Physicians have a motto that peacemakers would do well to adopt: "First, do no harm." Neither the United States nor the United Nations have quite grasped this. Since the end of the Cold War unleashed them to intervene in civil conflicts around the world, they have done reasonably well in some cases, but in others they have unwittingly prolonged suffering where they meant to relieve it.

How does this happen? By following a principle that sounds like common sense: that intervention should be both limited and impartial, because weighing in on one side of a local struggle undermines the legitimacy and effectiveness of outside involvement. This Olympian presumption resonates with respect for law and international cooperation. It has the ring of prudence, fairness, and restraint. It makes sense in old-fashioned U.N. peacekeeping operations, where the outsiders' role is not to make peace, but to bless and monitor a cease-fire that all parties have decided to accept. But it becomes a destructive misconception when carried over to the messier realm of "peace enforcement," where the belligerents have yet to decide that they have nothing more to gain by fighting.

Limited intervention may end a war if the intervenor takes sides, tilts the local balance of power, and helps one of the rivals to win - that is, if it is not impartial. Impartial intervention may end a war if the outsiders take complete command of the situation, overawe all the local competitors, and impose a peace settlement - that is, if it is not limited. Trying to have it both ways usually blocks peace by doing enough to keep either belligerent from defeating the other, but not enough to make them stop trying. And the attempt to have it both ways has brought the United Nations and the United States - and those whom they sought to help - to varying degrees of grief in Bosnia, Somalia, and Haiti.


Wars have many causes, and each war is unique and complicated, but the root issue is always the same: Who rules when the fighting stops? In wars between countries, the issue may be sovereignty over disputed territory, or suzerainty over third parties, or influence over international transactions. In wars within countries the issue may be which group will control the government, or how the country should be divided so that adversaries can have separate governments. When political groups resort to war, it is because they cannot agree on who gets to call the tune in peace.
A war will not begin unless both sides in a dispute would rather fight than concede. After all, it is not hard to avert war if either one cares primarily about peace - all it has to do is let the other side have what it claims is its due.

A war will not end until both sides agree who will control whatever is in dispute.

Is all this utterly obvious? Not to those enthusiasts for international peace enforcement who are imbued with hopes for global governance, unsympathetic to thinking of security in terms of sovereignty, or viscerally sure that war is not a rational political act. They cannot bring themselves to deal forthrightly in the currency of war. They assume instead that outsiders' good offices can pull the scales from the eyes of fighting factions, make them realize that resorting to violence was a blunder, and substitute peaceful negotiation for force. But wars are rarely accidents, and it is no accident that belligerents often continue to kill each other while they negotiate, or that the terms of diplomatic settlements usually reflect results on the battlefield.

Others sometimes proceed from muddled assumptions about what force should be expected to accomplish. In a bizarre sequence of statements last spring, for instance, President Clinton threatened air strikes against the Bosnian Serbs, then said, "The United States is not, and should not, become involved as a partisan in a war." Next he declared that the United States should lead other Western nations in ending ethnic cleansing in Bosnia, only to say a moment later, "That does not mean that the United States or the United Nations can enter a war, in effect, to redraw the lines . . . within what was Yugoslavia."

This profoundly confused policy, promulgated with the best of lawyerly intentions, inevitably cost lives on all sides in Bosnia. For what legitimate purpose can military forces be directed to kill people and break things, if not to take the side of their opponents? If the use of deadly force is to be legitimate killing rather than senseless killing, it must serve the purpose of settling the war - which means determining who rules, which means leaving someone in power at the end of the day.

How is this done without taking someone's side? And how can outside powers pretend to stop ethnic cleansing without allocating territory - that is, drawing lines? Yet Clinton and U.N. Secretary General Boutros Boutros-Ghali did not make threats to protect recognized or viable borders, but to enforce naturally unstable truce lines that made no sense as a permanent territorial arrangement. Such confusion made intervention an accessory to stalemate, punishing either side for advancing too far but not settling the issue that fuels the war.

Some see a method in the madness. There are two ways to stop a war: having one side impose its will after defeating the other on the battlefield or having both sides accept a negotiated compromise. The hope for a compromise solution accounts for a misconceived impartiality.

When is compromise probable? When both sides believe that they have more to lose than to gain from fighting. Because leaders are often sensible, this usually happens before a war starts, which is why most crises are resolved by diplomacy rather than combat. But peaceful compromise has to seem impossible to the opponents for a war to start, and once it begins, compromise becomes even harder. Emotions intensify, sunk costs grow, demands for recompense escalate. If compromise was not tolerable enough to avert war in the first place, it becomes even less attractive once large amounts of blood and treasure have been invested in the cause.

