Political Contexts, Doctrinal Moves, and Framing Principles for a Code of Practice

In thinking about the value of a code of practice, it is worth remembering its real limitations, which are political. Codes are technical – but they depend upon some prior moment of decision. A code, however detailed, will have no effect if not implemented, and its implementation – though sometimes subject to procedures that themselves can be technical – depends upon political will.

The Political Context

In Europe, as in the world more broadly, states’ territorial sovereignty is notionally a domestic, political matter. But the predominant sentiment is suspicious of territorial sovereignty disputes, seeing them as challenges to European and international order. That hostility rests on assumptions that have the quality of unexamined prejudices; they are the real challenge this project faces, which it must overcome to succeed.

This article is a contribution to the third, more technical part of our project. But taken as a whole, that project should tackle head-on the assumption that TSDs are antithetical to Europe and its ideals. There are clear, compelling arguments against this prejudice, but they need to be made strongly and unapologetically. When they are, resistance to them often collapses. Articulating a formal framework of practices may help catalyze that political shift – but it is the shift we should keep in focus. Our project is a code, but its purpose – our real project – is to radically change minds, so that a code like this might make sense. Many parts of this project should be cautious, careful and modest – but that goal is not one of them.

Two Doctrinal Moves

Two doctrinal moves would be critical to creating useful space for a code of practice and framing any European intervention as legitimate.

Subsidiarity

Subsidiarity as a basis for legitimacy, discourse and decision [1]1 — I discuss this at length in “A World Elsewhere: Secession, Subsidiarity, and Self-Determination as European Values,” 23 Revista d’Estudis Autonòmics i Federals 11-45 (2016) “Shifting States: Secession and Self-Determination as Subsidiarity,” Percorsi costituzionali 751-64 (3/2014). : There is a model within the European institutional and political context that, although not itself a detailed code, could provide the justification for a code and shape its details. Subsidiarity (Art. 5(3) TEU) is a framework for deciding on the distribution of power among different levels. In the EU context, this means deciding whether Brussels or states (and their regions and local levels) ought to exercise substantial decision-making power.

We typically find secession treated as the breaking of a political unit, and thus irrelevant to the more genteel norms of subsidiarity; this is part of what contributes to the animus against secession and the belief that it is antithetical to European values. But secession is also a form of subsidiarity – a claim about the right level for governance within a multi-layered system.

Reframing secession as subsidiarity would challenge the mistaken assumption that TSDs are zero-sum calculations – that secession must and should mean withdrawal from the EU and its value system

Secession breaks a state, but not necessarily relations at other levels, such as the regional and transnational. For example, a seceding state would still be subject to important human rights treaties and institutions. (See next section.) Seen that way, secession is just a repositioning within the European order – an assertion that some community ought to have status of a state within Europe’s institutions. A subsidiarity framework allows TSDs to be understood and resolved within the European institutional context, rather than as a challenge to it.

Reframing secession as subsidiarity would challenge the mistaken assumption that TSDs are zero-sum calculations – that secession must and should mean withdrawal from the EU and its value system. That belief is a function of EU treaty law, not logic. It is a discretionary design choice.

Reinterpretation of EU Treaty Norms

This brings us to the second doctrinal move: rethinking the treaty norms regarding membership and withdrawal. Under current interpretations, a community seceding from an EU member also withdraws from the EU. So, one of the most valuable things that work towards a code might accomplish is to alter that practice, which in turn would make a code practically useful. A different interpretation is possible.

Under international law, while most treaty obligations are voided by the creation of a new state (the tabula rasa principle), treaties that are humanitarian in nature presumptively continue in force (the automatic succession principle). Their beneficiaries are individual human beings in their own right, not as subjects of a state – to borrow from Anglo-American property law, their rights run with the land. So rights derived from treaty regimes such as the ECHR would presumptively devolve onto the citizens of a newly independent Catalonia, Bavaria or either Galicia.

What about EU treaty rights? Large parts of the EU treaties aren’t humanitarian, but their fundamental values and rights commitments, and any parts necessarily connected to their operation, could plausibly be seen as humanitarian. (After all, humanitarian treaties often contain provisions that are technical in nature.) This could include membership.

