Blog of the International Journal of Constitutional Law

Independence Referenda Through the Prism of Kurdistan (I-CONnect Column)

Aslı Bâli, UCLA School of Law

[Editor’s note: This is one of our biweekly I-CONnect columns. Columns, while scholarly in accordance with the tone of the blog and about the same length as a normal blog post, are a bit more “op-ed” in nature than standard posts. For more information about our four columnists for 2017, see here.]

An independence referendum in Iraqi Kurdistan was held on September 25, 2017 with disastrous results. Given its significance comparatively little attention has been paid to the course and consequences of the referendum, particularly when contrasted with focus on the Catalan referendum—including a symposium in ICONnect—that occurred the following week. The differences in the conduct of the two referenda—both concerning the fate of a territory seeking independence—are striking. On the one hand, the Kurdish referendum went off with very few hitches on the day of. Campaigning for the referendum occurred over a period of two weeks without any significant incidents and polling day was also peaceful. The Iraqi constitutional court had ordered the suspension of the referendum to examine its constitutionality but the leading Kurdish political figure driving the referendum campaign, the President of Iraqi Kurdistan, Masoud Barzani, went ahead with the vote despite the legal challenge. The result was that 93% of votes cast favored independence. The Iraqi constitutional court declared the vote unconstitutional the same day.

In Catalonia, by contrast, the day of the referendum was marked by violence with nearly 15% of polling places shut down by police raids and over 1000 injured by riot police in the course of the voting. The results of the Catalan referendum—90% voting in favor of independence—were widely seen as evidence that only pro-independence voters turned out and the aftermath witnessed arrest warrants issued for elected leaders in the territory and the dissolution of the Catalan parliament by the Spanish prime minister. Catalan elections to replace the ousted parliamentarians resulted in a renewed majority for “separatist” parties, in a fresh rebuke to Madrid.

The fact that an independence referendum in the heart of Europe would be met by violence while a comparable referendum in Iraq would unfold peacefully is remarkable. But, in fact, a longer view of the aftermath of the referenda shows considerable convergence in their outcomes. Weeks after the vote, Baghdad undertook a violent intervention against the Kurdistan Regional Government (KRG). Like Catalonia, following the referendum, the KRG finds itself with fewer political options and facing far greater coercion from the center. Moreover, the territory within the regional control of Iraqi Kurdistan has been reduced, a result likely to have longer-term adverse implications for the KRG than any outcome in Catalonia.

The reaction to the Kurdish referendum amongst neighboring Kurdish communities was initially celebratory. But the celebrations soon turned to a more sober assessment of the implications of a referendum that Barzani had insisted would be “binding”. Most observers believe that Barzani campaigned for a referendum for two reasons: first, to shore up his own party, the Kurdistan Democratic Party, against opposition parties—the Patriotic Union of Kurdistan and the Gorran movement (or movement for change)—that had been gaining ground and challenging his continued rule. Second, Barzani seemed to believe that the independence referendum would strengthen the KRG’s hand in its negotiations with Baghdad. Barzani’s gamble failed on both fronts—just over a month after the referendum he tendered his resignation as president following the KRG’s loss of the oil-rich and multi-ethnic city of Kirkuk to Iraqi security forces. Iraqi Kurds today are split over how they should deal with the fallout from the referendum, which is now seen by most Kurds as a miscalculation that backfired.

This quick survey reveals a puzzle concerning the independence referenda in Catalonia and Iraqi Kurdistan. Neither had any real prospect of delivering territorial independence. Both proved costly to those who campaigned for the vote, resulting in crackdowns against the restive regions. Given that these outcomes were largely predictable why do regions pursue independence through referenda? The answer in both cases, and likely in most others, is a combination of domestic political calculations and frustration with the international context. Moreover in both regions few deny that at least a simple majority, perhaps even the overwhelming majority, support independence in principle and, at a minimum, would like greater autonomy in practice. Given this tableau, an obvious problem to tackle is identifying the best available mechanisms for pursuing meaningful autonomy that fall short of territorial independence.

