s), then only the physical separation of the enemies will suffice to ensure peace. Other types of settlement, such as the highly unbalanced federation created between the Serbs, Muslims and Croatians in Bosnia, will only be possible when outside powers step in to force such a solution and these solutions are unlikely to last.

Critics of the historical hatreds view of internal conflict, however, argue that the emphasis on fixed identities and animosities cannot explain why many ethnic groups do not go to war when restraints are removed ₁₀ . Why did the Czechs and the Slovaks, the Lithuanians and the Russians, the Romanians and the Hungarians not fight each other after 1991? If cleavages really are immutable, how can one account for those cases where different ethnic groups reached mutually-acceptable settlements without a balance of power and without resorting to partition. How do you account for the peaceful transfer of power between white South Africans and black South Africans? And how would you explain the French Canadian and Anglo-Canadian decision to peacefully resolve their differences?

Historical hatreds arguments also cannot justify why even the most hate--filled groups--groups that have fought viscous, long-running wars - can agree to compromise. How do you explain the compromise settlements between the Blacks and Whites in Zimbabwe, the Arabs and Africans in Sudan, and the Christian and Muslims in Lebanon? Finally, the emphasis on fixed identities and animosities also cannot account for the variation in the level of conflict and compromise across cases. Large empirical studies by Roy Licklider, David Mason and myself have all found that identity wars or ethnic wars were no more likely to resist settlement than those fought between the same ethnic or religious group. ₁₁ Identity does not appear to be a factor in the decision to settle or continue fighting.

The Conflicts of Interest Approach

The conflict of interest approach, on the other hand, tends to view civil wars as more rational events designed to obtain state control, force political and social reform or advance one's status in society. This approach argues that civil wars tend to be decisive, irreconcilable battles for one of four reasons. First, groups fighting civil wars often have contradictory goals that leave little or no room for a compromise. If one or both parties want nothing less than complete control of the state, then mutually-acceptable solutions will be difficult to reach. ₁₂ Moreover, even these groups did not have contradictory aims when a war first broke out (if, for example, the rebels simply wanted reform and not complete control) the act of fighting itself tends to generate totalistic war aims. ₁₃

Second, groups might also fight to the finish because the government and all its trappings are difficult if not impossible to divide. ₁₄ Only one government and one military force can exist at any given time and it is this fact that encourages decisive war. According to Paul Pillar,

This problem of divisibility emerges when groups must decide who will control key positions in many new government such as the presidency or the Ministry of Justice. Every group would like to see its leader become the new executive, its people appointed to important ministries, and they resent suggestions that another group take the lead. If both demand control of the same position as the black Patriotic Front and the white Rhodesians did during the Lancaster House talks, then mutually-acceptable solutions will be difficult to find.

Third, groups fighting civil wars are often encouraged to continue fighting due to exaggerated battlefield expectations. Groups fighting wars often hold overly optimistic views of their ability to win a war. If both groups simultaneously hold such unrealistic expectations, neither will see any reason to compromise and both will continue to fight in the mistaken belief that victory will ultimately be theirs. ₁₆

But realistic expectations might also stand in the way of settlement. Opponents usually prefer to initiate negotiations after they have made significant battlefield gains and prefer to put off negotiations when they have suffered a string of losses. This however, means that groups will rarely, if ever, wish to negotiate at the same time. I. William Zartman points to this specific problem when he writes that "...the asymmetries in power and commitment become a trap for their respective parties, preventing the negotiations that the parties seek to end the conflict." ₁₇ Only when groups have reached a costly military stalemate and their relative military positions become equal will they both have equal incentives to settle.

Finally, domestic groups might also reject compromise treaties because the benefits of obtaining full control of the state are perceived to be so high. A decisive win would not only give the victor full control of the state, it would also permanently eliminate a rival for limited resources. Decisive victory, therefore, offers both immediate and long-term returns for which adversaries are willing to incur great costs. As Jane Holl has argued, "the intensity or duration of combat may beexcessive because belligerents value the shape of the settlement more than the costs of disagreement." ₁₈ In short, each of these explanations agrees that groups often fight to the death because unavoidable and often irreconcilable conflicts of interest - in the form of greed, goals, expectations, and stakes - inevitably stand in the way of settlement.

Although proponents of the conflict of interest approach to civil war resolution offer pessimistic predictions about the chances for compromise, they do believe that outside intervention can help groups reach these elusive settlements. If groups can be convinced that fighting offers fewer benefits than settlement and if the stakes are fairly easy to divide than outside states can help groups locate fair settlements. Outside intervention can make war less attractive while at the same time enhancing the benefits of cooperation by withdrawing outside aid and support, refusing access to and passage through their territory, and placing economic sanctions on those parties that refuse to talk. And once adversaries agree to negotiate, outside intervention can then help locate a middle ground, smooth over tensions and navigate what are usually difficult talks. (Discuss mediation literature in more detail.)

"Conflict of interest" explanations capture some of the key problems groups face when deciding whether or not to compromise. Oftentimes groups have no interest in cooperating and sometimes there is no overlapping middle ground between the goals and principles of two competing groups. But in most cases a bargaining range does seem to exist within which even the most ideologically antagonistic groups might find a settlement. Chiang Kai-shek and Mao Tse-tung signed a peace treaty in February 1946 which included detailed plans for military integration and political power-sharing. The Laotian Government and the Pathet Lao guerrillas signed two agreements to end their war. And even the Greek Government and the Greek Communists were able to agree to sign the Varkiza Agreement on February 12, 1945. It guaranteed the free and open existence of the Communist party, the creation of a national army, the holding of a plebiscite to decide on the specifics of a new regime, and the organization of elections for a constituent assembly to draft a new constitution. Even these groups were not beyond political reconciliation.

"Conflict of interest" explanations cannot seem to account for the high rate of political reconciliation or the fact that negotiations often drag on long after the underlying issues of the war were resolved. Political issues were often not the most difficult to resolve; whatever groups discussed thereafter appeared to be the crucial, vexing problems. In short, these explanations seem to assume that cooperation failed because solutions could not be found or because the parties were not sincere in their concessions.

Clearly, not all settlements are signed with honorable intentions. Many are signed for tactical reasons - whether to placate an outside patron, buy time to re-supply or simply to counteract a reputation for aggression. But not all failed settlements appear to be "insincere" or purely tactical. Lengthy public negotiations are not costless to the leader's involved. Leaders(especially guerrilla leaders) often put their lives in danger when they emerge to publicly participate in negotiations. "You come out only when you have a message," Jonas Savimbi revealed to a reporter when asked if he was serious about Angola's most recent round of peace talks "When you don't, you don't come out." ₁₉ A readiness to accept supervision, public announcements of important concessions, a desire to discuss the details of a transfer of power and long-term participation in lengthy negotiations all generate costs to the rivals and indicate more than a tactical interest in appearing cooperative. It remains unclear, therefore, why groups and their leaders would incur the costs of bargaining only to ignore the resulting settlements later on.