If neither side manages to pound the other into submission and a stalemate emerges, does a compromise peace become more practical? Not for a long time, and not until many more lives have been invested in the contending quests for victory. Stalemates rarely seem solid to those with a strong stake in overcoming them. Belligerents conjure up one set of military stratagems and schemes after another to gain the upper hand, or they hope for shifts in alliances or outside assistance to tilt the balance of power, or they gamble that their adversary will be the first to lose heart and crack. Such developments often do break stalemates. In World War I, for example, trench warfare in France ebbed and flowed inconclusively for four years until the Russian capitulation. This allowed the Germans to move armies from the east and achieve a breakthrough that unglued the Western Front and almost brought them victory in the spring of 1918. Then the Allies rebounded, turned the tables with newly arrived American armies, and won the war six months later.

Stalemate is likely to yield to negotiated compromise only after it lasts so long that a military solution appears hopeless to both sides. In the Iran-Iraq War, where U.N. mediation was useful, the two sides had fought ferociously but inconclusively for eight years. The United Nations smoothed the way for both sides to lay down their arms, but it is hard to credit that diplomatic intervention with as much effect in bringing peace as the simple exhaustion and despair of war-makers in Tehran and Baghdad. Mediation is useful, but it helps peacemaking most where peacemaking needs help least.


If there is any place where peacemaking needs help most, and fails most abjectly, it is Bosnia. There, the West's attempt at limited but impartial involvement abetted slow-motion savagery. The effort wound up doing things that helped one side, and counterbalancing them by actions that helped the other. This alienated both and enabled them to keep fighting.

The United Nations tried to prevent the Serbs from consolidating their victory, but without going all the way to consistent military support of the Muslims and Croats. The main U.N. mission was humanitarian delivery of food and medicine to besieged communities, but this amounted to breaking the sieges - a military and political effect. It is hardly surprising that the Serbs interfered when they could get away with it. In line with the humanitarian rationale, the United Nations supported "safe areas" - pockets of Muslims and Croats hanging on in areas conquered by the Serbs. Apart from such limited action to frustrate the last phase of territorial rearrangement by force, U.N. and U.S. attempts to settle the war were limited to diplomatic mediation, an arms embargo, a "no-fly zone," and economic sanctions on Belgrade.

For over a year, the U.N. presence inhibited forceful reaction to Bosnian Serb provocations because French, British, and other units on the ground were hostage to retaliation. U.S. and U.N. threats were not just weak and hesitant; by trying to be both forceful and neutral, they worked at cross-purposes. First, after much dancing around and wringing of hands, the United Nations and NATO used force on behalf of the Bosnian government, albeit with only a few symbolic "pinprick" air raids against Serb positions. But the outside powers did this while refusing to let those they were defending buy arms to defend themselves. Given the awkward multilateral politics of the arms embargo, this may have been understandable. But as strategy, it was irrational, plain and simple.

Impartiality compounded the absurdity in August 1994, when the U.N. military commander also threatened the Bosnian government with attack if it violated the weapons exclusion zone around Sarajevo. U.N. strategy thus bounced between unwillingness to undertake any combat at all and a commitment to fight on two fronts against both belligerents. Such lofty evenhandedness may make sense for a judge in a court that can enforce its writ, but hardly for a general wielding a small stick in a bitter war.

Overall, U.N. pressures maintained a teetering balance of power among the belligerents; the intervenors refused to let either side win. Economic sanctions worked against the Serbs, while the arms embargo worked against the Muslims. The rationale was that evenhandedness would encourage a negotiated settlement. The result was not peace or an end to the killing, but years of military stalemate, slow bleeding, and delusionary diplomatic haggling.

The desire for impartiality and fairness led outside diplomats to promote territorial compromises that made no strategic sense. The Vance-Owen plan and later proposals mimicked the unrealistic 1947 U.N. partition plan for Palestine: geographic patchworks of noncontiguous territories, vulnerable corridors and supply lines, exposed communities, and indefensible borders. If ever accepted, such plans would create a territorial tinderbox and a perpetual temptation to renew the conflict. Yet the one case in which Washington said it was willing to thrust tens of thousands of American troops into the Bosnian tangle was to enforce just such an accord.

In Somalia, the United States succeeded laudably in relieving starvation. Then, fearful that food supply would fall apart again after withdrawal, Washington took on the mission of restoring civil order. This was less limited and more ambitious than the outside powers' "strategy" in Bosnia, but it stopped short of taking charge and imposing a settlement on the warring factions.

