WINNING UGLY

This page intentionally left blank

WINNING UGLYNATOs War to Save Kosovo

B I PWashington, D.C.

I H. DM E. OH

Copyright 2000

1775 Massachusetts Avenue, N.W., Washington, D.C. 20036www.brookings.eduAll rights reserved

Library of Congress Cataloging-in-Publication data

Daalder, Ivo H.Winning ugly : NATOs war to save Kosovo / Ivo H. Daalder and Michael

E. OHanlon. p. cm.Includes bibliographical references and index.

ISBN 0-8157-1696-6 (cloth)ISBN 0-8157-1697-4 (pbk.)1. Kosovo (Serbia)HistoryCivil War, 1998 2. North Atlantic

Treaty Organization. I. OHanlon, Michael E. II. Title.DR2087 .D33 2000 00-009198

949.71dc21 CIP

(First softcover printing, September 2001)

9 8 7 6 5 4 3 2

The paper used in this publication meets minimum requirements of theAmerican National Standard for Information SciencesPermanence of Paper

for Printed Library Materials: ANSI Z39.48-1992.

Typeset in Minion

Composition by Cynthia StockSilver Spring, Maryland

Maps by Parrot GraphicsStillwater, Minnesota

Map images by Mountain High MapsCopyright 1993 Digital Wisdom

Printed by Phoenix ColorHagerstown, Maryland

Board of Trustees

James A. JohnsonChairman

Leonard AbramsonMichael H. ArmacostElizabeth E. BaileyZo BairdAlan R. BatkinJames W. CicconiAlan M. DachsKenneth W. DamD. Ronald Daniel

The Brookings Institution is an independent organization devoted to nonpartisan re-search, education, and publication in economics, government, foreign policy, and thesocial sciences generally. Its principal purposes are to aid in the development of soundpublic policies and to promote public understanding of issues of national importance.

The Institution was founded on December 8, 1927, to merge the activities of theInstitute for Government Research, founded in 1916, the Institute of Economics, foundedin 1922, and the Robert Brookings Graduate School of Economics and Government,founded in 1924.

The general administration of the Institution is the responsibility of a Board of Trusteescharged with safeguarding the independence of the staff and fostering the most favor-able conditions for scientific research and publication. The immediate direction of thepolicies, program, and staff is vested in the president, assisted by an advisory committeeof the officers and staff.

In publishing a study, the Institution presents it as a competent treatment of a sub-ject worthy of public consideration. The interpretations or conclusions in such publica-tions are those of the author or authors and do not necessarily reflect the views of theother staff members, officers, or trustees of the Brookings Institution.

Robert A. DayLawrence K. FishCyrus F. Freidheim Jr.Bart FriedmanStephen FriedmanHenry Louis Gates Jr.Lee H. HamiltonTeresa HeinzSamuel HellmanRobert A. HelmanAnn Dibble JordanBreene M. KerrMarie L. Knowles

Thomas G. LabrecqueJessica Tuchman MathewsDavid O. MaxwellConstance Berry NewmanSteven L. RattnerRozanne L. RidgwayJudith RodinWarren B. RudmanMichael P. SchulhofRobert H. SmithJoan E. SperoVincent J. TrosinoStephen M. Wolf

Honorary Trustees

Rex J. BatesLouis W. CabotA. W. ClausenWilliam T. Coleman Jr.Lloyd N. CutlerBruce B. DaytonDouglas DillonCharles W. Duncan Jr.Walter Y. ElishaRobert F. ErburuRobert D. HaasAndrew Heiskell

F. Warren HellmanRoy M. HuffingtonThomas W. JonesVernon E. Jordan Jr.James T. LynnDonald F. McHenryRobert S. McNamaraMary Patterson McPhersonArjay MillerMaconda Brown OConnorDonald S. PerkinsSamuel Pisar

J. Woodward RedmondCharles W. RobinsonJames D. Robinson IIIHoward D. SamuelB. Francis Saul IIRalph S. SaulHenry B. SchachtRobert Brookings SmithMorris TanenbaumJohn C. WhiteheadJames D. WolfensohnEzra K. Zilkha

This page intentionally left blank

vii

Foreword

I , violence again erupted in the Balkans,this time in Kosovoan ethnic cauldron in the heart of theformer Yugoslavia. For the third time in a decade, the violence was causedby the nationalist politics of Serbias long-standing dictator, SlobodanMilosevic. Whereas Milosevic had relied primarily on local Serb surrogatesin the Croatian and Bosnian wars, the violence in Kosovo was directly con-trolled by him and involved land within Serbia itself that was of great his-torical, cultural, and religious importance to all Serbs.

For years, it had been widely feared that a violent conflict in Kosovo waslikely to be both bloody, given the long-held animosity between Serbs andAlbanians, and highly destabilizing, in view of the large Albanian popula-tions just to Kosovos south. For these reasons, the United States and its Eu-ropean allies were determined to avoid an escalation in the conflict. Onceviolence erupted in Kosovo in March 1998, Washington and its allies soughtto stop it, and to forge agreement on a political solution that would providethe Kosovar Albanians with increased autonomy (though short of the inde-pendence virtually all Albanians sought). These efforts failedand the re-sult was NATOs decision in March 1999 to intervene militarily in order toachieve by force what it could not achieve through negotiations.

NATO made the decision to go to war in the belief that a few days oflimited bombing in the Balkans would likely suffice to persuade Milosevicto end the attacks on the Kosovar Albanian population and accept a politicalformula for restoring Kosovos autonomy. That proved to be a major miscal-culation. Rather than bowing to NATOs will, Milosevic escalated his violentcampaign against the local populationforcibly removing 1.3 million peoplefrom their homes and pushing 800,000 people entirely out of Kosovo. Rather

viii

than undertaking a limited use of coercive force, NATO became engaged inthe most extensive combat operations in its fifty-year history.

In this book, Ivo Daalder and Michael OHanlon provide the first com-prehensive examination of the causes, course, and consequences of the con-flict in Kosovo. Their focus is the immediate period of crisis of 1998 and1999 that culminated in the eleven weeks of bombing. Daalder and OHanloncritically assess the policies of the Clinton administration and other NATOgovernments over this time period, with particular emphasis on three mainquestions: Was the war avoidable (and if so, how)? Did NATO win the war,not only in narrow military terms but also in a broader political and strate-gic sense? And why did Slobodan Milosevic, after seventy-eight days of bomb-ing, finally relent and accept NATOs terms for a cessation of the war? Fromthis historical and policy analysis, they draw lessons for future situations inwhich the United States and its allies might consider using force to save livesand shore up the stability of critical regions of the world.

The authors draw for their analysis on an extensive body of primary andsecondary literature. But this book would not have been possible withoutthe assistance of many U.S. and European officials who were directly in-volved in making and executing policy and consented to be interviewed.These officials include Joachim von Arnim, Samuel Berger, Edgar Buckley,General Wesley Clark, Ambassador Charles Crawford, Ambassador RobertGelbard, Ambassador Sir John Goulden, Ambassador Phillippe Guelluy,General Sir Charles Guthrie, Morton Halperin, Ambassador RichardHolbrooke, Colonel Gregory Kaufmann, General Klaus Naumann, JamesOBrien, Ambassador Francisco Olivieri, Christian Pauls, John Sawers, Gre-gory Schulte, William Shapcott, Jamie Shea, General Rupert Smith, Ambas-sador Alexander Vershbow, Lieutenant General David Weisman, andBrigadier General Klaus Wittmann. Daalder and OHanlon are exceedinglygrateful for their cooperation.

This book could not have been written in such a short period of timewithout the indefatigable research assistance of Karla Nieting and MicahZenko, who never tired of searching for additional sources and informationand provided critical assistance in tracing the chronology of the conflict.Additional assistance was provided by Brian Finlay and Jason Forrester, forwhich the authors are grateful. They would also like to thank Susan Jacksonfor her work in verifying the manuscript, and Jessica Friedman for prepar-ing the verification materials that enabled the process to be completed inshort order. Kate OHanlon also helped her brother with useful and impor-tant insights. The production staff at the Brookings Institution Press, nota-bly Diane Hammond, who edited the volume, and Carlotta Ribar and Sherry

ix

Smith, who proofread and indexed the pages, also made possible the veryshort publication turnaround.

The authors benefited greatly from reviews of a draft of the manuscriptby Robert Art, Michael Brown, and Warren Zimmermann, as well as infor-mal reviews by Hans Daalder, David Ochmanek, and Richard Ullman. Asidefrom reading and commenting on the manuscript twice, Richard Haass as-sisted in innumerable ways throughout the project. They all have the au-thors heartfelt thanks.

Just as important, the authors would like to thank both the GermanMarshall Fund of the United States and the John M. Olin Foundation fortheir financial support of this effort.

The views expressed here are those of the authors and should not be as-cribed to those acknowledged above or to the trustees, officers, or other staffmembers of the Brookings Institution.

