George W. Bush, Barack Obama and Donald Trump all promised the same thing: The United States would not get stuck with the burden of “nation-building” in Afghanistan.
In October 2001, shortly after ordering U.S. forces to invade, Bush said he would push the United Nations to "take over the so-called nation-building."
Eight years later, Obama insisted his government would not get mired in a long "nation-building project," either. Eight years after that, Trump made a similar vow: "We're not nation-building again."
Yet nation-building is exactly what the United States has tried to do in war-battered Afghanistan - on a colossal scale.
Since 2001, Washington has spent more on nation-building in Afghanistan than in any country ever, allocating $133 billion for reconstruction, aid programs and the Afghan security forces.
Adjusted for inflation, that is more than the United States spent in Western Europe with the Marshall Plan after World War II.
Unlike the Marshall Plan, however, the exorbitant nation-building project for Afghanistan went awry from the start and grew worse as the war dragged on, according to a trove of confidential government interviews with diplomats, military officials and aid workers who played a direct role in the conflict.
Instead of bringing stability and peace, they said, the United States inadvertently built a corrupt, dysfunctional Afghan government that remains dependent on U.S. military power for its survival. Assuming it does not collapse, U.S. officials have said it will need billions more dollars in aid annually, for decades.
Speaking candidly on the assumption that most of their remarks would not be made public, those interviewed said Washington foolishly tried to reinvent Afghanistan in its own image by imposing a centralized democracy and a free-market economy on an ancient, tribal society that was unsuited for either.
Then, they said, Congress and the White House made matters worse by drenching the destitute country with far more money than it could possibly absorb. The flood crested during Obama's first term as president, as he escalated the number of U.S. troops in the war zone to 100,000.
"During the surge there were massive amounts of people and money going into Afghanistan," David Marsden, a former official with the U.S. Agency for International Development (USAID), told government interviewers. "It's like pouring a lot of water into a funnel; if you pour it too fast, the water overflows that funnel onto the ground. We were flooding the ground."
By some measures, life in Afghanistan has improved markedly since 2001. Infant mortality rates have dropped. The number of children in school has soared. The size of the Afghan economy has nearly quintupled.
But the U.S. nation-building project backfired in so many other ways that even foreign-aid advocates questioned whether Afghanistan, in the abstract, might have been better off without any U.S. help at all, according to the documents.
"I mean, the writing is on the wall now," Michael Callen, an economist with the University of California at San Diego and a specialist in the Afghan public sector, told government interviewers. "We spent so much money and there is so little to show for it."
Callen and others blamed an array of mistakes committed again and again over 18 years - haphazard planning, misguided policies, bureaucratic feuding. Many said the overall nation-building strategy was further undermined by hubris, impatience, ignorance and a belief that money can fix anything.
Much of the money, they said, ended up in the pockets of overpriced contractors or corrupt Afghan officials, while U.S.-financed schools, clinics and roads fell into disrepair, if they were built at all.
Some said the outcome was foreseeable. They cited the U.S. track record of military interventions in other countries - Iraq, Syria, Libya, Yemen, Haiti, Somalia - over the past quarter-century.
"We just don't have a post-conflict stabilization model that works," Stephen Hadley, who served as White House national security adviser under Bush, told government interviewers. "Every time we have one of these things, it is a pickup game. I don't have any confidence that if we did it again, we would do any better."
Troubles plaguing many reconstruction programs in Afghanistan have been well documented, but the interviews obtained by The Washington Post contain new narratives from insiders on what went wrong.
"Once in a while, OK, we can overspend," Douglas Lute, an Army lieutenant general who served as the White House's Afghan war czar from 2007 to 2013, told government interviewers. "We are a rich country and can pour money down a hole and it doesn't bust the bank. But should we? Can't we get a bit more rational about this?"
In comments echoed by other officials who shaped the war, Lute said the United States lavished money on dams and highways just "to show we could spend it," fully aware that the Afghans, among the poorest and least educated people in the world, could never maintain such huge infrastructure projects.
