COIN Versus CT

20. COIN Versus CT

On January 20, 2009, Barack Obama moved into the White House with a path already laid for an exit from Iraq. Two months earlier, the Bush administration and Nouri al-Maliki’s government had signed a Status of Forces Agreement, which required all American troops “to withdraw from all Iraqi territory, waters, and airspace no later than the 31st of December of 2011.” The details of the transition still needed to be worked out, but the American side of the war was practically over.

Now, though, the other, forgotten war, in Afghanistan, was spiraling out of control.

During his presidential campaign, candidate Obama had repeatedly called for sending more troops to Afghanistan, which he called the center of global terrorism. As far back as October 2002, when he was still an Illinois state senator, Obama, in a speech against the then-impending invasion of Iraq, had said, “I don’t oppose all wars. What I am opposed to is a dumb war . . . a rash war, a war based not on reason but on passion, not on principle but on politics.” Iraq was the dumb war. But there was a smarter war that Obama favored waging. “You want a fight, President Bush?” he said in that same speech. “Let’s finish the fight with bin Laden and al Qaeda.”

When he gave that speech, serious fighting hadn’t yet resumed in Afghanistan. But by the time Obama was running for president, the Taliban had come out of hiding and were attacking American and NATO troops in the eastern and southern parts of the country, and so he spoke out explicitly for sending in more troops. Many thought he was playing politics: Iraq was Bush’s war, and therefore bad; Afghanistan was the war that Bush had neglected, and therefore good. There may have been some truth to this, but another, substantive factor was at play: the fact that one of Obama’s foreign-policy advisers was Bruce Riedel.

Riedel was a recently retired career CIA analyst with a specialty in South Asia and a deep knowledge of terrorist groups in Afghanistan and Pakistan. Still a consultant, and plugged into a network of associates, Riedel knew that the Taliban fighters were on the rise in Afghanistan, that they had ties to al Qaeda, and that a restoration of their rule in Kabul could destabilize neighboring Pakistan, which possessed a nuclear arsenal and an intelligence service rife with Islamist extremists. Riedel joined the Obama campaign because of his speech against the war in Iraq, which he opposed for the same reason as Dave Kilcullen and other strategically minded analysts: because it distracted resources and attention from the real center stage, Afghanistan and Pakistan. Obama heeded Riedel’s briefings on the dangers in that part of the world; his call for sending more troops there was based on more than electoral calculations.

Soon after he won the November 2008 election, Obama asked Admiral Mike Mullen, the chairman of the Joint Chiefs of Staff, to come talk with him at his transition headquarters in Chicago. Mullen told him the same thing about Afghanistan that Riedel had, with the additional admission that the Bush administration had devoted too few resources to the region (they’d all been diverted to Iraq) and had devised no strategy for it whatever.

In early January, Obama sent his vice president–elect, Joe Biden, to Afghanistan on a fact-finding mission. Biden had long been a senior member of the Senate Foreign Relations Committee; this was far from his first trip. When he returned, he told Obama, with his usual theatricality, “I bring back one big headline for you: if you ask ten of our people over there what we’re trying to accomplish, you get ten different answers.”

So, one week after his inauguration, Obama phoned Riedel and asked him to chair a White House review of America’s policy in Afghanistan and Pakistan. He would need the report in sixty days, in time for a NATO summit in early April. Riedel had told Obama that he wanted no administration job, but he agreed to do this.
Riedel’s report painted a gloomy, though not quite dire, picture. His bottom line: “We are losing the war in Afghanistan, but it is not yet lost.” He rejected Bush’s notion of creating a Western-style democracy in Kabul and recommended instead a set of challenging though feasible strategic goals: to “degrade, dismantle, and destroy al Qaeda”; to train the Afghan military, so that it could secure the country after Western troops leave; and to stabilize Afghanistan, build up its government, and thus leave no pockets of anarchy for al Qaeda to exploit.
The best way to accomplish these goals, his report concluded, was to pursue, at least in the southern part of the country, a “fully resourced counterinsurgency strategy.”
Riedel had no background in counterinsurgency. He was influenced on this point by one of the key members of the interagency group that worked on the report—General David Petraeus.

•  •  •

Petraeus had wrapped up his tour as commander in Iraq the previous September on a note of triumph. George Bush promoted his deputy, General Ray Odierno, to take over the job of commander, and he put Petraeus in charge of US Central Command, which supervised all American military operations in the Persian Gulf and Southwest Asia, including Iraq and Afghanistan. But the command’s headquarters were at MacDill Air Force Base in Tampa, Florida, giving Petraeus closer access to the powers in Washington.

The Riedel group had started meeting in February 2009, just weeks after Obama took office. As with most policy-review groups of this sort, it included second- and third-tiered officials from all the agencies with any stake in the policy: CIA, AID, Justice, Agriculture, and several others. But the three main players, besides Riedel, were Richard Holbrooke, the State Department’s envoy to Afghanistan and Pakistan; Michèle Flournoy, the undersecretary of defense for policy; and Petraeus, representing the military.

Flournoy backed Petraeus on a counterinsurgency approach with some authority. The two had known each other from the COIN field manual workshop at Fort Leavenworth three years earlier and from the Clinton administration’s policy reviews of Haiti back in the mid-1990s. In 2007 Flournoy had cofounded a Washington think tank called the Center for a New American Security. Its purpose was to build bridges between Republicans and Democrats on defense policy, but the two wars going on happened to be insurgency wars, so it evolved into a locus of thinking and writing on counterinsurgency; some called it “COIN Central.” (Fittingly, when John Nagl retired from the Army in 2008, he joined the center as a senior fellow; when Flournoy left to join the Obama administration, Nagl replaced her as the center’s president.) Flournoy thus approached the Riedel group’s tasks as an analyst well versed in COIN doctrine and, though not as fervent as some, an advocate.

On March 12, Obama’s top national-security advisers met to discuss the Riedel report and touched off an argument that wouldn’t be settled completely for another three years.