The EU treaty regime does not have an essential nature – its nature is a function of its provisions and the interpretation given them. Article 50 is a design and policy choice; it would be entirely possible to create a European constitutional structure that presumed continuity of membership [2]2 — The principal focus of our efforts should be the EU. However, the Council of Europe – some of whose treaties already presumptively continue – might be a valuable alternative point of entry, and might be marginally more amenable to the idea that its democratic values require some regard for TSD claims. CoE institutions such as the Venice Commission could scrutinize constitutional commitments to ensure that protections for territorial integrity do not license the use of military or public order powers against peaceful internal challenges to territorial sovereignty. If the Commission were even half as assertive about identifying limits on states’ ability to use their police powers in TSDs as it is in protecting democratic principles in other contexts, it could make significant contributions towards altering the doctrinal and political presuppositions underpinning resistance to TSD resolution. . There is no conceptual obstacle to amending the treaties to provide explicitly for continuation, reading the supposed Article 50 limit out [3]3 — After all, Article 50 in fact says nothing at all about secession: It refers to a member state withdrawing. At least since Brexit, it is obvious that a sub-state territory might secede from a member state in order to remain in the EU.) Equally, since the continuator state’s membership does not lapse, if the seceding state was also treated as a continuator – and nothing in international law prevents this – then membership could continue for both. or interpreting the existing treaties as humanitarian for purposes of treaty succession after secession, to benefit the EU citizens living within.

Reinterpretation might also allow conditional succession – provisional or suspended membership to preserve the essential core of fundamental values and rights while technical aspects are worked out to ensure a reactivated full membership. Although an enormous task, conceptually it would not differ from what already happens: a new state succeeds automatically to humanitarian treaties, even though some elements must be modified to account for the change (names, composition of committees, contribution formulae, etc.) It would be a technical job for lawyers to revise, rather than a contested political question.

There are serious obstacles to such a move. It would have to be a decision by the member states themselves, either in an explicit amendment of the treaty, or a kind of authoritative commitment. Either would require considerable democratic legitimacy. And it is not as if a formal reinterpretation of the treaty would reshape political norms, rather the other way round: Once the political norms shift, the reinterpretation will follow.

It is not clear what the better strategy is. It might be that running straight at this most political of questions is unwise – instead, building a code could lay the groundwork for change more readily and quickly than tilting at the political windmill. I don’t know. But it is sure that before any code is successfully promulgated, the political question will have to be answered, sooner or later. And if it were answered, then the process of creating and implementing an effective code would follow much more easily: A change in the treaty text, or in its interpretation, would clear the space in which such a code could develop.

If the European Union truly believes in an ever closer union among the peoples of Europe and in its own values, why does it say that the resolution of a domestic political dispute necessarily leads to the expulsion of one side of that dispute?

Ultimately, sovereignty conflicts are political conflicts, and their resolution will not depend on codes or technical practices. Our strategic objective has to be to challenge the legal and political culture that views TSDs with hostility and enables resistance to peaceful resolution. Unless those norms are challenged, no code will achieve much – but perhaps drafting a code can help focus attention on the argument that European norms of democracy and participation require European institutions to rethink their reflexive commitment to existing territorial arrangements.

I think a direct challenge to the current interpretation of Article 50 might help in this goal. If the European Union truly believes in “an ever closer union among the peoples of Europe” (Art. 1 TEU) and in the attractive force of its own values, why does it keep a rule denying people who share those values the right to remain in that union? Why does it say that the resolution of a domestic political dispute – in which the Union notionally has no interest – necessarily leads to the expulsion of one side of that dispute? The default presumption of the EU should be that the protections, benefits and obligations of membership continue unless repudiated.

Framing Principles for a Code

To Whom Should a Code of Practice Be Addressed?

The present draft imposes many conditions on the state. But it might be as productive – and strategic – to put the obligations on the claimant. Groups imposed standards on themselves to purchase legitimacy. Claimants are proposing to break apart existing states and create new ones – new coercive political projects. This requires considerable justification – even if we admit that existing states need justification. Claimants are in a weak position, and by imposing high standards on themselves, they can marginally improve their position – at least by inoculating themselves against easy objections. Hewing to high, defined standards could improve the diplomatic prospects of claimants within European institutions.

