The Referendum and other Essays on Constitutional Politics 9781509929290, 9781509929320, 9781509929313

Table of contents :
Preface
Contents
PART I: REFERENDUM
1. The Referendum: Non basta più
2. The World History of Referendums 1400–2018
I. Introduction
II. The Earlier History of the Referendum
III. Democracy in America 1776–87
IV. The Referendum 1789–1920
V. The Referendum in Switzerland in the Nineteenth Century
VI. The Referendums After the First World War
VII. The Referendum in the 20th Century
VIII. The Use of Referendums After 1945
IX. The Referendum in the Age of Dealignment
X. Conclusion
3. Plebiscites in Dictatorships – Voting as a Signifiant of Total Control
I. The Paucity of Plebiscite Research in Dictatorships
II. The Plebiscite as a Signifiant of Total Control
III. The Plebiscite as a Signifiant
IV. Qualitative Evidence of the Plebiscite as a Signifiant of Total Control
V. Conclusion
4. The Law and Politics of Independence Referendums
I. Once Upon a Time in Norway…
II. The Context and Development of Independence Referendums
III. Legal Recognition
IV. The Legal Argument
V. When are Referendums on Independence Recognised?
VI. Statistical Analysis
VII. Conclusion
Appendix A: Independence Referendums 1980–2017
5. Consociationalism and Direct Democracy: The Paradox of Direct Democracy and Elite Accommodation
I. Introduction
II. Pouvoir Constituant Referendums and Consociationalism
III. The Referendum in Switzerland
IV. Optional Referendums and Consociational Democracy
V. The Period of Radicalism 1848–1919
VI. The Referendum After 1919
VII. Conclusion
6. The Stupidity of the People or the Wisdom of Common Folks? Can Referendums Make Societies Richer and More Equal – or the Opposite?
I. Introduction
II. The Evidence
III. Summing Up
7. Prediction or Prophecy: Forecasting the Outcome of EU Referendums
I. Introduction
II. Empirical Tendencies and EU Referendums
III. Hypotheses
IV. Data and Methods
V. The Development of a Predictive Model
VI. Conclusion
Appendix A: Actual and Predicted Outcomes of Referendums 1972–2016 (Results in Red Are Outliers)
Appendix B: Ballot Questions (Emotive Words Underlined)
8. Brexit as an Inelastic Good: A Case Study in Political Economy of Referendums
I. Introduction
II. An Economic Theory of Direct Democracy
III. The Political Economy of Brexit: Demand for an Inelastic Good
IV. The Political Economy of Bremain
V. The Framing of the Brexit Debate
VI. Other Factors: The Sociology of the Campaign
VII. Last Thoughts
PART II: CONSTITUTIONAL POLITICS
9. Comparative Constitution Making: An Introduction
I. Constitutions in History
II. Constitutions: Idealist Expressions or Pragmatic Frameworks?
III. When?
IV. How?
V. Excursus: The Role of Electoral Systems
VI. Conclusion
10. The Consent of the Governed: The American Revolution and the US Constitution
I. Introduction
II. From Colony to the Consent of the Governed
III. The War and the Articles of Confederation
IV. The Philadelphia Convention
V. Ratification
11. Total Recall: Revocation of Electoral Mandates
I. Introduction
II. The Early History of the Recall
III. The Recall in the United States
IV. Marx and the Recall in Europe
V. Lenin and the Recall
VI. Gramsci, Luxemburg and Other Socialist Advocates of the Recall
VII. Conclusion
12. Last Thoughts
Index

Citation preview

THE REFERENDUM AND OTHER ESSAYS ON CONSTITUTIONAL POLITICS Until recently, referendums were little used. After the Scottish independence and Brexit referendums, they have come to the fore as a mechanism with the potential to disrupt the status quo and radically change political direction. This book looks at the historical development of the referendum, its use in different jurisdictions, and the types of constitutional questions it seeks to address. Written in an engaging style, the book offers a clear, objective overview of this important political and­ constitutional tool.

ii

The Referendum and Other Essays on Constitutional Politics Matt Qvortrup

HART PUBLISHING Bloomsbury Publishing Plc Kemp House, Chawley Park, Cumnor Hill, Oxford, OX2 9PH, UK HART PUBLISHING, the Hart/Stag logo, BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2019 Copyright © Matt Qvortrup, 2019 Matt Qvortrup has asserted his right under the Copyright, Designs and Patents Act 1988 to be identified as Author of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. While every care has been taken to ensure the accuracy of this work, no responsibility for loss or damage occasioned to any person acting or refraining from action as a result of any statement in it can be accepted by the authors, editors or publishers. All UK Government legislation and other public sector information used in the work is Crown Copyright ©. All House of Lords and House of Commons information used in the work is Parliamentary Copyright ©. This information is reused under the terms of the Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/ open-government-licence/version/3) except where otherwise stated. All Eur-lex material used in the work is © European Union, http://eur-lex.europa.eu/, 1998–2019. A catalogue record for this book is available from the British Library. A catalogue record for this book is available from the Library of Congress. ISBN: HB: 978-1-50992-929-0 ePDF: 978-1-50992-931-3 ePub: 978-1-50992-930-6 Typeset by Compuscript Ltd, Shannon To find out more about our authors and books visit www.hartpublishing.co.uk. Here you will find extracts, author information, details of forthcoming events and the option to sign up for our newsletters.

‘I must follow them, the people, for I am their leader’ Alexandre Ledru-Rollin, French Republican Politician (1807–74)

vi

Preface

T

his little book is a collection of essays. As such, many of them are inconclusive and preliminary contributions to an ongoing debate. As the dictionary definition of the word ‘essay’ derives from the French infinitive essayer, ‘to try’ or ‘to attempt’, the present scribblings – based on papers, articles and lectures I have delivered in the period from 2010 – cover various aspects of the referendum (its rationale, its use and its consequences), the drafting of constitutions and the often overlooked institution, the Recall (which allows voters to revoke an elected politician’s mandate). In keeping with the genre of the essay, the articles do not have a single theme or message. The aim of the collection is rather more modest, namely to present reflections which other scholars can use to formulate more coherent and perhaps more scientific papers. The critic György Lukács once wrote: ‘the essayist is the John the Baptist who goes out to preach in the wilderness about another who is still to come, whose shoelace he is unworthy to untie’.1 Without claiming that these scribblings will pave a way for the Messiah, the hope is that the papers will spur thoughts and theories – and that others will be introduced to some of the lesser known aspects of referendums and constitutional politics. The book was put together at a time of considerable constitutional upheaval; not merely here in the United Kingdom, but also in the wider world, in Turkey, in the United States, in Hungary, and many other places. Issues such as populism, Brexit, and a lack of trust in the political class suggest that new mechanisms are needed to restore the public’s faith in constitutional democracy. Should we change the whole constitutional framework? And, if so, what is the most democratic way of changing the constitution? Are more referendums a solution – or part of the problem? The essays look at referendums on independence, at what actually determines the outcome of referendums, and at the rules and regulations that ought to underpin such direct democratic exercises. The essays 1 György Lukács, quoted in Judith Marcus, George Lukács and Thomas Mann: A study in the sociology of literature (Amherst MA, University of Massachusetts Press, 1987) 27.

viii  Preface also touch upon the legality of referendums – especially on the so-called ‘right’ to self-determination, which – as we shall see – does not imply that regions have a legal right to vote on independence. The essays – being exactly this – draw on different approaches and do not rely on a single methodology or the insights of one discipline. Some have a chatty, perhaps even journalistic feel; others follow a more conventional social science format. Thus, the essays in the book draw on economics, history, law and political theory – or the disciplines which collectively constitute the science of politics. Overall, however, the book is based on the assumption that good research should ‘make a specific contribution to an identifiable scholarly literature by increasing our collective ability to construct verified scientific explanations of some aspect of the world’.2. The same writers went on to say that it is the goal of ‘scientific research’ is to make ‘causal inferences on the basis of empirical information’.3 Thus, those who seek to understand and explain the occurrence, outcomes and consequences of referendums, constitutional changes and the effect of other political institutions, should follow the same prescription. But is there a particular format or methods that must be followed? As the essays in this book show, these phenomena can be studied using a wide range of methods, though with a tendency towards using qualitative approaches. The focus on qualitative approaches is not surprising. It is often objected that quantitative analyses fail to uncover the essential issues. In Clifford Geertz’s words, the quest for statistical patterns often fails to unearth the truly significant aspects of social relations and ‘bleach human behaviour of the very properties that interest us before we even begin to examine [them]’.4 While this writer is inclined to agree, there is nevertheless a place for statistical analysis. If we want to understand a pattern of when referendums are won or lost or when new constitutions are likely to endure, then a quantitative approach is useful – even if it cannot stand alone. But above all, the essays are based on one overall – or transcendent – condition, that everything in social science is based on the ability to draw causal inferences. In the poetic words of the biologist Thomas H Huxley, ‘the one act of faith in the convert to science, is the confession of the universality of

2 G King, RO Keohane and S Verba, Designing social inquiry: Scientific inference in qualitative research (Princeton NJ, Princeton University Press, 1994) 15. 3 King et al, Designing social inquiry, ibid, 7. 4 C Geertz, The Interpretation of Cultures: Selected Essays (New York, Basic Books, 1973) 17.

Preface  ix order and the absolute validity in all times and under all circumstances, of the law of causation’.5 Whether the world is subject to this ‘law’, is, of course, a philosophical question.6 Yet for us and in our discussions, we always base discussions on this fundamental axiom. It is not, as Huxley noted, something we can prove; it is ‘an act of faith’. Yet, without causation, without the ability to say that some things follow from others, we can have no rational political debate. Thus, in writing these essays, I hope to have made a case for evidence-based policy making and a case against ‘fake news’ and the absurd belief in ‘alternative facts’, which recently have been espoused by a prominent adviser to the current President of the United States.7 Whether these essays draw the right or the correct conclusions is a matter for the reader to decide. My only hope is that they will lead to a vigorous – but respectful – debate so that we can improve upon our political institutions, and hence – ultimately – perfect our individual and collective lives. In writing these essays I have benefitted from discussions and suggestions from many colleagues. These include, in particular, Sir John Curtice, Julia Smith (Baroness Smith of Newnham), Arend Lijphart, Yanina Welp, John Matsusaka, Mogens Hansen, Mark Beissinger, Carlos Closa, Renata Lodge, Klaus Rasmussen, Laurence Morel, Lee Daly and Alastair Campbell. I am grateful for their suggestions, but I alone am responsible for the outcome; the good, the bad and the ugly. Matt Qvortrup London, 30 November, 2018

5 Thomas H Huxley, Collected Essays: vol 2 Darwiniana (London, Macmillan, 1893–94) 252. 6 Immanuel Kant, of course, pointed out that ‘causality’ is a quality for us (für uns) but not necessarily in the external reality (An sich). I will leave this issue to philosophers. See Immanuel Kant, Kritik der reinen Vernunft 1 (Frankfurt am Main, Suhrkamp, 2014) 226. 7 Aaron Blake, (22 January 2017), ‘Kellyanne Conway says Donald Trump’s team has “alternative facts”. Which pretty much says it all’, The Washington Post (accessed 22 January 2018).

x

Contents Preface����������������������������������������������������������������������������������������������vii PART I REFERENDUM 1. The Referendum: Non basta più���������������������������������������������������3 2. The World History of Referendums 1400–2018���������������������������11 3. Plebiscites in Dictatorships – Voting as a Signifiant of Total Control�������������������������������������������������������������������������34 4. The Law and Politics of Independence Referendums��������������������47 5. Consociationalism and Direct Democracy: The Paradox of Direct Democracy and Elite Accommodation��������������������������63 6. The Stupidity of the People or the Wisdom of Common Folks? Can Referendums Make Societies Richer and More Equal – or the Opposite?���������������������������������������������80 7. Prediction or Prophecy: Forecasting the Outcome of EU Referendums���������������������������������������������������������������������94 8. Brexit as an Inelastic Good: A Case Study in Political Economy of Referendums��������������������������������������������������������� 115 PART II CONSTITUTIONAL POLITICS 9. Comparative Constitution Making: An Introduction����������������� 129 10. The Consent of the Governed: The American Revolution and the US Constitution����������������������������������������������������������� 149 11. Total Recall: Revocation of Electoral Mandates������������������������� 174 12. Last Thoughts�������������������������������������������������������������������������� 192 Index���������������������������������������������������������������������������������������������� 197

xii

Part I

Referendum

2

1 The Referendum Non basta più What is a rebel? It is a man who says no: but whose refusal does not imply a renunciation … A slave who has taken orders all his life, suddenly decides that he cannot obey some new command. What does he mean by saying ‘no’? He means, for instance, that ‘this has been going on for too long’, ‘so far but no farther’, you are going too far, or again ‘There are certain limits beyond which you shall not go’. In other words, his ‘no’ affirms the existence of a borderline. Albert Camus1

N

on basta più – roughly translated – ‘enough is enough’, said the Roman taxi driver as we drove up Via Pietro Roselli towards the American University of Rome. He was not, he said, a ‘populist’, but he had voted against the proposal for a reformed constitution put forward by Prime Minister Matteo Renzi in 2016. Not, the man told me, ‘because he was against the welfare state or the EU’, but simply to send a message to la classe politica. In short, non basta più. His was a signal to the elected politicians that there were certain limits to what they could get away with; that immigration was perceived to be out of control; that growing inequality was unacceptable; and that he disapproved of cuts to healthcare spending. Some might view these utterances as examples of irrationality, just as many saw the Brexit vote in Britain in 2016 and the Colombian voters’ rejection of the peace plan in the same year as indications of democracy gone amok; as conclusive proof that government by the people, should mean representative government; that too many referendums is a folly. There are, of course, examples of referendums that threatened and

1 Albert

Camus, L’homme révolté (Paris, Gallimard, 1951) 1.

4  The Referendum discriminated against minorities. But the often unquestioned faith in the ‘elected aristocracy’ – to use ­Rousseau’s term2 – is on closer inspection in need of revision. While politicians by and large are able to deliberate, there are a fair number of Acts of Parliament which, at the very least, question the sagacity of elected representatives. In the United Kingdom, the Dangerous Dogs Act 1991 and the Local Government Finance Act 1988 (­introducing the Community Charge or ‘Poll Tax’) are examples of ill-considered legislation that reached the Statute Book. It is to avoid such legislation that it can sometimes be necessary to complement representative democracy with mechanisms that allow the voters to say non basta più – and the referendum is, perhaps, the most appropriate means of doing so. Populism can be crude and its proponents can be rude and obnoxious but often they are a symptom of an underlying malaise. In the 1890s, the original Populists urged the introduction of referendums that allowed the people to veto laws passed by legislatures and initiatives that could force recalcitrant politicians to introduce legislation that was opposed by big business. And they were often successful. The Populists were denounced as socialists when they proposed to introduce anti-trust legislation. The Los Angeles Times predicted that ‘radical legislation would result and business and property rights would be subject to constant turmoil at the hands of the agitators’.3 In reality this legislation, perhaps, saved capitalism from itself and made sure that the monopolies – that Karl Marx had predicted in Das Kapital4 – never emerged, and that America did not experience the extreme ideological battles that characterised countries with mechanisms for direct legislation by the people or ‘people’s vetos’ through referendums. As I began to studying the previous examples of Non basta più, a kind of pattern began to emerge, namely one of a self-confident elite defending an abstract ideal which was opposed by ‘ordinary people’, whose everyday experiences were out of sync with the prevailing scientific orthodoxy. Economics professors, senior civil servants and business

2 Jean-Jacques Rousseau, ‘The Social Contract’ in Frederick Watkins (ed & tr), ­Rousseau: Political Writings (Madison WI, University of Wisconsin Press, 1986) 74. 3 LA Times cited in Thomas Cronin, Direct Legislation: The Politics of Initiative, ­Referendum and Recall (Cambridge MA, Harvard University Press, 1989) 52. 4 Karl Marx, Das Kapital Band III in Karl Marx/Friedrich Engels – Werke, Band 25 (Berlin, Dietz Verlag, 1983) 211.

