Earlier this year, Iraq’s parliament approved a new power-sharing government led by Prime Minister Haidar al-Abadi. The move toward political inclusion was encouraging, especially as Iraqi forces continued to battle the militants of the Islamic State of Iraq and al-Sham (ISIS). For weeks, though, two key cabinet positions—Minister of Defense and Minister of the Interior—remained unfilled. Onlookers held their breath as they waited to see whether Shia, Sunni, and Kurdish politicians would be able to compromise on these all-important security posts.

The appointments, which were finally made in October, might represent a genuine attempt at reconciliation. The choice of Mohammed al-Ghabban, who belongs to the Badr Organization, as the new interior minister could be an adept way to bring hardliners into the fold. So, too, could the selection Khaled al-Obeidi as the new minister of defense be a genuine attempt at inclusivity. Obeidi previously served as an adviser to the governor of Nineveh Province. When Mosul fell to ISIS this summer, he emphasized the need for granting Iraq’s regions greater autonomy and even creating a separate Sunni Arab army.

Alternatively, the appointments might be run-of-the-mill bartering. And with violence showing no signs of abating, Iraq’s leaders will likely face considerable pressure from sectarian extremists. If the government of Iraq proves unable to get beyond sharing power through high-level appointments, it will only fuel the ongoing civil war and could lead to territorial partition along sectarian and ethnic lines.

At this point, partition might sound preferable to persistent sectarian conflict. U.S. policymakers were tempted by the idea at the height of Iraq’s sectarian war in 2006, when Joseph Biden, who was a senator at the time, and Leslie Gelb, the President Emeritus of the Council on Foreign Relations, advanced a plan for the “soft partition” of Iraq. In a 2006 Foreign Affairs roundtable focused on policy options for Iraq, Chaim Kaufmann, a well-known scholar of international relations, argued that only through separating the population would the violence end. This summer’s bloodshed seemed to revive the idea. Writing in the Washington Post, columnist Fareed Zakaria advocated that the United States adapt to the reality of sectarian enclaves. Others, like the Council on Foreign Relations’ Steven A. Cook, hinted that the United States might need to come to terms with a full partition of Iraq, however “bloody and protracted” the process would be.

Events in Syria, meanwhile, have further revived the partition debate. ISIS has kept a firm grip on its Syrian territory in the face of a U.S.-led air campaign, and forces loyal to Syrian President Bashar al-Assad have proved unwilling to back down. Appearing on Zakaria’s CNN show, GPS, in November, Syria expert Joshua Landis promoted a partition plan for Syria. Landis argued that partition would “accept [the] reality” of a Sunni state spanning Syria and Iraq. Partition would be more stable, and, as Zakaria added, would “reflec[t] the realities of sectarianism.”

The usual argument for partition is that, once ethnic or sectarian fighting gets too bloody, nobody can put Humpty Dumpty back together again. War reveals the fault lines in a country’s social terrain, the thinking goes, and redrawing official borders along those lines is the only way out of a perpetual cycle of identity-based bloodletting.

The argument seems intuitive, but it rests on a flawed premise. It treats social identities as givens and ignores the fact that it was politics—not identities in and of themselves—that brought Sunnis and Shias to blows in the wake of the U.S. invasion of Iraq. Far from resolving disputes, partitions can actually activate dormant fault lines. After all, the formation of a new state typically turns once-allied groups against each other, as it did in South Sudan, which has been plagued by civil conflict following its partition from Sudan in 2011. The country now faces insurgent violence from the South Sudan Liberation Movement and the South Sudan Democratic Movement, both of which include fighters who had previously fought against the north in the fight for independence. In addition, South Sudan has also been plagued by inter-ethnic clashes over how power will be divided in the new state.

Indeed, an objective and systematic assessment of the history of civil war-related partitions shows that partitions have usually not promoted peace, and in some cases they have made things worse.

Since 1945, about 180 civil wars (of which more than half were ethnic or sectarian) have produced 14 to 24 partitions, depending on how we count. These cases range from the de facto loss of sovereign control over small enclaves of territory, as in Cyprus in 1964 or Somaliland in 1991, to full-blown declarations of sovereignty with international recognition, as in Eritrea or Croatia in 1991. When comparing those cases to conflicts that did not end with any kind of partition, we found no evidence that partitions prevent the recurrence of war. The data simply do not support the view that partitions are a good solution to ethnic or sectarian civil wars.          

A mild form of partition—the building of separate enclaves—did not seem to settle conflicts, either. In Cyprus, after a brief period of inter-communal conflict in 1963, Turkish Cypriots began concentrating in ethnic enclaves that they could easily defend. They forged a separate administrative structure in these areas and imported arms and fighters from Turkey. This was partition in the making. The enclaves made national reconciliation in Cyprus that much harder and served as footholds that Turkey would later use for an invasion in 1974 that ultimately divided the island.

A similar sequence has played out elsewhere, particularly in more formal cases of partition. In some, the parties started fighting again soon after a territory was divided. That happened in Croatia, where that country’s partition from Yugoslavia in 1991 was followed by another bout of separatist war from 1992 to 1995 as Croatian Serbs tried to secede and unite with Serbia. Bosnia’s secession from Yugoslavia and rapid international recognition in 1992 was followed by the bloodiest war in Europe since 1945. In this case, partition caused the war, rather than the other way around. A second war between Russia and secessionist rebels started in 1999 even though the first war between them had ended in 1996 with de facto recognition of Chechnya. Puntland and Somaliland may have experienced relative stability after they separated from Somalia in the middle of that country’s civil war in the 1990s, but dividing Somalia in this way has not dampened the chaos rampant in the rest of the country.