This is how the Quebec court described the politics of recognition, which applies equally to the politics of negotiating TSDs:

  • [T]he viability of a would-be state in the international community depends, as a practical matter, upon recognition by other states. That process of recognition is guided by legal norms. . . .
  • [O]ne of the legal norms which may be recognized by states in granting or withholding recognition of emergent states is the legitimacy of the process by which the de facto secession is, or was, being pursued [4]4 — Reference re Secession of Quebec, [1998] 2 S.C.R. 217 ¶¶ 142-3 (“Quebec Reference”). .


The court is describing conditions a unilateral secession must meet to deserve recognition. If a claimant fails to negotiate in good faith – defined in part by a code of practice – it would risk losing support:

  • [A]n emergent state that has disregarded legitimate obligations arising out of its previous situation can potentially expect to be hindered by that disregard in achieving international recognition. . . . On the other hand, compliance by the seceding province with such legitimate obligations would weigh in favour of international recognition [5]5 — Quebec Reference ¶ 143. .


So instead of speaking primarily about what states must do, a code could describe what the claimants must do to be taken seriously. Implicit here is the expectation that existing states would then be under greater pressure to take these claims seriously. For what the Quebec court said about secessionists if true of states too: If the state doesn’t negotiate in good faith with a claimant following legitimate standards, its own case becomes weaker [6]6 — The current draft standards largely presume a cooperative state – sensibly enough. But a claimant-focused code could also define standards-based behavior for dealing with a resistant or hostile state. .

Obedience – Minimal Interference with the State

In any TSD, the claimant doubtless has particular complaints – certain laws or practices – that it would like to change. But the core of a TSD is a dispute is about the sovereignty of the existing state – not the rules, but the power to make rules. The result of a TSD may well be the termination of the existing state’s legal authority, but the dispute itself is not properly a license to nullify the state’s laws, and a claimant weakens its own legitimacy if it conflates the TSD with nullification.

Claimants’ legitimacy is at its highest in asserting the right to contest the state’s sovereignty as such – the right to decide. The farther from this core, the less legitimate the claimant’s actions will seem. It follows that to the degree possible, claimants should respect and obey state law, deviating from it only in the most minimal ways necessary to assert and achieve their right to contest sovereignty [7]7 — These will inevitably be difficult questions of interpretation: in Catalonia, the use of public funds was prosecutable on grounds formally unrelated to the substantive TSD. But even there we can discern a difference between core and peripheral claims and their relative legitimacy: European outrage was focused on the violent suppression of attempts to vote, not on Spain’s resistance to legislative and administrative actions of the Catalan government aimed at entrenching its local power or nullifying Spain’s in advance of the referendum. . In turn it is this obligation on claimants to obey the states’ norms in all other matters that justifies the duty of the state to engage in good faith negotiations. The claimant should obey in all things, asking only one, and the state may demand obedience in all things, allowing only one: the right to contest sovereignty.

A claimant already possessing governmental power under the existing legal order should make only those minimal changes to governance whose purpose is to make a decision on resolving the TSD possible, and to prepare for independence in the event it is successful [8]8 — One corollary is that claimants should separate TSD referenda from regular elections if possible; see below. .

In the European context, there is no good alternative to advancing TSD claims on the basis of law and democratic principle

Polite, rule-abiding behavior can be frustrating, especially when the deck is so clearly stacked against TSD claimants. And radicalization can be effective, just as being the victim of state oppression can greatly strengthen sympathy and support. However, waiting for threshold violations of democratic norms or violence is not a real strategy, and certainly not a pleasant one. Radicalization is also as likely to generate resistance as support, and gives the state a pretext to invoke its public order powers. In the European context, there is no good alternative to advancing TSD claims on the basis of law and democratic principle. Entrenching a ‘no unnecessary nullification’ standard in the code will strengthen that approach.

The Nature of the Territory at Issue

A code of practice might include two separate processes: for TSDs based on existing sub-state units, and for TSDs that propose novel units [9]9 — Another plausible frame is to stress the protections for national minorities – something that would invoke the competencies of the Council of Europe and OSCE (including its High Commissioner on National Minorities). But this is a less promising route: It makes the issue one of ethno-national status, rather than popular sovereignty, and this inevitably raises the problem of remaining members of the state’s majority ethno-national groups (as well as other minorities); it ends up giving the state, and outside groups, a reason to reject the claim as illiberal. Moreover, the very idea of a minority is, by definition, a group not entitled to its own state. This is a nonsensical subordination of substance to arbitrary definition, but it is a social and political reality: the language of national minorities is likely to prove self-limiting, even self-defeating, in a TSD. . The former includes such disputes as Scotland and Catalonia, the latter Padania or, outside of Europe, the Kurds of Turkey or Iraq.