"One poignant example of this is a ribbon-cutting ceremony complete with the giant scissors I attended for the district police chief in some Godforsaken province," Lute said. He recalled how the U.S. Army Corps of Engineers had overseen the design and construction of a police headquarters that featured a glass facade and an atrium.
"The police chief couldn't even open the door," Lute said. "He had never seen a doorknob like this. To me, this encapsulates the whole experience in Afghanistan."
Ever since the war started, U.S. officials have debated - and decried - the expense of rebuilding Afghanistan. In 2008, as reports of fraud and excessive spending piled up, Congress created a watchdog agency to follow the money.
Since then, the Office of the Special Inspector General for Afghanistan Reconstruction, or SIGAR, has carried out more than 1,000 audits and investigations, exposing wasteful projects and highlighting $2 billion in potential savings.
In 2014, SIGAR launched a special $11 million project - titled "Lessons Learned" - to diagnose policy failures in Afghanistan. Agency staffers interviewed more than 600 people with firsthand experience in the war.
SIGAR published two Lessons Learned reports that focused on nation-building, but they were steeped in jargon and omitted the most critical comments from the interviews.
“The U.S. government’s provision of direct financial support sometimes created dependent enterprises and disincentives for Afghans to borrow from market-based financial institutions,” concluded an April 2018 report on development of the Afghan private sector. “Furthermore, insufficient coordination within and between U.S. government civilian and military agencies often negatively affected the outcomes of programs.”
In one of the hundreds of Lessons Learned interviews obtained by The Post, Robert Finn, who served as U.S. ambassador to Afghanistan from 2002 to 2003, said Bush administration officials dismissed his early warnings that they needed to do far more to stabilize Afghanistan.
"This is a systemic problem of our government," he said. "We can't think beyond the next election. When we went to Afghanistan everybody was talking about a year or two, and I said to them that we would be lucky if we were out of here in 20 years."
Of the three commanders in chief, Bush may have been the unlikeliest nation-builder. When he first campaigned for the presidency, he derided the Clinton administration for committing to unpopular "nation-building exercises" in Somalia and Haiti.
"I don't think our troops ought to be used for what's called nation-building," he said during a debate with Democratic nominee Al Gore in October 2000. "I think our troops ought to be used to fight and win war."
A year later, Bush ordered U.S. forces to invade Afghanistan. Victory on the battlefield came swiftly. Coping with the aftermath would take longer than most expected.
No nation needed more building than Afghanistan. Desperately poor, it had been consumed by continuous warfare since 1979, when it was invaded by another superpower, the Soviet Union.
Few Afghans knew much about the outside world. A large majority was illiterate. The country's ousted rulers, the Taliban, a movement of religious zealots, had banned many hallmarks of modern civilization, including television, musical instruments and equal rights for women.
"We were dealing with parts of a society who thought the king was still in power, never knew the Russians came, or that the Americans were here," Jordan Sellman, who spent several years in Afghanistan working for USAID, told government interviewers. "They didn't even use currency, but bartered for items. We were bringing 21st-century stuff to a society living in a different time period."
Mindful of Bush's campaign rhetoric, his administration initially tried to avoid responsibility for Afghanistan's reconstruction, according to people interviewed for the Lessons Learned project.
The administration tried to get the United Nations, NATO and other countries to take charge of humanitarian aid and reconstruction. The United States agreed to help train a new Afghan army but pushed to keep it small, because the Pentagon and State Department did not want to bear the long-term costs.
Eventually, however, officials interviewed for the Lessons Learned project said, the Bush administration recognized it had a duty to help Afghanistan build a new economy from scratch. Although Afghanistan had scant experience with free markets, the United States pressured the Afghans to adopt American-style capitalism.
Yet several U.S. officials told government interviewers it quickly became apparent that people who would make up the Afghan ruling class were too set in their ways to change.
"These people went to the communist school," said Finn, the former ambassador. A common Afghan fear, he recalled, was "if you allow capitalism, these private companies would come in and make profit."
Richard Kraemer, a former senior program officer for Afghanistan at the National Endowment for Democracy, told government interviewers that Afghan bureaucrats "were in favor of a socialist or communist approach because that's how they remembered things the last time the system worked." Afghanistan was run by communists from 1978 until 1992.