"Conflict of interest" arguments also tend to underestimate the high costs of internal war and overestimate the final benefits of victory. But civil wars are extremely costly and can leave a state so despoiled that it is no longer worth controlling. (Example: what is left in Angola? What is left in Liberia) What this approach cannot explain is why complete control of a decimated country with exhausted resources can be a preferable solution to preserving the country through cooperation. ₂₀

Finally, the argument that compromise solutions do not exist because the government is difficult to divide is simply untrue. Territory can be divided through secession, and political powercan be shared through a federal or confederal structure. East and West Pakistan, the Czech and Slovak Republics, and Ethiopia and Eritrea were all divided through partition or secession. And power-sharing arrangements were negotiated in Colombia, Zimbabwe, Lebanon, Sudan and Yemen during their civil wars. In short, although conflict of interest explanations offer compelling reasons why many domestic groups might refuse to initiate negotiations, they do not offer equally plausible reasons why groups would still reject compromise even after they've reached agreement and even when there is little to be gained materially by fighting on.

The Greedy Elites Approach

A third approach to the problems of civil war resolution focuses on the individual leaders and their power to unilaterally evade settlement. These "elite" explanations argue that leaders of rival factions tend to benefit disproportionately from gaining complete control of a state while also remaining insulated from the costs of defeat. Leaders often have exit options - they can escape to villas in neighboring countries - and they often have personal armies who can do the fighting for them. (Ian Smith had dual citizenship and money. Mobutu of Zaire had Mediterranean villas. Latin American dictators had Miami.) As a result, they often hold different preferences for settlement than the large mass of citizens who are more directly affected by war. Angolans now speak with disgust when describing how both President Jose Eduardo dos Santos and Jonas Savimbi profited greatly from years of black market dealings in oil, diamonds and arms. And they blame the government for perpetuating the war because "generals and politicians profit from their skim of the arms trade." ₂₁ As long as war remains lucrative for the elites involved, there is little incentive for them to invite or accept settlement.

Domestic conflicts can also turn immutable because leaders themselves have much more to gain from total victory than from a shared state, and these benefits convince them to fight on. Milosevic is often held up as an example of a leader driven by his own ambition to become leader of a "greater Serbia" rather than accept the more tenuous position of a candidate for leadership of the Federal Republic of Yugoslavia. ₂₂

Finally, it has also been argued that civil wars can become decisive wars to the death because a leader's greed, paranoia, or stubborn attachment to principle convinces him or her to continue fighting. In these cases, as Stephen Stedman has argued, "a leader' s drive for power can make compromise solutions impossible." ₂₃ . The Ibo rebel leader Ojukwu ignored pleas from many of his supporters to accept what were very generous terms from the Nigerian government and end that devastating secessionist war, yet Ojukwu refused even to meet with his enemy. Similarly, Jefferson Davis never wavered in his rejection of all peace feelers from President Lincoln during the American civil war. In both cases, fairly generous terms were offered by their opponents, terms that appeared less costly to Biafra and the American South than continued resistance.

Can outside intervention in any way help facilitate settlement in the face of greedy elites or "spoilers" as Stephen Stedman calls them? In situations where incumbents have few incentives to compromise (they had enjoyed authoritarian control and now must open up the political system)and few deterrents against using war to protect themselves and their positions (they have the capabilities), compromise settlements will have a greater chance to succeed if these leaders are replaced, if they can be marginalized within their party, or if their war-making supplies can be cut off. ₂₄ In situations where rebel leaders stubbornly refuse to give in on key issues like Ojukwu on secession), these leaders will also have to be replaced. (Expand.)

Critics of "greedy leader" explanations have argued that outward actions and statements by elites are not always perfect indicators of their preferences and intentions, and other motives besides greed might be driving behavior. In fact, even peace-loving leaders often have incentives to appear uncompromising during negotiations in order to ensure the best possible outcome for themselves and they often take what appear to be inflexible positions only to settle for compromises later on. ₂₅ In many cases, taking a hard stance early in negotiations maximizes one's leverage and promises a better settlement.

Explanations that blame the leader for intransigence or greed also underestimate his or her ability to change and adapt as wars evolve. Even the most extreme, inflexible leaders have been known to change their views. So, while Jonas Savimbi might have stated that he "would not accept defeat" in the 1991 election and it appeared by all objective accounts that he was not serious about sharing the state, this does not mean that by 1993, the continuation of the war had not convinced him otherwise. Painting a leader as unambiguously and unchangingly greedy or stubborn might not always adequately reflect reality and it is important to remember that most negotiating successes are preceded by numerous negotiating failures.

"Greedy leader" explanations also have difficulty explaining why the rank and file would continue to fight even when it becomes clear that victory is impossible and even after they had suffered tremendous losses. Ambitious leaders whose lives are not directly on the front line might have good reasons to continue fighting, but what explains persistent support from the masses? Finally, arguments that quickly blame the leader for intransigence also often overlook the very real fears some leaders face from a settlement. The fact that a leader will reject a treaty or violate it once its signed should not be taken as an unequivocal sign that this leader is incorrigible and militant. Often he or she might simply be trying to avoid retribution.

In short, the "historical hatreds", "conflict of interest" and "greedy leader" approaches appear to leave large parts of the puzzle unanswered. They might explain that subset of cases where groups really do not want to compromise with each other, but they cannot explain the set of cases where groups wish to avoid or resolve a war but cannot seem to reach and implement this goal.

The Security Dilemma

Recently, a number of scholars have argued that security fears might play a far greater role in the decision to fight unconditionally than policymakers or social scientists have thus far recognized. These arguments allege that ancient hatreds, irreconcilable differences and stubborn leaders might play a role in the persistence of internal violence, but they are all exacerbated by fears that take over when a state and all of its rules, restraints and regularized activity collapses. Ultimately, it is this uncertainty and fear that creates destabilizing security dilemmas that make it difficult to accept reconciliation.

The fourth possible explanation for why internal conflicts tend to be so intractable, therefore, argues that domestic adversaries might reject settlements because of the fear and insecurity they produce among competing populations. ₂₆ When settlements are signed, domestic groups face a world of uncertainty and unanswered questions. Can I that my enemy sincerely wants to compromise? Will they attempt to annihilate me when my defenses are down? How willpower and resources be redistributed in any new government? Who will win the election? Will the winner respect the position and the rights of others once safely installed in power or will they use their new position to repress all opposition? Will the moderates remain moderate? How can you be sure that leaders can deliver the rest of their party? Domestic groups can never be sure how relations with each other will evolve. The fact that leaders will change, rules will be rewritten and party discipline will vary only adds to the uncertainty and insecurity of the time. ₂₇ All of these uncertainties can leave citizens too anxious and distrustful to cooperate and thus leave them no choice but continue the war.

