The Surge

16. The Surge

In mid-September David Petraeus received a phone call from General Peter Pace, chairman of the Joint Chiefs of Staff. A consensus was forming among senior officers in the Pentagon that a spurt of new thinking might help turn the war around. Pace was putting together a group of creative colonels to advise the generals on their options, and he wanted Petraeus’s input on which colonels should be called.

Petraeus needed only a few seconds to respond. Here was a rare chance, still a few months before the COIN manual’s publication, to install followers of the cause into the Pentagon’s inner sanctum. He doubted that Pace would take anyone off the battlefield for this project, but there were two colonels—both acolytes of sorts to Petraeus and, in his eyes, brilliant—who now had some time on their hands.

One was Pete Mansoor, a West Point graduate with a concentration in history, the top-ranking cadet in the class of 1982. The other choice—an obvious one—was H. R. McMaster, the 3rd Armored Cavalry commander who had made counterinsurgency work in Tal Afar.

Mansoor could head off to Washington right away with Petraeus’s blessings. But McMaster had just moved with his family to London to take a yearlong Army fellowship at the International Institute for Strategic Studies, doing research and writing on the future of US military policy in the Persian Gulf and South Asia. It was only his second day on the job when he got the call from one of General Pace’s aides, asking him to come work at once on a top-secret project for the chairman.

McMaster wasn’t sure what to do; the aide who called him couldn’t discuss any details about the project on the phone. So McMaster called General John Abizaid, the head of US Central Command, who’d arranged for him to get the London fellowship. Abizaid had heard a little about the project. He told McMaster that it concerned the war in Iraq but that he should feel no obligation to take it; he could stay in London and do his long-term research if he preferred.

Two days later, McMaster left for Washington.

When he checked in at the Pentagon, he was taken to “the Tank,” the Joint Chiefs’ conference room on the building’s second floor, where he found himself joined by fifteen other colonels from all the services. They were told that their work, even the very existence of the group (which came to be called the “council of colonels”), must be kept secret. Their mission was to give the chiefs unvarnished advice about the war in Iraq. They were to examine all scenarios and options—political, strategic, tactical, operational—and would be allowed access to everything they wanted, from the daily intelligence briefings to the most highly classified data on weapons systems. Once a week, sometimes twice, they would meet with the chiefs, not just to brief them on the work but to discuss it, without regard to the formalities of rank, as if they were peers.

McMaster had never heard of a setup like this, couldn’t have conceived of such a thing. But he was heartened by the arrangement. Not only did it promise to be interesting, it might signal that the top generals finally recognized the urgency of their problems.

Petraeus had the same reaction when he first learned about the project. It seemed so unlike the JCS. Who prodded them into opening their vaults and opening themselves to self-criticism? He did have one suspicion. He called a retired four-star general, a longtime friend and mentor named Jack Keane.

“Jack,” Petraeus asked, “was this council of colonels your idea?”

Keane laughed and admitted that it was.

•  •  •

The seeds of the council had been planted just a month and a half earlier, on the night of August 3, when Jack Keane settled into the easy chair in his McLean, Virginia, living room and watched C-Span’s rebroadcast of the day’s congressional hearings. Rumsfeld, Pace, and Abizaid were testifying on Iraq before the Senate Armed Services Committee, and they were claiming that all was well, that their present strategy of drawing down US forces and turning over security tasks to the Iraqis was yielding good results.

At one point in the hearing, Senator Hillary Clinton accused Rumsfeld of papering over a “failed policy” with “happy talk and rosy scenarios.” Rumsfeld, with his customary condescension, exclaimed, “My goodness!” He then acknowledged that there were problems and setbacks, but, he went on, “I don’t know that there’s any guidebook that tells you how to do it. There’s no rule book, there’s no history for this.”

Keane sat up, appalled. First, he’d been to Iraq, and he’d spoken at great length with other officers, trusted friends, who were stationed there. He knew that things were not going well; that, in fact, a disaster was in the works. Second, what was Rumsfeld talking about? There was history; there were guidebooks that told you how to do this. Keane had read the books, by Galula, Kitson, Nagl, and others. He’d studied—he’d lived—the history.

Forty years earlier, Keane had fought as a paratrooper in Vietnam. When his tour ended in 1968, he was sent to Fort Benning, Georgia, to take an advance course for captains. One of his electives was a class on counterinsurgency, in which he read books and monographs on Algiers, Malaya, the Philippines, and the CORDS experiment in Vietnam. This was all new to him, but it rang true. He’d sensed that his commanders in Vietnam were fighting the wrong war. Now he understood the full context: that they’d been fighting against guerrillas with conventional tactics. The impression was deepened in 1975–76, when he was stationed at the Command and General Staff College and took a course on the lessons of Vietnam. The war had recently ended; all his classmates had fought in it, most of them as platoon or company commanders. The Army generals were abandoning those lessons, resuming their focus on the prospective big war in Europe and pretending that Vietnam never happened. But Keane paid attention to those lessons and tried to apply them to his soldiering, signing up for light infantry units and serving in the moot-wah wars in Somalia, Haiti, Bosnia, and Kosovo.

