HAS THE AMERICAN WITHDRAWAL from Iraq begun? The Defense Department has announced troop reductions there amounting to 29,000 soldiers almost immediately and has dropped broad hints that another 31,000 will come out by the end of 2006, "conditions permitting."

The conditions in Iraq, however, do not seem to support such reductions. A series of spectacular attacks in recent weeks highlight the continuing vigor of the insurgency. And the disenchantment of the Sunni Arabs with the results of the December 15 elections portend a critical time in the months ahead. The president has repeatedly declared that the withdrawal will not adhere to any "artificial" political timeline. How can such statements be squared with a reduction in the U.S. presence at a time many regard as the tipping point in this war?

There were approximately 160,000 U.S. troops in Iraq in the months leading up to the October referendum and the December elections. This represented an increase from the "normal" baseline of 138,000, intended to secure those momentous votes. The extra soldiers were not just pulling guard duty, however. On the contrary, the coalition used the additional forces to conduct a series of intelligent and aggressive operations along the Upper Euphrates valley and elsewhere in the Sunni Triangle to clear towns and villages of insurgents and establish Iraqi Security Forces in their wake to hold them. Coalition commanders and spokesmen have subsequently claimed that these operations played a critical role in allowing peaceful elections and in reducing the overall level of insurgent violence in the country (at least until recently). They are probably right.

In the wake of the elections, the Department of Defense announced that U.S. forces in Iraq would come down to the level of 138,000--a reduction of 22,000 soldiers. Early this month, it announced a further reduction of 7,000 soldiers. The administration has attempted to minimize the significance of these reductions, claiming that the drop from 160,000 to 138,000 was simply a return to normalcy and that only the ensuing cut of 7,000 troops was really a reduction. But, from the standpoint both of military operations and of perceptions, what matters is the 18 percent cut from the levels of December 15. The further drop to 100,000 mooted for the end of 2006 is not what many would consider a measured withdrawal.

The effectiveness of American forces in Iraq does not result simply from the number of soldiers, of course, but also from what they are doing. Here the news is even more disturbing. Instead of exploiting the successes in the Euphrates Valley and elsewhere, coalition commanders seem to foresee a dramatically reduced role in fighting insurgents and have announced their intention to concentrate the remaining U.S. forces on training Iraqis. Once again, coalition commanders and spokesmen are bombarding the media with the numbers of trained or training Iraqi Security Forces and police recruits. Military news releases since the election have described no large-scale counterinsurgent operations at all.

Those releases have focused, instead, on the numbers of operations led by the Iraqis or conducted jointly with U.S. forces. The growing number of such operations is in one sense positive. There can be no question that the development of a robust Iraqi counterinsurgency capability is essential to success in this war. But the operations the Iraqi Security Forces are carrying out differ dramatically from the clear-and-hold operations carried out by U.S. forces in the months leading up to the election. ISF troops are not, on the whole, capable of planning and conducting such complex operations, and U.S. military releases describe instead "cordon-and-knock" missions that tend to net relatively few suspects.

The result of this shift in military operations is worrisome. From the beginning of the war, the coalition has lacked the number of forces that would be needed to clear and hold the Sunni Triangle, let alone the major population centers in Iraq. It will likely be many months before the ISF is capable of conducting such missions on a significant scale. If U.S. forces withdraw to training areas and cease operations against insurgents except for the odd joint raid or "cordon-and-knock," the insurgents may once again begin to establish safe havens in which to train and operate. The longer safe havens persist, the harder it will be to clear them out--and the longer it will be until the ISF troops are able to undertake the mission.

Backing off now assumes that the insurgency is already broken and that the political process will inexorably reduce the violence to a level the ISF can handle. It is possible that this assumption is valid. Recent evidence that Sunni rejectionists have turned on, and even fought, al Qaeda groups is promising, as is the evidence that some rebels have reached out to the coalition and the government to negotiate an entry into the political process. That these groups seem to have held back from attempting to disrupt the elections is also promising. But these trends are by no means irreversible.