Perhaps this inquiry might begin with asking why referenda are not a preferred vehicle. After all, international best practice in making determinations about border demarcations or minority rights nearly always involve referenda among the affected population. In an age of democracy, having a regional majority favor a particular position should, in principle, strengthen the hand of regional representatives. The challenge is that international law largely frowns on secession as a means of realizing the right of self-determination. The principle of uti possidetis provides that changes of borders—including through secession—are only legitimate if they are consensual. When a majority of the country as a whole opposes loss of territory through regional declarations of independence, the absence of requisite mutual consent renders secession unlawful. More recently, the Kosovo decision of the International Court of Justice suggests a possible exception to the rule in cases of a long-standing record of extreme human rights abuses by the center against the population of the region. The chemical weapons attacks against the Kurdish population by the Iraqi government might satisfy such a criterion, but the fact of regime change in Iraq since the Halabja massacre cuts in the opposite direction.

Because independence referenda raise the specter of secession, they almost always trigger a violent backlash from the central government in question. That said, the 2017 independence referendum was not the first time Iraqi Kurdistan voted on the question of independence. In 2005, an informal independence referendum conducted by the Kurdistan Referendum Movement produced an outcome of 98.98% of votes being cast in favor of independence. Yet the 2005 referendum did not have the destabilizing effects of the 2017 vote. Perhaps this was a consequence of international oversight of Iraq at the time, with both the UN and the U.S. overseeing aspects of the 2005 elections. Or perhaps the more significant difference was the characterization of the referendum as “informal” signaling that it would gauge the preferences of the Kurdish electorate without treating the results as a binding commitment or declaration of independence. Indeed, the idea of convening a non-binding vote might be a useful mechanism for seeking greater autonomy within a federal system. On the one hand, such a vote signals the preferences and underlying power of a majority of the regional population. On the other hand, it might avoid tripping the wires that trigger coercive reassertions of control from the center.

In retrospect, this might have been a valuable avenue for Barzani to consider once he began to go down the path of an independence referendum. After all, he himself was at pains to allay the fears of regional governments like Turkey by assuring them that the real purpose of characterizing the referendum as “binding” was to increase pressure on Baghdad rather than actually to declare independence. If the referendum was indeed intended to bolster claims for greater decentralization in a federal Iraq rather than independence, it was poorly communicated. In fact, the threat of an independence referendum prompted the Americans and the British to offer greater support to the KRG’s position in negotiations with Baghdad in exchange for a cancellation of the vote. Seizing that offer might have accomplished Barzani’s goals of strengthening the KRG’s hand through external support while avoiding the current crisis. Moreover, the U.S. and UK also reportedly suggested that should negotiations fail to make progress over two years, they would then lend their support to a referendum. For the Kosovars, a similar commitment followed by years of Serbian intransigence resulted little international opposition when they finally broke off negotiations and declared independence unilaterally. The reality today is that Kosovo’s declaration has secured de facto autonomy and some de jure recognition.

But Barzani insisted on going forward with the referendum. Why was this strategy so problematic? Referenda capture the democratic preferences of those who are canvassed. So long as only one side of a particular question is polled—say, the regional population of a province seeking independence, but not the population of the larger state—a referendum will remain a flawed vehicle for pursuing greater decentralization or devolution of power within an existing territory. And the worry that the referendum will translate into support for secession carries with it all the usual risks of conflict attendant to non-consensual territorial change.

In an earlier blog post, I argued that referenda conducted under undemocratic conditions cannot be treated as a valid expression of the preferences of those polled. There I argued for adopting clear standards to distinguish among referenda that represent valid democratic exercises and those that do not. Here I am more concerned with identifying circumstances in which referenda may be ill-advised regardless of their democratic pedigree. Indeed, there is little reason to doubt that a valid democratic poll that asks Iraqi Kurds to express a simple preference between independence and the status quo will systematically result in an overwhelming and sincere preference for independence. But the Kurdish population of Iraq—like those in Turkey, Iran and Syria—are not able to make this choice with all else held in check. Moreover, these populations have regularly demonstrated their sophisticated appreciation for the regional realities of existing territorial boundaries and the distribution of power.

