Yesterday, on 15 April 2014, Ukrainian interim president Turtschinov considered to hold, simultaneously with the presidential elections, a referendum on regional competences in Ukraine. On 8 April 2014, separatists in the Ukrainian region of Donetsk proclaimed that they would hold a referendum on the independence of that Eastern region of Ukraine. Some days before, representatives of the Crimean Tatars announced that they sought to hold a referendum on their political autonomy within Crimea.
On 16 March 2014, the population of Crimea had overwhelmingly voted in favour of joining the Russian Federation. The population was asked to choose between the following alternative: “1) Are you in favour of Crimea joining the Russian Federation as a subject of the Russian Federation?” or “2) Are you in favour of re-establishing the 1992 constitution of the Republic of Crimea and Crimea’s status as a part of Ukraine?” The maintenance of the territorial and status quo was not given as an option in that referendum, and no international observers were admitted. With a voter turnout of 83.1 %, 93 % answered with a “Yes” to the second question, and thus pronounced themselves in favour of joining the Russian Federation.
The spokespersons of the Tatars now declare that their ethnic group had boycotted the referendum of 16 March, and assert that the majority of Tatars would have preferred to stay within Ukraine. Tatars currently form about 10 percent of the Crimean population. Probably hundreds of thousands of Tatars were killed, starved, and were deported from the 1920s to the 1940s under Soviet policy. The new government of Crimea rejects the idea of a politically autonomous territory for the Crimean Tatars but holds that the Tatars can only claim “cultural autonomy”.
The 16 March referendum, and announced further territorial referendums in Ukraine, place in the limelight the problématique of this legal institution. Are not the outcomes of referendums in ethnically mixed units most often ethnically pre-determined? And does not the resort to a referendum lead to ever smaller subgroups which again seek to detach themselves from a larger one? After all, the Ukrainian people, including the Crimean population, had some 20 years ago voted in favour of independence from the Soviet Union. (See on the 1991 referendum in Ukraine Anne Peters, Das Gebietsreferendum im Völkerrecht (Baden-Baden: Nomos 1995), 184-88; specifically on previous Crimean referendums ibid., 190-91, 211-15). That Ukrainian referendum of 1 December 1991 had been at the time widely appreciated as having rung the death knell for the dissolution of the USSR one week later, when the Agreement Establishing the Commonwealth of Independent States of Minsk of 8 December 1991 declared that the Soviet Union had ceased to exist. But even before that date, and later, Crimean politicians had several times (in 1991, 1992, 1994, and so on) planned and sometimes held “polls” on a special status of Crimea.
This post postulates that, as a matter of international customary law, and as a matter of legal consistency and fairness, a free territorial referendum is emerging as a procedural conditio sine qua for any territorial re-apportionment. However, the 16 March referendum was not free and fair, and could not form a basis for the alteration of Crimea’s territorial status.
In the case of Crimea, the territory broke away only to unite itself one split second later with the neighbouring state Russia. Such a transfer could be qualified, in traditional terms of territorial realignment, depending on what viewpoint one takes and what the technical details of the operation were, as a secession (maybe such as the secession of Eritrea from Ethiopia in 1991), but here with a subsequent fusion of two states (such as the fusion of Northern and Southern Yemen to form a United Yemen in 1991); as an integration of one entity into a neighbouring state (just like the German Democratic Republic integrating into the Federal Republic of Germany in 1990); as a cession of territory from Ukraine to Russia (such as the cession of Louisiana to the USA by France in 1803 or the cession of Alaska to the USA by Russia in 1867); as a dereliction of Crimea by Ukraine; or finally as an annexation (such as the annexation of the Baltic states by the Soviet Union in 1990, or of East Timor by Indonesia in 1975).
What happened with Crimea is probably best qualified as a seizure of territory under threat of force, i.e. as an unlawful annexation. Were it to be understood as a secession with an ensuing immediate fusion with Russia, the very strict preconditions for the exceptional ex post-toleration or acceptance of a secession would have had to be met, which was not the case (see on this below).
