Exit Lessons

In the midst of negotiating the Soviet withdrawal from Afghanistan in the mid-1980s, a Western diplomat confided to United Nations mediator Diego Cordovez, “The Russians would like to get out of Afghanistan, but they don’t know how. And we in the West would like to cooperate and help them, but we don’t know how either.” The Soviet experience is not unique. Historically, it has always been easier to launch a military intervention than to end one, especially when the effort has not gone well. From the United States in Vietnam to the Soviet Union in Afghanistan to Israel in Lebanon, intervening powers have often found it exceedingly difficult to extricate themselves from bad situations. As the United States is learning in Iraq, even when you are determined to make an exit, it is easier said than done.

The debate over exit strategies originated in America’s painful experience during the Vietnam War, which led some foreign-policy thinkers to conclude that an exit plan should be a prerequisite for any military intervention. The debate intensified in the 1990s, after the end of the Cold War, as the United States undertook interventions that appeared to be matters of choice more than necessity. In laying down what came to be called the Powell Doctrine, then-chairman of the Joint Chiefs of Staff Colin Powell included an exit strategy on his list of conditions that should be met before the United States committed forces overseas. But from Somalia to the Balkans and Haiti, none of the subsequent conflicts to which U.S. forces were committed in the 1990s met this condition, much less Powell’s chief principle that interventions must be directly tied to the long-term security of American interests. These costly and inconclusive efforts led critics to put even greater emphasis on questions about how the story was going to end. A year before he was elected president, George W. Bush questioned President Bill Clinton’s 1999 decision to intervene in Kosovo: “Victory means exit strategy, and it’s important for the president to explain to us what the exit strategy is.”

Yet for all the talk of exit strategies, there has been little attempt to review their history and assess their effectiveness. The results of such a study are chastening. Since the end of World War II, the United States has been very active in the world, but it has had no monopoly on large-scale intervention. The Soviet Union was the other obvious player in the postwar period, but there were also interventions by Egypt (in Yemen), Cuba (in Angola and Ethiopia), and India (in Sri Lanka), among others. The majority of these states’ interventions did not end well. (Among the notable exceptions, at least from the perspective of the intervening power, were the U.S. defense of South Korea from 1950 to ’53, the Soviet Union’s 1956 invasion of Hungary, and the American overthrow of President Manuel Noriega’s corrupt Panamanian regime in 1989.) A survey of nearly two dozen major military interventions since the end of World War II reveals that intervening powers were able to craft effective exit strategies in only about a third of the cases—and those happen to be the cases in which the goals of the interventions had already been met. Both the successes and the failures yield a handful of clear lessons about getting out.

Lesson 1: How you leave doesn’t matter very much.

During the 2008 presidential campaign, former U.S. national security advisers Zbigniew Brzezinski, a Democrat, and Brent Scowcroft, a Republican, published a much-noted book in which they debated the future of U.S. foreign policy, strongly disagreeing on Iraq. Brzezinski called for a rapid withdrawal, arguing that it would encourage the Iraqi people to take responsibility for their own governance. Scowcroft contended that a quick pullout would have disastrous consequences for Iraq, the region, and the United States.

The assumption by both these eminent foreign-policy thinkers that it matters a great deal how you leave is widely shared, and it has a surface logic. An abrupt departure could unleash a civil war or create a power vacuum that adversaries could exploit. Leaving according to a considered plan, on the other hand, could allow an intervening power to withdraw with stability intact and its reputation unscathed.

Whether or not an exit leads to instability or the erosion of the intervening power’s reputation depends much more on the conditions under which the intervention ended than on the character of the pullout. There is no such thing as a failed intervention capped by a successful exit strategy.

What might such a smart exit strategy look like? It might be conditional on the creation of credible institutions that could govern the post-intervention state. It could require a slow-and-gradual withdrawal of troops rather than cutting and running, as the United States supposedly did in leaving Somalia in 1993 after the “Blackhawk Down” incident. Perhaps it involves leaving behind a residual force to keep the peace, as the United States did after the Korean War and as a number of states did in Bosnia and Kosovo in the 1990s.

