The most important thing to understand about Iraq today is also the hardest. The country has fallen once more into civil war, a recurrence of the civil war of 2006-2008. In 2007-2008, the United States committed tremendous military and economic resources to pull Iraq out of that first instance of civil war. This time around, Washington has made clear that it will not devote anything like the same resources and there is no other country that can.
All other things being equal, therefore, the most likely future Iraqis face is a new civil war fought largely along the internal ethno-sectarian divisions of the country. Towns and cities may change hands, but the broad frontlines could remain largely unchanged for years. Meanwhile, tens or hundreds of thousands will die in fruitless conflict.
The history of such civil wars demonstrates that there are really only three ways that they end. The first is that one group wins. In the case of Iraq, the Shi’a will have the best chance because of their demographic weight, but history has seen many civil wars in which numbers alone have proven inadequate to produce either a swift or even an eventual victory. However, when one identity group does prevail, it typically does so in a brutal and bloody fashion. Therefore, such an outcome should not be the goal of American policy.
A second possibility is partition. Neither the Sunni Arabs nor the Shi’a Arabs of Iraq seem willing to share power in Iraq. Both might be willing to allow the Kurds to go their own way (although Kirkuk could remain a stumbling block), but they seem equally determined to dominate the Arab lands of Iraq. That suggests a long and costly civil war. Over time, both might come to realize that neither can conquer the other outright, especially if both are receiving considerable outside assistance from neighboring states. In those circumstances, it might become possible to convince them to divorce from each other, separating into two different states as in Sudan and Yugoslavia.
Nevertheless, as those two examples suggest, partition is still likely to be difficult, bloody and a long time in coming. There is a dangerous mythology taking hold in Washington that partition might be easy because Iraq has since been sorted out into neat, easily divided cantonments. That is false. While there are far fewer mixed towns and neighborhoods, they still exist, and even the homogeneous towns and neighborhoods remain heavily intermingled across central Iraq, including in Baghdad. Moreover, both the Sunni and the Shi’a militias are claiming territory largely inhabited by the sects of the other. All of that indicates that it would probably take years of fruitless bloodshed to convince both the Sunni and Shi’a leaderships to agree to partition, let alone on where to divide the country.
The final possibility is the best—or perhaps just the least bad—is what the Obama administration is trying now: Engineering a new Iraqi government that Kurds, Shi’a and moderate Sunnis can all embrace, so that they can then wage a unified military campaign (with American support) against ISIS and the other Sunni militant groups. That needs to remain Washington’s priority until it fails because it is the best outcome for all concerned, including the United States.
Of course, this is also a long shot. It will not be enough just to depose Iraqi Prime Minister Nouri al-Maliki, anoint new leaders, and declare a national unity government. What brought Iraq to this impasse is the fear in Iraq’s periphery—the Sunnis, Kurds, and the Shi’a of the deep south—of the efforts by its center, literal and figurative, to oppress them. The leaders of these communities might agree to join a new government, but they are only going to be willing to make it work if there are far-reaching, structural changes in Iraq’s political, military and economic systems to decentralize power, security and wealth from the center to the periphery.
In other words, they are only going to be willing to participate in a new Iraqi political process if there is real federalism.
A federal solution would preserve the nominal unity of Iraq and some key functions of the central government, but would decentralize the vast majority of powers and responsibilities from Baghdad to the provinces and regions.
For the moment, let’s set aside the Kurds. First, it is worth sketching out what federalism might mean for Arab Iraq. The keys would be both redistributing power from the central government to the provinces and regions, and redistributing power within the central government to prevent the re-emergence of another overly powerful Prime Minister like Maliki.
Shifting power from the center to the periphery needs to start with money. The central government would retain the function of redistributing oil wealth according to a new oil revenue sharing law. Such a law would also apportion a certain percent of the funds to the federal government to perform its (limited) functions, or might simply set parameters that would limit the extent to which the federal parliament could tap Iraq’s oil revenues to pay for the activities of the federal government. However, the exploration, production, and even export of hydrocarbons would belong to the provinces and regions.