In 1987 Keane helped set up the JRTC, the Joint Readiness Training Center, in Fort Chaffee, Arkansas. (It would soon move to Fort Polk, Louisiana.) This was the “light” version of the National Training Center in California. Instead of armor officers fighting simulated tank battles on an open range in the desert, the JRTC pitted airborne infantry soldiers, Rangers, and Special Forces against guerrillas and insurgents in a mock village—to simulate the kind of battle where enemy fighters and innocent civilians mixed, almost indistinguishable from one another, and the goal was to kill or capture the former while protecting the latter.

Two decades later, after the insurgency in Iraq took off in undeniably full force, the National Training Center would follow suit, overhauling its test range, erecting mock villages, and hiring Iraqi exiles to play the parts of local citizens and terrorists. But in its beginnings, the JRTC was a novel, even controversial, enterprise. Generals came through Keane’s center with their aides, often out of curiosity, to take a look and sit in on the after-action reviews.

General Gordon Sullivan, the Army chief of staff, attended one of the early reviews, in 1991. A training exercise had just taken place, and a captain with the 82nd Airborne Division was brought into a small conference room to discuss how he’d done. In the scenario, the captain and his men were to enter a village and find the gunman who’d killed one of their soldiers. The captain had decided to accomplish this mission by storming the village before daylight, guns cocked, pounding down doors, dragging the local men out of their houses, locking them in handcuffs, and interrogating them harshly, all while their families watched in horror. Of course, he came up with nothing.

The after-action review’s chairman was a British officer who’d fought in Northern Ireland. Trying to prod the captain into discovering for himself where he’d gone wrong, he asked if there might have been some alternative approach to finding the killer. The captain shrugged and said, “No, I don’t think so.”

At this point, the man who’d played the town’s mayor asked if he could say something. He gently told the captain that a better way might have been to come to the village quietly the night before and to consult with him. He would have advised the captain to wait until later in the morning, after the kids had gone to school. He then would have personally asked the townspeople to leave their homes, just for a few minutes, so that the captain’s men could search for weapons. But to come in so brusquely, to let children witness such horrifying treatment of their parents—this sort of thing alienates the townspeople, makes them sympathize with, maybe join, the insurgents.

The captain was flustered. “Lookit!” he said, pounding a nearby table. “My job is not to deal with this people thing! My job is to kill the enemy!”

General Sullivan, who’d risen through the ranks as an armor officer but was beginning to sense that the end of the Cold War might spawn a different kind of warfare, watched this exchange with rapt attention. He leaned over to Keane with a smile and whispered, “Jack, this is real powerful stuff!”

Twelve years later, Keane was stunned to see the same scenario played out for real in Iraq. He was the Army’s vice chief of staff when the invasion got under way in March 2003. Rumsfeld had asked him to replace General Eric Shinseki, who was about to retire as chief of staff in June. But Keane turned down the offer: his wife had recently been diagnosed with Parkinson’s disease; he wouldn’t have time to care for her if he took such a demanding job. (He would retire from the Army in October.) Meanwhile, in the two months between Shinseki’s departure and the confirmation of his eventual successor, General Pete Schoomaker, Keane stepped up as acting chief. One of the first things he did, in late June, was to go see, six weeks into the occupation, what was happening in Iraq.

His first stop was Mosul, where he was greeted by David Petraeus, commander of the 101st Airborne Division. The two men had known each other for more than a decade. Back in 1991, when Petraeus led a battalion of the 101st at Fort Campbell, Keane was the division’s assistant commander. The day when Petraeus was accidentally shot during a live-fire training exercise, Keane was the one who rushed him to the hospital, keeping up a calm chatter on the way, arguably saving his life. Ever since, Keane had remained a close friend and, perhaps second only to General Galvin, a mentor.

Petraeus gave Keane the works: the PowerPoint briefing, followed by a tour of the market, city hall, the university, the police station, then drop-ins on the brigade commanders, some of them working with the people, others going after the enemy. Keane was impressed. This was what counterinsurgency was all about. Mosul gave him hope.

His next stop was Tikrit, where Ray Odierno was commanding the 4th Infantry Division, and the news here was depressing. They had met only a couple of years earlier in the Pentagon, when Keane was vice chief of staff and Odierno was a one-star general in charge of planning the Army’s force structure. In the summer of 2001, when Rumsfeld announced that he was chopping off two of the Army’s ten divisions, Keane had Odierno work up a briefing overnight, explaining how force requirements were calculated and why eight divisions weren’t enough. Odierno delivered the briefing the next day to Rumsfeld’s deputy, Paul Wolfowitz. After Odierno finished and left the room, Wolfowitz turned to Keane and asked, “Who was that guy?” It was the best briefing he’d ever heard from an Army general, though, granted, the competition wasn’t fierce. Wolfowitz urged Rumsfeld not to cut two divisions. Rumsfeld acceded.