Michael H. ArmacostPresident

Washington, D.C.April 2000

To our wivesElisa Harris and Cathryn Garland

Contents

1 Introduction 1

2 The Escalating Crisis 22

3 The Road to War 63

4 Losing the War 101

5 Winning the War 137

6 Conclusions and Policy Implications 182

Appendixes

A Chronology 227

B Military Issues in Operation Allied Force 235

C Documents 247

Notes 279

Index 333

xi

Kosovo and Its Neighbors

Ad r i a t i c

S e a

Sarajevo

LjubljanaZagreb

Belgrade

Skopje

Tirana

Podgorica Pristina

ALBANIA

GREECE

ITALY

ROMANIA

BULGARIA

HUNGARY

AUSTRIA

MONTENEGRO

SLOVENIA

CROATIA

BOSNIA ANDHERZEGOVINA

MACEDONIA

KOSOVO

S E R B I A

0 100 km

0 100 mi

Pristina

Vucitrn

Zitorada

KosovskaMitrovica

Durakovac

Klina

Vitomirica

Shalqin

Brod

Decani

Junic

Poljance

NovoBrdo

KosovskaKamenica

JanjevoLipljan

KosovoPolje

Obilic

Srbica

Glogovac

Istok

Pec

Zubin Potok

Tutin

Rozaj

BajramCurri

Dakovica(Djakovica)

Kukes Dragas

Oragivac

Malisevo

Lapusnik

Prizren

Racak

Strpce

Stimlje

Urosevac

Gnjilane

Presevo

Kumanovo

Bujanovac

Tetovo

Kacanik

Vitina

Blace

Skopje

Kamenjane

VrapcisteGostivar

Sjenica

ZvecanZvecan

Prokuplje

Doljevac

Kursumlija

Blace

Medveda(Medvedja)

Lebane

Bojnik

Brestovac

Leposavic

Raska

NoviPazar

K O S O V O

S E R B I A

A L B A N I A

M A C E D O N I A

MONTENEGRO

Kosovo

This page intentionally left blank

Introduction

O , , NATO went to war for the first timein its fifty-year history. Its target was not a country, but aman. As the Serb leader of Yugoslavia, Slobodan Milosevic had been mostresponsible for a decade of violence that accompanied the breakup of Yugo-slavia. Well over a hundred thousand people had been killed and millionsdisplaced in Croatian and Bosnian wars during the first part of the 1990s.Now a similar humanitarian catastrophe threatened in Kosovo, part of Serbia,the heart of the former Yugoslavia. Milosevics security forces were arrayedagainst the Kosovo Liberation Army (KLA), a small insurgent force, and theethnic Albanians who dominated the areas population. In the previous yearof fighting, nearly two thousand people had been killed and many hundredsof thousands were driven from their homes. A full-scale war in Kosovo be-tween Serbs and Albanians would likely have been particularly brutal, leav-ing untold death and destruction in its wake. Compounding the likelyhumanitarian disaster was the potential for large numbers of refugees en-gulfing the fragile border countries of Macedonia and Albania, with conse-quences for stability and security across the entire region.

So NATO went to war. For a decade, the alliance had wavered in its re-solve to confront Milosevic. At times, the Serb leader had proven a willingpartner in negotiating a halt to the regions violence. More often, he hadbeen the source of that violence. For more than a year, the United States andits principal European allies had tried to head off a military confrontationby seeking to engage the man most responsible for the carnage that hadbefallen Kosovo, an approach similar to that followed in Croatia and Bosniaearlier in the decade. The Kosovo effort failed, not least because Milosevicdisplayed little interest in defusing a confrontation with a NATO alliance heassumed would soon founder in disagreement over how and to what extentto prosecute a war.

Late on the night of March 24, NATO warplanes began what was ex-pected to be a brief bombing campaign. The purpose of the campaign wasto force Milosevic back to the negotiating table so that NATO could find away short of independence to protect Kosovos ethnic Albanian populationfrom Serb violence and political domination. This bombing campaign, itwas emphatically stated, was not a war, and none of the NATO leaders hadany intention of waging one.

Politics at home and abroad were believed to constrain the United Statesand its partners in the use of force. When hostilities began, President BillClinton had just survived his impeachment ordeal. He faced a Congress thatwas not just politically hostile but also increasingly wary of U.S. militaryaction designed to serve humanitarian goals, including in the Balkans. Al-though Clinton had authorized the use of military force several times in hispresidency, he had not ordered American soldiers into situations in whichsome were likely to be killed since the Somalia operation had gone tragicallywrong in late 1993. Against this backdrop, the president failed to prepare thecountry for the possibility that NATOs initial bombing raids might be theopening salvo of a drawn-out war. Nor were he and his top advisers reallyprepared for this possibility themselves.

As alliance aircraft revved up their engines to start a short air campaignfocused primarily on Serb antiaircraft defenses, the expected operation hadthe flavor of a number of other recent, short, and antiseptic uses of Westernairpower. Three months earlier the United States and Great Britain had con-ducted a four-day bombing campaign against Iraq, and in August 1998 theClinton administration had launched cruise missile strikes against suspectedterrorist facilities in Afghanistan and Sudan in retaliation for the bombingsof two U.S. embassies in Africa by the Osama bin Laden network. Neither ofthose recent military operations had achieved core U.S. strategic objectives.Saddam Hussein had not allowed weapons inspections to resume in Iraq,and the bin Laden network remained intact and, by all accounts, poised tostrike again. These generally unsuccessful attacks did little to enhance thecredibility of the United States. They were designed more to punish, and tosend a message, than to compel an adversary to change his behavior ordirectly achieve concrete strategic objectives.

NATOs security interests seemed even less engaged in the Balkans thanthey had been with Saddam Hussein or bin Laden. This was apparent in thealliances goals for Kosovo, which were quite nuanced. NATO did not seek todefeat the Serb-dominated Yugoslav armed forces, cause a regime change inBelgrade, or gain Kosovo its independence. Rather it sought to convince

Milosevic to resume negotiations that would allow an armed internationalforce into Kosovo to quell the violence that had erupted there in March1998. Beyond that the alliance objective was nothing more than autonomyfor the ethnic Albanian majority in Kosovo. As a major inducement toMilosevic, it even promised to disarm the KLA. Under such circumstances,a protracted NATO bombing campaign seemed disproportionateand thusunlikelyto most onlookers, be they in Belgrade, Brussels, Washington, orelsewhere.

However, what had started very much as a foreign ministers battle soonbecame NATOs first real war. Seventy-eight days later, it finally ended asSerb forces left Kosovo and a NATO-led international force of 50,000 beganto move in. But over the intervening weeks a great deal of destruction waswrought, by Serbs against ethnic Albanians and by NATO against Serbia.

Despite the fact that most of the worlds best air forces were conductingcombat missions over Yugoslavia from March 24 onward, the early phasesof the conflict were dominated militarily by Serb units in Kosovo. NATOlost the war in the initial going, and the Kosovar Albanian people paid theprice. Up to 10,000 or so died at Serb hands, mostly innocent civilians; thou-sands more were raped or otherwise brutalized. Some 800,000 people wereforcefully expelled from Kosovo, and hundreds of thousand more were dis-placed within the territory. Ultimately perhaps 1,000 to 2,000 Serbs per-ished as well, both civilians killed inadvertently and regular and irregularSerb forces killed on the battlefield.

In the end, NATO prevailed. Although there was no clear turning point,the NATO summit in Washington on April 2325, 1999organized origi-nally to celebrate the alliances fiftieth anniversarymay represent the bestdividing line between losing and winning the war. Before that time, the vastmajority of Kosovar Albanians were forced from their homes. Despite anintensification of its air campaign, NATO remained powerless to preventatrocities on the ground or to establish a public perception that it was trulycommitted to winning the war.

But the summit revealed an alliance unified in its conviction that thewar against Serbia must be won. War planning became more systematic,and further increases in NATOs air armada were authorized. The alliancesteeled itself sufficiently that even the accidental bombing of the Chineseembassy on May 7 by a U.S. B-2 bomber did not seriously threaten con-tinuation of the war effort. Perhaps most significantly, on April 25 RussianPresident Boris Yeltsin called Bill Clinton, resuming U.S.-Russian ties thathad been effectively frozen when the war began. Yeltsin, though still upset,

committed to do what he could to end the war, setting in motion a nego-tiating process that would ultimately put a 360-degree diplomatic squeezeon Milosevic.

When it was all over, the alliance was able to reverse most of the damageSerbia had caused in the early period of the war. Notably, of the nearly 1.3million ethnic Albanians driven from their homes, virtually all were able toreturn within a few short weeks of the end of the war. Serb forces left Kosovo,with NATO-led units assuming physical control of the territory. An interna-tional administration was set up to run Kosovo, effectively wresting politicalcontrol over the area from Belgrade.

Although overall political momentum began to shift in NATOs favoraround the time of the Washington summit, the military tide of battle turnedmost dramatically in late May. By then, NATO air assets had nearly tripledand the weather had improved, making precision bombs far more effective.In addition, the alliances political leadership had authorized attacks againsta much wider range of targets in Belgrade and elsewhere. The KLA, thoughstill a modest militia force, had begun to conduct limited offensives againstSerb positions within Kosovo, in some cases forcing Serb troops to exposethemselves, at which point they became more vulnerable to NATO attackaircraft.

By early June Serbia was reeling. In Kosovo Serb forces had lost substan-tial amounts of the equipment with which they had begun, and Serb sol-diers were finding themselves at considerable personal risk. In Serbiaelectricity grids were being severely damaged, water distribution was ad-versely affected in all major cities, and the businesses and other assets ofMilosevics cronies were being attacked with growing frequency.

