Learning From Our Defeat: The Assumptions of Donald Rumsfeld

This review essay is part of the “Learning From Defeat” project. It is the first of thirteen. An introduction to the project can be found here.

One hopes for statesmen chastened by defeat. In this world of our hopes, the authors of catastrophe would discuss their mistakes with the humility, introspection, and sense of disgrace these mistakes deserve. Decisions that led to death—death in its thousands and hundreds of thousands—would be examined with probing honesty. The decision makers behind them would be seized with a fierce guilt and urgency. They would quest to understand the nature of their errors. They would incessantly press upon us the lessons of experience, gripped with fear that the next generation might repeat their calamities.

One can imagine such a statesman, chastened by defeat. Douglas Feith is not he.

Feith served as the Under Secretary of Defense for Policy (USDP) in the Pentagon of Donald Rumsfeld. This position is sort of the #3 spot in the civvie side of the DoD. Where the Secretary of Defense is responsible for managing the entire Department, as well as serving as a key “principal” on the president’s National Security Council, and the Deputy Secretary of Defense usually takes charge of the nuts and bolts of budgets, acquisitions, personnel, and daily operations, the USDP is tasked with issuing, reviewing, and refining the strategic guidance that commanders in the field receive from civilian leadership, representing the DoD in interagency “deputy meetings,” drafting memos directed to other organs of government, and developing the intellectual framework that guides military operations and defense posture.

For the first five years of the Bush presidency, Feith was this man. He was the bureaucrat at the center of America’s immediate response to 9/11 attacks, the invasion of Afghanistan, the prolonged preparations for the invasion of Iraq, and the mismanaged occupations of both countries. Feith was at the key meetings where decisions of war and peace were decided; his office developed the main strategic rationale behind the Bush administration’s “Global War on Terror.” War and Decision: Inside the Pentagon at the Dawn of the War on Terrorism is his memoir of these years. It is a full throated defense of his—and Secretary Rumsfeld’s—role in these affairs.

Feith’s book is equal parts fascinating and frustrating. Fascinating because its full 674 pages offers a far more detailed look into the nuts and bolts of policy making than most DC memoirs  dare. Feith quotes liberally from policy memos that he or his department sent out, as well as from notes he took in various National Security Council (NSC) and NSC deputies meetings. This level of detail and the book’s reliance on primary source documents is unusual—I have not found a memoir from the Obama years that is as forthright in its presentation or as detailed in its reporting as War and Decision. To top it off, Feith is sharp in his thinking and concise in style. Everyone who reads this book will come away with an appreciation for the intelligence, clarity, and careful thought that imbue Feith’s writing.

But Feith’s obvious intelligence is also a source of frustration. This man is simply too smart to not see the consequences of the policy he crafted. Feith wrote this book in 2008, by which time several hundred thousand Iraqis lie dead and Afghanistan had transmogrified from international success story to international quagmire. The connection between the administration’s policies in the first half of the Bush presidency and their ghastly fruits in the second half were then quite clear. Feith does not deny these realities so much as dodge culpability for them. He is more than willing to acknowledge various mistakes by his administration—but these mistakes are admitted in a funny way: every error, all the problems with Iraq, Afghanistan, or anywhere else, came about because the American government was not willing to go all in on DoD’s plans. The Bush administration’s fatal flaw, in his view, was that it just was not Rumsfeldian enough.

This is an extraordinary position to take. It brooks no compromise. It offers little in way of humility or introspection. It is utterly unrepentant.  But this position is useful in its own fashion—by all accounts, the way Feith wrote his book mirrors how he ran DoD’s policy team. War and Decision is thus not only a window into the thinking behind early Bush era foreign policy, but the attitudes and leadership style that determined how this thinking was turned into bureaucratic action.

Feith’s book is about much more than Afghanistan. The majority of the book, in fact, deals with Iraq: the rationale for invasion, the interagency battles over intelligence, war planning in the lead up to Iraq, and the bungling of the post invasion occupation each take up a hundred pages or more of the book’s narrative. The Pentagon of Donald Rumsfeld was the driving force behind all of these. But this review is about Afghanistan, not Iraq. To Secretary Rumsfeld’s chagrin, DoD played second fiddle in initial invasion planning—until the Taliban was driven into Pakistan, that role was played primarily by the CIA, which had the contacts on the ground and the necessary area knowledge needed to support the Northern Alliance’s southern offensive. 1


A fluid account of the CIA’s planning is found in Steve Coll, Directorate S : The CIA and America’s Secret Wars in Afghanistan and Pakistan (New York: Penguin, 2018), 12-112. In his memoirs Rumsfeld laments that the Department of Defense had to rely on decades old British maps for their own planning; No such complaint can be found in any of the CIA memoirs from the same time period. Donald Rumsfeld, Known and Unknown: A Memoir (New York: Sentinel, 2011), 369-70. For Rumsfeld’s mounting frustration with DoD reliance on CIA planning, see Feith, War and Decision, 104, 106.