Joe Biden instigated the dispute. Biden could go off on tangents at self-indulgent length; many saw him as a blowhard. But Obama had privately asked his vice president to raise questions in these sorts of meetings, to challenge every assumption, to highlight a policy option’s unstated implications. Obama was a famously fast learner, but military issues were new to him. And he was preoccupied, to the exclusion of nearly everything else, with the economy, which was plunging in steeper free fall than anyone had calculated just a few months earlier.

So Biden asked hard questions and, in this case, not merely as a devil’s advocate. He’d first been elected to the Senate when he wasn’t quite thirty, in 1972, during the final throes of the Vietnam War, and so he was suspicious of counterinsurgency, suspicious of any sort of military escalation based on a theory. If the strategic goal was to defeat al Qaeda, whose leaders were based mainly in Pakistan, he asked, why should we deploy a lot of troops to do nation-building in Afghanistan? Why not just send drones and small teams of Special Ops forces to go directly after al Qaeda fighters along the border?

Admiral Mullen, the Joint Chiefs chairman, explained patiently that chasing insurgents wouldn’t solve the problem and, worse still, would allow them the initiative. The best way to defeat insurgents, he said, reciting the standard COIN argument, was to protect the population, earn its trust and loyalty, and thus dry up its support for the insurgents.

Biden said he understood all that, acknowledged that Mullen’s case had logic. But, he argued, this course wasn’t politically sustainable; it would require a lot of troops, a lot of money, a lot of time, and the American people wouldn’t put up with it.

At this point, Hillary Clinton, Obama’s erstwhile rival in the Democratic primaries and now his secretary of state, spoke up. She thought that the mission was sustainable, that the American people would put up with it, if we succeeded at meeting our objectives.

Mullen seemed to win the day on substance. Biden, besides being the sole outright dissenter in the room, had objected mainly on grounds of domestic politics; and one could argue that Clinton—who’d crushed Biden in the Democratic primaries and had spent eight years in the White House as a very policy-driven First Lady—possessed a firmer grasp of domestic politics than he did.

Obama endorsed the Riedel report, but with caveats. He didn’t take his vice president’s side, but he shared his apprehensions. So, importantly, did his secretary of defense, Robert Gates.

When Gates agreed to replace Donald Rumsfeld in the final two years of the Bush administration, he had no intention of staying any longer. He’d taken the job in part as a favor to the Bush family, in part out of a patrician sense of obligation to the nation in a time of crisis. And he found the experience wearying. He hated Washington and longed to get back to his two houses in the Pacific Northwest: one lakeside, the other overlooking the ocean. A friend had given him a digital “countdown meter” displaying how many days remained before Bush’s term was over, and Gates showed it to visiting legislators, journalists, and lobbyists with mischievous glee. The sentiment was far from reciprocal. Gates was respected in all corners of the capital for his experience and well-worn gravitas. Several senators, on both sides of the aisle, hoped out loud that he would stay on, no matter who won the upcoming election. Gates discouraged all such speculation, telling one reporter, “The circumstances under which I would do that are inconceivable to me.”

He later confessed that his dismissiveness had been part of a “covert action” (a serious phrase for a man who’d spent the bulk of his career in the CIA) to convince everyone that he didn’t want to stay, so that no one would ask him to. But he knew all along that if the president did ask, he would feel duty-bound to comply. Besides, as much as he hated to admit it, he was growing comfortable in the job; he liked being in the thick of things and having influence with those at the zenith of national power.

Finally, when he met with Obama, he found himself, to his surprise, liking him. Gates was a Republican, but he saw in this Democrat a man with a similar temperament: a thinker and a reformer, but one with a pragmatic bent and an impatience for party dogma of whatever stripe. He sensed that he could work with Obama and make a difference. All those speeches Gates had given under Bush, outlining the changes that his successor should put in place—it might be best if he were that successor. Gates agreed to stay on, for one year; he ended up staying for two and a half.

For all his agitated efforts to change the Pentagon, Gates was at heart a conservative. He killed or slashed thirty weapons systems in his first two years as Obama’s defense secretary, including some of the services’ dearest projects (the Air Force’s F-22 stealth fighter and the Army’s Future Combat Systems vehicle, among others), but he had no interest in cutting back the military’s size or America’s global reach.

Yet when it came to Afghanistan, Gates was leery. He’d been deputy director of the CIA when the Soviet Union crashed and burned on the country’s rugged terrain, despite nearly a decade of pummeling the mujahedeen rebels with all the firepower it could muster. And he’d been the Agency’s director when, partly as a result of that fiasco, the once-mighty Communist empire collapsed. Contemplating America’s options, he had to consider that there was still some way to go in Iraq; he wasn’t at all eager to step up yet another war in another Muslim nation.

Gates didn’t say much at the White House meeting on Riedel’s report; he rarely said much in large meetings, holding his fire for key moments or, better yet, for the smaller sessions, preferably face-to-face, with the decision-maker. But at Senate hearings a month and a half earlier, Gates had testified that he would be “deeply skeptical” of a request to deploy a lot more troops. It would be best, he said, to keep America’s “footprint” relatively small. If the Afghan people begin to see the war as our war, not theirs, he warned, “we will go the way of other imperial occupiers.”

There was one other aspect of Afghanistan that weighed on President Obama: the war’s cost. His senior advisers met one last time on March 17 to approve the Riedel report. The next day, Obama flew on Air Force One to Los Angeles for a political event and to appear on Jay Leno’s Tonight Show. His chief of staff, Rahm Emanuel, told Riedel to get on that plane; the cross-country flight would pose a perfect opportunity for a discussion about the report—and, more to the point, a discussion with no generals present.

A couple of hours into the flight, Riedel was called into the president’s cabin to go over the whole report: its grim analysis of the situation in Afghanistan and Pakistan, and its twenty recommendations of what to do about it. But one remark in his briefing seemed to impress Obama particularly. Riedel told him that the annual cost of arming and supporting one American soldier in Afghanistan was $250,000 (he’d been given the number by the Pentagon, but, as he discovered later, the real cost was at least three times as much). By contrast, the cost of one Afghan soldier was $12,000. You could pay twenty Afghan soldiers for the same price as a single American soldier.