Security dilemma arguments that concentrate on the problems of civil war resolution come in two forms: those that argue that groups fight because they fear the aftereffects of any fully implemented settlement, and those that argue that groups fight because they fear the actual transition period that precedes full implementation.

Barry Posen has argued that security dilemmas can be triggered when competing groups find themselves geographically intermingled in ways that make it difficult to defend themselves. ₂₈

If a peace settlement cannot promise that former adversaries will enjoy defensible positions in a new state, they will reject this unstable outcome and choose instead to fight. Adversaries must be able to design settlements that offer them at least some degree of future security if they can hope to avoid destabilizing and dangerous security dilemmas in the future. Civil war negotiations, therefore, so often fail to produce successful settlements because populations within states tend to be interspersed and unorganized and therefore cannot adequately defend themselves. ₂₉ The lengthy debate in Bosnia over the exact boundaries of the ethnically-based Muslim, Serb and Croat "states" is an excellent recent example of this concern over geography and its role in future security.

But geography is only one of many possible sources of insecurity. Security dilemmas can also be triggered by concern that changes in the balance of power might eventually leave groups vulnerable to attack. When settlements are negotiated, adversaries must not only design a cease-fire, they must also determine how the material remnants of the former state will be reallocated. ₃₀ But reallocation rarely benefits all groups equally and power inevitably shifts from some groups to others. Those groups who will most likely lose in this redistribution, can become anxious of the rising power of a neighbor. James Fearon has argued that even the fairest democratic system can still pose a threat to groups within society if that system creates ruling majority parties that can then shut out smaller minority parties. Such a settlement could appear quite threatening to minority groups who now face a lifetime of entrenched opposition. This fear appears to have convinced Chiang Kai-shek to repeatedly reject offers from the Chinese Communists for a coalition government. The Generalissimo argued to U.S. mediator Patrick Hurley that the proposal of a coalition government would be "an acknowledgment of total defeat of his party by the communists, and that the proposed agreement would eventually lead to Communist control of the government." ₃₁ In short, domestic groups clearly worry about their long-term political position in a new government and will continue to fight to avoid the possibility of becoming permanently disadvantaged in any new government. ₃₂

Fears can also be constructed by unscrupulous elites who play on people's natural fear of the new and unknown to further their own ambitions. ₃₃ As mentioned above, leaders often have greater incentives to resist negotiation and prolong wars than their followers who bear a disproportionate share of the costs. Leaders who wish to continue fighting in order to pursue their dream of victory, therefore, must somehow convince their followers that early settlements would be more detrimental to them than war. These "war entrepreneurs" can use fear to convince supporters that any peace offering is simply a dirty trick designed to bring their defenses down.(Ojukwu is an excellent example - created a pr campaign designed to convince supporters that they would be exterminated by the government if they settled.) In short, by convincing people that settlements would mean exploitation, these leaders can ensure that any appeals for more open multi-party systems go unheeded.

Artificially-generated fears can explain why citizens blindly support intransigence even after it becomes clear that they cannot win a war and even after they have suffered tremendous costs. Ambitious leaders whose lives are not directly on the front line might have good reasons to continue fighting. But a cleverly constructed "security dilemma" might be necessary to obtain persistent support from the masses. Settlement, in these cases, will require the destruction of these damaging myths through the non-partisan radio broadcasts, mediation and conferences. (Discuss importance of accurate information in overcoming the security dilemma.) ₃₄ But wars can also break out because leaders (especially incumbent leaders) themselves face a security dilemma when deciding whether to open up the political system or fight for absolute control. Incumbent leader's refusal to cooperate is often interpreted as greed, arrogance or even irrationality. Their desire to maintain control at all costs, however, can emerge from their own fear of the aftereffects of settlement. Incumbent leaders who agree to incorporate rebels into a new multi-party state could be viewed as traitors by the more extreme elements within a faction and by those who shed blood for the original goals. These followers are unlikely to support the leader in any new multi-party administration and could perceive any compromise as a betrayal that deserves death. ₃₅ (Example: Anwar Sadat) The problem leaders face for "selling out" to the opposition was recently illustrated in the continuing conflict in Northern Ireland. Two representatives of the Northern Ireland Women's Coalition - Monica McWilliams, a Catholic, and Pearl Seger, a Protestant - "were called "scum" and "traitors" by their unionist colleagues simply because their stated aim was inclusiveness." ₃₆

A fear of "revenge" might also convince leaders that war is still a better option than inviting their former enemy into power. A leader who opens the government to opposition groups or voluntarily steps down from power also invites retribution from those parties new to power. This fear of revenge appears to have played a role in the Hutu decision to massacre the Tutsis following the acceptance of the Arusha Accords. The Arusha Accords promised the previously unrepresented Tutsis a ?% representation in a new multi-party government; a solution that would have completely shut the militant arm of the Hutus (the akazu) out of power. It has been argued that the akazu, fearing Tutsi reprisals, chose genocide as the best method to avoid this outcome. Security dilemmas, therefore, can also be created in situations where leaders fear retaliation and do not believe they will have the power to protect themselves.

In short, settlements to civil wars can have the unanticipated and unnerving effect of making groups less secure than they would be if they continued the war, either because a settlement will leave groups physically intermingled with their former enemies, a settlement would institutionalize an unsettling balance of power or it would leave leaders vulnerable to attack. In each of these scenarios, it is this fear of the negative consequences of settlement that convinces groups to choose the more secure but ultimately more violent path of war.

[If settlements can place groups in such perilous situations, what would it take to convince groups to actually go through with an agreement? Scholars in the security dilemma camp are not wholly pessimistic about the possibility of compromise settlements to civil wars and they believe that they can succeed if the conditions are right. Barry Posen, Chaim Kauffmann and John Mearsheimer suggest creating a reassuring military balance of power to make settlement more appealing and war less attractive. If this isn't possible, they recommend resettling groups into their own defensible regions. If conflict seems to be driven by deeply-instilled myths spread by ambitious elites, then groups will only agree to settle when these myths have been debunked. If a leader consistently refuses to consider compromise, then intransigence will continue until this leader can be reassured or protected. Replacing him won't be enough in some cases. It might avoid retribution by the incoming opposition, but it will most likely not prevent punishment from his own party since he would still be viewed as a traitor by other members of his party. And finally, if conflict is driven by minority fear that any compromise government will shift the political advantage decisively in favor of their opponent, then settlements will only succeed if these weaker groups are given adequate guarantees of future representation. Create institutional checks and balances and you will get settlement. In short, security dilemma arguments offer a number of specific recommendations to the policy maker looking for effective ways to help end civil wars.]