But toting troops on paper was different from commanding them in the field, and, watching these troops in action, Keane concluded sadly that Odierno didn’t measure up. If Petraeus in Mosul was the textbook case of how to do counterinsurgency right, Odierno was the sidebar on how to do it wrong. His PowerPoint briefing boasted how many offensive operations he’d conducted each week, how many suspected bad guys he’d killed or captured. The operations themselves, which Keane briefly observed, were no different, focusing almost exclusively on raiding, arresting, and killing—which, though clearly necessary, had little bearing on the outcome of this kind of war.

Keane thought back on the simulation at the Joint Readiness Training Center a dozen years earlier, when the captain had dragged men out of their houses and explained afterward that he had no patience for “this people thing,” that his job was to “kill the enemy.” Keane decided to pose the same Socratic questions that the British officer had that day at the after-action review. He asked Odierno to bring in one of his brigade commanders.

After the commander, a colonel, entered Odierno’s office and saluted, Keane asked him what he’d been doing.

Cordoning off a village, sir, the colonel replied.

Why? Keane asked.

The enemy is in there, sir, he answered.

How do you know that? Keane asked.

Intelligence, sir.

Do you know where the insurgents are in the village?

No, sir.

How do you find them?

We go door to door, sir.

Do you knock?

No, sir. We go through the door.

What’s on the other side?

Families, sir.

What do you do?

We take all the males outside and handcuff them, sir.

How young are these males?

As young as teenagers, sir.

What’s the scene like, emotionally?

The colonel paused. Sir, it’s terrible.

After the colonel left, Keane turned to Odierno. “What’s going on here, Ray?” he asked. “We’re breeding an insurgency here. We’ve got to see the people as part of the solution.”

Later on, during the same trip, Keane met with Lieutenant General Ricardo Sanchez, the commander of all US forces. He’d rarely seen a general officer so in over his head. Keane spent the Fourth of July weekend with Jerry Bremer, the head of the Coalition Provisional Authority, and realized that nobody, civilian or military, was in charge at the top.

A week later, back in the Pentagon, Keane briefed the other chiefs on his trip. His first point was that the American troops were facing “a low-level insurgency.”

Richard Myers, the Air Force general who was chairman of the Joint Chiefs of Staff, warned Keane that they all had to be careful about using that word: “insurgency.”

Keane didn’t know what to make of this. If we don’t define the war accurately, he said, we’re never going to be able to fight it effectively. We haven’t been around this kind of thing for a long time. Keane admitted that his own service, the Army, was ill prepared for these wars, lacking any doctrine or training. Only one of the generals over there had any idea how to deal with the challenge (and, though he didn’t say so, Keane knew that the exception—Petraeus—was hardly a favorite among the top brass).

Myers replied, “I’m just saying, the people on the third deck”—meaning Rumsfeld and his top aides, whose offices were on the Pentagon’s third floor—“don’t want to hear it.”

End of conversation.

Keane was flustered. On his way to Iraq, he’d assumed that the war was pretty much over; that, at worst, it might be something like Bosnia on steroids. Now he realized that this fight might drag on for years, because the officers at the top lacked a strategy and seemed unaware that they needed one. He’d spent the bulk of his career preparing for this kind of war; now, just as the Army found itself fighting one, and in desperate need of leadership, he was about to leave the scene.

Like most retiring modern generals, Keane joined a few corporate boards and nabbed a lucrative job as a strategic planning executive on K Street, Washington’s lobbyists row. But he ached to be a player in the war. Over the next three years, he kept in touch with former colleagues and subordinates who were in the fight; he made several trips to the war theater as a consultant of one sort or another; he saw the situation worsening and felt his frustrations mounting.

Finally on that night in August 2006, while watching the abysmal Senate hearings, he hit a boiling point. Somebody needed to fix American strategy. Clearly, the secretary of defense, the chairman of the Joint Chiefs of Staff, and the top commanders in the field weren’t doing it, weren’t even acknowledging the strategy was broken. There was no getting around it, in Jack Keane’s mind: this was a job for Jack Keane.

The next day, Keane called Rumsfeld’s office to ask for a meeting. It was an unusual thing for a retired general to do, but the two had gotten along when Keane was on active duty; Rumsfeld had asked him to be chief of staff, after all. The meeting took place on Tuesday, September 19, in the secretary’s office. General Pace sat in, though he didn’t speak, just took notes.

Keane came straight to the point: “We are edging toward strategic failure.” The number of attacks was soaring, civilian casualties were skyrocketing, the Iraqi government couldn’t keep the country from unraveling, and our own forces didn’t have the right strategy.

Rumsfeld asked what was wrong with the strategy.

“It is not designed to defeat the insurgency,” Keane replied. The only way to fight this war was through classic counterinsurgency strategy: protect the people, isolate them from the insurgents. He advised Rumsfeld to read the after-action report on Colonel H. R. McMaster’s operations in Tal Afar and David Galula’s book Counterinsurgency Warfare: Theory and Practice. Rumsfeld hated condescension when he was on the receiving end but had no leverage to fire back at a retired general. He shifted uncomfortably in his chair throughout the conversation.