Riedel could practically see a lightbulb switch on above the president’s head. This would be the logical exit strategy: to turn the fight over to the Afghans. And the calculation—this ratio of the two soldiers’ costs—would be a sobering constraint, especially given America’s financial crunch, on any pressure to send tens of thousands of more US troops into battle.

In the end, buffeted by the crosswinds of the military’s advice, Biden’s dissent, Gates’s qualms, and Riedel’s recital of the costs, Obama took a middle course.

At a press conference on Friday, March 27, one week before the NATO summit in Strasbourg, Obama announced what he called a “comprehensive strategy” for Pakistan and Afghanistan; but, in fact, it was not a strategy at all. His declared goal for the war—taken straight from Riedel’s report, and a sober scaling-back from Bush’s dreams of erecting a Western-style democracy with the full panoply of human rights—was simply “to disrupt, dismantle, and defeat al Qaeda in Pakistan and Afghanistan, and to prevent their return to either country in the future.” How to do this—whether to pursue the COIN approach, which had been endorsed by nearly everyone in his cabinet, or Joe Biden’s more limited counterterrorism approach, which some dubbed CT—the president didn’t exactly say. Or, rather, he straddled both.

Even before the Riedel report was finished, Obama had approved an urgent request from the Joint Chiefs for twenty-one thousand more troops to be sent right away to Afghanistan, on top of the thirty-seven thousand already there. Presidential elections in Afghanistan were scheduled for August. One reason the Taliban insurgents were gaining strength was that Hamid Karzai—who had been installed in his office by Western powers—wasn’t seen by many of his own people as quite legitimate. The elections could change that, but first there had to be elections, and they had to be seen as fair. The areas around polling places had to be secure so that people would feel safe going to vote. More troops were needed for that security, and only the United States could provide them. To get them there in time for the election, they had to be mobilized now. Obama had hoped to wait for Riedel’s report before making any decisions on troops, but the request had logic; Riedel thought so, too.

Now, at his March 27 press conference, Obama declared that these extra troops “will take the fight to the Taliban in the south and the east and . . . go after insurgents along the border” between Afghanistan and Pakistan—with one exception, this urged on him by Bob Gates: one brigade of the 82nd Airborne Division, four thousand troops in all, would beef up the training of the Afghan army.

Had he stopped there, the new policy would have amounted to an acceptance of Biden’s CT strategy and a desire, perhaps prompted by Riedel’s economic argument, to jump-start the transition to the Afghans.

However, Obama went on to say, “To succeed, we and our friends and allies must . . . promote a more capable and accountable Afghan government.” This goal, he said, will require “a dramatic increase in our civilian effort,” including the deployment of “agricultural specialists and educators, engineers and lawyers,” all of whom must, of course, be protected by soldiers as they go about their work. He also said that the United States and its NATO allies must compel Karzai’s government to reduce the “corruption that causes Afghans to lose faith in their own leaders.”

This part of the speech sounded like an endorsement of COIN. In fact, at a press briefing afterward, Michèle Flournoy called the new policy “very much a counterinsurgency approach,” although she added that military operations against “high-value targets”—meaning al Qaeda and Taliban leaders—would also remain “a central part of this mission.”

The problem was that twenty-one thousand extra troops wouldn’t be enough to carry out both CT and COIN strategies. Besides, these troops would have little to do with either mission; they were being sent to boost security before the election. One thing the Riedel report did not deal with, at all, was how many troops would be needed to do whatever it was that the president decided to do: this sort of calculation lay outside Riedel’s lane and beyond his competency; he left it to the military.

Obama put off all these questions until after the Afghan elections—a deferral that stemmed from his indecision on the basic question but also made sense. If, as he’d been told, a key issue in COIN was whether the host government was seen by its people as legitimate, it would be wise to wait for the results of the election—to see whether or not they legitimized Karzai or whichever candidate won—before upping the stakes any further.

The president went off to Strasbourg, his first official trip to Europe, to declare the new policy and to rally allied support. Then he went home to deal with his own country’s gravest economic crisis since the Great Depression.

•  •  •

Meanwhile, Bob Gates continued to deal with all the relevant players, including the US commander in Afghanistan, General Dave McKiernan. For some time, McKiernan had a request on the table for an additional ten thousand troops beyond the twenty-one thousand that Obama had approved. But the more Gates talked with him, the more he realized that McKiernan had little idea of what he would do with them. He was devoting all of his effort to battling al Qaeda and Taliban militants along the eastern border with Pakistan, while doing almost nothing in the southern cities, where the Taliban were strengthening their hold, especially in Kandahar. McKiernan also acknowledged that he’d been devoting all his intelligence assets to pinning down the locations of the Taliban fighters, and none at all to analyzing the Afghan people: their tribal allegiances, sectarian splits, or any of the other sociocultural tags that would have been vital inputs in a counterinsurgency campaign. Petraeus, who also spoke often with Gates, was particularly annoyed at this neglect, which amounted to a violation of his own directive as CentCom commander to shift more to a COIN strategy.

McKiernan wasn’t entirely to blame; he barely had enough resources to fight along the border. But he was seen as old-school Army, an armor officer who hadn’t made the conversion to the new way of thinking that some other tank men, such as Pete Chiarelli, H. R. McMaster, and John Nagl, had made.

On May 11, after consulting with Mullen, Petraeus, Clinton, and the president, Gates announced that he was relieving McKiernan of duty one year before his term in Afghanistan was due to expire—effectively ending his Army career. It was the first time that an American wartime commander had been fired since Harry Truman dismissed General Douglas MacArthur during the Korean War a half century earlier. (It was the second time Gates had fired any sort of general, but the earlier dismissal, handed to T. Michael Moseley, the Air Force chief of staff, had been on grounds of malfeasance.) At a press conference, Gates said that McKiernan hadn’t done anything wrong specifically; it was just that the war needed “fresh thinking.”

By “fresh thinking,” Gates meant a working knowledge of COIN. He announced at the same news conference that McKiernan’s replacement would be General Stanley McChrystal, whose main qualification, Gates said, was his “unique skill set in counterinsurgency.”