During the eleven-week air campaign, NATO flew nearly 40,000 combatsorties, about one-third the number flown in six weeks during the 1991Desert Storm campaign. Fourteen of the alliances nineteen members par-ticipated in the attacks.1 The air campaign was conducted very profession-ally and precisely by the armed forces of the United States and other NATOmember countries. Although some 500 Serb and ethnic Albanian civilianswere killed accidentally by NATO bombs, that toll is modest by the stan-dards of war. Moreover, only two alliance jets were shot down in combat,and only two NATO troops diedU.S. Army pilots who perished in anApache helicopter training accident in neighboring Albania.

NATOs air war had two main thrusts: a strategic campaign against theSerb heartland and a tactical campaign against the Serb forces doing the

killing and the forced expulsions in Kosovo. NATO supreme commanderGeneral Wesley Clark rightly argued that, for understanding how NATO wonthe war, the indispensable condition for all the other factors was the successof the air campaign itself. The Pentagons report on the war reached a simi-lar conclusion.2 But neither ultimate victory nor historically low losses dem-onstrate that the air war was well designed or properly conceived by topdecisionmakers in Washington and other NATO capitals.

Final victory required more than bombing. Two critical factors occurredon the political front: NATOs demonstrated cohesion as an alliance andRussias growing willingness to cooperate in the pursuit of a diplomatic so-lution. On the military front, NATOs talk of a possible ground war (whichalliance leaders had unwisely ruled out when the bombing began) and thewell-publicized decisions to augment allied troop strength in Macedoniaand Albania proved to be crucial as well. Whereas the air war inflicted mount-ing damage, these other factors probably convinced Milosevic that no plau-sible escape remained. Once that became clear to him, capitulation becamehis best course, both to minimize further damage to Serbia and its militaryand to secure his position in power.

Although U.S. domestic politics complicated the conduct of the conflictat times, and did much to shape the limited way in which it was initiallyfought, they did not fundamentally threaten the operation once it was un-der way. Most polls showed clear, though hardly overwhelming or impas-sioned, majorities of the U.S. public supporting NATOs air campaign.3

Indeed, once the war started, the Clinton administration faced more criti-cism from those who felt its war plan to be excessively cautious than fromthose who believed the use of force to be wrongheaded in the first place. Ascolumnist E. J. Dionne Jr. wrote, liberals in particular supported this war forits humanitarian dimensions, getting over their Vietnam syndrome in theprocess.4 And while some conservatives objected to the war as not servingU.S. interests, much of the Republican foreign policy elite felt strongly thatthe conflict had to end in a NATO victory.

In this book, we trace the causes, conduct, and consequences of the Kosovoconflict, analyzing the prelude to war in 1998 and 1999, the period when theKLA first came into direct conflict with Serb forces in Kosovo and the latterbegan a deliberate escalation of the conflict. We also critically assess thekey decisions in NATO policymaking over the eighteen-month period fromearly 1998 through mid-1999: the October 1998 Holbrooke-Milosevic agree-ment to place unarmed international monitors in Kosovo, the Rambouillet

negotiations of early 1999, the alliances decision to begin bombing, and itsgradual realization that it would have to win a war, whatever that ultimatelyrequired. We conclude by drawing lessons from the conflict that may berelevant to managing similar crises and conducting other such interven-tions in the future.

Our basic thesis is summarized in the books title, Winning Ugly. NATOdid meet reasonable standards of success in its 1999 war against Serbia. Theoutcome achieved in Kosovo, while hardly without its problems, representeda major improvement over what had prevailed in the region up to that pointand certainly over what would have happened had NATO chosen not tointervene. It is in that relative sense that the policy was successful, not be-cause it was properly designed at most major stages and not because itachieved the best plausible outcome to which NATO might have aspired.

Operation Allied Force was far from a perfect diplomatic and militaryaccomplishment. The United States and its allies succeeded only after muchsuffering by the ethnic Albanian people on the ground. They prevailed onlyafter committing a number of major mistakes, which future interventionsmust seek to avoid. In fact, NATOs mistakes were so serious that its victorywas anything but preordained. Had Milosevic not escalated the conflict dra-matically by creating the largest forced exodus on the European continentsince World War II, and had alliance leaders not then realized they had toradically overhaul their military strategy, NATO could have lost the war. Thatwould have held very serious implications for the future of the alliance andeven worse implications for the peoples of the region.

The Roots of War

The immediate cause of the conflict in Kosovo was Slobodan Milosevic, andhis oppression of Kosovar ethnic Albanians in the preceding decade. Op-pression ultimately gave rise to violent opposition to Serb rule, first in theformation of the KLA and then in the spiral of violence of 1998 and 1999.But the antecedents of the war go back many centuries. The most famoushistoric event of the millennium in the territory was probably the 1389 Battleof Kosovo, in the Field of the Blackbirds, near Kosovos present-day capitalof Pristina. There, Serb forces attempted to fend off the invading Turks, withethnic Albanians probably fighting on both sides in the battle. A subsequentbattle in Kosovo in 1448 between the Ottoman Turks and the Hungarians,together with the fall of Constantinople to the Ottomans in 1453, sealed thefate of the region. The Ottoman Empire would soon dominate the region

and it in fact controlled Kosovo into the twentieth century. Looking back onthese momentous events, nineteenth- and twentieth-century Serb nation-alists mythologized the 1389 battle and, more generally, the role of Kosovoin their nations history. In the process, they portrayed the primarily MuslimAlbanians essentially as sympathizers of the victorious Turkish invaders. Thecomplex interaction of Serbs, Albanians, and Turks over the ensuing centu-ries provided the ground for all parties competing historical perceptions,myths, fears, and vendettas.5

Kosovos population became increasingly ethnic Albanian during theperiod of Ottoman rule. A decisive turning point, politically and demographi-cally, was the large Serb exodus from the region (ultimately into Hungary)in the late seventeenth century. It was caused by Ottoman armies pressingnorth, ending in their defeat at Vienna against the Habsburg dynasty duringthe Ottoman-Habsburg War of 168399. That war spelled the beginning ofthe end for Ottoman rule in the Balkans, though as noted it survived inKosovo for another two hundred years.

In the early twentieth century, the Ottoman Empire was driven out ofthe Balkans by Serb, Macedonian, and Bulgarian forces, shortly before itscomplete collapse. Serbia, having itself regained de facto independence inthe early 1800s and formal state status in the 1878 Treaty of Berlin, assertedcontrol over Kosovo in 1912. That was the same year in which an indepen-dent Albanian state was created for the first time, with many of the key moveson the road to independence occurring in Kosovo, conferring on the terri-tory a historic importance for the Albanian people comparable to that forthe Serbs.6 Serbia lost control of Kosovo during World War I. After the warboth Serbia and Kosovo were integrated into the new country of Yugoslavia,with Kosovo a province of Serbia.

At various times over the last century, Serbs drove large numbers of eth-nic Albanians from Kosovo in what would be antecedents to SlobodanMilosevics 1999 campaign to effectively empty the territory of nearly all ofthem.7 Nonetheless, ethnic Albanians remained the majority populationthroughout the century, representing an increasingly high percentage of allKosovars in recent decades. Serbs and Montenegrins constituted slightly lessthan 30 percent of the population in the early years of Titos rule, whichlasted from 1945 until his death in 1980. They gradually declined to lessthan half that percentage in recent times due to Serb departures and highAlbanian birth rates.8

Whatever the recent population proportions, Kosovo is a land to whichboth Serbs and Albanians have important and long-standing claims.9 For

that reason, claims by extremists on both sides that they have exclusive rightsto the land are false, as are claims that the peoples are so different from eachother as to be innately incapable of coexistence.10 The fact that Kosovos Al-banians are now effectively in charge in the provinceand that they shouldremain in control of at least most of it, whether through autonomy withinSerbia, republic status within Yugoslavia, or eventual independencehasnothing to do with original claims to the land. It has instead to do with thetreatment of the Kosovar Albanians by Slobodan Milosevic and his fellowSerb nationalists in recent times.

Problems became serious even before the rise of Milosevic in 1987. Asearly as 1981, a year after the death of Tito, a student uprising in Pristinagave rise to provincewide demonstrations against Yugoslav authorities andperhaps dozens or even hundreds of deaths of ethnic Albanians. For Serbsthe uprising was surprising in that Kosovo had been granted greater au-tonomy and rights, including Albanian-language schools, under the 1974revision of the Yugoslav constitution. For Albanians these new rights onlymade them hunger for more, and the deteriorating economic conditions inthe province together with their second-class status exacerbated the politi-cal tension. Additional incidents through the 1980s further divided Serbsfrom Albanians.11

In early 1989, as part of his effort to consolidate power as president ofSerbia (the position he held from 1987 until becoming president of Yugo-slavia in 1997), Milosevic stripped Kosovo of its autonomy. That denied theterritory the special status within the Yugoslav Federation that it had en-joyed since the adoption of the 1974 constitution. In response to growingSerb oppression, Albanians established parallel state structures that werechampioned by Ibrahim Rugova and his Democratic League of Kosovo(LDK), officially founded in December 1989 and committed from the out-set to opposing Serb sovereignty over the Kosovar Albanian people. In thefollowing years, Rugova would be elected and reelected as Kosovos presi-dent in unofficial elections among the regions ethnic Albanian population.12

Rugova and the LDK hoped that in demonstrating their ability to run theterritory in all but name, the West would come to recognize Kosovos rightto be independent, as it had for four Yugoslav republics in the early 1990s.Support for separation from Yugoslavia was essentially universal. Accordingto a 1995 survey, 43 percent of all Kosovar Albanians wanted to join Albania,and the remaining 57 percent desired outright independence, with nonefavoring any other solution (including the status of an independent repub-lic within Yugoslavia, which by then included only Serbia and the much

smaller Montenegro).13 That hope and expectation proved to be misplaced.Kosovo was not at the center of U.S. and European Balkan policy and, aslong as violence did not escalate, it would not be at the center of that policy.