Afterward, Pace asked Keane to come see him separately. When he did, two days later, Keane opened up the double barrels. It was widely known throughout the Pentagon that Pace, following Rumsfeld’s lead, had taken his eye off Iraq, leaving the war to Casey and Abizaid while spending his own time on politically safer subjects. Keane now told him that he had to dive in, immerse himself in the issues, learn what was happening directly, and fire the commanders at fault.

Pace seemed unnerved. A few hours later, he postponed a scheduled trip to South America and started calling around to a few select generals, including Petraeus, asking for names of suitable advisers. The council of colonels held its first meeting in the Tank just six days later, on Wednesday, September 27.

When McMaster was asked to join the project, he was told it would last “weeks, not months,” but it went on for three months. Workdays in the Tank began before dawn and ended after sundown, with several weekend afternoons consumed as well. McMaster felt weird on those Saturdays or Sundays, driving into a nearly empty Pentagon parking lot. There was a war going on—two wars, if you counted Afghanistan—and this was the Department of Defense. Why was almost nobody else working long hours?

The colonels covered a vast array of issues, not just the nitty-gritty of warfare (how many troops, and what sorts of tactics and strategies, were needed to clear and hold key cities) but also the political and strategic questions: the sectarian breakdown of Iraqi political parties, the alliance structure of the government, the preconditions of stability, the nature of US interests in the country and the region, what military objectives needed to be achieved to meet those interests. They had access to the Joint Staff’s entire trove of classified files and software; they did computations, ran simulations, read reams of reports, and called in outside experts, including Eliot Cohen, who sat on the Defense Policy Board, and David Kilcullen, who was now working at the State Department. And they prepared lots of PowerPoint briefings.

At their first session with the chiefs, McMaster, Mansoor, and a Marine colonel named Tom Greenwood—who all agreed with one another on most matters—tested the limits of the top generals’ tolerance for frankness. The first slide of their briefing read: “We are losing because we are not winning. And we are running out of time.”

This equation was straight out of Galula (the insurgent has only to sow disorder anywhere, while the counterinsurgent must maintain order everywhere), but it came as a shock to the chiefs. They seemed not to have considered this concept: that stalemate for the United States meant victory for the insurgents.

As the study progressed, almost all the colonels reached the same conclusions: that Iraq was devolving into civil war; that the Iraqi government was a faction in this war; and that, therefore, merely turning over authority to the government—as Generals Casey and Abizaid were planning to do—wouldn’t break, and might escalate, the cycle of sectarian violence.

On what to do about the looming disaster, though, the colonels could form no consensus. McMaster emerged as the hardliner, adamant in arguing that the United States should mobilize as much military power as possible in order to fight and “win” the war. Mansoor was less confident about victory; he admitted that he’d never been an enthusiast of the war and didn’t really know whether a good outcome was possible under any circumstances. Still, he did support a brief troop surge, to bring down the violence and restore a semblance of order, followed by a long-term counterinsurgency strategy pursued by a much smaller level of American troops. He saw this as the most that could be managed politically. Greenwood sided with Mansoor.

However, the other Army and Marine colonels, most of whom worked on the Joint Staff, were worried more about the war’s effect on the military as an institution: the draining of resources, the dwindling number of recruits, and the diminishing rates of reenlistments. The Air Force and Navy colonels, who had less of a stake in the war to begin with, saw it as not only a sinkhole but also a lost cause worth no further sacrifice, and they urged getting out as quickly as possible.

The chiefs were torn. They were by nature inclined to accept the judgment of the commanders in the field. And while they were shocked by the colonels’ critique of the war strategy, they also took solace at the same critics’ failure to agree on a remedy. Some of the colonels urged doing more, others doing less—making it easy for the chiefs to convince themselves that a prudent compromise would be to do nothing, to stay the course.

But in the weeks ahead, McMaster’s work on the council would have a profound impact, much more than even he could have imagined.

•  •  •

Roughly midway into McMaster’s work with the council of colonels, a defense analyst named Frederick Kagan embarked on his own study to promote a surge of American troops in Iraq and the adoption of a counter-insurgency strategy.

Kagan and McMaster had known each other for almost ten years. In the mid-1990s they were both instructors—initially, they shared an office—in the History Department at West Point. Kagan was now thirty-six, eight years younger than McMaster, and he had never served in the armed forces. (West Point hired him after Congress passed a law requiring the academy to place a few civilians on its faculty.) But he came from a family of military historians. His father, Donald Kagan, a classics professor at Yale, was the author of a seminal four-volume history of the Peloponnesian War. His older brother, Robert, a prominent defense adviser in Republican circles, had written a bestselling book, Of Paradise and Power, about the growing rift between the United States and Western Europe regarding national-security matters. In the late 1990s, all three Kagans signed various petitions of the Project for the New American Century, the neoconservative group urging regime change by force in Iraq.

