The Six Strategic Mistakes That Led to U, S, Defeat in Afghanistan

WASHINGTON—When the Afghanistan war was launched twenty-five years ago last month, there was no indication that it would become the longest conflict in U.S. history. The Taliban regime collapsed in two months, and a new government was in place in three. The light-footprint U.S. strategy in which special operations forces and Central Intelligence Agency (CIA) teams enabled local militias seemed revolutionary, and a smart counterpoint to the plodding, brutal Soviet war of the 1980s. But then, the U.S. war, too, plodded, through two decades of waxing and waning U.S. attention, until U.S. troops were finally flown out of Kabul’s international airport in 2021 in defeat.

Could it have turned out differently? Iraq, after all, had turned the corner in 2008, helped by a strategic decision by President George W. Bush to surge additional troops. But Afghanistan had—and continues to have—far weaker state institutions than Iraq.  When the United States invaded, it was the third, possibly fourth, chapter in a decades-long civil war that started after the communist coup of 1978—and arguably even earlier, after King Mohammad Zahir Shah was overthrown in 1973. Afghanistan may have been a less violent war than Iraq, but it was much more hopeless.

The allies’ decision at the Bonn conference in December 2001 and then at the June 2002 Emergency Loya Jirga, a grand assembly held in Kabul, to back Hamid Karzai instead of the former king encapsulated this mistake. A modern Pashtun leader, with ties to the mujahedin period and anti-Taliban legitimacy, might well become a threat to both militia leaders and Pashtun irredentists. After more power was centered in Karzai’s hands by the 2003-2004 constitutional Loya Jirga, that is exactly what happened. Karzai gained the ability to appoint provincial and district governors, and the national army and the police largely replaced the ethnic militias.

This arrangement was triply harmful. First, it hollowed out the core fighting strength of the existing government. Second, it elevated Karzai into a role where he sought to extend his authority over the regions—meaning over his former peers—rather than shape alliances with them. And third, by eschewing localism, he presented Afghans with a stark national choice between his government and the Taliban. The problem with Karzai was not Karzai himself, necessarily, but the highly imperfect Kabul supremacy that a Karzai-led government represented. Every bit of corruption, every bit of brutality, every mistake by the coalition forces could be shown to Afghans by the Taliban in an unflattering light, particularly as comparisons with Taliban-ruled areas emerged. The propaganda wrote itself.

Ironically, the choice of Karzai had been meant to preclude exactly this problem. The Northern Alliance saw the former king as a potential threat, since he had independent legitimacy. But the king, living in Italy, had few modern ties to Afghanistan and would have been by definition a figurehead only. That likely would have been the best option. Karzai seemed more malleable, more dependent, and, to the West, more democratic—until he wasn’t.

Without question, the Bush administration was distracted from Afghanistan by its war in Iraq. After initially being routed in the first few months, the Taliban began regrouping in the borderlands of Pakistan. By 2004 and 2005, attacks on Afghan security forces, Western nongovernmental organizations, and U.S. troops began to increase significantly. By 2006, a full-blown insurgency was underway.

Had the Bush administration been focused on Afghanistan, it almost certainly would have been more attuned to the signs of trouble. However, it is not clear that the United States would have significantly increased its force presence in Afghanistan at that point even if every available soldier had not been needed in Iraq. Washington was loath to repeat the mistakes of the Soviet Union, which had gone into Afghanistan heavy and alienated the populace, and the success of the initial U.S. campaign seemed to validate an unconventional mix of special operations forces, CIA paramilitary teams, airstrikes, and existing local allies.

But that force posture was impossible to square with the national political agenda the United States had set for itself of building a new Afghanistan. In the end, the United States would rely on only a handful of existing forces in country to build a new political and security order, rather than restoring the one that had existed before. And for that, the light footprint was not nearly enough.

More attention on Afghanistan during the critical early years might have reconciled this disparity and generated innovations sooner to help expand security. The Bush White House might have increased U.S. troop levels earlier, as security declined, or set more limited goals and retained key Northern Alliance fighting forces, many of which were disbanded between 2003 and 2005. More focus on Afghanistan might also have led to innovative security solutions, as it did later. These included militia programs like the Afghan Local Police, whose members were recruited from the villages where they served and were far cheaper—and more tuned in—than national soldiers. Instead, violence intensified as the war drifted.