Yet if the CIA took the operational lead in bringing American forces into Afghanistan, it was the Pentagon that developed the strategic rationale for the American presence there. The 2001 campaign was nested in a larger effort. The architects of this effort—which we now know as the “Global War on Terror”—were concentrated in the Pentagon. It is an exaggeration to credit Feith’s policy shop for the entirety of the administration’s grand strategy. Yet they shaped it more effectively than any other body in the federal government. The same memos that Feith draws on to write this book were one of the secrets to their success. Feith explains:

Defense officials did have influence, but this owed much to a mode of operation that was the opposite of conspiracy: We created a transparent record of the facts and reasoning we used to support our proposals. Bush often complimented Rumsfeld’s memos, which addressed him at the level on which he liked to operate—that of strategy, not tactics. They were analytical and mercilessly edited, giving the results of much thought in few words. And they showed confidence that the ideas they contained, when reduced to print on a page, could retain potency and withstand scrutiny over time, unlike arguments that derive their force from the personality of the advocate. 2


Feith, War and Decision, 62; Rumsfeld makes a similar point in Known and Unknown, 324.

George W. Bush was contemptuous of detail. He often derided the minutia of policy as “small ball” not fit for a commander in chief; he wanted bold, large picture conceptions that presented clear yet creative policy options to decide between. He demanded concision and admired powerful rhetoric.3 These demands were well suited to Donald Rumsfeld, whose natural mode of analysis was the bold-yet-pithy memo. Rumsfeld was also unafraid to reach above his station and write from a presidential (as opposed to departmental) perspective.4 As with much of Rumsfeld’s behavior, this posturing was extremely effective both at shaping the president’s thinking and in fomenting resentment in other executive departments and agencies. Departments less eager to battle out the big picture—especially State, whose leaders were not the cerebral idea-smiths running DoD but were instead practical operational types—struggled to influence policy. For the first five years of Bush’s presidency, Rumsfeld and his underlings drove the agenda.


On the phrase “small ball” see George W. Bush, Decision Points (New York: Crown Publishers, 2010), 195; Ari Fleischer, Taking Heat: The President, the Press, and My Years in the White House (New York: HarperCollins, 2005), 31.

This description of Bush’s attitude towards briefings and memos in the early years of his presidency draws from Decision Points, 272; Fleischer, Taking Heat, 31-33; Bob Woodward, Bush at War, large reprint ed.(New York, Thorndike Press, 2003),, 243; Michael Mazarr, Leap of Faith: Hubris, Negligence, and America’s Greatest Foreign Policy Tragedy (New York: PublicAffairs, 2019), 100-101; Karl Rove, Courage and Consequence: My Life as a Conservative in the Fight ((New York: Simon and Schuster, 2010), 106-108; Nicolas Lemann, “Without a Doubt: Has Condoleezza Rice Changed George W. Bush, or Has He Changed Her?” New Yorker, October 14, 2002; James Bennet, “C.E.O., U.S.A.,” New York Times Magazine, January 14, 2001. Robert Draper’s Dead Certain and To Start a War also provide valuable pictures of these dynamics in action.


Feith, War and Decision, 52-3.

To understand this agenda, we must go back to 9/11. Rumsfeld, Wolfowitz, Feith, and their allies in the White House staff believed that the world must undergo radical changes. The stakes were incredibly high: Feith, along with many administration officials, believed that these terrorist attacks were a fundamental challenge to the American project. Feith lays this view out at greatest length in a comment on Bush’s September 20th, 2001 address to a joint session of Congress:

In that speech, Bush twice used the phrase “way of life.” These struck me as his weightiest words. The President noted that our terrorist enemies killed in order to destroy a way of life, and he described the importance of “defeat[ing] terrorism as a threat to our way of life.” …it encapsulated my thinking about 9/11’s strategic significance and I championed it at Deputies Committee meetings and in extended conversations with Rice’s deputy, Stephen Hadley. The phrase called attention to the stakes in the terrorist challenge—the openness, humane liberality, and personal freedom that define our society. I looked at our new terrorism problem this way: If America were not a liberal democracy, a country that recognizes the worth and political equality of its individual citizens and respects their civil liberties, it would cease to be America. Our country is not so much a land and a people as it is a way of life that embodies an idea—the idea of individual freedom….