Fred Kagan had left West Point just a year earlier, in 2005, and was now ensconced at the American Enterprise Institute, Washington’s most prominent neocon think tank, its roster including some of the Iraq War’s most fervid supporters. Kagan was keen not only on the war but on counterinsurgency doctrine. For a few of his ten years at West Point, he taught the Revolutionary Warfare course. Now it would be through Kagan that AEI emerged as the nexus joining the neocon movement and COIN.

It was a logical convergence: the COIN revivalists saw the doctrine as a tool for fighting the small wars that were looming on the horizon; the neocons were eager to fight those wars as a way of asserting American power.

Kagan had been one of the outside critics invited to President Bush’s war-strategy meeting at Camp David in June. Around the same time, he’d been one of the two dozen expert analysts invited to speak at a plenary session of the Iraq Study Group chaired by James Baker and Lee Hamilton. He’d come away from that session deeply worried. Congress had authorized the creation of the Baker-Hamilton Commission explicitly to find a way out of Iraq. Its ten members represented the cream of the Washington establishment. Their report was due in December, and it was likely to add a tone of gravitas to the growing calls for bringing the troops home.

To counter those calls, Kagan decided to form his own Iraq study group—which he called the Iraq Planning Group—with the idea of analyzing the war the same way that an Army commander’s staff would: as a detailed problem of military strategy, tactics, and operations.

In early December, Kagan got in touch with his old West Point office mate H. R. McMaster and invited him to his apartment for dinner. McMaster couldn’t talk directly about his work in the council of colonels, which Kagan vaguely knew about, but there was no prohibition against discussing his thoughts about the issues that the council happened to be covering; much of his thinking, after all, stemmed from his earlier work at Central Command, in Tal Afar—even back at West Point. So they talked, in a general way, about the conclusions he was reaching: that the Iraqis had neither the capacity nor the will to quell the violence on their own; that nothing could improve until the Iraqi population felt secure; that the way to make this happen was to adopt a counterinsurgency strategy, which involved, among other things, stationing American troops among the people instead of “commuting” them from large, heavily protected bases on the outskirts of the cities; and, finally, that such a shift would require more troops, not fewer. Kagan, from his own training and experience, was inclined to agree.

At one point during dinner, Kagan asked McMaster if he knew any recently retired Army officers who might be suitable for the project, trained staffers who had real experience in analyzing how many troops were needed to deal with a specific threat, how quickly the troops could be deployed, what they needed for support, that sort of thing.

McMaster knew exactly the right guy for the job: Colonel Joel Armstrong. He’d been McMaster’s deputy commander at Tal Afar and had done much of the fine-tuned analysis that made the campaign a success. And he’d retired from the Army just a few months earlier.

The next day, Kagan sent Armstrong an email describing the project, noting that he saw Tal Afar as the model of what American forces should be doing all across Iraq and asking him to phone if he was interested in further discussion. Armstrong called back at once. Kagan offered him a contract. He took it.

Before hanging up, Kagan asked if he knew anyone else with similar skills. The job had to be professional, but it also had to be done quickly; an extra hand might be necessary. Armstrong thought at once of Dan Dwyer, a major who’d been McMaster’s chief of planning in Tal Afar and, before that, an analyst in Baghdad. Dwyer was the one who had done the number-crunching, and he’d retired from the Army just a few days earlier. Kagan called Dwyer and hired him, too.

The work began right away. The Baker-Hamilton Commission was set to release its report on December 6. Kagan couldn’t beat that deadline, but he hoped he could put out his alternative study very soon after. He scheduled a four-day conference to present and discuss the work, beginning on Friday, December 8—less than a week away. (Before then, Kagan, Armstrong, and Dwyer would have one more dinner with McMaster, at an Italian restaurant near Dupont Circle, to discuss how the study was going.)

Even before hiring his two new aides, Kagan and some of the staff at AEI had already gone through the vast array of open-source material—news stories, official reports, studies by private groups that assembled data on war casualties—to pinpoint the Iraqi neighborhoods with the most intense violence. They circled those areas on a large-scale map, focusing particularly on Baghdad and Anbar Province, the areas deemed most critical to the stability of Iraq overall. They then clicked on overhead images of those areas on the Google Earth website. Armstrong and Dwyer examined the images closely and calculated how many troops would be needed to secure—to clear and hold—each area.

After computing the requirements for all of these neighborhoods, running some simulations, and double-checking by altering a few assumptions and seeing if the answers came out roughly the same, they added up the numbers. It turned out that the areas could be secured with an additional five Army brigade combat teams and two Marine combat regiments (each regiment roughly equivalent in size to a brigade).

The next step was to figure out how quickly these seven extra units—about twenty-four thousand extra troops—could be mobilized to Iraq from wherever they currently were. This was Dwyer’s specialty. He was one of the few people, in or out of the Army, who understood the workings of “the batting order”: the sequence of which combat units were scheduled to move into Iraq, which were moving out, and which were on home leave or in training. There was a hitch: the Army’s “force-generation model”—the mathematical formula that let planners plug in the numbers and find out the answers—was classified. Dwyer searched the internet, hoping to find an approximation of the real thing in the open literature. To his astonishment, he found the Army’s actual, secret model on Wikipedia.