Pakistan played an increasingly negative role as the conflict continued. Its president during the U.S. invasion, Pervez Musharraf, agreed to support the war after a direct ask from the United States. But Islamabad almost immediately welched on the deal, allegedly airlifting hundreds of senior Taliban fighters, Pakistani intelligence officers, and Islamic militants out of Kunduz in late 2001.

That basic deal between Bush and Musharraf evolved into one in which the United States had free rein to target al-Qaeda members in Pakistan while turning a blind eye to Islamabad’s support for the Taliban. This support evolved in two directions: tolerance of a loose border and organized sanctuaries for high-level Taliban leaders in places like Quetta, and more operational support to Taliban elements like the Haqqani network. However, as Taliban activity increased, along with Pakistani collaboration, that deal became untenable for several reasons. Al-Qaeda activity was commingled with Haqqani network operations against U.S. forces in Afghanistan, while the Haqqani network was also collaborating with Pakistani intelligence services. President Barack Obama was repeatedly restrained from taking action against the Haqqani network by this collaboration, and concurrently against a critical al-Qaeda facilitator.

Internationally, the increased pressure on the United States—and subsequent boost to Taliban stature—was almost indistinguishable from enhanced prestige of al-Qaeda forces. The narrative was that Islamic militants were beating Washington. The more the battlefield in Afghanistan expanded, the more room al-Qaeda had to operate. And U.S. strikes in Pakistan proper against al-Qaeda angered Pakistanis and made the population more sympathetic to the militants—or at least more angry at the United States.

Something seemed bound to give at the height of Obama’s troop surge in Afghanistan, exacerbated by a succession of diplomatic crises with Pakistan in 2011 and 2012. But then the pressure abated. Ultimately, the United States did not want to face Pakistan as an enemy, as the Soviets had. The White House’s logic was that a Pakistan 50 percent onside was better than one that was 100 percent offside. But the problem with that was that the U.S. could not mitigate that 50 percent. It was reluctant to strike the Taliban inside of Pakistan, meaning the best it could do was ask Pakistan politely to end its support, which Islamabad did not. It also underestimated how much Pakistan needed the United States—its economic aid, supply lines, and military support—especially at the apogee of American power.

In any case, meaningful Pakistani action against the Taliban sanctuaries, promised for so long, never materialized, and by the time a change of policy was demanded, the moment had passed.

The war, like most counterinsurgencies, was never going to end with the Taliban defeated. It was going to end with the Taliban slowly losing influence as Afghans went about their lives, with most fighters ultimately reintegrating into society and others being killed. But to achieve that, the United States and the Afghan government would have needed to provide security for the population.

This was effectively the plan when the Obama administration began to surge more than fifty thousand additional troops to Afghanistan in 2009, and there was a shining moment to achieve it. But the first major push of these troops was not to Kabul or Kandahar, but to Helmand, an agricultural province to the southwest.

Operation Moshtarek, as it was called, started with the deployment of fifteen thousand troops to seize Marjah, a town of around eighty thousand residents in a poppy-producing area of central Helmand. U.S. defense officials promised a quick victory in Marjah, but a new civilian government never took hold. Three months into the operation, the commanding general at the time, Stanley McChrystal, called Marjah a “bleeding ulcer.” Security problems there persisted for years after.

That misfire took much of the air out of the new campaign. Making Marjah and Helmand more broadly the centerpiece of the United States’ renewed focus on Afghanistan meant that victory would depend on U.S. forces fighting in a remote Taliban heartland far from major population centers. It was a difficult measure of success, and one that did not result in victory.

And the United States urgently needed a victory. The U.S. public was tired of war, even with Obama behind the plan. And as in Iraq, the administration wanted an early success story. Three years after Operation Moshtarek launched, Americans were still waiting.

A different choice—seizing the beating heart of the country around Kabul and Kandahar, and the road in between—might have built enough stability that security in the outlying provinces could have been farmed out to local militias and strongmen. A central government would never control northern Helmand effectively. Nor did it need to. It needed instead to provide governance for Helmand at the lowest possible cost. That meant local Afghan allies and forces who would meet the United States’ minimum objectives on the ground.

