Are sovereignty referendums but a tool to legitimize territorial claims of the powerful?

Written by and

This is the impression one could be left with in the wake of the popular votes organized by Russia in the occupied Ukrainian regions of Donetsk, Kherson, Luhansk, and Zaporizhzhia. This impression might be reinforced by the fact that, as recently shown by Sze Hong Lam on these pages, these were by far not the first referendums to be (mis)used to justify the reallocation of sovereignty. Invocations of popular self-determination to legitimize annexations of territories can be traced back to the time of the French Revolution.

The wider picture

Considering that more than 600 referendums on sovereignty have been held across the world, it can, however, hardly come as a surprise that some of them should have been manipulated by those with the power to do so. As with any other instrument, one cannot do justice to the sovereignty referendum by focusing exclusively on its misuses. That voters are forced at gunpoint to cast their ballot is the exception, not the rule.

Instead of Donetsk, Kherson, Luhansk, and Zaporizhzhia (2022) or, for that matter, Crimea (2014), one could point to Eritrea (1993), Montenegro (2006), Scotland (2014), or Bougainville (2019). In fact, one does not need to look further than Ukraine itself to find a rather uncontroversial example of a legitimate referendum: its existence as a sovereign state is rooted in the approval of independence from the Soviet Union in a popular vote held in 1991.

These, and numerous further, referendums allowed the inhabitants of the territories concerned to freely and genuinely express their will regarding their political, economic, social, and cultural destiny. This is, after all, what, according to the International Court of Justice, the application of the right to self-determination requires. Holding a referendum is the most obvious method of consulting people on questions relating to self-determination, as it gives them the opportunity to directly express their opinion on a specific proposed course of action.

The importance of procedure

Thus, a referendum may be a perfectly appropriate – and, we would argue, in many instances the only appropriate – means to decide on the territorial contours of a polity. However, for this to be the case, the conduct of a given referendum must meet a number of requirements that follow, in particular, from the guarantees of universal, equal, and secret suffrage and the freedom of voters to form and express an opinion set forth by Article 25(b) of the International Covenant on Civil and Political Rights.

What these guarantees entail, has been spelled out in more detail by the Venice Commission in its Revised Guidelines on the Holding of Referendums. To highlight only some of the most important requirements under international law: referendums need to be held in a peaceful environment; civil and political rights must be effectively protected; voter qualification must be based on reasonable criteria; the referendum question must be clear; and compliance with these requirements should be monitored by international observers.

A necessary condition for secession

In cases of secession, respect of these procedural requirements has proven to be decisive when it comes to international recognition. An analysis of international practice since 1990 demonstrates that recognition only followed in cases where secession was backed by popular support and, more specifically, where this popular support was expressed in a referendum conducted in accordance with international standards. As a consequence, a rule of customary international law has crystallized: a (successful) referendum on the question of independence that meets the requirements described above has come to be regarded as a necessary condition for achieving statehood.

The international response to the referendums in Russian-occupied territories of Ukraine arguably gives further impetus to such a norm of customary international law. In the case of the referendum held in Crimea in 2014, states and international organizations in unison declined to recognize the result, citing reasons such as military intimidation, a manipulated media environment, implausibility of the results reported, and the lack of credible international observation. The UN General Assembly passed Resolution 68/262, underscoring the invalidity of the referendum and calling on states not to recognize any alteration of Crimea’s status on its basis. Similarly, the popular consultations in Donetsk, Kherson, Luhansk, and Zaporizhzhia were widely decried as farce that could not be considered as the free expression of the will of the people living in these regions. On 12 October 2022 the Assembly condemned Russia’s organization of the ‘illegal so-called referendums’, declared them to have ‘no validity under international law’, and called on states not to recognize any alteration of the status of any of these regions (Resolution ES-11/4). While already the resolution on Crimea was adopted by a large majority of states (100 in favour, 11 against, 58 abstentions), support for this year’s resolution was overwhelming (143 in favour, 5 against, 35 abstentions).