In what follows, I argue that problems of security offer important insights into why some civil wars resist settlement but that current security dilemma arguments ultimately overlook a crucial barrier to success. All of the security dilemma arguments described above point to the fears groups have of the consequences of settlement. None of them point to a debilitating event that occurs before a new society is formed: the transition. ₃₇ Because current security dilemma explanations have concentrated on the final outcome of settlement and not the process by which combatants get there they have missed the overriding fear groups have regarding the actual implementation of the treaty. And because they have underestimated the enormous insecurity groups feel during this transitional period, they have also underestimated the extent to which domestic groups themselves can solve this problem.

When one looks at the specific cases of negotiation between 1940 and 1992 one finds that the successful civil war settlements had one thing in common. Almost all of them were underwritten by an outside state willing to provide security guarantees during the vulnerable implementation period. The successful settlements did not all have a balance of power, they did not rely on partition, they did not replace leaders or necessarily include extensive checks and balances. However, once adversaries agreed to negotiate, every case where a third-party stepped in to guarantee a treaty resulted in a successful settlement.

The fact that there does not appear to be a strong correlation between these variables and the ultimate success or settlements means that the initial puzzle remains unanswered. A balance of power does appear important in getting groups to the negotiating table; it does appear to convince them that settlement might be a better option than continued fighting. Incorporating elites in a new government and bolstering moderate elements also appears important in obtaining agreement on a treaty (mention Cambodia, Somalia, Rwanda). Multiple check and balances, minority vetoes and generous political terms also occupy much debate during negotiations. But none of these factors seems sufficient to ultimately bring about the successful implementation of settlements.

3. The Dilemma of Domestic Cooperation

In what follows I argue that civil war opponents seem to walk away from so many promising negotiations because the prerequisites of rebuilding a collapsed state make any bargain -no matter how sincerely intended - almost impossible to sustain without some form of outside enforcement. Adversaries retreat from promising negotiations and repudiate even generous settlements because the prospect of entering untested political arrangements in the context of their own disarmament becomes too frightening to accept. It is the actual terms of the treaty, therefore, and the resulting security dilemma they would create that convinces groups to aim for hegemony.

Domestic groups are likely to fight decisive wars and violate peace treaties more often than states in the international system because they are asked to do the impossible. If domestic groups wish to end their war they must rebuild a single, united government. But in order to rebuild a single, united government former enemies must not only give up their independent assets and armies but they must do this at exactly the same time as no established government exists to enforce compliance or protect those who do. The ultimate challenge facing civil war opponents at the negotiating table, therefore, is how to reestablish a functioning government under conditions of anarchy while simultaneously shedding all means to protect themselves.

The fact that any peace settlement is also a contract for long-term social, political, economic, and military collaboration means that ultimately the problem of civil war resolution is a problem of long-term cooperation. But the question then becomes: how do you rebuild collapsed states when rebuilding always takes place in an anarchic environment that strongly encourages self-reliance, not integration?

The fact that civil war settlement will always require groups to consolidate their assets and relinquish military forces if they decide to stay together has three deleterious effects on cooperation. First, groups will ultimately be unable to enforce any state building agreement. Second, groups will become far more vulnerable to attack and exploitation than independent states that are allowed to remain armed and independent. And third, groups will rarely be able to convince each other that they will abide by an agreement that offers enormous rewards for cheating and enormous costs for being cheated upon. As a result, even the most equinanimous agreement between the most willing partners cannot withstand what will always become an overwhelmingly dangerous implementation period.

Problems with Enforcement

Encouraging and then maintaining cooperation under anarchy is never easy. When no overarching enforcer exists to demand compliance, cheating often offers a cheap and easy means to obtain at the bargaining table what couldn't be obtained on the battlefield. States in the international system have devised a number of military and economic strategies to counteract these temptations and encourage the long-term maintenance of peace treaties. When states in the international system settle their wars, they can withdraw to separate territories and defensively reinforce their militaries. They can build early warning systems, create sophisticated monitoring and verification procedures and form buffer zones to check each other's behavior. They can forge external alliances and create reciprocal punishment strategies to make aggression more costly. They can also manipulate trade policies, use side-payments or withhold key goods to enhance payoffs for cooperation. This mixed use of carrots and sticks helps build binding agreements and encourage long-term cooperation.

Groups sharing a single state can do no such thing. If groups share common ownership over everything within their state--over the armed forces, over nuclear arsenals, over strategic oil reserves and important trading commodities--then none of these assets can be used as independent leverage to keep each other in line.

Problems of Vulnerability

The fact that domestic groups would have great difficulty enforcing agreements, however, should not necessarily doom cooperation, for even the best enforcement strategies cannot guarantee that agreements will be honored. This risk is tolerable as long as the participants who are contemplating cooperation can limit the damage caused by cheating. If groups can survive cheating so that they are no worse off than they would have been had they continued to fight, then the fact that they cannot deter or punish cheating should not discourage them from proceeding. A surprise attack might renew the war but if groups had time to established strong defenses and maintain active forces they would be no worse off than before. In fact, they might actually come out stronger since even short cease-fires would allow them to refurbish forces and restock supplies. Under these conditions, the risks of exploitation are manageable and well worth the potential benefits of long-term peace.

But domestic groups have no such luxury. If domestic groups wish to cooperate, they must waive their right to maintain separate armies, control personal armories, hold individual resources or form independent alliances and in so doing, relinquish their only remaining means for protection or enforcement. Thus, groups will be asked to give up the very defenses that would allow them to survive a sneak attack. Even states that surrender unconditionally are rarely required to go so far. ₃₈

Problems with Credibility

But the fact that domestic groups cannot independently enforce a peace agreement and cannot successfully sustain a surprise attack still should not unequivocally prevent settlement. Domestic groups should still be willing to sign and implement a peace treaty if they can make credible commitments to honor the terms. Such commitments should be possible if implementation can be structured in a way that limits vulnerability, institutions are set up that can effectively enforce the agreement, or costly signals of peaceful intentions can be sent. The tragedy of civil war resolution, however, is that none of these strategies are possible.

First, domestic groups will never be able to structure the agreement so that both sides can avoid vulnerability during the transition phase. Building a multi-party state will always require competing groups to eventually disarm and integrate their individual assets. Although any number of different implementation plans can be designed to reduce vulnerability - disarmament can be postponed, it can be implemented incrementally and in a reciprocal fashion--in the end it can never be avoided. And as long as groups realize that an avaricious opponent need only wait for this opportune time to attack, any promise to remain faithful to the original terms is simply not believable.

