The Once and Future Defeat in Afghanistan


“Everyone is failing us.” These were the first words that Ashraf Ghani uttered — not as he fled the advancing Taliban on Aug. 15, 2021, but in March 2002 as we sat down to dinner on a chilly and wet night on my first post-9/11 visit to Kabul. I had known him since 1984, first as an academic colleague and then as a World Bank official. We had collaborated as informal advisors to U.N. envoy Lakhdar Brahimi during his first assignment from 1997 to 1999, and then as part of the United Nations team in 2001, when the two of us served as Brahimi’s advisors during the Bonn Conference that set the framework for the post-Taliban government.

Soon after, Ghani had left for Kabul, where he established the Afghanistan Aid Coordination Agency, which he said would align international aid with Afghan priorities. Over dinner, he described how the Bush administration had allocated no new money for rebuilding the country, which was then devastated by (a mere!) 24 years of war. International agencies had presented gross underestimates of reconstruction costs, and their uncoordinated operations were marginalizing the destitute government. The U.N. political mission was virtually alone in pressing for the broadening of the interim administration at the upcoming Emergency Loya Jirga. Ghani became finance minister at that jirga three months later and ultimately president in 2014, only to discover that the policies that he prescribed in his book, Fixing Failed States, ignored political realities in both Afghanistan and America. Even after the United States signed a February 2020 deal with the Taliban, promising to leave by May 1, 2021, he refused to believe in or plan for the withdrawal of U.S. forces. As late as Aug. 2, 2021, he boasted to the Afghan parliament that he would get the situation “under control” in six months.



A few days after my 2002 dinner with Ghani, following a few thin leads, I found Amrullah Saleh, who I had first met in the northern Afghan city of Kunduz in January 1996. Defense Minister Ahmad Shah Massoud had sent him to meet me as I traveled on a mission for the Open Society Foundation to evaluate the situation of refugees from Tajikistan in northern Afghanistan. Saleh was then the 23-year-old deputy spokesman of the Defense Ministry. I found him again five years later in his office in the upscale (for 2002 Kabul) Shahr-i Naw neighborhood, from which he directed the counter-terrorism department of Afghanistan’s intelligence agency. Then, Saleh said of Ghani, “To us [Tajiks] it is obvious that he has an ethnic [Pashtun] agenda.” In 2004 he became Afghanistan’s intelligence chief, only to be fired by President Hamid Karzai in 2010. In 2019, he apparently set his prior reservations aside and agreed to become Ghani’s first vice president in his second term. In August 2021, as Kabul’s defenses crumbled, Saleh evacuated the capital with some trusted men, alongside Massoud’s son Ahmad, to their native Panjshir Valley to make a stand against the Taliban. After Ghani fled, Saleh briefly claimed to be acting president, until he and Massoud were forced to flee to Tajikistan.

After our initial meeting in his office in March 2002, over lunch in the Marco Polo Chinese restaurant across the street, Saleh told me that he was spending half of his time trying to repair the political harm caused by the civilian casualties and mistaken detentions that had resulted from U.S. raids based on false intelligence. In September 2002, when I quoted his warnings without attribution at a conference in the United Kingdom, a confident British minister responded sarcastically, “I’ve just heard the voice of doom!” Saleh came up to me afterwards and said, “You used my words!” “I spoke, because you could not,” I answered.

During that same visit to Kabul, on the eve of Nawruz in 2002 (1381), I dined at the palace with Karzai, whom I had known since 1987. Over green tea before dinner, I observed that in the aftermath of the overthrow of the Taliban, provinces and localities had been taken over by local commanders. While political power needed to be separated from military force, I wondered if this arrangement might evolve into a less centralized Afghan state. He asserted that as soon as he could, he would reestablish the old centralized system, appointing governors and district governors from outside the areas that they were to administer to ensure that they represented the central government rather than local factions.

Contrary to a widespread misconception, the United States did not “impose” this centralized structure on a “traditionally decentralized” Afghan government. The centralization of the Afghan state dated back to the regime of Amir Abdul Rahman Khan (1880 to 1901), when Britain subsidized a strong army to keep Afghanistan as a stable if isolated buffer between the British and Russian empires. By 2002, that state was back in its familiar role, helping a great power to maintain “security,” this time from “terrorism,” in return for aid.