The Sunni Arabs in Iraq have so far tried three approaches to regaining the control that many of them see as their birthright. First they boycotted the elections of January 2005 in an attempt to delegitimize them. Then they attempted to vote down the constitution in the October referendum. Now they have attempted to participate in the political process in an effort to regain some measure of control over the state. The results of the elections disappointed many in this community (who mistakenly believe that their percentage of the Iraqi population is much higher than it actually is). They have responded by denouncing the elections, staging protests, and renewing violence and threats.

It is by no means clear that the negotiations leading to the formation of a new government will satisfy them, or that that new government will make sufficient changes in the Iraqi constitution to address their fears and demands. The Sunni Arabs therefore must decide whether to commit themselves to a political process that is unlikely to give them what they desire. It is vital as they do so that the coalition demonstrate that violence will not improve their bargaining position. The net reduction in U.S. forces accompanied by a fundamental shift in coalition operational patterns does not send this message. It sends the contrary message: There is a window in which violence might be productive. That is why these reductions and operational changes are unwise.

THERE IS IN REALITY a number of U.S. troops that must be present in Iraq, below which the situation will collapse. American forces continue to be essential in sustaining the ISF and police, and in handling the more dangerous and complex missions against insurgent groups and regions. It is not possible to know what that number is at any given moment, because it depends on too many variables for which we cannot begin to find values: attitudes of the Iraqi populations, the insurgents' capabilities and intentions, reactions to random events, and so on. But think of that number as a red line that we will cross at our peril.

It seems clear that the Bush administration intends to keep U.S. force levels in Iraq as close to that red line as it possibly can. The desire to withdraw results not only from domestic pressure but also from the belief, widespread among senior officers and in the administration, that the presence of U.S. forces in Iraq is perhaps the single most important element fueling the insurgency. Rebel attacks on Iraqi police stations, Iraqi citizens, Iraqi ministries, Iraqi gasoline trucks, Iraqi pipelines, and even the relatives of Iraqi leaders call this assumption into question. The comparative effectiveness of joint U.S.-Iraqi operations in the Euphrates Valley also undermines its validity. The fact that increasing American forces in Iraq from 138,000 to 160,000 in preparation for the elections led to a dramatic improvement in the security situation should have ended the notion that the U.S. presence is the most important factor fueling this insurgency.

The problem with attempting to cleave to the red line is not simply the weakness of the assumptions underlying the policy, however, but also the difficulty of pursuing it effectively. It is quite likely that the desire to reduce forces as rapidly as possible will lead to excessive reductions. Violence will then increase; the situation will begin to deteriorate. The administration may respond by increasing troop strength again, pushing it back over the line, in which case the situation will improve and reductions begin anew. The pattern will be repeated.

That, at least, is one possible scenario. But of course it is much easier to pull forces out than to put them back in. Any such reversal of position will look like an admission of error by the administration. It will anger those Iraqis (not a small number) for whom the U.S. presence is a significant issue, much more so than the initial reduction will have pleased them. It will exacerbate inconsistencies in operational approach and discontinuities in intelligence-gathering. It will enable the insurgents to entrench themselves in safe havens. In sum, it will make the task of counterinsurgency much more difficult.

We have already had a taste of what might result from too hasty a transition in the evidence of Shiite officials' torturing Sunni detainees, and the accusations of Shiite commando units' complicity in an array of atrocities. The coalition responded by attempting to reassert its control over both detention centers and commando units, but at the cost of overcoming significant resistance by the commanders of those units and their superiors. It is much easier to relinquish control than to regain it, and much less controversial to continue to oversee operations than to reassert oversight. A pattern of premature transfers of control and repeated reversals generates ill will and resistance--to say nothing of the damage done to the counterinsurgency effort by the atrocities themselves.

In short, it is better to risk having too many troops than too few. It is better to maintain active pressure on the insurgents than to wait until the Iraqis can do so. It is better to remain focused on the goal of this war. America's objective is not to withdraw from Iraq--we could do that tomorrow if we did not care about the consequences. America's objective is to establish a stable government there, which requires defeating the insurgency. Training Iraqi Security Forces is not a proxy for that goal, but one of several necessary preconditions. The temptation to subordinate our strategy to the establishment of that single precondition has always been the greatest threat to victory in Iraq.

Frederick W. Kagan, a military historian and coauthor of While America Sleeps, is a resident scholar at the American Enterprise Institute.

Next Page