by Steve Coll
The most remarkable aspect of Holbrooke’s involvement in the election remained his indiscretion. On August 21, before any preliminary results had been released, but after Karzai had already claimed victory, Holbrooke organized a conference call with special representatives for Afghanistan from other N.A.T.O. and international governments. He held the call over open lines. From an intelligence collection perspective, Holbrooke might as well have stood in the middle of Kabul streets and spoken with a megaphone. Operatives from N.D.S. listened in and took notes. Karzai soon learned of the details.
Holbrooke laid blame for the mess on the Bush administration, because it had effectively forced the Obama administration to support a presidential election in this summer that was destined to make things worse. He emphasized that his fellow envoys should use their influence to make sure that no frustrated candidates or allied militias took their protests into the streets. “We have to respect the process,” Holbrooke said. The outcome of the vote “will be disputed” because of fraud allegations already being voiced by Abdullah, Ashraf Ghani, and other trailing candidates. The envoys’ common position should be to wait for a certified outcome, which might well include a second round of voting.
Only Abdullah had a prayer of winning in a second round, and even his prospects looked dim. Ghani, who had received less than 5 percent of the vote, urged Holbrooke into action, arguing that the election had been “entirely illegal.”22
The next day, Holbrooke returned to the Arg to meet Karzai. Citing the conference call they had just listened to, Karzai’s aides had informed the president that Washington and London were “pushing” for a second round. This may have been a distortion of Holbrooke’s comments but it reinforced what Karzai already believed. Furious, Karzai had already tried to telephone world leaders, including Obama, to protest outside interference in the election. His advisers assured him that he had crossed the 50 percent threshold and had been reelected—any effort to suggest otherwise was an American conspiracy to unseat him.
Karzai’s face darkened. He declared to Holbrooke, “I will not accept fake facts based on foreign interference.”23
—
It required two months and countless meetings and threats to resolve the election. The essential problem was simple in form but very difficult to fix. Even setting aside the massive fraud carried out by Karzai’s allies, the president had won close to half the vote. In a second-round face-off with Abdullah, Karzai would almost certainly prevail. Yet if the two election commissions overseeing the vote certified that Karzai had won outright in the first round, Abdullah and others “will say he stole it,” as Holbrooke put it.
Karzai held firm, telling just about every diplomat who met him that the West was “trying to defeat him.” Gradually, through September, the Obama administration came to accept reluctantly that Karzai would likely be president for another five years, and an even unhappier partner than before. “One way or another Karzai is going to be president of Afghanistan,” Holbrooke told a private meeting of former Clinton administration foreign policy officials and other specialists on Afghanistan on September 12. “It’s a fact.”24 The truth was that Holbrooke’s improvisations during the election had not removed Karzai and only destabilized further Karzai’s strained relationship with the Obama administration.
Karzai was also suspicious of Karl Eikenberry, the retired general who became U.S. ambassador to Kabul in the spring of 2009, succeeding Bill Wood. Eikenberry knew Afghanistan from his three tours in uniform but in meetings with Karzai as ambassador he was stiff and formal, marching through structured agendas and taking it upon himself to educate Karzai about how he should conduct himself as a statesman. Eikenberry had a well-grounded skepticism about Karzai but he was too much a general to manage such an insecure and moody client.
Senator Chuck Hagel and Vice President Joe Biden recalled the role John Kerry, now Biden’s successor as chairman of the Senate Foreign Relations Committee, had played in mollifying Karzai during their difficult dinner in the Arg early in 2008. Kerry flew in to Kabul in October. The plan was to persuade Karzai to accept the recount; if he could achieve that, he was confident that Abdullah would forgo a second vote, to spare the country the risk of violence and instability.