The evidence suggests that these choices do not much matter. Whether or not an exit leads to instability or the erosion of the intervening power’s reputation depends much more on the conditions under which the intervention ended than on the character of the pullout. There is no such thing as a failed intervention capped by a successful exit strategy. This is not to say that exit strategies are completely inconsequential—some clearly are executed better than others. But exit strategies are far less important than the overall success of the interventions that precede them. America’s intervention in South Korea produced good results because it ended reasonably well—in stalemate, if not victory—not because of a brilliant exit strategy. The residual forces left in place there, and later in Bosnia and Kosovo, would not have been effective if the intervention had not established a degree of underlying stability. Exits of all shapes and sizes—from the hasty U.S. departure from Somalia to the prolonged Soviet withdrawal from Afghanistan—have failed to achieve much in the way of results that suggest that the particular form of exit strategy can have a significant impact.

Lesson 2: It’s hard to make a defeat look like a victory.

Leaders sometimes recognize Lesson 1, but they still spend a great deal of time debating exit strategies. Even if a struggling military campaign cannot be transformed into a success, appearances can be manipulated, they hope, so that the nation’s (and the leader’s) honor, reputation, and status can be salvaged. Nobel Prize winner Thomas Schelling lent weight to this argument in Arms and Influence (1966), contending that “face” is not merely a matter of a “country’s ‘worth’ or ‘status’ or even ‘honor,’ but . . . its reputation for action. If the question is raised whether this kind of ‘face’ is worth fighting over, the answer is that this kind of face is one of the few things worth fighting over.”

Attempts to save face usually take the form of arduously negotiated pre-withdrawal agreements that allow all sides to claim that they have accomplished something positive in the course of the fighting. That is what happened in Vietnam, Angola, and Afghanistan, where negotiations over pre-withdrawal agreements dragged on for years. Face-saving agreements include elaborate plans for post-withdrawal governance and security institutions as well as economic support and other aid to a fledgling local government. What is most striking about these agreements is that they are rarely credible, rarely last very long, and are almost always known to be flawed by the signatories. Two years after the endlessly negotiated 1973 Paris Peace Accords, North Vietnamese tanks rolled into Saigon. The Tripartite Accord, signed by Angola, South Africa, and Cuba, concluded Cuba’s 13-year presence in Angola in 1988 with a withdrawal stretched over another 18 months, yet Angola’s civil war continued for a decade. The 1988 Geneva Accords, which brought an end to the Soviet intervention in Afghanistan, proved equally ineffective, as the country slipped into chaos and then, in 1996, the hands of the Taliban.

It should not be surprising that such agreements quickly fall apart. Once the outside force has departed, the balance of power on the ground shifts, opening up opportunities for some contending groups and leaving others vulnerable. Opposition groups that may have been biding their time can also find fresh opportunities in the wake of evacuation, as the Taliban did in Afghanistan.

If face-saving agreements are rarely credible and the signatories know it, why do they continue to pursue them? History gives two answers. First, the agreements offer short-term domestic rewards to political leaders. Even if the pacts are likely to collapse eventually, extricating forces in a way that appears honorable may offer enough cover to allow a leader to get reelected or stay in power. Second, an agreement may hold together long enough to allow leaders to disclaim any responsibility for violence that recurs after their troops have left.

Once the outside force has departed, the balance of power on the ground shifts, opening up opportunities for some contending groups and leaving others vulnerable. Opposition groups that may have been biding their time can also find fresh opportunities in the wake of evacuation, as the Taliban did in Afghanistan.

Both of these logics were evident in the strategy the Nixon administration adopted in Vietnam: The United States would withdraw from South Vietnam in such a way that a “decent interval” would be guaranteed before things fell apart, as National Security Adviser Henry Kissinger famously put it. In June 1972, he told Chinese premier Zhou Enlai, “While we cannot bring a communist government to power . . . if, as a result of historical evolution it should happen over a period of time, we ought to be able to accept it.” Ultimately, the accords were signed too late to aid in Richard Nixon’s reelection—he still swamped Senator George McGovern in their 1972 contest—but they allowed him to claim he had achieved “peace with honor.” It was left to others to cope with the hollowness of those words.

Face-saving requires a sleight of hand. Sometimes a clever leader can buy a bit of time in domestic politics, but that kind of magic is harder to pull off on the international stage. In the end, the truth will out. While a prolonged exit may generate an immediate perception that a withdrawal has been carefully planned and conducted on the intervening power’s terms, the long-term attempt to convince others that a defeat is actually a victory is not likely to succeed.