Incongruously, the international operation in Somalia worked at throwing together a local court and police organization before establishing the other essential elements of government, an executive or legislature. Then U.S. forces set out to arrest General Mohamed Farah Aidid - who was not just a troublemaker but one of the prime claimants to governing authority - without championing any other contender. The U.S. attempts failed, but killed a large number of Somalis and further roiled the political waters in Mogadishu. Stung by casualties to U.S. forces, Washington pulled out and left U.N. troops from other countries holding the bag, maintaining an indecisive presence and taking casualties of their own.

It may have been wise to avoid embroilment in the chaos of conflict between Somali clans. But then it was naive to think that intervention could help to end the local anarchy. As author Michael Maren asked in The New York Times, "If the peacekeepers aren't keeping the peace, what are they doing?" - especially since the cost of intervention topped $1.5 billion. The U.N. operation was not only indecisive, Maren argued, but fueled the fighting by letting the feuding factions compete for U.N. jobs, contracts, and cash. In areas where U.N. forces were absent, the parties reached accommodation in order to reestablish commerce, rather than jockeying for U.N. resources.

After early forceful rhetoric by President Clinton had raised expectations about U.S. action, Bosnia and Somalia helped to brake enthusiasm in New York and Washington for taking on other peace operations. It is indeed wise to be more selective than in the heady days of hope for collective security that followed the end of the Cold War, but it will be unfortunate if the Western powers and the United Nations abandon such missions altogether. If they do not quit completely, the same problem - an attempt to bring peace turning out to postpone peace - can arise again if the misconceptions that produced it are not recognized.

Of course, not all problems are due to impartiality. In Haiti, for example, the United States and United Nations clearly did choose sides, supporting the exiled president, Jean-Bertrand Aristide. And eventually the unmistakable U.S. willingness to invade forced the junta in Port-au-Prince to back down. Even here, however, suffering was prolonged by the initially limited character of the intervention.

For over a year after the junta reneged on the Governor's Island agreement, Washington relied on economic sanctions against Haiti, a "trickle-up" strategy of coercion that was bound to hurt the innocent long before the guilty. The blockade gradually devastated the health and welfare of the country's masses, who were powerless to make the policy changes demanded by Washington and on whose behalf the sanctions were supposedly being applied. Yet sanctions offered no incentive to Haiti's kleptocratic elites to cut their own throats, and sanctions were not what made the generals sign the accord brokered by former President Jimmy Carter. Instead, the many months during which sanctions were left to work were used by the junta to track down and murder Aristide supporters at a steady pace.

Meddling in the long tragic saga of Haitian misgovernment is a dubious gamble for the United States, considering the formidable durability of the country's predatory political culture. While the relatively peaceful entry of the occupying U.S. forces was welcome, the fledgling Haitian experiment in democracy may not outlive the departure of U.S. armed forces. The September agreement did not disband the Haitian military or even completely purge its officer corps, whose corruption and terror tactics have long been most of the problem. Even worse, the accord hinted that - for the first time in the crisis - Washington might have again erred on the side of impartiality. American leaders spoke of the generals' "military honor," U.S. troops were ordered to cooperate with the usurpers' security forces, and many of the anti-Aristide gangsters were left free to plot to regain power. Deciding whether to intervene in Haiti was agonizing. Once that was done, however, picking a side was certainly wise. But that choice was weakened by dithering too long with sanctions, then appearing to waver in support for the chosen side when U.S. military force was finally applied.

Impartiality nonetheless remains a norm in many other cases. It has worked in cases that lie beyond traditional peacekeeping, such as the cease-fire mediation between Iran and Iraq, or the political receivership of the U.N. Transitional Authority in Cambodia (UNTAC). When looking at the reasons for their success, however, it becomes apparent that impartiality works best where intervention is needed least: where wars have played themselves out and the fighting factions need only the good offices of mediators to lay down their arms. Impartiality is likely to work against peace in the more challenging cases - where intervention must make the peace, rather than just preside over it - because it reflects deeper confusion over what war is about.


If outsiders such as the United States or the United Nations are faced with demands for peace in wars where passions have not burned out, they can avoid the costs and risks that go with entanglement by refusing the mandate - staying aloof and letting the locals fight it out. Or they can jump in and help one of the contenders defeat the other. But can they bring peace sooner than exhaustion from prolonged carnage would, if they remain impartial? Not with a gentle, restrained impartiality but with an active, harsh impartiality that overpowers both sides: an imperial impartiality. This is a tall order, seldom with many supporters, and it is hard to think of cases where it has actually worked.