Other cases of partition had a similar fate, and the few examples where the peace was successfully kept seem too different from the situation in Iraq today. Russia guarantees Abkhazian and Transdniestrian autonomy from Georgia and Moldova, respectively. Bangladesh has remained at peace with Pakistan following a bloody 1971 war of secession. But India stands between the two, making it hard for Pakistan and Bangladesh to go back to war. And even there, the partition sowed the seeds of more conflict: the Baluchis in southwestern Pakistan rose up in a rebellion that simmers to this day soon after Bangladesh won independence.

Does the United States have the appetite to be the external guarantor of peace in a partitioned Iraq? Conflict between the new territories is not the only danger. Internally, the prospect of partition could set off a demographic scramble, with increasing bloodshed as sects and ethnic groups in the border regions sort themselves out. A Kurdish state would not be immune, since the capture of Kirkuk this summer added a sizeable Arab minority to the Kurdish region. If partition works only when there is complete population separation, as some of its advocates claim, then even Kurdistan could see an ethnic cleansing. Once their territory is sufficiently homogenous, Kurdish factions could turn on each other as they compete for power and political control.

The new Shia entity would face violent contests for control of the state as well. The massive response to the Iraqi Shia leader Ayatollah Ali Sistani’s call to arms this summer to prevent ISIS from marching on Baghdad was not a sign of a united Shia political front. Rather, deep fault lines divide Shia political parties and the militias that rushed to join the fight this summer. Parties which were repressed under Saddam Hussein’s dictatorship are often at odds with former exiles, and the thorny issue of alliance with Iran lurks in the background.

Abadi, Iraq’s new prime minister, comes from the Dawa party, the Shia Islamist party that was once under former Prime Minister Nouri al-Maliki. Dawa’s main rivals are the Sadrist Movement and the Islamic Supreme Council of Iraq (ISCI). Although Maliki asserted Dawa’s primacy when his State of Law Coalition handily won parliamentary elections this spring, opposition to Maliki pushed the Sadrists and ISCI into an alliance. In a new Shia state, militias such as the Sadrist Jaysh al-Mahdi and the Badr Brigades—the latter independent from ISCI but often allied with it—would be among the main contenders for power with state security forces loyal to Dawa. Even groups directly linked to Iran such as Asa’ib Ahl al-Haq and Kataeb Hizballah (which have also been deployed in Syria) would become important players in power struggles.

Adding to the potential for conflict escalation, parties opposed to Dawa have in the past successfully forged political alliances with Sunni and Kurdish leaders. Militarized conflict in a new Shia state could easily draw in forces from the newly-formed Sunni and Kurdish states. Pre-existing political ties would provide a precedent for military cooperation. And, other Sunni and Kurdish blocs would have an incentive to support Dawa to hit back against their co-sectarian rivals.

In a new Sunnistan, which would probably include part of Syria, things wouldn’t look much better. Here, the 2006–07 Awakening is instructive. Sunni tribes allied with al Qaeda in Iraq (AQI) extremists to oppose to the U.S. occupation and the Shia-led central government. But, in a recently forged alliance with the United States, the tribes turned on AQI, whose tactics and ideology had proved incompatible with their interests. For now, Sunni interest in taking out a domineering Shia-led government in Baghdad gave tribal fighters and Baathist networks common cause with ISIS. But these factions would find themselves at odds within a Sunni state when power is up for grabs. And ISIS would need to try and eliminate the independent military capability of these other groups in order to advance its radical state-building project, which would ensure escalating conflict. Assassinations of former Baathists during ISIS’ advances over the summer foreshadow future purges.

Iraq’s internal cleavages reflect divisions across the region, so partition could fuel a regional war. A new Shia state could enflame fears of Iranian hegemony, which would invite further attempts at balancing by major powers on the Sunni side. Meanwhile, an Iraqi Sunnistan would likely annex part of Syria and spur its splintering into ethnic and sectarian mini-states with cross-border connections. This could draw neighbors into the fray. Turkey, for example, might try to prevent the creation of a pan-Kurdish state. Sunni, Alawite, and Druze populations in Lebanon could rally to their respective mini-states in Syria as well. In addition to Iran and Israel, Jordan, Kuwait, Turkey, Qatar, Saudi Arabia, the United Arab Emirates, and even the Assad government could each find reasons to back factions within the new states. Furthermore, an ISIS-controlled Sunnistan would provide a platform for the growing Sunni radical movement to translate its ideals into political action throughout the Levant and the Gulf. Fueled by foreign resources, political tensions in a Shiistan, Sunnistan, or Kurdistan could escalate rapidly into a broader conflagration. Control of oil and gas resources could make matters worse on all sides, motivating attempts at territorial conquest in Shiastan, Sunnistan, and Kurdistan.

The way the security positions in Iraq’s cabinet were filled is difficult to interpret. If things really did come down to political horse-trading, some observers will conclude that partition is around the corner. But it would be a mistake to jump to that conclusion—or to try to hurry the process. Partition advocates claim that their position is hard-nosed—it confronts artificial Middle Eastern borders drawn by imperial powers, acknowledges the reality of enclaves on the ground, and deals with the polarizing effects of ethnic war. But the historical record on partitions after civil war does not bode well for Iraq. In a region which is unusual in that the sectarian cleavages within countries also divide states externally, partition is doubly wrong: It will not stop the violence among Iraqis and it will also relocate Iraq’s internal conflict to the international arena.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now