Relying on existing units offers strategic advantages: predictability and stability of expectations; preexisting governance infrastructure that can be used to advance a claim; some measure of recognition. But a code built only for existing units would exclude legitimate claims. After all, existing units are functions of the existing state, and some were designed in part to frustrate sovereignty claims – the very question at issue.

Any code needs to identify thresholds: What population or territory is required? How is it to be determined? When an existing unit is used, these questions are easier; they are considerably more difficult when there are no pre-existing borders. One possibility is to devise rigorous general norms – such as a supermajority requirement or higher standards of evidence to initiate a process of negotiations – but allow more relaxed thresholds for existing units.

Claimants’ legitimacy is at its highest in asserting the right to contest the state’s sovereignty as such – the right to decide

And on many particulars, the two tracks overlap. Even for existing units, a sensible and humane code would include provisions to ‘right-size’ the territory, to ensure that populations not wishing to exit might remain, and likewise that populations outside the existing unit can join it in exiting, through a process of border plebiscites or adjustments. Each of these modifications requires thinking about populations that aren’t defined by some pre-existing territorial boundary.

Conditions of Exercise – Elaborating the Principle of Clarity

As the Canadian Supreme Court’s Quebec Reference declared, starting negotiations to resolve a TSD requires a clear majority answering a clear question. (Quebec Reference (1998) ¶ 100) Other elements are implicit in these two: majority requires standards for participation and implies some measure of democratic deliberation. And clarity implies a third element: a clear context in which the question is asked and the majority answers.

The Question

The question should meet standards of clarity – “the clear question’ standard to which the Quebec court referred – or as the Scottish and UK governments agreed, the “question must be fair, easy to understand and capable of producing a result that is accepted and commands confidence” [10]10 — Memorandum of Agreement, U.K.–Scot., October 15, 2012, ¶ 5. . In particular, it should avoid leading phrases and historical or ideological preambles; be self-contained (i.e. it should not refer to other documents); and present a simple, binary choice (one of which should be the status quo):

  • The wording of the referendum question must be clear and should leave no room for ambiguity. The question must not be misleading, must not suggest an answer, especially by mentioning the presumed consequences of approving or rejecting the proposal and it must not ask an open question [11]11 — Vanessa Rüegger & Rekha Oleschak-Pillai, “State Secession in International Law and the 2011 Referendum in the Sudan,“ in Peters Dreiblatt: Föderalismus, Grundrechte, Verwaltung, Festschrift für Peter Hänni zum 60. Geburtstag (M. Gredig, et. al., eds; Bern, Stämpfli, 2010), citing Venice Commission Referendum Guidelines 1.3.1(c)), at 61. The Venice Commission referendum guidelines can be applied to TSD referenda. .


As we can see in the UK-Scotland agreement noted above, there are three basic elements to a well-designed question: fairness, clarity, and consequence or legitimacy. The Scottish referendum question provides a strong model that meets these elements: ‘Should Scotland be an independent country? Yes/No.’ The question posed a binary choice, one prong of which was the status quo, the other a preference for independence.

Fairness requires that the question not be biased or leading. In the Scottish case, even the phrase ‘Do you agree’ was seen as encouraging a positive response. Preambles encourage a particular frame for the question that can bias voters; they should be avoided.

Clarity is best achieved by short and simple phrasing. There will often be reasons for more complex formulae (such as asking voters to make a second choice contingent upon a first) [12]12 — The question and any preamble or ancillary materials should make clear if a positive vote would have immediate effect or lead to negotiations for a decision, and if so if these would be subject to a second confirming referendum, which should be the preferred pratice. (The Brexit process suggests the serious problems that arise if a sovereignty referendum is undertaken without a plan for the subsequent process. Much of the paralysis and dysfunction of that process could have been mitigated if the initial referendum had indicated that a second vote would be required at the end of negotiatons.) , but a multi-part question (or a question referring to other documents or decisions) makes it hard for individuals to understand what they are deciding on. This in turn makes it hard to make consequential claims about what has been decided [13]13 — The 1995 Quebec referendum included an extremely unclear question: “Do you agree that Quebec should become sovereign after having made a formal offer to Canada for a new economic and political partnership within the scope of the bill respecting the future of Quebec and of the agreement signed on June 12, 1995?” See “Clarity Act,” in Parli – The Dictionary of Canadian Politics (Peter Donolo, ed., Campbell Strategies, 2017). The question mentions sovereignty (a notoriously unclear concept) but not independence; and it refers to two other documents and a contingent event. . Multi-part, complex questions (and questions asked in connection with regular elections, see below) make it harder to give a sufficiently clear signal.

