No soldier wants to be the last one to die in a war, goes the old saw. In Afghanistan, make that: No coalition member wants to be the last one to quit. And so, competitive withdrawal has practically begun. According to President Obama, the United States will start drawing down next year. Canada is preparing to exit in 2011. Poland is talking complete withdrawal by 2012. Britain wants to end it all by 2015.

This is no way to win. It’s like a trainer announcing in round three that he will throw in the towel in round six. His man will take no more risks, while the other boxer will wait it out or go for the knockout punch.

Already, President Karzai seems to be preparing for our departure by catering to the Taliban. And, with every platoon that heads home, our wards will increase their bets on whoever they think will step into the vacuum—the Taliban, the warlords, the Iranians, the Pakistani ISI (the intelligence service that practically invented the Taliban in the days of Charlie Wilson and the Soviets).

This problem is built into any counterinsurgency strategy. Whose commitment is more sustainable, hence more credible: the insurgents’ or the intruders’? The “bad guys” fight where they live, and so, they have nowhere else to go. But we, who fight for abstract reasons only remotely related to our core security, can always ship out. The locals—be they wards or warriors of God—know it.

They know that democracies fight wars of choice only if victory is swift, bloodless, and reasonably priced. They don’t like operations that are indecisive, and this one has lasted even longer than the war in Vietnam. The asymmetry has become crueler over time. Thirty-five years after the fall of Saigon, postmodern Western society is horrified even by blood we shed on the other side. This our enemies have learned as well, hence the tactics of Hamas, Hezbollah, and the Taliban, which lure our forces into killing (either real or make-believe) civilians. Nothing has soured Germans more on the war in Afghanistan than the scores of Afghans killed near Kunduz in a German-ordered bombing run by U.S. aircraft late last year. How many were civilians? We’ll never know.

Democracies will fight as fiercely as totalitarians when their own lands are at stake. But they won’t fight to the end in a difficult war of choice, as Afghanistan surely is. Yet the willingness to stay as long as it takes is the alpha and omega of any counterinsurgency strategy. If we go in, we have to be willing to stay sine die. We must not think like a traditional army that knocks out the enemy and then goes home. We have to think like a police force. The police stay on the beat forever. Only then can they tell the good guys from the bad guys. Only then can they gain vital intelligence from the locals. Only because they reliably serve and protect can they conquer “hearts and minds.”

But why would we stay where interests (remote) and costs (high) are so unbalanced? There are lots of good reasons. Our interests may be abstract, but they are not unreal. The greater Middle East, from the Levant to Kashmir, will be in the twenty-first century what Europe was in the twentieth: the arena where endless vicious conflict—strategic as well as ideological, within as well as between states—will come home to haunt us if it remains unchecked. Indeed, the Taliban pale against a failing, deeply anti-Western nuclear state like Pakistan or a revolutionary regime like Iran’s that believes it is on a mission from God.

Here, then, is our conundrum: We must never set an exit date, as we did not in Kosovo. But, for the last decade, Western forces have stayed in Kosovo only because nobody dies, neither “them” nor us. American troops are still in Germany because there are no IEDs on the autobahns. The only way, therefore, for us to stay in Afghanistan is to go with our advantages and dodge our weaknesses as democracies, which recoil sooner or later from the sight of blood—theirs and ours.

Our advantages are technology and training: skilled soldiers, “eyes in the sky,” information processing, and standoff weapons ranging from drones to aircraft carriers and long-range bombers. Civil War General Nathan Bedford Forrest is (erroneously) credited with the counsel to “git thar fustest with the mostest” as a guarantee of military success. Today, the key is to “git thar fustest with the bestest”—be swift and precise. Keep enemies off balance, exploit surprise, rely on air- and space-borne intelligence, disrupt their command and logistics networks (yes, even irregulars have supply lines), immobilize them, keep them from massing, avoid “collateral damage.” Deny them sanctuaries and stay away from the population, which also means: Forget nation-building. There is no nation in Afghanistan.

No, you can’t “win” that way—in the sense of enshrining a preferred political order or routing the enemy for good. But you can constrain and deter your foes by maximizing their costs and minimizing yours. Best of all, a combination of watchful presence and nimble offensive can be sustained indefinitely. And indefinite the twenty-first century’s “Great Game” will be. The tactical payoff is the enemy’s growing conviction that we won’t go home. The strategic benefit is that he might eventually reconsider and start talking in earnest. That’s the best we can do, and it is better than throwing in the towel in round six. 

Josef Joffe is editor of Die Zeit and a senior fellow of the Institute for International Studies, and an Abramowitz Fellow at the Hoover Institution, at Stanford.

For more TNR, become a fan on Facebook and follow us on Twitter.