The post-invasion phase of the Iraq mission has been the least well-planned American military mission since Somalia in 1993, if not Lebanon in 1983, and its consequences for the nation have been far worse than any set of military mistakes since Vietnam. The U.S. armed forces simply were not prepared for the core task that the United States needed to perform when it destroyed Iraq’s existing government—to provide security, always the first responsibility of any sovereign government or occupier.
The standard explanation for this lack of preparedness among most defense and foreign policy specialists, and the U.S. military as well, is that Secretary of Defense Donald Rumsfeld and much of the rest of the Bush administration insisted on fighting the war with too few troops and too Pollyannaish a view of what would happen inside Iraq once Saddam was overthrown. This explanation is largely right. Taken to an extreme, however, it is dangerously wrong. It blames the mistakes of one civilian leader of the Department of Defense, and one particular administration, for a debacle that was foreseeable and indeed foreseen by most experts in the field. Under these circumstances, planners and high-ranking officers of the U.S. armed forces were not fulfilling their responsibilities to the Constitution or their own brave fighting men and women by quietly and subserviently deferring to the civilian leadership. Congress might have been expected to do more as well, but in fact it did a considerable amount of work to highlight the issue of post-invasion planning—and in any case, it was not well positioned to critique or improve or even know the intricacies of war plans. On this issue, the country’s primary hope for an effective system of checks and balances on the mistakes of executive branch officials was the U.S. armed forces.
The broad argument of this essay is that the tragedy of Iraq—that one of the most brilliant invasion successes in modern military history was followed almost immediately by one of the most incompetently planned occupations—holds a critical lesson for civil-military relations in the United States. The country’s Constitution makes the president commander in chief and requires military leaders to follow his orders. It does not, however, require them to remain mute when poor plans are being prepared. Nor does it require them to remain in uniform when they are asked to undertake actions they know to be unwise or ill-planned.
This argument is not intended to suggest that military leaders are always right on matters of war simply because they are professionals in that arena. Often they are wrong. Eliot Cohen’s book Supreme Command (Free Press, 2002), dramatizing several periods in history in which civilian leaders have usefully challenged their military establishments not just on military strategy but on operations and tactics, was convincing in its main thesis. Reportedly the book was read before the Iraq war by President Bush and received a good hearing elsewhere in the administration as well.
But if military affairs are too important to be left to the generals, they are also too important for key decisions to be left just to the civilians. In the ongoing debate over the proper roles of uniformed personnel and their constitutionally superior civilian bosses in American national security decisionmaking, it is probably now time for a correction in favor of an enhanced role for the military voice.
It is of critical importance to the United States that civilians and military personnel share responsibilities. They must not pretend that their jobs can be neatly separated into two broad and distinct bins—high strategy, the primary province of civilians, and military operations, where the uniformed services possess the nation’s principal expertise. There are usually no clear red lines separating strategy from operations. Clausewitz depicted war as a continuation of policy by other means; Sun Tze wrote that the greatest form of military victory was the one that required the least battlefield action. What these observations have in common, from these two great yet very different military theorists, is a recognition that broad strategy and military operations are inherently intertwined.
So civilians and military personnel must of necessity encroach on each other’s policymaking territory. The question of how wars are conducted affects decisions on whether to fight them, meaning that civilians must concern themselves with the technical subjects in which the armed forces specialize. Likewise, the political goals of the nation’s conflicts—and the political assumptions on which plans for them are shaped—fundamentally affect the tactics and operational plans available to the military to prosecute them, meaning that military planners and commanders must also think about and understand strategy.
Missing: “Phase IV” in Iraq
Unfortunately, in the Iraq operation, the U.S. defense planning system did not work. Indeed, it failed badly in planning for the aftermath of Saddam’s fall from power. The first three phrases of the operation, including the buildup, initial preparatory actions (largely by covert teams), and the main air-ground thrust, were impressive. But what is now commonly called Phase IV was handled so badly that its downsides have now largely outweighed the virtues of the earlier parts of the operation. In other words, while it has achieved a worthy goal in the removal of Saddam, on balance the U.S. operation in Iraq has probably become a sub-par performance of the U.S. armed forces.
