Partitioning Iraq: Make a Detailed Case, or Cease and Desist

May 16, 2016

For special access to experts and other members of the national security community, check out the new War on the Rocks membership.

Iraq is once again in political turmoil, and once again we are hearing calls to partition the country into three ethno-sectarian cantonments: Shi’a, Sunni, and Kurd. The partition trope resurfaces periodically, most often while Iraq looks “too hard to fix.” Advocates of partition suggest that Iraq is a false construct of the century-old Sykes-Picot treaty, and that Iraqis are incapable of sustaining a heterogeneous state. Putting aside the fact that the Sykes-Picot narrative is at best contested, it is time to put the partition trope to the test and then, hopefully, to rest. The mostly non-Iraqi voices who want to divide the country into thirds owe the Iraqi people and the rest of the world extensive, detailed clarification. Surely, any plan to drastically restructure Iraq must be more thoughtful and detailed than the widely condemned 2003 plan to invade Iraq. At the very least, advocates for partition should address some fundamental questions. If they cannot answer these satisfactorily then they should pause before reissuing what many Iraqis view as disheartening, and even inflammatory, positions about their state.

First, who wants to break the state into three parts, either under “loose federalism” or as separate states? There appears to be no evidence that the current Sunni revolt seeks sectarian partition. Other than the outlying Islamic State terrorists, Sunni Arab Iraqis want to be part of and, in some cases to control, the state. Most Sunni Arabs I have spoken with are terrified by the idea of partition. It does not appear that leaders from Iraq’s powerful Da’wa party, or even Muqtada al-Sadr, seek partition. While the two major Kurdish parties—the PUK and the KDP—do seek eventual partition or confederation for themselves, and while the head of the PUK has suggested three way partition, neither party has pushed hard for this solution and neither party can claim to represent Iraqi Arab interests. Arguments for partition cannot be predicated on the idea that this is what the Iraqis want. If Iraqis do eventually seek three-way partition, then there is no need to advocate the position, as they will get there of their own accord.

If neither the Iraqi Arab polity nor Iraq’s most powerful political factions seek three-way partition, then the case should be closed. If external powers still decide to encourage or force partition on the Iraqis then these options must also be explained in detail. One way to force the issue is to simply allow Iraq to collapse. This might involve the full withdrawal of American aid and combat power, and ostensibly Iranian support as well. Iraqis would violently settle into a so-called “natural state” of ethno-sectarian homogeneity. Perhaps the Americans and the Iranians could collude to make this happen. But a more likely scenario is that American withdrawal would lead to greater Iranian influence, just as it did between 2011 and 2014. There is little evidence that Iran seeks three-way partition in Iraq. Therefore, it is not clear that a U.S. withdrawal would lead to anything more than greater Iranian hegemony, more destabilization, and a more powerful Islamic State. Alternatively, the United States or another external power could order Iraqi political leaders to divide the state, but why would they obey that order? Exactly how would partition be triggered?

If it could be triggered, how would a tri-partite state be ruled, and why would it be viable? Under current conditions, Baghdad’s relationship with Sunni and Kurdish areas is poor at best. Under partition, the Shi’a-led government would be under less pressure to work with or support the Sunni Arabs. Incentives for cross-sectarian collaboration would be less under partition than within a unified government. Therefore, it would be entirely unreasonable to assume that the relationship between the Shi’a and Sunni Arab Iraqis would improve under a devolved system of federation or confederation. In all likelihood, devolution would lead to mass ethnic cleansing and even more fighting as all three sides sought to consolidate power. The bloodshed in Baghdad and along the disputed territories line — running loosely from the southeast of Kirkuk to the west of Mosul — could be horrendous, possibly far worse than anything happening now. There is no reason to believe that the results would be better than any of Iraq’s prospective futures under a single state. But post-partition requires even more analysis.

Which historical analogies recommend this approach? A strong argument would require some consistent precedent. A positive exemplary case would be one in which a contested heterogeneous state was divided into ethnic or sectarian parts that were secure, economically viable, and politically stable—and that were not at war with their former co-nationalists. “Positive” cases would also have to demonstrate that the new homogeneous states (federal or national) could stand on their own or with minimal central support. This would allow the full U.S. withdrawal sought after by many pundits critical of ongoing involvement in Iraq. Further, a strong argument would have to be made that the negative cases, in which such experiments failed, did not apply to Iraq. Would Kosovo—where post-war segregation remains challenging and where NATO and the European Union retain a long term presence—be a positive or negative case? What about Sudan and South Sudan, where partition led to two years of brutal war that killed over 1.5 million people, displaced over 2 million people, and left behind a dictatorship and a decimated state? How about Rwanda, where even the ethnically-driven slaughter of approximately 800,000 people could not prevent a return to tense but stable heterogeneity down to the village level? How would these cases, or more “positive” ones, relate to Iraq? Paul R. Williams and Matthew T. Simpson argue that there is no historical trend to suggest that ethno-sectarian segregation is viable, or that such a solution would be viable in Iraq. Their arguments and others that question the efficacy of separation should be addressed by anyone who still thinks partition is a good idea.