But Kraemer said U.S. officials suffered from an equally narrow mind-set. "We had all good intentions," he added, ". . . but we had plenty of hubris. Dogmatic adherence to free-market principles led to our inability to adopt a nuanced, balanced approach to what Afghanistan needed."
When it came to economics, others said the United States too often treated Afghanistan like a theoretical case study and should have applied more common sense instead.
Donors insisted that a large portion of aid be spent on education, even though Afghanistan - a nation of subsistence farmers - had few jobs for graduates.
"We were building schools next to empty schools, and it just didn't make sense," a Special Forces officer told government interviewers. He said local Afghans made clear "they didn't really want schools. They said they wanted their kids out herding goats."
U.S. and European officials also insisted that Afghanistan embrace free trade, even though it had almost nothing of value to export.
"What could we sell?" an Afghan official said in a Lessons Learned interview in March 2017. "A few grapes here or something of the like."
Economic policies that might have helped Afghanistan slowly emerge from penury, such as price controls and government subsidies, were not considered by U.S. officials who saw them as incompatible with capitalism, said Barnett Rubin, a former adviser to the United Nations and State Department.
In developing countries, "the idea that there are perfectly functioning markets without subsidies is pure fiction, fantasy," Rubin, a New York University professor and leading academic on Afghanistan, told government interviewers. "Every late-developing country happened by government picking winners."
It didn't take an Ivy League political scientist to see that Afghanistan needed a better system of government. Riven by feuding tribes and implacable warlords, the country had a volatile history of coups, assassinations and civil wars.
The Bush administration persuaded the Afghans to adopt a made-in-America solution - a constitutional democracy under a president elected by popular vote.
In many ways, the new government resembled a Third World version of Washington. Power was concentrated in the capital, Kabul. A federal bureaucracy sprouted in all directions, cultivated by dollars and legions of Western advisers.
Under American tutelage, Afghan officials were exposed to newfangled concepts and tools: PowerPoint presentations, mission statements, stakeholder meetings, even appointment calendars.
But there were fateful differences.
Under the new constitution, the Afghan president wielded far greater authority than the other two branches of government - the parliament and judiciary - and also got to appoint all the provincial governors. In short, power was centralized in the hands of one man.
The rigid, U.S.-designed system conflicted with Afghan tradition, typified by a mix of decentralized power and tribal customs. But with Afghanistan defeated and broke, the Americans called the shots.
"In hindsight the worst decision was to centralize power," an unnamed European Union official said in a Lessons Learned interview.
A German official echoed the point: "After the fall of the Taliban, it was thought that we needed a president right away, but that was wrong."
An unidentified USAID official said he was astounded that the State Department thought an American-style presidency would work. "You'd think they've never worked overseas," he said. "Why did we create centralized government in a place that has never had one?"
A big reason is that U.S. leaders had a potential Afghan ruler in mind. Hamid Karzai, a tribal leader from southern Afghanistan, belonged to the country's largest ethnic group, the Pashtuns.
Perhaps more importantly, Karzai spoke polished English and was a CIA asset. In 2001, a U.S. spy had saved his life, and the CIA would keep Karzai on its payroll for years to come.
At first, to American eyes, the new system of government led by Karzai worked. In 2004, after serving as interim leader, Karzai was elected president in Afghanistan's first national democratic election. He built a personal rapport with Bush; the two leaders chatted frequently by videoconference.
But relations gradually soured. Karzai grew outspoken and criticized the U.S. military for a surge of airstrikes and night raids that inflicted civilian casualties and alienated much of the population. Meanwhile, U.S. officials chafed as Karzai cut deals with warlords and doled out governorships as political spoils.
"After 2005, my impression was that the warlords were back because Karzai wanted them back and he only understood the patronage system," Hadley, the Bush administration national security adviser, told government interviewers. "Karzai was never sold on democracy and did not rely on democratic institutions."
Richard Boucher, who served under Bush as the State Department's chief spokesman and later its top diplomat for South Asia, told government interviewers that Karzai's governing instincts "were to rely on his friends. That is how Afghanistan works - relying on his friends, supporters and local potentates; powers that be, not just powers that the Americans created."