The Referendum  5 leaders would tell the voters that ‘you cannot stop progress’ and would denounce popular movements that called for lower taxes or opposed nuclear power stations. Sometimes, as these groups were on the centre-right, as in the case of the Californian Proposition 13 anti-tax revolt or in the case of the similar movement in Norway, where Anders Lange’s Party for Considerable Reduction in Taxation and Public Intervention (Anders Langes Parti til sterk nedsettelse av skatter, avgifter og offentlige inngrep) won ­representation in the Stortinget (Parliament) in 1974. Like his counterpart in America, 40 years later, Mr Lange was unpolished, expressed extreme views in an uncompromising language. Yet, he had a point. Soon, more mainstream political parties took notice and nudged their policies towards lower taxation. They did not follow the more extreme proposals of Mr Lange, but they adjusted their policies – and soon after support for the radical anti-tax party withered away. To paraphrase ­Friedrich Schiller’s play, the ‘the anti-tax movement had done its duty. The anti-tax movement could go’.5 In the same way, the referendum on nuclear power in Austria in 1978. The Conservative Österreichische Volkspartei (ÖVP) and the ­Freiheitliche Partei Österreichs (FPÖ), then in opposition, were opposed to the opening of the Zwetendorf nuclear power station. Sensing that the issue was controversial with a year to go before the next parliamentary election, the Social Democrat Chancellor, Bruno Kreisky, decided to hold a ­referendum on the subject. He did not need to do so, as his party had a majority in the Nationalrat (the lower house), but he sensed that it was necessary for political reasons. Kreisky narrowly lost the referendum but he did not resign. An issue that had the support of the ‘elite’ was denounced by the people. And, just as importantly, the consequence of the vote was not chaos. Normality soon returned. Kreisky won an absolute majority of the vote at the general election the following year and increased his party’s representation in the Nationalrat. One is tempted to agree with the conclusion drawn by Larry Bartels and Christopher Achen, ‘The historical record leaves little doubt that the educated, including the highly educated, have gone as wrong in their moral and political thinking as everyone else’.6

5 Friedrich Schiller, Verschwörung des Fiesco zu Genua (Fiesco’s Conspiracy at Genoa) (1783), Second Act. Part IV. 6 Larry Bartels and Christopher Achen, Democracy for Realists (Princeton NJ, P ­ rinceton University Press, 2016) 310.

6  The Referendum But this is very much a minority view – among commentators, that is. Examples of ‘ordinary people’ revolting against the elite on certain issues are not uncommon and are often denounced as ‘populist’. Indeed, there is a long tradition of lamentations about the supposed stupidity of the ordinary voters. ‘People in masses are like children, easy to influence and even easier to steer if the message is well-packaged and repeated’ wrote the French sociologist Gustave le Bon (1841–1931).7 Perhaps so for little children, but not for teenagers who are not easy to influence and certainly not easier ‘to steer’. At this time in history we are witnessing the collective kind of adolescence in which large numbers of people – like youths on the threshold of adulthood – reject the diktat of their elders, in this case the political class. The referendum plays a key role in this but perhaps not in a way that most commentators appreciate. For referendums do not allow far-right or far-left demagogues to have free rein – though referendum can occasionally perform the role of a democratic safety valve. This has even happened in Hungary – a country balancing on the brink of becoming a repressive and autocratic state. Yet, even in a country without a viable parliamentary opposition, the referendum has arguably been the only effective check on the semi-authoritarian rule by the Fidesz party. Thus, in 2016, Viktor Orbán failed to win support for his tough stance on immigration in a referendum on the matter as ‘the turnout was only 41.32 % and the poll … thereby invalid’.8 But more importantly, opposition groups also utilised the provisions in the constitution for citizens’ initiatives to halt Orbán in his tracks. For example, a ‘sudden strong public support for a referendum [led] to the cancellation in 2017 of Budapest’s bid to host the 2024 Olympic Games after the new opposition group collected 266,000 signatures for a referendum’.9 Used in this way, the referendum is a shield against ­populists and a safeguard against rulers. But the problem with referendums is that they are a heterogenous category. When initiated by the people or by opposition groups, the institution can be a check on power. However, when a referendum can be

7 Gustave le Bon, quoted in Paul Lendai, Orbán: Europe’s New Strongman (Oxford, Oxford University Press, 2017) 193. 8 Marcin Pomaránski, ‘Direct Democracy in Hungary’ in Maria Marczewska-Rytko (ed), Handbook of Direct Democracy in Central and Eastern Europe After 1989 (Opladen, Barbara Budrich Publishers, 2018) 116. 9 Lendai, Orbán, n 7 above, 186.

The Referendum  7 called by the president or the head of government, it is a different story. In this case the referendum has the potential to be used as a plebiscitary instrument that gives the government carte blanche to enact policies with little or no deliberation – especially if the head of government is a ­popular figure. This is the way the referendum was often used by Charles de Gaulle (who adorns the cover of this book). The former French President (and war hero), stated that he was convinced the ‘sovereignty belongs to the people, provided they express themselves directly and as a whole’. For this reason, he went on, ‘I introduced the referendum system, made the people decide that henceforth its direct approval would be necessary’.10 But what de Gaulle failed to add was that he was to decide when this ‘direct approval’ was ‘necessary’. There can be no doubt that a referendum can make a decision ­legitimate and that a direct appeal to the voters can be a way of overcoming parliamentary opposition. Charles de Gaulle used the referendum when he was unable to win approval of controversial policies (such as the direct election of the president in 1962). He was not the first French executive to do so. A century before, Napoleon III had done the same. On that occasion, Karl Marx noted that the president was in a stronger position and this was why he could use (and abuse) the appeal to the people: While the votes of France are splintered to pieces upon the 750 members of the National Assembly, they are here, on the contrary, concentrated upon one individual. While each separate Representative represents only this or that party, this or that city, this or that dunghill … the President, on the contrary, is the elect of the nation, and the act of his election is his trump card, that the sovereign plays out every four years. The National Assembly stands in a metaphysical relation to relation to the nation; the President stands in a personal relation to the nation.11

Giovanni Sartori – the eminent Italian political scientist – thus concluded that the ‘more patent usurping of presidential powers has occurred with referendums’.12 While the referendum could be justified as a tie-break (as in the case of the Danish referendum on the Single European Act in 1986), Sartori maintained that ‘the line that separates use and abuse is – in the case of referendums – a very fine line. Though, he went on, 10 Charles de Gaulle, Memoirs of Hope (London, Weidenfeld and Nicolson, 1971) 6–7. 11 Karl Marx, Der achtzehnte Brumaire des Louis Bonaparte (Frankfurt am Main, Suhrkamp, 2016) 26. 12 Giovanni Sartori, Comparative Constitutional Engineering: An Inquiry into ­Structures, Incentives and Outcomes, 2nd edn (London, Macmillan, 1997) 137.

8  The Referendum ‘the French Fifth Republic has employed the referendum judiciously (and ­parsimoniously)’.13 Yet the same cannot be said of the United Kingdom, where the ­referendum – with the exception of the referendum on Scottish independence in 2014 – has been used almost solely for party political purposes  – first to keep the Labour Party together in 1975,14 then to maintain unity in the Conservative–Liberal Democrat coalition in 2011,15 and lastly (and fatally) to maintain unity within the Tory Party in 2016.16 When constitutions allow politicians to initiate votes merely to safeguard their own political interests (as in the case of the unofficial referendums in Mexico on President AMLO’s pet projects17) there is a potential that the referendum can be abused. The referendum ought to be a ‘people’s veto and for this reason it was appropriate that a referendum was held on Scottish independence as an election victory for the Scottish National Party (SNP) did not in itself give Alex Salmond’s party a mandate for leaving the United Kingdom. Indeed, 38 per cent of backers of union voted SNP in 2011 according to the Scottish Social Attitudes survey.18 Hence a vote for the Scottish National Party did not imply support for a break-up of the United ­Kingdom. For this reason, a referendum was necessary. But who decides when a referendum should be called? Who decides if a question is of a constitutional nature that requires it to be put to a referendum? In a country without a written constitution, such as the United Kingdom, the constitutionality of an issue is bound to be a political one. Albert Venn Dicey proposed a law enacting that a … referendum … be required by a resolution of either House [of Parliament], in respect of any Act e.g. affecting, 1) The Rights of the Crown, 2) The Constitution of Parliament, the Acts of Union & and other large constitutional topics which might easily be enumerated.19 13 Sartori, Comparative Constitutional Engineering, n 12 above, 165. 14 The United Kingdom European Communities membership referendum, 5 June 1975. On this see: M Qvortrup, Government by Referendum (Manchester, Manchester University Press, 2018) 25ff. 15 The United Kingdom Alternative Vote referendum (on whether to replace the existing ‘first-past-the-post’ system with the ‘alternative vote’ method), 5 May 2011. 16 United Kingdom European Union membership referendum, 23 June 2016. 17 Though note that the otherwise popular AMLO (Andrés Manuel López Obrador) was unable to secure a high turnout and that the legitimacy of the poll was therefore in doubt. El País, 18 August 2018, ‘López Obrador someterá a referéndum el nuevo aeropuerto de Ciudad de México’. 18 I am grateful to Sir John Curtice for providing me with this statistic. 19 AV Dicey to Leo Maxse, 2 February 1894, Maxse Papers, West Sussex County Record Office, Chichester.

The Referendum  9 In countries with a written constitution – like Ireland or Denmark – this means that a referendum is automatically held every time the constitution is amended. This is right and proper but there are important issues which are not in the constitution. Issues like nuclear power, abortion, immigration and same-sex marriage are not usually in the constitution (though in this regard, Ireland is an exception); yet these issues are hotly debated and failure to allow people a say in these matters can lead to a legitimacy crisis. Is there a solution to this? Might it be provided by the referendum? Perhaps we should allow the people a say on the matter – like in some US States, Uruguay, Italy and Switzerland – where a specified number of voters have a right to demand a referendum on existing laws (Italy and Uruguay) or on newly enacted laws (California and Switzerland). These polls, of course, should only be held if this is consistent with the rule of law (such as codified human rights and treaty obligations), and there are many examples of referendums and initiatives that have been nullified due to these concerns.20 But would this not lead to demagogue-fuelled populism and illconsidered decisions? That is always a possibility. Democracy is not risk free. But it is a myth that voters always vote for populist measures such as the death penalty. Indeed, the voters in Ireland voted overwhelmingly to abolish capital punishment in 2001, and ‘Oregon … twice voted to abolish the death penalty … It abolished the death penalty by a large majority on a referendum in 1964’.21 The attraction of the referendum is not that it leads to more liberal or more conservative legislation. Institutions should be neutral and provide equal opportunities for the enactment of policies that reflect the preferences of the voters – the ultimate sovereign. The problem with pure representative democracy is that it does not allow voters to have views that differ from those of the political parties. In Scotland this problem was resolved through a referendum – though only because this was granted by the British government.

20 Kenneth P Miller, Direct democracy and the courts (Cambridge, Cambridge University Press, 2009). 21 JL Kirchmeier, ‘Another place beyond here: The death penalty moratorium movement in the United States’ (2002) 73(1) University of Colorado Law Review 1, 17, fn 103. It should be noted, however, that the otherwise liberal Californian voters rejected Proposition 62 in 2016. The initiative would have put a moratorium on the use for the death penalty in the Golden State. See Jim Miller, ‘California Votes to Keep Death Penalty’, The Sacramento Bee, 9 November 2016.

10  The Referendum But the system should not rely on the whim of politicians. Whether an issue merits a referendum should be decided by the voters. This is how it is in most American States. Thus in 2018, a majority of the voters in Florida voted for a conservative Republican, Ronald DeSantis, in the gubernatorial election. Yet, at the same time the voters in the Sunshine State also endorsed a citizen-initiated proposal to enfranchise former convicts (excluding those who have served time for murder or serious sexual offences). Sixty-four per cent endorsed the so-called ­Amendment  4. In other words, the Floridians showed that they could distinguish between measures and men. The system in Florida allows voters to have their democratic cake and eat it. But for this to happen, there must be a mechanism which takes the initiative away from the executive and returns it to the People. Writing at the height of the Populist revolt in America around 1900, Morsei Ostrogorski believed this could be achieved through direct democracy. He suggested that the referendum was ‘aimed at the arbitrariness and the corruption of the parties, which are entrenched behind legislatures’, and that this device provided a mechanism for ‘holding in check the permanent and crystalized forces, in which selfish designs of every kind so easily take refuge’. ‘These designs’, he went on, in an almost M ­ adisonian fashion,22 can be ‘combatted by temporary combinations, which interpret more faithfully the aspiration of current opinion, the real and not the factious opinion’,23 in short by a referendum. It is difficult to see what has changed.

22 See James Madison’s Federalist Paper No 51, ‘the society itself will be broken into so many parts, interests, and classes of citizens, that the rights of individuals, or of the minority, will be in little danger from interested combinations of the majority.’ James Madison in The Federalist Papers (New York, Signet Books, 1961) 321. 23 Morsei Ostrogorski, Democracy and Political Parties II (London, Macmillan, 1902) 690.

2 The World History of Referendums 1400–2018 Unchanging consistency of standpoint has never been considered a virtue in great statesmen. Cicero1

I. INTRODUCTION

T

he sickly looking cardinal was worried. He looked out of the window of the third floor of Palazzo del Quirinale in Rome and sighed. Carlo Rezzonico had cause to be concerned. It was September in the year 1791. The Pontiff had asked his trusted aide to compose a letter opposing the ‘manifest violation of the law of nations’ resulting from the outbreak of ‘democracy’.2 The Vatican bureaucratcum-cleric was in no doubt as to the graveness of the situation. A week before, the peoples in the Papal States of Avignon and Comtat Venaissin had voted ‘Oui’ to joining France, in a referendum. And then, five days later, on 14 September, the revolutionary French Assemblée nationale had adopted a decree stating, ‘in conformity with the freely and solemnly expressed wish of the majority of the communes and the citizens [of Avignon] are from this moment, an integral part of France’.3 The Cardinal put pen to paper and wrote to the Pontiff, ‘Henceforth, everybody will be able to choose a new master in accordance with one’s

1 MT Cicero, Cicero’s Letters to His Friends vol 1, DR Shackleton Bailey (ed) (London, Penguin, 1978) 67. 2 Pope Pius VI quoted Johannes Mattern, The Employment of the Plebiscite in the Determination of Sovereignty (Baltimore MD, Johns Hopkins University Press, 1921) 58. 3 Archives parlementaires 1791, Series I, Vol xxvi, 362.

12  The World History of Referendums 1400–2018 pleasure. For such would be the consequence of the principle adopted by the National Assembly’. He paused, and then he wrote a single word: ‘absurd’.4 And France was a dangerous case. Indeed, the cleric’s premonition was correct: a year later, ‘a referendum with universal male suffrage approved a constitution’, that emancipated the Jews and even legalised homosexuality.5 The referendums in the Papal States were hardly what modern election observers would call ‘free and fair’. Contemporaneous accounts reported that the voters were ‘extorted under the menace of death’.6 Notwithstanding the, arguably, dubious levels of fairness, the referendums were considered to be legitimate by the people and a growing number of intellectuals.7 If the Cardinal had cared to read it, the German philosopher Immanuel Kant, already considered a giant thinker in his day, had pronounced that ‘the civil constitution of every state should be republican’,8 and the German was not alone in this. The works of Jean-Jacques ­Rousseau and lesser mortals were being distributed all over the Christian world, and even beyond. The Genevan philosopher’s creed, ‘the holders of executive office are not the people’s masters but its officers and the people can appoint them and dismiss them as it pleases’ was accepted by wealthy and well-connected citizens as well as by a fair number of aristocrats.9 This was dangerous for kings, clerics and others who claimed to rule by divine grace. As an historian surveying the period wrote a century and a bit later, The French Revolution proclaimed the dogma that we now term selfdetermination … the mental and logical process was simple. The people are the state and the nation; the people are sovereign. As such they have the right to decide, as the ultima ratio, by popular vote and simple majority, all matters affecting the state and the nation.10 4 Cardinal Rezzonico, cited in F Freudenthal, Die Volksabstimmung bei Gebietsabtretungen und Eroberungen. Eine studie aus dem völkerrecht (Erlangen, Th Blaesing, 1891) 4. 5 Jonathan Fenby, The History of Modern France: From the Revolution to the War with Terror (New York, Simon & Schuster, 2015) 14. 6 Freudenthal, Die Volksabstimmung, n 4 above, 4. 7 See Mattern, The Employment of the Plebiscite, n 2 above, for an overview of the contemporary debate: 58ff. 8 Immanuel Kant, Towards Perpetual Peace (New Haven CT, Yale University Press, 2006) 71. 9 Jean-Jacques Rousseau, ‘Du Contrat Social’ Jean-Jacques Rousseau Oeuvres completes, vol III, Bernard Gangebin and Marcel Raymond (eds), 1st edn (Paris, Gallimard, 1964) 434. 10 Mattern, The Employment of the Plebiscite, n 2 above, 77.