Even if the historical case can be made, the devil may remain in the idiosyncratic details of Iraqi politics. If the Shi’a controlled much of Iraq’s oil and the capital of Baghdad (they would, just as they do now) and they viewed the world through a purely sectarian lens—a necessary condition for ethno-sectarian partition—then why would they share anything with a weak, impoverished, and volatile Sunnistan? If past performance is prologue, they would not. As a result, the Sunni would keep fighting and supporting groups like the Islamic State; they have told us as much. If the Shi’a state retained control of the increasingly powerful Iraqi Army, why would they allow the Kurds to retain control of Kirkuk or of any of the oil fields in the disputed territories? They almost certainly would not, and once the Shi’a re-cleansed the Army of residual Sunni and Kurds, war would most likely erupt. Meanwhile, there is a fair chance that Basra would declare semi-independence and try to take oil produced in the region away from the central state, throwing the entire system into further turmoil. More sanguine prospects might abound, but these require detailed and evidence-based defense.

Why is it time to give up on the idea of an Iraqi state, and why do Westerners get to foist weak ethno-sectarian federalism or dissolution upon them? Iraq’s difficult experiment with democracy — thrust upon them by the United States and its allies — has been running for about 13 years. Iraqi democracy is in its infancy. While the violence there is terrible, why is now the time to give up and try something new? Why not three years, five years, or perhaps 10 years from now? A detailed explanation is needed of the risks and rewards associated with partitioning now and giving the Iraqis more time to work out their own future. More importantly, a strong case is needed as to why foreigners get to tell the Iraqis how they should govern their state. Most importantly, if it is assumed that meddling with Iraq’s borders and government in the early 1900s and in 2003 was disastrous, why is external meddling good now? The partition argument needs better logic.

The Iraqis that I have spoken with during my research for RAND and with the Atlantic Council’s Task Force on the Future of Iraq do not talk about three-way partition. They do not view Iraq through a stark, primordialist lens like some external pundits. Instead, they talk about provincial federalism. Provincial, or geographic, federalism with divisions at a smaller and more manageable scale, might allow for improved governance, security, and service delivery. It might help move Iraq beyond ethno-sectarian division, at least to some extent. Iraq will always be a very difficult state to manage, but it can be managed in ways that do not risk mass slaughter and continuous war.

No answer to Iraq’s problems can be perfect, but the Iraqis deserve a clear explanation of any proposal that will so deeply affect their future. In 2007 Edward P. Joseph and Michael O’Hanlon made a structured argument for ethno-sectarian partition. This appears to be the most comprehensive effort to describe the pathway to tripartite segregation of Iraq. Their brief 2014 update needs a 2016 update with much more detail. At the very least, they have to make a better argument as to why partition will lead to less war, and why Shi’astan would indefinitely support an economically nonviable Sunnistan, as no Sunni or Shi’a I have met believes this would happen. They should refute strong arguments by experts on Iraq, including Reidar Visser, who, in his strong condemnation of ethno-sectarian federalism, wrote, “Iraqi sects do not have sectarian homelands.” If Joseph and O’Hanlon cannot make a compelling case and clearly answer all of the questions above and refute scholars who argue against ethno-sectarian federalism or full partition, then someone else must.

Otherwise, the sovereign Iraqi state should be supported, and the unsettling arguments for state devolution should be set aside. As Iraq’s well-respected ambassador to the United States, Lukman Faily argued, a unified Iraqi state offers the best hope to defeat the Islamic State. A unified state can in turn help achieve America’s strategic objectives in the Middle East. The alternative prospect of ethno-sectarian federalism seems destined to sustain the Islamic State indefinitely and to undermine U.S. objectives. If that’s not the case, prove me wrong — in detail.


Ben Connable is a senior political scientist at the nonprofit, nonpartisan RAND Corporation, a professor at the Pardee RAND Graduate School, and a retired Marine Corps intelligence and Arabic-speaking Foreign Area officer.