It would be unfair to blame all the troubles in post-Saddam Iraq on the lack of a proper stabilization plan. Given the history of Western colonialism in the region and other factors, it was bound to be difficult for any coalition of Western countries to invade, occupy, and help rebuild Iraq. Moreover, many of the mistakes made during the occupation—dissolving and then not attempting to reconstitute the Iraqi army, taking de-Baathification to an extreme, keeping the occupation in the hands of Washington and reconstruction contracts in the hands of supporters of the war, frequently changing plans for elections and the transfer of sovereignty—cannot be fairly blamed on the military plan. Rumsfeld may share some responsibility for these mistakes as well, but war planners generally do not.
Nor was the problem excessive civilian intervention in the war plan per se. Indeed, Rumsfeld was on solid ground in demanding a fundamental reassessment of a plan for invading Iraq that, when last formally approved in the mid-1990s, would have required half a million troops and more than half a year of preparation.1 Rumsfeld also allowed himself to be talked out of some of the bad ideas he or others around him may initially have held about how to win such a war. Notably, the initial hopes among some civilians that a war plan could be executed with only a few tens of thousands of American troops were ultimately dashed by a responsible military planning process. By this standard at least, the invasion force that was ultimately employed was in fact not small.2 Rumsfeld and, to an even greater degree, Director of Central Intelligence George Tenet had usefully involved themselves in war planning in Afghanistan as well.3
The problem was simply this: The war plan was seriously flawed and incomplete. Invading another country with the intention of destroying its existing government yet without a serious strategy for providing security thereafter defies logic and falls short of proper professional military standards of competence. It was in fact unconscionable.
Lest there be any doubt about the absence of a plan, one need only consult the Third Infantry Division’s after-action report, which reads: “Higher headquarters did not provide the Third Infantry Division (Mechanized) with a plan for Phase IV. As a result, Third Infantry Division transitioned into Phase IV in the absence of guidance.” A broader Department of Defense report on the war similarly observed that “late formation of Department of Defense [Phase IV] organizations limited time available for the development of detailed plans and pre-deployment coordination.”4
Why did the Bush administration fail to recognize that overthrowing Saddam could shatter Iraq’s security institutions and thus leave responsibility for maintaining civil order in the hands of the American-led coalition? It is not hard to hypothesize. The explanation surely includes the administration’s desire to portray the Iraq war as a relatively easy undertaking in order to assure domestic and international support, the administration’s disdain for nation-building, and the Pentagon leadership’s unrealistic hope that Ahmed Chalabi and the rather small and weak Iraqi National Congress might somehow assume control of the country after Saddam fell. But it is harder to understand why the uniformed American military effectively went along with this set of assumptions.
Many people outside the Pentagon did recognize and emphasize the centrality of the post-Saddam security mission. Some were at the State Department, though State’s Future of Iraq Project produced an extremely long and somewhat unfocused set of papers.5 Other analysts were also prescient, and much more cogent, in their emphasis on the need to prepare for peacekeeping and policing tasks. One of the more notable was a study published in February 2003 by the Army War College. It underscored the importance not only of providing security but also of taking full advantage of the first few months of the post-Saddam period when Iraqi goodwill would be at its greatest.6
These think tank studies and reports did not, of course, develop precise estimates of how many troops would be needed to stabilize post-Saddam Iraq or lay out detailed rules of engagement for restoring security. But General Shinseki, the Army Chief of Staff, provided Congress some clear advice on the former point when he estimated that “several hundred thousand” troops might be needed for the overall operation. And formal planning mechanisms were available for such purposes, notably the large planning staff at Central Command in Tampa, Florida, if only they had been properly instructed to develop detailed plans.