Karzai was under enormous pressure. His standoff with Abdullah revived the tension with the Panjshiris going back to 2002. Karzai’s Pashtun aides worried that the Panjshiri groups backing Abdullah might act rashly. Karzai’s closest bodyguards included two Panjshiris who walked beside him with loaded rifles. Panjshiri snipers manned positions on the palace rooftops. They could take him out in an instant. The situation recalled Indian prime minister Indira Gandhi’s vulnerability after she ordered a violent raid on a Sikh temple to root out armed separatists. Two of Gandhi’s Sikh bodyguards assassinated her. Karzai asked Rahmatullah Nabil, the trusted Pashtun who oversaw his personal protection, whether he should be worried. “Mr. President,” Nabil told him, “there are more than seventy Panjshiris around you. They are snipers, drivers, they are manning the I.D. checkpoints.” If he removed Panjshiris in the inner circle, he would “create a trust deficit” with all the others. That would only make things worse. In fact, to signal trust and his faith in the bodyguards’ professionalism, Nabil added Panjshiris to Karzai’s inner protection force. But the tension remained. Few presidential transitions in Afghanistan during the past century had occurred bloodlessly.25
Kerry walked with Karzai in the Arg gardens to try to resolve the stalemate. He discovered, however, that Zal Khalilzad was staying in the palace, dining with Karzai and counseling him as a friend. Khalilzad was now a private citizen, and he was in Afghanistan to work on a foundation to support education in the country. He always stayed at the palace as Karzai’s guest, because of their friendship and for the sake of security. Yet the situation bordered on the absurd: Both Kerry and Khalilzad had access to Karzai at a decisive moment of crisis. Kerry told aides that he was unhappy about Khalilzad’s presence because of the risk that it would create multiple channels of communication and confusion about who was talking to whom. He feared that whatever arguments he made to Karzai during his five-day marathon of garden walks, Khalilzad would unpack them over dinner and complicate Karzai’s thinking.
Khalilzad met with Karl Eikenberry, who told him that the policy of the Obama administration was that no one had won the first round of the election; that there should not be a second round; and that, instead, Karzai should remain as president of Afghanistan and should appoint Abdullah as chief executive, with powers that would be negotiated with the help of the United States and the United Nations.
Kerry asked to meet Khalilzad at the U.S. ambassador’s residence. They went out on the roof and talked. Kerry said his mission was to persuade Karzai to agree that he had not won the first round, to skip the second round, and to work out a power-sharing deal with Abdullah. Khalilzad advised against this strategy. He did not think Kerry could possibly persuade Karzai to go along.
Kerry said that President Obama had approved his approach. Then you should reengage with the White House and get new marching orders, Khalilzad insisted. Kerry said that would be difficult and asked Khalilzad to help him with the plan he had. Khalilzad declined. They agreed to remain in touch, but recognizing that Kerry was not receptive to his advice, Khalilzad left Kabul.
Kerry had always believed that Karzai was a nationalist at heart, and that the best course would be to appeal to his sense of Afghanistan’s national interest. During their walks, he cited his own decision to accept his close defeat in the 2004 American presidential election, when some supporters had urged him to challenge electoral snafus in the decisive state of Ohio. “Look, we’ve all had some tough decisions to make about the outcome of elections,” Kerry said. In the end, Karzai agreed to concede that he had not won the first round, and to allow the Obama administration to persuade Abdullah to stand aside, to avoid the violence and chaos of a second round. He would not conside
r the power-sharing plan Kerry had in mind.26
Despite their outrage at Karzai, Abdullah and other opposition powers such as Governor Noor had little incentive to attempt a coup. The Obama administration and N.A.T.O. and other allied governments had promised a major escalation of aid and military support if the election could be sorted out, a massive infusion of funds, manpower, and technology that might improve security and would certainly create economic opportunity for Kabul’s elites. Twelve days after Kerry persuaded Karzai to agree to a second round of voting, Abdullah withdrew and declared in a press conference in his Kabul garden that he would lead the opposition to Karzai’s second-term government peacefully.