Lesson 3: When in doubt, leave.

The only good reason to prolong a struggling military intervention is if a change in strategy can bring victory. Of course, few politicians believe their strategy is going to fail, but it is very rare for leaders who lack the ability to be honest with themselves to succeed. There are no sure things. Any strategy is going to be highly debatable, as the Bush administration’s “surge” strategy was in 2007. The point is that leaders must evaluate strategies based on whether they are likely to produce a successful outcome, not just whether they may yield some political advantage.

If a plausible strategy for victory is unavailable, it is best to withdraw sooner rather than later. This reality is often difficult for leaders to recognize. They fear the domestic political costs of leaving without victory, but prolonging matters promises to increase human and material costs without concomitant gains. Some 20,000 American lives were lost in Vietnam as the Nixon administration searched for a way out after 1970.

There are reasonable things an intervening power can and should do when it is leaving to reduce the chances of a catastrophic outcome. It can press other powers in the region to refrain from meddling in the power vacuum created by its departure, and it can appeal to the United Nations and other institutions to step in to protect vulnerable minorities and preserve the peace. But the fact is that such institutions are rarely eager to clean up the mess left by an unsuccessful intervention, and the record when they try to do so is not encouraging. The UN forces that took over from departing Americans in Somalia in 1994 and Haiti in 1995, for example, were helpless to prevent worsening violence and political chaos. Somalia today is an anarchic haven for pirates, and Haiti, despite the presence of 9,000 UN troops, is plagued by political instability and violence.

If a plausible strategy for victory is unavailable, it is best to withdraw sooner rather than later. This reality is often difficult for leaders to recognize. They fear the domestic political costs of leaving without victory, but prolonging matters promises to increase human and material costs without concomitant gains.

Leaving sooner rather than later may be the best available option, but that is not to say it will produce a good outcome. The damage caused by a misguided intervention is not easily repaired, and it is usually layered on top of nasty preexisting conditions. The notion that the people of an occupied country will get their political act together once the intervening power threatens to leave is based more on hope than experience. Political instability is often what draws in an outside power in the first place. In many cases, the underlying tensions can only be resolved over the long haul, after considerable bloodshed.

Lesson 4: Beware the domestic politics of exit.

The key motivating forces behind exit strategies are as often domestic political considerations as they are military or diplomatic ones. Consider the end of Israel’s 1982–85 intervention in Lebanon, marred from the beginning by the 1982 massacres at Sabra and Shatila, which saw more than a thousand Palestinian and Lebanese civilians killed by Phalangist militia as Israeli troops stood by. The primary impetus to withdraw came from Israeli political groups that were critical of what they saw as their government’s lack of strategic clarity in Lebanon and unhappy about the human and moral costs of the occupation.

Two kinds of domestic political pressure usually appear, pushing in opposite directions. One kind of pressure comes in the form of demands to cut costs and reduce the human suffering caused by an intervention, reflected in highly visible poll numbers and protest marches. The other is less obvious and thus less discussed by analysts: the imperative to avoid appearing to be the author of failure. Leaders face a strong temptation to “kick the can down the road” and pass on the dilemma of how to get out to a subsequent government.

In the end, it is impossible to disentangle the domestic politics of exit from strategic decisions, but good leaders put more emphasis on the likelihood of military victory than the likelihood of their own political survival. Concerns about the domestic political consequences of withdrawal have more often pushed leaders to drag out interventions than to end them quickly.

Lesson 5: Concerns about how to get out shouldn’t preclude intervention, but they must be a consideration.

Public demands for an exit strategy are often simply camouflage for a particular speaker’s opposition to the intervention. When President Bill Clinton announced plans to send 20,000 U.S. troops to serve as peacekeepers in Bosnia in 1995, for example, Senate Republican leader Robert Dole quipped, “If Bill Clinton is going to have the entry strategy, the rest of us should have the exit strategy.”

It is usually impossible to fix the terms of exit in advance; they must be shaped by conditions on the ground. In any case, an ex ante exit strategy is not likely to be credible, since battlefield realities will inevitably require strategic adjustments.