Previous CENTCOM plans for overthrowing Saddam had indeed given attention to this issue. They were not perfect. Throughout much of the 1990s they relied on a brute-force approach to seizing Iraq that had its own downside, such as the likelihood that surprise would be totally sacrificed during a lengthy buildup period. And a major revision of the war plan begun late in the decade by General Tony Zinni may not have ever been completed.7 But they laid a groundwork that could have been built upon in the year leading up to the March 2003 invasion. Instead, they were effectively discarded. According to General Tommy Franks, while planners spent many hours in discussions about Phase iv, and while Franks himself always cautioned that this stage of the operation could take years, it was ultimately assumed that much of the regular Iraqi army would survive and be available to play a large role in keeping postwar order.8
Recent American experience in the Balkans provided modern-day experience with the challenges of post-conflict stabilization missions, making naïvete about their typical nature an unacceptable excuse. To be fair, it should be noted that many aspects of the Balkan peace plans and ensuing stabilization and reconstruction effort could be developed sequentially—after battlefield victories were achieved in the more classic sense. That is because there were still cohesive, if battered, indigenous security forces in place in Bosnia and Kosovo, allowing for something akin to a classic surrender and transfer of responsibility. Power vacuums did not result as they did in Iraq. In Haiti, a robust stabilization plan was lacking as well.9 However, these examples are no excuse for the poor planning for the Iraq operation. For one thing, much had been learned from these experiences that should have been applicable to Iraq. For another, a power vacuum should have been foreseeable in Iraq. Even if the invasion plan was designed to “cut off the head of the snake” and leave much of the rest of Iraqi security institutions unscathed, there was no prudent way to assume that this goal would be neatly achieved. Military force is a blunt instrument of policy; the hope that it can be surgically precise is usually wrong and always dangerous.
Many basic tasks that should have been seen as necessary in Iraq—policing the streets, guarding huge weapons depots, protecting key infrastructure, maintaining public order—were simply not planned for.10 Instead, such planning as there was, conducted largely out of the office of Under Secretary of Defense Douglas Feith, was reportedly unfocused, shallow, and too dependent on optimistic scenarios that saw Ahmed Chalabi (or perhaps some of Saddam’s more moderate generals) taking charge without the need for a strong U.S. role in the stabilization mission. Even as it became apparent that the initial assumptions were wrong, the Pentagon was unresponsive. The initial post-invasion chaos was famously attributed by Donald Rumsfeld to the fact that “freedom’s untidy.” In fact, only the U.S.-led coalition military forces were in a position to stabilize the anarchic conditions.11
Admittedly, many of the critical tasks involved in stabilizing Iraq had more of a State Department flavor to them than a military one—getting reconstruction going quickly, employing unemployed Iraqis, figuring out a proper de-Baathification strategy, determining a process to select new Iraqi leadership. So again, military planners do not deserve principal criticism for the shortcomings here. In his recent book, General Franks is thus on reasonable ground in saying he wished that Secretary Rumsfeld and Secretary Powell had forced their respective departments to work better together, since the Pentagon needed the State Department’s help. But uniformed planners knew the command arrangements and were aware of the relative marginalization of the State Department in the process. They also knew how important it was that someone have responsibility for these types of political and economic tasks. In that sense, they were too willing to be quiet in the hope that somehow the problem would sort itself out.12
How much did it really matter that the coalition got off to a poor start in securing post-Saddam Iraq, allowing chaos to reign in much of the country for weeks before fully responding? In fairness, we cannot know for sure. In a country rife with weapons and still plagued by the continued presence of thousands of Baathists from Saddam’s various elite security forces who had melted into the population rather than fight hard against the invading coalition, violent resistance was likely. Porous borders and foreign fighters exacerbated the situation.