Much of what makes Americans happy—our self-governance, economic prosperity, domestic tranquility, and opportunity to better ourselves—derives from the liberal and democratic nature of our society and the degree of mutual trust (sometimes called social capital) that such a society engenders. It is hard to overstate the moral and material benefits rooted in that trust, in our freedom. And this is what terrorism had the potential to undo. Beyond the cost in lives and property, the 9/11 attack—or, rather, our reaction to it—could have far-reaching consequences, especially if it were followed by more such attacks.5


ibid.,, 68-9.

In framing things this way, Feith is not so far removed from the many Bush administration critics who argued that the security theater of the post 9/11 moment threatened to erode the integrity of the American regime. But Feith takes a different tack from these civil libertarian: to ensure that the threat of terrorism did not deform our politics here, the threat of terrorism had to be defeated over there:

To preserve civil liberties, the President had to adopt a strategy of disrupting terrorist networks abroad, where they do much of their planning, recruiting, and training. He had to adopt a strategy of initiative and offense as well as defense. As I saw it, the President decided that, in dealing with the terrorists, he had the choice of changing the way we live or changing the way they live.6


ibid., 71.

This is the metric by which U.S. officials believed they should be judged. “It would be sensible for U.S. officials to understand their mission as defeating terrorism as a threat to American freedom and openness,” Feith writes, concluding that “if we fought the terrorists so effectively that they no longer threatened the nature of our society, the United States would have achieved a substantial victory.”7 In Feith’s own mind this was the basic rationale behind everything the Rumsfeld Pentagon did—from the occupation of Afghanistan to the invasion of Iraq. Keep this goal in mind as we read through the various statements of strategy offered by Feith, Rumsfeld, and their counterparts in this book, as well as those to come in the other books in this series of reviews.

Before we move to those statements, let’s return to Feith’s conception of the essential task of the GWOT (as with all quotes in this post, the bolded emphasis is my own):


ibid., 11.

“[Bush] had the choice of changing the way we live or changing the way they live.”

 There are several assumptions embedded in this phrase that are worth teasing out. Who exactly is the “they” Feith speaks of?  Did “they” mean bona-fide terrorists active in anti-American terrorist cells at that exact moment? Or did it also include dictators or authoritarian leaders currently sheltering terrorist groups? What about those who did not shelter terrorists themselves, but hosted radical imams that justified terror tactics? What about Shia outfits that the perpetrators of the 9/11 attacks would have considered enemies to their cause? Or is the “they” whose world that must be changed meant to include every young man who might be tempted to one day join a terrorist organization in the future? Does “they” mean the entire Middle East? Must Washington find some way to transform the course of an entire civilization–the grand totality of the political, social, and religious history that produced the ideology that inspired and conditions that allowed the 9/11 attacks to occur?

The question matters: if American policy is designed to disrupt terrorism at the source, the specific level at which this disruption should occur matters. Trying to degrade the capabilities of a terrorist cell is a fundamentally different project from trying to dissuade foreign leaders from tolerating the presence of terrorists on their soil, which in turn is different from trying to reshape the culture of the entire Middle East. A successful tactic on one of these planes might be detrimental on the others.

Feith dances around the problem in his discussion of the phrase “Global War on Terror”:

 The U.S. government could not simply define the enemy as a set of terrorist organizations together with states that helped them in one way or another. If we did, we could find ourselves declaring war against all countries that gave safe haven, funds, or ideological and other types of support to terrorists—a list that would include Afghanistan, Cuba, Iran, Iraq, Libya, North Korea, Pakistan, Saudi Arabia, Sudan, and Syria. This was clearly an unrealistic idea. It further complicated matters that the United States considered some of these states important friends. Moreover, a formal list of terrorist enemy organizations would require continual revision, reflecting the mergers, acquisitions, splits, and name changes that were common among them. We needed a better way to define the enemy, one that would cover all the relevant bases but preserve our flexibility regarding how, when, and against whom we should act.

