The Kurdistan Independence Referendum and Constitutional Self-Determination
Iraq’s Kurdistan Regional Government (KRG) held a non-binding independence referendum on 25 September 2017. Voters were asked: ‘Do you want the Kurdistan Region and the Kurdistani areas outside the region’s administration to become an independent state?’ Voting occurred in Kirkuk and the Kurdish-controlled parts of other territories in northern Iraq whose disputed status is recognized in the Iraqi constitution. In retrospect, Kurdish leaders seem to have overreached politically, as the Iraqi armed forces and allied militias have in recent days seized Kirkuk Governorate from Kurdish control. But was it legal overreach?
Drawing from the Quebec secession case, Martin Scheinin recently argued in the Catalan context that it is always lawful for a group peaceably to ‘come forward with claims of peoplehood, self-determination and secession,’ and to design modalities for internal consultation, including a referendum. Presentation of that claim obligates the existing state to act in good faith, an obligation it violates by declaring its constitution ‘a non-negotiable bar’ to concessions or even negotiations. With respect to İlker Gökhan Şen’s thoughtful analysis of the Kurdistan referendum, which he finds ultra vires under the Iraqi constitution and groundless under international law, I agree with Scheinin on these points.
Thus the question: assuming the referendum was lawful within the present boundaries of the Kurdistan Region, was it also lawful with respect to the disputed territories?
Considering this question may strike of putting the cart before the horse. Anne Peters has argued convincingly that the question of boundaries arises only once independence has been obtained, and as such, uti possidetis is agnostic on the legal grounds of statehood. In that view, there is no clear reason why an expressly non-binding (i.e. consultative) referendum among a putative ‘people’ for self-determination purposes must hew to established administrative boundaries. On the other hand, the democratic legitimacy of self-determination referenda, especially questionable in the unilateral secession context, demands a sensible correlation between the population consulted and the territory at issue. For present purposes, it suffices to assume without deciding that a group’s right to consultation on self-determination should be, if not coextensive with the territory on which its self-determination might be realized, closely related to that territory.
The constitutional status of the Kurdistan Region’s boundaries also matters because Kurdistan’s claim to secession derives indirectly from the Iraqi constitutional order. Marc Weller has observed that since the 1990s, international law has given effect to self-determination claims derived from constitutional orders that gives separate legal personality to a constituent part or parts of the state. This doctrine of constitutional self-determination developed in response to the negative, disenfranchising effect of the classical doctrine of self-determination, which was available only to peoples seeking independence within existing colonial boundaries, but not to peoples, such as the Kurds, who opposed their integration at the point of decolonization and seek to reply its decision. According to Weller, a claim of constitutional self-determination may arise through an express constitutional provision granting the right to self-determination, as with Ethiopia and Liechtenstein, the effective dissolution of a federal state, as occurred in the former Yugoslavia, or may be implied from the recognized ‘nationhood’ of a constitutionally-defined territory, as with Quebec and Scotland.
The Iraqi constitution does not give regions an express right of secession. However, the constitutional order, including the special status and treatment it accords the Kurdistan region, suggests an implied constitutional self-determination status. Previously defined under the British-era constitution as a unitary state of indivisible territory, the current constitution recognizes Iraq as a state of multiple nationalities and brought together in ‘free union.’ Although Kurdish autonomy in Iraq has existed de jure since 1970, the constitution recognizes the Kurdistan Region as it has existed since 1992, including its authorities and its laws, while all other regions were to be established anew in accordance with the constitutional framework.
Thus, the Iraqi constitutional order recognizes the ‘nationhood’ of the Kurdish population, the autonomy of the Kurdistan Region, and its free association in a federal Iraq. A right of constitutional self-determination might be implied on the basis of this asymmetric federalism.
But the Iraqi constitution goes further. It requires the central government to complete the process established in the 2004 transitional administrative law to reverse the Baath regime’s ‘Arabization’ of northern Iraq, i.e. its systematic repression of the Kurdish population. This process specifically includes remedies for forced population transfers, including repatriation of the non-Arab population, and changes to administrative boundaries altered during the Baath era. The constitution stipulates that the process should culminate by the end of 2007 with a census and referendum in Kirkuk and other territories disputed by the Kurdistan region and the central government (the boundaries of which were left to a committee to decide and remain in controversy) ‘to determine the will of their citizens’.