However we qualify this alteration of territorial status, it cannot be justified by the 16 March Crimean referendum, which did not satisfy international legal standards. Importantly, holding a free and fair referendum is only a necessary, but not a sufficient condition for a territorial realignment to be accepted as lawful by international law.
The operation could therefore not constitute a legal basis for the new territorial status quo. The announced further referendums by subgroups of the Ukrainian population do not promise to fulfill international standards either.
Contemporary international law moves in the direction of requiring that all territorial realignments be democratically justified, and preferably through a direct democratic decision, i.e. by a territorial referendum. The state practice founding this principle started with the plebiscites after World War I, the decolonisation referendums of the 1950s and 1960s, and has been much intensified by the numerous referendums during the dissolution of the Soviet Union and Yugoslavia after 1991. Since then, probably all territorial changes and re-drawing of boundaries were preceded by (and justified by) referendums, or at least by democratic elections in which the territorial issue was the main or only agenda item. Examples for such indirect democratic justifications were the re-unification of Germany in 1990, and the secession of Kosovo from Serbia in 2008. An atypical case with a doubtful democratic basis was the dissolution of Czechoslovakia (CSFR) in 1993.
The international legal obligation to conduct a territorial referendum flows from the principle of self-determination of peoples. That principle has its historic roots in the principle of popular sovereignty and has a democratic component, even if it does not outrightly amount to a right to democratic government. It is generally acknowledged that the right to self-determination should be exercised democratically.
In other words, a free and fair referendum is (potentially) one procedural exigency for the exercise of the right to self-determination. It is my basic claim that a territorial referendum is admissible and has even emerged as the standard procedure to exercise the right to self-determination. Moreover, contemporary international customary law seems to mandate that the collective right to self-determination (notably when it seeks the extreme result of secession) should be exercised through a referendum.
Further procedural requirements for the exercise of this collective right, besides the (direct-) democratic quality of the exercise, are the peacefulness of the process and the exhaustion of negotiations about the territorial issue with all stakeholders, in order to find a consensual solution. And furthermore, a number of material requirements exists in order to render a secession acceptable or tolerable under international law.
If the procedural and material pre-requisites are not met, the principle of territorial integrity and stability − which is presumed to serve best the interests of humans − prevails. In contrast to the ICJ’s view in the Kosovo Advisory Opinion, para. 80, the principle of territorial integrity is best understood to protect states also against disruptions from inside the state, and this understanding also informs state practice. In any case, the principle of territorial integrity was applicable to the status change of Crimea, because the threat to territorial integrity emanated (also) from a neighbouring state, and not only from the inside.
Let us assume that the population of Crimea, despite its multi-ethnic composition, and despite its close ties both to the Ukrainian people on the one hand, and the Russian people on the other, qualifies as a separate “people” in the sense of the international right to self-determination. The collective holder of the right to self-determination need not be ethnically defined. It is sufficient, and in normative terms preferable, to ascribe the collective right to a group of persons who live on a given territory and who are united by their political aspiration to form a political community with its own territorial basis. This is the concept of “people” or “nation” in the “French” tradition, and it is in fact lived by many multi-ethnic and multi-lingual peoples in the world, for example the Swiss, the Nigerians, or the Chinese peoples. The entire process of decolonisation, which was legally based on the principle of colonial self-determination, always took populations (independently of their ethnic composition) as the bearer or subject of the collective right.
However, the intention of the group to form a “people”, which will then constitute the “personal” element of a new state (consisting in a people, a territory, and a government) must be expressed in a “free” way (cf. common Art. 1 of the UN Human Rights Covenants of 1966), and this is where the procedural standards kick in. The self-constitution of the population of Crimea might have occurred over some time, manifest in various political moves, even before the 16 March referendum. But for that (assumed) political actor and potential holder of the collective right to self-determination to lawfully exercise an extreme form of this right (namely to secede and join another state), specific procedural and material conditions must be fulfilled.
Even proponents of a principle of remedial secession (as the extreme outcome of an exercise of self-determination) accept that such an action must in any case remain a means of last resort which may come into play only when other strategies to realise internal self-determination within a given state, without disrupting territorial integrity, have failed. This means that negotiations on this matter must have been seriously tried out and failed. In material terms, resort to this ultima ratio can only be triggered by persistent and massive human rights violations, and by a long-lasting denial of the right to internal self-determination which could be realised by establishing mechanisms of political autonomy within one state. All these conditions were absent in Crimea.