He downloaded the model, plugged in the numbers, and it turned out that five Army brigades and two Marine regiments were also the number of combat units that the military could put in Iraq; five plus two were all that the Army and Marines had left in reserve.

It was an unlikely coincidence that the number of troops required matched the number of troops available. When Petraeus learned of the results, he assumed that Kagan had rigged the model to make the two numbers identical, a common practice in military analysis. Kagan was a bit astonished, too; he had Armstrong and Dwyer redo the calculations, to make sure they were right. They were.

Kagan’s team then laid out all the data and conclusions, replete with charts, tables, and bullet points, in fifty-five PowerPoint slides. The briefing made clear that the surge was a gamble. It would take the better part of a year to get all the extra troops in place. To sustain the surge, the standard twelve-month tour of duty would have to be extended to fifteen months for all American troops in Iraq, not just the newly deployed ones. Finally, because these troops would be living among the people, out in the open constantly, American combat casualties would rise, perhaps dramatically, at least for the first several months. But in the end, the group concluded, the surge, along with the shift in strategy, could make the crucial difference.

•  •  •

To help roll out the study and enhance its credibility, Kagan asked two retired generals to speak at the upcoming AEI conference and to lend their names to the group’s roster.

One of them was David Barno, a member of the West Point Sosh network, class of 1976, and a former US commander in Afghanistan who, back in 2003–05, had put in motion a short-lived counterinsurgency strategy against the Taliban.

The other retired general was Jack Keane. Kagan hadn’t known about Keane’s personal campaign to change the military strategy in Iraq; the two had met only a couple of months earlier, after Keane called him to praise an article Kagan had written.

Kagan, Armstrong, and Dwyer delivered the full briefing to the two generals on the morning of Friday, December 8, just before the conference began. Both generals found it compelling. Keane saw it as precisely the sort of hard-core, detailed analysis that he needed to make his case to higher circles.

And the timing was serendipitous. Just the day before, Stephen Hadley, President Bush’s national-security adviser, had invited Keane to a meeting in the Oval Office. The president was coming around to the view that changes were needed in the war strategy, and some staffers on the National Security Council had put together a list of outside experts to discuss the options with him. The meeting was scheduled for the following Monday, December 11. As a further coincidence, Vice President Cheney had asked to see Keane later the same day; the two had known each other since the days when Cheney was secretary of defense for Bush’s father. Keane asked Kagan if he could take a copy of the briefing slides with him and share them with the president and vice president. Kagan, of course, had no problem with that.

The White House meeting began around three thirty Monday afternoon. It had been put together by the same NSC staffers who’d organized the Camp David meeting with Bush, the war cabinet, and a few outside critics the previous June. The earlier meeting had had no effect; Bush used it as mere pretext to fly to Iraq without public notice. But the staffers had reason to believe this meeting would be different. Several crucial events had occurred since the summer: the midterm elections in November had turned over control of the House and the Senate to the Democrats; Bush’s popularity was at an all-time low, mainly because of the war; he’d fired Rumsfeld right after the elections and seemed ready to make more changes still.

In the room were Bush, Cheney, Hadley, and several aides, including Karl Rove, Bush’s top campaign adviser, who had started to take a more active role in policy matters. The outsiders this time around, besides Keane, were retired generals Barry McCaffrey and Wayne Downing, and two civilians: Stephen Biddle, a former professor at the Army War College, now with the Council on Foreign Relations, who had written an article calling for a new strategy; and—the one repeat from the Camp David session—Eliot Cohen.

Cohen had regretted not taking a firmer stance back in June, so this time he steeled himself and advised the president to fire some of his generals. He knew that Bush was disinclined to challenge the judgment of commanders, so Cohen rattled off a list of presidents and prime ministers in history—most notably Winston Churchill and Abraham Lincoln—who had done just that, with fruitful results.

When Bush asked who should replace General Casey as the commander in Iraq, Cohen had a ready answer: David Petraeus.

Some of the NSC advisers had mentioned Petraeus as a strong candidate for the job, especially Meghan O’Sullivan, Bush’s top aide on Iraq, with whom Petraeus was still in regular contact. The president now went around the room, asking the other outsiders if they agreed with Cohen. They all nodded. Petraeus, they said, was the right man.

Keane launched into a very abbreviated version of his case for a counterinsurgency strategy and more troops. Cohen and Biddle, the civilians, agreed with him. McCaffrey and Downing, the other retired generals, did not, arguing that a surge probably wouldn’t work and couldn’t be sustained, in any event.

After the meeting, Keane followed Cheney down the hall into his office. He went through the fifty-five slides from Kagan’s briefing. Cheney said he’d show them to the president.