We did not solve this puzzle [in his first planning discussions on September 12th]. The President eventually dealt with it by coining the term “war on terror,” declaring, in effect, that the enemy was not a list of organizations and states but certain inherently evil activities that included both terrorism and state support for terrorism. Though the term was imperfect—many commentators have noted the peculiarity of declaring war against a method of attack—I considered it an intelligent and useful stopgap that acknowledged the unprecedented nature of the challenge represented by 9/11. It avoided the problem of lists gave the President flexibility, and called attention to differences between us and our enemies on the issue of respect for human life…. The coinage of the term “war on terror” allowed the Administration to defer naming the enemy while considering these perplexing questions. The threat we faced was not abstract or remote. It was as real as the corpses and debris of 9/11.”8


ibid., 8-9.

The trouble is that the Bush administration never really did “consider these perplexing questions.” The “stop-gap” solution persisted to administration’s last day. Feith, for his part, never seriously attempts to define “they.” Nor did his superiors. Consider the following exchange, which Feith records as having occurred during a September 13th NSC meeting (where the CIA presented the first version of the eventual invasion plan):

Cofer Black, the head of the CIA’s Counterterrorism Center, gave a response [to the President] that jibed with Rumsfeld’s thinking: We could coordinate a U.S. military attack with the Northern Alliance in the north. Sending only Tomahawks would be like sending a letter saying, “We surrender.” We must have troops on the ground. Otherwise they’ll think we’re weak. Rumsfeld concurred: We don’t want to run the risk of being laughable, he noted.


ibid., 16-17.

This concern with looking “laughable” or “weak” is a recurring feature of the early Bush era debates which we shall return to later. But at the moment our focus is on the ambiguous audience whose perceptions of American strength matter so much to Rumsfeld et. al. Who is it the United States must not look weak to? Al Qaeda? The Taliban? Other terrorist groups? All of Central Asia and the Middle East? The imperative to impress the reality of American power on America’s enemies is a dominating theme of the 2001-2003 period; just who needed to be impressed is far less stable.  

Differing interpretations arose early. Later that same day—September 13th—Feith attended a Deputies committee where at least two conflicting versions of ‘they’ were presented:

The Deputies Committee (the highest subcabinet-level committee under the National Security Council) met on September 13 to discuss those options in preparation for the upcoming Camp David meeting. Deputy Secretary of State Richard Armitage declared in favor of Option 2, commenting that it was important to create a “shockwave” that would “disrupt the worldwide network” of the terrorists. Paul Wolfowitz, on the other hand, was unsatisfied with the options Rice’s staff had crafted, and endorsed none of them. Rather, he favored having a new paper written for Camp David, one that would clarify that the chief purpose of U.S. military action was not punishing those behind 9/11 but attacking those who might launch the next 9/11.

Wolfowitz warned against focusing narrowly on al Qaida and Afghanistan. The next 9/11, after all, could come from other organizations and places in the global terrorist network. He questioned whether Armitage’s preferred option—actions targeting only al Qaida and the Taliban—would be sufficient to create the “shockwave” Armitage was calling for. If we should take hasty action that produced only meager effects, he warned, it could embolden rather than discourage regimes that were assisting our terrorist enemies.10


ibid., 49.

Armitage and Wolfowitz present two different versions of the “they” whose lives must be changed by the American response to 9/11. Armitage’s “shockwave” presupposes a global network of terrorist organs working in concert, organizations whose operational tempo and OODA decision cycle might be disrupted by decisive military action. Wolfowitz’s “they” are state leaders that give safe haven—or might give safe haven to—anti-American terrorist groups.

Wolfowitz articulated what was to become the default DoD view of the question. The Bush national security team were old Cold Warriors; both their academic training and their specific Cold War experiences biased them to see the world through the lens of states-based actors. 11Their framing was hardly inevitable: in the late Bush administration the blob would come to see the root problem not as dark regimes aiding terrorists out of malice, but weak regimes unable to control terrorists operating in their own borders. The late Obama administration would try to extricate itself from the regime shaping game altogether, focusing their efforts on decapitation strikes and raids designed to dismantle terrorist organizations with the smallest military footprint and political commitment possible. But the early Bush team had different priorities:  


On this point see James Mann, The Rise of the Vulcans (New York: Penguin, 2004).

