Next year’s referendum on Scottish independence raises many questions. Do countries have the right to hold referendums? Should they hold them? Does the wording of the question determine the outcome? Who wins?

Matt Qvortrup presents an overview of the history of independence referendums.

In a government building in Montreal, a man sat wrapped in the Quebecois flag Le Fleurdelisé. His eyes were red. He stared glumly into the air and wiped away a single tear. He – and hundreds of thousands of supporters of independence for Quebec – had lost. Just over 40 per cent voted for independence for the Francophone Canadian province in the 1980 referendum; a credible result, but not enough. Years of campaigning had been in vain. The failed referendum on ‘sovereignty association’ is one of my earliest political memories. The referendum was, as I recall it, a battle royale between the captivating French-speaking Canadian Prime Minister Pierre Trudeau and the charismatic leader of the separatist Parti Québécois, René Lévesque. It was one of those rare political campaigns that – for a brief moment – replaced discussions about ice hockey, soap operas and celebrities in smoky bars and over dinner tables across Quebec. The verdict of psephologists is that the referendum in Canada was won because a French-speaking politician (Trudeau) was able to return to his native land and appeal to his brethren. The showdown between proponents and opponents of Scottish independence next September will follow a different pattern to Quebec in 1980, but many things will be the same. As the date of the referendum approaches, politicians, pundits and the public are increasingly looking to other examples of independence referendums to understand, analyse and make sense of the forthcoming Scottish vote. Who will win? David Cameron may not have the celebrity qualities of Pierre Trudeau – and, in any case, is not Scottish.Furthermore, the coalition government is unpopular in Scotland, whereas Trudeau was popular in his native Quebec. But these factors, important though they are, do not mean that the vote is in the bag for Scotland’s First Minister Alex Salmond and his followers. The forthcoming referendum on Scottish independence promises to be a defining moment in the history of the hitherto United Kingdom. But the referendum is likely to be more than just a bare-knuckle political fracas; it also raises fundamental questions about national self-determination and democracy, as well as mundane issues about the importance of the question on the ballot.

Patterns of Independence Referendums

To date there have been 49 independence referendums across the world. Some have been official, others unofficial. Some have been successful, and others have ended in failure. The first independence referendums were held in Texas, Tennessee and Virginia in 1861. All voted for secession, but their wishes were not granted and were resolved on the battlefield in the American Civil War. Forty-four years later, Norway voted for secession from Sweden. But these votes are outliers. Generally speaking, referendums on independence have come in waves (see Figure 1). Without going into a deep discussion about overall patterns, it seems that momentous changes in the international system are correlated with independence referendums.

Quebec independence supporters in 1995. Unionists won the referendum that year, but by the thinnest of margins

In the wake of the second world war, during the process of decolonisation, several countries broke free from their erstwhile colonial overlords following successful referendums. Similarly, in the period following the collapse of the Soviet Union there was an explosion in the number of independence referendums. This pattern might be understandable: countries that were previously able to prevent independence due to a combination of legal and power-political factors were unable to stem the tide of national awakening. Generally speaking, turnout in independence referendums has been high. At 79 per cent, the average turnout is above that for general elections. Perhaps just as interesting, the vast majority of these votes have resulted in large majorities for the separatists. (The average yes vote is 83 per cent.) But this figure disguises the fact that many referendums were held under less than ideal conditions and in countries with less than impressive track records of free and fair elections. If we limit our analysis to independence referendums held in democratic countries after 1945, the average yes vote is 62 per cent. Still, there is a clear tendency that independence is a popular option. 


Figure 1: Independence referendums, 1861-2011
Source: Qvortrup, 2012


Referendums and International Law 

‘Why is it,’ pondered Herman Hesse, ‘that you are only in favour of national-self determination when you can profit from it?’ The debate about the people’s rights to democratically decide their fate has entertained many lesser minds than the author of the Steppenwolf, Siddhartha and The Glass Bead Game. Hesse’s cynicism was as warranted in the wake of the first world war as it is roughly a century later. Very few countries have freely accepted the results of referendums on independence that have taken place. The Soviet Union did not accept the secession of Latvia, Lithuania and Estonia; the break-up of Yugoslavia, which was preceded by popular votes, was likewise rejected by Belgrade. True, neither Yugoslavia nor the Soviet Union were democratic states, and might not be expected to be committed to the ‘self-determination of the peoples’. But opposition to secession through referendums is not confined to authoritarian states: for example, in 1946 the Danish government did not accept the outcome of a referendum on independence for the Faroe Islands. After negotiations, the Danes accepted that the Faroese kept their MPs in Copenhagen but were granted legislative power in all areas bar foreign affairs and defence. In effect the Faroe Islands got what has been called ‘devolution max’. This deal was sealed when unionist parties won the hastily organised general election to the Lagtinget (the Faroese Parliament) shortly after. More recently, the Royal Supreme Court of Canada ruled that ‘any attempt to effect the secession of a province from Canada must be undertaken pursuant to the Constitution of Canada, or else [would] violate the Canadian legal order’. Consequently, a referendum on the independence of Quebec would not be permitted due to the absence of a constitutional amendment. Similarly, in the United States, the Supreme Court of Alaska held in Kohlhaas v. Alaska that a referendum on secession would be unconstitutional. The jurisprudence of the US Supreme Court is simple: it is unconstitutional to break up the union. Legally speaking, such rulings are neither new, nor surprising. As far back as the 1860s, in White v. Texas, the US Supreme Court had held that the ‘Lone Star State’ was not constitutionally permitted to hold an independence referendum. Until David Cameron and Alex Salmond reached the so-called ‘Edinburgh Agreement’ last year, there was considerable discussion about the legality of a Scottish referendum. Technically, the Scotland Act 1998 does not permit referendums, but in the Edinburgh Agreement, the coalition government in London and the Scottish National Party administration in Edinburgh agreed to a so-called ‘Order in Council’ under Section 30 of the Scotland Act. In plain English, the two governments agreed to make an exception to the general ban on independence referendums. Such legal considerations, of course, have not prevented would-be countries from holding referendums; nor have the legal issues hindered the international recognition of these states following a referendum. But, as hinted by Hesse, existing states have been rather selective in their decisions to recognise these new states. Politics is not always a very principled profession.

