In his speech before the Catalan regional parliament on 10 October 2017, the Catalan President Carles Puigdemont suspended a declaration of independence but stated that the referendum of 1st October gave the Catalans a mandate for creating a sovereign state. This post examines whether this assertion is borne out by international law. I submit that neither the Catalans and their leaders nor the central government act in an international law-free zone.
A declaration of independence would not violate international law
The International Court of Justice, in its Kosovo opinion of 2010, found that a unilateral declaration of independence does “not violate general international law” (para. 122) ─ if such a declaration is not “connected with the unlawful use of force or other egregious violations of norms of general international law, in particular those of a peremptory character (jus cogens)” (para. 81; see also paras 84, 119-121 on non-violation). The ICJ in that Opinion inverted the legal question placed before it (which had been whether the declaration of independence was “in accordance with international law” (para. 1)). The Court had also shied away from saying anything meaningful on secession (as opposed to the speech act of declaring independence). In result, the Advisory Opinion came out as a parsimonious if not meagre restatement of the law.
Disproportionate use of force (police and military) is prohibited by international law
However meek, the Kosovo Advisory Opinion is relevant for Catalonia also with regard to the prohibition on the use of force. The Court here said that “unlawful use of force” would taint a declaration of independence and make it violative of international law (para. 81), but did not say when such resort to force would indeed be “unlawful”. Also, the ICJ did not say whose use of force although it probably had the separatists themselves in mind.
I claim that Spain is not allowed to use disproportionate violence against separatists. Surely, there is ─ under the law as it stands ─, no general ius contra bellum internum. The prohibition of the use of military force (as enshrined in Art. 2(4) of the UN Charter and parallel customary law) normally applies only in the relations between states. In addition, the use of military force is prohibited in constellations of “green lines”, among stabilised de facto regimes, or when a Security Council resolution specifically prohibits resort to force, or where separatist armed group and a central government committed themselves not to use force in a treaty of armistice (e.g. as in Georgia in its relations to separatist Abkhazia and South Ossetia). (See for this legal proposition GA Res. 2625 (XXV) of 24 October 1970: “Every State likewise has the duty to refrain from the threat or use of force to violate international lines of demarcation, such as armistice lines, established by or pursuant to an international agreement to which it is a party or which it is otherwise bound to respect”; the Eritrea-Ethiopia Claims Commission, Partial Award, Jus Ad Bellum, Ethiopia’s Claims 1-8 of 19 December 2005, para. 10; Independent International Fact-Finding Mission on the Conflict in Georgia, report (‘Tagliavini-Report’) of September 2009, Volume II Chap. 6, at 239-241 and 291).
But this does not mean that the parent state may repress a rebellious or secessionist group by police or military action without legal constraints flowing, inter alia, from international law. Internationally recognised human rights constitute the yardstick in constellations of internal unrest below the threshold of non-international armed conflict (cf. Art. 1(2) AP II of 1977 to the Geneva Conventions of 1949). Should the situation in Catalonia escalate to a NIAC, the pertinent rules of international humanitarian law will apply.
No “remedial secession” here
Finally, just a reminder that the preconditions for tolerating an extraordinary remedial secession are not satisfied in the case of Catalonia.
Most observers accept the Catalan proposition that they form a “people” in terms of international law which is entitled to self-determination and which could constitute the “personal” element of a new state (consisting of a people, a territory, and a government). This self-constitution must be expressed in a “free” way (cf. common Art. 1 of the UN Human Rights Covenants of 1966), and this is where procedural standards kick in. For a political actor and potential holder of the collective right to self-determination to lawfully exercise an extreme form of the right to self-determination (namely to secede), both material and procedural conditions must be fulfilled.
In its decision on the question on the secession of Québec, the Canadian Supreme Court recapitulated the state of international law of the 1990s (Reference re Secession of Quebec,  2 S.C.R. 217) of 20 August 1998). First of all, the international right to self-determination of peoples (cf. Art. 1(2) of the UN Charter and the mentioned common Art. 1) must be exercised in an “internal” fashion, that is through arrangements of political participation and representation within the framework of an existing territorial state) without touching the state’s territorial integrity. The Canadian Supreme Court then mentioned three (more or less well defined) contexts in which the international right to self-determination of peoples could be exercised “externally”, potentially meaning secession. Besides colonial self-determination and other “alien subjugation, domination or exploitation outside a colonial context” (paras 132-133), the Court named “remedial secession” ─ but left explicitly undecided whether this was covered by international law as it stands:
Para. 134. A number of commentators have further asserted that the right to self-determination may ground a right to unilateral secession in a third circumstance. (…) [T]he underlying proposition is that, when a people is blocked from the meaningful exercise of its right to self-determination internally, it is entitled, as a last resort, to exercise it by secession. The Vienna Declaration requirement that governments represent ‘the whole people belonging to the territory without distinction of any kind’ adds credence to the assertion that such a complete blockage may potentially give rise to a right of secession.” Para. 135: “Clearly, such a circumstance parallels the other recognized situations in that the ability of a people to exercise its right to self-determination is somehow being totally frustrated. While it remains unclear whether this third position actually reflects an established international law standard, it is unnecessary for present purposes to make that determination (emphasis added).
I personally lean towards the view that remedial secession is nowadays part and parcel of international law, not as a right but as a legal “defence”, or as a moral (as opposed to a legal) right of resistance or a means of revolution. In any case, the degree of “exploitation” that the Catalans seem to be suffering is not sufficient to justify such a legal defence, at least not without further negotiations which are now under way.