If we confined our response to al Qaida’s presence in Afghanistan, the terrorists there might simply migrate to other safe havens. Rumsfeld was intent on undoing not just al Qaida’s refuge in Afghanistan, but the policies of states around the world that supported or tolerated terrorist groups. He thought our goal should be to make it increasingly difficult for terrorists to find governments willing to support them.12


Feith, War and Decision, 19.

I suspect that this focus on “state sponsors of terror” was in part an unconscious attempt to make a hard problem more tractable. Feith presents 9/11 as a rupture point in the American understanding of security and deterrence:

The 9/11 attack made the Iraq problem appear more grave and urgent, though not because any officials assumed Iraq was involved in the hijackings. Rather, it was because al Qaida had demonstrated that America’s enemies had the opportunity, will, and means to cause devastating harm to our country without having to defeat us militarily. With this in mind, U.S. officials reevaluated all major national security threats—and especially the danger that terrorists might acquire weapons of mass destruction. There was intense concern that a hostile state, working with a terrorist group, could precipitate a catastrophic attack against us.13


ibid, xi.

The invasion of Afghanistan (and later Iraq) was in some ways an attempt to reverse this process. The United States national security complex was extremely skilled at deterring conventional military forces from aggressive action; the Bush team of NatSec A-listers had a distinguished track record bending foreign states to the American will. Conceptualizing the dilemmas of the war on terror as a problem of rogue regimes put new developments back into familiar territory. It also made these developments more tractable to the military tools at Bush’s disposal. Hammer, meet nails.

This decision to conceptualize the Global War on Terror as a multi-tiered campaign to dissuade dictators from supporting terrorism had far reaching consequences. The most important of these was that initial operations in Afghanistan were not really about Afghanistan. As the goal of the campaign was just as much to frighten other potential sponsors of terror as it was to change “the way they lived” in Helmand or Kandahar, Afghanistan did not even need to be the first theater of operations:

If the President directed strikes, however, Rumsfeld wanted to reiterate [in a September 2001 memo] that they should be designed to produce “impressive results,” which they might if they hit:

• Al-Qaida forces and assets in more than one country, including some outside the Middle East.
• At least one non-al-Qaida target—e.g., Iraq [and other specified locations].
Rumsfeld described surprising the enemy as “a crucial operational value.” Accordingly, the widespread expectation that the United States would hit Afghanistan “argues for the initial strike to be directed some place else and preferably someplace like South America or Southeast Asia.” 14


ibid, 66.

South America! That response seems to have been a one-off thought on Rumsfeld’s part. One of the other targets on his list, however, was raised repeatedly and raised early. Two days after the attacks, we find Rumsfeld

Looking beyond bin Laden and Afghanistan. Rumsfeld mentioned Saddam Hussein’s Iraq as a threat to both its region and to the United States. Iraq, he observed, was a state that supported terrorism, and that might someday offer terrorists weapons of mass destruction to use against us. Unlike Afghanistan, however, Iraq also had substantial infrastructure and military capability. In Iraq, he noted, we could inflict the kind of costly damage that could cause terrorist-supporting regimes around the world to rethink their policies. 15


ibid., 16-17.

 We know how that story turned out. But there were other, more immediate consequences of the GWOT’s war aims. Conceptualizing American operations in Afghanistan as a sort of performance before an audience of rogue regimes changed the way those operations were conducted. What kind of political settlement these operations produced in Kabul was far less important to Rumsfeld than how “impressive” a demonstration they were to the rest of the world:  

For weeks, Rumsfeld had been concerned that our initial military action after 9/11 would look puny. If our efforts were “not confidence-inspiring,” he feared, we would signal to the world—in particular to state supporters of terrorism—that the United States was still not serious about destroying terrorist network.16


ibid., 96-7.

These goals would resurface in October, when NSC met to plan out the length of the coming campaign. Colin Powell argued that the campaign might need to extend through the winter. Feith did not like this idea:

When I briefed Rumsfeld for the October 11, 2001, Principals Committee meeting on Afghanistan, I suggested that the Principals review the U.S. government’s goals. I thought it would be good to clarify, for example, that one key purpose of ousting the Taliban was “to make an example of them” as a state sponsor of terrorism. That would help explain why our operation was urgent: We wanted to disrupt the international terrorist network—to shock the system—as formidably and quickly as possible… If our only goal was to punish the Taliban, however, it would matter far less how swiftly we beat them.