Put another way, the Iraqi constitutional order obligates the federal government to take measures towards restoring the pre-existing demographic character of northern Iraq, and thereafter to facilitate the internal self-determination of the citizens of the disputed territories claimed by the Kurdistan Region. Effectively, the constitutional order recognizes as illegitimate the existing internal boundaries of northern Iraq, including the Kurdistan Region, and directs the federal government to reconstitute them through remedial measures and a referendum, that is, through an act of internal self-determination. New internal boundaries established through an expression of self-determination such as a status referendum are inescapably internal self-determination units.
This constitutionally-mandated remedial program, including the census and referendum, was never implemented. In lieu of a referendum, Staffan de Mistura, while head of the UN mission in Iraq, proposed that part but not all of the Kirkuk district be incorporated into the Kurdistan Region; the Iraqi government rejected the proposal. Although the official Iraqi government position (at least before the independence referendum) supports holding the Kirkuk status referendum once the security situation allows, that is empty rhetoric. With all key foreign stakeholders opposed, including the US, Turkey, Iran and Saudi Arabia, the Kirkuk status referendum likely would never have happened, irrespective of the independence campaign.
If we take constitutional self-determination seriously, the non-implementation of his remedial program must have international legal consequences. Constitutional self-determination gives effect on the international plane to features of a domestic constitutional order that define a ‘people’, its territory and self-governance within the broader state. The international legal order should likewise give effect where the domestic constitutional order disclaims the validity of the existing boundaries and mandates the federal government to draw new ones. As Weller has noted, ‘a self-determination entitlement for an entity may be established in quasi-constitutional law, such as self-determination settlements concluded within states, at times with international involvement.’ The Iraqi constitutional order – reconceived as a federal state comprised of ‘regions’ to accommodate Kurdistan, then and now the one existing region – was precisely such a self-determination settlement, negotiated with international involvement. If constitutional self-determination gives international legal effect to the settlement, surely it also gives effect to noncompliance with that settlement.
Whatever the precise contours of these legal consequences, I would suggest the Iraqi government’s failure to ascertain the will of the people of the disputed territories entitles the Kurdistan regional authorities to undertake such a consultation for itself. Insofar as the domestic constitutional order recognizes that the population of the disputed territories might choose to affiliate with the Kurdistan Region, and the validity of that choice, the constitutional order gave the Kurdistan regional authorities, at minimum, a good faith basis for including that population in its consultation on self-determination. This is true irrespective of whether that consultation posits affiliation with the Kurdistan Region or an independent Kurdistan as the ultimate expression of self-determination; while the latter is politically explosive, legally both are simply consultations on whether to advance a claim of self-determination vis-à-vis the federal government.
The referendum was a consultation involving a constitutionally-recognized ‘people’ and territory to which it has a constitutionally-recognized claim. On that basis, I conclude the referendum lawfully extended to the disputed territories.
This territorial claim was provisionally satisfied through a constitutional framework that recognized the illegitimacy of the territory’s existing boundaries and obligated the federal government to take remedial measures culminating in the claim’s ultimate resolution through new boundaries. If international law derives a basis for self-determination from this constitutional order, legal consequences must also flow from non-implementation of these same constitutional provisions.
I would suggest, for further discussion, that ascertaining these legal consequences requires considering the relationship between constitutional self-determination and remedial secession: does the former regime’s persecution of the Kurdish population, coupled with the current regime’s failure to implement agreed remedial measures, create an entitlement to secession? Might the Kurdish claim to remedial secession have been subsumed into but not extinguished by the self-determination settlement reflected in the constitution? If so, might the federal government’s noncompliance with that settlement reactivate the secession claim, with similarities to the way Serbia’s failure to enshrine irrevocable autonomy for Kosovo in its 2006 consolidated support for Kosovo’s ultimate secession?
Good thread! Thanks. best regards
Christian Oliver Scheich