•  •  •

Several officers who’d worked with McChrystal were puzzled by his appointment. As far as they knew, he had no skill set in counterinsurgency, much less a unique one, at least not as the COINdinistas defined the term. He was an exceptionally creative officer, but his creativity seemed to reside in the art and science of finding and killing bad guys—and, at least in Iraq and Afghanistan, he hadn’t done this by winning hearts or minds.

McChrystal came from a military family: his father was a two-star general who’d fought in Korea and Vietnam; all four of his brothers enlisted in the Army; even his sister wound up marrying an Army officer. But Stan was the only son who’d gone to West Point, graduating in the class of 1976. Afterward, he entered the force as a parachutist in the 82nd Airborne, then rose through the ranks in Ranger and Special Forces units, climaxing in the fall of 2003, when he took control of the Joint Special Operations Command. JSOC was the most clandestine branch of the American military, the black-ops ninjas—Delta Force, Navy SEALs, Night Stalker pilots, and various specialized Task Forces identified only by their numbers (TF 11, TF 6-26, TF 121, among others)—that swooped down on terrorists in the dead of night. After the September 11 attacks, JSOC expanded enormously. In September 2003, Donald Rumsfeld signed an executive order authorizing the command to take any action against al Qaeda without prior approval of the president or secretary of defense and without notifying the intelligence committees of Congress.

McChrystal became commanding general of JSOC the same month that Rumsfeld’s order took effect. Still, for all his command’s power, he found its reach limited. Each branch was cut off from one another, not to mention from the rest of the military and the national intelligence agencies. McChrystal saw that al Qaeda was a network, each cell’s powers multiplied by its ties with other cells. It would take a network to fight a network, so McChrystal built one of his own.

Before taking command at JSOC, he had worked in the Pentagon as the Joint Staff’s vice director of operations. He saw how different elements of the national-security bureaucracy did or did not connect. Applying those lessons, he now reached out to the CIA, the National Security Agency, the National Geospatial-Intelligence Agency, regional specialists from the State Department, and the top officers of Central Command: making deals and exchanging services, while at the same time remaining sensitive to each fiefdom’s turf. As a result, when JSOC was searching for a terrorist, tracking an insurgent, or planning and executing a raid, everyone involved had continuous access to a vast range of real-time intelligence—drone imagery, satellite signals, cell phone intercepts, and much more—as well as the support of conventional Army forces. By fusing all this intelligence, McChrystal’s men in Iraq were able to perform dozens of raids a night; it was his team that found Saddam Hussein’s hideout and that killed Abu Musab al-Zarqawi, the leader of al Qaeda in Iraq.

McChrystal was driven, even obsessed. It was well known that he ran seven miles a day, ate just one meal, and slept only four hours a night. After his men killed Zarqawi, he insisted on personally inspecting and identifying the corpse.

JSOC’s nature—elite, hard-core violent, and unaccountable—sometimes sowed a toxic atmosphere. Several raids targeted the wrong house, killing innocent people. A top-secret detention center known as Camp Nama, run by JSOC’s thousand-man Task Force 6-26 at Baghdad airport, came to resemble a torture chamber, so brutal that the US Army stepped in and disciplined thirty-four of its members. McChrystal wasn’t involved personally in its operations; the first time he toured the center and saw what was happening, he muttered, “This is how we lose.” After the disciplining, he set new rules for interrogation. Still, he had at least tolerated the climate that produced such practices. Members of the Task Force had posted placards all over the center reading, “No Blood, No Foul.” In other words: feel free to do anything to a prisoner, as long as the pain and bruises you inflict don’t make him bleed.

And yet McChrystal knew more about counterinsurgency than his career path suggested. As a cadet at West Point, his favorite course had been the History Department’s offering on Revolutionary Warfare—the academy’s sole concession, at the time, to the study of COIN—which included readings from Galula, Mao, The Centurions, and the assorted classics that David Petraeus and others had also read in their formative years.

When he was named commander of US and NATO forces in Afghanistan, he knew it was time to reactivate that aspect of his education. Up till then, though, McChrystal’s experience in this side of insurgency warfare had been, literally, academic, derived strictly from reading books. He hadn’t maneuvered to find the right balance, hadn’t navigated the shoals between killing and co-opting, that Petraeus and others had done in sustained operations—and, even then, with uneven success—in Iraq. The consequences would soon show.

•  •  •

McChrystal won Senate confirmation as commander on June 10. The first thing he did was to emulate his old friend and fellow West Point alumnus Dave Petraeus: he recruited a group of experts to come to Afghanistan for a month and write an assessment of the situation. Petraeus had put together these groups to get political buy-in for what he’d already decided to do. McChrystal had the same idea in mind, but he also really wanted the outsiders’ advice and reality-check.

The questions he asked the “assessment team” to answer came in a directive from Petraeus himself, in his role as head of Central Command: Can ISAF—the International Security Assistance Force, the formal name of NATO’s coalition in Afghanistan—achieve its mission? If so, how should it go about the task?

In choosing his team, McChrystal again followed Petraeus’s example by picking a dozen specialists, many of them PhDs from leading Washington think tanks. Among them were Fred Kagan, the former West Point history professor at the American Enterprise Institute, who’d made the crucial case for a surge in Iraq; his wife, Kimberly Kagan, who’d since formed her own think tank to tout and document the surge; and Stephen Biddle of the Council on Foreign Relations, who’d served on a similar team when Petraeus took command in Iraq.

McChrystal asked John Nagl to participate, too, but he had just taken over as president of the Center for a New American Security. Most of its top analysts—not just Michèle Flournoy, whom he’d replaced—had gone off to join the Obama administration; so he was too busy raising money and recruiting new staff, both unfamiliar tasks. Instead, he offered up one of his new hires, Andrew Exum, an intense, scholarly former Special Ops officer who’d not only earned a master’s degree at the American University of Beirut and a PhD in war studies at the University of London but had also commanded Ranger platoons in Iraq and Afghanistan.