•  •  •

One week later, on December 18, Robert Gates was sworn in as secretary of defense. The next day, he flew to Iraq. Among the aides flying with him was Eric Edelman, the undersecretary of defense for policy. For the previous year and a half, Edelman had been handing out copies of John Nagl’s Learning to Eat Soup with a Knife to every Iraq policy hand he knew. He’d also been instrumental in pushing DoD Directive 3000.05—the policy statement that declared stability operations were “a core US military mission,” to be given “priority comparable to combat operations”—through the Pentagon’s top echelons.

From his White House contacts, Edelman knew that Jack Keane had given Cheney the slides from Fred Kagan’s briefing and that Cheney had, in turn, shared at least their conclusions with Bush. Edelman knew Kagan. He’d been friends with his father. And as a Soviet-affairs specialist in the Pentagon during George H. W. Bush’s presidency, he’d hired young Fred as an intern. Knowing that he was going to be on the plane with Gates, Edelman called Kagan and asked for a set of the same slides. Kagan sent them over. En route to Iraq, Edelman brought them over to the new secretary of defense and said, “The president has seen these. You should, too.”

What none of these people knew—not Gates, Edelman, Casey, Abizaid, nor, for that matter, Kagan—was that Keane had also shared the slides with Ray Odierno.

It had been announced in the spring that Odierno would be returning to Iraq at the end of the year to replace Pete Chiarelli as deputy commander of US forces.

Odierno had come a long way in the two-and-a-half years since his first tour, when he was commander of the 4th Infantry Division and Keane criticized him (as others would later do) for running operations that inflamed the insurgency. He left Iraq in the summer of 2004, knowing that he’d made mistakes and determined to learn as much as he could about Iraq and counterinsurgency before going back. He spent the next year and a half as the JCS chairman’s liaison to the State Department and accompanied Condi Rice, Phil Zelikow, and Eliot Cohen on their trips to the war theater, gaining a broader perspective than he’d managed either as a division commander or, years earlier, as a rising artillery officer. In May 2006 he was named commanding general of Army III Corps in Fort Hood, Texas, and spent the next seven months—before the multidivision corps was scheduled to rotate into Iraq—organizing seminars, conferring with experts (including Cohen and David Kilcullen), reading lots of books (including the draft of David Petraeus’s COIN field manual), and taking the COIN-heavy Battle Command Training Program that Petraeus had created at Fort Leavenworth.

A crucial moment in this reeducation took place late that fall, when an old friend, General J. D. Thurman, came to Fort Hood for a visit. Thurman was on break from a tour in Iraq, where he was serving as Odierno’s successor, commanding the 4th Infantry Division. Over cigars and brandy, they talked about the war’s frustrations. By this time, many officers were at least somewhat versed in counterinsurgency theory, but they were ill equipped to apply it in practice. Thurman’s troops had tried to “protect the population,” and they’d swatted away the insurgents from an area with little problem. But then they’d turn over the piece of land to the Iraqi army, and within a month, the insurgents would return and resume full control.

Suddenly, in Odierno’s mind, everything clicked into place. He remembered the memo he’d helped Zelikow write on the plane back to Washington from Baghdad in the fall of 2005, the memo that introduced the phrase “clear, hold, and build.” Thurman was now telling him, essentially, that his troops could clear an area of insurgents but that the Iraqi soldiers couldn’t hold it. Until this moment, Odierno had found much of the COIN reading material a bit abstract. Now it struck him as stunningly tangible. The problem in Iraq seemed obvious: the American commanders were turning over operations to the Iraqis too quickly. The solution seemed equally clear: we had to stay in place longer; we had to clear and hold.

Keane sent Odierno the data from the Kagan-Armstrong-Dwyer study at just the right moment. How to manage and sustain a clear-and-hold strategy (the build part, as in clear-hold-build, seemed at this point a long way off) would be the key task and biggest challenge of Odierno’s new job. And from his own background in quantitative analysis, he knew that the only way to meet this challenge was somehow to get more troops.

Odierno did not know that Keane was sharing the same data with the White House or that Edelman had passed on the same information to the new defense secretary. So when Gates met with Odierno in Baghdad and asked if he could use some more troops, he quickly affirmed that he could. He didn’t realize—nobody in the room realized—that his answer and the secretary’s question had stemmed from the same analysis, the same source. Neither of them knew that Keane had been pushing the top officials on both sides of the water toward the same conclusion.

Casey was also in the room. He was still the commander and would remain so for another seven weeks. (His replacement hadn’t yet been announced, but rumors were flying that it would be Petraeus.) Odierno had privately told some officials that he needed an additional five Army brigades and two Marine regiments—reflecting the data Keane had passed along. But Casey, while ceding somewhat to the new political realities, had told his new deputy before the meeting with Gates that he would agree to no more than two extra brigades—one in Baghdad and a Marine equivalent in Anbar Province. So Odierno, who’d been on the job for barely a week and was respectful of hierarchy (while knowing that the scene was about to change very soon), parroted his commander’s reply.