Yet the United States and its European allies recognized that Kosovo wasa powder keg in the middle of a highly volatile region. With Albanians livingin at least four countries (Albania, Greece, Macedonia, and Yugoslavia), any-thing that stoked Albanian nationalism could be highly destabilizing forKosovos neighbors. Probably of most concern was Macedonia, whose popu-lation is a potentially volatile mix of Slavs with a large minority of Alba-nians. The fragility of Albania itself was also a reason for concern. Widespreadviolence in Kosovo was therefore to be avoided, even if that required directU.S. military action. This together with congressional pressure explains whythe Bush administration, which otherwise had a hands-off policy towardYugoslavias breakup, decided in late 1992, in response to indications thatSerbia might be contemplating a violent crackdown against the Albanianpopulation in Kosovo, to issue a stern warning that such action would leadto U.S. military action. In a letter to Milosevic, President Bush warned thatin the event of conflict in Kosovo caused by Serbian action, the UnitedStates will be prepared to employ military force against the Serbs in Kosovoand in Serbia proper.14 This so-called Christmas warning was reiterated bythe Clinton administration within a month of taking office in 1993, whenthe new secretary of state, Warren Christopher, stated, We remain preparedto respond against the Serbians in the event of conflict in Kosovo caused bySerb action.15

On the whole, however, Kosovo occupied a distinctly secondary place inU.S. and Western policy toward the region. Indeed, unlike Bosnia, Croatia,and the other former Yugoslav republics, Kosovo was regarded as an integralpart of Serbia rather than as a constituent part of the federation that brokeup in the early 1990s. Whereas the republics were regarded as new states thatemerged from Yugoslavias dissolution and thus enjoyed sovereign rights,this status did not apply to Serbias autonomous provinces (Kosovo andVojvodina), even though these had enjoyed many of the same prerogativesthe republics had, including their own constitutions, governments, judicia-ries, central banks, and seats alongside the republics in Yugoslavias eight-member federal presidency.16 As a result, the issue for Western policy inKosovo was not self-determination or national rights but how to protectminority and human rights.

Almost from the beginning, therefore, U.S. and European policy towardKosovo was limited to increasing pressure on Belgrade to improve the hu-

man rights situation in the territory and establish conditions for greaterautonomy and self-government. These demands were raised as a matter ofcourse in all diplomatic dealings with the Belgrade government. The chiefU.S. negotiator for ending the Bosnian war, Richard Holbrooke, also raisedthe issue of Kosovo with Milosevic during their Bosnia negotiations in Day-ton in 1995. However, not only were the issues to be resolved in Daytonhighly complex and the negotiations intense, the fact that Milosevics coop-eration was critical to success weakened the negotiators leverage in exactingthe concessions that would have been necessary for progress on Kosovo.17

Still, even after sanctions on Serbia were lifted as the reward for Milosevicsimportant role in concluding the Bosnian peace agreement, an outer wallof sanctions remained in place, partly to encourage the Serb government toimprove its policies in Kosovo.18 In 1996, moreover, the United States estab-lished an official government presence in Pristina, when it opened a culturalcenter run by the U.S. Information Agency that, in Holbrookes words,amounted to a virtual U.S. embassy. This step demonstrated U.S. concernover the deteriorating conditions in Kosovo. Finally, during the few meet-ings he had with Milosevic after successfully concluding the Dayton nego-tiations, Holbrooke consistently repeated the Christmas warning, eventhough he says that he was not 100 percent clear on what it was or exactlywhat it would mean in practice.19

Despite these limited efforts, there was no concerted attempt to resolvethe Kosovo issue before it exploded in full violence (for example, by threat-ening to impose new sanctions unless Milosevic restored autonomy to theKosovar Albanians). Although the failure to address Kosovo in Dayton wasunderstandable, the lack of international attention to the issue dealt a majorsetback to Rugovas strategy of nonviolence. It became increasingly clear inthe second half of the 1990s that the Serbs would not stop their repressionof Kosovos majority population and that the international community woulddo little to effect a change in Serb policy, let alone endorse the Kosovarsdemand for independence. For many ethnic Albanians, one conclusion wasinescapable: only violence gets international attention.20 A previously un-known groupthe Kosovo Liberation Armytook advantage of this real-ization and started to engage in sporadic violence, harassing and even killingSerb policemen and other authority figures. Its levels of violence were fairlymodest; the KLA claimed to have killed ten Serbs in the two-year period upto early 1998. Nonetheless, the situation was deteriorating.21

Meanwhile, Rugovas efforts at nonviolent resistance were leading no-where after Dayton. In 1996, he negotiated an agreement with Milosevic

(with the assistance of Communit di SantEgidio, an Italian Catholic char-ity) that would have given the educational systems run by ethnic Albaniansaccess to official government buildings. Milosevic failed to implement theaccord. Rugova was also hurt when the Sali Berisha government in Albania,from which he had received support, fell in 1997 as a result of the spectacu-lar collapse of a nationwide pyramid scheme. The ensuing chaos in Albanialed to looting of weapons stocks in many parts of Albania, some of themundoubtedly winding up in the hands of the KLA.22

In March 1998 Serb security forces stoked the fires by massacring eighty-five people in a brutally indiscriminate attempt to stem the KLAs growingimportance in Kosovo. At that point, the violence in Kosovo reached a criti-cal threshold and demanded sustained international attention. Unlessstopped by a third party, the ethnic Albanian population and SlobodanMilosevics Serb nationalists were headed for war. The only remaining ques-tions were two. First, did the United States and its allies care enough aboutKosovo to be true to Washingtons 1992 Christmas warning? Second, wouldthey find an integrated strategy involving both diplomacy and the threat offorce that would succeed in getting the local parties off their collision course?The respective answers to these questions were yes and no.

This book explains how these and other questions were raised, and an-swered, during the critical eighteen-month period from early 1998 throughmid-1999, when Kosovo was a top concern of Western policymakers. Chap-ters 2 and 3 focus on the prewar period, up to and including the Rambouil-let negotiations of early 1999. Chapters 4 and 5 trace the seventy-eight-dayviolent conflict, including the ups and downs of NATOs, Serbias, and theKosovar Albanians fortunes, and identify its key elements and milestones.The conclusion addresses three key policy questions: Was the war inevitable?Did NATO win? And why did Milosevic ultimately capitulate? It also drawsseveral lessons for future policy regarding coercive diplomacy, the use offorce, and humanitarian intervention.

Before addressing these questions, however, it is necessary to spell out themain argument of the book: that NATOs cause was worthy, and its effortsultimately were reasonably successful, but that the strategy it chose to pur-sue its cause was seriously flawed.

NATOs Worthy Cause

NATO had moral and strategic rectitude on its side in using military powerin the Balkans. First, upholding human rights and alleviating humanitarian

tragedy are worthy goals for American national security policy. Doing soreinforces the notion that the United States is not interested in power for itsown sake but rather to enhance stability and security and to promote cer-tain universal principles and values. Second, the United States and its allieshave a special interest in upholding these values in Europe, a continent thathas become generally free and undivided since the cold war ended but thatremains conflict ridden in the Balkans. Third, in addition to these humani-tarian and normative rationales, traditional national interest argues for quell-ing violence in the Balkans because instability there can affect key allies moredirectly than instability in most other parts of the world.

A number of critics of NATOs approach to the Kosovo crisis either dis-agree with these arguments or argue that the alliance should have found away to solve the problem without going to war. More specifically, critics ofthe alliance argue the following: that the level of violence in Kosovo did notjustify NATO military action, especially given the predictable fallout suchan action would have on relations with Russia and China; that a more tacti-cally creative and balanced negotiating strategy in the years before the con-flict, and particularly at Rambouillet in the winter of 1999, might have avertedwar; and that even after Milosevic capitulated to most of NATOs main de-mands in June, conditions in Kosovo did not improve enough to deem theoutcome successful. On all of these points, however, critics have overstatedtheir case.23

Certainly the levels of violence in Kosovo before March 24, 1999, weremodest by the standards of civil conflict and compared to what ensued dur-ing NATOs bombing campaign. The violence had caused the deaths of anestimated 2,000 people in the previous year. This was not an attempted geno-cide of the ethnic Albanian people. However, there was good reason to be-lieve that, without intervention, things would have gotten much worse.Milosevic and his fellow Serb extremists had already displayed their truecolors earlier in the decade in Bosnia, where at least 100,000 people, mostlyMuslims and Croats, had been killed. More recently, and more to the point,in 1998, before the October agreement between Milosevic and U.S. specialenvoy Richard Holbrooke, Milosevic and his henchmen had driven some300,000 Kosovar Albanians from their homes, with 50,000 winding up, vul-nerable and exposed, hiding in the hills of the province. It was one of theworlds five largest crises involving refugees and internally displaced per-sons in 1998 and the only one in a country subject to cold winters.24 Whatthe alliance has since learned about Milosevics planned Operation Horse-shoe only confirms these judgments. Recognizing that the KLA was prob-

ably becoming too strong and too popular within ethnic Albanian societyfor him to defeat using classic counterinsurgency techniques, he chose toexpel much of the civilian population instead. If this policy was conceivedto serve a military purpose, it had a real political appeal for Milosevic too,being the surest way to restore Kosovo to complete Serb control and to freeland for Serbs displaced from other parts of the former Yugoslavia in recentconflicts.