To try to repair the damage with Karzai, the C.I.A. dispatched Greg Vogle for another tour as Kabul station chief soon after the election was settled. Vogle remained close to McChrystal, the new American war commander. Karzai might be a maddening partner, but any objective reading of his performance in 2009 had to account for the fact that he had outwitted his American doubters, including Holbrooke. He would be president for five more years. His half brother remained in power in Kandahar. His supporters in the south and east had gotten away with fraud. He had established himself with N.A.T.O. governments as intractable and independent minded, no longer pliant or passive, with new room to maneuver in Afghan domestic politics. The U.S. embassy, the State Department, and the White House might be frustrated and even disgusted with Karzai, but through his personal relationships with Vogle and McChrystal, he had also renewed private channels to the C.I.A. and the Pentagon, legendarily the true centers of American power. It was an outcome that other South Asian politicians with a bent toward conspiratorial thinking about America could only envy.
Holbrooke understood his own marginalization in Kabul. The State Department would be a “backseat driver in Afghanistan,” he told his aides. American policy in the country was becoming a “runaway car” steered mainly by the Pentagon, under Petraeus’s sway. He still saw room for diplomacy. With Mike Mullen, he could work to change the American relationship with Pakistan. And on his own, protected by compartmented secrecy, he could try to negotiate a way out of the war directly with the Taliban.27
TWENTY-TWO
A War to Give People a Chance
In June 2009, Major General Michael Flynn arrived in Kabul as Stan McChrystal’s J-2, or chief intelligence officer, in the Afghan war command. Flynn was a dark-haired, wiry, direct-speaking Rhode Islander who had grown up in a roughhousing family of Irish American brothers. He had been at McChrystal’s side since 2004, largely engaged in door-kicking special operations in Iraq and Afghanistan. Those hard experiences had schooled Flynn about tactical intelligence for counterterrorist and counterguerrilla operations—tracing insurgents, managing interrogations, the surveillance of routes for attack or patrol, and looking out for enemy infiltrators.
Sunni and Shiite militias in Iraq, like the Taliban in Afghanistan, had bled American forces by embedding in local populations, deploying suicide bombers, and implanting improvised explosives where soldiers patrolled. By 2008, in much of Baghdad, at least, McChrystal and Flynn, along with Petraeus and his theater high command, had stitched together “I.S.R.” systems—for intelligence, surveillance, and reconnaissance—that allowed commanders on secure bases to watch on television screens as guerrillas in nearby neighborhoods buried mines or maneuvered for ambushes. The surveillance systems included drones, blimps with digital cameras tethered to ground stations, blimps tethered to patrolling armored vehicles, and more traditional reconnaissance aircraft. Their continuous video feeds shone on flat screens in “fusion centers” filled with targeting analysts. The setup resembled the Global Response Center at C.I.A. headquarters in Langley or the joint intelligence operations center at Central Command in Tampa. The analysts in Iraq had developed what McChrystal termed an “unblinking stare” at active battlefields or risky neighborhoods. On the inside, the facilities looked every bit as high tech and spookily omniscient as the imagined versions conjured up in Hollywood thrillers like the Mission: Impossible series. The I.S.R. systems had not provided a decisive answer to Al Qaeda in Iraq or Shiite militias but they had certainly helped improve the safety and offensive lethality of American and allied Iraqi forces in Baghdad.
In Kabul, as he set up shop around the rose gardens and hammered-together trailer parks of I.S.A.F. headquarters in the summer of 2009, Flynn was stunned to discover how comparatively weak the Afghan battlefield’s I.S.R. systems were. American soldiers and Marines, as well as British, Canadian, Dutch, and other European soldiers, were dying and suffering catastrophic injuries in I.E.D. strikes because they could not see from hour to hour what was happening around their bases or on the roads they were ordered to patrol.
In Kabul that summer, a single, highly visible, $20 million aerostat blimp hovered above Bala Hissar, a fourth-century fortress. Drifting on a white line over crumbling mud-rock parapets, the blimp presented an ominous image of N.A.T.O. technology, something from science fiction. In fact it was a fairly straightforward aerial surveillance machine. Its cameras were linked to a ground station that fed the imagery to I.S.A.F. operations centers, providing live coverage of Kabul streets around the clock, much like security cameras in office buildings or subways in New York or London. But elsewhere in Afghanistan, there was not a single blimp in operation. In addition, that summer, I.S.A.F. had but one or two “lines” of drones, meaning pairs of machines and linked ground stations that could provide continuous, twenty-four-hour surveillance wherever the drones were directed to fly. And, of course, in comparison with Iraq, the Afghan battlefield was much larger and more dispersed. Flynn’s first advice to McChrystal was “We are blind on the battlefield and we had better move fast.”