The best example of imperial impartiality is the U.N. operation in Cambodia - a grand-scale takeover of much of the administrative authority in the country, and a program for establishing a new government through supervised elections and a constituent assembly. Despite great obstacles and tenuous results, UNTAC fulfilled most of its mandate. This success should be given its due. As a model to rescue the ideal of limited and impartial intervention, however, it falls short.

First, the United Nations did not nip a horrible war in the bud. As with Iran and Iraq, it capitalized on 15 years of exhaustion and bloody stalemate. The outside powers recognized that the main order of business was to determine who rules, but they did not act before the local factions were weary enough to agree on a procedure for doing so.

Second, U.N. intervention was limited only in one sense: it avoided direct enforcement of the transition agreement when local contenders proved recalcitrant. Luckily, such incidents were manageable, or the whole experiment would have been a fiasco. In other respects, the scale of involvement was too huge to provide a model. Apart from the wars in Korea and Kuwait, UNTAC was the most massive U.N. intervention in history. It involved thousands of personnel from a host of countries and billions of dollars in expenditures. The Cambodia operation proved so expensive, at a time when other demands on the United Nations were escalating dramatically, that it cannot be repeated more than once in a blue moon.

Third, although UNTAC should count as a success - especially after the election it conducted against all odds in 1993 - the operation's results have been unstable. Despite a tremendous U.N. presence, the terms of the transition agreement were never faithfully followed by all the local combatants and continued to erode after UNTAC's departure. For example, because the Khmer Rouge reneged, none of the Cambodian factions disarmed to the degree stipulated in the agreement; after the election, the constituent assembly never seriously debated a constitution, but more or less rubber-stamped Prince Norodom Sihanouk's demands; and sporadic fighting between the Khmer Rouge and other parties continued before and after UNTAC left.

Fourth, and more to the point, the U.N. success was linked with impartiality only in principle, not in effect. The real success of the transition overseen by UNTAC was not in fostering a final peaceful compromise among the parties in Cambodia, but in altering the balance of power among them and marginalizing the worst one. The transition did not compel an end to violent strife, but it did facilitate the realignment of parties and military forces that might bring it about. The old Cold War alignment of Sihanouk, Son Sann, and the Khmer Rouge against the Vietnamese-installed government in Phnom Penh was transformed into a new coalition of everyone against the Khmer Rouge. Any peace Cambodia is likely to achieve will come from the new balance of power.


The "peacekeeping" that has been the United Nations' forte can help fortify peace, but it does not create peace as "peace enforcement" is supposed to do. Since the end of the Cold War, the United States and the United Nations have stumbled into several imbroglios where it was not clear which of the two missions they were up to, and there has been much head-scratching about the gray area between operations under Chapters VI and VII of the U.N. Charter. Washington and New York have responded to rough experiences by remaining mired in indecision and hamstrung by half-measures (Bosnia), facing failure and bailing out (Somalia), acting only after a long period of limited and misdirected pressure (Haiti), or holding back from action where more awesome disaster than anywhere else called for it (Rwanda). To do better in picking and choosing, it would help to be clearer about how military means should be marshaled for political ends.

Recognize that to make peace is to decide who rules. Making peace means determining how the war ends. If Americans or U.N. forces are going to intervene to make peace, they will often have to kill people and break things in the process. If they choose to do this, it should be after they have decided who will rule afterward.

If claims or capabilities in the local fracas are not clear enough to make this judgment, then they are not clear enough for intervention to bring peace. By the same token, international forces should not mix in the dangerous business of determining who governs without expecting deadly opposition. An intervention that can be stopped in its tracks by a few dozen fatalities, like the U.S. operation in Somalia was, is one that should never have begun.

Avoid half-measures. If the United States or the United Nations wish to bring peace to violent places before tragedy unfolds in full, gruesome detail, they should act decisively - by either lending their military weight to one side or forcing both to compromise. In either case, leaders or outside powers should avoid what the natural instincts of successful politicians and bureaucrats tell them is sensible: a middle course.

Half-measures often make sense in domestic politics, but that is precisely because peace already exists. Contending interests accept compromises negotiated in legislatures, adjudicated in courts, and enforced by executives because the state has a monopoly on organized force; the question of "Who rules?" is settled. That is the premise of politics in peace. In war, that premise is what the fighting is all about. A middle course in intervention - especially a gradual and symbolic use of force - is likely to do little but muddy both sides' calculations, fuel their hopes of victory, or kill people for principles only indirectly related to the purpose of the war. If deadly force is to make a direct contribution to peace, it must engage the purposes most directly related to war - the determination of borders and the distribution of political power.