At the very best, had Milosevic been left to his own devices, hundreds ofthousands of ethnic Albanians would have been driven permanently out ofthe province, possibly causing serious economic and political consequencesin Albania and Macedonia. Fewer Kosovar Albanians might have died in aslower campaign of forced expulsion than the number who died in the ac-tual war, but even that is not certain. It is equally plausible that, once Milosevicsaw NATOs lack of resolve in protecting the Kosovars, he would have re-verted to the style of warfare perpetrated in Bosnia, killing far more than the10,000 or so who ultimately perished.

As for the effects on Russia and China, both of these countries had sup-ported previous UN Security Council resolutions demanding an improve-ment in the human rights situation in Kosovo. Yet neither offered anyconstructive or serious alternatives to NATOs adopted strategy. The dam-age to relations with Moscow and Beijing was deeply regrettable; the NATOMay 7 bombing of the Chinese embassy in Belgrade was both tragic andincompetent; and the need to act without a UN Security Council resolutionwas unfortunate. But NATO could not allow itself to be prevented from stop-ping a mass murderer in its own backyard by unreasonable demands fromforeign capitals. In fact, there may even have been a silver lining, in thatNATO demonstrated to Russia and China that it would not be intimidatedby their protests over a matter that did not concern them directlyand thatthey seemed uninterested in trying to solve in any case.

NATOs war against Serbia will remain an irritant in Western relationswith both China and Russia, perhaps for years, but it was already fading insalience by late 1999 and early 2000. By that point, most normal ties be-tween Western capitals and Moscow and Beijing had been resumed, the U.S.-China relationship had moved on to trade issues and Taiwan security matters,and the U.S.-Russian agenda had refocused on nuclear and economic issuesas well as on achieving a smooth transition from Boris Yeltsin to VladimirPutin. The UN Security Council had restored its ability to function effec-tively, as evidenced by resolutions proposing a new weapons inspection re-gime for Iraq and peacekeeping forces for East Timor and Congo. Despite

the charges of some, it is not plausible that NATOs war over Kosovo was achief cause of Russias resumed conflict in Chechnya or of Chinas tenserelations with Taiwan. Both of those problems are far too central to thosecountries core interests to be blamed on Western military action in theBalkans. Operation Allied Force was clearly bad for the Wests relations withChina and Russia, but it was hardly a turning point. Ultimately, both Chinaand Russia have more important things to worry about and more impor-tant matters to discuss with the United States and its allies.

What about the possibility of protecting the ethnic Albanians withoutgoing to war? Critics argue, for example, that NATO essentially ignored theKosovo problem throughout the 1990s, doing little to make the 1992 Christ-mas warning credible. Others argue that NATO made a deal with Milosevicimpossible when it demanded at Rambouillet that its troops have access toall Yugoslav territory and that a plebiscite on Kosovos future be held threeyears after the signing of an accord. Still others assert that partitioning Kosovobetween Albanians and Serbs might have prevented war or that armingthe KLA might have allowed NATO to stay out of the violent conflict once itdid begin.

It is true that NATOs tendency to neglect Kosovo throughout the 1990swas a mistake. At a minimum, once the 1995 Dayton Accords were signedand NATO troops deployed to Bosnia, the United States and its allies shouldhave threatened to impose sanctions as conditions in Kosovo deterioratedin the ensuing years. They no longer needed Milosevic to establish or keepthe peace in Bosnia and could thus hold him up to reasonable standards ofbehavior in his own backyard. However, it is doubtful that doing so wouldhave convinced Milosevic to back down; he had already demonstrated hiswillingness to accept sanctions as the price of trying to gain land in Bosniafor Serbs, and if anything he cared more about Kosovo than about Bosnia.Moreover, by the time of Rambouillet Milosevic knew the tide had turnedagainst him in Kosovo and that to keep his hold over the territory he wouldhave to wipe out the KLA rather than simply restore autonomy to the ethnicAlbanians (which it is doubtful that he wished to do in any event). Havingwatched NATO troops become ensconced in Bosnia, he probably alsodoubted whether NATO troops, once allowed in, would ever leave Kosovo.

NATO did err in insisting on military access to all of Serbia at Rambouil-let. It clearly failed to recognize that the kind of language common in status-of-forces agreements it had negotiated previously (including with Croatia)might be wrongly interpreted, but this provision was almost certainly notthe decisive factor in Milosevics thinking. The Serbs never raised it during

the negotiations, focusing their opposition instead on the proposed deploy-ment of a NATO-led force inside Kosovo. Had they objected to the provi-sion, negotiators would surely have recommended that alliance militaryauthorities change their position. NATO made this demand as a matter ofmilitary convenience and nothing else. As the later Kosovo Force (KFOR)operation showed, NATO did not need access to northern and central Serbiato carry out its mission in Kosovo.

The conflict in Kosovo was fundamental: only NATO troops could haveprotected the Kosovar Albanians reliably enough to convince the KLA todisarm, but such troops were anathema to Milosevic, as was the idea ofKosovos autonomy or independence. The parties were on a collision coursefor war, and a different negotiating strategy could not itself have changedthat fact. It may still have been worth making one last offer to Milosevic,partly to allow Russia a somewhat greater role in setting the terms of NATOsultimatum. But the practical purpose of doing so would have been more tolimit the fallout between NATO and Moscow once the war began than toavoid war. Only a very credible threat of massive NATO military actionademonstrated willingness to achieve by force what negotiations could notmight have been enough to convince the butcher of Belgrade to relent. Yetcritics do not generally make that argument, focusing instead on secondaryor tactical issues, where their case is unconvincing.

What about partitioning Kosovo between Serbia and the ethnic Albanianpopulation? It would have run counter to the views of virtually all NATOgovernments, which believed that partitioning within the various republicsof the former Yugoslavia, such as Bosnia, was a bad idea. That is not itselfsufficient reason to dismiss the idea analytically. However, while partition-ing can help produce stable peace accords in civil conflicts, it is doubtful itwould have worked in this particular instance. Ethnic Albanians, who werethe overwhelming majority of Kosovars, would have insisted on, and hadrights to, the majority of Kosovos territory. Yet that would surely have beentoo high a price for Milosevic, who viewed Kosovo as key not only to hisown rise to power a decade before but also to the territorial integrity of theSerb nation. Both sides essentially wanted all of the province. Perhaps parti-tion would have been moderately preferable, in Milosevics eyes, to havingNATO throughout all of Kosovo. But either way, serious NATO threats orthe use of force would have been necessary to convince him to give up mostof a territory that he and his countrymen held very dear.

Arming the KLA might have contributed to an eventual battlefield suc-cess once war was under way, but it could hardly have prevented violence or

allowed the Kosovar Albanians to win a quick victory. Leaving aside the factthat the KLA was an organization with goals for Kosovos independence thatNATO governments and Russia did not share, arming it would not haveachieved any of the stated goals. An organization with only a few thousandragtag fighters into early 1999, it could not have become strong enough totake on tens of thousands of Serb soldiers and policeor to preventMilosevics campaign of ethnic cleansingin the space of a few shortmonths of equipping and training by NATO.25 Arming the KLA might haveprevented Serbia from consolidating its control over the province but onlyat the price of turning Kosovo into another Afghanistan or Angola. Thatwas not the right way to pursue a policy focused first and foremost on hu-manitarian goals.

Finally, many critics argue that postwar Kosovo, beset by problems likethe exodus of local Serbs, is little improved from conditions that prevailedwhen Serbia ran the province. This claim is wrong. The level of per capitaviolence in Kosovo remains too high, but it dropped tenfold within the ninemonths after the war ended on June 10, 1999.26 Serbs have left in great num-bers, many out of a very real fear for their lives, but the displacement ofsome 100,000 Serbs since the end of the war is a far less severe violation ofhuman rights than what Milosevic did to the ethnic Albaniansand for themost part it happened much less violently. Two wrongs do not make a right.But people who have been discriminated against for decades, oppressed forthe last decade, brutalized for a year, and then driven from their homes andtheir landoften with the collaboration of local Serbscan be forgiven acertain paranoia, even if their revenge attacks against Serbs cannot be con-doned. To be sure, enough problems remain in Kosovo that NATOs victorycannot yet be called permanent. But the international community is now ina very favorable position to maintain basic military peace in the territoryand to gradually improve its economic and social conditions. With goodpolicymaking, it should be able to consolidate its victory, at least by stan-dards that are reasonable to apply to a place, such as Kosovo, that has re-cently suffered a vicious civil war.