A variety of other functions would also need to be divided up, with some migrating to the provinces and regions and others remaining with Baghdad. For instance, the management of Iraq’s power grid, its water resources (and therefore its agricultural policy), and its transportation network would likely remain with the federal government. On the other hand, education policy, housing, labor policy, and other similar functions would probably become the purview of the provinces and regions, and paid for by their own individual funds.
Security would be particularly tricky. Baghdad would have to cede the maintenance of internal law and order to the provinces and regions, which would be allowed to raise local forces to do so. They would also have to receive a share of the federal budget—proportionate to demographics, but wholly under the control of the provinces and regions—to fund their internal security units. However, one solution for external defense would be to assign this responsibility to the central government. In this case, because the Iraqi Army will always retain the latent capacity to subjugate any or all of the provinces, there will have to be a redistribution of power within Baghdad to make it difficult, if not impossible, for an Iraqi Prime Minister to do so. Alternatively, some Iraqis argue that each federal region should maintain its own forces, like the Peshmerga, and that the regions should then coordinate their security forces and activities to ensure external security.
Whichever path was followed, it would be critical to create checks and balances that would constrain the government’s ability to employ Iraq’s security forces to repress any of Iraq’s communities. The most useful but also the most ambitious step would be a constitutional amendment to shift the conduct of foreign affairs and national security (including command of the Iraqi armed forces) from the prime minister to the president. The prime minister would retain control over domestic politics and economic policies, and both would be subject to a limit of two terms in office.
Iraqi leaders also need to agree to implement constitutional provisions (and enact reforms when necessary) to depoliticize the security services and enhance parliamentary oversight. The constitution already includes significant checks and balances that have been circumvented. It requires parliamentary approval of senior appointments, namely the Army chief of staff, his assistants, division commanders and above, and the director of the intelligence service. Laws are still needed to define the powers of the minister of defense (appointed by the president) and minister of interior (appointed by the prime minister), resubordinate Iraq’s various military operations centers under the ministry of defense chain of command (they currently report directly to the prime minister), and specify the treatment of Iraq’s hydrocarbon wealth in the non-Kurdish lands. Other constitutional amendments should redefine the appointment of Iraqi judiciary and election officials and should delegate control of local security forces and appointment of all provincial officials to the provincial and regional governments.
The new government should also reestablish constitutional checks and balances on the premiership by enhancing the legislative powers of the parliament (repealing the court ruling that restricts the power to present bills to the Cabinet and president), restoring the independence of key institutions such as the central bank, election committee, media committee, and establishing the Supreme Federal Court. The future Cabinet should also have clear bylaws to regulate its work and the authorities of ministers.
Even such a heavily revised Iraqi political structure is not going to be good enough for the Kurds. Sunni and Shi’a Arabs disaffected with Maliki’s premiership might see them as a reasonable alternative to protracted civil war, but the Kurds will not see them as a reasonable alternative to what they consider a historic opportunity to achieve independence.
The Kurds already have an independent nation in virtually every way, at least every way noticeable to the average Kurd or Iraqi. They have their own schools and power, their own foreign policy and security services, their own language and culture, their own oil export channels and airports. And the list can go on. Not surprisingly, President Masoud Barzani has announced that the Kurdish Parliament will be asked to establish a Kurdish independent election commission. This commission is expected to organize a referendum asking the residents of the region to decide on sovereignty and independence.
Given the Kurdish leadership’s decision on a referendum, federalism may be an answer for Arab Iraq, but to keep Kurdistan even a nominal part of Iraq will require something more. It will require something closer to confederation.
The difference here is that in a typical federal system, resources and authorities are generated from the center and delegated to the periphery for all but a limited number of constrained functions. However, keeping the Kurds on board will likely require a shift to one in which resources and authority begin in the periphery and then are shared with the center for specific purposes and under specific constraints.