If these material (and overlapping) three procedural conditions (democratic procedure, peacefulness, exhaustion of negotiations on internal political autonomy) are not fulfilled, then the right to self-determination has not been exercised properly and for that reason cannot justify – under international law – a territorial alteration.
The legal basis of the 16 March referendum was a resolution adopted by the Supreme Rada (Council) of the Autonomous Republic of Crimea “On the all-Crimean referendum” on 6 March 2014. That resolution had been passed on the basis of Articles 18.1.7 and 26.2.3 of the Constitution of the Autonomous Republic of Crimea. (Article 18.1.7 provides that among the powers of the Autonomous Republic of Crimea is “calling and holding of republican (local) referendums upon matters coming under the terms of reference of the Autonomous Republic of Crimea”. According to Article 26.2.3 “passing of a resolution upon holding of a republican (local) referendum” belongs to the powers of the Supreme Rada). These provisions are based on Article 138.2 of the Constitution of Ukraine according to which the “organising and conducting local referendums is within the competence of the Autonomous Republic of Crimea”.
The 16 March referendum was not allowed by the Ukrainian Constitution, which enjoys supremacy over the Constitution of Crimea as an Autonomous Republic. The Venice Commission reached this conclusion based on the following analysis (Venice Commission, Opinion no. 762/2014 of 21 March 2014 (Doc. CDL-AD(2014)002), “Whether the decision taken by the Supreme Council of the Autonomous Republic of Crimea in Ukraine to organise a referendum on becoming a constituent territory of the Russian Federation or restoring Crimea’s 1992 constitution is compatible with constitutional principles”). First of all, Ukraine is a unitary state. According to Article 132 of the Constitution of Ukraine, “the territorial structure of Ukraine is based on the principles of unity and indivisibility of the state territory, (…).” Under Article 134 of the Constitution, “the Autonomous Republic of Crimea is an inseparable constituent part of Ukraine and decides on the issues ascribed to its competence within the limits of authority determined by the Constitution of Ukraine”. The Autonomous Republic of Crimea therefore enjoys autonomy only to the extent that powers were transferred to it by the Constitution of Ukraine. Article 135 of the Constitution of Ukraine holds that, “regulatory legal acts of the Verkhovna Rada of the Autonomous Republic of Crimea and decisions of the Council of Ministers of the Autonomous Republic of Crimea shall not contradict the Constitution and laws of Ukraine (…)”. Since Article 134 of the Constitution of Ukraine defines Crimea as an inseparable constituent part of Ukraine, the secession of Crimea would require amending the Constitution of Ukraine. Such a constitutional amendment is, however, prohibited by Article 157.1 of the Constitution of Ukraine which contains a kind of freezing clause. The Venice Commission – in my opinion correctly – concluded that “the Ukrainian Constitution prohibits any local referendum which would alter the territory of Ukraine and that the decision to call a local referendum in Crimea is not covered by the authority devolved to the authorities of the Autonomous Republic of Crimea” (ibid., para. 15).
However, from an international law perspective, the constitutional admissibility or inadmissibility of the referendum is irrelevant. Therefore, any potential international legal value of the Ukrainian 16 March referendum is not tainted by its unconstitutionality. It is actually typical that territorial referendums conducted in the exercise of the right to self-determination are unconstitutional under the law of the mother state. For example, prior to the Lithuanian referendum of 9th February 1991, then president of the Soviet Union, Gorbachev, had declared these referendums illegal and their result void. Nevertheless, the European Community and numerous other international actors welcomed the decision to hold referendums on Baltic independence (i.e. their restoration of statehood).
International law does not only ask for a democratic decision-making process on territorial questions, and to that extent demands a referendum (or elections), but also sets up the rules on the modalities of conducting such referendums. If these are not observed, if a referendum is not free and fair under international standards, it cannot constitute a basis in international law for the sought territorial change.