Karzai, like most of those Afghan elites with ties to the old regime who regained power in alliance with elements of the anti-Soviet mujahidin who had fought against the Taliban through the political arrangements concluded at Bonn, supported this centralized structure as a means to maintain control over and prevent foreign powers from exploiting differences among the diverse communities of Afghanistan. However, that structure was — and is — wildly unsuited to the functions that it acquired: providing public services or, supposedly, “good governance” to the people of Afghanistan. As abuses of power and civilian casualties increased after 2002, the Taliban, flush with local recruits seeking revenge for attacks on their villages and tribes by the United States and its allied warlords, took advantage of the inability of the state to penetrate rural society as they won over, intimidated, or eliminated local leaders.

In July 2006, after sensing during my repeated visits that the American project in Afghanistan was starting to fall apart, I returned to Kabul. Saleh, who by then was Afghanistan’s intelligence chief, gave me a report that he had written two months earlier, entitled “The Strategy of Insurgents and Terrorists in Afghanistan.” Alluding to Sun Tzu, Saleh warned, “The pyramid of Afghanistan government’s legitimacy should not be brought down due to our inefficiency in knowing the enemy, knowing ourselves and applying resources effectively.” That, however, is exactly what happened.

The failure of the U.S.-led mission in Afghanistan arguably started on Dec. 6, 2001, the day after the signing of the Bonn Agreement, when U.S. Secretary of Defense Donald Rumsfeld, under repeated questioning from reporters at the Pentagon, rejected a political agreement between Karzai and the remaining Taliban leadership in Kandahar. In shock from their rapid and seemingly total defeat, the Taliban had agreed to acknowledge Karzai’s leadership and hand over the four provinces that they still controlled in return for an amnesty that would allow their leader Mullah Omar to live in Kandahar “with dignity.” This would have permitted them to participate in the process set up by the Bonn agreement to establish a permanent government. Instead of being sent to Guantanamo or to those famous Afghan graveyards, they could have participated in proportion to their true numbers and influence (small, but real) in the drafting and implementation of the constitution. Rumsfeld, however, responded that there would be “no negotiated solution.” Nearly 20 years later, he was proved right, though not in a good way.

For Rumsfeld, the core aim of U.S. policy in Afghanistan was not to stabilize the country or help it to create a government that could deal with terrorism on its own, but to set an example for all the evil people who might think of attacking the United States by capturing and killing the “bad guys.” This remained policy even when, as Rumsfeld complained, “I have no visibility into who the bad guys are.”

By 2006, Saleh thought that the Taliban were “succeeding strategically.” Unless the Afghan government and international coalition pursued “a clear and well-defined strategy,” he wrote, “[the Taliban] will be invincible.” Four months after Saleh finished his report, in September 2006, the Taliban launched their first offensive against a major city: Kandahar. They suffered a tactical defeat at the hands of Canadian NATO forces, but they established a permanent presence in the districts around Kandahar, from where they rode their pickups, motorcycles, and abandoned U.S. Humvees into the city when the Afghan army collapsed 15 years later.

President Barack Obama recognized the deteriorating situation behind the public happy talk. He promised to end the war in Iraq and focus on Afghanistan. The first draft of his administration’s policy on Afghanistan was a March 2009 report written by Bruce Riedel, a retired CIA counter-terrorism analyst and National Security Council official working at the Brookings Institution. The policy review that followed in the fall of 2009 was largely based on an assessment of the situation on the ground by the newly appointed commander, Lt. Gen. Stanley A. McChrystal. Though written in a different style, its content was virtually identical to that of Saleh’s report three years before, except the situation had grown worse.

Riedel’s report called for a “fully resourced counter-insurgency” to produce “a self-reliant Afghanistan that will enable a withdrawal of combat forces while sustaining our commitment to political and economic development.” That sentence in the introduction, however, was the last — and only — reference to self-reliance in the report. The Riedel report pretended to have a plan to strengthen local government in Afghanistan — “increasing civilian capacity” by bringing to bear American and other international “expertise,” as if the problems of the Afghan state resulted from a lack of training by Americans — but it did not even pretend to have a plan to make Afghanistan self-reliant. It seemed to assume that organizations built with foreign aid on foreign models would function just like institutions that countries designed and financed themselves.