Certainly, on one of the most important aspects of statehood, security, the Kurds already provide for themselves and will not agree to subordinate the Peshmerga (or their intelligence services, the Asayish) to Baghdad’s control. At most, they might allow some Iraqi army units to reinforce the Peshmerga if they were hard-pressed by a foreign invasion. Thus, the security structure in Iraq’s constitution which in theory gives Baghdad’s army the right to go into Kurdish territory is not suitable anymore. The Kurds will most certainly insist on retaining independent control of the Peshmerga, perhaps with an agreement that would specify under what circumstances Baghdad might call on Kurdish military assistance, and when the Kurds might ask for military help from Baghdad.
The Kurds will also insist that the KRG maintain the current lines of control in disputed territories unchanged until a referendum can be conducted in accordance with article 140 of the Iraqi constitution. The two sides would need to hammer out an agreement that would include a fixed timeline for the process—perhaps within 4-6 months of the formation of a new federal government—and ensuring them the right to hold the referendum even if Baghdad fails to meet the deadline.
The Kurds may also seek to exercise control over their own airspace to protect against unilateral actions by Baghdad—such as the recent decision to close Kurdistan’s airspace to cargo flights—that can cause significant damage to their economy.
Oil revenues pose a similar challenge. Earlier this year, the Kurds began exporting oil directly to the Turkish port of Ceyhan, bypassing Iraqi federal control. Then, just before the ISIS offensive in early June, the Kurds began marketing that oil on their own, again bypassing Baghdad’s state oil marketing organization. The Kurds are reluctant to give up what they see as an insurance policy that they can always sell oil themselves and not have to rely on disbursements from Baghdad. Thus, to keep the Kurds in Iraq, Baghdad is likely to have to recognize Erbil’s right to develop and market the oil it produces as the new status quo.
Another difficulty lies in determining how Iraqi oil revenues should be shared. There are many ways to skin this cat. One way would be for the Kurds to simply pump as much oil as they like and pay all of their expenses—including those of the international oil companies producing the oil—on their own from those revenues. Under this scenario, Baghdad would do the same with the oil exports under its control. An alternative approach would be for Baghdad and Erbil to determine an agreeable formula for splitting these revenues. (It is currently set at 17 percent for the Kurds, but may have to increase to reflect the fact that the KRG is now also responsible for Kirkuk and other areas.) The two sides would then account for the net balance of all the sales revenue according to the new revenue sharing agreement. Either way, the Kurdish budget and the Iraqi budget would become separate and independent of each other.
The Kurds will also demand as part of an agreement that Baghdad settle the outstanding balance from the government’s decision to slash or withhold past budget payments to Erbil.
While a one-country-two-system solution (federalism for the Arab provinces and confederation with Kurdistan) is certainly a plausible future for Iraq, it may well be the least plausible. It’s less that federalism-confederalism would be complex (which it would be) and therefore prone to dysfunctionality, and more that so far it seems hard to imagine Iraq’s various warring factions agreeing to it. Maliki is determined to rule a highly centralized Iraq, and is determined to bring the Sunnis to heel. For their part, the Sunni groups either want to do the same to Maliki or else be left entirely alone—which may sound like federalism in theory, but in practice would probably amount to partition given the fear and antagonism they now nurture toward Maliki. The Kurds would have to decide between confederation and outright independence.
Yet here we are. The circumstances of Iraq have worsened so dramatically that we—and the Iraqis and their neighbors—have nothing but difficult choices.
There is no escaping the simple reality that restoring peace to Iraq will require a re-balancing or disengagement of the center from the periphery, either de facto or de jure. In the latter case, partition looms. Yet it would take a long and awful fight and still might not prove feasible. The alternative is federalism, which is both possible and attractive to most, but will be very difficult to pull off. Unfortunately, at this hour in Iraqi history, the question is not how to put humpty dumpty back together again, but merely what can be made of the broken pieces.
Zalmay Khalilzad is a Counselor at CSIS and has served as U.S. Ambassador to Iraq, Afghanistan, and the U.N. Kenneth M. Pollack is a senior fellow at the Brookings Institution and has served as a Persian Gulf Military analyst at the CIA and Director for Persian Gulf Affairs at the National Security Council.