Domestic groups will also have difficulty credibly committing to an agreement whose future depends on new and untested institutions. Although there are numerous safeguards that can be drawn into political contracts, initially it will be difficult to guarantee that these new government structures and institutions will be effective. Even the American constitution, a contract that included numerous detailed checks and balances, was dependent on the integrity of the first leaders in power to set an important precedence for the peaceful transfer of power. In short, political institutions can be made reliably self-enforcing over time, but early in the process they hold no guarantee of exacting compliance.

Finally, groups will also have difficulty making credible commitments to any civil war settlement because effective signals tend to be too costly for even peace-loving groups to send. ₃₉ If domestic groups wish to convince their neighbor that they truly want peace they can credibly commit to a settlement by either making it impossible for themselves to launch a sneak attack, or by making themselves more vulnerable than their neighbor during the fateful transition period. But both of these signals would require some form of unilateral disarmament and it is unlikely that even the most peace-loving group would agree to anything that could result in annihilation. (Discuss Weingast's "Building Trust..." argument)

And even if groups were willing to accept this risk, disarmament is too easy to mimic by groups with aggressive intentions. Many international police monitors in Bosnia have reported that local police forces have secretly stockpiled huge amounts of weapons and ammunition in that country. "They can hide anything they want to from us," said an American police officer and mid-level supervisor for the United Nations. "We're out here filling in forms that say everything looks good, but most of these police are ready to go into combat in a quick minute." ₄₀ The fact that weapons are easy to hide, easy to buy and difficult to monitor was confirmed by police chief Ivo Katic when he announced to a reporter that: "They can take my weapons today, and I'll get more tomorrow. Then I'll hide them, it's no trouble." ₄₁ Thus, even the potentially powerful signal of disarmament might not have the desired effect of relaying peaceful intentions.

In short, domestic groups realize how vulnerable they will become if they agree to form anew state and they understand the limitations of enforcement. But ultimately, what makes it impossible for them to cooperate is there inability to credibly commit to any treaty that must, by definition, demand a defenseless form of cooperation.

4. Can Outside Intervention Help?

Since the requirements of rebuilding a state will always make cooperation risky and therefore either unattractive or inherently unstable, outside powers could play an important role in those situations where groups have begun to negotiate. Outside intervention can help domestic groups succeed in three important ways. First, and most importantly, they can provide protection to ensure that groups will survive the most vulnerable disarmament or demobilization phase that occurs during implementation. Second, outside powers can sensitize the parties to the importance of self-enforcement strategies and they can help design constitutions that include detailed promises of political participation. And third, they can allow groups to retain a measure of self-help by not demanding disarmament and by offering important "escape hatches" to groups such as dual citizenship, open borders or generous asylum provisions.

Provide Short-term Protection

Third parties can guarantee that groups will be protected, terms will be implemented and promises kept (or at least they can ensure that groups will survive until a new government and a new national military is formed). Outside states can step in to provide the level of security needed to reassure groups until a new non-partisan military can be formed. This is the most important function outside intervention can perform since groups ultimately must be able to weather this period before true state building can take place. If outside powers are not willing or able to guarantee the security of these groups, groups will most likely refuse to cooperate and the war will continue until one side or the other decisively wins.

Provide Strategies for Long-Term Self-Enforcement

Successfully integrating forces during the transition period is different from successfully maintaining the treaty over the long-term. Thus, while short-term vulnerability might overshadow all other concerns when groups first consider cooperation, it is the long-term enforcement that dominates their thoughts thereafter. Groups will only successfully maintain the treaty over time if their otherwise poor ability to enforce a contract can be enhanced or if they can somehow protect themselves from post-settlement opportunism. The type of institutions that are set up will make a difference in a party's decision to honor the agreement over time. ₄₂

Scholars have recently argued that groups will embrace democratic systems if they are allowed to compete in a fair and open manner. ₄₃ As long as parties are not completely shut out of "power", agreeing to such a system was assumed to be better than fighting a war. I argue that domestic adversaries require far greater reassurance before they will fully integrate themselves into a new state. Vulnerable minority groups with little independent ability to enforce a contract cannot be satisfied with the opportunity to compete and still feel certain that their rights and liberties will be protected. One can participate in the fairest, most open, most well monitored election and still be permanently shut out of power. Discussion over whether presidential or parliamentary system are better in collapsed states struggling to build democratic systems miss the point. Any one of these systems will still create far too much uncertainty to effectively prevent war. Groups fear not that they will lose an election, but that they will lose the ability to block terrible outcomes once an election is over.

Knowing that they are unlikely to have the personal power to enforce compliance, groups negotiating a peace treaty should demand very clear institutionalized rules. This could include a formal constitution with intricate checks and balances, a strong and independent judiciary able to punish violations, an apolitical, professional military, and free and open elections. Most importantly, however, it should include a guaranteed percentage of representation in all critical branches of government. (Examples: Colombia, Sudan, Lebanon, U.S. Constitution) The successful agreement in Colombia called for a 50-50 split of all government positions including the delegation of patronage jobs and an alternating presidency. Lebanon created a coalition cabinet with two Muslim and two Christian leaders in 1958. And white Zimbabwean were guaranteed 20% of the seats in the lower house of Parliament even though they comprised only 3% of the population. If groups cannot independently enforce their own political contracts, very detailed arrangements must be made to include all competing groups in key sectors of the government and military. In this way, each of the groups were assured an influential voice in future decisions.

[Other Suggestions: institutional rules that provide for extraordinary majorities or built-in protection like promises that certain posts won't be up for competition immediately, or that they can retain or obtain certain key posts (like head of security. See Nicaragua and Zimbabwe) The key is to strike a balance between greater access for the majority and guaranteed representation for the minority. Phase in liberal democracy over time while starting with more reassuring types of arrangements like pacts. (Colombia is a good example of both the advantages and disadvantages of a rigid pseudo-democratic system.)]

Since these arrangements are sometimes difficult to orchestrate and often highly contentious, groups will often agree to leave such precautions vague or undefined in order to reach agreement. They then hope that more detailed arrangements can be fashioned once the fighting has stopped. The Uganda Peace accord signed in 1985, the cease-fire agreement signed in Jordan in 1970, Greece's 1945 Varkiza Agreement and China's 1946 agreement included only the vaguest terms for a cease-fire and no transitional security provisions and each broke down within nine months. Angola's 1991 Estoril Accord called for winner-take-all elections. War resumed soon after UNITA was completely shut out of power. Outside states, sensitive to the fact that any ambiguity could breed renewed fear could insist that political arrangements be as specific and detailed as possible.