The team visited American and NATO troops all over Afghanistan and spoke with high-ranking Afghan officials in Kabul. By the time of its interim briefing to McChrystal on July 4, barely a week after the members’ arrival, the picture was clear. The country was falling apart, the government was corrupt and incompetent, the Taliban were gaining strength and popular favor, ISAF had no coherent strategy, and its troops were “commuting to the war”—just as most American troops had done in Iraq before Petraeus—patrolling the streets by day, usually in armored vehicles, almost never getting out to walk among the people, and returning to their large, gated bases at night. If nothing was changed, defeat was inevitable.

There was unanimity on this grim finding, but not on what to do about it. A few of the team’s members, including Jeremy Shapiro of the Brookings Institution, came away from the tour deeply skeptical that the mission was feasible under any circumstances. To varying degrees, the others saw the prospects as dim but not hopeless and thought that the stakes were high enough to justify giving the mission a shot—as long as there was a surge in troops and a shift to a counterinsurgency strategy, the same formula that seemed to have turned things around in Iraq (though none of the experts had any illusions that the two countries were otherwise similar).

Eventually the dissenters were brought around, on the condition that the report made clear that success was by no means certain even if its recommendations were followed. Nobody had a problem with that. Even Fred Kagan, the most enthusiastic team member, put the chances of success—with an adequate surge and a shift to COIN—at less than fifty–fifty.

McChrystal was briefed on the study periodically. During the drafting stage (and the document went through a few drafts), he sat with the team and went over every line, rewriting many.

The final sixty-six-page report began with an attention-grabber: “The overall situation is deteriorating . . . Success is achievable, but it will not be attained simply by trying harder or ‘doubling down’ on the previous strategy.” There was an “urgent need for a significant change to our strategy and the way that we think and operate.” This “new strategy,” it went on, must be “an integrated civilian-military counterinsurgency campaign.”

In language familiar to anyone who had studied COIN, the report outlined the implications: “This is a different kind of fight . . . Our strategy cannot be focused on seizing terrain or destroying insurgent forces; our objective must be the population.” To gain the people’s support, the troops must maintain “a persistent presence” in the neighborhoods (the report called this approach “shape / clear / hold / build,” a slight rephrasing of the by-now-standard phrase “clear, hold, and build”) and cultivate a “better understanding” of their needs.

Victory was defined as “a condition where the insurgency no longer threatens the viability of the state.” This suggested a two-pronged approach: fighting the insurgency directly, when that was necessary; and, at least as important, bolstering the Afghan state, making it a viable, legitimate entity that fulfilled the needs of its people and thus preempted or co-opted the insurgents’ appeal.

McChrystal and his team realized that bolstering the state was a bigger task than battling the Taliban. Their insight about the nature of Karzai’s government was reminiscent of the realization that dawned on American commanders in Iraq rather late in the day about the nature of Nouri al-Maliki’s regime. “Progress is hindered,” they wrote, by “a crisis of confidence in the government,” owing to “the weakness of state institutions, malign actions of power brokers, widespread corruption and abuse of power by various officials.” All of these failings “have given Afghans little reason to support their government” and have created “fertile ground for the insurgency.” To win the people’s support, ISAF “must protect the people from both of these threats”: the insurgents and their own government.

Yet the essence of COIN was to provide for the people’s needs “by, with, and through” the local government. Therefore, ISAF would need to shore up the legitimacy of Karzai’s regime: strengthen its institutions, clean up its corruption, help it deliver basic services. And none of this could be accomplished simply by killing insurgents.

The report stressed that another obstacle, besides the Afghan government’s malfeasances, was ISAF itself, in two ways. First, ISAF was “a conventional force . . . poorly configured for COIN, inexperienced in local languages and culture.” Second, it was “under-resourced.” A new COIN strategy would require more money, more civilian experts, and more troops.

How many more troops, the report’s authors didn’t say, and deliberately so. Fred and Kim Kagan argued during the team’s deliberations that it should recommend a precise number of additional brigades, as Fred had done two-and-a-half years earlier in his AEI report on the Iraq surge. But Steve Biddle argued that the team had neither the access to data nor the analytical tools to make this sort of judgment. Fred did, in fact, know intelligence officers at ISAF headquarters who could have supplied him with the data and the tools, just as H. R. McMaster’s former staff aides from Tal Afar had done on his AEI study. But Biddle figured that he and the other members might be called upon to defend this report in articles, media appearances, or congressional hearings. (He well knew that this was one of the functions of these teams: the commander gets buy-in from outside experts, who in turn make statements or write articles that get buy-in from the public.) And he had no desire to defend a number without knowing how it was reached or whether some other number might be better.

Biddle’s view won out. McChrystal would make his recommendation on troop levels separately, after an analysis by his headquarters’ staff.

On August 30 McChrystal signed the team’s report, formally titled “Commander’s Initial Assessment,” and sent it to Secretary of Defense Robert Gates, who, in turn, passed along copies to the Joint Chiefs and the White House.

•  •  •

A few days before Gates received McChrystal’s assessment, he took a brief vacation, during which he caught up on some outside reading. Friends and colleagues were always sending him articles that they thought he should read, on top of the mounds of official memos and reports that filled his working days. Away from the mounds, he flipped through some of these articles, especially those on Afghanistan, a topic with which he was still wrestling. One in particular caught his attention, a piece in the American Enterprise Institute’s magazine, the Weekly Standard, titled “We’re Not the Soviets in Afghanistan,” written by Frederick Kagan.

Gates knew the Kagan name. He remembered the slides from Fred’s briefing that called for a surge in Iraq. He knew of the book comparing American and European foreign policy by Fred’s brother, Robert. And Gates had long regarded their father, Donald, as one of the great military historians of all time. Once, after learning that Kimberly Kagan would be briefing him on developments in Baghdad, Gates turned to an aide and asked, “How many of these fucking Kagans are there?”