Introduction  13 Referendums were an issue of concern among the rulers of Europe in 1791 – as much as they were 220 years later. It is, in this light, tempting to draw a more or less straight line between these early manifestations of popular sovereignty and the referendums that have caused so much upheaval in the twenty-first century. Yet it is questionable if it is legitimate to do so. Certainly, the referendums in France and adjacent territories in the early 1790s are interesting, as they show that this institution was not just the product of the late twentieth century. And it is true that the referendums held at the time of the French Revolution did inaugurate a new way of thinking overnight. But much as one can go back to the origins, the referendum then was very different from the institution that, over two centuries later, was used to legalise gay marriage in Ireland in 2015 and take the United Kingdom out of the European Union in 2016; though perhaps not so different from the institution that was used to ratify ­President Erdogan’s new constitution in Turkey in 2017. The referendum back then was not born fully formed. Hence an understanding of it must take into account the twists and turns the ­institution has undergone in the course of the centuries. To fully appreciate its evolution we must look at its Wirkungsgeschichte: how it has unfolded, changed and at the same time maintained certain traits throughout its history. Without being too inhibited by the epistemological constraints we might draw on Michel Foucault’s notion of genealogy. The Frenchman – inspired by Friedrich Nietzsche’s Zur Genealogie der Moral – made a distinction between a concept’s Ursprung (origin) and its Herkunft (descent). Whereas the ‘origin’ of a phenomenon is the pursuit of the ‘immobile form’ and the ‘primordial truth’,11 the Herkunft is the study of the ‘myriad of events through which – thanks to which, against which – they [the concepts] are formed’. A genealogical approach to studying the discourses of referendums is not just about the ‘origin’ – ie the first time this type of political consultation was held – but also about the changes it (and the discourse about it) has undergone over a period of ­centuries, in short, its Herkunft. For the historian of ideas, as well as for the political scientist, the aim is to understand the different and sometimes interwoven ways in which thinking about a concept proceeded through different periods of time.

11 Michael Foucault, ‘Nietzsche, Genealogy and History’ in Paul Rabinow (ed), The Foucault Reader: An Introduction to Foucault’s Thought (London, Penguin, 1986) 81.

14  The World History of Referendums 1400–2018 II.  THE EARLIER HISTORY OF THE REFERENDUM

Historians of ideas have sometimes contended that submitting issues directly to the people was first associated with the French Revolution, and that this event marked the ‘first time that the modern notion of a plebiscite or a referendum had been raised’.12 It is certainly true that the institution took on a character that resembles its modern use in the years after fall of the Bastille. And, yet, if one looks carefully, there were earlier uses of this institution. At the time when French intellectuals like the Marquis de Condorcet eulogised the institution (see below), the very term ‘referendum’ had been used for over 100 years and the practice of submitting issues to the people could be traced back further still. In 1684, in the Swiss canton of Graubünden, the Bürger (all male citizens over the age of 16) were given the right to cast their votes on the policy issues that were submitted to them ad referendum by the elected representatives.13 The Swiss might have been the first to use the term but they were not the only pioneers of the practice. Several European countries had already experimented with referendums at this stage. One can always go back to the beginnings, to ancient Greece and Rome. In fourth-century Athens, for example, decrees ‘were passed by majority vote of those Athenians attending the meetings of the Assembly (ekklesia), which were held four times per civil month, or forty times per annum’,14 and this system of direct democratic involvement was also characteristic of the Athenian democracy after the Peloponnesian War, ie in the period between 400 and 320 BC.15 Further, it is possible to argue that the Romans employed a certain kind of direct democracy before the fall of the Republic in 49 BC. And, there are some suggestions that other peoples used what we today may describe as direct democracy. When Niccoló Machiavelli 15 centuries later described the Roman Republic as sustained by ‘the consent of a whole people’ – da uno commune consenseo d’una ­universalità – he was substantially describing a political system based on

12 R Tuck, The sleeping sovereign: The invention of modern democracy (Cambridge, Cambridge University Press, 2016) 143. 13 F Pieth, ‘Das altbündnerische Referendum’ in (1958) (5) Bündner Monatsblatt: Zeitschrift für Geschichte, Landes- und Volkskunde 137–53, 146. 14 Paul Cartledge, Democracy: A Life (Oxford, Oxford University Press, 2016) 210. 15 Mogens Herman Hansen, The Athenian democracy in the age of Demosthenes: ­Structure, Principles, and Ideology, JA Crook (tr) (Norman OK, University of Oklahoma Press, 1991).

The Earlier History of the Referendum  15 referendums; a system that several Italian city states in varying degrees had used in the Middle Ages.16 And yet, describing these early uses of popular involvement as referendums stretches the definition in the direction of meaninglessness. Historically as well as conceptually speaking, it was not until the early sixteenth century that the institution was established in anything ­resembling the present-day referendum with universal suffrage, secret ballot and the like. As was the case centuries later, referendums back then were held less as a result of idealistic concern and more as a consequence of ­Realpolitik. King Francis I of France (1494–1547) had suffered the misfortune to be captured by Spanish troops at the Battle of Pavia in 1525. His release was only secured after Suleiman the Magnificent (the ruler of the Ottoman Empire) put pressure on Charles I of Spain. The Frenchman was not in a position to make demands; it was a condition of his release that he gave up Burgundy. However, once freed the French monarch changed his mind. A bookish ruler with a taste for the arts (it  was he who acquired the Mona Lisa), the King read Erasmus of Rotterdam (1466–1536). The Dutch writer, almost en passant, had made a case for the view ‘what power and sovereignty soever you have, you have it by the consent of the people’.17 That gave Francis an idea. He could submit the transfer of Burgundy to the people. We do not know many details about the conduct of the vote, how ‘the people’ was defined, and, in any case, its fairness is likely to have been somewhat dubious.18 But the net result was a convenient, ‘non’. Francis I’s opportunistic use of Erasmus theory inspired his son Henri II, who held a plebiscite in the three dioceses (Metz, Verdun and Toul) after he had conquered them in 1552.19 Unlike the vote held by his father we know rather more about the latter exercise. Before the referendum in Verdun, Bishop de Lénoncourt is reported to have said, to the inhabitants of said city, ‘that the King of France had come as a liberator

16 Niccolò Machiavelli, Discorsi sopra la prima deca di Tito Livio di Niccolò Machiavelli (Firenze, Felice Le Monnier, 1848) III, 7, and Niccolò Machiavelli, The Discourses (London, Penguin, 1970) 435. 17 Disiderius Erasmus, Erasmus Against War, JW Mackail (ed) (Boston MA, The ­Merrymount Press, 1907 [1517]) 51. 18 Sarah Wambaugh, The Doctrine of National Self-determination: A Study of the Theory and Practice of Plebiscites with a Collection of Official Documents, vol 1 (Oxford, Oxford University, 1919) xxiii. 19 Eugene Solière, Le Plébiscite dans l’annexion. Étude historique et critique de droit des gens (Paris, L Boyer, 1901) 26.

16  The World History of Referendums 1400–2018 who will treat the citizens as good Frenchmen … He appealed to the vote of the people’.20 It is remarkable – and possibly a result of Solière’s enthusiastic reporting of the vote that Bishop Lénoncourt used words such as ‘bourgeois’ and ‘peuple’ at a time when Jean Bodin (1530–96) expounded his theory of divinely sanctioned absolutism by the grace of God in Six livres de la République (1576). History is not always linear and it is certainly the case that the use of referendums at this stage was far from being an indication of a spirit of democracy among those who initiated the votes. It is difficult to disagree with the conclusion that, We find in France in the sixteenth century a policy of opportunism which recognised, or even insisted upon, the principle of popular self-determination in the transfer of cities and territories if such self-assertion was favourable or could be forced into an expression favourable to France, but which refused to acknowledge any voice or opinion to those who wanted to conquer against their will, or to any section of the Kingdom which for some reason or other might wish to sever its former or forced connection to France.21

Yet for all the cynical and opportunistic use of the referendum, the emerging practice spurred thinking in the minds of theorists, which, little by little, gave birth to the expectation that the citizens were to be involved in matters of great importance. At this stage the referendum was  – apart from in Switzerland – almost exclusively used to settle territorial disputes. Hence the right to be consulted in popular votes was forged by international lawyers and not by political theorists and philosophers. In his magisterial The Law of War and Peace, Hugo Grotius (1583–1645) observed that ‘in the alienation of a part of sovereignty, it is required that the part which is alienated consent to the act’.22 Soon thereafter, Samuel Pufendorf (1632–94), another distinguished legal scholar, was even more explicit when he wrote in De jura n ­ aturae et gentium (1672) that ‘in the alienation of a part of the kingdom, there is required … the consent of the people’.23 And, roughly 100 years later, Emer de Vattel (1714–67) cited the aforementioned example of a vote held in Burgundy in 1527 in which the citizens had objected to a plan to transfer them to the Spanish King. Though Vattel added – rather

20 Solière, Le Plébiscite dans l’annexion, ibid. 21 Mattern, Employment of the Plebiscite, n 2 above, 53. 22 Hugo Grotius, The Rights of War and Peace, K Haakonsen (ed) (Indianapolis IN, Liberty Fund, 2005 [1625]) 570. 23 S Pufendorf, Of the law of nature and nations: eight books (London, J Walthoe, 1729 [1672]) 59.

Democracy in America 1776–87  17 soberly – ‘Subjects are seldom able to make resistance on such occasions; and, in general, their wisest plan will be to submit to their new master, and endeavour to obtain the best terms they can’.24 But this was very much a theory. The political order of the day was enlightened absolutism, or more often, just plain dictatorial rule. This changed with the American and the French revolutions. III.  DEMOCRACY IN AMERICA 1776–87

Even before the American Revolution, the citizens in the Thirteen Colonies were unusually involved in political action and had considerable say over public policy. As the Massachusetts Charter from 1691 stated, ‘councillors … may at any time be removed’.25 In short, the people were the masters and the representatives their servants. As a result of this, Alexis de Tocqueville later reflected: ‘democracy gradually permeated their customs, opinions and social habits’ and that the Americans had succeeded because ‘the practical education of the nation [had] reached the height of perfection’.26 The experience in America, as the Frenchman continued, provided evidence that, the most powerful and perhaps the only means that we possess of interesting men in the welfare of their country is to make them partakers in ­government … civic zeal seems to be inseparable from the exercise of political right.27

But this principled commitment to popular sovereignty did not lead directly to the widespread use of the referendum. Admittedly, James Madison noted, ‘the people is the fountain of all power’.28 And for this reason the constitution had to be ratified by them. The natural consequence of this commitment to ‘the people’ would have been a referendum to ratify the US Constitution. But Madison conveniently 24 E Vattel, Le droit des gens: ou, Principes de la loi naturelle, appliqués à la conduite & aux affaires des nations & des souverains, Vol 1 (Paris, Aux depens de la Compagnie, 1758) 264. 25 http://avalon.law.yale.edu/17th_century/mass07.asp (Accessed 18th November, 2018. 26 Alexis de Tocqueville, Democracy in America (London, Penguin, 2003) 276. 27 Tocqueville, Democracy in America, n 26 above, 252. 28 Madison, James (1977) [1787] ‘Method for Ratifying the Constitution’ in Robert A Rutland, Charles F Hobson, William ME Rachal and Frederika J Teute (eds), The Papers of James Madison, vol 10, 27 May 1787–3 March 1788 (Chicago IL, The University of Chicago Press, 1977) 27.

18  The World History of Referendums 1400–2018 ignored this option. He did so for a reason. Rhode Island – a State that had not sent delegates to the Constitutional Convention in Philadelphia in 1787  – had, in fact, held a referendum; and the voters rejected the Constitution by a margin of ten to one.29 This inconvenient truth was largely ignored by the federalists – and by the contemporary media. The latter, luckily for Hamilton, Madison and their cabal, supported the new Constitution, ‘Eighty out of the nation’s ninety-two n ­ ewspapers tilted towards the Federalists’.30 Like kings and emperors using the referendum for their own political purposes the defenders of representative democracy of the early American Republic, were willing to use the means available to win approval for the Constitution. They were engaged in Realpolitik. In most of the States, Madison and Hamilton’s allies had organisational advantages that went with controlling the political system – if not the majority of the legislatures. The true aim of the ratifying assemblies was, as a little known federalist noted in an unguarded moment, ‘to pack convention[s] whose sense would be different from the people’.31 Throughout the country, the federalists used all available means – not all of them democratic or legal, and certainly not ethical – to win support for the Constitution. All is fair in love and war – and when ratifying national constitutions. In Pennsylvania, they organised an election in haste, in order to ensure that only a small minority of well-educated (and wealthy) voters turned out to vote. (Only 18 per cent of the eligible electors voted in the election).32 And when two anti-federalists decided to boycott the proceedings, thereby ensuring it didn’t have quorum, the two were forcibly dragged into the chamber and a vote was taken; ‘bad measures’, as one observed, ‘in a good cause’.33 In other words, the referendum was not a prominent part of the ratification of the US Constitution but it was widely discussed and dismissed as a mechanism for letting populists rule. The Founding Fathers distrusted direct democracy, having seen the risks and dangers of submitting issues to voters who did not necessarily share their vision and goals. Those lamenting the use of referendums in the twenty-first century can find many of the arguments against this ­institution in the 29 See: IH Polishook, Rhode Island and the Union, 1774–1795 (Evanston IL, N ­ orthwestern University Press, 1969). 30 A Taylor, American Revolutions: A Continental History, 1750–1804 (New York, WW Norton & Company, 2016) 389. 31 George Minot, quoted in Jackson T Main, The Antifederalists: Critics of the Constitution 1781–1788 (New York, WW Norton, 1961) 202. 32 Taylor, American Revolutions, n 30 above, 389. 33 Minot quoted in Taylor, American Revolutions, n 30 above, 390.

The Referendum 1789–1920  19 writings of the Founding Fathers. Equally, those who decry the rule of the elites and their failure to heed the calls of ‘the people’ will find that established politicians in America in the 1780s did not differ much from those currently in Congress and State Houses across the USA at the present time. It was a very different story in France. There the referendum played a central role from early on – and it did so for better or for worse. And, once again, there are parallels with how referendums were later used and abused by dictators. IV.  THE REFERENDUM 1789–1920

The French Revolution heralded a new era of democracy and the new rulers in Paris enthusiastically embraced referendums. Indeed, no less a theoretician than Baron de Condorcet (1743–94), published a pamphlet in 1789 with the telling title Sur la nécessité de faire ratifier la constitution par les citoyens – roughly translated, ‘on the necessity of the people ratifying the constitution’.34 At this stage this was not mere idle talk. As we saw in Cardinal Rezzonico’s letter to the Pope, France’s annexation of Avignon in 1791 only took effect after a referendum had been held in the area. A contemporary report read: Considering that the majority of the communes and citizens have expressed freely and solemnly their wish for a union with Avignon and France … the National Assembly declares that in conformity with the freely expressed wish of the majority … of these two countries to be incorporated into France.35

Before the French Revolution degenerated into the reign of terror, the voters had voted to ratify the Constitution in the first ever nationwide referendum in 1793 and again in 1795 and 1799.36 The First Constitution envisaged widespread use of referendums as a check on the executive, and even the use of ‘the popular initiative in constitutional and legislative matters’.37 After the fall of Robespierre, the 1793 Constitution was 34 Nicolas de Condorcet, Sur la nécessité de faire ratifier la constitution par les citoyens et sur la formation des communautés de champagne (Paris, Nabu Press, 2012 [1793]). 35 Cited in GF von Martens, Recueil de Principaux traits d’alliance de paix (Göttingen, JC Dieterich, 1801) 401. 36 Maurice Duverger, Le system politique français (Paris, PUF, 1996) 252. 37 L Morel, ‘France: towards a less controversial use of the referendum?’ in Michael Gallagher and Pier Vincenzo Uleri (eds), The referendum experience in Europe (London, Palgrave Macmillan, 1996) 67.