There are, however, several powerful counterarguments to the claim that post-Saddam Iraq was destined to be chaotic. First, porous borders and large unprotected weapons caches were to a large extent preventable. A more complete “Phase IV” operational blueprint would have done much to secure them through better planning and, quite probably, more troops. As the head of the Coalition Provisional Authority, Ambassador Paul Bremer, later argued, “The single most important change . . . would have been having more troops in Iraq at the beginning and throughout.” Bremer claimed to have “raised this issue a number of times with our government” but to have been overruled.13
Second, although violent resistance from hard-core Baathists and jihadists was perhaps inevitable, the willingness of Iraqi “fence-sitters” to take up arms against the coalition out of frustration appears to have increased over time. Indeed, while estimates of the strength of insurgencies are never reliable, it is nonetheless striking that the Iraq resistance was estimated to number only 5,000 hardened fighters in mid-late 2003 but later thought by some U.S. officials to approach 20,000 in size by mid-2004. Wasting those precious first weeks and months gave this third group—the fence-sitters—a perceived rationale to take up violence too.14 It created a dynamic in Iraq in which high levels of street crime combined with the growing insurgency increased the population’s insecurities, which then also impeded economic recovery activities. With the security environment and the economy both stagnant, dissatisfaction grew, and the resistance thus had more potential recruits to draw upon.15
Finally, opinion polls in the occupation’s early months showed a general happiness among Iraqis that Saddam was gone. That translated into a certain goodwill towards occupation forces, or at least a willingness to tolerate their presence as a necessary means of ensuring stability.16 Wasting this moment of Iraqi cooperation was to lose something that could never be recovered thereafter. This was not just a matter of winning a popularity contest. The general population’s willingness to provide intelligence on the resistance, always a key ingredient in any successful counterinsurgency, is also always a function of the perceived risks of doing so. Citizens are more likely to provide information when convinced it will help defeat an insurgency; they are less likely to take such risky steps if they see the tide of battle favoring the rebels. If a major effort had been made to nip the resistance in the bud, that effort could have developed self-perpetuating momentum.
None of these arguments is conclusive, but, especially when taken together, they are highly suggestive: Establishing early momentum would have made a huge difference in the subsequent course of the coalition’s counterinsurgency operation.
Entering into this conflict without a readiness to quickly restore order in post-Saddam Iraq was the military equivalent of going into open heart surgery without extra units of blood on hand just because the head surgeon insisted that he had devised a new procedure that would make such precautions unnecessary. One might say to the surgeon, perhaps your new technique is good and worth employing, but why should we be so confident about an untried method that we throw caution to the wind? Why not prepare for the possibility of perfectly foreseeable complications, even if we hope that the new method will make such precautions unnecessary? The same sort of question should have been posed in regard to the Iraq invasion. Whatever forces were on hand should have been fully prepared for the mission of keeping order once the Baathists were defeated, with proper pre-war training and proper planning that was then conveyed to officers on the ground.
The mistake here was primarily of the Bush administration’s making. Indeed, much of the prevalent view within the uniformed military is that the Rumsfeld/Wolfowitz/Cheney vision of modern warfare, as well as their strong preconceptions about how easy it would be to depose Saddam, deserve the blame for CENTCOM’s lack of readiness to handle the challenges that began to present themselves in Iraq on April 9, 2003 when Saddam’s statue fell in Baghdad. This perspective is mostly right. It is also too simple.
The uniformed military in fact shares some of the blame for the mistakes made in planning the Iraq stabilization mission. That is partly because General Tommy Franks in the end was the author of the plan. Even if he was under pressure from Secretary Rumsfeld to produce a certain concept, he had every opportunity to voice his objections. It is also because the joint chiefs of staff, with the apparent exception of Army Chief of Staff Shinseki, reportedly blessed the plan as well. It is also because no member of the armed forces of the United States went public with his objections or resigned in protest even though the plan was the military equivalent of medical malpractice.