But the Weekly Standard article’s subject matter was what drew Gates’s interest. He had been leery of expanding the mission in Afghanistan precisely because he’d seen close-up how badly the same course had turned out for the Russians. Yet Kagan was arguing that history didn’t have to repeat itself: when the Soviet army entered Afghanistan, it had virtually no prior experience at counterinsurgency—in fact, no experience at any sort of combat since World War II. Its tanks and artillery rolled in with brute force; its arsenals contained no precision weapons; its troops, many of them drunk and half-literate, killed civilians with no compunction, destroyed their villages, defoliated their crops. An American campaign, particularly a counterinsurgency campaign, would be very different; the results needn’t be the same.

Gates had forgotten the details of the Russians’ brutality in the war, the crassness of their strategy and tactics. Maybe, he thought, we could do things differently.

Then, a few days later, McChrystal’s assessment arrived, laying out a path of how to do things differently, advocating a COIN strategy, which Gates himself had been pushing the Army to adopt as mainstream policy. His views on Afghanistan began to shift; he opened up to the idea of escalation.

It was a crucial shift. Over the next few months, in the president’s meetings with his national-security team, Joe Biden remained the only outright opponent of McChrystal’s push for a surge and a switch to COIN. By this time, Gates and Obama had grown close. Obama respected his views, trusted his judgment; Gates was one of just three cabinet officers—the others were Hillary Clinton and Treasury Secretary Tim Geithner—who met with the president at the White House one-on-one every week.

At NSC meetings on a variety of topics, Gates would frequently emerge as a center of gravity. He didn’t speak much, but when he did, he shifted people’s opinions or at least altered the boundaries of what differences in opinion they’d tolerate, and coaxed them into a consensus with others whose positions they’d previously opposed. He’d honed this craft during his time as deputy national security adviser in George H. W. Bush’s White House. Right before a meeting of the various departments’ deputy secretaries, who routinely set an administration’s foreign policy agenda, Gates would sit down with his boss, retired general Brent Scowcroft, Bush’s national security adviser, and work out what position they wanted the group to adopt. Gates would then guide the meeting in such a way that all the deputies came to precisely that view and walked away thinking that it had been their idea.

Several aides who observed the national-security meetings on Afghanistan through the fall of 2009 later reflected that if Gates had remained skeptical of escalation, if he and Biden had opposed McChrystal’s recommendations, President Obama might well have settled on a different policy, involving fewer troops and a less ambitious strategy.

•  •  •

On September 13, two weeks after McChrystal submitted his assessment, President Obama brought together his top national-security advisers—the same officers, officials, and aides who had met in March after the Riedel report came in—to discuss what to do about Afghanistan.

Obama was ambivalent about the war. He had been persuaded that the stakes were high; that if the Kabul government fell, the Taliban would easily take over, with al Qaeda not far behind, and Afghanistan would devolve once again into a haven for terrorists. He’d decided that he would not pull out altogether.

Yet as a student of history, he knew well the pitfalls of escalation and overconfidence. That summer, he’d invited a group of presidential historians to the White House for an off-the-record dinner. They included Robert Caro, Robert Dallek, Doris Kearns Goodwin, and Michael Beschloss, all of whom had written books about Lyndon Johnson. The conversation turned inevitably to the Vietnam War, which had taken down that earlier progressive president, and some of the historians told reporters afterward that Obama seemed mindful of the potential parallels.

The specter of Vietnam haunted conversations throughout the White House and the Pentagon that summer. Skeptics of escalation in Afghanistan brandished copies of a new book called Lessons in Disaster by Gordon Goldstein, about how some of Johnson’s smartest advisers, including McGeorge Bundy (for whom Goldstein had later worked as a research assistant), led the nation into a tragic quagmire. Advocates of sending in more troops were seen carrying around Lewis Sorley’s A Better War, the revisionist tome, published a decade earlier, which contended that the Vietnam War could have been won if only the commanders had stuck to a COIN strategy from start to finish.

The September 13 meeting was, in many ways, a reprise of the debates—COIN, CT, or something in between—from the previous March. Eleven days later, McChrystal submitted his follow-on report, stating how many troops he needed to carry out his assessment’s strategy. Obama and his advisers met again six days after that, on September 30, to discuss those numbers. McChrystal had listed a range of options: ten thousand more troops if the president wanted merely to accelerate the training of the Afghan army; forty thousand for a counterinsurgency strategy with some risk, sending the troops first to fight in the south, then to push their way to the east; eighty-five thousand for a “more robust counterinsurgency” strategy across both regions simultaneously.

It seemed to some a standard ploy: lay out three options—one too small, one too big, and one in the middle, which, by process of elimination, looks just right. And as his “professional military judgment,” McChrystal recommended forty thousand—the number in the middle. But, in fact, he’d played it straight. His staff had calculated that the highest number, eighty-five thousand, while preferable in theory, was neither practical nor worth the extra effort: by the time the military built all the additional bases, airstrips, and other logistical networks to support so many troops, the course of the war would already be decided. McChrystal’s forty thousand was not a political compromise; it was his genuine preference.

Still, the reaction inside the White House was sticker shock. Obama had approved the military’s request for twenty-one thousand extra troops just six months earlier; they’d barely started to arrive in country, and now—already—the generals were hitting him up for twice as many more. If the situation didn’t improve, would they ask for another forty thousand the next year and one hundred thousand the year after that? This had all the earmarks of the endless escalation he most feared.

Clearly, he was going to send some increment of more troops, and however small or large it was, the Afghanistan War would from that moment on be “Obama’s war.” He’d called it the war of necessity on the campaign trail, as distinct from Bush’s war of choice in Iraq. He’d authorized his secretary of defense to fire the commander he inherited, General McKiernan, on the grounds that he lacked a winning strategy. This new recommendation of forty thousand more troops and a COIN strategy came from the commander that Obama had hired as McKiernan’s replacement. There was pressure to take his commander’s advice. But he didn’t want to get boxed in. He had lots of questions, and he wanted more options.