20  The World History of Referendums 1400–2018 seen as too democratic and replaced by the 1795 Constitution. However, at this moment ‘the legitimacy of the constitutional referendum was ­instituted’,38 and it was deemed politically necessary to submit the text to the people for ratification. While this Constitution was soon superseded by the 1799 one, that too was submitted to a vote by le peuple. Given the importance of acquiring popular approval, it was not surprising that Napoleon Bonaparte used the institution. Having come to power in a (relatively bloodless) coup, the diminutive Corsican officer submitted several issues pertaining to himself to plebiscites: first on whether he should be Consul (1800), then on whether he should be Consul for life (1802) and on hereditary rule for the Napoleon family in 1804.39 Why Bonaparte (and other autocratic rulers) held (and hold) votes – the outcome of which was as predictable as their conduct was unfair – is a question we shall return to in the next essay. Table 1  Referendums in France 1793–1870 Year

Yes

Turnout

1793

Constitution

Issue

70.0

44.4

1795

Constitution

99.9

26.4

1799

Constitution

63.8

3.8

1800

Napoleon as Consul

99.9

43.6

1802

Napoleon as Consul for Life

99.8

51.2

1804

Heredity rule for Bonaparte family

99.9

43.3

1815

Restore imperial constitution

99.7

18.8

1851

Powers to Napoleon III (10-year presidency)

92.1

79.9

1852

Louis Napoleon as Emperor

96.7

79.9

1870

Parliamentary Rule

83.1

83.5

Source: Maurice Duverger, Le system politique français (Paris, PUF, 1996).

Napoleon’s commitment to the will of the people even extended to conquered ones, such as Switzerland: In June 1802 the Swiss people as a whole voted yes or no for the first time on the text of the Second Helvetic Constitution. It was clearly announced before the referendum that abstentions would be considered as ­affirmatives  …



38 Morel,

ibid. Le system politique français, n 36 above, 252.

39 Duverger,

The Referendum 1789–1920  21 The  Constitution was accepted with 92.500 votes against and 72.500 in favour because there were 167.000 abstentions.40

Though it should be added, Napoleon – realising the opposition to the Constitution – unilaterally scrapped it and replaced it with another less centralised one – and did so without submitting the document to a ­referendum. There were, after all, limits to direct democracy. The Emperor also used referendums when it was opportune. Such an occasion arose when Bonaparte returned from Elba and submitted the constitution penned by the philosopher Benjamin Constant (the document was also known as la Benjamine) to a popular vote. However, this constitution was short-lived and hardly came into force before Napoleon was defeated at Waterloo. It is trite to note that the Congress of Vienna dealt a blow to the doctrine of selfdetermination – and, as a consequence, the use of referendums; ‘The Congress of Vienna in 1815 did not accept self-determination as a basis for reshaping the map of Europe’.41 The victors in the Napoleonic Wars were ­conservatives who wanted to return to a time when the popular sovereignty was not the gold standard of political legitimacy. The excesses of revolutionary fervour and the horrors of the Napoleonic Wars gave democracy a bad name. This changed after the revolutionary year of 1848 when referendums once again became ­fashionable, especially in Italy, Switzerland and, once again, in France. Napoleon III, the nephew of the former emperor, seized power in a coup after he had been elected in 1850. Mirroring his uncle, ‘the little Napoleon’, as Victor Hugo called him,42 used dubious plebiscites to claim popular legitimacy for granting himself executive powers in 1851 and in the following year for becoming emperor. In this case, there was a conflict between the newly-elected president and the majority in the Assembly. The former decided to resolve the deadlock thorough a seemingly democratic mechanism: a referendum. This was a shrewd tactic, which worked. But it was in direct violation of the rule of law. The French plebiscites held by Napoleon III were predictable and indicative of the very French tradition of voting on issues that were a fait accompli. Matters were little different elsewhere. Generally the 40 Jean-François Aubert, ‘Switzerland’ in David Butler and Austin Ranney (eds), Referendums: A Comparative Study of Practice and Theory (Washington DC, The American Enterprise Institute, 1978) 39. 41 M Griffiths, ‘Self-determination, international society and world order’ (2003) 3(1) Macquarie Law Journal 29–49, 38. 42 V Hugo, Napoleon le petit (Paris, Hetzel et Cie, 1851).

22  The World History of Referendums 1400–2018 r­ eferendum has been held in order to give a dubious seal of democratic approval to matters that have been decided beforehand by the political elites. Opportunistic political leaders used the referendum as a strategic device. This was true in both Italy (where several referendums were held in the name of self-determination as a part of the process to unify the country) and in Schleswig-Holstein (between present-day Denmark and Germany) where a referendum was proposed – but not held – over the fate of the province. The Risorgimento referendums were held to put pressure on the great powers that were reluctant to change the status quo. In a series of votes held between 1848 and 1870 different parts of Italy voted to join the new unified state under the constitutional monarch Victor Emmanuel of Sardinia. The leading Italian statesman at the time Camillo Benso di Cavour (1810–61) expressed the consensus among those advocating the use of referendums at this time in a letter before the referendum in Tuscany and Emilia in 1860, in which he wrote, I await with anxiety the result of the count, which is taking place in Central Italy. If, as I hope, this last proof is decisive [questa ultima prova], we have written a marvellous page in the history of Italy. Even should Prussia and Russia contest the legal value of universal suffrage, they cannot place in doubt [non potranno mettere in dubbio] the immense importance of the event today brought to pass. Dukes, archdukes and grand-dukes will be buried forever beneath the heap of votes deposited in urns of voting places of Tuscany and Emilia.43

Cavour was perhaps correct in expressing doubt about the sincerity of the commitment on the part of more autocratic powers such as Prussia and Russia, yet even these countries were surprisingly positive towards referendums on self-determination in the 1850s and 1860s. As Sarah Wambaugh noted, ‘There was not one of the great powers, not even Austria or Russia, which did not participate in those years [1848–70] in some form of appeal to national self-determination to settle Europe’s numerous territorial questions’.44 Britain’s half-hearted mediation between Denmark and P ­russia following the early part of the First Schleswig War in 1848–51 is a case

43 Camillo Benso, Conte di Cavour, ‘Letter to Villamarina di, Minister of Sardinia at Naples, March 1860’ in Lettere edite ed inedite di Camillo Cavour, Chiala Luigi (ed) (Torino, Roux, 1883) vol 3, 211. 44 Wambaugh, The Doctrine of National Self-determination, n 18 above, xxxii.

The Referendum 1789–1920  23 in point. Lord Palmerstone (the British Foreign Secretary 1846–51) suggested to Christian von Bunsen, the Prussian ambassador in London, that the dispute should be decided ‘with reference to the ascertainable facts’, and that these could only be found through a referendum.45 The Prussian diplomat responded, Germany [It was technically Prussia, as Germany was not unified until 1870!] cannot give up the principle declared on all occasions that no separation of any part of Schleswig can ever be thought of, unless the population in the northern districts themselves declare, by an open and unbiased manifestation of their intention to that effect.46

The proposal was, however, rejected by the Danes, who militarily had the upper hand. This changed 15 years later. Under the pretext of a clumsy decision by Copenhagen to centralise powers, Otto von Bismarck used this ­thinnest of pretexts to declare war on the Danes in order to liberate the ­German-speaking minority in the small Nordic kingdom. The Danes were no match for the highly skilled and effective Prussian war machine, especially as Berlin was supported by Austria-Hungary. In order to maintain the balance of powers among European powers, Britain sought to mediate between the warring parties. In 1864, during an armistice, the Prussian Foreign Minister, Peter Graff von Bernsdorff, maintained at the London Conference that he was guided by the c­ onviction that the conference should be aware of the wishes of the people whose future they were debating and that ‘the inhabitants of Schleswig should be consulted on the subject’.47 The Danes erroneously believed that London would reject Prussian annexation and come to their rescue. Hence, they held tight and rejected the proposal. This was a mistake. The British were not in a position to stop Bismarck militarily and did not come to their rescue. The Prussian army marched on and the Danes lost. While, the Treaty of Prague, 1866, which formally ended hostilities did make annexation conditional upon the consent of the people, Prussia (having realised opposition to its rule in the Danish-speaking parts) annexed Schleswig-Holstein in toto without a referendum. Once again pragmatism had triumphed over idealism. The referendum on self-determination played a very minor role in the years following 45 Palmerstone to von Bunsen, 24 June 1848, British and Foreign State Papers, vol 40, 1321. 46 Graff von Bunsen to Palmerstone, 24 June 1848, ibid. 47 Bernsdorff – in Conference of London, Protocol No 10, 1864.

24  The World History of Referendums 1400–2018 the Franco-German War. Interestingly, given that the referendum was often used in an opportunistic way, leading German lawyers now rejected the use of referendums whereas French international lawyers and intellectuals rediscovered the attractions of letting the people decide. Political convenience has a tendency to get dressed up as philosophical principle. What is known as ‘spin’ in the 21st century was not unknown in the 1870s. V.  THE REFERENDUM IN SWITZERLAND IN THE NINETEENTH CENTURY

Meanwhile the referendum was gaining prominence in Switzerland, and not just in the rather arcane form of Landesgemeinde (public open-air meetings where male citizens voted on local laws). To be sure, the Swiss pioneered the referendum and invented the term itself in the seventeenth century. Yet, while the institution could be traced back hundreds of years, the modern use of the referendum was a result of a compromise that emerged in the period after the Sonderbund War (1848). The Radical Party (from 1894 renamed the Free Democratic Party) was the dominant force in Swiss politics immediately after the civil war. The party pursued a twin-track of muscular secularism and free-market, laissez-faire liberalism. Under the majoritarian (first-pastthe-post (FPTP)) electoral system the party was able to win a majority of the seats in the Bundesrat (the executive) without winning a majority of the votes. Realising that the Radical majority in the Bundesrat did not always have the support of the voters at large, and that Catholics and Protestants shared some of the same interests, their theological disagreements notwithstanding, the two religious groups began to push for the introduction of a popular veto; a referendum in which the voters could vote on legislation already passed. The Radicals, for their part, wanted to strengthen their stranglehold on power. After a constitutional reform was rejected in 1872, the Radicals accepted that the popular referendum would be introduced in return for federal control over legislation (something the Catholics opposed). It was not anticipated that the referendum would be widely used, as it would require the collaboration between Catholics and Protestants against the secular Radicals.

The Referendums After the First World War  25 But – contrary to the Radicals’ predictions – the two Churches were able to cooperate, their theological disagreements notwithstanding. Subsequent historians have pondered why the Radicals accepted the referendum. It was probably miscalculation. The changes resulted in a number of important changes that challenged the Radicals’ virtual monopoly on legislative power in the period before the First World War. For example, the rejection of a law on the establishment of a federal ministry of education in 1884 (the very thing the Radicals had wanted), the introduction of the constitutional initiative in 1891 and the rejection of a more liberal temperance law in 1903, are all examples of how non-liberal groups prevented radical legislation. Through the referendum, parties in opposition were able to shape public policy. In the view of one observer, ‘strong political minorities were able to threaten and mobilize for an activation of the optional referendum, until they were eventually co-opted into the government’.48 This tendency became stronger after Switzerland became a multi-party system. But at this stage – as well as today – Switzerland is sui generis – a special case of direct democracy; in the local vernacular, a Sonderfall. VI.  THE REFERENDUMS AFTER THE FIRST WORLD WAR

In the wake of the First World War – at the behest of the American President, Woodrow Wilson – eight referendums were held to determine the borders in Europe. Wilson’s commitment to self-determination was not – in seems – only a result of a study of the European doctrines espoused in the wake of the French Revolution, still less the ideals of the Italian Risorgimento or the doctrines of Grotius and Pufendorff. Rather, Wilson’s commitment was also inspired by his early years as a campaigner for more direct democracy. Earlier in his career, Wilson had stated his commitment to direct democracy in domestic politics. He had noted that, It must be remembered that we are contrasting the operation of … the referendum, not with representative government, which we possess in theory … but with the actual state of affairs, with legislative processes which are

48 U Serdült, ‘Referendums in Switzerland’ in M Qvortrup (ed), Referendums around the World. The Continued Growth of Direct Democracy (London, Palgrave Macmillan, 2014) 65–121, 85.

26  The World History of Referendums 1400–2018 carried out in secret, responding to subsidized machines, and carried through by men whose happiness it is to realize that they are not their own masters but puppets of the game.49

These ideals – so it seems – inspired the President in his espousal of national self-determination. Wilson did not – as is commonly assumed – mention referendums in his famous Fourteen Points speech to Congress on 8 January 1918. However, it is clear from the context that the 28th President wanted the decisions regarding the national borders to be taken by the peoples concerned through plebiscites. As Wilson noted, Peoples may now be dominated and governed only by their own consent. Selfdetermination is not a mere phrase. It is an imperative principle of action, which statesmen will henceforth ignore at their peril. The settlement of every question, whether of territory, of sovereignty, of economic arrangement, or of political relationship [must be] upon the basis of the free acceptance of that settlement by the people immediately concerned, and not upon the basis of the material interest or advantage of any other nation which may desire a different settlement for the sake of its own exterior influence or mastery.50

For all his idealism Wilson was not always true to his word. Indeed, a referendum organised by the council in the Tyrol was ignored – at the insistence of the French – despite the fact that more than 90 per cent voted for union with Germany. Not all the votes resolved the matters. However, it is worth noting that ‘It was precisely in those areas where plebiscites were refused (with the exception of Alsace-Lorraine) – Danzig, the Polish corridor and the Sudetenland – that were the subject of revisionist claims by the Nazis in the 1930s’.51 Tellingly, German revisionist claims were not made in areas that were ceded after a referendum, such as Nord Schleswig where there was a large German-speaking minority. This is possibly because ‘frontiers that were fixed by plebiscite could not easily be undermined’.52

49 Wilson quoted in William Munro, The Initiative, The Referendum and the Recall (New York, D Appleton, 1912) 87. 50 Wilson quoted in Lawrence T Farley, Plebiscites and Sovereignty: The Crisis of Political Legitimacy (Boulder CO, Westview Press, 1986) 3. 51 Vernon Bogdanor, ‘Referendums and Separatism II’ in Austin Ranney (ed), The Referendum Device (Washington DC, American Enterprise Institute of Public Policy Research, 1981) 145. 52 Bogdanor, ‘Referendums and Separatism II’, ibid, 145.

The Referendum in the 20th Century  27 VII.  THE REFERENDUM IN THE 20TH CENTURY

In the twentieth century, the use of the referendum took on somewhat different forms. In the United States so-called Populists in the 1890s had proposed and ensured the adoption of initiatives (referendums on laws proposed by the people) and referendums (votes on laws enacted by the legislature). In Europe – on the continent – the referendum eventually degenerated into abuse by Hitler and other dictators) and in the United Kingdom, the referendum was much theorised and little used. To appreciate this use of the referendum, it can be useful to use a distinction often found in French political and legal research, namely between referendums and plebiscites. According to Guillaume-Hofnung, ‘While the referendum allows the people to express themselves freely, plebiscite allows a person … to legitimize himself or herself by posing a question to the people who answered in the affirmative’.53 In Britain and Australia conservatives wanted to use the referendum as a brake on change; one that could hinder the Government’s sins of commission and not one that would rectify the Executive’s sins of omission (we shall return to this in the essays below). In Germany and other countries with a less than impeccable record of democratic pluralism, the involvement of the people was to be limited to plebiscites. What is interesting, though, was the appeal to and recognition of the ‘people’ even by those who sympathised with and actively espoused dictatorship. Thus, the leading scholar – and later Nazisympathiser – Carl Schmitt recognised that in the present day no one ‘would remain on the throne against the will of the people’.54 As a result of this, ‘institutions of direct democracy’ would always be ‘in a position superior to the so-called indirect democracy of the parliamentary state’.55 However, Schmitt pointed out, the voters were not in a position to pose the questions; ‘the people can only respond yes or no. They cannot advise, deliberate and discuss’.56 Hence, ‘plebiscitary legitimacy requires a government or some other authoritarian organ in which one will have confidence that it will pose the correct question’.57 53 Michèle Guillaume-Hofnung, Le référendum (Que sais-je?) (Paris, Presses universitaires de France, 1987) 14. 54 Carl Schmitt, The Crisis of Parliamentary Democracy, Ellen Kennedy (tr) (Cambridge MA, MIT Press, 1988); Carl Schmitt, Legality and Legitimacy, J Seitzer (tr) (Durham NC, Duke University Press, 2004 [1932]) 29. 55 Schmitt, Legality and Legitimacy, ibid, 60. 56 Schmitt, Legality and Legitimacy, ibid, 93. 57 Schmitt, Legality and Legitimacy, ibid, 90.

