President Obama wants to postpone a full-scale debate over Afghan policy until Congress passes health-care reform.

Unfortunately for Obama, Afghanistan won’t wait on health care. A debate on the direction of his Afghan policy is already brewing in Congress _ especially over whether to send more troops.

Yet this debate is being conducted in a vacuum. True, the president has laid out his aim: to disrupt, dismantle, and defeat al-Qaeda and other extremists in Pakistan and Afghanistan and prevent their return to either country.

The U.S. commander in Afghanistan, Gen. Stanley McChrystal, has submitted his (still classified) assessment.

Administration officials have briefed Congress on a list of 50 “metrics” by which to judge whether the policy is succeeding.

But the crucial element in the debate is missing: The president hasn’t resolved the dispute among advisers and within his party over his Afghan strategy. Until that strategy is set, the arguments over troop levels cannot be settled.

The strategy dispute basically revolves around three approaches. The minimalists want to use drones, missiles, and special-ops forces to go after the Afghan Taliban _ while decreasing U.S. troop levels. This is the classic counterterrorism model.

Unfortunately, it won’t work. Gen. David Petraeus has pointed out that we tried launching cruise missiles at al-Qaeda targets in Afghanistan in the 1990s and failed to eliminate the danger.

Were we to quit Afghanistan, and the Taliban to retake control, we would lack the intelligence and infrastructure on the ground to find jihadi targets _ or to effectively target terrorists inside Pakistan.

Moreover, as Ryan Crocker, the former ambassador to Iraq and Pakistan, noted in congressional testimony last week, McChrystal is a former special-operations commander.

“If he thought it could be done that way,” Crocker said, “I think we’d be seeing different sets of recommendations.”

The second approach is one on which most everyone agrees _ in principle: expanding and accelerating the training of Afghan security forces to replace NATO troops, while trying harder to woo some Taliban to stop fighting.

Sen. Carl Levin, the Michigan Democrat who is chairman of the Armed Services Committee, says these steps should precede any increase in U.S. troop numbers. This sequencing is likely to please many congressional Democrats.

Yet the Levin approach won’t work either _ at least until the security situation eases in Afghanistan.

To understand why, look back at the U.S. effort to train Iraqi security forces.

The numbers increased too quickly; when violence soared, the inexperienced Iraqi army nearly crumbled. It took an enormous U.S. effort to help Iraqi forces regain their footing.

Similarly, if the new Afghan army is thrust too quickly into major combat roles while Afghan violence is increasing, it will crumble. “I think we and the Afghans have to be careful not to put more of a burden on these developing forces than they can bear at this time,” Crocker said.

In other words, Afghan forces are far from capable of bearing the brunt of the fighting. A decision on U.S. strategy and force levels cannot wait until they are fully trained.Nor can that decision be postponed while pursuing a strategy of reconciliation with Taliban forces.

Such reconciliation efforts are key, but U.S. commanders believe there is little chance of major breakthroughs while the Taliban think they are winning.

Only when NATO forces, along with Afghans, regain the initiative, are more Taliban (and tribal leaders) likely to shift to the winning side.

Which brings us to the third approach: the classic counterinsurgency strategy favored by Petraeus and McChrystal. This one aims to protect and support parts of the Afghan population by clearing out Taliban and funneling in economic aid. The goal is to buy time to train Afghan troops, woo midlevel Taliban, and stabilize Pakistan.

This strategy would require more troops. But McChrystal has reportedly been told by the administration to delay a request for higher numbers. Any McChrystal testimony to Congress is on hold.

Were he to testify, he would face many questions about an Afghan counterinsurgency strategy. Skeptics such as Vice President Biden say they believe it will distract from stabilizing Pakistan. (Knowledgeable Pakistanis such as the top expert on the Taliban, Ahmed Rashid, tell me the opposite:

If we don’t curb the Afghan Taliban, the blowback will reinvigorate the jihadis inside Pakistan.)

Others worry about a weak, corrupt, possibly illegitimate government in Kabul, after disputed elections. They ask about the risk that Afghans will reject us as an occupying force if we increase our footprint.

My guess is that McChrystal and his civilian counterparts could give Congress solid answers. No doubt they would say U.S. military and civilian experts can work with local leaders and fund local aid projects, even if Kabul politics are unsettled.

They would add that Afghans, who still overwhelmingly dislike the Taliban, according to polls, will judge U.S. forces less by their numbers than by their behavior, and what they deliver. Under McChrystal’s new rules of engagement, Afghan civilian casualties have dropped.

But McChrystal cannot get out ahead of the commander in chief. He cannot make the case for more resources before Obama sets a strategy for him to implement.

So the debate over Afghanistan in Washington has a strange quality _ going in circles without an essential framework. It’s easy to understand why Obama wants to delay tough Afghan policy decisions. But he won’t be able to do so for long. 

• Trudy Rubin