IN AFGHANISTAN, the problem is doubled. Sparse funds must be extracted from Kabul and restricted from the poppy trade, while not expunging the poppy fields without a substitute crop. Poppy cultivation has decreased by 20 percent since the 2007 record crop, and eighteen of the thirty-four provinces are poppy-free. But 50 percent of the world's opium is grown in Helmand Province in southeast Afghanistan, where the Taliban taxes the farmers an estimated $70 million. Poppy fields account for an estimated five hundred thousand acres, with a yield of eight hundred tons. In a country with an unemployment rate of 40 percent, the heroin/opium industry provides jobs for 3 million workers and produces $4 billion in total revenues, about 40 percent of the gross national product.

In January of 2006, CENTCOM declared the drug trade to be "the number one threat" to Afghanistan's democracy and freedom. Improving governance requires firing or arresting corrupt officials. The opium trade has corrupted all segments of Afghani leadership-from tribal and military rulers to individual warlords. Afghanistan's President Karzai, however, has strictly limited NATO and U.S. efforts to investigate and arrest drug overlords because the web of corruption entangles officeholders at all levels.

This is, of course, nothing new. In fact, in many countries, corruption-theft and bribery-has lubricated a functioning government and insured a stable order. The leaders of most Arab oil-exporting states dispense sufficient treasure to motivate a loyal bureaucracy, including the security forces, to perform their jobs. In the 1860s, the British in India granted local authority and property rights to landlords called zamindars. Permitted to tax and thus benefit, the zamindar bureaucracy ruthlessly stamped out anti-British movements because a rebellion meant the end of their livelihoods. Corruption and bribery actually ensured stability.

The trouble with Afghanistan is that corruption is having the opposite effect. What happened under Chiang Kai-shek in China in 1949 is similar to the goings-on in Afghanistan today. Chiang's government is cited in counterinsurgency theory as an example of how corruption corroded a regime, leading to victory by the guerrillas. It wasn't the corruption itself that destroyed the ruling class, but rather the actions of Chiang's subordinates. As Nobel Prize-winning economist Roger Myerson argued, "The problem was that highly connected government agents took profits from their positions without providing the governance and services that were expected of them." So subordinates were rewarded when they performed their jobs correctly-and when they failed. Those failures then became the insurgents' gains, spreading the perception that Chiang's rule was doomed. The subordinates began to steal more and more before the window of opportunity closed. They did nothing to prevent the collapse they had engineered by their own ineptitude.

In Afghanistan, the warlords and the provincial officials profiting from the opium trade are working against the state by eroding its legitimacy. CENTCOM is correct in identifying the effects of the opium cabals, but has not advanced a strategy to counter them. The United States has to imprison insurgents and drug dealers if the Afghan government will not do so. This is the hardest part. Currently, there is no consensus inside NATO for prosecuting a serious strategy intended to destroy the opium trade. Afghan officials at all levels are dirty and have no moral credibility among the people. If American units, living alongside Afghan soldiers and local tribal gangs, are ordered to straighten out a corrupt mess, they can do it in a few years-if they have the power to arrest and to imprison. In essence, American units would be working against the interests of many in the Afghan government, which parallels the ironic situation of the U.S. military in Iraq.

AT THE end of the day, we have to face up to the reality of what it would take to achieve our goals, and to wrestle with our limitations. Afghanistan may be a war against the Taliban and al-Qaeda, but it is also a battle for the security, safety and prosperity of the Afghani people. This is about tracking down an enemy while building a society. And choices must be made. Is the goal of the U.S. military to shore up security along the eastern border, or to rebuild the entire nation? Can the former be done without the latter?

In 2001, the CIA and the Special Operations Command rented warlords to attack the Taliban. After al-Qaeda fled, America's goals changed radically. Five years later, the United States government was hatching ploys to manipulate those warlords into disarming, forsaking the profits of the drug trade and yielding power to officials in Kabul whose backgrounds were as shady as those of the warlords.

America's goal escalated into building a democratic, prosperous nation. In 2007, the NATO commander in Afghanistan explained that NATO's International Security Assistance Force facilitated reconstruction so the "Afghan people might enjoy self-determination, education, health, and the peaceful realization of their hopes and dreams." Along that same line, Admiral Mullen also said in September of 2008 that Afghanistan needs highways, electricity, commerce, alternative crops to poppy, foreign investment, reliable provincial governors and a justice system based on the rule of law. "These are the keys to success in Afghanistan," he said. "We can't kill our way to victory."

It's strange to hear a military commander during war saying he can't kill his way to victory. Our police don't tell us they can't catch all the criminals. We expect our police to provide security even in impoverished areas. Similarly, we expect our military to destroy al-Qaeda by killing its members. The American military mission is not nation building. If poverty and poor government were the causes of insurgencies, most of the countries in the world would be at war. To prevent more recruits for the Islamic extremists, we'd like to have a tolerant democracy and a thriving economy in Pakistan and in Afghanistan. But if we make that a precondition for crushing al-Qaeda, we will be in Afghanistan for decades.

Only a few years ago, the Bush administration vigorously rejected the idea of involving the Pentagon in nation building. With a 20 percent literacy rate among men and 7 percent among women, Afghanistan faces staggering obstacles to development, even without an insurgency. But the United States sank over $20 billion into development projects in Iraq that disappeared. It's not at all proven that the United States can build another sovereign nation in our image. The broad goals Admiral Mullen outlined might be achieved in twenty or thirty years, if a coherent development strategy were devised and funded at several billions of dollars a year-that is, after security was established.

But nation building and economic development are distinct from the fundamental American mission of preventing the resurgence of al-Qaeda or its extremist affiliates. Our military strategy should treat development funds as a tool in achieving the mission of area security. Provincial Reconstruction Teams embedded in each American brigade and led by the brigade commander add value because they focus their efforts on local projects that give the local people an incentive to align with the American and Afghan security forces in the area. We should trim back our goals and our rhetoric about the end-state in Afghanistan. The U.S. military should focus on security, with economic development as a useful tool but not a mission.

THERE ARE proven methods from Iraq to establish local security, cultivate local leaders and hand off control to local Afghan security forces. It won't be quick and easy; we're talking years of effort by tens of thousands of American soldiers. But the techniques exist. Still, we need to understand that doing so is simply playing defense. It does not directly place pressure on al-Qaeda inside Pakistan.

President Bush failed to lead and to hold the United States together during the Iraq War. Because the war inside Afghanistan, plus raids across the border into Pakistan, will go on for years, the major challenge for the new administration is to win the steady support of the public and both parties in Congress. To avoid divisiveness at home, Afghanistan cannot be handed off to the generals and ambassadors. The political pressures going forward to reduce the defense budget will be intense, requiring the president to lobby constantly for the resources to fight the war. He must remain involved daily as the commander in chief of a nation at war, because the "good war" is going to get a lot bigger.

Bing West is a former assistant secretary of defense for international security affairs and combat marine. His third book on the Iraq War is The Strongest Tribe: War, Politics and the Endgame in Iraq (Random House, 2008).