McChrystal doubled down on the counter-insurgency Ponzi scheme. Classic counter-insurgency consisted of three stages: clear (out the enemy), hold (the cleared territory with military force), and then build (the civilian institutions of governance that would substitute for the military). This model was derived from cases either of states fighting a domestic insurgency (India, Colombia) or colonial powers facing revolts (France in Algeria, the United Kingdom in Malaya). In all such cases, the state that commanded the military force that cleared and held was the same state that would build the new institutions. As the United States in Afghanistan was neither the national government nor, ostensibly, a colonial power that intended to occupy the country forever, the designers of U.S. counter-insurgency theory added a fourth stage: transfer (to the control of the national government). To do that, McChrystal wrote, the United States had to “prioritize responsive and accountable governance.” On Sept. 30, 2009, when I saw that meaningless phrase in a copy of the report that Richard Holbrooke shared with me in his office, I thought it was a plan to blame the civilians when it failed.

The U.S. government, military or civilian, had no idea how to build and transfer institutions that would work in rural Afghanistan. Neither did the Afghan elites who were Washington’s partners. Never in its history had the Afghan state built an effective local administration. The governors and district governors working for the centralized state relied on a partnership with informal institutions in the village (mosque, jirga, and shura). The state had rarely “extended its writ” into the villages without provoking revolts. The elites who ran the state feared the consequences of devolving power to the localities themselves. The Taliban worked through the non-state local institutions, primarily the mosques. With their local knowledge — most Taliban operated in the areas where they had grown up — they knew exactly whom to persuade, intimidate, or kill.

“Prioritize responsive and accountable governance” was like the reassurance a fund manager might give nervous clients as he asks them for new money to pay earlier investors the above-market returns he had promised. When the Afghan government failed to be responsive and accountable, and the insurgency kept spreading, the Defense Department would ask for more troops, more money, or more time. When the money was diverted or stolen, as was inevitable when the richest country in the world throws billions of dollars at problems it does not understand in one of the poorest countries of the world, it proposed “anti-corruption” efforts to uproot the corruption it continued to fund. But by trying to compensate for unresolved political issues with infusions of force, money, and training, the United States was just postponing the day when the whole enterprise would collapse like the Ponzi scheme it was.

The Obama administration did eventually authorize political talks, though the Taliban had greatly stiffened their demands since 2001. As senior adviser to successive Special Representatives for Afghanistan and Pakistan (Holbrooke, Marc Grossman, and James Dobbins) I advocated and helped to orchestrate the first direct talks between the United States and the Taliban. The talks made slow progress through 2010 and 2011, but at the insistence of Karzai and senior U.S. officials who had resisted the talks, Grossman (then the lead U.S. negotiator) told his Taliban counterpart in Doha in January 2012 that the negotiations would not resume until the Taliban talked to the Afghan government. At the NATO summit in Chicago in June 2012, Obama and Karzai resolved to “double down” on reconciliation. In a joint statement issued after Karzai’s official visit to Washington in January 2013, the two presidents “said that they would support an office in Doha for the purpose of negotiations between the High Peace Council and the authorized representatives of the Taliban.” When the office briefly opened in June 2013, however, the Taliban and Qatar violated what the United States thought had been agreed about the display of symbols of the “Islamic Emirate.” In retrospect, it is not surprising that a complex and politically delicate negotiation in which the parties did not meet for a year and a half and never agreed on a final document would result in a misunderstanding. It was my job to drive to the Taliban office at night and wait until the Qataris persuaded the Taliban to lower the flag and remove the emblems of the Islamic Emirate. I left the government soon after, and U.S. diplomacy then focused on negotiation of a bilateral security agreement with the Afghan government that provided the latter with a false sense of security.

Much of the leadership of the U.S. Departments of Defense (civilian and military) and State tried to delay these efforts repeatedly. They paid lip service to a political solution but always wanted to postpone negotiations until the United States was in a “position of strength.” There was, however, no evidence that waiting made the United States stronger and forced concessions from the Taliban. On the contrary, the U.S. position deteriorated, and the Taliban position kept hardening. The proponents of “strategic patience” or a “long-term” U.S. presence had no alternative strategy to offer. They were asking to prolong the scam. Their motto might have been, when you are in a hole, make a long-term commitment to digging.

