The study finds that the greatest threat to Iraqi stability and security comes from a possible Kurd-Arab armed conflict over contested areas, which in turn could result in armed intervention by Turkey if steps are not taken to avoid such strife.

The study examines how the United States, having recently pulled back troops from Iraqi cities and towns, might draw down all combat units and transition from combat operations to advising and assisting the Iraqi Security Forces. The study was sponsored by the Office of the Secretary of Defense in response to a congressional mandate initiated by U.S. Sen. Edward M. Kennedy.

The report details three drawdown schedules: removing all combat units from Iraq by May 2010; moving forward with President Obama's goal to withdraw all combat units by August 2010; and maintaining combat units until December 2011, when all U.S. forces are scheduled to leave Iraq.

"To ensure that Iraq has a peaceful national election in January, it is important that a sufficient U.S. force remains in place until the election is completed and the new government seated," said Walt L. Perry, the report's lead author and a senior information scientist at RAND, a nonprofit research organization. "So the number of forces that are withdrawn leading up to the elections must balance the need to provide security and demonstrate to the Iraqi people that the United States is, in fact, withdrawing its forces."

U.S. bases in Iraq will be closed or turned over to the Government of Iraq. It could take up to a year to close the larger bases — including Joint Base Balad and Victory Base Camp — so those bases will likely be among the last to close.

The first alternative described in the study would have all U.S. combat forces departing Iraq by May 2010, leaving approximately 44,000 American troops. The remaining force would include units to train Iraqi Security Forces as well as the associated support personnel. This force would depart no later than December 2011, consistent with the established Security Agreement between the United States and Iraq.

There are three significant risks with this proposal, the study finds: it would affect the potential security of remaining U.S. military and civilian personnel; it would quickly end the ability to pair U.S. combat units with their Iraqi Security Forces counterparts for training; and it diminishes the ability of the residual force to deal with unforeseen contingencies.

The second alternative is RAND's interpretation of President Obama's goal, announced in February 2009, to withdraw all combat units by August 2010. Approximately 12,000 combat forces would depart Iraq by November 2009, with all remaining personnel to stay through the national elections in January 2010. The drawdown would restart in February 2010. Once all combat units had withdrawn by August 2010, some 50,000 U.S. troops consisting of military trainers, support forces, combat brigades restructured as Advise and Assist Brigades, would remain in place. This transition force would be removed no later than the end of December 2011, in line with the Security Agreement.

While U.S. commanders believe the installation of the new government will take three months after the elections, waiting that long to restart the withdrawal under this second alternative could put the August 2010 completion date in jeopardy. This approach would, however, provide the Obama administration a clearer picture of Iraq's security situation by the time combat forces depart, possibly precluding the need to position U.S. forces nearby, the study finds.

The third alternative — maintaining some combat units until December 2011 — would entail a slower troop drawdown during 2010. This alternative leaves some 39,000 military personnel remaining in Iraq in March 2011. At this point, the remaining troops would be withdrawn over the next nine months.

All three drawdown schedules include estimates for the number of non-combat forces — trainers, support troops and enablers — that remain.

"Drawing down these forces is based on a complex combination of factors, some dealing with the need for continued support to the Iraqi forces, as well as the need to secure remaining U.S. forces," Perry said. "It's also understood that commanders will need to make schedule judgments based on the security situation and other conditions that could influence the withdrawal."

The study cites three key dangers that threaten Iraq's internal security and stability during, and after, the drawdown of U.S. forces:

• Extremists who reject the merging political order and would use violence to generate chaos
• Main political and sectarian groups that currently participate in the political process, but have the ability to use force to gain political advantage and control resources
• A politicized Iraqi security force may stage a coup or may be used by the Iraqi government to crush political rivals.

Of these, the study finds that the greatest danger arises from the tension between the Iraqi government and the quasi-independent Kurdish northern portion of the country. The report recommends that U.S. forces depart at a slower rate from the part of Iraq where the Arab and Kurdish populations meet and where the danger of clashes escalating to open conflict are most serious.

The study, "Withdrawing from Iraq: Alternative Schedules, Associated Risks, and Mitigating Strategies," was sponsored and prepared for the Assistant Secretary of Defense for International Security Affairs, Office of the Secretary of Defense.

Other authors of the RAND study are Stuart E. Johnson, Keith W. Crane, David C. Gompert, John Gordon, Robert E. Hunter, Dalia Dassa Kaye, Terrence K. Kelly, Eric Peltz and Howard Shatz.

The work conducted within the International Security and Defense Policy Center of the RAND National Security Research Division, which conducts research and analysis for the Office of the Secretary of Defense, the Joint Staff, the Unified Combatant Commands, the defense agencies, the Department of the Navy, the Marine Corps, the U.S. Coast Guard, the U.S. Intelligence Community, allied foreign governments and foundations.