Article

Pivotal Deterrence and the Kosovo War: Why the Holbrooke Agreement Failed

Timothy W. Crawford

Debate rages over whether the North Atlantic Treaty Organization’s (NATO) air war against Serbia was a success or failure. But one thing is certain: that war resulted from a prior failure of NATO policy. The goal was to prevent the escalation of conflict, and to achieve a political settlement that got “Serbia out of Kosovo, not Kosovo out of Serbia,” without having to use force. As intended, the KLA’s insurgency in Kosovo had provoked Belgrade’s ham-fisted repression tactics. The Serb military police (MUP) and the Yugoslav National Army (VJ) killed innocent civilians and loosed a flood of refugees wherever they went: 42,000 by June, 100,000 by August, and over 200,000 by October 1998. The Racak Massacre in January 1999 captured public attention and triggered a more vigorous international response.

At the Rambouillet (February 6-23) and Paris (March 15-19) Conferences, NATO threatened to bomb Serbia if it did not accept the Contact Group’s (composed of Britain, France, Germany, Italy, Russia, and the US) interim political settlement and NATO’s plan to enforce it with armed peacekeepers. These threats did not compel Serbia to sign the agreements, and they did not deter it from launching “Operation Horseshoe,” a brutal offensive that displaced hundreds of thousands of Kosovo Albanians (Kosovars).

Belgrade’s decisions were the most visible failures of NATO’s coercive diplomacy, but they were not the only ones. What made enforced negotiations at Rambouillet necessary was a breakdown of the cease-fire NATO imposed in October 1998. Under the threat of air-strikes, Serbia agreed to withdraw or stand-down most of its forces in Kosovo. Then, deterring the Kosovo Liberation Army (KLA) from provocations became crucial to NATO’s policy, a necessary condition for real progress in negotiations between Belgrade and Kosovar moderates. NATO failed to do this. That failure must be addressed if we are to understand why NATO did not prevent the war.

It is tempting to ignore it now that the war is won, refugees have returned, NATO controls Kosovo, and Milosevic is out of power. The temptation is strongest for those who uncritically subscribe to the official Clinton Administration view that “the responsibility for violence” in Kosovo lay “squarely with Belgrade,” that the KLA was “created” by the “extremist measures” of Serb forces. Comparing this view with boilerplate KLA propaganda (“the source and cause of all evil is in Belgrade—the military actions of KLA are only of a defense nature”) should at least give us pause.

If we want to understand why NATO’s policy to prevent the war did not work, we must explain why NATO failed to deter the KLA from escalating. This is not to assert a “moral equivalency” between what the KLA did in Kosovo and what Serb forces did. But it is wrong to infer from Serb atrocities that the KLA was not aggressive, that it did not also need to be deterred. If we do that we assume away NATO’s strategic problem, not understand it.