The argument within the Obama administration for a big troop withdrawal from Afghanistan over the next year goes roughly like this: We've killed Osama bin Laden. That means we've achieved the core goal for which we sent forces in 2001. We have a ticket out, and we should take it.

The counter-argument from administration hawks is that a quick departure is a guarantee of failure. It risks repeating the mistake made in the 1980s when, after pumping money and guns into Afghanistan to defeat the Soviet Union, America walked away -- creating a power vacuum that was exploited by warlords and their regional patrons.

Somewhere in the middle is President Obama. "We will begin a transition this summer," he said Monday. "By killing bin Laden, by blunting the momentum of the Taliban, we have now accomplished a lot of what we set out to accomplish 10 years ago."

Inevitably, this debate is partly a numbers game: The rapid-withdrawal advocates want a timetable for removing all 30,000 of the "surge" troops Obama decided to send in December 2009. The "stay the course" proponents want a modest reduction of 3,000 to 5,000 troops, which is all they think conditions allow. A "split the difference" caucus argues for a cutback that hits five figures -- around 10,000.

The problem with all these arguments is that they lack a clear strategic rationale. Do the "stay the course" proponents really expect that the Afghan army will be strong enough to stand on its own by 2014? That strikes many analysts as a dubious proposition.

Are the speedy-withdrawal advocates really comfortable with an Afghanistan that could quickly return to the pattern of the 1990s, with the regional powers, India and Pakistan, each manipulating their favorite ethnic groups and warlords? That sounds like a recipe for perpetual instability in South Asia.

The point is obvious: The troop level should be a function of the strategic plan, not the other way around. The three variables that U.S. policymakers have been discussing -- troop withdrawal, reconciliation with the Taliban and drone attacks in Pakistan -- are interrelated.

Let's take the question of political reconciliation: If Obama announces a big troop withdrawal, will this encourage Taliban concessions? Probably not, but if you think negotiation may work, then you want military leverage that enhances it.

The administration has begun secret talks with Taliban intermediaries. If this process is serious, it needs to move toward the practical test of a cease-fire, perhaps initially in one locale: The U.S. needs to show the Taliban there's a way to ease the pain through negotiations, and Taliban representatives need to show they can deliver on the ground.

The same pragmatic test should be applied to Predator drone attacks. The drones have been an invaluable weapon against al-Qaida. But if the drone attacks cause such severe political problems for Pakistan that they prevent Islamabad from playing a constructive role in reconciliation, then the policy may need adjustment.

The strategic goal is a regional framework for a post-America Afghanistan. That means, in essence, a coordinated effort by Pakistan, India, Afghanistan and the United States to make a political settlement work. Obama's challenge in framing his troop-withdrawal announcement is to enhance this regional process, not undercut it.

"We're not going to do anything precipitous," Obama said Monday, in a statement that embodied his governing style. Hopefully, this means an exit strategy that actually provides a reliable exit.