Scotland's First Minister, Alex Salmond, and his deputy, Nicola Sturgeon, will be hoping for a 'yes' vote next year. Press Association

The Question

There has been a considerable debate about the wording of the question on the ballot in independence referendums. The Scottish government’s proposed question on next year’s ballot was ‘Do you agree that Scotland should be an independent country?’ The inclusion of the word ‘agree’ led to criticism that they were trying to influence the result by using positive language that could sway voters. As a consequence, the Electoral Commission changed the wording to ‘Should Scotland be an independent country?’What was the fuss about? The argument, credibly enough, holds that a biased and one-sided question can prompt voters to vote yes to a question that – had they understood it – they would have rejected. Is this a real danger in referendums on independence, and will the voters be swayed by artful misrepresentations in the form of misleading statements? Or is the question of minor importance? It is difficult to answer this with anything like cast-iron certainty, but we can – perhaps – draw some conclusions if we compare some recent examples of the wordings used in referendums. In Northern Ireland, in 1998, the voters were asked to approve (or otherwise) the rather neutral question ‘Do you support the agreement reached in multi-party talks on Northern Ireland and set out in Command Paper 3883?’; 71.2 per cent did so. ‘Command Paper 3883’ was a coded reference to the official document containing the Belfast Agreement on power-sharing. In other referendums, however, voters have been asked more leading questions. For example, in March 1991, a referendum in the Soviet Union asked ‘Do you consider necessary the preservation of the Soviet Union as a renewed federation of equal sovereign republics in which the rights of the individual of any nationality will be fully guaranteed?’ Seventy-seven per cent voted yes, but the referendum was boycotted in many states. One of the objections against the SNP’s proposed referendum in 2014 was that the word ‘agree’ might create a bias and encourage voters to support independence. However, there are several examples of similar questions. In 1999, in East Timor, the voters were asked the question ‘Do you accept the proposed special autonomy for East Timor within the Unitary State of the Republic of Indonesia?’. A majority of the voters – close to 75 per cent – rejected the proposal, with the result that East Timor became independent. In this internationally monitored referendum, that valueladen word ‘accept’ did not swing the vote. A similar conclusion could be drawn from the referendum in Quebec in 1995, where voters were asked a question that included the word ‘agree’:

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 respectingmthe future of Quebec and of
the agreement signed on June 12, 1995?

While the result was very close (the proposal was rejected by just 51 per cent of the electorate), there was no indication that the wording of the question swayed the decision. Citizens had learned about the pros and
cons of the proposed ‘sovereignty association’ during the campaign. In both East Timor and Quebec, attempts to hoodwink the voters to support a proposition by using positive language failed. 


Former Chancellor Alastair Darling, chair of the No campaign, extols the virtues of the Union to an audience at the Scottish Conservative Party conference in Stirling
Press Association


‘Neverendum’ – that was the word suggested by Anglophone Canadians in the wake of the second unsuccessful referendum in Quebec. Having lost heavily in 1980, the Parti Quebeçois simply waited 15 years and
asked the voters anew. The close result – independence was defeated by less than 1 per cent – suggested that the word ‘neverendum’ was indeed appropriate. It seemed to many that the separatists would keep holding referendums until they got the result they wanted. The referendum in Scotland is unlikely to resolve the matter oncen and for all. The most likely outcome is a unionist win, but with a less than decisive majority. If independence supporters manage to get more than 40 per cent, there is every chance that the referendum in Scotland will also become a ‘neverendum’.


Hesse, H. (1997) Lektüre für Minüten (ed.mVolker Michels). Frankfurt: Suhrkamp. Qvortrup, M. (2012) ‘The History of Ethno- national Referendums 1791–2011’, Nationalism and Ethnic Politics, 18 (1), 129–50.

Legal Cases

Kohlhaas v. Alaska 147 P 3d 714 (2006).
Reference re: Secession of Quebec (1998) 161
DLR (4th) 385.
Texas v. White 74 US 700 (1868).

A lawyer and a political scientist, Dr Matt Qvortrup is a senior researcher at Cranfield University. He won the Political Studies Association’s BJPIR Prize for best article in 2012. His recent publications include The British Constitution; Nationalism, Referendums and Democracy; and Direct Democracy.