28  The World History of Referendums 1400–2018 Schmitt’s theory was adopted by the National Socialists, who submitted issues to the voters in order to acquire the ‘plebiscitary legitemacy’ of a ‘decision through one will’.58 The German voters thus voted to leave the League of Nations (1933), to make Hitler Führer (1934) and for the incorporation of the Saar into Germany in 1935. It was arguably this abuse of the referendum which prompted Clement Attlee to reject Winston Churchill’s proposal for a referendum on the postponement of the 1945 General Election. Attlee said: I could not consent to the introduction into our national life of a device so alien to our traditions as the referendum, which has only too often been the instrument of Nazism and Fascism. … Hitler’s practices in the field of referenda and plebiscites can hardly have endeared these expedients to the British heart.59

That these plebiscites were neither free nor fair somewhat undermines Attlee’s argument. VIII.  THE USE OF REFERENDUMS AFTER 1945

Immediately after the Second World War several countries needed to establish democratic regimes. Unlike Germany and Japan – who had constitutions imposed upon them by occupying forces – other ­countries faced a crisis of legitimacy. In France, Italy and Belgium – to name but three examples – large sections of the elites had been uncomfortably close to the previous collaborator regimes. In all three countries, referendums were held to deal with the past and to open a new chapter. Italy is a particularly interesting example. There, the King (King Victor Emanuel  III) had worked closely with Mussolini and although he had helped overthrow Il Duce he was viewed with suspicion among large sections of society. The first Post-War Italian Prime Minister, Alcide De  Gasperi, a Christian Democrat, acted resolutely. In the words of one observer, De Gasperi’s first real political masterpiece was the decision to refer the choice regarding the state’s institutional form, republic or monarchic, to a popular referendum. In 1945, following a crisis among the member parties of a coalition linked to the National Liberation Committee De Gasperi took 58 Schmitt, Legality and Legitimacy. ibid, 92. 59 Attlee’s letter to Churchill, quoted in VB Bogdanor, The People and the Party System: The Referendum and Electoral Reform in British Politics (Cambridge, Cambridge University Press, 1981) 35.

The Referendum in the Age of Dealignment  29 on a position of high rank … Very few realised they had put power into the hands of a man endowed with great political capabilities, who would be present throughout the process of reconstruction.60

The referendum was used to break a deadlock. De Gasperi’s resolute decision to let the people decide was bold and efficient, even though the margin of victory was less than resounding: 54–46 in favour of abolishing the monarchy. It was observed that the vote ‘revealed a dramatic split between North and South’61 and that ‘there followed a number of days of high tension in the Capital, with rumours of a possible coup in support of the King. But De Gasperi and other ministers remained firm’ and shortly thereafter the King flew into exile.62 This pattern would repeat itself many times subsequently, namely a narrow victory settled the matter. Other referendums, however, showed that even majority support is not always enough. The Belgian King Leopold III – who to be very diplomatic – equivocated in his stance towards the Nazi occupiers. He had left the country – on the orders of Heinrich Himmler – and before he returned a referendum was held. Although a majority of 57 per cent voted for his return in 1950, the referendum split the country. There were four deaths and a general strike. In the end, the monarch abdicated in favour of his son Baudouin. But overall there were relatively few referendums in the 1950s and in the decade after that. Admittedly, the Swedes voted on three alternative national pension systems in 1957 and on whether to drive on the right in 1955, and in France, Charles de Gaulle used the referendum in a somewhat plebiscitary fashion to win approval for, respectively, the introduction of the Fifth Republic (1958), withdrawal from Algeria (1961), and the direct election of the Executive (1962). (De Gaulle resigned when he failed to win a referendum on Senate reform in 1969). But these were the exceptions. IX.  THE REFERENDUM IN THE AGE OF DEALIGNMENT

It all began to change around 1980. In this decade, several referendums were held on ‘new’ issues. The Swedes and the Austrians voted on nuclear power in, respectively, 1979 and 1980. There was a sense that the 60 P Pombeni, ‘Christian Democracy in Power, 1946–63’ in Erik Jones and Gianfranco Pasquino (eds), The Oxford Handbook of Italian Politics (Oxford, Oxford University Press, 2015) 256. 61 Paul Ginsburg, A History of Contemporary Italy (London, Penguin, 1990) 98. 62 Paul Ginsburg, A History of Contemporary Italy (London, Penguin, 1990) 99.

30  The World History of Referendums 1400–2018 r­eferendum was being used to resolve issues that cut across traditional political cleavages. The political parties had failed to represent the electors as well as they had done before. Referendums became a ‘legitimization tool for constitutional changes that occasionally serve as a synchronization mechanism between ­politicians and citizens’.63 The same explanation held at the local level where local referendums contributed ‘an additional qualitative dimension to the public space of local politics by opening new channels for public deliberations on issues’.64 To understand the apparent change towards a greater use of direct democracy a bit of historical context might be useful. In the middle of the twentieth century, political theorists were sceptical of the people. The Austrian theoretician Joseph Schumpeter summed up the general consensus when he wrote, in Capitalism, Socialism and Democracy, [Democracy does] not mean and cannot mean that the people actually rule in any obvious sense of the terms ‘people’ and ‘rule’. Democracy means only that the people have the opportunity of accepting or refusing the men [!] who are to rule them. But since they might decide this also in entirely undemocratic ways, we have had to narrow our definition by adding a further criterion identifying the democratic method, viz., free competition among would-be leaders for the vote of the electorate.65

Even the most extensive use of referendums does not change the validity of Schumpeter’s conclusion. But gradually democracy began to go beyond ‘accepting and refusing men and women. More and more referendums were held and countries that had held few or no referendums before (such as the Netherlands, Luxembourg and Iceland) began to submit issues to referendums. Referendums were ‘used twice as frequently today compared with fifty years ago and almost four times more than at the turn of the twentieth century’.66 There is no simple and all-encompassing explanation for this trend. The decision to call a referendum – like all other human actions – is always shaped by unique circumstances and in response to context-specific challenges. But much as each referendum has its own unique logic, there seemed to be a tendency; a recurrent pattern across much of the W ­ estern

63 David Altman, Direct Democracy Worldwide (Cambridge, Cambridge University Press, 2011) 197. 64 Theo Schiller, Local Direct Democracy in Europe (Wiesbaden, VS-Verlag, 2011) 69. 65 J Schumpeter, Capitalism, Socialism and Democracy (London, Routledge, 1942) 242. 66 Altman, Direct Democracy Worldwide, n 63 above, 65.

The Referendum in the Age of Dealignment  31 World. Without claiming to have discovered a universal pattern, the referendum filled the vacuum left by the political parties when they no longer represented societal and class interests. Previously, specific groups had represented distinct factions of society, Labour parties representing the workers, Christian parties representing middle-class religious people, and so on. Figure 1  Referendums in Democratic Countries: Constitutional, Ad Hoc and Citizens’ Initiatives 1900–2016

Source: Aarau Centre for Research on Direct Democracy (C2D) and Qvortrup, Referendums Around the World.

But these ties began to break down in a process sometimes referred to as dealignment.67 The frozen party-system began to thaw at a time when the number of party-identifiers began to drop. Many people felt political parties were not willing and able to represent them. The views articulated by minority groups (especially on the ‘New Left’) failed to be addressed by the established political parties, and at the same time, voters on the centre-right felt that the traditional bourgeoisie parties were unresponsive to views of the New Right. The tendency can be clearly seen in Figure 1.

67 M Qvortrup, Government by Referendum (Manchester, Manchester University Press, 2018) 17.

32  The World History of Referendums 1400–2018 As Figure 1 also shows, there was an increase in all forms of referendums, both the ones held for political reasons (also known as ‘ad hoc referendums’) and the ones that were mandated by constitutional requirements. At the same time there was also an increase in the number of citizen-initiated referendums, though the single example of Italy is responsible for most of these. Some might see this increase as an indication of a crisis of representative democracy. This might be a hasty conclusion (though one we shall return to below). In fact, research seems to suggest that the voters were still broadly in agreement with the dominant political parties. Søren Holmberg thus found congruence between the voter’s preferences and the policy positions of the representatives in 79 per cent of the cases.68 However, there was disagreement in the remaining 21 per cent of the cases. This analysis is not only plausible in Western Europe and other developed nations; it is also corroborated by research in younger democracies, eg in Latin America, Because institutionalized party systems may … become overinstitutionalized, they have serious dilemmas for channelling social demands, they lack the required flexibility to do so, and ultimately they will be subject to massive demands for movement towards citizens’ preferences.69

Dealignment might also explain the growing use of mechanisms of direct democracy in the United States, where the initiative (legislation initiated by citizens and voted on by the whole electorate) was used to spearhead the ‘Reagan revolution’. (The so-called ‘Proposition 13’ in California is a famous example of how the ‘people’ by-passed a recalcitrant political elite to demand tax-cuts). But the story was, superficially, a different one in the former authoritarian countries, many of which held referendums after end of the Cold War and after the fall of right-wing authoritarian regimes in Latin America and Southern Europe. Yet, on closer inspection, these referendums, too, conformed to the pattern. In the absence of established political parties, a mechanism was needed to confer legitimacy upon the newly-enacted constitutions in former autocratic regimes (whether Communist or not); thus the referendums after the fall of dictatorships in Spain and Brazil – and in Russia, Hungary, and practically all other

68 Sören Holmberg, ‘Issue Agreement’ in P Eliasson and K Heidar (eds), Beyond Westminster and Congress: The Nordic Experience (Columbus OH, University of Ohio State University Press, 2000) 155. 69 Altman, Direct Democracy Worldwide, n 63 above, 197.

Conclusion  33 former Communist countries – are examples of how the referendum was used as an ‘alternative’ channel of legitimacy. Once the party-systems were established in these countries the number of referendums dropped slightly. X. CONCLUSION

The referendum has a long history. While the referendum in the form we know today originates in the eighteenth century, the institution can be traced back to the early part of the sixteenth century when France used the referendum to gain legitimacy for its foreign policy decisions. The referendum began to be used in earnest in the nineteenth century and grew in prominence around the turn of the century where countries introduced referendums to resolve complex political problems. In the decades following immediately after the Second World War, the referendum was little used – though votes were held in France, Italy and Belgium. This changed from the 1980s onwards. The reasons for this are manifold and cannot be reduced to a simple formula. But it could be conjectured that the referendum and its increased use are due to a failure of political parties to represent the voters. To relieve the political system from demands, the referendum and other forms of direct democracy have served ‘as institutionalized, sporadic safety valves of political pressure’.70 By developing mechanisms for letting off political steam in the form of referendums, the political systems seem to have become more legitimate. And, it appears – though hard evidence is difficult to come by – that countries with more referendums have suffered lower levels of political distrust in the political elites; ‘giving people more voice is widely considered a promising remedy against the current crisis of democracy’.71

70 Altman, Direct Democracy Worldwide, n 63 above, 198. 71 Laurent Bernhard, Campaign Strategy in Direct Democracy (Basingstoke, Palgrave, 2012) 199.

3 Plebiscites in Dictatorships – Voting as a Signifiant of Total Control

S

ometimes dictators submit issues to a vote in what appears to be a referendum. Yet, the outcomes of these ‘plebiscites’ are never in doubt. The dictators’ policies are always approved by the voters. So why have them at all? That is what this essay is about. But, first a story from real life … In 1993, a small delegation of electoral observers in South Ossetia (formally a part of Georgia) stopped. One of the party – to put it delicately – found it difficult to digest the local food. So, a toilet break was needed. In fact, several! The observer – an American who shall remain nameless – was rather surprised when the toilet paper was a stack of ballot-papers, all of them with the crosses next to ‘da’ (Russian for yes). Full of indignation, and for a moment forgetting about his digestive issue – he asked the interpreter: ‘What is this? These are ballot-papers? They are already marked?’ ‘Yes’, responded the interpreter in a thick Russian accent, ‘they had printed too many. So they use them as toilet paper. That’s what we’ve always done’.1 The story is not unique. On 24 October 1955 Jean Baptiste Ngô Đình Diệm, the first President of South Vietnam, hailed the result of the plebiscite of the previous day. ‘The plebiscite of which [the people of South Vietnam] took such an enthusiastic part’, he stated, ‘constitutes an approval of the policies pursued.’2 The ‘approval’ was echoed by the US State Department, which issued a statement saying that, ‘The Department of State is gratified that according to reports the referendum was conducted in such an orderly and efficient manner and that the people of Viet-Nam have made their 1 I am grateful to Mr Klaus Rasmussen, formerly of the OSCE for sharing this story with me. 2 JM Chapman, ‘Staging Democracy: South Vietnam’s 1955 Referendum to Depose Bao Dai’ (2006) 30(4) Diplomatic History 680.

Plebiscites in Dictatorships  35 choice unmistakably clear’.3 The result was certainly ‘unmistakably clear’. 600,000 supported Diệm. The registered number of electors was 450,000.4 In other words, the plebiscite was endorsed by 126 per cent of the electorate! But why hold such plebiscites? What was the aim and what was the reason for submitting the matter to a vote when the result was a foregone conclusion? What is paradoxical is that dictators such as Ngô Đình Diệm, who by their very nature are unrestrained by the shackles of accountability,5 who ‘rule by commands and prohibitions’,6 often resort to plebiscites. Dictators from Napoléon Bonaparte and his nephew Louis Napoléon, through Adolf Hitler and Benito Mussolini, to Ayatollah Khomeini and Bashar al-Assad, are but some of those who have felt impelled to submit issues to the voters in plebiscites.7 More generally, in the developing world ‘referendums have been utilized primarily by authoritarian regimes’.8 The question, given that most of these votes are anything but fair and free, is why do they do so? What is the reason for such a seemingly futile exercise? If we accept Guillaume-Hoffnung’s definition of a plebiscite as a vote that ‘allows a person … to legitimize himself by posing a question to the people which calls for an answer in the affirmative’,9 and if we limit ourselves to non-democratic (‘Not-Free’) states,10 the use of plebiscites has increased in recent decades. This essay investigates what role the plebiscite plays in an authoritarian or totalitarian regime. The general hypothesis is that plebiscites 3 W Brownell, The American Mandarin: a study of the life of Diem and of the origins of the American involvements (Ithaca NY, Cornell University Press, 1963) 158. 4 S Karnow, Vietnam: A history (New York City, Penguin, 1997) 239. 5 J Gandhi and A Przeworski, ‘Cooperation, Cooptation, and Rebellion under Dictatorship’ (2006) 18(1) Economics and Politics 1–26. 6 Ronald Wintrobe, ‘Dictatorship: Analytical Approaches’ in Carles Boix and Susan Stokes (eds), The Oxford Handbook of Comparative Politics (Oxford, Oxford University Press, 2007) 365. 7 David Altman, Direct Democracy Worldwide (Cambridge, Cambridge University Press, 2011) 88. 8 A Marques and TB Smith, ‘Referendums in the Third World’ (1984) 3(1) Electoral Studies 85, 103. 9 M Guillaume-Hoffnung, Le referendum (Paris, PUF, 1987) 14. It is, perhaps, in order to quote the original French, ‘Alors que le référendum permet au peuple de s’exprimer librement, le plébicite permet à une personne … de se légitimer en posant au peuple une question qui appelle une réponse positive’. 10 Non-democratic states are countries that were classified as Not-Free, according to Freedom House. For countries before 1974, these were polities where only one-party was allowed to contest elections.

36  Plebiscites in Dictatorships rather like mechanisms of policy-aggregation or legitimisation serve as instruments of control and repression, and that this dynamic can be fruitfully analysed through the prism of the semiology of Roland Barthes.11 Before turning to this explanation an overview of previous literature is useful. I.  THE PAUCITY OF PLEBISCITE RESEARCH IN DICTATORSHIPS

While there has been considerable research into elections in nondemocratic regimes,12 the role, rationale and occurrence of plebiscites in authoritarian and totalitarian regimes have largely gone unnoticed in the political science literature after the Second World War.13 As an instrument of totalitarian government, one would have expected that plebiscites had received more coverage in the growing literature on dictatorship. However, plebiscites are discussed neither in Gordon Tullock’s Autocracy14 nor in Ronald Wintrobe’s The Political Economy of Dictatorship,15 let alone in Ezrow and Franz’s Dictators and Dictatorships.16 Similarly, plebiscites are overlooked in the literature about referendums.17 The phenomenon of plebiscites under dictatorships constitutes a gap in the literature on non-democratic regimes. While there has been remarkably little research on plebiscites by political scientists in recent years, earlier generations of scholars acknowledged the plebiscite as a weapon in the armoury of totalitarian rulers, though the treatment of the topic was always somewhat marginal. Thus, prior to the Second World War, Robert Michels wrote in ­passing

11 Roland Barthes, Elements of Semiology (New York, Hill and Wang, 1964). 12 See, inter alia, J Gandhi, Political Institutions Under Dictatorship (Cambridge, Cambridge University Press, 2008), J Gandhi and E Lust-Oskar, ‘Elections under authoritarianism’ (2009) 12 Annual Review of Political Science 403–22, and, K Koehler, Authoritarian elections in Egypt: formal institutions and informal mechanisms of rule (2008) 15(5) Democratization 974–90. 13 For an important exception in German, see Otmar Jung, Plebiszit und Diktatur: Die Volksabstimmungen der Nationalsozialisten. Die Fälle, ‘Austritt aus dem Völkerbund’ (1933), ‘Staatsoberhaupt’ (1934) und ‘Anschluss Österreichs’ (1938) (Tübingen, Mohr Siebeck, 1995). 14 G Tullock, Autocracy (Dortrecht, Martinus Nijhoff, 1987). 15 R Wintrobe, The Political Economy of Dictatorship (Cambridge, Cambridge University Press, 2000). 16 NM Ezrow and E Frantz, Dictators and dictatorships: Understanding authoritarian regimes and their leaders (New York, Bloomsbury Publishing USA, 2011). 17 For an overview of the literature See M Qvortrup, Referendums and Ethnic Conflict (Philadelphia PA, University of Pennsylvania Press, 2014).