Consequence or legitimacy – though principally a function of other factors – can be undermined by a question that does not offer a meaningful choice. The Crimean referendum question was leading and biased. It offered restoration of the 1992 constitutional autonomy or union with Russia – but no option for the status quo. Leaving aside the other factors that made that referendum useless, the question alone made it impossible to determine voters’ preferences.

In the absence of state cooperation, claimants should seek interlocutors in civil society or draw on international expertise

If the state is cooperative, the claimant should negotiate the language with state authorities, and if possible rely on existing electoral supervisory mechanisms. In the Scottish case, the wording of the referendum was subject to review by the Electoral Commission, which influenced the final wording to ensure a more neutral phrasing. In the absence of state cooperation, claimants should still follow these principles, seeking interlocutors in civil society or drawing on international expertise that can act as a guarantee for the integrity of the process.

Decisional Majority

Any plausible model for resolving TSD disputes must rely on either institutional decision-making or assertions of popular will. Institutional approaches may not require an actual popular majority, so long as they have a democratically legitimated basis, such as a parliamentary majority. (This was the case in Catalonia.) But in a free society a purely institutional strategy is unlikely to succeed without democratic support, and in cases in which the state is unwilling to cooperate, clear proofs of popular, democratic support are an essential part of a TSD claim’s legitimacy.

Thus resolving a TSD generally requires a democratic majority – the clear majority of the Quebec standard. Different thresholds are possible: bare majority (as in Quebec or Scotland), 55 percent (Montenegro), or a higher supermajority – and interacting with these, the question of quorum or participation (see below). There is a trade-off in the choice: the higher the threshold, the higher the potential legitimacy – but also the higher the chance the referendum will fail.

TSD claimants face serious challenges to their democratic legitimacy, since they seek to replace the state-level democracy with a local one. A bare majority means that even locally, the claimant has only partial support. A high threshold more effectively meets this challenge: Although it is still not a majority in the country as a whole, a supermajority signals that in the affected region there is strong support. With a high enough level of votes in favor (and high participation, see next), even a referendum not supported by the state can generate considerable legitimacy.

The trade-off, of course, is that a high threshold denies smaller local majorities’ claims. To take a concrete case, Catalan separatists have not achieved a popular majority in an electoral contest, and if they had to meet a supermajority standard, their case would be all but foreclosed.

But guidelines would not be a permanent, unamendable constitution, and there is value in vindicating the principle that TSDs are norm-governed, rule-based disputes, even if the initial rules are not favorable to a particular claimant. (After all, right now Catalan separatists face an even higher threshold, since Spain rejects their claim on principle.)

So, while standards should indicate that claims based on a bare majority need to be taken seriously, it is worth considering ways to incorporate a higher threshold, because it purchases greater certainty and legitimacy. So, for example, standards might indicate that states must recognize and negotiate with supermajority claimants, and should negotiate with claimants based on a bare majority, especially where existing units are the focus of the TSD.

Whatever the question and the level of majority, it should be a true, popular majority – not merely a parliamentary majority. By their nature, TSDs are extra-constitutional – it will not do to rely on the constitutional order’s translation of popular will into decisional majority to make a claim against that order. (See below on stand-alone referenda.)

Quorum and participation

Even supermajorities cannot send a clear and consequential signal if participation is low. This is especially true if the process is not sanctioned by a cooperative state, because then only those committed to the cause may vote. Given the fraught nature of TSDs, groups opposed may consider the process illegitimate and refuse to take part, creating a false impression of support and uniformity (as in Bosnia in 1992). Any legitimate model must allow boycotting groups to ‘count’, placing the onus on the claimant to generate sufficient enthusiasm for its cause to reach the requisite quorum.

As with the majority itself, the higher the participation requirement – both in terms of the percentage and the number of individuals participating – the greater the legitimacy. But participation should be with reference to the disputed territory, not the whole territory of the state (such as the provisions in the Ukrainian constitution). All-state referenda effectively preclude all TSD claims.