Assuming that most states will abide by these resolutions and refuse to recognize the changes in sovereignty sought by Russia, this can be considered a continuation of previous state practice. States’ reactions to sovereignty referendums contribute to the formation of customary international law standards for their conduct. Seen in this light, the referendums in Russian-occupied regions of Ukraine, rather than giving reason to question the usefulness of sovereignty referendums, reinforce the importance of upholding due process standards in their organization to ensure voters’ freedom to express their will.

Fears of disintegration

Suspicion of sovereignty referendums seems to be mainly prompted by fears that they will lead to the disintegration of the established international order. These fears are unfounded. Giving people a say in the determination of their political future need not result in instability.

First, once actually given a genuine opportunity to express their wishes, voters will often enough reject changes in sovereignty and opt for continuity instead, as shown, for instance, by the referendums held in Quebec (1995), Scotland (2014), and New Caledonia (2018/2020/2021).

Second, a successful referendum is a necessary but not a sufficient condition for reallocations of sovereignty. In addition, these also depend on the recognition by the international community. Considering the recent practice described above, one can be reasonably confident that changes in territory will only be recognized if they were approved in popular votes that were mutually agreed upon by the key actors involved in a conflict.

Third, and importantly, the democratic exercise of the right to self-determination by way of a popular vote may in fact help to defuse or even resolve long-standing conflicts. A prominent case in point are the referendums held in Northern Ireland and Ireland on the Good Friday Agreement of 1998 that were instrumental in putting an end to decades of violent conflict. Further examples that could be referred to in this context are the independence referendums in East Timor (1999), Montenegro (2006), and South Sudan (2011). What may be regarded as anecdotal evidence has recently been bolstered by political science: A study found that self-determination referendums, at least when they are mutually agreed by the relevant minority and majority groups, tend to foster peace.

Giving people a say

Many existing territorial boundaries are the product of arbitrary decisions or, worse, violence. Denying people democratic ways of challenging such boundaries, or of otherwise having a say on questions concerning sovereignty, will not prevent conflict but, on the contrary, is in many cases its very source. That those affected are involved in such decisions is what is required by the rights to self-determination and political participation. As long as it is ensured that voters can freely express their will, holding a referendum is the best way of giving people a say in sovereignty issues.

Print Friendly, PDF & Email

Leave a Comment

Comments for this post are closed


Asher Rottenberg says

December 1, 2022

Thank you for this blogpost that perfectly completes Sze Hong Lam's recent piece.
However, I think the conclusion that there IS a customary rule that secession requires referendum is haphazard. First. it assumes that statehood is legally dependent on recognition which is an old debate in IL.
Secondly, you chose to focus on events from 1990 and onward. In the context of Statehood, given the nature of the subject, and according to the ICJ's rationale in the North Sea Continental Shelf case, the temporal perspective must be broader (albeit, I yield that the collapse of the Soviet Union contributes to your argument that the 1990 can be set as a new zero day for the practice).
Thirdly, I wonder what is the role of the original state with regard to the referendum, in your eyes? It is one thing to say that referendum held with state's consent is valid and another to say it has legal effect even if there is no consent (even if the referendum complied with the other requirements you present). Thus, the decisive factor for delineating the customary rule we have is the state's consent and not the referendum's conditions, which makes the end result similar to the legal situation we used to have for as long as history can remembers.

susan kemp says

December 1, 2022

Excellent and timely piece which I have shared with others here in Scotland discussing the issue.

Daniel Moeckli says

December 1, 2022

Dear Asher,

Thank you for your comment, you raise some important issues.

1) We don’t assume that a state may only come into being through recognition. Rather, our point is that secession is only lawful if it has been approved in a referendum, and thus that third states may not recognize a state that declares independence without first having held a referendum.

2) Duration is not the decisive component of state practice and there is no minimum period; 32 years is actually not that bad.

3) Whether the parent state has given its consent to the referendum or not may matter from a constitutional law perspective, but it is irrelevant from an international law perspective. Having said this, a popular vote that is organized against the will of the central government is quite unlikely to meet the international legal requirements for the conduct of referendums (peaceful environment, effective protection of civil and political rights etc.). Furthermore, third states will be reluctant to recognize the new state if the referendum was forced through against the will of the parent state.