Aid workers, members of Congress, ordinary Afghans, and ordinary Americans alike were angry and frustrated, but the situation regarding corruption tended to be misunderstood. Yes, there was waste and graft, and millions were embezzled. But it was also true that Afghanistan was still a tribal society in which tribal leaders and local bigwigs saw it as their duty to take from the state resources for their community. Karzai felt the need to satisfy that demand to survive at the top. That sort of corruption is not alien to politics, and certainly not in Afghanistan.14
Did Karzai’s corruption really matter to the ebb and flow of the insurgency? Yes, but not in the ways that we might think it does. There was never evidence that most who joined the Taliban did so in protest against the corruption of Karzai’s ministers. The problem was local. The corruption that really mattered, that angered the small peasant and drove him to pick up a gun and join the Taliban, was being shaken down by local police and government officials. We treated Karzai as if he was head of an independent sovereign government, but in reality his was no government at all. He was holed up in the capital, reliant even there on foreign protection for his physical security, and had a writ that could not run much of anything without U.S. help. Karzai was (as he remains) no more than a glorified “mayor of Kabul.”
His government was poorly designed, too. On paper it was overcentralized—the central government controlled the purse strings and made every decision on education, health care, or development. Yet in practice it was absent from large parts of the country, and where it was present people did not welcome it but wished that it would go away.15 The economy was a shambles, too. Infrastructure remained inadequate and industry nonexistent, and agriculture barely dented endemic poverty in the countryside. The country’s economy was a sum of the drug trade plus the money that international aid and military operations sloshed around.16
Afghans blamed the sorry state of the economy on Karzai’s failings and on America, his main backer. In growing numbers, they were lending the Taliban a hand to take back the country.17 The situation was particularly bad in the Pashtun heartland of southern Afghanistan, which had served as the Taliban’s power base in the 1990s. Southern Pashtuns felt excluded from Karzai’s government. They viewed the December 2001 Bonn Agreement—the result of an internationally sponsored conference to decide the shape of Afghanistan’s constitution and government—as having favored their enemies, the Uzbeks, Tajiks, and Hazaras of the Northern Alliance. They felt that Karzai, though a Pashtun of the Durrani tribe himself, had never done much to address their concerns. Feeling disenfranchised, many had thrown their lot in with the Taliban.
The Obama administration’s initial reading of the crisis in Afghanistan was to blame it on the spectacular failure of the Karzai government, paired with wrongheaded military strategy, inadequate troop numbers for defeating an insurgency, and the Taliban’s ability to find a safe haven and military and material support in Pakistan. Of these, Karzai’s failings and the need to straighten out the military strategy dominated the discussion. Above all, the Afghan conflict was seen in the context of Iraq. The Taliban were seen as insurgents similar to the ones whom the United States had just helped defeat in Iraq. And what had defeated the insurgency in Iraq was a military strategy known as COIN.
COIN, shorthand for counterinsurgency, was not new, and it had a checkered past. The British had adopted it while fighting rebellious Boers in South Africa at the dawn of the twentieth century, then used it again in Malaysia in the 1950s. The French had employed a version of it at around the same time with less success, in Algeria, and America had tried it, disastrously, in Vietnam. COIN strategy recognizes that a rebel group does not always organize into regular military units or hold on to territory. Insurgents avoid fixed positions and hide among the people, denying them to the adversary. An insurgency wins by controlling people. Its center of gravity is not a place on the map, but its support base among a sympathetic (or at least cowed) population.
To defeat an insurgency, therefore, you must secure the populace. They must be shielded from insurgent violence and their trust gained. Only then will they stand against the insurgency and help with its defeat. The keys to COIN are small, socially and politically aware units; local cultural and linguistic knowledge; and good relations with civilians, whose loyalties are the real prize.
In Iraq, American troops had fanned out into districts and villages, setting up small posts from which they could mount patrols, see to security and governance at the local level, and squeeze the insurgency out of villages, towns, and entire districts in restive Anbar Province.18 It worked. As more and more places were freed from insurgent control, community leaders asserted their authority and joined hands to form the so-called Sons of Iraq. They took over local politics and security, and, with American financial assistance, rebuilt the local economy. American troops protected these leaders, but eventually it fell to Iraq’s own American-trained security forces to provide that security and help local leaders finish off the insurgency.
Success in Iraq crowned COIN as America’s military strategy of choice for winning “asymmetric” wars against terrorists, tribes, and what used to be called guerrilla fighters in failed, or failing, states. The relevance of Iraq to Afghanistan seemed self-evident. COIN had won Iraq; it was the right choice for Afghanistan.
Still, COIN requires governance, and governance requires a government. The Afghan government did not have the means (or the will) to follow the marines into areas cleared of Taliban to provide governance, and thus COIN could go only so far and no more. President Karzai would prove singularly instrumental in dashing America’s hopes of anything good coming out of Afghanistan. But in early 2009, Washington was hopeful it could knock some sense into him. Failing that, Afghan presidential elections in the fall of 2009 might return a better partner.
The strategy review proved torturously long. The president sat with his national security team through ten meetings—twenty-five hours—over the course of three months, hearing analysis and debating facts. There were many more meetings of those advisers and their staffs without the president present to dig further into the relevant issues, and go through stacks of folders, each one the size of a phone book, answering every question that came down from on high. At SRAP (the office of the Special Representative for Afghanistan and Pakistan), we managed the State Department’s contribution to the paper deluge. We worked long hours preparing memos and white papers, maps and tables, and then summaries for each. The Pentagon and CIA had their own stacks. In fact there was a healthy competition between the agencies over who did a better job of producing more.
Early in the process, Holbrooke came back from a meeting at the White House, called us into his office to give us an update, and said, “You did a good job, the Secretary [Clinton] was pleased with her material but wants her folders to be as big as [those of Secretary of Defense Bob] Gates. She wants color maps, tables, and charts.” Clinton, continued Holbrooke, “did not want Gates to dominate the conversation by waving his colorful maps and charts in front of everybody. No one reads this stuff, but they all look at the maps and color charts.” Everyone in the office looked at him. “So who does read all this?” I asked him, pointing to a huge folder on his desk. “I’ll tell you who,” he said. “The president reads them. He reads every folder.”