NATOsand WashingtonsKey Mistakes

The story of the Kosovo crisis is largely a saga of NATO and its major inter-national partners doing the right thing but in the wrong way. From the be-ginning of the Kosovo crisis, U.S. and European leaders shared a commonbelief that they had to do something about the situation in this small ter-

ritory in the heart of the former Yugoslavia. They just could never agreewhat that something was. When Milosevics forces engaged in a brutalcrackdown on the KLA in early March 1998, the Clinton administrationknew it had to act for political, strategic, and moral reasons. Politically,Milosevics actions challenged U.S. and NATO policy in other parts of theBalkans, including the decision to reward the Yugoslav leader for his coop-eration in helping stabilize the situation in Bosnia. Strategically, widespreadviolence pitting Serbs against Albanians could rapidly spread to other partsof the Balkans, notably to Albania and Macedonia and even to Greece andBulgaria. Morally, after what the world had witnessed in Croatia and Bosniaearlier in the decade, it was not difficult to imagine what Milosevic and hishenchmen might be capable of doing against an ethnic Albanian popula-tion long despised by much of the Serb majority in Yugoslavia.

The choice for the Clinton administration and its European allies wasnot whether to act but how. The administrations fundamental failure indealing with the Kosovo crisis was that it never decided what it was preparedto do, except incrementally and reactively. It was likely not until May 1999six-plus weeks into NATOs war against Serbiathat President Clinton fi-nally determined that, if necessary, the United States would do whateverwas required to prevail, probably even including a U.S.-led allied invasion ofKosovo to end the war and ensure safety and autonomy for the KosovarAlbanians.

In the year and a half leading up to that point, the Clinton administra-tion resorted to speaking loudly and carrying a small stick. It threatenedlargely unspecified action, hoping that would be sufficient to influence theparties to enter a dialogue leading to a political settlement. That approachwas never adequate. Slowly but surely, the United States and its NATO alliesmoved down a slippery slope of making threats, planning for military ac-tion and demonstrations, backing up the threats by deploying military as-sets, issuing ultimatums, using airpower, intensifying strategic bombing, andfinally being on the verge of committing to a ground invasion of Kosovo. Noone thought that the policy would eventually end up there. Indeed much ofthe effort was designed to prevent the use of forceand certainly to preventthe use of ground forces, even in a peacekeeping mode.

The failure in Kosovo was the result of policymakers in Washington andelsewhere who proved unwilling or unable to set political objectives and toconsider how far they were prepared to go to achieve them militarily. To besure, the broad goals were widely agreed on: to end the violence and to es-tablish conditions for the political autonomy of Kosovo within Yugoslavia.

But since this broad objective proved unacceptable to the Serb authorities,who wanted to maintain the political status quo, and to many Kosovars,who wanted independence, an international protectorate of the territorywas the only real alternative. It took the alliance nearly a year to arrive at thatconclusion. Even when it finally did, it failed to develop a reliable strategyfor establishing such a protectorate-like arrangement in Kosovo. It had ahope, but not a plan.

NATO stumbled into war, unready either for countering Serbias massivecampaign to forcefully expel much of the ethnic Albanian population fromKosovo or to do militarily what it would take to achieve its stated objectives.Even if the war itself was not easily or demonstrably avoidable, NATO lead-ers should have been better prepared. That required knowing what the ob-jective was and then committing to achieve it with the necessary militarymight. Instead NATO went to war in the hope it could win without much ofa fight. It was proven wrong.

NATOs campaign plan was unsound in the wars early going. The faultdid not lie in the alliances decisionmaking processes or in specific foreigncapitals like Paris and Berlin, as some have argued.27 Rather, the fault lay inthe basic strategy espoused for the war by the United States and its alliedpartners as a group. The basic idea of using bombing as an element of coer-cive diplomacy against Milosevic was pushed incessantly by Washington,and most specifically by the State Department, with strong support fromNATOs military leadership. The U.S. government generally expected air strikesto last only a few daysa couple of weeks at the outsideas interviews withkey officials and other sources convincingly attest. The United States did noteven envisage hard-hitting attacks during that short period: on March 24 ithad made available only about one-third the number of aircraft it ultimatelydevoted to the war, and days earlier it had pulled its only nearby aircraftcarrier out of the Mediterranean region and thus away from the war zone.During the wars early going, NATOs limited number of strike sorties focusedlargely on attacking Serb air defenses, due largely to a body of U.S. Air Forcedoctrine that requires that air supremacy be established in the early phase ofany war.28 NATOs policy of keeping aircraft above 15,000 feet above sea level,which limited the effectiveness of the tactical bombingseverely so in thewars early going, given the predictably poor Balkan early spring weatherwas primarily due to Washingtons preference to avoid casualties at nearlyany cost. In short, the frequent postwar tendencies of Clinton administra-tion officials, particularly at the Pentagon, to blame the allies for the slowstart of Operation Allied Force is almost entirely without foundation.

Operation Allied Force was in its early weeks a textbook case of how notto wage war. The blindness of NATOs major members to the possibility thatthe war might not end quickly was astounding. As a result of that blindness,the alliance was caught entirely unprepared for what followed. Had NATOnot enjoyed such a huge military advantage over Serbia, the alliance mightwell have lost its first real war. The losers would have included the KosovarAlbanians as well as NATO itself, since a defeat would have called into doubtnot only NATOs raison dtre but even its basic competence in the postcold war world.

NATOs shortsightedness, and its cavalier attitude toward the use of force,could have had extremely serious consequences. Had Milosevics henchmenin Kosovo been more brutalfor example, on the scale of what RatkoMladics Bosnian Serb forces did in Srebrenica in July 1995or had foodsupplies in the hills and forests of Kosovo not held up for the many thou-sands of people who had to hide outdoors during the war, far more peoplemight have died, with NATO powerless to save them. Had these things hap-pened, NATOs ultimate victory would have been Pyrrhic.

Finally, had Milosevic not upped the stakes in the conflict by drasticallyescalating his forced expulsion campaign, NATO could easily have lost thewar. He so repulsed Western publics with his barbaric actions that the alli-ance found a resolve it would almost certainly not have otherwise displayed.If Milosevic had hunkered down and restrained his military and paramili-tary forces during the bombing, support within NATO countries for sus-taining the operation probably would have quickly dissipated.

Perversely, Milosevic came to NATOs rescue. In a way that alliance lead-ers did not anticipate, he shored up their resolve and cohesion by his brutaltreatment of the ethnic Albanians. Without it, NATO would probably havebombed for a few days and then been obliged to desist, even had Miloseviccontinued to resist an international armed presence in Kosovo. This argu-ment is supported by our interviews with numerous NATO governmentofficials, the Desert Fox precedent in Iraq from the previous December, andthe alliances limited enthusiasm for coming to the military aid of the KLA.True, Milosevic would have had to expect stronger economic sanctions inthe aftermath of such an unsuccessful bombing campaign, but he had provenhe could live with that. Moreover, Washington had already demonstratedthe December before in the bombing campaign against Iraqwhich wasboth short and ineffectivethat it could bomb and forget even more hei-nous and dangerous adversaries affecting even more important U.S. inter-ests. Milosevic had ample reason to think that NATO would bomb for a

spell, declare victory, and stop. He could have then proceeded to a morepatient form of ethnic expulsion, gradually, over a period of many monthsor even years, pushing ethnic Albanians out of large swaths of Kosovo whilealso weakening the KLAs hold in those areas. That he did not adopt such anapproach may have been his greatest mistake of the war.

Since the day the war ended U.S. officials from President Clinton on downhave concluded that the alliance did the right thingwhich is trueandthat it also did so in the right waywhich is not.29 Their argument restson the twin contentions that in the end NATO prevailed and that the alli-ance could not have fought any other way given its internal political con-straints. However, making war by accepting political constraints that impedesound military preparations can be a prescription for defeatand nearlywas in this case. Particularly for the United States, the alliances undisputedleader, accepting alliance political constraints rather than working to moldthem in support of the U.S. perspective was bad policy. It is true that NATOis, and must be, a committee. But in this war it was a committee without achairman, particularly in the conflicts early going.

What was NATOs real alternative to its policy of diplomatic caution andmilitary gradualism? Before the war, the proper approach would have beena muscular NATO threat to Milosevic, with the goal of convincing him toallow the establishment of a de facto international protectorate over Kosovothat would ensure the safety of civilians and demilitarize the KLA. Either inthe fall of 1998 or in the immediate aftermath of the January 1999 Racakmassacres, NATO should have promised a much more extensive and open-ended bombing campaign. Ideally it should also have deployed forces intothe region to conduct a ground invasion if necessary.

This approach might have produced a negotiated settlement allowing in-ternational peacekeeping troops into the province. If not, once it beganbombing NATO would have had the option of intervening quickly had mas-sive slaughters been undertaken by the Serbs or if a lack of food supplies hadled to widespread starvation. That it would have been extremely hard for theClinton administration to convince Congress and the NATO allies to sup-port such a strategy and that there was no guarantee that such a threat wouldthen have worked, we acknowledge. But there is no excuse for not trying.Whatever the outcome would have been on preparing a possible invasionoption, moreover, Washington could have and should have convinced NATOto pose a far more daunting aerial threat from the wars beginning. And itshould never have ruled a ground force option off the table.

Conclusion

Operation Allied Force, the last war in Europe in the twentieth century, wasultimately an accomplishment for which NATO, the Clinton administra-tion, and a number of other key actors can take satisfaction. The ethnic Al-banian people of Kosovo, who suffered significant oppression under SlobodanMilosevic, are today far better off than they would have been had NATOstood aside. Their violent reprisal against Serbs in Kosovo since the war ended,while highly regrettable, does not begin to compare to what had happenedbefore. The wars damage to Western relations with Russia and China, thoughreal, is generally reparable and has already been largely attenuated. More-over, NATO as an alliance distinguished itself by showing the political willto do what was right, on humanitarian and political grounds, even in theface of strident opposition from Moscow and Beijing. The demands the warplaced on NATO military forces, in budgetary terms and human terms, weremodest and were largely compensated for by important lessons the alliancesmembers learned about how to improve their individual and combinedmilitary capabilities for the future.