Flynn worked his classified e-mail and phones to call in favors from colleagues scattered worldwide. He tried to speed up the Pentagon’s procurement and supply lines. He had an ally in Robert Gates, the defense secretary, who had ordered a crash increase in drone and other I.S.R. production when he arrived at the Pentagon in 2007. Yet it had taken time for manufacturing to catch up with his orders, and two years later supply was still inadequate.
Like a manic project manager building a secret television network, Flynn ultimately imported about 175 fixed and mobile aerostat blimps to Afghanistan. He became partial to the smaller aerostats that could be tied to patrolling armored vehicles and toted across the vast stretches of southern Afghanistan, providing over-the-horizon reconnaissance. (The blimps were visible, tempting targets for the Taliban and other Afghans with guns. Small rounds from assault rifles generally didn’t damage the aerostats severely, but a rocket-propelled grenade could. Once, Flynn’s team was visiting Kandahar when an intelligence unit reeled down a blimp for maintenance and found an arrow stuck in its skin.)
Flynn also created secret flat-screen-filled fusion centers packed with targeting analysts in Kandahar, Mazar-i-Sharif, and Herat. Simultaneously, I.S.A.F. and the C.I.A. expanded the number of Omega bases along the Pakistan border to about eight and linked them to drone and other surveillance and cell phone intercept systems. When Flynn and McChrystal first arrived in June, American forces had perhaps two full “strike packages” for offensive raids against Taliban commanders or units. (A strike package was a sizable integrated unit that included not just the men and helicopters to carry out lethal raids or searches but dedicated helicopter support, artillery units, interpreters, and intelligence.) They soon increased the number of strike packages to about fourteen.1
Yet as Flynn and his intelligence aides flew about Afghanistan building out infrastructure to support ramped-up kill-or-capture operations against the Taliban, they also became increasingly disquieted by I.S.A.F.’s lack of political and social insights about the enemy. Since 2002, for all the captains and Green Berets who had taken tea and chatted about tribes and development with local elders, the I.S.A.F. intelligence collection and analysis system remained overwhelmingly focused on tactical eve
nts—a shoot-out here, a bombing there, or reports on the component parts of the latest I.E.D. that had blown up. As Flynn put it starkly in a paper he coauthored in 2010, criticizing not only Pentagon intelligence but the C.I.A. and National Security Agency as well:
Eight years into the war in Afghanistan, the U.S. intelligence community is only marginally relevant to the overall strategy. Having focused the overwhelming majority of its collection efforts and analytical brainpower on insurgent groups, the vast intelligence apparatus is unable to answer fundamental questions about the environment in which U.S. and allied forces operate and the people they seek to persuade. Ignorant of local economics and landowners, hazy about who the powerbrokers are and how they might be influenced, incurious about the correlations between various development projects and the levels of cooperation among villagers . . . U.S. intelligence officers and analysts can do little but shrug in response to high level decision-makers seeking the knowledge, analysis, and information they need to wage a successful counterinsurgency.2
It was a stark, impolitic, even shocking summary of the American position in Afghanistan. The paper’s boldness was characteristic of Flynn, a man not prone to self-censorship. It raised at least three fundamental questions about the Obama administration’s incipient war strategy. If kill-or-capture operations were inadequate by themselves and if I.S.A.F. lacked basic insights into how those raids affected the social and political landscape, why was the United States doubling down on lethal operations? Was I.S.A.F. actually capable of acquiring the subtle insights Flynn listed concerning a society as complex and opaque to outsiders as Afghanistan’s? And what were the most important insights about the Afghan public’s likely reaction to a surge of U.S. forces? Nobody had ever asked them whether they wanted so many ground troops in the first place.