The procedures and modalities of a referendum are very important, because it depends on them whether the idea of a free and fair territorial referendum is operational in real life. Only an operational rule of international law is credible and can deploy normative force. In fact, during the 20th century, and most of all in the extensive referendum practice after 1989, some international standards, rules and principles on how a territorial referendum must be conducted have emerged or are in the course of formation. Additionally, post-1989 international law, notably in Europe, has shaped standards on other kinds of referendums (not specific territorial ones), as a form of exercising direct democracy. (See Council of Europe, Parliamentary Assembly Recommendation 1704 (2005), “Referendums: towards good practices in Europe”) of 29th April 2005; Venice Commission, Code of Good Practice on Referendums (CDL-AD(2007)008rev) of 20th January 2009 which contains “Guidelines on the Holding of Referendums”).
Not all of these modalities of a referendum derive from binding international standards. Some (only) constitute “best practices” (Venice Commission, Opinion on the Compatibility of the Existing Legislation in Montenegro concerning the Organisation of Referendums with Applicable International Standards (CDL-AD(2005)041) of 19 December 2005, para. 11). Although existing international standards on territorial referendums are “open-textured” (ibid., para. 64), based on the varied practice of many countries, and leaving leeway to their judgment and traditions, some core principles can be said to form part and parcel of international customary law.
Importantly, the legal status of requirements on the modalities of territorial referendums are independent of the question whether a customary law requirement to hold a referendum (or to legitimise a territorial change in an indirectly democratic fashion) already exists (as I assume) as a matter of hard international law. Should this not be the case, a conditional scheme applies: even if there were no international law obligation to organise a referendum, international law could still regulate its modalities in a compulsory fashion. The “if … then”-scheme is well established in international law. For example, there is no right of option (i.e. the right to choose one’s nationality in the event of a realignment of territory) under contemporary customary law. However, if an option is granted either by treaty or by domestic law, time limits for the exercise of this right must be reasonable. The same “if … then”-scheme applies with regard to the admission of the entry of aliens to the territory of a state, and in the area of social rights. To conclude: If a state decides to hold a referendum, then it must satisfy international standards. And when these standards are not respected, a territorial referendum cannot serve as a legal basis for a territorial change.
The most important and arguably hard international legal standards concern, first, peacefulness; second, universal, equal, free and secret suffrage; third, the framework conditions of freedom of media and neutrality of the authorities; and finally an international referendum observation. Neither of these four sets of international legal standards has been respected in the Crimean 16 March referendum.
(1) First, the territory was not pacified. This is the classic requirement for unimpeded voting on territorial issues. In the historical plebiscites on cession, which developed with regard to the re-drawing of boundaries among neighbouring states in the aftermath of wars, the pacification was realised by the “neutralisation” of the territory through the withdrawal or reduction of the troops of both concerned states. Pacification may also require the imposition of a ceasefire.
This basic rule of pacification and neutralisation is of paramount importance for the assessment of the 16 March referendum in Crimea. It was held in front of the guns and tanks of the Russian army and of unidentified troops. For this reason alone, the referendum cannot deploy a legal value under international law. It cannot be said that, against the background of the history of Crimea, even in the absence of Russian and unknown troops, a majority of the voters would in any case have voted in favour of joining Russia, even if the exact rate of approval might have been less than 98 percent. Such an argument is unacceptable, because the prohibition of conducting a referendum in a non-pacified territory, under threat of force, is exactly a procedural and formal device to forestall speculations about an ostensible real will of the concerned population. What happened in Crimea corresponds to what the academic authority on territorial plebiscites, political scientist Sarah Wambaugh, wrote: “a plebiscite not effectively neutralized is a crime against the inhabitants of the area”. (Sarah Wambaugh, Plebiscites Since the World War: Vol. I, Washington 1933, 4429).
(2) Universal, equal, free and secret suffrage: these are the internationally recognised fundamental principles of electoral law as expressed in article 25 CCPR and article 3 of the First Protocol to the ECHR. The democratic component of the right to self-determination requires that these principles be observed in exercising the right to self-determination. In order to guarantee the universal and unfalsified vote, all voters must be orderly registered. Concerning the element of “universal” suffrage, difficult questions arise with regard to the delimitation of the voters. For example, it can be said that, in principle, transferred, dispelled and displaced former inhabitants of the territory should be in principle allowed to vote, too.