Do not confuse peace with justice. If outside powers want to do the right thing, but do not want to do it in a big way, they should recognize that they are placing a higher premium on legitimacy than on peace. Most international interventions since the end of the Cold War were not driven by the material interests of the outside powers but by their moral interests: securing peace and justice. Peace and justice, however, are not natural allies, unless right just happens to coincide with might.

Outside intervention in a civil war usually becomes an issue when the sides are closely enough matched that neither can defeat the other quickly. And when material interests are not directly involved, it is impractical to expect great powers or the United Nations to expend the resources for an overwhelming and decisive military action. So if peace should take precedence, intervention should support the mightiest of the rivals, irrespective of their legitimacy. If the United Nations had weighed in on the side of the Serbs, or had helped Aidid take control in Mogadishu rather than trying to jail him, there might well have been peace in Bosnia and Somalia long ago. If justice takes precedence, however, limited intervention may well lengthen a conflict. Perhaps putting an end to killing should not be the first priority in peacemaking, but interventionists should admit that any intervention involves such a choice.

Tension between peace and justice also arises in assessing territorial divisions like those proposed for Bosnia. If the aim is to reduce violent eruptions, borders should be drawn not to minimize the transfer of populations and property, but to make borders coherent, congruent with political solidarity, and defensible. This, unfortunately, makes ethnic cleansing the solution to ethnic cleansing. It also will not guarantee against later outbreaks of revanchism. But it can make war less constant. Better the model of India and Pakistan than that of Lebanon.

Do not confuse balance with peace or justice. Preventing either side from gaining a military advantage prevents ending the war by military means. Countries that are not losing a war are likely to keep fighting until prolonged indecision makes winning seem hopeless. Outsiders who want to make peace but do not want to take sides or take control themselves try to avoid favoritism by keeping either side from overturning an indecisive balance on the battlefield. This supports military stalemate, lengthens the war, and costs more lives.

Make humanitarian intervention militarily rational. Sometimes the imperative to stop slaughter or save the starving should be too much even for the most hard-boiled realists, and intervention may be warranted even if it does not aim to secure peace. This was a motive in Bosnia and Somalia, but interventions there involved presence in battle areas, constant friction with combatants or local political factions, and skirmishes that escalated without any sensible strategic plan. Bad experiences in those cases prevented rapid multilateral intervention in the butchery in Rwanda, which could have saved many more lives.

Operation Provide Comfort, the U.S. humanitarian intervention in northern Iraq, and the recent unilateral French action in Rwanda provide better models. In these cases, the intervening forces carved out lines within which they could take command without fighting, but which they could defend if necessary - areas within which the intervenors themselves would rule temporarily. Then they got on with ministering to the needy populations and protecting them from assault. Such action is a stopgap, not a solution, but it is less likely to make the war worse.

In Bosnia, by contrast, the "safe areas," weapons exclusion zones, and towns supplied by American airdrops were islands surrounded by hostile forces and represented messy territorial anomalies in what was effectively a Serb conquest. It was no surprise that the Serbs would hover, waiting to pounce whenever they thought they might be able to get away with it, probing and testing the resolve of the outsiders to fight, waiting for the international community to tire of the effort to keep the enclaves on life support.

Calling attention to mistakes, confusions, and uncomfortable choices is not intended to discredit intervention altogether. It is meant to argue for caution, because confusion about what is at issue can make such undertakings cause conflict rather than cure it. Doing it right is not impossible. The United States and United Nations have collaborated successfully in peacemaking in the past, most notably in the wars over Korea and Kuwait. Enthusiasm for widespread involvement in local conflicts in the early 1990s was based on expectations that it would require a small proportion of the effort of those two huge enterprises. Unfortunately, this was probably true in some cases where the United Nations held back, as in Rwanda, and untrue in some cases where it jumped in, as in Bosnia. Peacemaking will not always cost as much as it did in Korea and Kuwait. The underlying issues, however, are much the same - who is in charge, and in what pieces of territory, after a war ends. Intervention that proceeds as if the issues are different, and can be settled by action toward the belligerents that is both evenhanded in intent and weak in capability, will more likely prevent peace than enforce it.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Richard K. Betts is Professor of Political Science and Director of the Security Policy Program at Columbia University's School of International and Public Affairs. His latest book is Military Readiness, published by the Brookings Institution.
  • More By Richard K. Betts