The Taliban leadership’s safe haven in Pakistan — something absent from the classic models of counter-insurgency — meant that military efforts could not attack the Taliban’s core leadership. Many critics, including me, have argued that one reason for the failure of U.S. efforts in Afghanistan was the failure to exert sufficient pressure on Pakistan to close the Taliban sanctuary. This is only true if the U.S. defeat in Vietnam was due to Washington’s failure to exert sufficient pressure on the Soviet Union and China to stop supporting North Vietnam. But there was nothing the United States could do about the fact that it was a distant offshore power, while Pakistan is a deeply entwined neighbor of Afghanistan. In Vietnam, America tried — and failed — to defeat the enemy despite its external supporters. In Afghanistan, the United States either ignored Pakistan’s behavior or engaged in efforts to “change the strategic calculus” of Islamabad, as if Washington could alter the entrenched threat perceptions of Pakistan’s military with economic aid or threats it could not carry out. Claims that the United States “failed” to exert sufficient pressure on Pakistan assume that there is a level of pressure that the United States could have exerted that would have forced Pakistan to change its policy. That is one of many examples of Americans’ over-estimation of their own power. File under: failure to “know yourself.”

The United States found itself in the situation described by the Prussian general and scholar Carl von Clausewitz, who observed, “There is nothing more common than to find considerations of supply affecting the strategic lines of a campaign and a war.” Afghanistan is landlocked. All land and air routes through which Afghanistan might receive aid or engage in trade traverse Pakistan, Iran, or a combination of Russia or China with Central Asian countries. Unless the United States persuaded Russia, China, or Iran to support its war efforts on their doorstep, Pakistan provided the sole politically and logistically feasible route for any aid or military supply relationship between the United States and Afghanistan. There was a limit to how much pressure Washington could exert against its own supply lines. Exasperated U.S. officials occasionally proposed special operations raids into Pakistan to disrupt the Taliban leadership’s safe haven, but in addition to the many other reasons not to attack an ostensibly allied nuclear power backed by China, it did not appear feasible for a force whose logistics depended on Pakistan to attack Pakistan. As a Russian colleague said to me in Moscow in June 2017, “Pakistan certainly deserves whatever you do to it, but it won’t work.”

Throughout the war in Afghanistan, Washington was consistently losing the international leverage that it had enjoyed since 1945. Over those 20 years, the United States spent an estimated $14 trillion dollars on wars, largely funded by debt-financed supplemental appropriations, while ignoring the decay of its own fiscal soundness, physical infrastructure, healthcare system, education system, and political institutions. In 2001, according to World Bank figures, Chinese GDP in purchasing power parity was 41 percent the size of U.S. GDP, while in 2019 Chinese GDP was 12 percent larger. To the extent that the ratio of the two GDPs serves as a rough indicator of relative power, by 2019 China’s relative power compared to the United States had nearly tripled (1.12 versus 0.41). A comparison of U.S. and Chinese policies toward Pakistan shows the consequences.

On Feb. 2, 2009, shortly after he was appointed as U.S. Special Representative for Afghanistan and Pakistan, Holbrooke convened a consultation at the Asia Society in New York to discuss what it would take to turn Pakistan into a reliable partner for the United States. After discussion of the challenges posed by the Pakistani army and intelligence services and the proliferation of militants and terrorists in Pakistan, World Bank economist Shah Javid Burki estimated that donors would need to provide Pakistan with $40 to $50 billion over five years to safeguard the country from economic collapse. In response to these sorts of analyses, on Oct. 15, 2009, Obama signed into law the Enhanced Partnership with Pakistan Act, also known as the Kerry-Lugar-Berman bill, which provided a total of $7.5 billion of assistance for Pakistan over five years. The act expired in 2014 and was not renewed, by which time China had made the China-Pakistan Economic Corridor, established in 2013, the flagship project of its Belt and Road Initiative. The value of Chinese projects in Pakistan was estimated at $62 billion in 2020, slightly higher than what Burki had estimated would be necessary. The availability of an increasingly stronger and more reliable alternative to the United States weakened Washington’s ability to influence Pakistan.