The group would meet eight more times over the next two months. Obama’s Republican critics, especially former vice president Dick Cheney, accused him of “dithering.” In fact, Obama was asking the chiefs pointed questions, and the answers they’d bring to the next meeting only raised more questions. When one official raised the possibility of working less with Karzai’s central government and more with local officials across Afghanistan, Obama asked for a rundown of which provincial and district governors had the tightest and loosest connections with Karzai or seemed to be the most and least likely to cooperate with US interests. No one around the table knew the answers, so back they went to their offices at the Pentagon or the CIA and ordered an intelligence analysis.

And the chiefs never did satisfy Obama’s request for more options; they never answered his question of what could be done, say, with thirty thousand more troops or with twenty thousand. (The vice chairman of the JCS, Marine General James “Hoss” Cartwright, drew up such an option, figuring that his job was to comply with the president’s lawful orders. He got an earful from the chiefs; Mullen never sent his paper up the chain to the White House. Cartwright pulled an end run and took his paper to Biden’s office; the vice president at least put it on the agenda, but the chiefs discussed it only to dismiss it.)

Some of the aides sitting behind the officials around the table found the whole process a bit dumbfounding. Petraeus, the most persuasive advocate for McChrystal’s numbers, barely mentioned Afghanistan as a distinct country; his main interest seemed to be defending his own COIN philosophy as a principle of warfare. Mullen and Gates backed their commanders. In some past administrations, the secretary of state might have risen in opposition, but Obama’s team players were remarkably unified in their views; Hillary Clinton herself was more hawkish than some had anticipated, so she went along as well.

•  •  •

Two events outside the Situation Room intensified the political tensions and pressures within.

First, on September 21, a few days before McChrystal submitted his recommendation on troop levels, the Washington Post printed a front-page story by Bob Woodward summarizing the main points of McChrystal’s sixty-six-page assessment, and the Post’s website reproduced a leaked copy of the entire document. There was no evidence that the leak had come from McChrystal or his staff, but it compounded the feeling among White House political advisers that the generals were trying to box in the president.

The second event amounted to pressure from the opposition. On November 9, the ambassador to Afghanistan, Karl Eikenberry, a retired three-star general who had been commander of US forces in Afghanistan for eighteen months between 2005 and 2007, sent a secret cable to Secretary of State Hillary Clinton titled “COIN Strategy: Civilian Concerns.”

It was a lengthy critique of McChrystal’s entire presentation: his assessment of the security situation, his request for more troops, and his call for a shift to a counterinsurgency strategy. Eikenberry stated the reasons for his objections in a simple, shocking sentence: “President Karzai,” he wrote, “is not an adequate strategic partner.”

He elaborated that COIN

assumes an Afghan political leadership that is both able to take responsibility and to exert sovereignty in the furtherance of our goal—a secure, peaceful, minimally self-sufficient Afghanistan hardened against transnational terrorist groups. Yet Karzai continues to shun responsibility for any sovereign burden . . . He and much of his circle do not want the US to leave and are only too happy to see us invest further . . . It strains credulity to expect Karzai to change fundamentally this late in his life and in our relationship.

Eikenberry also criticized McChrystal’s analysis for assuming that once US forces clear and hold an area, the Afghan government will quickly step in and provide basic services. “In reality,” he argued, “the process of restoring Afghan government is likely to be slow and uneven, no matter how many US and other foreign civilian experts are involved.”

Finally, there was the matter of the safe havens across the porous border with Pakistan. “More troops won’t end the insurgency,” the ambassador wrote, “as long as Pakistan sanctuaries remain.”

He acknowledged that McChrystal’s analysis was “logical and compelling within his narrow mandate to define the need for a military counter-insurgency campaign within Afghanistan.” However, he went on, it failed to take into account several “real-world variables”: not only Pakistan’s sanctuaries and Karzai’s weak leadership but also America’s own “national will to bear the human and fiscal costs over many years.” Any one of these variables, he observed, could potentially “block us from achieving our strategic goals, regardless of the number of additional troops we may send.”

Eikenberry urged instead that the president conduct a “comprehensive, interdisciplinary analysis of all our strategic options,” going well beyond an up-or-down verdict on COIN—in essence, that Obama start his review all over. “For now,” Eikenberry summed up, “I cannot support DoD’s recommendation for an immediate Presidential decision to deploy another 40,000 troops here.”

In one sense, Eikenberry’s cable wasn’t surprising. He had been making some of the same points, though less forcefully, in some of the national-security meetings, which he attended via video teleconference. Nor were most of his points particularly novel: McChrystal’s own assessment had spelled them out as serious problems; Petraeus and Mullen had testified in congressional hearings that cleaning up Karzai’s corruption and shutting down Pakistan’s sanctuaries were just as important as battling the Taliban. At a Senate hearing in mid-September, Mullen had acknowledged that the main problem in Afghanistan was “clearly the lack of legitimacy in the government.”

Senator Lindsey Graham, Republican of South Carolina, asked, “We could send a million troops, and that will not restore legitimacy in the government?”

Mullen replied, “That is correct.”

But Petraeus, Mullen, and McChrystal saw these matters as obstacles to overcome, not as reasons to call off the campaign. All three were furious when they heard about Eikenberry’s cable, and more so when it leaked to the press a few days later. It wasn’t just that they disagreed with his conclusions. They felt ambushed by his abruptness; he hadn’t consulted with any of his former brothers in arms or alerted them that he was writing such a document. How was the United States supposed to run an “integrated civilian-military counterinsurgency campaign,” as McChrystal’s memo recommended—or, for that matter, any kind of coherent campaign, COIN or otherwise—when the top diplomat in the country was rejecting the game plan and, more than that, hadn’t bothered to discuss his qualms with the top officers?

Another reason for the officers’ ire may have been that the ambassador’s main points were so clearly accurate. The Afghan president’s behavior was increasingly erratic and self-destructive; Eikenberry’s description of him (“not an adequate strategic partner”) barely began to capture the concerns about his shortcomings. There were reports that Karzai was on medication for bipolar disorder, that his brother was a major drug runner, that he and his entourage were siphoning money abroad through corrupt state banks.