Like Marjah, the departure of McChrystal threw the renewed momentum behind the strategy into doubt. It was always a fragile thing. Obama had imposed an eighteen-month window before withdrawals from the surge began, anyway, and his Democratic Party was lukewarm about more war. But McChrystal was hand picked to be the face of the new strategy. It was his plan and his operation. His call for additional troops had dominated the policy dialogue in 2009.

Along with General David Petraeus, who commanded U.S. forces in Iraq and Afghanistan, and Lieutenant General Michael Flynn, who served as McChrystal’s intelligence chief at Joint Special Operations Command (JSOC) and in Afghanistan, McChrystal was one of the three military officers who had probably done more to defeat al-Qaeda and Sunni extremists than any other living American. At JSOC, McChrystal and Flynn had revolutionized counterterrorism by increasing the tempo of intelligence exploitation, enabling U.S. forces to conduct sequential raids overnight and roll up entire networks.

His firing reflected understandable political sensitivity, given McChrystal’s and his staff’s poor judgment in criticizing the Obama administration on the record to a journalist. What was less understandable was that the administration had promoted a special operations general who had come out of that community’s informal, even insubordinate traditions and given him command of the ship. He did not run the major command like a corporate leader, and nobody should have expected him to.

In retrospect, creating an intermediate command during his tenure—the International Security Assistance Force Joint Command (IJC)—to run the war and having McChrystal focus on allies, partners, Americans, and the media, was backward. His loss drained momentum from Obama’s push. Petraeus was there and gone in a year, and by the time General John Allen arrived to take over, U.S. focus was already shifting away from the war.

By the start of 2017, the Afghanistan war was in its final stages. U.S. troop levels had declined to about nine thousand, from a peak of around a hundred thousand in 2010-2011, and most security functions had been handed over to the Afghan National Army. The Afghan government and the United States had already set Afghanistan’s strategic orientation through compromises over its leadership under President Ashraf Ghani, the shape of its military, and its foreign policy.

Pakistan’s strategy had weathered the U.S. storm, and America’s appetite for additional investment in Afghanistan to change course was nonexistent. However, history offered a clue to a potentially different ending. The last communist leader of Afghanistan, Najibullah, did not fall when Soviet troops withdrew in 1989, but rather when Soviet aid ran out in 1992. Until the collapse of the USSR, he used billions in Soviet assistance to buy loyalty and play off different factions against each other.

The same idea was theoretically available to the United States in 2017. A significant pool of operational assistance for the government, plus contractors and potentially a continued high-end counterterrorism capability at Bagram, might have allowed the Kabul regime to rent the loyalty of different commanders in the east and south with a grab bag of carrots and sticks.

This would have been a variation of what presidential adviser Steve Bannon and others in the first Trump administration reportedly advocated: a contractor-based approach that faced opposition from the Pentagon. But a strategy knitting together Kabul, plus Herat, plus Mazar-e-Sharif, plus a few Pashtun strongmen in the east and south such as the feared Kandahar police chief Abdul Raziq might have kept Afghanistan onside relatively cheaply.

That would have only worked, however, with the existence of ready-to-fight ethnic militias—something like the original grouping that brought down the Taliban in 2001. It would also have required the narrowest possible limits on U.S. goals and a hefty tolerance for ugliness among its rougher allies. The United States might have been able to prevent al-Qaeda from returning to Kabul and keep a friendly regime in power. But gone would have been talk of anything bigger. Perhaps this option had been foreclosed by the decision to back Karzai at Bonn and the subsequent Loya Jirga in Kabul, or by the further centralization of power under his presidency in 2003-2004. Rather than a set of reliable ethnic fighters, the United States was left with a national government led by a Pashtun president that tried to rival the Taliban. And after decades of corruption and fear, that was not enough for the people to fight for it.

Andrew L. Peek is the director of the Scowcroft Center's Adrienne Arsht National Security Resilience Initiative and GeoStrategy Initiative.

This article appeared originally at Atlantic Council.