(3) Framework conditions: in order to realise the basic principles of democratic suffrage, a number of typical practical measures need to be taken. In order to allow for a free vote (which includes the freedom of voters to form an opinion and their freedom to express their wishes), freedom of expression and of the press, free campaigning, including freedom of assembly, freedom of association for political purposes, and free movement must be guaranteed. The administrative authorities must espouse a neutral attitude, in particular with regard to the referendum campaign, coverage by the media, public funding, and the right to demonstrate.
In its Opinion on the territorial referendum in Montenegro which justified the separation from the Union of Serbia and Montenegro, the Venice Commission summarised these framework conditions as follows: “the authorities must provide objective information; the public media have to be neutral, in particular in news coverage; the authorities must not influence the outcome of the vote by excessive, one-sided campaigning; the use of public funds by the authorities for campaigning purposes must be restricted.“ (Venice Commission, Opinion on the Compatibility of the Existing Legislation in Montenegro concerning the Organisation of Referendums with Applicable International Standards (CDL-AD(2005)041) of 19 December 2005, para. 12).
(4) Notably the lack of international observation: the adamant international legal precondition for a valid territorial referendum is robust international oversight, ideally encompassing a transfer of authority over all matters connected with the referendum to an international institution. At the very least, international observers and facilitating personal must be deployed. The basic rule of international oversight already formed in the context of the plebiscites organised by the League of Nations and the United Nations. These referendums had been prepared and organised, or were observed by international institutions.
The Venice Commission’s Guidelines on the Holding of Referendums (CDL-AD(2007)008rev) of 20 January 2009) spell out the requirement of international observation as following: “b. Observation must not be confined to election day itself, but must include the referendum campaign and, where appropriate, the voter registration period and the signature collection period. It must make it possible to determine whether irregularities occurred before, during or after the vote. It must always be possible during vote counting. c. Observers should be able to go everywhere where operations connected with the referendum are taking place (for example, vote counting and verification). The places where observers are not entitled to be present should be clearly specified by law, with the reasons for their being banned.” (point 3.2. of the Guidelines., p. 11).
The rationale of international observation is obvious: even if representatives of international organisations are merely passive observers, their reports will decide about the value which is ascribed to the referendum by the international community. Their presence is a guarantee both for the organisers and for the voters that the international standards on procedures, organisation, and side conditions of territorial referendums, are complied with.
Historical experience shows that international observation of territorial referendums is feasible. Organisations involved in the organisation and/or observation of territorial referendums have so far been notably the United Nations, the European Union, the Organisation of African Unity, and the CSCE/OSCE.
In State practice, only those territorial referendums which were conducted under international observation have been subsequently recognised by other states. A counterexample is the referendum in the Bosnian Krajina of the Serbian population of Bosnia-Herzegovina of 10 November 1991, which was not internationally monitored. The Arbitral Commission established by the European Community was “of the opinion that the will of the people of Bosnia-Herzegovina to establish a SRH [a Serbian Republic of Bosnia-Herzegovina] as a sovereign and independent state cannot be held to have been fully established” (Opinion No. 4 of the Badinter Commission, paras 3-4, repr. in ILM 31 (1992), 1488 et seq., at 1503). In the case of Ukraine, it OSCE referendum observers were not granted entry, and therefore a fundamental condition for the international legal relevance of the territorial referendum was lacking.
The Venice Commission had issued an opinion before the 16 March referendum was held, and therein qualified that referendum as problematic from the perspective of “European constitutional principles” (in the sense of a European Common law). The Venice Commission also stated (before the referendum) that “a number of circumstances make it questionable whether the referendum of 16 March 2014 could be held in compliance with international standards.” Venice Commission, Opinion no. 762/2014 of 21 March 2014 (Doc. CDL-AD(2014)002), paras 21-22). These circumstances were the following:
(1) Lack of legal clarity: The legal rules according to which the referendum was carried out were unclear, because Ukraine did not have a law regulating local referendums.