Because Kurdish independence is treated as a casus belli by each of the countries in which the majority of Kurds reside, the affected Kurdish communities have largely opted to throw their support behind strategies that will give them a greater measure of independence from within existing borders in practice. Indeed, there are few communities in the world where average citizens have a greater appreciation for theories of devolution, decentralization and deconcentration of power than the Kurds of Turkey, Syria and Iraq in particular. Indeed, the eager embrace of democratic confederalism should make the political experiment in Rojava an important case-study for comparative law and comparative politics scholars. But all of this makes Barzani’s referendum gamble that much more irresponsible. The Kurdish and Catalan people clearly prefer far greater autonomy than they currently enjoy under the political arrangements in Iraq and Spain. The legitimacy of that aspiration is beyond doubt. The open question, rather, is what mechanisms are best suited to realizing that goal. The answer may depend on how susceptible a central government is to international pressure and how willing the international community is to exert such pressure. In contexts like Kurdistan or Catalonia, there may well be substantial international support for finding formulae that will enable significant regional autonomy within the existing boundaries of a nominally united, federal state. If that is so, then an important contribution that comparative law/politics scholars might offer is clearly identifying the best mechanisms for meaningful devolution of power in cases like Iraqi Kurdistan and Catalonia. The violent reactions from Baghdad and Madrid to the independence referenda are totally inexcusable. Rather than countenancing the consolidation of control by each capital over the restive regions, the international community would do far better to hold the governments to account and insist that they return to serious talks about decentralizing power. But these recent experiences also suggest that devolution strategies may hold far more promise for these regions than the momentary satisfaction of a referendum premised on the fantasy that all else can be held equal.

Suggested citation: Aslı Bâli, Independence Referenda Through the Prism of Kurdistan, Int’l J. Const. L. Blog, Dec. 27, 2017, at:


One response to “Independence Referenda Through the Prism of Kurdistan (I-CONnect Column)”

  1. Nuhat Muğurtay Avatar
    Nuhat Muğurtay

    This is one of the most comprehensive op-eds that I have read since Kurdistan referendum. Especially, the author described the existing domestic political tensions within Kurdish politics sufficiently and also its relation to the independence referendum. The remarks on comparative law and comparative politics also are substantial.

    But,I am not sure if the argument below is a good.

    “More recently, the Kosovo decision of the International Court of Justice suggests a possible exception to the rule in cases of a long-standing record of extreme human rights abuses by the center against the population of the region. The chemical weapons attacks against the Kurdish population by the Iraqi government might satisfy such a criterion, but the fact of regime change in Iraq since the Halabja massacre cuts in the opposite direction.”

    It is true that Halabja is irrelevant today. But let’s make a though experiment for further implications of Kosovo experience:

    Timeline of Kosovo

    1. There was an armed conflict in Kosovo between 1998 and 1999. (recall Kosovo Liberation Army)
    2. UNSC 1199, as a consequence of 200.000-300.000 displacement of Albanians.
    3. NATO intervention in 1999, because of the danger of ethnic cleansing.
    4. Milosevic withdrew his forces from Kosovo after NATO intervention.
    5. In 2002 – election of Ibrahim Rugova as president by the Kosovan parliament.
    6. 2008, Kosovo declaration of Independence.

    For Iraqi Kurdistan,
    1. There was an armed conflict between Kurds and Saddam during the Iran-Iraq war (1980-1988)
    2. An ethnic cleansing by chemical weapons. (Halabja, 1988)
    3. US-led coalition-intervention in 1990, as a response to Iraqi invasion of Kuwait. (not because of Saddam’s human right abuses against Kurds)
    4. UNSC 688, as a consequence of 200.000-300.000 displacement of Kurds and danger of further ethnic cleansing. (No-fly zone)
    5. In 1991, Saddam Hussein withdrew his forces.
    6. UNSC provided Kurds a kind of autonomy, external patronage and political legitimatization, and later a quasi-state with the re-opening of Kurdish parliament in 1992.

    As you see, the cases somehow were very similar.

    My question: what would happen if Kurdistan Parliament declared an independence during the high level of external patronage, would de-facto autonomy lead to a de-jure recognition?

    This is only a thought experiment that we need to focus on. It is true that Catalonia does not fit into Kosovo case, but what will be the opinion of ICJ if we witness very similar cases when compared to Kosovo today? Or why do many states abstain from recognizing Kosovo as an independent state? Which stance is consistent with the international law? Recognizing or not recognizing? (Regarding Kosovo, or further cases which are very similar to Kosovo?)

Leave a Reply

Your email address will not be published. Required fields are marked *