The Paucity of Plebiscite Research in Dictatorships  37 about plebiscites, and observed that ‘a Führer [sic!] would lead the people astray through unclear questions, and would himself be solely entitled to interpret the result afterwards’.18 Eric Voegelin likewise recognised that of the ‘screen devices’ used by Hitler to gain legitimacy, ‘the most important … [had] been the plebiscite’.19 He went on to say, that ‘the plebiscite is, by principle, a democratic device, and by using this method … measures could be taken that in substance have little to do with democracy’.20 However,­ Voegelin did not elaborate on, let alone develop a theory of, why dictators resorted to plebiscites at particular times. Roughly at the same time, a more empirically oriented political scientist writing in American Political Science Review analysed the patterns of voting in Hitler’s plebiscites.21 But this work contained little by way of generating hypotheses. At the same time, the controversial German legal theorist Carl Schmitt developed a normative political theory of plebiscites, though he did not hypothesise as to when these were likely to occur.22 After the Second World War, the writings on plebiscites were sparser. Friederich and Brzezinski did make passing reference to ‘rigged plebiscites’,23 and acknowledged ‘Hitler’s Volksbefragung through ­plebiscites’.24 In a similar vein, Leonard Shapiro observed that decisions in totalitarian states were characterised by ‘approval by plebiscite’.25 And Latey noted that dictators in tyrannical or totalitarian states ‘exploited to the full the modern totalitarian plebiscite … as a means of demonstrating [and] … of indoctrinating the people and of testing …

18 R Michels, Zur Soziologie des Parteiwesens in der modernen Demokratie: Untersuchungen über die oligarchischen Tendenzen des Gruppenlebens (Leipzig, Alfred Kröner, 1925) 431. 19 E Voegelin, ‘Extended Strategy: A New Technique of Dynamic Relations’ (1940) 2(2) The Journal of Politics 189, 193. 20 Voegelin, ‘Extended Strategy’, ibid. 21 AJ Zurcher, ‘The Hitler Referenda’ (1935) 29(1) The American Political Science Review 91–99. 22 Schmitt proposed that a dictator, as a ‘single trusted representative’, could use plebiscites to ‘decide in the name of the … people’ (Carl Schmitt, The Crisis of Parliamentary Democracy, Ellen Kennedy (tr) (Cambridge MA, MIT Press, 1988) 34). Through the plebiscite the dictator, as a kind of Hobbesian ‘mortall God’ (Thomas Hobbes, The Leviathan, R Flathman and D Johnston (eds) (New York, WW Norton, 1997, 95), could by-pass the political parties and, in a deus ex machina fashion, go directly to the people. 23 CJ Friedrich and ZK Brzezinski, Totalitarian dictatorship (Cambridge MA, Harvard University Press, 1965) 9. 24 Friedrich and Brzezinski, Totalitarian dictatorship, ibid, 13. 25 Leonard Shapiro, Totalitarianism (London, Macmillan, 1972) 41.

38  Plebiscites in Dictatorships control over them’.26 But the observations were disjointed and did not amount to a general hypothesis. This observation was also alluded to by Juan Linz who observed that ‘plebiscites in totalitarian systems’ were held to ‘test[s] the effectiveness of the party and its success in getting out the vote’.27 The same author, working with Houchang Chehabi, also suggested en passant that ‘in the communist world plebiscites are not customary’,28 but that they are often held in sultanistic regimes.29 However, these authors did not provide examples, let alone statistical evidence, in support of this assertion. In recent years historians specialising in Nazi Germany and Fascist Italy have undertaken case-study research into the use of plebiscites in these dictatorships. Beginning with Jung (1995) and more recently continued by Omland (2010),30 historians have provided historical accounts of the plebiscites held under Hitler. Similar research has been carried out in Italy, above all by Enzo Fimiami.31 However, in both cases historians of these totalitarian regimes concede that ‘until now, the problem of plebiscites … has only been dealt with marginally’.32 To date, no testable propositions have been developed and weaved into a coherent theory and no analytical statements as to the causes and effects of resorting to such votes have been developed. This essay seeks to remedy this gap in the literature. In the following, this essay will develop a framework of plebiscites as a mechanism of ‘control’. The analysis will be based on countries with totalitarian or tyrannical governments as defined by Wintrobe.33 It will be taken as an axiomatic assumption that totalitarian systems are 26 M Latey, Tyranny: a study in the abuse of power (London, History Book Club, 1969) 145. 27 JJ Linz, Totalitarian and Authoritarian Regimes (Boulder CO, Lynne Rienner Publishers, 2000) 92. ‘Getting out the vote’ refers to increasing the voter turnout. 28 HE Chehabi and JJ Linz, ‘A Theory of Sultanism’ in HE Chehabi and JJ Linz (eds), Sultanistic Regimes (Baltimore MD, Johns Hopkins University Press, 1998) 19. 29 Sultanistic systems are defined as ‘regimes based on personal ideology and personal favor to maintain the autocrat in power; there is little ideological basis for the rule except personal power’ (Chehabi and Linz, n 28 above, 1). 30 Otmar Jung, Plebiszit und Diktatur, n 13 above, and Frank Omland, ‘Plebiszite in der Zustimmungsdiktatur – Die Nationalsozialistischen Volksabstimmungen 1933, 1934 und 1938: das Beispiel Schleswig-Holstein’ in Lars Feld et al (eds), Jaahrbuch für direkte Demokratie 2009 (Baden-Baden, Nomos, 2010) 131–59. 31 E Fimiani, ‘La legittimazione plebiscitaria nel fascismo e nel nazionalsocialismo. Un interpretazione comparata’ (1997) 32 Quaderni storici 183–224. 32 Gustavo Corni, ‘Il Nazionalsocialismo: Una dittatura plebiscitaria’ in Enzo Fimiani (ed), Vox Populi? Pratiche plebiscitarie in Francia, Italia, Germania (Bologna, Clueb, 2011) 179–202, 179. 33 Wintrobe, The Political Economy of Dictatorship, n 15 above.

The Plebiscite as a Signifiant of Total Control  39 c­ haracterised by ‘a violent passion for unanimity’ and that their raison d’être is to limit all manifestations of dissent.34 II.  THE PLEBISCITE AS A SIGNIFIANT OF TOTAL CONTROL

Any scholarly study must be based on conceptual distinctions that correspond to real world phenomena. Borrowing Aristotle’s distinction between ‘Kingship’ and ‘Tyranny’, we can distinguish between polities in which a dictator rules over ‘willing subjects’, and one in which he rules over ‘unwilling subjects’.35 It is conceivable that dictators – under certain circumstances – can win the support of the people, such that ‘willing subjects’ may give their consent in a plebiscite.36 These votes do not concern us here. What we are concerned about in this essay is plebiscites in totalitarian states (ie countries with ‘unwilling’ subjects). The fundamental question is why do totalitarian rulers submit policy issues and support for their own rule to a vote in countries based on repression? What, for example, explained Saddam Hussein’s decision to conduct a plebiscite in which he won the support of 100 per cent of voters on a 100 per cent turnout only a few weeks before the allied invasion? It seems, at first sight, pointless to hold a vote if the result is a foregone conclusion. However, this apparent contradictio in adjecto is based on the premise that the plebiscite is a mechanism by which the demos exercises a choice and potentially can veto a decision by the rulers. But under tyrannical or totalitarian government this may not be the case. Plebiscites in ‘Not Free’ countries afford the voters no choice, merely an opportunity to acclaim the dictator. For example, according to case studies of plebiscites under tyranny, plebiscites often perform a kind of symbolic function.37 In the words of Leonard Shapiro, the fact that 34 Friedrich and Brzezinski, Totalitarian Dictatorship, n 23 above, 132. 35 Aristotle, The Politics (Oxford, Oxford University Press, 1905) 145. 36 This is arguably what happened in Bangladesh in 1977 when Major General Ziaur Rahman sought and won popular support for the military coup. While hardly free and fair, the plebiscite apparently did legitimise Rahman’s rule and was a reasonable reflection of a general support for the General. See M Rashiduzzaman, ‘Bangladesh in 1977: Dilemmas of the military rulers’ (1978) 18(2) Asian Survey 126–34. Another example of a plebiscite that arguably legitimised authoritarian rule is the Greek plebiscite in 1968. See N A ­ livizatos, Les Institutions politiques de la Grèce à travers les crises: 1922–1974 (Bibliothèque constitutionnelle et de science politique, vol 60) (Paris, Librairie générale de droit et de jurisprudence, 1979). 37 L Blaydes, ‘Who votes in authoritarian elections and why? Determinants of voter turnout in contemporary Egypt’ at Annual Meeting of the American Political Science Association. Philadelphia, PA, 31 August–3 September 2006.

40  Plebiscites in Dictatorships the vote was not a ‘spontaneous or rational expression [was] of no immediate significance … what we are concerned about here is the symbol of legitimacy’.38 But the fact that the plebiscite is ‘more than the act of depositing a single name ballot in the electoral box’39 begs the question: what is this symbolic function? III.  THE PLEBISCITE AS A SIGNIFIANT

Borrowing a distinction from semiology, as developed by Roland Barthes, the central thesis in this essay is that the plebiscite can be understood as a floating signifier.40 To understand this, a brief understanding of the basic concepts of semiology is necessary. To the observer from a Western democracy, the word referendum – le signifiant – conjures up an image of electoral and democratic processes – le signifé.41 But for the individual living under a totalitarian regime the same word conjures up a very different signifé, namely one of intimidation and control; of voting at gun-point and facing a severe fine for not voting.42 Of course, totalitarian dictators do not admit that plebiscites are in reality mechanisms of control. Indeed, Hitler, on the occasion of one of his four plebiscites, declared that he was ‘steeped in the conviction that the authority of the state proceeds from the people and must be ratified by them in a free and secret (geheimer) referendum’.43 Everybody knows that plebiscite is a sham and that the vote is severely controlled but, as a voter in Nazi Germany put it, ‘no one will dare not to vote, and no one will respond with a “no” … because 1) nobody believes in the secrecy of the ballot and 2) a “no” will be taken as a “yes” anyway’.44

38 Leonard Shapiro, Totalitarianism, n 25 above, 41. 39 Friedrich and Brzezinski, Totalitarian Democracy, n 23 above, 163. 40 I am using this term in a general way and not in the specific sense employed by LéviStrauss. See C Lévi-Strauss, ‘Introduction à l’œuvre de Marcel Mauss’ in Levi-Strauss, Marcel Mauss, Sociologie et anthropologie (Paris, Presses universitaires de France, 1950) ix–lii. 41 F Saussure, Course in General Linguistics, Wade Baskin (tr) (London, Fontana/Collins, 1974) 102. 42 For examples of the consequences of non-voting see Marques and Smith, ‘Referendums in the Third World’, n 8 above. 43 A Hitler, 1934 Reichsgesetzblatt, Berlin, 751–52. 44 Viktor Klemperer, I Shall Bear Witness, The Diaries Of Victor Klemperer 1933–41 Vol 1, M Chalmers (tr) (London, Phoenix, 1999) 33.

Qualitative Evidence of the Plebiscite as a Signifiant  41 But the practical use of the plebiscites is – as case studies suggest – to ensure that ‘voting is not just a duty but an opportunity to express publicly, visibly and preferably joyously the identification with the regime’.45 IV.  QUALITATIVE EVIDENCE OF THE PLEBISCITE AS A SIGNIFIANT OF TOTAL CONTROL46

The contention that the plebiscite carries a symbolic meaning and serves a repressive function rather than one of legitimisation, has often been hinted at by academic historians studying the phenomenon in Italy and Germany during the years of dictatorship. In Italy, under fascism in the 1920s and 1930s, the plebiscites were ‘a symbolic fact, a plastic testimony of the link between the new power and popular consensus’,47 and the plebiscites in Germany, were designed to ‘express publicly the power and unity of state and party and to compel the population to make gestures of homage and devotion’.48 While it is possible that ‘the plebiscites staged in 1933, 1934, 1936 and 1938 … reflected genuine widespread approval’,49 the archival evidence, as well as other qualitative data provide support for the thesis that the plebiscite was used to communicate that the totalitarian regime was in total control, and that the regime actively used the plebiscite to tighten their grip on power. The political campaigns were orchestrated by the party and citizens were compelled to take an active part. Thus, ‘for shop-keepers it was obligatory to decorate their stores, for business associations to call for support for the regime, for newspapers to proclaim the regime’s apparent success and their devotion to the Führer’.50 45 Linz, Totalitarian and Authoritarian Regimes, n 27 above, 92. 46 For a quantitative analysis that reaches the same conclusion see: Matt Qvortrup, Brendan O’Leary and Ronald Wintrobe (forthcoming) ‘Explaining the Paradox of Plebiscites’ (2019) 54 Government and Opposition, published online 10 August 2018. 47 Enzo Fimiani, ‘Elections, Plebiscitary Elections and Plebiscites in Fascist Italy and Nazi-Germany: Comparative Perspectives’ in Ralph Jessen and Hedwig Richter (eds), Voting for Hitler and Stalin: Elections Under 20th Century Dictatorships (Frankfurt am Main, Campus, 2011) 233. 48 Matthias Hörtnagel, Regionale Kultur in Zeichen des Hakenkreuzes, PhD-­Dissertation, University of Kiel, 1998, 134–35. 49 Ian Kershaw, The Hitler Myth. Image and Reality (Oxford, Oxford University Press, 1987) 258. 50 Frank Omland, ‘Germany Totally National Sozialist – Nationalist Socialist Reichstag Elections and Plebiscites, 1933–1938: The Example of Schleswig Holstein’ in Ralph Jessen and Hedwig Richter (eds), Voting for Hitler and Stalin: Elections Under 20th Century Dictatorships (Frankfurt/New York, Campus, 2011) 260.

42  Plebiscites in Dictatorships There is ample evidence from primary sources that these assessments are accurate and, consequently, that thesis that the plebiscite is a signifiant of repression is corroborated. For example, in Italy, in the plebiscite held in 1929, those who voted ‘yes’ had to put a ballot-paper designed to look like the Italian flag into the ballot box, whereas those who voted ‘no’ were required to put a plain coloured ballot-paper into a transparent box under the watchful eyes of election officials.51 While testimonies from individuals opposed to the regime must necessarily be treated with a degree of caution, they may contribute to the overall trend. In Italy, contemporary accounts report that much ‘publicity was given to the rumours that spoke of how the votes would be checked and controlled and of the consequences a negative vote would have for the voter’.52 The same conclusion was reached in Germany where a declared opponent of the regime reported that ‘the fear that there might be manipulations [of the result of the 1934 plebiscite] induced many of my acquaintances, who were declared enemies of National Socialism, to vote “yes” in the plebiscite’.53 And that same source’s fear that ‘polling would be closely checked … and that anyone who voted against the government would be found out’ was – as the archives suggest – well founded.54. The instructions sent out to local councils by the Nazi party required the former to communicate if anyone requested a postal-vote (Stimmschein), something which would enable them to avoid voting and which made it difficult for the authorities to check their compliance with the state ideology. As the letter sent out by NSDAP made clear, In order to prevent Marxist or other politically unreliable elements from using absentee ballots to avoid voting, the Ortsgruppenleiter must immediately contact local police officials … Absentee ballots are only to be issued in the most exceptional cases and only to politically reliable individuals.55

Indeed, so all-pervasive was the control exercised during plebiscites that even those interned in concentration camps were forced to vote. And it is clear from the evidence that massive ‘re-education’ was used in an

51 Archivio storico mediateca, No M136, March 1929. 52 Alfonso Leonetti, ‘No’ Come si è votato il 24 merzo in Italia. (Fatti e documenti sul plebiscite fascista) (Paris, Edizioni di coltura sociale, 1929) 49. 53 Quoted in Norbert Frei, Der Führerstaat. Nationalsozialistische Herrschaft 1933 bis 1945 (Munich, Deutscher Taschenbuchverlag, 2001) 221. 54 Norbert Frei, Der Führerstaat. Nationalsozialistische Herrschaft 1933 bis 1945 (Munich, Deutscher Taschenbuchverlag, 2001) 221. 55 KSF 9/26 NSDAP-Ortsgruppe Schleswig to Schleswig Magistrates, 6 November 1933.