Provide a Means of Self-Help or an Exit Option

Groups would also benefit from the added reassurance of knowing that they will not be left completely helpless even after foreign troops leave. Many of the successful settlements since 1940 did more than simply incorporate short-term security guarantees and long-term institutions.(Examples) They also allowed groups to retain a measure of self-help and contained important escape clauses. [ Suggestions: a. let the population keep their arms, do not insist on complete disarmament. (Mention American Constitution and the logic behind the right to bear arms.) b. Try to even out imbalances in creative ways. Offer dual citizenship to minority groups. This has proven to be an important escape hatch for some groups. (Examples: Whites in Zimbabwe, Jews in regard to Israel) d. Maintain open borders to enable besieged groups to escape and obtain refuge. Allow former leaders to retain one crucial appointment: i.e.: Daniel Ortega was retained as commander in chief of the re-formed Nicaraguan army after Chamorro won the election.] Each of these options will help groups feel more secure even if their future political situation remains uncertain and could help assure the long-term success of a settlement. Since outside states are unlikely to retain a long-term presence in any country it is important to offer other creative means for groups to protect themselves.

Policy Recommendations

If correct, the theory presented in this chapter could offer important guidelines to policymakers searching for effective ways to resolve civil wars. It appears that groups will walk away from any peace settlement because once they begin to implement the terms they will not be able to enforce continued cooperation, they will not be able to survive attack, and they will not be able to credibly commit to honoring it and that outside intervention can help in alleviating each of these problems. Given this, a number of recommendations can be made:

To be effective, intervention must guarantee the safety of all domestic groups as they begin to demobilize, disarm and prepare for peace. It must be credible and, most importantly, it must be backed by force. ₄₄ Outside powers who are unwilling to remain involved through the establishment of a new government and who are unwilling to use force will have little effect on successful cooperation. They might halt the bloodshed temporarily, but by definition, such intervention cannot enforce a peace, since maintaining any settlement and rebuilding a stable community occurs only after a war is over. (Do not simply push for elections and then leave.)

Long-term enforcement, however, should be the job of institutions and not an outside guarantor. [This is where institutional design comes in/the actual state-building part of the equation.. Institutions can make it difficult for any one group to fortify a temporary advantage. Might include a short-term administrative option: take over day to day administration during the transition, set up judicial system and jails in order to ensure punishment of violations. Run elections. Make sure elections are free and fair. (Why wasn't this enough in Cambodia?) Over the long-term outside states can then reassure groups that new political rules will be enforced by promising both positive inducements for continued compliance (aid, recognition, membership in regional organizations), and negative inducements for cheating (cut off aid, recognition, membership; close borders; economic sanctions).] It is important to first set up good institutions that in and of themselves make it difficult to coopt power, and then secondly, to make sure any attempt will be punished.

What criteria should be used for evaluating the success of a settlement? Intervention aimed at successful resolution of a conflict should ensure that the violence of war ends and that central order is reestablished in the form of a new government. Ultimately, what this will require is (1)that the competing armies are demobilized and a brand new professionalized and non-partisan military formed, or soldiers from each of the formerly warring parties are consolidated under a fully integrated or brand new command structure. And (2) a new political order and previously fighting factions are incorporated as actors in the new regime. (Expand)

[Caveats: Is this a goal we want to pursue? Or should be simply stick to humanitarian concerns rather than larger political goals? What are the drawbacks? It's hard to find a stable outcome, it's costly to intervene, and unrealistically costly to set up. If cooperation is so difficult and order can be established through a decisive war, should outside powers intervene at all? Early compromise treaties might produce fewer casualties in the short-run but if they often break down and the war resumes, then early negotiation might not be the most humanitarian option to pursue.]

Conclusions

This chapter attempted to uncover the underlying causes of persistent internal conflict and attempted to identify when outside intervention could help facilitate cooperation. It argued that domestic groups choose to battle "to the death" before they reestablished order because any power-sharing arrangement would, by definition, require them to relinquish any independent military forces and this fact made them too helpless to survive attack and too weak to enforce long-term cooperation.

Until now, the problem of civil war resolution has been viewed mainly as a problem of war termination - how do you get groups to stop fighting - and not as the more difficult but accurate problem of state rebuilding from a situation of severe anarchy. Resolving a civil war is never simply a matter of agreeing to a cease-fire. That would only be the case if functioning, legitimate state structures already existed and a new leader merely needed to be appointed. But the settlement of civil war will always require the accommodation of new groups into government, new soldiers into the national armed forces and a new political system to support multi-party rule.

The disturbing problem this chapter points out is how a particularly severe security dilemma always arises whenever a peaceful settlement is attempted because the general requirements of state rebuilding force groups to enter a strategic game for which there is only one rational outcome: continued war. ₄₅ So even if a balance of power exists between the competing groups, even if elites are incorporated into the new government and are protected from vengeful extremists, even if you know your opponent sincerely wants peace and all the checks and balances are in place to guarantee future transfers of power groups will still walk away from a settlement once they realize the strategic implications of implementation.

In short, domestic cooperation under anarchy is difficult and it often breaks down. Once the international community understand why new states are so difficult to reconstruct, then outside assistance can be targeted in ways that effectively help rather than hinder cooperation. The cases of "attempted negotiation" analyzed in this chapter suggest that at least one lesson can be drawn from past experiences. When attempting to resolve civil wars, enforcement does matter. But it only matters in the short-term. If outside states truly wish these settlements to last over time, they must also consider how the institutional parameters of any new government shape groups' expectations about their future security and factor into decisions to fight or cooperate. Military force might be crucial for implementation, but institutional design matters far more in the long run.

Knowing what we now know, it is up to political leaders to decide what their goals and aims are. Intervention aimed at stopping the killing can be done relatively cheaply and with a minimum commitment of time and forces; something the American public appears eager to do. But these cease-fires will only be temporary for they do not solve the underlying problem of how to rebuild the state and truly find peace. The failed interventions in Somalia, Rwanda and the mixed success of Cambodia, has signaled to the international community that civil war resolution requires more than a simple cease-fire, more than passive peace keepers. True success requires a three-pronged strategy of short-term guarantees, stable institutions and a measure of self-help.

Bibliography

Assefa, Hiskias. Mediation in Civil Wars: Approaches and Strategies - the Sudan Conflict.(Boulder: Westview Press, 1987).

Baumhoegger, Goswin. The Struggle for Independence: Documents on the Recent Development of Zimbabwe (1975-1980). (Hamburg: Institute of African Studies, Africa Documentation Center, 1984).

Charlton, Michael. The Last Colony in Africa: Diplomacy and the Independence of Rhodesia. (Oxford & Cambridge: Basil Blackwell, 1990).