The U.S. military’s overall performance in the Iraq mission has been impressive, courageous, and admirable. Only a very small fraction of military officers had the opportunity to understand and object to the shortcomings in the war plan; only they can be even partially blamed. Unfortunately, their willingness to turn a blind eye to a badly constructed Phase IV plan is costing the country—and most of all their own comrades in arms, who are still fighting and dying in Iraq—very dearly. Whether or not it was an understandable mistake, it is one we must collectively strive to learn from now so that we can avoid similar mistakes with comparably tragic consequences in the future.
Who in the military officer corps should have done a better job? Primary responsibility must lie with General Tommy Franks and those elements of the joint chiefs of staff who blessed the warfighting concept. They gave their professional imprimatur to a military strategy that was innovative and solid for the invasion phase of the war yet negligently incomplete for the aftermath.
Second, those officers who did limited planning for the post-Saddam phase of operations at CENTCOM should have realized that their efforts were not receiving proper emphasis, support, or visibility. In the end, their efforts were not successful. CENTCOM did not have an overall framework for ensuring at least a modicum of security throughout most of the country that was conveyed to forces on the ground in advance. As a result, most division commanders had not worked out concepts of operations for the period after the Baathist regime fell. Their key officers had to resort to improvisation in whatever localities they found themselves; lower-ranking individual unit officers had no idea what was expected of them. Hospitals were looted, major buildings destroyed, shops ransacked, and chaos allowed to reign as American soldiers stood by. Iraq quickly became one of the most violent places on earth—not so much in terms of the ongoing resistance, which while brutal was limited in scale (especially at first), but particularly in the growing prevalence of street crime throughout the country.
What should the military have done differently? For one thing, it could have carried out detailed planning for the stabilization phase of the operation informally at CENTCOM even without explicit civilian permission. To avoid direct confrontation with Rumsfeld, it could have been viewed not as development of a formal plan but as backup analysis.
But that would probably not have been enough here. Leaks to the media would therefore also have been justified under the circumstances. Leaks are used on many matters of defense policy, sometimes for reasons no better than to embarrass political opponents. The uniformed military has itself leaked information about the readiness rates of individual divisions (data that are supposedly classified), about its unmet funding needs for weapon x or weapon y, and about other such issues, so it is hard to believe that the uniformed military is against such unauthorized releases of information as a matter of principle. In this case, the leaks would have been designed to improve the country’s war plans and core security interests and to help save the lives of its troops — certainly worthwhile objectives by comparison with the more frequent usages of leaks.
Finally, some military personnel probably should have resigned. If my charges are right that the willful neglect of planning for the post-invasion phase of the conflict amounted to professional malpractice, it was unprofessional and unconscionable for the uniformed military to stand by while such a plan was finalized and put into effect. This measure would have been extreme for the people resorting to it, requiring enormous individual sacrifice, but it would have been less costly than seeing so many lives lost, at least some of them needlessly and as a result of poor policy, since the invasion. And it would hardly have been unconstitutional or un-American. Military personnel cannot legally and constitutionally disobey orders while in uniform. That does not mean, however, that they must remain in uniform when asked to do things that violate either their ethics or their sense of proper professional conduct.
Peter Feaver and Richard Kohn have made a powerful argument against resignations by military officers, which as they point out can become tantamount to mutiny if done widely at a given moment in time. Feaver and Kohn have an especially convincing case in regard to officers in the field of operations. But when a relatively small group of military personnel, in this case CENTCOM planners for the post-Saddam mission in Iraq or any others with particular insight into preparations for that period, are witness to neglectful and irresponsible preparation, the situation is different. Under extreme circumstances, their response should be too.17
It is also worth remembering that military officers do not serve just the civilians of the executive branch in their advisory capacities. Of course, they do owe loyalty to that branch, and in particular the chain of command going from the secretary of defense to the president, on matters of war and in response to direct orders. But they also owe the Congress and the country their best advice. When General Shinseki offered the Congress his estimate that several hundred thousand troops could be needed to stabilize post-Saddam Iraq, he may or may not have been substantively correct (and Secretary Rumsfeld and Secretary Wolfowitz were within their rights to publicly debate the facts of the matter with Shinseki in the ensuing days). But he was categorically in the legal and political right to offer the Congress his unvarnished counsel.