…The Northern Alliance fighters would have their own ideas about how to proceed [with how to control Afghanistan]. We had influence, but not necessarily control. So it would help to get government-wide agreement on what our goal was not: “Creating a stable, post-Taliban Afghanistan is desirable, but not necessarily within the power of the US,” I wrote. “The US should not allow concerns about stability to paralyze US efforts to oust the Taliban leadership.” Rather, we should “facilitate capture before winter, if possible, of key cities—Mazar-e-Sharif, Herat and Kabul.” 17


ibid, 100-1.

This is a remarkable passage. It encapsulates the priorities and problems of U.S. strategy in the opening act of America’s intervention in Afghanistan. Feith saw military operations in Afghanistan as a choice between a punitive campaign of retribution and a preemptive campaign designed to disrupt future Al-Qaeda operations and build an international environment inimical to terrorist activity. Rumsfeld was laser focused on the latter aim. No matter how many times I read them, his articulation of the campaign’s war aims amazes me. Feith quotes a memo from September 30th, sent by Donald Rumsfeld to the president, that lays them out:

If the war does not significantly change the world’s political map, the U.S. will not achieve its aim. There is value in being clear on the order of magnitude of the necessary change. The USG [U.S. government] should envision a goal along these lines:

• New regimes in Afghanistan and [some other states] that support terrorism (to strengthen political and military efforts to change policies elsewhere).
• Syria out of Lebanon.
• Dismantlement or destruction of WMD in [key states].
• End of many other countries’ support or tolerance of terrorism.18


ibid, 89.

 It is unlikely that toppling the Taliban could ever achieve such lofty aims. That realization was perhaps the most important driver of the administration’s decision to invade Iraq the following year.19 A more impressive display of power was needed to create the international environment the Bush team sought. Yet the parallels between the two invasions are legion: Just as Feith and his team were did not see a “stable, post-Taliban Afghanistan” as a central goal of Operation Enduring Freedom, so they refused to plan for building a stable, post-invasion Iraq. Rumsfeld was resolutely against anything that smelled of ‘nation building.’ In both Iraq and Afghanistan Rumsfeld argued against taking sides in inner political disputes or expending resources to strengthen the capacity and authority of the newly installed regimes. Unlike President Bush, Rumsfeld did not subscribe to the “you break it, you own it” theory of intervention. Breaking things was the entire point. Fixing them, on the other hand, was an uncertain venture sure to suck up time and dollars, but not guaranteed to produce political stability—much less liberal democracy—on the long run. Better to just break things and leave.


See especially Mazarr, Leap of Faith, 109-168.

This is a crude but accurate simplification of Feith and Rumsfeld’s position.20 In Iraq the problems of this logic would quickly become apparent: an Iraq governed by a weak, slap-dash state was far more hospitable to Salafi-Jihadist terrorist outfits than an Iraq suffering under the iron hand of Saddam Hussein. The world’s dictators might have quaked in fear as they watched the Baathists topple; the world’s terrorists, quite unafraid to sacrifice their own lives for the anti-American cause, would not be deterred by displays of military might, and only found opportunity in the anarchic conditions of the post-invasion Iraq.


Donald Rumsfeld, “Beyond Nation Building,” Speech given at 1th Annual Salute to Freedom,
Intrepid Sea-Air-Space Museum, New York City, Friday, February 14, 2003; Known and Unknown, 483-4; Feith, War and Decision, 149-150.

A similar error marred the administration’s approach to Afghanistan. Feith juxtaposes the punitive approach (‘we go into Afghanistan for retribution’) with the demonstrative (‘we go into Afghanistan to coerce untrustworthy states across the region with its example’). It is interesting that it never occurs to him to discuss—even if to disprove or dismiss—a third option: going into Afghanistan to create a stable political environment in Afghanistan, ensuring that the country offered no spot for outside terrorist outfits to take root in. Rumsfeld and company never seriously considered this approach. It did not mesh well with their general distaste for nation building. Nor did it address their concern that unless change happened at the level of the international system, fighting terrorists would devolve to an unending game of whack-a-mole as terrorists migrated from one country to another.

What if they had made Afghanistan’s political stability an explicit goal of American foreign policy? This would require the administration to take two steps that cut against their wider goals in the Global War on Terror.