(2) Absence of peacefulness and impediment to a free formation of the voter’s will due to at least implicit threats of the use of military force emanating from the massive public presence.
(3) Concerns with regard to the respect for the freedom of expression in Crimea.
(4) Difficulty for democratic deliberation and opinion forming due to the excessively short period of only 10 days between the decision to call the referendum and the referendum itself.
(5) Lack of neutrality of the Crimean authorities due to the 11 March declaration of Crimean independence by the Supreme Rada of Crimea.
(6) Absence of negotiations about a consensual solution among all stakeholders, especially with participation of all ethnic groups of Crimea (Russian, Ukrainians, Tatars and others).
These observations are pertinent. Overall, because of the disrespect for the existing international rules on territorial referendums, the 16 March referendum in Crimea could not justify the breakaway of Crimea and its joining with Russia under international law.
The unequivocal international reactions to the Crimean referendum confirm this assessment. The UN General Assembly passed a resolution entitled “Territorial Integrity of Ukraine” (UN GA Res. A/68/L39 of 27 March 2014) which “underscores that the referendum held in the Autonomous Republic of Crimea and the city of Sevastopol on 16 March 2014, having no validity, cannot form the basis for any alteration of the status of the Autonomous Republic of Crimea or of the city of Sevastopol”(para. 5, emphasis added). States participating in the General Assembly plenary debate pronounced themselves explicitly in this sense, too (for example Ecuador). Other actors held that the referendum was illegal without specifying whether under Ukrainian constitutional law, under international law, or both (Georgia, Iceland), or that it was illegal under Ukrainian constitutional law (the EU and the Venice Commission); or that the referendum was in violation of international law (Moldova and Turkey). Only one state in the General Assembly debate opined that the referendum was legal, and this was North Korea.
One day before the referendum, the Security Council Draft Resolution (UN Doc. 189/ 2014 of 15 March 2014) was tabled by 42 states. The text was “noting with concern the intention to hold a territorial referendum on the status of Crimea on 16 March 2014” (preamble), and “declares that this referendum cannot have any validity, and cannot form the basis of any alteration of the status of Crimea, and calls upon all States, international organizations and specialized agencies not to recognize any alteration of the status of Crimea on the basis of this referendum (…) ” (para. 5). That resolution was vetoed by Russia, with China abstaining.
These condemnations stand in stark contrast to the international reactions to the previous Ukrainian referendum of 1 December 1991. That vote had been explicitly and officially welcomed, inter alia by the then EC and its member states and by the United States. Comparing these reactions reveals that territorial referendums are deemed to be a crucial factor for legalising territorial alterations, but only if they are conducted properly.
The Crimean referendum of 16 March 2014 could not justify the Crimean secession and the territory’s integration into Russia. Neither the procedural nor the material conditions for secession of Crimea (and the immediately ensuing union with Russia) have been met in this spring.
The modalities and side conditions of that referendum were not in conformity with the European and international standards on that matter. In addition, the substantive conditions for a remedial secession have not been met either. As it looks now, the announced further referendums by the Tatars and by activists in the Donetsk region, will – should they take place – not satisfy these conditions either. Overall, it rather looks as if the 16 March vote and planned future referendums abuse the institution of the territorial referendum.
The alteration of the territorial status of Crimea remains illegal under international law, and third states are (under Art. 41 ILC Articles on state responsibility) obliged not to recognise it.
Nicolas Boeglin, Profesor de Derecho Internacional Público, Facultad de Derecho, Universidad de Costa Rica (UCR).…
Journal of Conflict Resolution Volume 69 Issue 1, January 2025 ISSN: 0022-0027 Online ISSN: 1552-8766…
Nicolas Boeglin, Professeur de droit international public, Faculté de droit, Université du Costa Rica (UCR).…
Gaza / Israel: a propósito de la solicitud de intervención de Chile en la demanda…
Nicolas Boeglin, Profesor de Derecho Internacional Público, Facultad de Derecho, Universidad de Costa Rica (UCR).…
El Impacto de los Estudios de Derecho Internacional Público en la Era Digital El Derecho…