How Referendums Challenge European Democracy: Brexit and Beyond (Palgrave Studies in European Union Politics) 3030441164, 9783030441166

Citation preview

PALGRAVE STUDIES IN EUROPEAN UNION POLITICS SERIES EDITORS: MICHELLE EGAN · NEILL NUGENT · WILLIAM E. PATERSON

How Referendums Challenge European Democracy Brexit and Beyond Richard Rose

Palgrave Studies in European Union Politics

Series Editors Michelle Egan American University Washington, USA Neill Nugent Manchester Metropolitan University Manchester, UK William E. Paterson Aston University Birmingham, UK

Following on the sustained success of the acclaimed European Union Series, which essentially publishes research-based textbooks, Palgrave Studies in European Union Politics publishes cutting edge research-driven monographs. The remit of the series is broadly defined, both in terms of subject and academic discipline. All topics of significance concerning the nature and operation of the European Union potentially fall within the scope of the series. The series is multidisciplinary to reflect the growing importance of the EU as a political, economic and social phenomenon. To submit a proposal, please contact Senior Editor Ambra Finotello [email protected]. Editorial Board Laurie Buonanno (SUNY Buffalo State, USA) Kenneth Dyson (Cardiff University, UK) Brigid Laffan (European University Institute, Italy) Claudio Radaelli (University College London, UK) Mark Rhinard (Stockholm University, Sweden) Ariadna Ripoll Servent (University of Bamberg, Germany) Frank Schimmelfennig (ETH Zurich, Switzerland) Claudia Sternberg (University College London, UK) Nathalie Tocci (Istituto Affari Internazionali, Italy)

More information about this series at http://www.palgrave.com/gp/series/14629

Richard Rose

How Referendums Challenge European Democracy Brexit and Beyond

Richard Rose Centre for the Study of Public Policy University of Strathclyde Glasgow, UK Robert Schuman Centre European University Institute Florence, Italy

ISSN 2662-5873 ISSN 2662-5881 (electronic) Palgrave Studies in European Union Politics ISBN 978-3-030-44116-6 ISBN 978-3-030-44117-3 (eBook) https://doi.org/10.1007/978-3-030-44117-3 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG, part of Springer Nature 2020 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover credit: Magic Lens/Shutterstock This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface: Spectres Haunting Europe

The peoples of Europe are today challenged by three political spectres. The first asks: How democratic is our political system? The second spectre wants to know: Who should make major political decisions—voters in referendums or members of Parliament? Thirdly: Where is power located—in national capitals, in Brussels or dispersed across continents? Each spectre presents a twenty-first century challenge to the late nineteenth-century belief that if citizens can elect a parliament this is enough to ensure democratic government. Democracy in Europe is today unbalanced. There is not enough democracy in the European Union and a surfeit in national political systems. In national systems citizens can participate in free elections that choose governors and hold them accountable. Opinion polls provide frequent popular judgements about how well the government is performing, and referendums occasionally give citizens the power to take a decision on a major issue. The EU has one essential democratic institution: free elections choose members of the European Parliament (MEPs). However, they are not elected on the basis of one person, one vote, one value, but by a system of disproportional representation. Nor can voters use their ballots to hold accountable the European Commission, the so-called eurocrats who form the executive branch of the EU. Who should govern is also in dispute. Representative democracy is challenged when many citizens see their representatives as advancing their

v

vi

PREFACE: SPECTRES HAUNTING EUROPE

own careers rather than the views of their voters. In consequence, politicians are having their monopoly of office challenged by outsider populist parties claiming that they represent the will of the people. Supranational European institutions are particularly vulnerable to challenge because they do not represent people. The European Commission represents supranational European values; the European Council represents member states; the European Court of Justice represents the principle of the rule of law; and the European Central Bank represents monetary economists. Referendums offer an alternative to government by representative democracy. Decision-making power is directly placed in the hands of an absolute majority of voters. The result shows whether most voters accept or reject the views of their elected representatives. Those who believe that democratic participation is desirable favour holding more referendums to increase the involvement of citizens in politics. Although the European Union does not have the authority to call a pan-European referendum, it cannot ignore the effect of a national ballot on a European issue. Brexit, the British referendum vote to leave the EU, is an extreme example. Assuming power is located in one place is misguided. Every national government is part of a network of political and economic interdependencies that extend across Europe and, as globalisation increases, across continents as well. The impact of decisions by European Union institutions is felt in all its 27 member states. National governments share power with the EU in the belief that making some policies in Brussels is more effective than going it alone. While every national government participates in the EU’s policy process, no government can control it. The intricate system of checks and balances institutionalised by EU treaties requires approval by super-majorities composed of many small states and a few big states. States that are small in population and economic clout welcome the opportunity to participate in decision-making rather than being ignored by big neighbours. The three large founder states–France, Germany and Italy–were reacting against defeat, depression and dictatorship. They saw European institutions as a means of settling their differences by what Winston Churchill described as jaw, jaw rather than war, war. Britain was approaching the end of its late Victorian status as the world’s leading political and economic power when Prime Minister Harold Macmillan put forward the country’s first application to join Europe. In a 1962 statement he argued, ‘In the past, a great maritime power, we might have given way to insular feelings of superiority over foreign