Qualitative Evidence of the Plebiscite as a Signifiant  43 attempt to get them to support the regime. In some cases, however, this indoctrination failed to yield the desired effect. A civil servant, who was in charge of the voting at a concentration camp in Schleswig-Holstein, reported with Orwellian candour that although two-thirds of the inmates had voted ‘yes’, ‘a third have still not understood, or are unwilling to understand, what today was all about’.56 These examples, which are taken from notoriously totalitarian regimes in the first part of the twentieth century, may not, of course, be representative of all plebiscites held under severely repressive regimes. Yet we have similar examples of voting under duress in the 1979 referendum on the newly established Islamic Republic in Iran that similar tactics were used and where the citizens were forced to vote at gun-point.57 The examples of close monitoring of who voted and who did not were also a feature of the plebiscites carried out in East Germany under communism. In one Stasi-file, we can read that Pastor Horst Kassner (the father of Angela Merkel!) failed to vote.58 By controlling the plebiscite, the regimes, in the words of an ­Italian specialist in Fascism, ‘organised consensus, oppressed and at the same time made the people participate’.59 Never mind popular support or legitimacy, the plebiscite was a perfect mechanism for control. In a regime where collective punishment for not toeing the party-line was the order of the day, non-voting – itself an act of disloyalty – as Jessen and Richter have demonstrated with regard to communist regimes, ‘could … provoke the punishment of [the voter’s] shop floor brigade or housing collective in that he or she [the non-voter] put the brigade’s premium or the renovation of the house at risk’.60 Such knowledge was, as the same authors comment, ‘both demotivating and demoralizing’.61 Perhaps the best indication that the plebiscite is perceived not as a free choice but as a mechanism of control and repression is provided by literature. Fiction, of course, does not empirically prove theories, but it may point to and reflect a deeper and widely share perception.

56 Landesarchiv Schleswig-Holstein 309/22574. 57 A Marques and TB Smith, ‘Referendums in the Third World’, n 8 above, 87. 58 Landeshauptarchiv Schwerin 7.11-7, Bezirkstag Z48/1989/398 – Rat des Kreises ­Templin, Analyse über die kirchlichen Amtsträger, 25 October 1961. 59 Vittorio Foa, Questo Novecento (Turin, Einaudi, 1996) 126. 60 Ralph Jessen and Hedwig Richter, ‘Introduction: Non-Competitive Elections in 20th Century Dictatorships: Some Questions and General Considerations’ in Ralph Jessen and Hedwig Richter (eds), Voting for Hitler and Stalin: Elections Under 20th Century Dictatorships (Frankfurt/New York, Campus, 2011) 9–38, 27. 61 Jessen and Richter, ibid.

44  Plebiscites in Dictatorships George  Sand’s observation – written after one of Napoleon III’s plebiscites – that the ‘plebiscite is an attack on the liberty of the people’ is an indication of this.62 So too, is Giuseppe di Lampedusa’s account of the Italian plebiscites in The Leopard. In an exchange about the vote between the protagonist, Don Fabrizio, and one of his servants, Don Ciccio, Lampedusa wrote: ‘You know that everyone in Donnafugata voted “yes”?’ [said Don Ciccio]. Don Fabrizio did know this; and that was why his reply merely changed a small enigma into an enigma of history. Before voting many had come to him for advice; all of them had been exhorted sincerely to vote ‘yes’. Don Fabrizio, in fact, could not see what else to do; whether treating it as a fait accompli or as an act merely theatrical and banal, whether taking it as a historical necessity or considering the trouble these humble folks might get into if their negative attitude were known.63

Repressive (‘the trouble these humble folks might get into’) and ‘theatrical and banal’, ‘everyone in Donnafugata voted “yes”’; the plebiscite described by di Lambedusa was not a mechanism of public deliberation. V. CONCLUSION

Seen in the light of this, the paradox dissolves and the plebiscite – under tyrannies – becomes a straight, if ingenious, mechanism of repression and control. This aspect of the plebiscite explains some of the – at first sight – absurd examples of vote-rigging and indoctrination often witnessed in connection with plebiscites. In a totalitarian regime, securing endorsement becomes a question of effective implementation, and steps are taken to secure maximum compliance, for example by issuing instructions on how to fill out the ballot-paper (as in Austria in 1938), or as in the case of France under Napoléon III, offering voters only one option, namely, ‘Oui’. ‘A revolution’, wrote Benito Mussolini, can be ‘consecrated by a plebiscite, but never be overthrown’.64 Unlike referendums – which can lead to defeats or resignations of governments – plebiscites in repressive regimes are not held to gauge the voters’ opinions, still less to provide 62 Georges Sand, Journal d’un voyageur pendant la guerre (Paris, Michel Lévy Frères, 1871) 306. 63 Giuseppe Tomasi di Lampedusa, The Leopard (New York, Random House, 1960) 105. 64 Benito Mussolini, ‘La Diana del nuovo tempo’ in Benito Mussolini, Opera Omnia. Vol XXIII (Florence, La Fenice, 1957) 267–73, 272.

Conclusion  45 them with a potential veto over controversial decisions. Plebiscites are votes of confidence, acts of acclamation, held in countries where the voters are not free to express support. That such votes are held at all is a bit of a paradox. What is so puzzling about the plebiscite is that it produces almost blatantly absurd results. In this essay it has been argued that this apparent absurdity is the very raison d’être of the plebiscite. The reason dictators hold plebiscites is because these provide a unique mechanism of tightening the grip on the population. To use the terminology of Saussurean linguistics the signifiant ‘plebiscite’ – at least in an autocratic regime – conjures up an image of intimidation and control– le signifé. Anyone knows that the plebiscite is not fair, that 99 per cent has not endorsed the proposal, and that the acclamation is a sham. The significance of the plebiscite lies in the impression rather than in the result; it lies in the regime’s ability to deliver unanimity under conditions of repression. Using intimidation – voting under duress with transparent ballot boxes for ‘no’ votes (as in Italy), voting at gun-point (as in Iran) or even requiring that political prisoners endorse the regime’s proposals (as in Germany in 1934) – dictators’ use of this ostensible institution of direct democracy is a largely misunderstood mechanism of Orwellian repression. When seen through this prism, the paradox of plebiscites dissolves itself; the puzzling propensity of dictators to submit themselves or their policies to votes becomes rational and in its own diabolical way a rather clever means of increasing suppression. There have been a small number of referendums which resulted in defeats (to wit in Uruguay in 1980, in Chile in 1988 and in Zimbabwe in 2000), but apart from these examples – three out of more than 200 plebiscites held since 1793! – plebiscites are usually held to control the population, as a mechanism of repression. ‘Words get their meaning in the context of their use’.65 The plebiscite is an example of an institution that follows this contextual principle. While plebiscites are seen as methods of aggregating preferences in democratic countries, the reverse is the case in totalitarian states. Seen in this light is there is nothing paradoxical in Juan Linz’s assertion that ‘voting is not just a duty but an opportunity to express publicly, visibly and ­preferably joyously the identification with the regime’.66 The ­plebiscite 65 Ludwig Wittgenstein, Philosophical Investigations, GEM Anscome (tr) (Oxford, Basil Blackwell, 1953) 181. 66 Linz, Totalitarian and Authoritarian Regimes, n 27 above, 92.

46  Plebiscites in Dictatorships in a totalitarian regime is a mechanism for control. But when and why is it necessary for the totalitarian ruler to hold such plebiscites? La tyrannie consiste au désir de domination universel et hors de son ordre – ‘Tyranny consists in the desire of universal power beyond one’s scope’, Pascal remarked.67 Universal power requires that the tyrant is capable of showing the existence of absolute support. Rigged plebiscites are an excellent way of giving the impression that the tyrant has this power.



67 Blaise

Pascal, Pensées (Paris, Aug Renouard, 1812) 243.

4 The Law and Politics of Independence Referendums Covenants without the Sword are but words.1 Thomas Hobbes

I.  ONCE UPON A TIME IN NORWAY…

W

e start with a story from northern Europe. On 29 July 1905, the Norwegian Department of Ecclesiastical Affairs (Kirkeministeriet in the local vernacular) issues a statutory instrument – a circulære – urging the Lutheran Pastors to hold shorter sermons on Sunday, 13 August. Perhaps Norwegian clergy were prone to being long-winded. In any case, on this particular Sunday, they were requested to ‘so arrange the service that it will be finished in good time before 1 pm’.2 Given that church services started at 10 am the request was not unreasonable, yet this is hardly a matter that needed government regulation. But, 13 August was not just any Sunday. It was the day of the referendum on independence from the dual monarchy of Norway-Sweden. In the months before, matters had come to a head. The Norwegians were reluctant to pay roughly half of the costs of the consulates, especially as most of these were in countries Norway did not trade with. After the Swedes had done nothing, both chambers in the bicameral Norwegian Stortinget had passed a unanimous resolution on establishing separate Norwegian consulates. When the Swedish King, Oscar II, refused to sign the law, the Norwegian Prime Minister, Christian Michelsen, declared that no new government could be formed and the Stortinget immediately 1 Th Hobbes, Leviathan (Cambridge, Cambridge University Press, 1986) 117. 2 Quoted in Sarah Wambaugh, A Monograph on Plebiscites (New York, Oxford University Press, 1920) 1069.

48  The Law and Politics of Independence Referendums conferred the powers formerly vested in the monarch on themselves. This hurt Swedish pride, especially among aristocrats. Some called for armed intervention. Yet, there were fears that an invasion would be a pretext for Russia (Sweden’s eastern neighbour to attack) and, just as importantly, the Swedish army was not skilled in what were likely to be years of ­guerrilla warfare in the Norwegian mountains. Further, the Social Democrats and trade unions in Sweden were sympathetic towards the Norwegians. The Swedes had maintained – correctly – that a unilateral declaration of independence was legally void, as Norway had freely entered into an agreement. But this was politics rather than black-letter law, and Michelsen – a provincial shipping magnate from Bergen – was out of his depth and had likely bitten more off than he could chew, or so the Swedes believed. The Swedes decided to call his bluff and challenge the Norwegians to hold a referendum. After a week of deliberations, the Swedish Riksdag (parliament) issued a communiqué that called for a vote to ascertain the Norwegians’ views. If the result showed the ‘unequivocal expression of the people’s view of the matter [Norwegian independence]’ (otvetydigt uttryck for sin uppfatning af saken), then the Swedes would be willing to negotiate.3 Michelsen acted resolutely! The Swedes had underestimated him. In fact, he had already secured support for a plebiscite on 23 July – five days before the Swedes issued officially their challenge. It is not entirely clear from the records and archives, but it seemed Michelsen got wind of the Swedish decision before it became official. As a modern historian has put it, ‘Michaelsen had teft – this strange and almost primordial ability to sense, gauge and feel things as opposed to the ability to analyse, calculate and rationally assess matters’.4 The Norwegians immediately began to organise the referendum. Only one day after the Swedish Riksdag’s ultimatum had been issued the Stortinget issued a Cirkulære (statutory instrument), which in considerable detail instructed municipal and parish-council officials that a referendum [folkeafstemning] would be held on ‘the 13th of August from one o clock in the afternoon’ (Cirkulære Justitsministeriet 29 July 1905). 3 Sweden’s Riksdag Till Kungun, 28 July 1905, translated by the author. 4 PE Hegge, ‘Christian Michelsen: Slagkraft Ja – Dokumentlesning Nei’ in G Forr, PE Hegge and O Njolstad (eds), Mellem Plikt og Lyst: Norske Statsministre 1873–2010 (Oslo, Domino Forlag, 2010) 97.

The Context and Development of Independence Referendums  49 The ­Norwegian authorities showed no complacency and were keen that as many people as possible voted. Hence the aforementioned request that Pastors give short sermons. The result was astounding. 99 per cent voted for independence on an 85.6 per cent turnout. The Stortinget sent notification to the Swedes. The latter sought reconciliation and sent a request to The Hague Tribunal. But Stockholm knew – not least because of the overwhelming result  – that the game was up. They invited the Norwegians to negotiations in the Swedish border town of Karlstad on 31 August. The parties negotiated until 23 September, when it was agreed that the Norwegians would be granted independence if Sweden was allowed to keep two fortresses. The Norwegians – albeit grudgingly – accepted. On 16 October, the [­Swedish] Riksdag – after two months of negotiations – approved a government resolution to annul the Act of Union.5 But not all independence referendums follow this pattern. Far from it, as we shall see below. II.  THE CONTEXT AND DEVELOPMENT OF INDEPENDENCE REFERENDUMS

Before looking at the factors determining the fate of independence referendums, it is useful to put these into context. Overall, there are three types of referendums: • Ad hoc referendums (questions to solve a perceived political issue – such as David Cameron’s decision to hold a referendum on UK membership of the European Union in 2016); • Initiatives (votes initiated by a specified number of electors on (a)  already enacted legislation (as in Switzerland) or (b) new laws (as in Hungary); and • Constitutional Referendums (see next paragraph). By convention, there are three types of Constitutional Referendums. The Constitutional doctrine normally distinguishes between three types of constitutional referendums: on the approval of the constitution, on its revision, and on sovereignty issues (like the foundation of a new state).6

5 Wambaugh, A Monograph on Plebiscites, n 2 above, 168. 6 Laurence Morel, ‘Referendum’ in M Rosenfeld and A Sajo (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford, Oxford University Press, 2012) 504.

50  The Law and Politics of Independence Referendums While there has been a considerable debate on independence referendums and scholarly writings about them,7 the number of them is comparatively small. Thus, out of the 1200 referendums only 61 have pertained to independence (of which only four have returned a ‘no’ vote (Quebec in, respectively, 1980 and 1995 in Scotland in 2014 and New Caledonia in 2018) – though other referendums have failed because they did not satisfy super-majority requirements (eg in Nevis in 1998 and in several referendums in Palau in the 1980s). Historically, the independence referendums have come in waves, a few in the early 1860s, when the US States of Arkansas, Tennessee, Texas and Virginia – held referendums on independence following the election of Abraham Lincoln to the US presidency. All the referendums were won but no country recognised the results. The result of the referendum in Virginia is particularly noteworthy. Before the vote, representatives from counties in Western Virginia declared that they, in the event of a ‘yes’ vote for independence, would establish a new State and that the constitution of this new State would be approved by the voters in a referendum. Virginia as a whole voted for secession: 21,896 were in favour – 16,646 were against. However, in the Western counties 8,375 out of the 9,758 votes cast were against secession.8 The Western counties sent delegates to a specially convened convention, which declared that the referendum in Virginia was ‘illegal, inoperative, null, void and without force and effect’.9 They then ‘passed an ordinance providing for the “formation of a new state out of a portion of the territory of this state [Virginia],” … This ordinance was submitted to a plebiscite’.10 18,408 voted for a new State; 781 voted against it.11 After the American Civil War the US Supreme Court established in Texas v White that unilateral declarations of independence are unconstitutional,12 a case that was most recently used by the Alaskan Supreme Court in 2006 in Kohlhaas v Alaska to ban a constitutional initiative for independence for that State.13 After the American secession-votes there was gap of a few decades before the aforementioned Norwegian referendum, then a hiatus again 7 L Leduc, The Politics of Direct Democracy: Referendums in Global Perspective (Toronto, Broadview Press, 2002), J Laponce, Le référendum de souveraineté: comparaisons, critiques et commentaires (Quebec City, Les Presses de l’Université Laval, 2010). 8 J Mattern, The Employment of the Plebiscite in the Determination of Sovereignty (Baltimore MD, Johns Hopkins University Press, 1921) 120. 9 Quoted in Mattern, The Employment of the Plebiscite, ibid, 123. 10 Mattern, The Employment of the Plebiscite, ibid. 11 Mattern, The Employment of the Plebiscite, ibid. 12 Texas v White 74 US 700 (1868). 13 Kohlhaas v Alaska 147 P 3d 714 (2006).