Requiring a solid advance plan for getting out would amount to precluding almost any intervention. Yet there are cases in which intervention makes sense, including some humanitarian ones. It is usually impossible to fix the terms of exit in advance; they must be shaped by conditions on the ground. In any case, an ex ante exit strategy is not likely to be credible, since battlefield realities will inevitably require strategic adjustments.

That is not to say that political leaders do not need to consider scenarios for withdrawal before taking the plunge. It may be politically impossible to publicly contemplate defeat, but good strategists plan for the possibility. Leaders should be expected to establish the conditions under which they would consider disengagement, and they should have contingency plans in place should a speedy pullout be necessary. The time for a leader to think about the cost of losing face in a hasty retreat is emphatically before the fighting starts.

The implications of the five lessons can be applied across a wide range of cases, but especially in Iraq and Afghanistan. Barack Obama promised as a presidential candidate to withdraw U.S. troops from Iraq, and under the plan he has adopted, U.S. forces have already moved out of Iraqi city centers and are scheduled to leave the country entirely by the end of 2011. The framework for the withdrawal, the U.S.-Iraqi Status of Forces Agreement (SOFA), was actually negotiated by the Bush administration in 2008, at a moment of optimism about Iraq. The level of violence had fallen, and though progress on the political front was halting, there was hope that sound political and security institutions would be in place before the United States left the country.

In pushing his exit strategy, President Obama seems to have avoided most of the pitfalls that have tripped up other leaders. The SOFA and the Obama administration’s subsequent withdrawal plan appear to reflect a genuine hope for a positive outcome in Iraq rather than an effort simply to save face. The real test of this strategy, however, will come if the situation in Iraq deteriorates. Will the United States remain committed to its plan or will it prolong its intervention? If the Obama administration extends the U.S. presence, will its decision be driven by sound strategic logic or by a desire to save face? The history of interventions suggests that a turn for the worse in Iraq should be met with a continued determination to pull out. The United States should do whatever it can to minimize the risk of catastrophic events in Iraq as its troops depart, but if the situation turns sour, its speedy withdrawal, however bloody, could very well be the best option for both the United States and Iraq.

In Afghanistan, the United States does not yet have an exit strategy. It remains committed to that intervention, with Obama calling in February for an immediate increase of 17,000 troops on the ground, bringing the total to 62,000. And military commanders may soon ask for more reinforcements. If the increase does not generate signs of progress toward stability, then it will likely be in everybody’s best interest for the United States to ratchet down its goals, withdraw its troops, and reorient its military effort to very specific counterterrorism missions designed to contain future threats that may germinate on Afghan soil.

There is no easy way out of an intervention gone awry. There are no clever exit strategies that can convert failure into success, and not even the most detailed agreements can effectively alter others’ perceptions of an effort that simply has not accomplished its goals. The lesson of past military interventions is that it is almost always better to leave sooner rather than later.

It is time to recognize that history does supply useful exit lessons. In January 1973, Henry Kissinger discussed the Paris Peace Accords with his assistant John Negroponte (U.S. ambassador to Iraq from 2004 to ’05) and John Ehrlichman, a top aide to President Nixon. “How long do you figure the South Vietnamese can survive under this agreement?” Ehrlichman asked. “I think that if we’re lucky they can hold out for a year and a half,” Kissinger replied. Negroponte asked whether such an agreement was in the best interest of South Vietnam. Kissinger’s response: “Do you want us to stay there forever?”

That question is familiar to students of military intervention. There is no easy way out of an intervention gone awry. There are no clever exit strategies that can convert failure into success, and not even the most detailed agreements can effectively alter others’ perceptions of an effort that simply has not accomplished its goals. The lesson of past military interventions is that it is almost always better to leave sooner rather than later. Leaving is not likely to produce a pretty result, but international politics often necessitates choosing between bad options. When things go wrong, door number one—the exit door—is likely to be the least bad option.

* * *

David M. Edelstein is an assistant professor in the Edmund A. Walsh School of Foreign Service, the Security Studies Program, and the Department of Government at Georgetown University, and the author of Occupational Hazards: Success and Failure in Military Occupation (2008). This article is drawn from his research on exit strategies as a fellow at the Woodrow Wilson Center during 2008-2009.

Photo courtesy of The U.S. Marine Corps