The U.S. government also thought that its firepower and money would enable it to ignore some simple realities. Afghanistan was by a large margin the poorest country in Asia, poorer than most countries in sub-Saharan Africa, and poorer even than Haiti. It also had the lowest literacy rate in Asia — the only Asian country with literacy under 50 percent and a much lower rate for women. It had the lowest life expectancy, highest maternal mortality rate, and highest infant mortality rate in Asia. Its population was scattered among a few cities and some mountain and river valleys separated by deserts. It had less access to electricity than any other Asian country. With no alternative for survival, its farmers had accepted offers from traffickers to make Afghanistan into the world’s leading provider of opium products to supply the demand from the developed world. The Afghan government had never completed a census and did not know how many people were living in Afghanistan or in the various provinces and districts, or how many of its citizens belonged to the various tribes, clans, ethnic groups, and sects. The Afghan state did not routinely issue birth or death certificates. People’s names and addresses were not registered with the government or required to conform to any standardized format. Without basic demographic data it was impossible to construct a valid list of eligible voters or to evaluate election outcomes against demographic data. Dr. Abdullah Abdullah once told a U.N. electoral official that he would accept tribal voting, in which an elder brought, say, 600 cards to the polling site and claimed all the members of his clan were voting for a particular candidate — but only if the clan actually had 600 eligible voters, which no one could determine. Under these circumstances, a winner-takes-all presidential election that gave the victor access to billions of dollars of aid and lethal force supplied by the United States could not possibly be free and fair. The only way to convince the apparent loser to accept the results was to dispatch senior U.S. officials to Kabul — the chairman of the Senate Foreign Relations Committee (John Kerry) in 2009, the secretary of state (again Kerry) in 2014, and again the secretary of state (Mike Pompeo) in 2019. A government whose election outcomes are adjudicated by a foreign power with troops stationed in the country is hardly a democracy. Efforts to address some of these problems led to temporary improvements in education, healthcare, energy supply, and other indicators, but they also turned Afghanistan into one a handful of the most aid-dependent countries in the world, along with a few war-torn African countries and some island microstates. Today, Afghanistan depends for electricity on transmission lines from its neighbors, for which it now cannot pay the bills. With aid embargoed, it cannot pay the salaries of teachers and health-care workers. The infusion of dollars made Afghanistan’s exports uncompetitive and increased the country’s dependence on food imports, part of the heritage of the Soviet war. With the foreign exchange assets that financed imports blocked, and repeated droughts attributed to climate change destroying up to 40 percent of Afghanistan’s own crops, Afghan children are starting to die of hunger. The U.S. embargo is destroying the “gains” of the past 20 years before the Taliban have even announced their policies.

The Taliban presented a threat to Afghanistan so serious that American military planners estimated that it would require defense and security forces of over 300,000 men to defend the state, plus a long-term presence of U.S. military and contractors. They equipped the Afghan forces with weapons systems interoperable with NATO (though Afghans had been using Soviet/Russian-model weapons since the 1950s), and sophisticated intelligence, surveillance, and reconnaissance capabilities, all to be operated by a combat force that could not read the equipment manuals, if they could read at all. The United States wanted to spend about $4 billion per year on security forces (not to mention the cost of its own presence) in a country whose estimated GDP peaked at $20.5 billion in 2013 and stagnated and fell after the reduction in the foreign troop presence. America supplied the troops’ wages, but the Afghan government’s payment systems were so ineffective and corrupt that, by some estimates, by the end there were as few as 50,000 real soldiers, while the rest of the 300,000 listed on paper were “ghosts” whose wages were skimmed off by their superiors. The Afghan forces that, according to Holbrooke’s summary of a White House discussion in December 2009, would be “our ticket out of there,” were built in such a way that they collapsed before the United States had even completed its withdrawal.

The main victims of this scam were the Afghan people. While the U.S. forces were there, some Afghans benefited from the foreign aid and the freedoms that came with it. These were real gains, for as long as they lasted, but there was no plan to make them last. Other Afghans, in war-torn rural areas, were caught between America’s poorly targeted drone strikes and night raids and the Taliban’s violence. The promises that the United States made to stand shoulder-to-shoulder with Afghans, formalized in the 2014 Bilateral Security Agreement, gave a false sense of security to the elites who gained power and the new professional class, especially the women, who gained previously unheard-of freedoms. To those parts of the population decimated by U.S. bombs those promises always rang false. Eventually the United States betrayed the Afghans who had supported its presence, claiming that it had never really meant those promises of a better life — it had come only to fight terrorism aimed at its homeland and people. Those who broke the impossible promises bear some of the responsibility, but those who recklessly made them in the first place bear more.