Dasgupta, Partha. "Trust as a Commodity", in Diego Gambetta, ed., Trust: Making and Breaking Cooperative Relations. (Oxford: Basil Blackwell, 1988).

Downs, George W., David Rocke and Randolph Siverson. "Arms Races and Cooperation", in Kenneth Oye, ed., Cooperation under Anarchy. (Princeton: Princeton University Press, 1986).

Fearon, James D. "Ethnic War as a Commitment Problem," unpublished paper, University of Chicago, May 16,1993.

Fearon, James D. "Rationalist explanations for war", International Organization. Vol. 49, No. 3, Summer 1995.

Gagnon Jr., V.P. "Ethnic Nationalism and International Conflict: The Case of Serbia," International Security. Vol. 19, No. 3, Winter 1994-1995.

Gambetta, Diego. "Can We Trust Trust?", in Diego Gambetta, ed., Trust: Making and Breaking Cooperative Relations. (Oxford: Basil Blackwell, 1988).

Gilpin, Robert. War and Change in World Politics. (Cambridge: Cambridge University Press,1981).

Glenny, Misha. The Fall of Yugoslavia: The Third Balkan War. (New York: Penguin Books,1993).

Hardin, Russell. One for All: The Logic of Group Conflict, (Princeton: Princeton University Press, 1995).

Horowitz, Donald L. "Ethnic Conflict Management for Policymakers", in Joseph V. Montville,ed., Conflict and Peacemaking in Multiethnic Societies, (New York: Lexington Books, 1991).

Jervis, Robert. Perception and Misperception in Interational Politics. (Princeton: Princeton University Press, 1976).

Jervis, Robert. "Cooperation under the Security Dilemma", World Politics. Vol. 30, No. 2,January 1978.

Kydd, Andrew. "Signalling and Structural Realism", unpublished paper, UCLA, March 1996.

Lake, David A. and Donald Rothchild. "Containing Fear: The Origins and Management of Ethnic Conflict". International Security, Vol. 21, no 2 (Fall 1996.

Lemarchand, Rene. "Managing Transition Anarchies: Rwanda, Burundi, and South Africa inComparative Perspective", The Journal of Modern African Studies, Vol. 32, No. 4, December1994.

Licklider, Roy. "The Consequences of Negotiated Settlement in Civil Wars, 1945-1993," American Political Science Review. Vol. 89, No. 3, September l99S.

Niou, Emerson M.S. and Peter C. Ordeshook. "Preventive War and the Balance of Power: A Game-Theoretic Approach", Journal of Conflict Resolution. Vol. 31, No. 3, September 1987.

Pillar, Paul. Negotiating Peace: War Termination as a Bargaining Process. (Princeton: Princeton University Press, 1983).

Posen, Barry R. "The Security Dilemma and Ethnic Conflict," in Michael E. Brown, ed., Ethnic Conflict and International Security. (Princeton: Princeton University Press, 1993).

Przeworski, Adam. Democracy and the market: Political and economic reforms in Eastern Europe and Latin America. (Cambridge: Cambridge University Press, 1991).

Reiter, Dan. "Exploding the Powder Keg Myth: Preemptive Wars Almost Never Happen," Intemational Security. Vol. 20, No. 2, Fall 1995.

Rogowski, Ronald. "Trade and the variety of democratic institutions", International Organization. Vol. 41, No. 2, Spring 1987.

Snyder, Jack. "Nationalism and the Crisis of the Post-Soviet State", in Michael E. Brown, ed., Ethnic Conflict and Intemational Security, (Princeton: Princeton University Press, 1993).

Stedman, Stephen John. "Mediation in Civil War," in Michael E. Brown, ed., The International Dimensions in Internal Conflict. (Cambridge: MIT Press, forthcoming 1996).

Stedman, Stephen John. Peacemaking in Civil War: International Mediation in Zimbabwe. 1974-1980. (Boulder & London: Lynne Rienner Publishers, 1991).

Stremlau, John. The International Politics of the Nigerian Civil War: 1967-1970. (Princeton: Princeton University Press, 1977).

Telser, L.G. "A Theory of Self-enforcing Agreements", Journal of Business. Vol. 53, No. 1,1980.

Van Evera, Stephen. "Why Cooperation Failed in 1914," World Politics. Vol 38, No. 3, October1985.

Wagner, R. Harrison. "Peace, War and the Balance of Power", The American Political Science Review. September 1994.

Wagner, R. Harrison. "The Causes of Peace", in Roy Licklider, ed., Stopping the Killing: How Civil Wars End. (New York & London: New York University Press, 1993).

Walter, Barbara F. "The Critical Barrier to Civil War Settlement", International Organization (July1997).

Weingast, Barry. "Constructing Trust: The Political and Economic Roots of Ethnic and Regional Conflict," unpublished manuscript, Hoover Institute, October 1994.

Note 1:  See Table 1 (Presently unavailable.) Back.

Note 2:  It is interesting to note that if all attempted negotiations would reach and implement peace settlements, the number of successful settlements to civil war would closely approximate that of interstate wars. Back.

Note 3:  A war was coded as having experienced negotiations if both sides had enough bargaining power to elicit important concessions from each other, if factions actually held face-to-face talks, if issues relevant to resolving the war were discussed and if talks appeared to be undertaken in good faith. These qualifications eliminated scheduled talks that never took place, meatiness where no substantive issues were discussed and talks which excluded key participants in the war. Attempts were also made to exclude meetings undertaken in bad faith. A readiness to accept supervision, public announcements of important concessions, a desire to discuss the details of a transfer of power and long-term participation in lengthy negotiations all generated costs to the rivals and indicated more than a tactical interest in appearing cooperative. This qualification helped exclude bogus meetings where delegates refused to talk to each other, set ridiculous pre-conditions, stalled for time or feigned interest in settlement. Back.

Note 4:  To be considered "successful" a settlement had to (1) be jointly drafted by all combatants through give and take bargaining. This qualification helped exclude agreements where one side essentially capitulated or surrendered to the winning side. (2) The agreement had to keep the opposition intact as a bargaining entity. And (3) it had to end the war for at least five years. If a formal peace treaty was signed but broke down within this time period, it was considered a failed attempt and the outcome in these cases was coded on the basis of the eventual military results. Back.

Note 5:  This phenomenon has been documented in Barbara Walter, "The Critical Barrier to Civil War Settlement," International Organization, forthcoming July 1997. This pattern of decisive civil war resolution has also been recorded by George Modelski.1964. "International Settlement of Internal War", in James N. Rosenau, ed.,International Aspects of Civil Strife. (Princeton: Princeton University Press); Paul Pillar. 1983. Negotiating Peace: War Termination as a Bargaining Process. (Princeton: Princeton University Press, p. 25); and Stephen John Stedman. 1991. Peacemaking in Civil War: International Mediation in Zimbabwe. 1974-1980. (Boulder & London: Lynne Rienner Publishers, pp. 6-7). Back.

