The South Asia Channel

Withdrawal without winning?

By Robert Jervis

Most discussion about Afghanistan has concentrated on whether and how we can defeat the Taliban. Less attention has been paid to the probable consequences of a withdrawal without winning, an option toward which I incline. What is most striking is not that what I take to be the majority view is wrong, but that it has not been adequately defended. This is especially important because the U.S.has embarked on a war that will require great effort with prospects that are uncertain at best. Furthermore, it appears that Obama's commitment to Afghanistan was less the product of careful analysis than of the political need to find a "tough" pair to his attacks on the war in Iraq during the presidential campaign. It similarly appears that in the months since his election he has devoted much more attention to how to wage the war than to whether we need to wage it.

The claim that this is a "necessary war" invokes two main claims and one subsidiary one. The strongest argument is that we have to fight them there so that we don't have to fight them here. The fact that Bush said this about Iraq does not make it wrong, and as in Iraq, it matters what we mean by "them." Presumably if we withdrew the Taliban would take over much of southern and eastern Afghanistan. This would be terrible for the inhabitants, but would it harm us? I don't think anyone believes that the Taliban would launch attacks against us or our allies, so that the menace is not a direct one.

Instead, the fear is of a repetition of the pre-2001 situation in which al Qaeda would have bases that would facilitate attacks. Obviously, this is a danger, but how great a one? The Taliban would not want to repeat what happened after 2001, and so I do not think one can simply assume that Taliban control would automatically lead to al Qaeda control. Nor is Afghanistan the only country that might permit an al Qaeda presence. Somalia is perhaps as troublesome, and yet noone calls for the U.S. to re-intervene there. Furthermore, al Qaeda has some sort of base ofoperations in Pakistan now (and is not likely to lose it even in the best outcome across the border); how much worse would it be if we withdrew?

In part, this might depend on exactly what "withdrawal" means and on what "bases" mean. Clearly al Qaeda grew by having large training camps in Afghanistan before 9/11, more than they have now in Pakistan and which they might be able to reestablish, presumably on a smaller scale, if we left. But are these still needed? The fact (assuming it is a fact) that 9/11 could not have happened without those camps does not mean that their reestablishment wouldlead to renewed terrorism. To put this another way -- and this is a genuine question and not a rhetorical one -- what sort of facilities might be established in Afghanistan that would increase the danger to the U.S.? It presumably would be easier for al Qaeda to operate, but would this translate into more and more effective attacks?

The second part of the question is exactly what withdrawal means. What would we keep in the region? What could we achieve by airpower? How much intelligence would we lose, and are there ways to minimize this loss? It is often said that we withdrew before 9/11 and it didn't work. True, but the circumstances have changed so much that I don't find this history dispositive. While al Qaeda resurgence is a real danger, I am struck by the thinness of theargument that in order to combat it we have to fight the Taliban and try tobring peace if not democracy to Afghanistan.

A second argument, made most recently by Frederick Kagan inthe September 5-6 Wall Street Journal, is that, to quote from its headline, "A stable Pakistan needs a stable Afghanistan." But does it really? Are there reasonable prospects for a stable Afghanistan over the next decade no matter what we do? Isn't there a good argument that part of the problem in Pakistan stems from our continued presence in Afghanistan? We are told that bases in Pakistan are used to support the insurgents in Afghanistan, while simultaneously being told that it is the fighting in Afghanistan that is endangering Pakistan.

Reciprocal causation is certainly possible, but this modern version of the turbulent frontier doctrine is not backed by solid logic. Pakistan's ISI and army clearly maintain ties to the Taliban in Afghanistan, and although they cannot exert anything like complete control, once the danger of a Taliban defeat by the U.S. passes they would have every incentive to reign in their clients.

Furthermore, the stability of Pakistan does not depend on pacifying the tribal areas. While the recent efforts by Pakistan to regain control of some of its territory may owe something to our combating the Taliban, I wonder if the effect is a large one. In parallel, it can be argued that we gain general influence over Pakistan by fighting in Afghanistan, but here not only the magnitude of the effect but its sign is open to question.

A third but subsidiary argument is that withdrawal would undermine American credibility around the world. Again, the fact that this is an echo of Vietnam does not make it wrong, but it does seem to me much less plausible than theother arguments. Who exactly is going to lose faith in us, and what are theygoing to do differently? Much could depend on the course of events in othercountries, especially Iraq, which could yet descend into civil war. But if it does, would American appear more resolute -- and wiser -- for fighting in Afghanistan? Of course if we withdraw and then we or our allies suffer a major terrorist attack many people will blame Obama, and this is a political argument that must weigh more heavily with the White House than it does with policy analysts.

It is worth noting that these issues are much less ideologically-charged than those surrounding the war in Iraq (or in Vietnam). This means that it shouldbe easier for the concerned community to address them seriously, although notnecessarily to come up with (correct) answers, and for people to change their minds. This makes it particularly unfortunate that we have not had a searching and thorough discussion. Although some deeply-rooted beliefs are involved, such as those involving the propensity for dominoes to fall and perhaps an estimate of how great a danger terrorism is, we are mostly in a more pragmatic realm.

Of course Yogi Berra was right when he said that prediction is difficult, especially about the future. But once we move beyond the alluring but unsustainable claim that our inability to exclude the possibility that withdrawing would be very harmful means that we must fight, it becomes clear that we are building a large and risky war on predictions that call for closer examination.

Robert Jervis is the Adlai E. Stevenson Professor of International Affairs at Columbia University.

MANAN VATSYAYANA/AFP/Getty Images