Before the Taliban recaptured Kandahar in 2006 (substantially increasing the Taliban’s manpower reserves), total Taliban insurgents numbered less than 10,000. This was a threat a properly sized Afghan National Army could have overwhelmed and then defeated. Rumsfeld refused to fund an Afghan security force this large. 21 A larger post-invasion presence or substantially larger dollar commitment to training and equipping the Afghan National Army might have made a difference then, when the Taliban was a few and scattered, than it could have in later years, when the Taliban was a far more potent insurgent force. Yet large post-invasion commitments to Afghanistan were not compatible with Washington’s determination to reshape the entire region. Money and troops that might have been used in securing Afghanistan’s future were needed for “changing the way they lived” elsewhere. Tying American resources down to one locale dampened their coercive potential.


Carter Malkasian suggests a 150,000 man Afghan army would have been sufficient for this purpose. See The American War in Afghanistan: A History (Oxford: Oxford University Press, 2021), 156.

There is an even more important decision made in these years: the decision to refuse all negotiations with the Taliban. Feith does not consider this decision notable enough to mention in his memoir. Only once does Feith describe a debate over what American policy should be towards a defeated Taliban:

Jalalabad surrendered on November 14. That day, the Deputies Committee discussed the Taliban’s ongoing loss of control of the country. Anticipating the discussion at an upcoming National Security Council meeting, Stephen Hadley raised the question of how the United States should respond if Taliban leaders offered to surrender and we came under pressure to halt U.S. military operations. Wolfowitz noted that we hadn’t yet completed our military mission against the Taliban and al Qaida, so a cease-fire would be premature. Who, he asked, would be in a position to accept a Taliban surrender?22


Feith, War and Decision, 133.

Later that year Hamid Karzai would be in a position to accept a Taliban surrender. He pushed for a negotiated settlement with Taliban leaders that would offer them amnesty and perhaps a role in the new Afghan government. If the primary ‘military mission’ in Afghanistan was securing a stable political order that would not shelter terrorists, a negotiated surrender of the remaining Taliban leaders would be the ideal solution. Karzai understood this. The Americans would have none of it. They not only announced that they would not allow the Taliban any role in the new Afghanistan, they made a list of some 30-plus high-level Taliban leaders that American and Afghan officials were banned from negotiating with.23 This stand makes sense in the context of the actual ‘military mission’ of U.S. forces–not post-invasion stability, but “making an example of [the Taliban] as a state sponsor of terror.” Refusing to give defeated Taliban a role in the new order was a powerful way to enforce this lesson.24


Coll, Directorate S, 101-2, 140-44; Malkasian, The American War in Afghanistan, 73-76; 99-101.


Rumsfeld makes this explicit in Known and Unknown, 368.

I suspect American intransigence on this point reflected very human emotions: roiling anger, a thirst for retribution, indignation at justice unserved. The administration was never able to admit that these things were driving U.S. policy—indeed, that they should have been driving U.S. policy. Witness Feith’s frustration with a comment by Zbigniew Brzezinski on the need to ‘punish’ the Taliban:

Zbigniew Brzezinski, President Carter’s erudite, though sometimes erring, National Security Adviser, told the Christian Science Monitor: “When we started out, we were going to smash Al Qaeda and punish the Taliban. Now we seem to be getting engaged in an Afghan civil war, almost as an end in itself. That could be a quagmire.”

Brzezinski was wrong, not just about the quagmire, but about the fundamental strategic goal of our attack—which was never to punish the Taliban, but to pressure state supporters of terrorism globally and thereby disrupt terrorist planning and operations. Brzezinski wasn’t entirely at fault: His misstatement reflected, at least in part, the Bush Administration’s failure to explain our actions clearly to the public. It was an early sign of the public communications problem that would persist for years, with ill consequences of strategic importance—especially later in Iraq.25


Feith, War and Decision, 125.

Feith was not the only one to take this stand—his boss would often say that the Global War on Terror should not be about the “Three R’s:’ Retribution, Retaliation, or Revenge.” 26 Feith viewed ‘punishment’ like this as a matter for “law enforcement”—and that is how terrorism was treated before 9/11, as a question for the FBI, not the DoD, to handle. As Feith saw it, “considering 9/11 an act of war, not just a crime, was a way of recognizing that retaliation would be an insufficient response.”27


Jay Nordlinger, “The Commanders and our Culture,” National Review (5 January 2005). See also Known and Unknown, 342, 355.


Feith, War and Decision, 125.