"Getting him to use that governing structure that we put in place, that we told him he had to have, was really hard," Boucher added. "We said, you have to work through this democratic, bureaucratic system just like we have in America."
In 2009, Karzai won reelection, narrowly avoiding a runoff thanks to a massive ballot-box stuffing campaign that tainted the outcome. Many U.S. officials were appalled and pressed for an independent investigation. Karzai, in turn, privately accused the Obama administration of violating Afghan sovereignty and plotting to oust him from power.
In the end, U.S. officials swallowed their objections. After all, they had built the new nation and put Karzai in charge.
A few weeks after Karzai's reelection, Obama announced he would send 30,000 more U.S. troops to the war zone as part of a new strategy to defeat the Taliban and bolster the Afghan state.
In a December 2009 speech at the U.S. Military Academy at West Point, New York, Obama told Americans this would not mean a drawn-out extension of the nation-building campaign that had already dragged on for eight years.
"Some call for a more dramatic and open-ended escalation of our war effort, one that would commit us to a nation-building project of up to a decade. I reject this course," Obama said. "Our troop commitment in Afghanistan cannot be open-ended, because the nation that I'm most interested in building is our own."
Obama's generals, however, held no illusions.
During a June 2010 hearing on Capitol Hill, Army Gen. David H. Petraeus was asked point-blank by skeptical lawmakers whether the United States was nation-building in Afghanistan.
"We are indeed," said Petraeus, who at the time was the chief of the U.S. Central Command. "I'm just not going to evade it and play rhetorical games."
In fact, a cornerstone of Obama's counterinsurgency strategy was to build the Afghan government at breakneck speed - with unprecedented sums from the U.S. treasury.
Petraeus and other U.S. commanders were betting the Afghan people would choke off support for the Taliban if they felt Karzai's government could protect them and deliver basic services.
But there were two big hurdles.
First, there was not much time for the counterinsurgency strategy to work. Obama had given the Pentagon just 18 months to turn the tide of the war before he wanted to start bringing troops home.
Second, across much of Afghanistan, there was hardly any government presence to begin with. And where there was, it was often corrupt and hated by the locals.
As a result, the Obama administration ordered the military, the State Department, USAID and their contractors to build up the Afghan government as quickly as possible. In the field, soldiers and aid workers were given a virtual blank check to construct schools, hospitals, roads - anything that might win loyalty from the populace.
"Petraeus was hell-bent on throwing money at the problem," an unidentified U.S. military officer told government interviewers. "When Petraeus was around, all that mattered was spending. He wanted to put Afghans to work."
An unnamed USAID official complained that he was always being asked “How much are you spending?” instead of “Are you winning the battle?” He added, “We were always chasing the dragon - always behind, never good enough.”
Another unidentified aid worker told government interviewers that the pace was unrealistic and unsustainable: "It was difficult to bring a region that is [a] hundred years . . . behind out in a few years."
In a Lessons Learned interview, Petraeus acknowledged the spendthrift strategy. But he said the U.S. military had no choice given Obama's order to start reversing the surge in 2011.
"What drove spending was the need to solidify gains as quickly as we could knowing that we had a tight drawdown timeline," he said. "And we wound up spending faster than we would have if we felt we had forces longer than we did."
Amid the haste to spend, U.S. agencies wasted large sums of money on ghost projects that never took shape.
Tim Graczewski was a Navy Reserve officer who oversaw economic development projects in southern Afghanistan from 2009 to 2010. When he arrived for the surge, he told government interviewers, he had to hunt for a 37-acre project that appeared to exist only on paper.
Before his arrival, the U.S. government had signed $8 million in contracts to build an industrial park near Kandahar for 48 businesses. But after reviewing the files, Graczewski said, he could not even find the site.
"It blew my mind how much we didn't know about the park in the first place when we embarked on this project," he said. "It was impossible to get info on it, even where it was located. It was that much of a blank spot. Nobody knew anything about anything."
Graczewski said he finally located the property, but there were no buildings - only some empty streets and sewer pipes.