Referendums as a procedure to exercise the international right to self-determination
In the Catalan independence referendum of 1st October 2017, 96 % voted for independence, but with only a 42 % turnout. Such resort to a referendum has followed a meanwhile familiar pattern, well established in the post-1989 era of great territorial realignments, marked by the dissolution of the Soviet Union and the Socialist Republic of Yugoslavia on the one hand, and the deepening and widening of the European Union on the other hand. Almost all of these events were prepared, accompanied, and justified by referendums, often formally binding ones, sometimes only consultative, but followed by the governments involved.
Prominent examples are the Ukrainian referendum of 1 Dec. 1991 (which inaugurated the “parade of sovereignties” of the Socialist Soviet Republics), the no-vote of the French and the Dutch on the European Constitutional Treaty in 2004 (which killed the Treaty), the referendum on independent South Sudan in January 2011 (which led to the creation of a new state now fraught with civil war), and the referendum on the transfer of Crimea from Ukraine to the Russian Federation in 2014 (which occurred under gun-point of Russian soldiers and was declared null and void by the UN General Assembly (Res. A/68/L39 of 27 March 2014, para. 1)). Famously, Opinion No. 4 of the Badinter Commission on Bosnia-Herzegovina asked for a referendum as a pre-condition for the recognition of a new state by the European Community (repr. in ILM 31 (1992), at 1501-3).
Asking the people to decide directly on founding a new state (Kurdistan), on splitting off from a state (Catalonia), or from a highly integrated polity (UK), seems democratic at first sight. But what about the international rule of law? I am deliberately writing “international”, because the constitutional admissibility or inadmissibility of the referendum is irrelevant here. It is typical that territorial referendums conducted in the exercise of the right to self-determination are unconstitutional under the law of the parent 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).
It is controversial whether a customary law requirement to hold a referendum already exists as a matter of hard international law. But the legal status of requirements on the modalities of territorial referendums are independent of the question. In any case, a conditional scheme applies: even if there were no international law obligation to organise a referendum, international law still regulates its modalities and procedures. Notably the Venice Commission (here, here, and here) and the Council of Europe’s Parliamentary Assembly have established “soft” international standards on referendums as a form of exerting direct democracy, including territorial referendums, e.g. the one in Montenegro (2006).
These procedural requirements can be summarised as follows: Use of force is prohibited, while peaceful and democratic procedures are prescribed. One of the appropriate procedures is notably recourse to a free and fair referendum on independence or a democratic election, ideally under international supervision (cf. the written statement by Switzerland in the Kosovo proceedings before the ICJ, of 17 October 2008, paras 69-80).
To conclude, even if a (properly conducted) independence referendum might be a necessary precondition for lawfully asserting the independence of Catalonia, it is not a sufficient condition under international law. But this does not entitle the Spanish government to use disproportionate physical force to discipline the Catalans.
The populist politics of referendums
Despite their facial legal appeal, resort to referendums in matters of international law is deeply ambivalent. Reliance on a popular vote on territorial realignments has often been used as a populist device, prone to manipulation and demagoguery. In the context of the dissolution of the Soviet Union and Yugoslavia, the outcomes of the referendums basically always followed ethnic lines. The votes seemed to perpetuate and even reinforce illiberal democracies, based on ethnically homogeneous peoples. The classic international legal term was, not coincidentally, “plebiscite” (as opposed to the modern term “referendum” which was mainly introduced by the United Nations in the decolonisation context). The term “plebiscite” clearly has the negative overtone of populism.
I have myself consistently defended referendums (or other mechanisms of democratic decision-making) as a procedure for territorial realignment. It is surely preferable to determine the territorial contours of a polity on the basis of the consent of the governed following public debate among all affected groups, not on the allotments made by hegemonic powers on the drawing board or in green rooms. But we must not forget that the referendum was first of all designed as a procedure to confirm, define, or reject the drawing of a boundary where the basic decision that there should or could be an international boundary was already agreed upon (such as in the case of the Scottish referendum based on an agreement between the British and the Scottish government).
This was the Wilsonian inception:
And there is a deeper thing involved than even equality of right among organized nations. No peace can last, or ought to last, which does not recognize and accept the principle that governments derive all their just powers from the consent of the governed, and that no right anywhere exists to hand peoples about from sovereignty to sovereignty as if they were property. (Woodrow Wilson, Address of the President of the United States of 22 January 1917, p. 6).
In reaction to this Wilsonian claim, there has been some controversy in international law on a possible requirement of a referendum. But that controversy related to territorial changes consented to by the governments involved. Put differently, the question was only whether a given executive consent needed to be supported by an additional popular consent. This structure of the debate had to do with the fact that the early referendums after 1914 only related to cessions, i.e. to transfers of territory on the basis of international treaties between the states concerned. The second type were decolonisation referendums on the legal basis of the colonial right to self-determination, where the release into independence in the end also happened with the consent of the then-colonial powers (not against their will). Only after 1989, most referendums accompanied the dismemberment of a state (the Soviet Union, which disappeared as a subject of international law), or successive secessions (the case of Yugoslavia).
Granted, the formal distinction between consensual and non-consensual territorial re-ordering is eroded in practice: The breakup of a state or the breakaway of a part of its territory normally is a protracted process during which the political attitudes of the actors, including that of the central government, change. For example, the Soviet Republics initially declared their independence against the will of the central government, but in the end that central government agreed. The same is true for all cases of decolonisation, likewise for the splitting off of South Sudan from Sudan in 2011, and arguably even for Yugoslavia.
Now can we say that once we accept a requirement of a democratic justification, this rule must extend to all types of territorial changes, especially against the background that a neat categorisation is not possible in practice? I have defended this position in the past but I am not so sure anymore. Where there is no underlying political acceptance, also among the other populations who will be directly affected by a secession (and who should therefore also have a say on the matter), e.g. the Spanish people, a territorial referendum seems more populist than democratic.