Image: Mobili, CC

We have retired our comments section, but if you want to talk to other members of the natsec community about War on the Rocks articles, the War Hall is the place for you. Check out our membership at!

8 thoughts on “Partitioning Iraq: Make a Detailed Case, or Cease and Desist

  1. The article makes good sense as long as we are only viewing “Iraq” AS “Iraq.” The determinant characteristic there, however, isn’t going to come from any three-way settlement; rather, it will all come down to what the Kurds want.

    If/when they decide to go fully independent, what the other two portions of the “country” want will be irrelevant. Further, since “Greater Kurdistan” extends outside the boundaries of “Iraq,” “Iraqi-Kurdish” independence will likely bring quick uprisings and petitions to join the new state from the “Turkish-Kurds,” the “Syrian-Kurds” and the “Iranian-Kurds.”

    It’s hard to imagine any new regime in Irbil that would/could refuse such petitions.

    It’s not a question of maintaining “Iraq”; that’s already a dead-and-gone polity. The question is: what’s next? And no detailed answer can be given to that until the Kurdish situation, in particular, is ironed out.

  2. The fact that anyone in the U. S. is talking about partition shows what a load of BS Washington wanted us to swallow ten to twelve years ago, when it kept talking about Iraqi sovereignty. Remember how the tiresome office of propaganda kept pushing out images of Iraqis with purple fingers? Everything was going to be good because Iraq would have its own constitution, it’s own elections, its own parliament, its own decision making apparatus. Everyone was going to get along, because that’s what you do in a constitutional democracy.

    The violence never stopped: it just became worse and worse over time, the lull in Anbar after the so-called surge notwithstanding. Talk of partition now, from the same quarters that talked of sovereignty in 2004 – 2006, shows all the previous pronouncements about a sovereign, constitutional democracy were a sham. The people responsible for the invasion thought they could control Iraqi outcomes then, and they still seem to think they can do that now.

    They are just fools. Worse, they do not know it.

    1. Right — it seems to me the situation now is such that anyone calling for the further maintenance of “Iraq” thereby has to explain in detail how that’s supposed to be done, and at what further amazing cost in US/Western blood and treasure — the opposite of the approach taken in the article.

      It seems to me the last realistic decision point when such a policy could’ve been kept in place was at the time of the US pullout in 2011.

  3. To the extent that a central state implies the ability to control and influence a “nation,” there is none in Iraq. Power is divided in so many ways, including intra-Shia. The U.S. has now reentered a war we had once left, again with no real plan on how it will end, only vaguely supporting the concept of an “Iraqi state.” The U.S. imposing a unified Iraq “solution” without details is as empty as the U.S. imposing a tri-state solution without details. Why, it is almost as if the U.S. should stop thinking about imposing anything anymore in Iraq…

  4. We have been and continue to fight a losing battle in Iraq and Afghanistan. This battle, actually a strategic defeat for the United States, began when G.W. Bush invaded Afghanistan but failed to commit those forces necessary to occupy and secure the country until the Afghani could do so themselves. Then, doubling down on one strategic defeat, G.W. invades Iraq and, again, fails to commit those forces/resources necessary to secure and govern Iraq until the Iraqi were capable of doing so themselves. In both instances, the United States did not meet its responsibility to assume control of those countries. But that is ancient history, although we continue to swim upstream in hope(s) of the situation in each country favorably swinging in our direction. It will not. As such, I advocate for allowing these peoples to determine their respective fate and future type governance. The only impediment to this course of action is: how does the US militarily disengage from these countries and retain some semblance of national honor? Perhaps we need a couple more Peace With Honor (i.e., our disengagement from Vietnam), a withdrawal with our national honor somewhat not destroyed then let the chips fall where they might.

  5. More importantly, a strong case is needed as to why foreigners get to tell the Iraqis how they should govern their state

    you write this as though this is not what the US has been doing since 2003

    if its not up to the US, why did they support maliki to be PM when he lost the election in 2010?
    why did they force him out and get him replaced with abadi?

    you write this as though the US is not flying over Iraq at this very moment supporting the domination of the shia sectarian state over sunni lands
    you are telling the sunnis to submit to sectarian cleansing by the shia

    anything short of complete non-interference in Iraq is an biased and unacceptable policy
    US out of Iraq

  6. The fight against the IS is being used to permanently alter iraq’s population. By the time the last town is retaken, expect each to be:
    1. Blown to bits
    2. Devoid of people
    3. Have the surviving population get pushed into concentration camps ran by the Shiite militias

    You have to hand it to the Iranian Mullahs – they’re getting a hell of a return on investment.