"Don't know who did it, but figured it was there, so let's try to use it," he recalled. Despite efforts to revive the project, he said, it "fell apart" after he left in 2010.
U.S. auditors visited the site four years later and found it largely deserted. A single company, an ice cream packing outfit, was open for business.
Undaunted, the U.S. government tried to connect the industrial park to an even more ambitious nation-building project - to generate electricity for Kandahar, Afghanistan's second-biggest city, and surrounding areas.
Hamstrung by a primitive electrical grid, Kandahar suffered from a scarcity of power. U.S. military commanders saw an opportunity. If they could generate a reliable flow of electricity, grateful Kandaharis would support the Afghan government and turn against the Taliban.
To do that, the U.S. military wanted to rebuild an aging hydroelectric power station at the Kajaki Dam, about 100 miles north of Kandahar. USAID had built the dam in the 1950s and installed turbines in the 1970s, but it quickly fell into disrepair.
The U.S. government had been trying to jump-start the project and add capacity since 2004 without much to show for it. The Taliban controlled the area surrounding the dam, as well as some transmission lines. Repair crews needed armed convoys or helicopters to access the site.
Despite the risks, by 2010, U.S. generals were lobbying to invest hundreds of millions of dollars in the project, calling it a critical part of their counterinsurgency strategy.
Some development experts pushed back, arguing that it made no sense to finance a giant construction project in enemy territory. They noted that the Afghans lacked the technical expertise to maintain it in the long run.
They also questioned whether it would really help win the hearts and minds of Afghans accustomed to life without central power.
"Why did we think providing electricity to communities in Kandahar who had no concept of what to do with it would convince them to abandon the Taliban?" a senior USAID official said in a Lessons Learned interview.
In the end, the generals won the argument. Ryan Crocker, the U.S. ambassador to Afghanistan at the time, told government interviewers he had deep misgivings about the dam project but approved a portion of it anyway.
"I made the decision to go ahead with it, but I was sure it was never going to work," Crocker said. "The biggest lesson learned for me is, don't do major infrastructure projects."
For the generals, the dam project was not enough.
It would take years to complete, and with the clock ticking on their counterinsurgency strategy, they wanted to supply electricity to the Kandaharis right away. So they drew up a temporary plan to buy giant diesel-fueled generators that could start humming in a matter of months, not years.
It was a horribly inefficient and costly way to generate electricity for an entire city. Expenses would run to $256 million over five years, mostly for fuel. Again, some people tried to push back.
An unidentified NATO official told government interviewers that he was given the task of trying to secure financing for the generators from international donors but got nowhere.
"Anyone who looked at this more closely could see that the math didn't add up, that it was all nonsense," he said. "We went to the World Bank [and] they didn't want to touch it. . . . People look at it and they think it's crazy."
Another former U.S. ambassador said that he opposed the diesel generator plan, too, but that U.S. military commanders prevailed.
"Petraeus got the power back on in Iraq and wanted to do the same in Afghanistan," the unnamed former ambassador told government interviewers. "But in Iraq, it made more sense; they had oil, engineers, and indigenous capacity; it was doable."
By December 2018, the U.S. government had spent $775 million on the dam, the diesel generators and other electrical projects in Kandahar and neighboring Helmand province, according to a SIGAR audit.
Power generation at the dam has nearly tripled, but the projects are still plagued by dysfunction; last year, USAID determined that the Afghan public utility for Kandahar was not commercially viable and may never be able to operate without foreign subsidies.
Jeffrey Eggers, a retired Navy SEAL and White House official under Bush and Obama, told government interviewers that such projects failed to achieve their objective - bringing peace and stability - and that U.S. military officials were guilty of "biting off more than they can chew."
"There is a bigger question here - why does the U.S. undertake actions that are beyond its abilities?" Eggers said. "This question gets at strategy and human psychology, and it is a hard question to answer."
The nation-building campaign was undone not only by white elephants. According to the Lessons Learned interviews, as well as audits from SIGAR, one of the most mismanaged pots of money was the Commanders' Emergency Response Program, or CERP.