Soon after this session, Gates met with Prime Minister Maliki to sell him on the idea of accepting two extra American brigades. After three hours of suasion, Maliki finally signed on.

On the plane ride home, Edelman and some of the other staffers, mainly from the NSC, wondered if they’d sold themselves short. Were two extra brigades just enough to fail? Should they have gone for the full seven, for everything they could muster? Gates said he wasn’t opposed to the idea but that he would have a hard time convincing Maliki.

•  •  •

Meanwhile, back at Fort Leavenworth, Petraeus was following the whole chess game through his elaborate network of contacts in the national-security bureaucracy. He’d kept track of every move: the council of colonels, McMaster’s influence on Kagan, Keane’s pushing the Kagan study through the White House and to Odierno.

Now that Rumsfeld was gone, Casey on his way out, and Bush on the verge of pushing for a surge and a shift in strategy, some of the holdouts among the top brass were gearing up a case that the two extra brigades allowed by Casey were all that the Army could send; there weren’t any more available. The campaign was beginning to take its toll. In a recent interagency meeting, Secretary of State Condoleezza Rice had opposed a surge because her counselor, Philip Zelikow, who might have favored the idea in principle, had been told by one of his Pentagon contacts that a surge was simply infeasible. Though Zelikow possessed enormous self-confidence, he would not have presumed to challenge an Army general on such an esoteric calculation.

Petraeus rallied his own resources at Leavenworth and elsewhere to identify the five brigades that could be mobilized over a six-month period to Iraq. Keane, who by this time was delving into the issue and its politics with the imprimatur of the president and vice president, finally prodded General Pace, the JCS chairman, to concede that the task was not physically impossible.

Bush scheduled a meeting of all his national-security advisers for December 28 at his ranch in Crawford, Texas, where he was spending the holidays. The question at this point, everyone realized, wasn’t whether to send more troops but how many.

Petraeus heard from Meghan O’Sullivan, his main source in the White House, that Casey was holding firm on no more than two extra brigades, with the further condition that only one of those brigades should deploy directly to Iraq; the other one should stay in Kuwait as a backup, to be moved on to Baghdad or Anbar Province only if needed. Although he didn’t say so, Petraeus had heard the same report from his contacts in the operations directorate of the Pentagon’s Joint Staff.

O’Sullivan asked Petraeus what he thought of the idea. He replied that it would be a disaster. First, he said, two brigades weren’t enough. Second, rolling them in one at a time would make the situation worse; every time he asked for an extra brigade, it would look like an admission of failure. Petraeus knew that he topped the extremely short list of candidates to replace Casey as commander. But, he told O’Sullivan, if the president approved Casey’s idea—if, in fact, he decided on any sort of surge smaller than five brigades—he would turn down the offer, because he simply couldn’t succeed with such skimpy resources.

He knew that O’Sullivan would pass this message on to Hadley, who would in turn convey it to Bush. It was an audacious ploy, a bit of a gamble. Would he really say no if the president asked him personally to take the job? Three-star generals weren’t supposed to reject a request from their commander in chief. Certainly they didn’t do so and retain hopes of earning a fourth star. Still, Petraeus wanted to clarify, dramatize, the stakes. And he was telling her the truth: he didn’t think he could do his job under Casey’s conditions.

Meanwhile, Keane was hearing from his contacts that Pace would be briefing Casey’s proposal at Crawford as the Joint Chiefs’ official position. He called John Hannah, Vice President Cheney’s national-security adviser, to give him a heads-up on what was going on and to make the same case against Casey’s position that Petraeus had laid out to O’Sullivan. He threw Hannah one additional line of argument in case Pace went ahead with the briefing. He should tell Cheney to ask Pace whether those extra two brigades would constitute “a decisive force” for victory. The key word was “decisive”: Pace would have no choice but to answer, “No.”

Keane also circulated word that Odierno, with whom he was talking on the phone once or twice a week, would much prefer a larger surge force and that he and his staff were developing a campaign plan to exploit the full five-plus-two brigades. If the deputy commander wanted a full surge, if the top candidate for new commander was threatening to turn down the job if he didn’t get the full surge, and if even the nation’s top general had to admit that something less than a full surge wouldn’t be a “decisive force,” Keane figured the game might be won.

•  •  •

On January 4, 2007, administration officials told reporters that President Bush would award David Petraeus a fourth star and name him as the new commander of US and coalition forces in Iraq. Six days later, in a nationally televised prime-time speech, Bush declared that the situation in Iraq was “unacceptable” and that “we need to change our strategy.” To facilitate the change, he had decided to send “more than twenty thousand additional American troops” to the battlefield—five Army brigades to Baghdad and another four thousand marines to Anbar Province (a bit less than Kagan’s two Marine regiments but deemed enough).

“In earlier operations,” Bush said, “Iraqi and American forces cleared many neighborhoods of terrorists and insurgents, but when our forces moved on to other targets, the killers returned. This time we will have the force levels we need to hold the areas that have been cleared.”

COIN was now the official policy.