Finally, the presidential election, which had been expected to bolster Karzai’s legitimacy, in fact diminished it further. So many ballot boxes had been so blatantly stuffed that nobody believed the official outcome. Western inspectors calculated that fully one-third of the ballots were fraudulent. Subtracting those from Karzai’s tally gave him a plurality but one that fell just under 50 percent of the vote—the bar he needed to cross to avoid a second-round runoff against the second-place candidate, former foreign minister Abdullah Abdullah. To make matters worse, Karzai condemned the inspectors as imperialist trespassers, declared himself the winner, and refused to hold a runoff. Obama dispatched John Kerry, the Democratic chairman of the Senate Foreign Relations Committee, who had traveled extensively in that part of the world, to calm Karzai down. It took five days of soothing talk—and, as Kerry later put it, “three hundred cups of tea”—before Karzai agreed to accept the monitors’ report and to abide by his own country’s election laws.

In the end, Abdullah dropped out before the runoff took place. Karzai retained the presidency by default, lending some hope for his redemption.

Yet the harsh dilemmas remained. On the one hand, the case for COIN in Afghanistan seemed far from plausible. On the other hand, the case for Biden’s position—to do little more than keep sending drones and Special Ops commandos to kill insurgents along the AfPak border—was still less compelling; that’s what McChrystal’s predecessor, Dave McKiernan, had been doing, with little to show for it.

There was another dilemma. On the one hand, Karzai was now the duly reelected president of Afghanistan and thus the only national leader with whom Obama could deal. On the other hand, Obama would have to pressure Karzai into reforming his government if COIN had a chance of succeeding; he needed to make Karzai realize that America’s patience wasn’t endless.

Back in March, when he announced his first escalation of the war, Obama had said, “We will not blindly stay the course” and “We will not, and cannot, produce a blank check.” Eight months later, after nine meetings of intense deliberations on the subject, he was still insistent on those points.

On Sunday, November 29, around five in the afternoon, Obama held his final meeting on the new strategy. This one was in the Oval Office and involved a smaller group than usual, restricted mainly to his top defense advisers: Gates, Mullen, the service chiefs, and Petraeus, along with Biden, national-security adviser Jim Jones, and the president’s senior adviser on Afghanistan, Doug Lute.

In a little over forty-eight hours, Obama would deliver a much-anticipated address at West Point, spelling out his decision on troop levels and a new strategy. He called the Sunday afternoon meeting to inform his advisers of what he’d decided and to give them each a copy of his orders, which he’d written himself.

In the order, he wrote that the “core goal” of the war would be the same as the one he’d announced back in March: “disrupting, dismantling, and eventually defeating al Qaeda and preventing al Qaeda’s return to safe haven in Afghanistan and Pakistan.” However, he now emphasized that his approach for doing this was “not fully resourced counterinsurgency or nation-building.”

Rather, the strategy would be to deny the Taliban access to key population centers, but merely to disrupt the Taliban outside those areas and, overall, not to defeat but rather to degrade their forces to levels that the Afghan army and police could eventually handle by themselves after an enlarged and reinvigorated training program.

In short, Obama was calling for a mix: COIN in the large cities, CT elsewhere.

To this end, Obama had decided to deploy a “surge”—he called it by that name—of thirty thousand additional American troops, plus three thousand “enablers” (headquarters staff, extra logistical support, and so forth), while also pushing the NATO allies to send seven thousand more of their troops. That would add up to the forty thousand that McChrystal had recommended.

Finally, Obama would begin to withdraw these surge troops in July 2011, just eighteen months hence, assuming that the Taliban would be degraded and the Afghans would get better at securing and governing their own country.

This last part was the shocker; none of the officers in the room had seen it coming. Obama was coupling the surge and the drawdown for two reasons. First, he wanted to convey a sense of urgency to Karzai; the Americans weren’t going to be around forever, so he had to make reforms now. Second, Obama regarded this whole business—the escalation and the partial adoption of COIN—as an experiment, a useful experiment, probably unavoidable, and, in any case, less risky than not trying. But if it didn’t achieve results after what he saw as a reasonable length of time, he wasn’t going to extend it.

Obama spoke directly to his top defense advisers: Gates, Mullen, and Petraeus. With these additional forces, he asked them, will you be able to clear, hold, and transfer to the point where—within eighteen months—the Afghan security forces can take the lead in the fight? He emphasized that this was a pivotal moment. If they said they could do this, and it turned out they couldn’t, he wasn’t going to double down. There wasn’t going to be another round of escalation. This was all they were going to get.

So he asked them, one at a time, if they could perform this mission with these resources: Bob? Mike? Dave?

Each of them replied, one after the other, “Yes, sir.”

Doug Lute, the three-star general who had been brought over from the Pentagon to serve as Bush’s “war czar” and who stayed on to work for Obama, was taking notes. He was startled that all three agreed to the terms. Lute had been a West Point cadet, class of 1975, and, a decade later, had taught international relations in the Sosh department for three years, overlapping with Petraeus’s tenures as a cadet and as an instructor. He had read the COIN literature, including Petraeus’s field manual, and he knew that no serious scholar or practitioner in the field would promise that a campaign of this sort, especially in Afghanistan, could succeed so rapidly. Counterinsurgency wars were by nature protracted wars; they usually took years, sometimes decades, to resolve.

Petraeus was asked afterward by a number of colleagues why he didn’t speak up and tell the president that the timetable wasn’t plausible. Petraeus replied that it wasn’t that kind of meeting; he got the clear sense that the advisers were there to take orders, not discuss options; that what Obama put on the table was a take-it-or-leave-it deal, and that if you wanted to stay on the team, you would take it. He also calculated, in his own head, that if enough progress had been made by the eighteen-month deadline, even if not quite as much progress as the president was now demanding, a case could be made to stay on a bit longer.

Everyone in the room was making a gamble. Obama was making at least four: that Karzai would reform his government and thus remake himself; that the Afghan army would shape up to a competent force; that Pakistan would alter its calculus of national-security interests and go after the Taliban on the border more forcefully; and that the NATO coalition would hold together.

Petraeus and the other top officers were making just one gamble: that, in the end, Obama would soften his position.

None of the bets would pay out.