Where possible, TSD questions should be asked in a free-standing referendum, not as part of an existing electoral process. This allows the clearest possible signal

In cases in which a TSD concerns an existing governance unit, it will usually be appropriate to rely on that unit’s electoral processes, including voter rolls. Even when the claim involves a novel territory, it should still be possible to use the norms and processes of the state’s electoral institutions, such as election commissions. But the process of determining the appropriate electorate – and therefore the required quorum and majority – will be different and more difficult. The model I advocate would allow the claimant to define the relevant voting territory, and uses population minima and cascades to adjust the parameters of the territory [14]14 — I develop this model in Timothy William Waters, Boxing Pandora: Rethinking Borders, States, and Secession in a Democratic World (Oxford University Press 2020). . But there are several models, each with its qualities and demerits, and each affects the democratic legitimacy of the process by altering the chances for a sizable local majority and reducing the number of individuals who might be forced out of the existing state against their will.

Clear Context – A Stand-Alone Referendum

In addition to a clear question and a clear majority, resolving TSDs requires a clear context: Where possible, TSD questions should be asked in a free-standing referendum, not as part of an existing electoral process. This allows the clearest possible signal.

Obviously, if a state is uncooperative, coopting the existing electoral system is an understandable strategy. This is what Catalans did after 2014, by running in provincial elections on a secession platform. But this produced an ambiguous mandate, since voters, whatever their personal motivations, were in fact choosing a government for the existing unit within Spain’s constitutional order – a decision radically different from the extra-constitutional challenge of a TSD.

Where the existing state is supportive of the process, claimants should rely on existing electoral mechanisms – but not regularly scheduled elections. If the regular elections must be used, it would be best to ensure that there is a separate, clear expression of approval by a popular majority, such as a separate ballot question distinct from votes for the legislature or administration. (The very claimants who won the Catalan elections understood this, which is why they proceeded with a separate independence referendum.)

If the state does not support the process, claimants should attempt to organize the process with the least use of governmental resources possible – organizing the vote privately, for example (even though this risks lowering participation rates).

Even the clearest process contains enormous uncertainty in practice. Contextual clarity will not always be possible. Sometimes a legitimately initiated process will lead to a crisis, or sometimes there may be no option except to proceed despite unclear processes and information. But the TSD process should not further contribute to an unclear context. (For example, in certain respects the electoral strategies adopted by the Catalan separatists made it more, rather than less difficult to identify a legitimate majority for their goals.)

Contextual clarity requirements should be rigorous, but should not be framed in ways that preclude the possibility of a viable TSD simply because not all conditions can be met. Doing so would simply prejudge the process, given the bias of the system towards territorial integrity and the power of the state to enforce public order. Contextual standards should be understood as ideals or norms that ought to shape the behavior of all parties, even if, as I have suggested, they are primarily or formally addressed to the claimants.

  • Referències

    1 —

    I discuss this at length in

    • “A World Elsewhere: Secession, Subsidiarity, and Self-Determination as European Values,” 23 Revista d’Estudis Autonòmics i Federals 11-45 (2016)
    • “Shifting States: Secession and Self-Determination as Subsidiarity,” Percorsi costituzionali 751-64 (3/2014).
    2 —

    The principal focus of our efforts should be the EU. However, the Council of Europe – some of whose treaties already presumptively continue – might be a valuable alternative point of entry, and might be marginally more amenable to the idea that its democratic values require some regard for TSD claims. CoE institutions such as the Venice Commission could scrutinize constitutional commitments to ensure that protections for territorial integrity do not license the use of military or public order powers against peaceful internal challenges to territorial sovereignty. If the Commission were even half as assertive about identifying limits on states’ ability to use their police powers in TSDs as it is in protecting democratic principles in other contexts, it could make significant contributions towards altering the doctrinal and political presuppositions underpinning resistance to TSD resolution.

    3 —

    After all, Article 50 in fact says nothing at all about secession: It refers to a member state withdrawing. At least since Brexit, it is obvious that a sub-state territory might secede from a member state in order to remain in the EU.) Equally, since the continuator state’s membership does not lapse, if the seceding state was also treated as a continuator – and nothing in international law prevents this – then membership could continue for both.