On its way out, the last operation by the U.S. military was a “counter-terrorism” drone strike that killed ten innocent people including seven children. After the United States left, its embargo on the country’s financial assets and suspension of all aid amounted to an even more massive attack on the wrong targets — in this case, the entire population of Afghanistan.

When the Taliban were completely defeated, in December 2001, Washington saw no reason to negotiate with them. When the idea gained traction after years of military and diplomatic failure, military leaders wanted to wait until the United States was in a position of strength. Finally, President Donald Trump negotiated with them only because he had decided to liquidate the Afghan War like one of the Trump Organization’s non-performing assets. The Taliban knew it and bargained accordingly. The Doha Agreement of Feb. 29, 2020 was a weak agreement that reflected America’s declining position in Afghanistan and the world.

On Aug. 15, 2021, the army that America built disappeared and the president whom America had supported fled the country. On Sept. 5, 2021, after the Taliban had finally driven him from not only Kabul but also Panjshir and, ultimately, Afghanistan, Saleh, who had always opposed negotiating with the Taliban, for the first time appealed to that mythical entity, the West, “to press for a political settlement with the Taliban.” The ignominious U.S. exit and the fates of Ghani and Saleh illustrate the result of self-deception about one’s own strength and capabilities. The United States negotiated seriously only after Trump had already decided to withdraw troops. Ghani boasted he would defeat the Taliban and then fled. Saleh sought a political settlement only when he was in the same position as the Taliban in December 2001.

Now the Taliban are making the same mistake. They have announced an “interim government” with no procedure or timetable for becoming a permanent government. Besides a very few individuals, it is composed entirely of Pashtun Taliban officials. Pakistan’s favored Taliban, the Haqqanis, dominate. Taliban leaders who sought to gain some independence from Pakistan or to seek a negotiated solution have been marginalized. Under a superficial calm reminiscent of 2002, things are simmering, and the alternative to the Taliban is not Karzai, Abdullah, Saleh, or Massoud. It is the Islamic State in Afghanistan. Even as it terrorizes Afghanistan’s minority Shi’a with mass killings, it is recruiting members from across Afghanistan’s Sunni groups: Tajiks, who predominated in the former government’s officer corps and intelligence services, and who find themselves marginalized like the Sunni former officers in 2003 Iraq, along with the disaffected young from other ethnic groups, are joining. Highly trained former members of the Afghan military and intelligence services have been identified among those fighting. One former senior officer was recently killed by the Taliban in a firefight with the Islamic State. “That’s exactly how it started in Iraq – with disenchanted Saddam Hussein generals,” a senior Western official warned the Wall Street Journal.

Afghanistan needs a political settlement as much today as it did on Aug. 14, 2021. If the Taliban persist in doing something that they themselves, in their more lucid moments, have said is impossible — imposing the rule of one group over Afghanistan — their “Islamic Emirate” will meet the same fate as its predecessors.

Over a late-night dinner in the fall of 2009, Holbrooke commented that the United States was using the same programs in Afghanistan as it had in Vietnam, where he had started his career in the early 1960s. It had changed the names, but what hadn’t changed, he said, is that they still didn’t work. The Taliban, too, may change the names of all their government’s programs and operations. They may even change the name of the state, but the victims will remain the same.



Barnett R. Rubin is a nonresident fellow at the Center on International Cooperation of New York University and the Quincy Institute for Responsible Statecraft. He has taught at Yale and Columbia universities, headed the Center for Preventive Action at the Council on Foreign Relations, and served as senior adviser to both the U.S. special representative for Afghanistan and Pakistan (2009 to 2013) and the U.N. special representative of the secretary-general for Afghanistan (2001 to 2002). His most recent book, Afghanistan: What Everyone Needs to Know, was published by Oxford University Press in 2020.

Image by iPostcross