The Context and Development of Independence Referendums  51 until the mid-1930s when the number of independence referendums began to pick up with the unrecognised, but successful, independence referendum in Western Australia in 1933 – 68 per cent voted in favour, but the vote was ignored as the secessionist party lost the State election on the same day14 – and the vote for independence for the Philippines in 1935. Especially after the Second World War, the referendum was increasingly used to show popular approval for decolonisation, though not all countries held plebiscites before they broke with their erstwhile colonial overlords. After a drop, in the 1970s and 1980s, there was an explosion of independence votes in the years immediately following the fall of the Berlin Wall and the collapse of Soviet Communism. Figure 1  Referendums on Independence 1860–2017

Number of Referendums

18

6 18 0 70 18 8 18 0 9 19 0 0 19 0 1 19 0 2 19 0 3 19 0 4 19 0 5 19 0 6 19 0 7 19 0 8 19 0 9 20 0 00 20 10

30 25 20 15 10 5 0

Source: Based on M Qvortrup, Referendums and Ethnic Conflict (Philadelphia PA, University of Pennsylvania Press, 2014). Note this Figure does not include the four multioption referendums in Puerto Rico (1968, 1993, 1998 and 2012), which formally included ‘independence’ as one of the options. However, the table includes the two-round multioption referendum in Newfoundland in 1948, as independence was one of the choices in the run-off. The independence options lost to ‘statehood’, and the former British territory became a Canadian Province. (See Qvortrup, Referendums and Ethnic Conflict, 69).

Not all independence referendums are comparable. Different historical and legal circumstances play a role. Further subdividing the independence referendums, we can distinguish between three forms,15 namely: 1) Post-Colonial (eg Philippines 1935); 2) By agreement (Montenegro and Bougainville); 3) Unilateral: Catalonia, Quebec and Estonia.

14 M Qvortrup, Referendums and Ethnic Conflict (Philadelphia PA, University of ­Pennsylvania Press, 2014) 29. 15 Ilker Gökhan Şen, Sovereignty Referendums in International and Constitutional Law (Heidelberg, Springer, 2015) 213.

52  The Law and Politics of Independence Referendums III.  LEGAL RECOGNITION

Not all of these different types of independence referendums have been equally conducive to the establishment of a new independent state. As Figure 2 shows, all the referendums in post-colonial territories (such as the Philippines in 1935, Micronesia 1983 and most recently in East Timor in 1999) have been recognised. The same is true for referendums held following an agreement, such as in the cases of the Montenegro referendum in 2006 and the referendum in South Sudan in 2011. It seems that the international community – which oversees these two  types of referendum – has been keen to ensure that its endeavours have not gone to waste, though it should be noted that some international agreements on referendums have not resulted in actual referendums, such as the referendum on the future of Kashmir and Western Sahara. These two latter referendums deserve to be mentioned although – or perhaps because – no referendum has taken place in either of the jurisdictions. Figure 2  Types of Referendums and International Recognition

Source: Based on Ilker Gökhan Şen, ‘Sovereignty Referendums: People Concerned and people entitled to vote’ in Laurence Morel and M Qvortrup (eds), The Routledge Handbook of Referendums and Direct Democracy (Abingdon, Routledge, 2018) 210–26 and M Qvortrup (ed), Referendums Around the World (London, Palgrave, 2017).

Thus, despite being condemned by the UN Security Council for its illegal annexation of Western Sahara, Morocco has delayed holding a ­referendum on the future status of the annexed area, in flagrant contravention of international law, due to uncertainties over the electorate.

The Legal Argument  53 Similar delaying tactics have been deployed by India over the disputed territory of Kashmir. The UN Security Council called for a referendum in Resolution 47, which stated that ‘A plebiscite will be held when it shall be found by the Commission that the cease-fire and truce arrangements set forth in Parts I and II of the Commission’s resolution of 13 August 1948’. That was 70 years ago at the time of writing. To date no ­referendum has been held. However, when referendums are held, the outcome has been accepted by the international community and by the parent states. Indeed, even when the result of the referendum was not legally binding (due to the doctrine of parliamentary sovereignty) the outcome of the referendum has been ratified by parliaments. Thus, the parliament of Indonesia – after considerable pressure from the international community – r­ecognised the outcome of the 1999 East Timorese independence referendum. The situation is markedly different for unilateral independence referendums. This type of independence referendum constitutes the ­ majority of the 42 independence referendums held since 1980. 36 or 85 per cent were in this category. Only in one twelve cases was the referendum followed by international recognition of the new state. Why is it that some referendums – even unilateral ones – result in the establishment of a new state (such as in the case of Bosnia, Estonia and the Ukraine) but not in other cases such as in Catalonia, Tartarstan and Somaliland? To answer this, we need to look at the legal aspects pertaining to – what misleadingly – is called the ‘right to self-determination’.16 IV.  THE LEGAL ARGUMENT

Legal philosophers may disagree at the theoretical level as to whether the law is the way acts or parliament or legal precedents are interpreted by the courts or whether there is some higher legal principle – or natural law – which overrides ‘black-letter law’. Proponents of the latter view – such as Thomas Aquinas (1224–74), but the doctrine can be traced back to Sophocles’ Antigone in the fifth century BC – hold that statutes and precedents that are at odds with natural law must give way. That is, ‘a legal norm fails to be valid if it goes against the human reason, regardless of the fact that it has been adopted by the state’.17 16 J-F Dobelle, ‘Référendum et droit à l’autodétermination’ (1996) 77(1) Pouvoirs 41–60. 17 Thomas Aquinas quoted in Ilker Gökhan Şen, Sovereignty Referendums in International and Constitutional Law (Heidelberg, Springer, 2015) 59.

54  The Law and Politics of Independence Referendums Whatever the philosophical merits of this view, courts and governments do not tend to be persuaded by legal theory or jurisprudence. Thus, while the late legal theorist Neil MacCormick, in the case of the United Kingdom, believed one could answer the question: ‘Is there a constitutional path to Scottish independence?’ affirmatively,18 this is very much a minority view among practising lawyers. Thus, while one may philosophically disagree with the ethical and moral tenants of legal ­positivism, this doctrine holds sway in practical politics. Hence, the following is based on a reading of the black letter law pertaining to independence referendums. The black-letter law of the ‘right’ to self-determination referendums is, in a sense, very simple. In the words of James Crawford, ‘there is no unilateral right to secede based merely on a majority vote of the ­population of a given sub-division or territory’.19 Those who espouse a similar legal positivist approach will further stress that this is consistent with the jurisprudence of international counts. Thus, in an obiter dicta in the Kosovo Case Judge Yusuf, opined: A radically or ethnically distinct group within a state, even if it qualifies as a people for the purposes of self-determination, does not have the right to unilateral self-determination simply because it wishes to create its own ­separate state.20

This view regarding the legality of independence referendums is nearly identical to the doctrine followed by domestic courts. In the Canadian case of Bertrand v Québec, it was held per Justice Robert Lessage that a referendum on a unilateral declaration would be, ‘manifestly illegal’.21 This is still the legal position notwithstanding the reasoning in the much cited (and little often misunderstood) Re Quebec.22 Thus, the general rule is that referendums have to be held in accordance with existing constitutions (such a provision exists in Article 39(3) of the Ethiopian Constitution but in few other states) or following an agreement between the area that seeks secession and the larger state of which it is part (this is what happened in the very different cases 18 N MacCormick, ‘Is There a Constitutional Path to Scottish Independence?’ (2000) 53 Parliamentary Affairs 725–26. 19 James Crawford, The Creation of States in International Law, 2nd edn (Cambridge, Cambridge University Press, 2006) 417. 20 Re Kosovo, ICJ Advisory Opinion, 22 July 2010, (2010) 49(5) International Law Materials 1410. 21 Bertrand v Québec (Procureur général), 1995 CarswellQue 131, 127 DLR (4th), 408 (per Lessage J). 22 Re Secession of Quebec (1998) 161 DLR (4th) 385.

The Legal Argument  55 of East Timor 1999, South Sudan 2011, Scotland 2014 and, a fortiori, ­Bougainville 2019).23 Following this logic, it would seem that the referendums in both Catalonia and Kurdistan, to take two recent examples, were both illegal and unconstitutional. Based on this reasoning, the Soviet leader Mikhail Gorbachev was well within his right to claim that the Latvian, Estonian and Lithuanian referendums on independence in the Spring of 1991 were illegal and that he was the guarantor of Pravovoe gosudarstvo – the equivalent of the rule of law in Soviet jurisprudence. Of course, some would say, p ­ reviously, under the so-called Stalin Constitution of 1936, individual Soviet states did indeed have the right to hold self-determination referendums under Article 48. But this provision had been dropped in the Khrushchev Constitution of 1956. Consequently, the Baltic republics were in breach in the early 1990s. (Though some claim that their annexation by the Soviet Union in 1939 was illegal and hence their declaration of independence was merely a statement of a reassertion of sovereignty). Consider the Catalan and Kurdish cases. As, respectively, the Iraqi and the Spanish constitutions do not allow for independence referendums, the two referendums held in these two entities referendum were, it would seem, ipso facto, unconstitutional. Yet matters are not that simple. Admittedly, all other things being equal a country only has a right if it follows the rules. However, when a region is part of an undemocratic constitutional order matters are a bit more complex. Antonio Cassese has argued, When the central authorities of a sovereign State persistently refuse to grant participatory rights to a religious or racial group, grossly and systematically trample upon their fundamental rights, and deny them the possibility of reaching a peaceful settlement within the framework of the State structure … a group may secede – thus exercising the most radical form of external self-determination – once it is clear that all attempts to achieve internal selfdetermination have failed or are destined to fail.24

As Iraq is not a well-functioning democratic state, it could be argued that Kurdistan meets these criteria. Again, the comparison with the Soviet Union is illustrative. Notwithstanding Gorbachev’s reforms, the USSR 23 P Radan, ‘Secessionist Referenda in International and Domestic Law’ (2012) 18(1) Nationalism and Ethnic Politics 14. 24 Antonio Cassese, Self-Determination of the Peoples: A Legal Reappraisal (Cambridge, Cambridge University Press, 1995) 119–20.

56  The Law and Politics of Independence Referendums was not a democratic regime, which consequently, provided the Baltic States with a justification for holding referendums. But, given that Spain is a democratic state, this rule hardly covers Catalonia. While the Spanish government, arguably acted in a way that appeared grossly disproportionate (to wit Police violence and arrest of democratically elected politicians), the legal argument remains the same. Catalonia is not currently part of a non-democratic state. Based on the situation, as it stands now, the Catalan referendum in 2017 was from a purely legal perspective extra constitutional. In a legal system under the rule of law, the powers of state institutions have to be enumerated in law. The basic principle of D’état du Droit is that citizens can do anything unless it is expressly prohibited. Public bodies or e­ manations of the state (to use a lawyerly expression) can only do things that are expressly permitted in law. Thus, the latter cannot legally speaking take actions that are not prescribed in enabling legislation. To pass legislation outside the boundaries of the constitution or enabling ­legislation is the very definition of being ultra vires. But does the law have to be that inflexible? Not necessarily. In Canada, the two referendums held in Quebec in, respectively, 1980 and 1995, were not strictly speaking within the powers granted to the Provinces by the Canadian Constitution.25 Technically speaking, the referendums were ultra vires. Yet, the Canadian judges, realising that legality ultimately rests on a modicum of legitimacy followed a more pragmatic logic. In the celebrated case, Re Quebec, the Court was asked the question, ‘Under the Constitution of Canada, can the National Assembly, legislature or Government of Quebec effect the secession of Quebec from Canada unilaterally?’ The Court held that while the ‘secession of Quebec from Canada cannot be accomplished … unilaterally’, a referendum itself was not unconstitutional but a mechanism of gauging the will of the ­francophone Province. Consequently a referendum, provided it resulted in a ‘clear majority’, ‘would confer legitimacy on the efforts of the Quebec ­government’.26 In other words, a result in favour of secession would require the rest of Canada to negotiate with Quebec. Needless to say, this ruling does not apply in Spain. But the Canadian example suggests that other

25 Ilker Gökhan Şen, Sovereignty Referendums in International and Constitutional Law (Heidelberg, Springer, 2015). 26 Re Secession of Quebec (1998) 161 DLR (4th) 385.

When are Referendums on Independence Recognised?  57 c­ ountries’ courts have shown a flexibility and appreciation of nuances that is conducive to compromises. These examples would seem to suggest that the international law pertaining to independence referendums is clear and simple. Alas, this is very far from being the case.27 While governments may confidently cite principles, the practice of independence referendums seemingly owes more to national interest than to adherence to principles of jurisprudence. For example, the states of Western Europe readily recognised the secessions of several former Yugoslav republics in the early 1990s – although these new states did not adhere to the aforementioned legal principles. And yet, in other cases international recognition has been less forthcoming even if the countries have seemingly followed the established norms. To wit, no state has to date recognised the outcome of Nagorno-Karabakh’s referendum in 1991, although Azerbaijan is very far from being a democratic state (the country has a Freedom House Score of 7 – the same as North Korea!28) and despite the greater freedoms for the citizens/inhabitants of the break-away republic. Similarly, no state recognised the referendum in Somaliland although this enclave is considerably more democratic, peaceful and respecting of the rule of law than Somalia, which at the time of the referendum was an arch-typical failed state. For all the legal arguments, acceptance of referendum results is ultimately a political rather than a legal decision. In other words, are all these arguments just examples of what IR scholar Stephen Krasner with an apt phrase called ‘organised hypocrisy’?29 Or, are states actually recognised if they follow the rules of the game? Or, it is simply a matter of power politics? V.  WHEN ARE REFERENDUMS ON INDEPENDENCE RECOGNISED?

Lawyers are interested in what is – or is not – legal and in accordance with more or less rigid rules. Political scientists, by contrast, are interested in what actually happens and the causes effecting this.

27 For a more general discussion see Ilker Gökhan Sen, Sovereignty Referendums in International and Constitutional Law (Heidelberg, Springer, 2015) 77ff. 28 https://freedomhouse.org/report/freedom-world-2018-table-country-scores. 29 Stephen D Krasner, Sovereignty: Organised Hypocrisy (Princeton NJ, Princeton University Press, 1999).

58  The Law and Politics of Independence Referendums Are there from a political science – or international relations – point of view causes and tendencies associated with recognition of referendum results? Or, are independence referendums simply recognised when the rules are followed? Alternatively, do we now live in a democratic age in which the gold standard of legitimacy is popular support? And, if the answer is in the affirmative, do independence referendums tend to be recognised when secession is supported by a large majority of the new demos on a large turnout? Or is it all down to power politics? Even 150 years ago, when the first independence referendums were held, democratic endorsement had a legitimising effect. And this effect was even stronger a couple of generations later. In the light of the latter it would seem reasonable and plausible that outcomes of referendums on independence would have an even stronger legitimizing force in an age where ‘democracy’ – to use a term from analytical philosophy – is an illocutionary speech-act, a term that demands unconditional observance. Yet, recent votes – such as the one in Kurdistan in 2017 – it seems that independence referendums, despite this near universal acceptance of the rhetoric of democracy, only tend to lead to independence and recognition when this is in the national interest of major powers. Whether it is one or the other – or more likely a combination of the two – is an empirical question. The hypothesis in the following is that power politics is the more important factor and that this can be demonstrated statistically.

VI.  STATISTICAL ANALYSIS

Since the 1980s there have been 44 successful referendums on independence. This analysis is based on the referendums held since the break-down of the Soviet Union. Before that date there had been relatively few independence referendums (only a handful in each decade). The first independence referendums were held in the US Confederate States of Texas, Virginia, Tennessee and Arkansas, where narrow majorities voted for independence in 1861. Other independence referendums include Norway (1905), Iceland (1944), Jamaica (1961) Algeria (1962) and Malta (1964). Of these, 16 have resulted in the establishment of a new state. What are the factors associated with the establishment of these new states?

Statistical Analysis  59 Factors associated with recognition are the legal one ‘the seceding entity was part of a non-democratic state. But there are also more political ones, eg a high turnout and a massive yes-vote. And then, there is the factor, which I think is the most important one – whether the new state has the support of the international community – or, more specifically, the three ‘democratic’ permanent members of the UN Security Council. In the analysis below we have measured some of the factors that statistically could be conducive for when states are recognised using what is known as a multiple logistic regression analysis. Without going into technical detail, this analysis measures the strength of the different given factors behind a phenomenon. The dependent variable is whether the state was recognised and took up a seat in the UN. The independent variables are the official yes vote, the turnout, the Freedom House score of the country from which the entity sought to secede and lastly a dummy variable for whether there was a support for secession among the five permanent members of the Security Council (in practice the USA, Britain and France). As the Table 1 shows Security Council Support from the three permanent Western powers is the key determining factor (statistically significant at p