President Obama faces tough decisions on Afghanistan, but his party is on the hot seat too. Afghanistan is the first test since Vietnam of Democrats' collective ability to manage a major armed conflict. Just how to do that is the subject of an intense internal debate. However it is resolved, the party must avoid a convulsive split that would cast doubt on its ability to defend the country.

Having declared Afghanistan a "war of necessity," Obama last March said his policy is to prevent al Qaeda from attacking the United States and to keep the Taliban from returning to power. According to his handpicked commander, General Stanley McChrystal, achieving those goals will require an Iraq-style counterinsurgency campaign and up to 40,000 additional U.S. troops.

McChrystal's request and Afghanistan's rigged Presidential election triggered another White House review, leading many to wonder if Obama is having second thoughts. The President's military advisers have endorsed McChrystal; Vice President Joe Biden and other White House officials are said to advocate a counterterrorism strategy with a smaller military footprint.

The administration's deliberations proceed amid eroding public support for the war. A new Washington Post poll found that Americans increasingly doubt that Obama has a clear strategy for Afghanistan. And while the public narrowly opposes sending more troops, fully 61 percent of Democrats say it's a bad idea. Skepticism is just as pervasive among Congressional Democrats. If he chooses to send more troops, Obama could find himself in the awkward position of waging war with more backing from Republicans than from his own party.

So what should Obama and his party do? None of the strategy options gleam with promise, and all entail big risks.

On balance, though, we believe a counterinsurgency campaign stands the best chance of achieving the President's strategic goals. More drone-fired missile attacks and Special Forces raids will kill more militant leaders, but they won't check Taliban advances in the south and east, hold territory or shield Pashtun civilians from intimidation. While a counterterrorism approach would likely limit U.S. casualties in the short term, it risks abandoning large swaths of Afghanistan to the Taliban. That would surely boost the insurgency's momentum and underscore the central government's impotence - even as the Pakistani army on the other side of the border has finally attacked insurgent havens in South Waziristan.

It may seem counterintuitive, but a surge of U.S. troops now could allow us to hand off the war to Afghans sooner rather than later. It would buy time to build up more capable Afghan security forces even as it keeps the insurgency from getting too big for those forces to handle.

Some in the counterterrorism camp have questioned whether a Taliban restoration would really lead to al Qaeda's return to Afghanistan. This seems a risky proposition: Why would al Qaeda leaders continue to skulk underground in Pakistan when they could live again as honored guests of Mullah Omar and other leaders of the Afghan Taliban? As Peter Bergen has pointed out, al Qaeda and the Taliban have fused ideologically, and the Afghan-Pakistani borderlands are both their traditional refuge and their irreplaceable operational base. What's more, the Taliban's return to power, would be hailed by Islamists everywhere as a second miraculous victory of the faithful over an infidel superpower.

Concentrating narrowly on al Qaeda overlooks the fact that the Taliban are our enemy too. Americans are under no illusions on this score: According to the Pew Research Center, 76 percent view a Taliban restoration as a "major threat" to American security; while the Post poll found that four of five Americans say we should prevent the Taliban from regaining power, even if some elements do not support terrorism against the United States. This probably explains why the public, despite its qualms about escalation, is not clamoring for a prompt U.S. withdrawal from Afghanistan.

We back the counterinsurgency option not because we're highly confident of its success, but because stabilizing Afghanistan is more consistent with America's long-term security interests and our responsibilities to the Afghan people, who overwhelmingly reject the Taliban. But we acknowledge that it could cost more money and lives, take more time, and still not succeed if the Afghan government fails to consolidate its authority, reduce corruption and deliver basic services.

Whatever course he chooses, the President will need his party's understanding and support to succeed. If Democrats fall out over Afghanistan, he won't be able to sustain a coherent policy, and the public will likely lose confidence in the party's ability to manage the nation's security.

This would be a calamity for the country. And it could throw a lifeline to the Republicans, who would have a field day resurrecting the old bugaboo of Democrats as "soft on defense." Only recently have Democrats shed that image. A Democracy Corps poll found that the 40 year national security confidence gap that has dogged Democrats has substantially closed. But some of the old doubts are creeping back: A September Gallup poll found that the public still prefers Republicans for fighting terrorism by a 49-42 margin.

Let's be clear: We're not arguing that Obama should make his decision based on a desire to "look tough" on national security. We're saying Democrats ought to think long and hard before forsaking a war that Obama has defined, both during the campaign and as president, as necessary to Americans' security.

President Obama and his party have a once-in-a-generation opportunity to consolidate a new progressive governing majority. But to do that, they must govern effectively. In today's terror-struck world, that means prosecuting America's fight against Islamist extremism with stamina and resolve. For his party no less than for President Obama, Afghanistan will be the acid test.