Don't rush out of Afghanistan

With Afghan President Hamid Karzai and his top advisers visiting Washington this week, huge questions about the future of the NATO mission there consume Afghan and American minds. How fast can we draw down our current total of 68,000 U.S. troops (and another 30,000 or so from other outside countries) before the mission formally concludes at the end of next year? And how many forces do we have to keep in Afghanistan afterward? These questions come on top of other decisions we have been making lately, about the long-term size of the Afghan army and police and about foreign aid levels the international community will provide to Afghanistan for many years.

We should slow this process down. It is important that Afghans — and other interested parties in places like Pakistan — see evidence of a long-term partnership between the outside world and our Afghan friends. That much is true. So some clarity about long-term plans is useful. But we cannot and should not try to answer all questions now.


Making specific, long-term plans is unrealistic for two main reasons. First, we cannot foresee battlefield conditions in Afghanistan in 2015 and beyond. Second, we cannot foresee who will win the Afghan presidential elections next year and what kind of partner the new president will become for the international community.

This rush to decide everything prematurely has already had some unfortunate consequences. In our efforts to be sure that future Afghan security forces get enough outside aid after 2014 to function effectively, we have already begun to assume that they must downsize by one-third — after we just spent half a decade building them up. Moreover, NATO will have pulled out most of its own remaining forces by that point, making it even harder for Afghans to also scale back drastically. It is in fact doubtful that such downsizing of the Afghan army and police should occur so soon.

Another example is what happened in Tokyo in July. International donors promised large-scale aid to Afghanistan for four years. At one level, this has some very beneficial effects, allowing long-term planning. But it also creates the bizarre situation in which, even if Afghans elect a poor leader in 2014, we have promised in advance that Washington will give Afghanistan more foreign aid than any other country receives from the United States for several more years.

In fact, for all his flaws and the corruption of many of those around him, Hamid Karzai has his strengths and is far from the worst Afghan leader we can imagine. He is not believed to be personally corrupt. He does not run his own death squads. He is the legitimate, popular and democratically elected leader of his country (despite the fraud that occurred in the 2009 presidential elections, which did not change the outcome). He respects the constitution and intends to step down from office in 2014. And Mr. Karzai, a Pashtun, has tried to unite the country ethnically. His top peace negotiators with the Taliban have been Tajik. In various periods his foreign and interior ministers have had Tajik roots, his two current vice presidents are Tajik and Hazara, and his minister of mines is Uzbek.

If the next leader of Afghanistan has less popularity, integrity or respect for the constitution than Mr. Karzai, we could be in trouble. No aid package promised now could be expected to be honored in full if a warlord or blatantly corrupt leader became president of the country next year. Again, long-term plans are OK as long as they are kept flexible and notional.

President Barack Obama, war commander Gen. John Allen and other administration officials are currently trying to decide how many American forces should stay in Afghanistan starting in 2015. Independent analysts like retired Lt. Gen. Jim Dubik have argued for 30,000 GIs. Former war commander Lt. Gen. Dave Barno has argued for fewer than half that number; press reports suggest that some administration officials favor a number in the range of perhaps 5,000 to 8,000.

Of course, these disputes do not come out of thin air. Much of the disagreement is over strategy. Should our remaining forces just carry out commando raids and provide limited training for Afghans after 2014? Or should they also help with air power, logistics, intelligence, mentoring in the field, and even some limited ground combat capability in one or two parts of the country?

This debate rapidly becomes very ideological and very political. Those most tired of the war lean toward minimalist options. Those most worried about future possible threats to U.S. security emanating from Afghanistan tend to lean toward the larger forces. And the desire to design the "permanent" U.S. force now, so that we can establish it within two short years, adds a lot of pressure and tension to the debate.

Let's take the pressure off. We should aspire to a small U.S. presence in Afghanistan of no more than 10,000 troops — but be in no hurry to move to that number immediately. Afghanistan's military is still developing, and it will not be completely built by the end of next year. For the period of 2015 to 2017 or so, it will need help with a broader range of tasks. I would suggest a U.S. force of around 20,000 to 25,000 troopers for 2015, phasing down toward about 10,000 by 2018 — subject to conditions on the ground.

There is a surreality about long-term scheduling of big decisions and milestones in war, which after all is a competitive process in which the enemy's actions and the behavior of one's own allies have great impact on the future course of events. We do not have a crystal ball for Afghanistan; let's stop pretending we do. And let's let the Karzai visit focus more narrowly on discussion of the next key steps for this year and next, rather than trying to solve everything all at once.

Michael O'Hanlon, a Maryland resident and a senior fellow at the Brookings Institution, is coauthor with Martin Indyk and Kenneth Lieberthal of "Bending History: Barack Obama's Foreign Policy," and coauthor with Hassina Sherjan of Toughing It Out in Afghanistan. He can be contacted at