Note 6:  Source: Correlates of War Back.

Note 7:  See Cvijeto Job "Yugoslavia's Ethnic Furies", Foreign Policy, no. 92, Fall 1993. Cite examples in both the academic literature and the press. Back.

Note 8:  For solutions based on these assumptions see: Chaim Kauffmann, "Possible andImpossible Solutions to Ethnic Conflict", International Security, Vol. 20 no. 4 (Spring1996); John J. Mearsheimer and Robert A. Pape, "The Answer: A Three-way PartitionPlan for Bosnia and How the U.S. Can Enforce It," New Republic (June 14, 1993); John J. Mearsheimer and Steven Van Evera, "Hateful Neighbors," The New York Times, (September 24, 1996, p. Al9). Back.

Note 9:  Fred Charles Ikle, Every War Must End, (New York: Columbia University Press, 1991) p. 95. Back.

Note 10:  For examples, Susanne Rudolph and Lloyd Rudolph, "Modern Hate: How ancient animosities get invented," The New Republic, (March 22, 1993). Add other cities. Back.

Note 11:  See Roy Licklider, Stopping the Killing: How Civil Wars End, (New York: New York University Press, 1994); David T. Mason, "How Civil Wars End: A Rational Choice Approach," Journal of Conflict Resolution, Vol. 40 (December 1996). and Barbara Walter, "The Critical Barrier to Civil War Settlement", International Organization, July 1997. Back.

Note 12:  See Stephen John Stedman, "Negotiation and Mediation in Internal Conflict," in Michael E. Brown, ed., The International Dimensions of Internal Conflict (Cambridge, MA: MIT Press, 1996). Back.

Note 13:  These totalistic aims can result from a strict adherence to principle as appeared to occur in the American civil war when President Lincoln' s refused to consider any solution that split the Union and Jefferson Davis refused to discuss anything less than secession. Back.

Note 14:  See Pillar, Negotiating Peace:, Fred Ikle, Every War Must End, George Modelski's "International Settlement of Internal War", in James Rosenau's, International Aspects of Civil Strife. Back.

Note 15:  Paul Pillar, Negotiating Peace: War Termination as a Bargaining Process, (Princeton: Princeton University Press, 1983). Back.

Note 16:  Cites. Back.

Note 17:  I. William Zartman, "Dynamics and Constraints in Negotiations in Internal Conflicts," in Zartman ed., Elusive Peace: Negotiating an End to Civil Wars, (Washington, D.C.: The Brookings Institute, 1995) p. 11. See also Zartman, "The Unfinished Agenda: Negotiating Internal Conflicts," Roy Licklider, ed., Stopping the Killing Back.

Note 18:  Jane Holl, "When War Doesn't Work," in Licklider ed., Stopping the Killing. Back.

Note 19:  Bill Keller, "Will Angolan Rebel Give Peace a Chance?", The New York Times,(February 14, 1995) p. Al, A6. Back.

Note 20:  See Fearon, "Rationalist Explanations for War," International Organization, Vol. 49, No. 3, (Summer 1995), p. 380. Back.

Note 21:  Keller, p. A6. Back.

Note 22:  Cities. Back.

Note 23:  Stedman, p.348-349. Back.

Note 24:  Stedman... Back.

Note 25:  See James Fearon, "Rationalist Explanations for War," International Organization. Back.

Note 26:  Herz, Jervis, Posen, Fearon, Lake & Rothchild, Weingast, Kauffman, Walter. Back.

Note 27:  Cite Jervis. Back.

Note 28:  Barry R. Posen, "The Security Dilemma and Ethnic Conflict," in Michael E. Brown,ed., Ethnic Conflict and International Security. (Princeton: Princeton University Press, 1993). Back.

Note 29:  See Chaim Kaufmann, "Possible and Impossible Solutions to Ethnic Civil Wars". Back.

Note 30:  Cite Posen. Back.

Note 31:  Tang Tsou, America's Failure in China: 1941-1950 Chicago: The University of Chicago Press, 1963. p. 291 Back.

Note 32:  Fearon, p. 401. Barry Posen also views internal conflict as a form of prevention war, fought to stop a competitor from becoming more powerful. If groups believe they will lose status in an upcoming government, then violence will erupt because groups wish to take advantage of a strong but declining position. According to Posen, groups have strong incentives to grab as much military equipment as possible since a failure to do so would result in potentially significant power losses and ultimately a weaker position in the new political system. Back.

Note 33:  See Jack Snyder, "Nationalism and the Crisis of the Post-Soviet State", in Michael E. Brown, Ethnic Conflict and International Security. See also Weingast 1996; Gagnon 1994. Back.

Note 34:  On the importance of accurate information in overcoming the security dilemma see Jervis, Posen, Snyder, Lake & Rothchild. Back.

Note 35:  See Henk Goemans, "War Termination"... Back.

Note 36:  Edan O'Brien, "Twisted Talk, Ulster's War," The New York Times, January 25, 1997,p. A17. Back.

Note 37:  Discuss Weingast and Stedman as exceptions. Back.

Note 38:  Even mandatory force reductions usually included in international peace treaties do not leave the adversaries defenseless. Back.

Note 39:  For an excellent analysis of the effects of signalling on security dilemmas see Andrew Kydd, "Signalling and Structural Realism," unpublished paper, University of California, Riverside, March 1996. Back.

Note 40:  Mike O'Connor, "Bosnia's Military Threat: Rival Police", The New York Times,(January 12, 1997) p. A6. Back.

Note 41:  O'Connor, p. A6. Back.

Note 42:  For an excellent discussion of the importance of institutional design in settling guerrilla wars see Matthew Soberg Shugart, "Guerrillas and Elections: An Institutionalist Perspective on the Costs of Conflict and Competition", International Studies Quarterly, (1992) 36, pp. 121-152. Back.

Note 43:  (Missing.) Back.

Note 44:  How exactly should it be handled? Begin with a military option (12 months?) and then pull out but replace it with an internal police force could be sub-contracted out to the Finns, the Canadians. Tax the state itself to provide this service. Station soldiers at assembly areas. Back.

Note 45:  These findings should generalize to any situation of state failure or collapse; any case where states must be rebuilt under conditions of anarchy. Interestingly, most civil wars can be traced back to some form of institutional failure; some time when groups had to rebuild their government without any central authority to help enforce a new contract. In these cases groups face the same problem that civil war adversaries who wish to find democratic solutions to their war, albeit at a much lower level of fear. And they will find it extremely difficult to rebuild these states for exactly the reasons outlined above. Back.