But this book is not, primarily, a laudatory history of NATOs first realwar. It is a critique and, in places, a rather severe one. NATO in general, andthe Clinton administration in particular, missed key opportunities in 1998and early 1999 to reduce the odds of war. The alliance then undertook armedhostilities when it was unprepared for real combat, unwisely confident thatits short campaign of coercive bombing would work. Its poor preparationand early lack of resolve extended the conflict; luckily that did not exact anenormous price in civilian or military lives lost, but it was riskyand un-necessary.

NATOs war in Kosovo was difficult enough, and unpleasant enough onmany grounds, that it is unlikely to be seen by Western governments as aprecedent for frequent humanitarian intervention. But the postcold warworld has already seen major Western military interventions designed tosave lives or uphold democratic principles in Panama, northern Iraq, Soma-lia, Haiti, Bosnia, Kosovo, and East Timor. That track record, and the con-tinuing prevalence of civil conflict around the world, suggests that Westerncountries need to learn as much as they can from NATOs 1999 war againstSerbia, for better and for worse. This war will not be the last time that NATOgovernments use force to save lives.

TheEscalating

Crisis

L , the Kosovo conflict finally turnedviolent in early 1998 when Serbia decided to crack downon the Kosovo Liberation Army, a small group of rebels bent on achievingindependence through violence. From the moment violence exploded, thebattle lines were drawn. On one side stood the Belgrade government, led byYugoslav president Slobodan Milosevic and backed by the armed might ofhis Interior Ministry police troops (the MUP) and the Yugoslav army (theVJ). Its goal was both to prevent any further disintegration of what remainedof Yugoslavia and to keep Kosovo firmly within Serbias orbit. On the otherside stood the Kosovar Albanians, who constituted the overwhelming ma-jority of Kosovos population and who had suffered under Serb repressionfor nearly a decade. Although divided over the means to be employedwithmuch of the political leadership under Ibrahim Rugova favoring peacefulresistance and the KLA preferring the use of violencethe Kosovar Alba-nians were united in their quest for the territorys independence not onlyfrom Serbia but also from the Federal Republic of Yugoslavia.

The United States and European countries had long been concerned aboutKosovo and the consequences of a violent eruption in the area. But they hadfailed to translate this concern into effective action to prevent the conflictsslide into bloodshed. Once the Serb assault commenced in early March 1998,however, international attention was immediate and involved the highestlevels of the U.S. and allied governments. From the start, these governments

shared several key assumptions on how best to approach the conflict, in-cluding that they had to act promptly to avoid what had happened in Bosnia,that unity of effort and American leadership were necessary to forge a viablestrategy, that concerted pressure on Milosevic would be necessary to reach asettlement, and that a solution short of Kosovos independence would haveto be found.

Throughout much of 1998, however, U.S., European, and NATO policytoward Kosovo was haphazard and marked by a tendency to avoid makingdifficult decisions. The focus of the initial effort was on economic sanctionsand on encouraging a dialogue between the parties in order to arrive at asettlement of Kosovos political future. The desire to halt an escalation in thefighting and to forge a political solution did not, at least initially, extend to awillingness to consider a forceful intervention to bring these results about.Rather than exposing themselves to the inevitable risks that any militaryoption entailed, the Clinton administration and its European allies werecontent to defer for as long as possible having to make a choice among theunpalatable options of letting the Serbs get away with their murderous cam-paign against the KLA, supporting ethnic Albanian aspirations for indepen-dence, and taking the decisive action necessary to avoid either of these twooutcomes. And once they were prepared to threaten force, they did so onlytentatively and irresolutely.

This indecisiveness continued even in October 1998, when RichardHolbrooke, the Clinton administrations special Balkans envoy, traveled toBelgrade in an effort to persuade Slobodan Milosevic to end his forces at-tacks on the Kosovar population (which by that time had forced some 300,000people from their homes, including upward of 50,000 into the mountains).Backed by the threat of NATO airpower, Holbrooke succeeded in gainingMilosevics apparent commitment to cease attacks on civilians, withdrawsome of his security forces from the region, grant access to humanitarianrelief agencies, allow refugees and displaced persons to return to their homes,and acquiesce to an international presence in Kosovo to verify his compli-ance with these commitments. However, key details of the Holbrooke-Milosevic agreement remained ill defined, and whatever NATO commitmentexisted to enforce compliance with its terms was effectively nullified by thedecision to deploy unarmed monitors to the region. Consistent with U.S.and European efforts throughout the crisis, the Holbrooke mission was onemore indication that the aim was less to find a viable and lasting solution tothe conflict than to push the final reckoning as far into the future as pos-sible. In doing so, however, this overall approach may well have sown the

seeds of an even more violent conclusion to the conflict, as developments in1999 would make clear.

Policy Assumptions and Contradictions

When violence exploded in the Kosovo hillsides in the first days of March1998, the six Contact Group nations (France, Germany, Great Britain, Italy,Russia, and the United States) reacted swiftly. Within days of the Serb crack-down that left eighty-five people dead, foreign ministers from the six coun-tries met in London to condemn the attack and take stock. The ministersdemanded that Milosevic agree to cease all action by Serb security forcesagainst the civilian population, withdraw Serb special police units from theterritory within ten days, allow humanitarian groups to enter Kosovo, andcommence an unconditional dialogue with the Albanian community. Fail-ure to meet these demands, the ministers warned, would lead to impositionof an arms embargo, a denial of travel visas for senior Yugoslav and Serbofficials, a moratorium on export credits for trade and investment, and afreeze on funds held abroad.1 The six countries also committed themselvesto remain on top of the issue, agreeing to meet by the end of March to assessBelgrades actions and, if necessary, to agree to additional steps.

This prompt, high-level, and unified response to the eruption of violencein Kosovo reflected a number of assumptions the six Contact Group coun-tries had derived from their shared experience in dealing with the Bosnianconflict earlier in the decade. First, all agreed that they had to act rapidly inorder to avoid a repeat of the Bosnian horrors. As U.S. Secretary of StateMadeleine Albright noted just days after the massacres: We are not going tostand by and watch Serb authorities do in Kosovo what they can no longerget away with doing in Bosnia.2 President Bill Clinton suggested much thesame in his first public comments: We do not want the Balkans to havemore pictures like weve seen in the last few days so reminiscent of whatBosnia endured.3 The need for a strong condemnation of and response tothe Serb crackdown was underscored by Albright in the London meetingwith her European colleagues on March 9, 1998. History is watching us,she told the gathering assembled in Lancaster House. In this very room ourpredecessors delayed as Bosnia burned, and history will not be kind to us ifwe do the same.4

A second assumption was that successful intervention required unity ofeffort as well as American leadership. Both had been absent in Bosnia. Inthat case, U.S. and European perceptions of the conflict and how best toapproach its resolution diverged as a result of the decision by European coun-

tries to deploy troops on the ground. Instead of joining them in this effort,Washington decided (with Europes full support) to cede leadership to Eu-rope. With Europe exposed to risks that the United States avoided, the twosides perspectives on the conduct of the war consequently diverged.5 In theKosovo case, the Clinton administration determined not to make this mis-take. From day one, Albright forcefully took the lead in devising an appro-priate response to the violence. Aside from a strong response, her goal wasto ensure allied unity rather than division and to forge a path that wouldkeep the Russians on board. This explains why the six-nation Contact Group,which had been founded to minimize disagreements over Bosnia, was cho-sen as the vehicle for developing a common approach to Kosovo.

A third assumption of Western policy was that only concerted pressureon Milosevic would prove effective in convincing Belgrade to end the vio-lence and commence a dialogue with the Albanian community in Kosovo. Apolicy of carrots had failed. In November 1997, Milosevic spurned a Euro-pean Union offer to improve diplomatic and trade relations with Belgradeand support its reentry into international institutions in return for accept-ing a negotiating process between the Kosovar, Serb, and Yugoslav authori-ties that would be supported by third-party mediation.6 Days before theviolence in Kosovo in March, the United States had lifted a number of sanc-tions as a reward for Milosevics assistance in moderating the Bosnian Serbleadership, only to find Belgrade planning a major crackdown in Kosovo.7

As Albright put it, by his actions in Kosovo, Slobodan Milosevic has madeit clear that he is spurning incentives that the United States and others haveoffered him in recent weeksunfortunately the only thing he truly under-stands is decisive and firm action.8

A final assumption guiding the international response to violence inKosovo was to rule out independence as a solution to the conflict. Inde-pendence posed two major problems. First, it could destabilize the rest ofthe region. The fragile interethnic consensus in Macedonia would be se-verely shaken, providing the minority Albanian population there with anincentive to join their Kosovar and Albanian neighbors in the quest for aGreater Albania. Macedonias destabilization, in turn, could reignite terri-torial and other ambitions among its neighborsBulgaria, Greece, andSerbiawhich in the early twentieth century had fought two wars overthe territory. A second concern fueling opposition to independence wasthat independence in Kosovo would set a precedent for Bosnia, whereBosnian Serb and Croat claims for independenceor for merger withneighboring stateswere at least as strong as those of the Kosovar Alba-nians. Having found what still appeared to be a viable solution to the Bosnia

crisis, no one wished to promote a policy for Kosovo that would call thatresolution into question.