Authorized by Congress, it allowed military commanders in the field to bypass normal contracting rules and spend up to $1 million on infrastructure projects. But most cost less than $50,000 each.
Commanders told government interviewers that they were under so much pressure to spend that they blindly copied CERP paperwork from past projects, knowing that it was unlikely anyone would bother to inspect it afterward.
"You'd see the same picture of a [health] clinic posted to a hundred different clinic project reports around the country," said one senior officer who worked at military headquarters in Kabul.
An Army brigade commander in eastern Afghanistan told government interviewers that he often saw CERP proposals that referred to "sheikhs" - a giveaway that they were cut-and-pasted from reconstruction projects in Iraq. ("Sheikh" is an Arabic title of respect but is generally not used in Afghanistan.)
At one point, the brigade commander recalled telling his staffers that if they could not show that a CERP project would be beneficial, "then the smartest thing to do is nothing." In response, he said: "I got crickets. 'We can't build nothing,' they said. I told them we might as well throw our money away."
Brian Copes, an Army National Guard general who served as a civil-affairs commander in Khost province in eastern Afghanistan, likened the flood of aid to "crack cocaine," calling it "an addiction that affected every agency."
In a Lessons Learned interview, he said he came across a U.S.-built greenhouse that cost $30,000 and had fallen into disuse because the Afghans could not maintain it. His unit built a replacement greenhouse out of iron rebar that worked better and cost only $55 - despite pressure to spend far more.
"Congress gives us money to spend and expects us to spend all of it," Copes said. "The attitude became we don't care what you do with the money as long as you spend it."
Despite its best efforts, the U.S. military could spend only about two-thirds of the $3.7 billion that Congress funded for CERP, according to Defense Department figures. Of the $2.3 billion it did spend, the Pentagon was able to provide financial details for only about $890 million worth of projects, according to a 2015 audit.
Officials from other agencies told government interviewers they were appalled at the waste and mismanagement.
"CERP was nothing but walking-around money," said Ken Yamashita, USAID's mission director for Afghanistan from 2011 to 2014, likening the payments to cash handouts for political purposes.
An unidentified NATO official called the program "a dark pit of endless money for anything with no accountability."
Of all the flaws with the Afghanistan nation-building campaign - the waste, the inefficiency, the half-baked ideas - nothing confounded U.S. officials more than the fact that they could never tell whether any of it was actually helping them win the war.
An Army officer assigned to U.S. military headquarters in Kabul during the surge told government interviewers that it was hard enough to track whether CERP projects were actually built, let alone whether they made a difference on the battlefield.
"We wanted hard quantitative metrics that would tell us that X project is producing the desired outcomes, but we had a hard time defining those metrics," he said. "We had no idea how to measure if [a] hospital's existence was reducing support for the Taliban. That was always the last 10 yards that we couldn't run."
Even some of the most well-intentioned projects could boomerang.
Tooryalai Wesa, who served as governor of Kandahar province from 2008 to 2014, said U.S. aid workers once insisted on carrying out a public-health project to teach Afghans how to wash their hands.
"It was an insult to the people. Here people wash their hands five times a day for prayers," Wesa told government interviewers. "Moreover, hand wash project is not needed. Think about employment, and think about enabling people to earn something."
But that could backfire, too.
For one project in Kandahar, U.S. and Canadian troops paid villagers $90 to $100 a month to clear irrigation canals, according to Thomas Johnson, a specialist on Afghanistan who works as a professor at the Naval Postgraduate School.
It took a while for the troops to figure out their program was indirectly disrupting local schools. Teachers in the area earned much less, only $60 to $80 a month.
"So initially all the school teachers quit their jobs and joined the ditch diggers," Johnson said in a Lessons Learned interview. He served as a political and counterinsurgency adviser to the Canadians from 2009 to 2010.
A similar problem arose in eastern Afghanistan, where one gung-ho Army brigade was so determined to make a difference that it promised to build 50 schools - but unwittingly ended up helping the Taliban, according to an officer in the brigade.
“There weren’t enough teachers to fill them, so buildings languished,” the unnamed U.S. military officer told government interviewers, “and some of them even became bomb-making factories.”