According to Cliff Mumm, who heads up Bechtel's infrastructure projects, the security situation in Iraq has deteriorated to the point where continuing is not possible. Bechtel's decision follows the decision by Kroll Security International to sell or abandon -- it was not clear which from media reports -- its operations in Iraq following the loss of some of its personnel.

When companies like Bechtel and Kroll begin to withdraw from Iraq, the situation has clearly reached a new level of instability. These firms are used to working in unstable environments, and security threats are simply a part of the business they are in. When they have to start calculating that the threat is greater than the potential profit, the situation is indeed serious.

There is a deeper aspect to this. The U.S. Army was designed, during the 1990s, to be a force that was dependent on the private sector to operate. Put another way, the standing Army was not designed to go into combat without integrating Reserve and National Guard components and without outsourcing support services to the private sector. It was not an Army that could undertake combat operations without this support.

During the 1990s, it appeared to some that the world had reached a new level of stability, and that economics had replaced geopolitics. The assumption was that there would not be extended combat. It made sense to depend on the Reserves and the National Guard for additional manpower during short combat situations, and to use contractors to provide many of the services that the military had provided for itself in the past.

The force structure was not designed for multi-year, multi-divisional combat. The Reserve and National Guard components were not expected to sustain the regular force for years. And the contractors did not expect to have to operate in a world of extreme risk.

The combat capability of the U.S. Army is therefore breaking in two ways. First, its manpower base is being exhausted through multiple deployments. Second, it is now going to find that the contracting support it relies on won't be there if the security risk becomes too extreme. Unlike combat support drawn from the ranks of the military, the contractors can't be ordered and expected to carry out their duties in high threat circumstances. But the Army is not built to operate without them.

The decision to outsource key support functions made sense in the 1990s. It shifted the cost of standby capabilities to private companies, and allowed the military to focus on its core mission. In the course of the Iraq war, the challenges have gone beyond feeding the troops to include rebuilding infrastructure, providing security to the firms doing the rebuilding, and so on. The Army could not provide security to engineering companies, so private companies like Blackwater were bought in. As the situation developed, the dependency on these contractors expanded, until the war effort -- understood in the broad sense of nation-building -- became enormously dependent on these contractors.

But they have a different appetite for risk than the military. They are free to leave, and they are leaving. It is unlikely that a decision reached by Bechtel and Kroll is so unique that others won't follow. They will. And that now poses a new problem for the U.S. effort: It does not have the military capability of filling in for the contractors. There are just not the numbers or skills. That means that if the security situation worsens, we will see a spiral in which contractors withdraw, the security situation further deteriorates and more contractors withdraw.

Given the structure of the force that has been fielded, the level of deployment cannot be controlled by the Department of Defense. When you depend on contractors looking to make money, a lot of them will bail when the risks get too great. Defense planners in the 1990s did not count on this scenario, when the enablers of the Army decide to leave the theater of operations. But it seems that that is what is happening.