But if retribution was an insufficient response, it was still an undeniable motivation behind the American response. When Bush highlighted a desire for revenge (or “justice”) in his public speeches, he was not failing to communicate to the public what the administration was doing—he was expressing what the public wanted to hear. We wanted revenge. We supported the war in Afghanistan so that we might have it.

The celebrals running DoD could not bring themselves to admit a base emotion like this was driving their policies. This was a terrible mistake. American officials should have enshrined ‘revenge’ as an official war aim. Had they done so, they could have then had an open conversation about which targets were the proper objects of revenge and what actions might satisfy American demands for retribution. Dodging these discussions did not cause American soldiers or statesmen to act in a less vengeful fashion28, but it did allow the bloodlust-driven target list to expand irresponsibly. The spirit of vengeance fueled the American war machine even as its wrath fell on targets at further and further remove from the 9/11 attacks. This was true both inside Afghanistan (where little distinction was made between Al-Qaeda and low level Taliban) and outside it (culminating ultimately in the invasion of Iraq).  

The strategy that Rumsfeld, Feith, and their compatriots in the Pentagon devised to prosecute the War on Terror was built on a series of flawed assumptions. In Iraq, the consequences of similarly flawed assumptions were apparent within months. In Afghanistan these consequences festered slowly over time, only manifesting themselves after the lot had been cast out of office. But their assumptions killed people all the same.

There is a sad irony to this outcome. As Secretary of Defense, Donald Rumsfeld was famous for subjecting generals and other military briefers to excruciating, hour-long dissections of every assumption Rumsfeld could identify in their presentations. Rumsfeld believed that the failure to identify and examine the assumptions behind operational planning would lead to defeat. 29What we lacked was a president capable of subjecting Rumsfeld’s own memos to Rumsfeld’s own exacting standards. On this, Rumsfeld was right: bad assumptions have paved the way to American defeat.

This has been the first essay in a series titled “Learning From Our Defeat.” Next week’s review essay will consider Dov Zakheim’s A Vulcan’s Tale.


For an an especially destructive example, see Malkasian, The American War in Afghanistan, 110-113.


Feith, War and Decision, 100-101. One senior official who observed many of these briefings described briefing the Secretary as akin to “volunteering to be wire-brushed in the nude.” Mazzar, Leap of Faith, 91. See also Michael Gordon and Bernard Trainor, Cobra II: The Inside Story of the Invasion and Occupation of Iraq (New York: Pantheon Books, 2006), passim.

Leave a Comment


It sounds like they learned and absorbed very little about Al-Qaeda and such groups even after the 9/11 attacks, if they were still fixated on the idea of “state sponsors of terrorism” when the main form of Taliban “sponsorship” of Al-Qaeda was just to give them a place to hang out.

That to me is the most damning part of what you said. I can understand a bunch of former Cold Warriors not really understanding an international terrorist threat that didn’t have state sponsors before 9/11, but after 9/11, when they had the full apparatus of the intelligence and military service on hand to give them analysis of what Al-Qaeda truly was?

“Better to just break things and leave.” Punitive expeditions have a long history, and frequently work. They are economy of force exercises which recognize the limitations on the capabilities of the punisher. The wrongdoer is harmed and deterred from similar future actions. Punishment of wrongdoers for a particular wrong is a legitimate and achievable policy framework. Changing the world so that an entire category of wrongful deeds can never occur again is not achievable and should not have been tried.

Christopher Lawrence of the Dupey Institute put the Afghan insurgency numbers at ~20-80,000 in 2004.

I believe he gets this number from the observed (known) deaths caused, combined with various population metrics. I am not sure what algorithm he was using. But he does number crunching like this for a living, so he would have a good feel for it.

He also notes that the various powers that be seemed to consistently underestimate the number of insurgents within the country and the extent of the problem.

Which means that it would have been a very tough slog at best to “win” in Afghanistan by brute combat. Coopting rival elites (as Kharzai wanted to do) would clearly have been the better option. Jacqueline L. Hazelton in her recent “Bullets Not Ballots: Success in Counterinsurgency Warfare” (Cornell Studies in Security Affairs) by Jacqueline L. Hazelton pretty much stated that you win by cooping rival elite groups.

Very interesting post.

Also interesting to compare this to killing Bin Laden during the Obama administration. Even though Pakistan may have been involved in letting Bin Laden hang out, the decision was to just attack Bin Laden and his compound, instead of the entirety of Pakistan. I wonder if this would have been a winning strategy in Afghanistan, or if it was just a lucky choice for the moment.