    4 —

    Reference re Secession of Quebec, [1998] 2 S.C.R. 217 ¶¶ 142-3 (“Quebec Reference”).

    5 —

    Quebec Reference ¶ 143.

    6 —

    The current draft standards largely presume a cooperative state – sensibly enough. But a claimant-focused code could also define standards-based behavior for dealing with a resistant or hostile state.

    7 —

    These will inevitably be difficult questions of interpretation: in Catalonia, the use of public funds was prosecutable on grounds formally unrelated to the substantive TSD. But even there we can discern a difference between core and peripheral claims and their relative legitimacy: European outrage was focused on the violent suppression of attempts to vote, not on Spain’s resistance to legislative and administrative actions of the Catalan government aimed at entrenching its local power or nullifying Spain’s in advance of the referendum.

    8 —

    One corollary is that claimants should separate TSD referenda from regular elections if possible; see below.

    9 —

    Another plausible frame is to stress the protections for national minorities – something that would invoke the competencies of the Council of Europe and OSCE (including its High Commissioner on National Minorities). But this is a less promising route: It makes the issue one of ethno-national status, rather than popular sovereignty, and this inevitably raises the problem of remaining members of the state’s majority ethno-national groups (as well as other minorities); it ends up giving the state, and outside groups, a reason to reject the claim as illiberal. Moreover, the very idea of a minority is, by definition, a group not entitled to its own state. This is a nonsensical subordination of substance to arbitrary definition, but it is a social and political reality: the language of national minorities is likely to prove self-limiting, even self-defeating, in a TSD.

    10 —

    Memorandum of Agreement, U.K.–Scot., October 15, 2012, ¶ 5.

    11 —

    Vanessa Rüegger & Rekha Oleschak-Pillai, “State Secession in International Law and the 2011 Referendum in the Sudan,“ in Peters Dreiblatt: Föderalismus, Grundrechte, Verwaltung, Festschrift für Peter Hänni zum 60. Geburtstag (M. Gredig, et. al., eds; Bern, Stämpfli, 2010), citing Venice Commission Referendum Guidelines 1.3.1(c)), at 61. The Venice Commission referendum guidelines can be applied to TSD referenda.

    12 —

    The question and any preamble or ancillary materials should make clear if a positive vote would have immediate effect or lead to negotiations for a decision, and if so if these would be subject to a second confirming referendum, which should be the preferred pratice. (The Brexit process suggests the serious problems that arise if a sovereignty referendum is undertaken without a plan for the subsequent process. Much of the paralysis and dysfunction of that process could have been mitigated if the initial referendum had indicated that a second vote would be required at the end of negotiatons.)

    13 —

    The 1995 Quebec referendum included an extremely unclear question: “Do you agree that Quebec should become sovereign after having made a formal offer to Canada for a new economic and political partnership within the scope of the bill respecting the future of Quebec and of the agreement signed on June 12, 1995?” See “Clarity Act,” in Parli – The Dictionary of Canadian Politics (Peter Donolo, ed., Campbell Strategies, 2017). The question mentions sovereignty (a notoriously unclear concept) but not independence; and it refers to two other documents and a contingent event.

    14 —

    I develop this model in Timothy William Waters, Boxing Pandora: Rethinking Borders, States, and Secession in a Democratic World (Oxford University Press 2020).

Timothy William Waters

Timothy William Waters

Timothy William Waters és professor i investigador a la Facultat de Dret Maurer School of Law de la Universitat d'Indiana. També és director associat del Center for Constitutional Democracy. Els seus interessos d'investigació inclouen l'estructura del sistema interestatal, el conflicte ètnic, els drets humans, la justícia de transició i el dret comparat, especialment en contextos europeus i islàmics. La seva recerca principal consisteix en redefinir l’autodeterminació per idear un dret efectiu de secessió pacífica. Ha publicat en nombroses revistes destacades de dret internacional i relacions internacionals, incloses Yale, Harvard, NYU, Virginia i George Washington. Ha estat consultor per a l'Open Society Institute, el PNUD i el Ministeri de Justícia de Letònia, i també ha treballat a Human Rights Watch. Va supervisar la implementació dels Acords de Pau de Dayton a Bòsnia per a l'OSCE i, en el marc del tribunal iugoslau de crims de guerra, va ajudar a redactar l'acusació de Slobodan Milosevic. També va ser voluntari del Cos de la Pau a Hongria.