Although these four assumptions of Western policy toward Kosovo werewidely shared, they contained at least three contradictions. First, there was aconflict between the desire to act quickly and decisively, and the perceivedneed to forge a consensus on policy not only with key NATO allies but alsowith Russia. Given the range of views within the Contact Group on how torespond to the crisiswith the Russians and, to some extent, the Italians,favoring a policy of relying on positive incentives and the Americans andBritish preferring a more confrontational policyforging a consensus ondecisive action proved difficult. At times, however, this conflict was bypassedwhen the need for action outweighed getting the agreement of all ContactGroup members, as when Russia abstained from decisions to impose eco-nomic sanctions on Belgrade and when NATO, in threatening to use force,bypassed the UN Security Council and a certain Russian veto.

A second contradiction concerned the belief that a solution to the Kosovocrisis lay in pressing Milosevic to end the violent crackdown in Kosovowhile at the same time NATO relied on him to negotiate a final settlementwith the Albanian community. This meant that the ability to apply pressureon Milosevic to end the violence would at every turn be constrained by therealization that in the end the Yugoslav president was central to any success-ful negotiations. This constraint was compounded by the belief that the suc-cess of Western policy in Bosnia also depended crucially upon Milosevicscooperation. It was therefore never really apparent who had leverage overwhom: the United States and its allies over Milosevic, or vice versa.

A third contradiction was pressuring Milosevic to end the violence whilehoping not to encourage the ethnic Albanians in Kosovo to push their maxim-alist claims for independence. Having rejected independence as an accept-able outcome of the conflict, Western policymakers had to constantly balancetheir pressure against Belgrade with the need to discourage the Kosovarsfrom pressing their case for secession.

These policy contradictions were not inevitable.9 Some of the assump-tions underlying Western policy could have been relaxed or even abandonedaltogether. For example, the decision to rely on consensus within the Con-tact Group inevitably resulted in least-common-denominator policies, hardlythe kind of approach necessary to convince Belgrade to change course. More-over, the best way to pressure Milosevic without encouraging the Kosovarsto seek immediate independence would have been if NATO countries hadbeen prepared to deploy ground forces in Kosovo as part of an agreement

with Belgrade to protect ethnic Albanians, demilitarize the KLA, and guar-antee their territorys continued inclusion in the Federal Republic of Yugo-slavia. None of these options would have ensured success, but all would haveimproved the prospects for a peaceful settlement. Conversely, if the use offorce was to remain a true option of last resort so as to maintain domesticand allied support for the policy, then abandoning opposition to Kosovosindependence in favor of pursuing partition might have been worth an at-tempt. Rather than making these choices, however, the United States and itsEuropean partners sought to defer making difficult decisions, preferringinstead to muddle through in the hope that somehow and someway a solu-tion would present itself that would at once end the violence, provide a firmpolitical basis for a settlement, and avoid confronting the international com-munity with the need to use massive force.

Responding to Violence

By early 1998, there were indications that the low-level violence that hadbeset Kosovo the previous year would escalate into a major conflagration.The Central Intelligence Agency warned in mid-January that the Serb au-thorities were contemplating a crackdown on ethnic Albanians. A monthlater it warned senior U.S. officials that Serb armored units were beginningto mobilize and that Interior Ministry police troops were moving towardthe Kosovo border.10 In late February, Robert Gelbard, the Clintonadministrations senior envoy for the region, traveled to Belgrade to warnboth sides that violence would solve nothing. In private talks with Milosevic,Gelbard indicated that Belgrade faced a choice: either cooperate in resolvingthe Kosovo problem (as it had in helping to bring peace in Bosnia by pro-moting a more moderate Bosnian Serb government), and be rewarded witha relaxation of sanctions, or face further isolation into a downward spiralof darkness if violence in Kosovo mounted.11

Gelbards warnings notwithstanding, Serb security forces launched amajor offensive against key KLA strongholds just days after his departure.On February 28, Serb forces killed some two dozen people in Qirez andLikosane in central Kosovo. Days later a major assault on Srbica and othervillages in the Drenica valley left another fifty-one people dead, including aKLA leader, Adem Jashari, and twenty members of his family. Those killedincluded at least eleven children and twenty-three women. Within a week,eighty-five people were murdered by Serb security forces.12 Serb officialsjustified their use of force as a necessary response to a terrorist group, ar-

guing that the attack on the village of Donji Prekaz where the Jashari familylived, had wiped out . . . the biggest and most important one.13

Strong Rhetoric and Weak Sanctions

Within days of the first reports of the Serb massacres in Kosovo, Secre-tary Albright traveled to Europe to consult with her Contact Group coun-terparts to set the course for a united response. Her main purpose was topush the European allies, American public opinion, and even her own gov-ernment toward concerted action designed to avert the kind of human trag-edy that had happened in Bosnia. Her leverage was neither a plan of actionnor a U.S. commitment to threaten or use force but rather her strong rheto-ric.14 It was a rhetoric steeped in a determination to avoid the appearance ofanother Munich (and the delay witnessed in Bosnia) and lacked the fear ofanother Vietnam-type quagmire that seemed to haunt some of her colleaguesin Washington and elsewhere.15 In London, Albright urged her colleaguesnot to delay and to push for immediate action rather than rely on rhetoricand diplomacy alone:

When the war in the former Yugoslavia began in 1991, the international com-

munity did not react with sufficient vigor and force. Each small act of aggres-

sion that we did not oppose led to larger acts of aggression that we could not

oppose without great risk to ourselves. Only when those responsible paid for

their actions with isolation and hardship did the war end. It took us seven

years to bring Bosnia to this moment of hope. It must not take us that long to

resolve the crisis that is growing in Kosovo; and it does not have to if we apply

the lessons of 1991. This time, we must act with unity and resolve. This time,

we must respond before it is too late.16

The Contact Group accepted the logic of Albrights position. In a state-ment issued on March 9, 1998, the ministers not only condemned Milosevicsactions but also demanded that he act within ten days, halting attacks bySerb security forces against the civilian population and withdrawing all Serbspecial police units from the province. They also insisted that Belgrade co-operate with the international community to monitor implementation ofthis demand by allowing the UN High Commissioner for Refugees (UNHCR)to visit Kosovo, granting access to investigators of the International Crimi-nal Tribunal for the former Yugoslavia (ICTY) to gather evidence of pos-sible war crimes, supporting the return of the Organization for Security andCooperation in Europe (OSCE) to Kosovo and other regions of Yugoslavia(where it had been briefly deployed in the early 1990s), and providing access

to Kosovo for Contact Group diplomats to monitor implementation of thesecommitments. To back up their demands, five of the six Contact Group na-tions agreed to consider additional measures, including instituting a com-plete arms embargo, denying visas to senior Serb government and securityofficials, placing a halt on export credit financing, and freezing Serb-heldfunds abroad.17

It would take another two months for all Contact Group members butRussia to accept the need to impose the economic sanctions first mentionedin the London meeting. Verbal commitments by Belgrade to accede to theContact Groups demands just before the meeting in Bonn in late Marchwere sufficient to delay imposition of the sanctions though not to forestallagreement on an arms embargo.18 The latter was imposed by the UN Secu-rity Council on March 31 (with China abstaining), after Russia succeeded ineliminating any reference to the situation in Kosovo as constituting a threatto international peace and security (which might have justified subsequentenforcement action). Instead, the embargo was to remain in effect untilMilosevic initiated a substantial dialogue with the Kosovar leadership, with-drew Serb police, and allowed access to Kosovo by UNHCR, the OSCE,humanitarian organizations, and Contact Group diplomats.19 After theUnited States threatened to abandon the Contact Groups process becauseof its least-common-denominator approach, the European members did fi-nally agree to freeze Serb funds abroad in late April and to tighten a ban onnew investments in early May.20

The Absence of Force

Conspicuously absent from the list of sanctions was any threat or discus-sion of military force. This absence was the more notable since both theBush and the Clinton administrations had warned Milosevic in 1992 and1993 that the U.S. response to the kind of violent crackdown that occurredin Kosovo in early March 1998 would be air strikes, including against targetsin Serbia proper. Yet although senior Clinton administration officials neverruled out using force and insisted that all options were open, no one re-peated this so-called Christmas warning, and any question about whetherthe warning was still in effect was consistently deflected.21 Given the beliefthat only pressure could work to convince Milosevic, why not explicitly in-clude the ultimate form of pressure: military force?

There were at least two reasons that the United States did not push theissue of threatening the use of force. First, there was very little appetite inWashington for another potential military adventure at this particular time.

Quite apart from the fact that the White House was then embroiled in amajor scandal involving the president, few top administration officials be-lieved that Milosevics transgressions warranted threatening force, at least atthis stage.22 Aside from the relatively small scale of atrocitiesthen still lessthan a hundred killed and no more than a few thousand made homelessthe Clinton administration was still recovering from its previous round ofthreatening force (in Iraq in the spring of 1998, due to an impasse over weap-ons inspections). If only for that reason, the White House, led by Clintonsnational security adviser, Samuel (Sandy) Berger, was wary about threat-ening force. This became apparent during a meeting in Bergers office inmid-May 1998 that Secretary Albright had requested in order for the na-tional security adviser to hear the case for threatening the use of force as away to increase the pressure on