If there is still any doubt about whether military officers should be shy about entering into national debates on matters of strategy and the use of force, a quick review of two other crises may help solidify the point.
First, consider NATO’s 1999 war against Kosovo. General Wesley Clark, who contributed to the prevalent but incorrect U.S. belief that Slobodan Milosevic would be easy to intimidate through a few days of light bombing, subsequently got it right. Realizing that NATO had to prevail in the conflict and that his own civilian superiors allowed for no other possibility in their public utterances, he therefore had his staff examine options for escalation up to and including a possible ground war. He also sought to discuss the ideas with the White House and the Congress. Livid with Clark’s unsanctioned planning activities and his willingness to share his military advice with others, Secretary of Defense William Cohen fought Clark at every bureaucratic turn and ultimately relieved him of command early. But Clark was right, putting the nation’s need to win its wars ahead of standard decorum. If the Clinton administration did not want to hear about ways to ensure that the war could be reliably won, that was its mistake, and Clark was correct not to accommodate that mistake. After all, he never questioned any command given to him; he simply did his homework to develop backup military strategies to serve the Clinton administration’s own stated and unflinching goal of winning the war.18 To give the uniformed military too much credit here would be a mistake; most of the joint chiefs reportedly fought the idea of escalation (and perhaps of any intervention in the Kosovo crisis in the first place) and may even have leaked their views to newspapers. But all the more reason why Clark was right to voice his opinions.19
In 1993, Chairman of the Joint Chiefs of Staff General Colin Powell objected to the casual talk he was hearing about using military force in unspecified ways for unspecified battlefield objectives in the Bosnian war. The Clinton administration came into office rightly convinced, in my eyes, that the first Bush administration was too callous in its views of the Bosnia conflict. Secretary of State Jim Baker’s statement that “we don’t have a dog in this fight” was morally wrong and strategically mistaken, and any military officer who categorically took this position, resisting proper planning or discussion of any and all military options as a result, was going too far, even to the brink of insubordination—as Admiral Leighton (Snuffy) Smith reportedly did.20
But Powell’s position was different and more defensible. He recognized that the Clinton administration, still searching for its national security sea legs in its first year or two in office, was more intent on “doing something” than on developing a serious plan for intervention that had a good chance of succeeding, and after Vietnam and Lebanon he was opposed to using the American armed forces in such a militarily vague way. Soldiers’ lives and the country’s security interests were at stake.
Powell was right to voice his objections clearly and in a way that became public. He was not straying into strictly civilian territory; he was not disobeying orders; he was not violating proper civil-military relations. Nor was he even opposing the use of force for limited purposes, even though his famous “Powell Doctrine” is often misinterpreted as arguing against such endeavors. Before the Powell Doctrine was famous, Powell underscored that, as Clausewitz argued, all wars are limited and that many limited uses of force are appropriate and successful.21
Rather, Powell was addressing a subject in which military tactical considerations were of critical importance for reaching an initial strategic judgment on whether to intervene or not in the first place. Perhaps there was a reasonable way to intervene early in the Bosnian civil war, but all Powell did, rightly, was to point out that he had never heard a convincing proposal. Had he resisted helping the Clinton administration explore such proposals, he would have been wrong; had he actually disobeyed orders, he would have been insubordinate. But he did neither.
On a final note, and a somewhat different matter, some have criticized the increased foreign policy role of the U.S. military’s regional combatant commanders during the 1990s.22 Indeed, Secretary of Defense Rumsfeld appears to have been one of the critics, as he objected to their previous title of regional “CINC” or commander in ch