PREFACE: SPECTRES HAUNTING EUROPE

vii

breeds and suspicion of our neighbours across the Channel. But we have to consider the world as it is today and will be tomorrow and not in outdated terms of a vanquished past.’ While Macmillan saw entering European institutions as leveraging Britain’s political influence, his successors in Downing Street have tended to view it in transactional terms. Joining the European Economic Community in 1973 was viewed as paying an economic dividend and its limited political powers as imposing little cost. The views of founder members and Britain have increasingly diverged since the Maastricht Treaty of 1992 gave the renamed European Union powers to deepen and broaden its impact on member states and conferred European citizenship on Britons. This was followed in 1997 by the single-issue United Kingdom Independence Party starting to contest British elections and promote a referendum with the option of the UK withdrawing from the European Union. The June 2016 referendum majority vote to leave the EU was the culmination of two decades of anti-EU campaigning in a country that had always been a reluctant European as European member states have increasingly endorsed steps toward an evercloser Union. The Brexit referendum has turned Britain’s parliamentary system inside out. The sovereignty of Parliament has been replaced by the principle that the people are sovereign. A majority of referendum voters have been granted a legitimacy superior to that of a majority of MPs, who cast their referendum ballots in favour of remaining in the EU. Although the referendum vote was not legally binding, it has been accepted as politically binding. Downing Street and Parliament had such difficulty in implementing the referendum decision that the date for withdrawal was twice postponed. The deadlock was broken by a new prime minister, Boris Johnson, fighting and winning a December 2019 general election by framing the electorate’s choice in populist terms: the People (that is, the 51.9% who had voted for the referendum more than three years earlier) against Parliament. Johnson won a comfortable majority in Parliament, and in seven weeks the United Kingdom ceased to be an EU member state. The Brexit referendum is a symptom of challenges facing European democracies. Even though some features are distinctively British, many social science theories assume that the underlying causes of political and economic problems are common to all European countries. Moreover, the British political system has long been treated as a textbook example of representative democracy. The fact that withdrawal is seen by most

viii

PREFACE: SPECTRES HAUNTING EUROPE

European governments as a mistake adds to the value of examining it as a lesson of how to avoid political self-harm. The greatest challenges to European democracy do not come from the one country that leaves the European Union but from the 27 countries that remain in. Populist parties have capitalised on widespread dissatisfaction with the governing parties that represent the whole of their country in EU institutions. With few exceptions, populist parties are in opposition, but they often win enough votes to put pressure on mainstream parties. Without ever winning a seat in the British Parliament, the votes won by anti-EU parties led by Nigel Farage succeeded in forcing three Conservative prime ministers to hold a referendum on EU membership and then get Brexit done. The political eminence of EU policy-makers does not absolve those at the top from accountability to citizens at the bottom. The EU has no authority to call referendums but it also has no power to stop national ballots that affect its affairs. Almost every member state is able to call a referendum on an EU issue. Since treaties require the unanimous consent of all member states, a vote against a treaty in a single national referendum kills it. A Constitution for Europe was drafted by a convention of European elites in 2004, but failed to come into effect when it was rejected by referendums in France and The Netherlands. Since then referendums in Denmark, Greece, Hungary, Ireland, The Netherlands, Switzerland and the United Kingdom have rejected EU policies. For EU policy-makers, national referendums or the threat of a referendum are unwelcome reminders that European citizens are divided rather than united when it comes to applying European values to real, existing problems of public policy. Moreover, Europe’s system of multi-level governance institutionalizes a two-way flow of influence. At best a referendum is seen as a waste of time and redundant, simply confirming the judgement of governors. At worst, rejecting major policy is a public rebuke to the claim of representatives that they know what the people want. If a government turns a referendum into a plebiscite in which the outcome is predetermined by imposing unfair and unfree conditions on the ballot, this undermines its claim to be a democracy. The purpose of this book is to integrate the analysis of democratic elections, the institutions of the European Union and the politics of Brexit. In doing so it offers a unified understanding of three fields usually kept apart in the literature of contemporary political science. While academics can gain a Ph.D. by focusing on just one of these topics, it can be fatal for

PREFACE: SPECTRES HAUNTING EUROPE

ix

a prime minister to do, as David Cameron and Theresa May have learned the hard way. The opening chapters compare and contrast the two chief institutions of electoral democracy: parliamentary elections that decide which party governs and referendums that directly decide what government does about a major issue. The normative debate about whether referendums are a good thing pits politicians against citizens. The view that the European Union’s democratic deficit is a problem that can be met by reforms is shown to be mistaken. It is a structural condition that must be accepted. It is caused by the EU being a league of states extremely unequal in population. The potential for conflict arising from the difference between national and EU elections is structural too. Moreover, it is shifting from votes about joining the EU to votes about whether to accept or reject EU policies. Three chapters focus on elections about Europe that differ institutionally and in which conventional economic and class concerns compete with cultural values as influences on voters. Between 2016 and 2019 British voters were caught up in a natural experiment: four different elections were held in which Brexit was relevant. Comparing referendums that rejected EU policies in Greece, The Netherlands and the UK tests the extent to which there are common or country-specific influences when voters react against what Brussels does. Given the failure of pro-European politicians to win these votes, experimental evidence offers suggestions about how politicians ought to talk about Europe if they want to win support. Implementing a referendum decision is but one event in a lengthy and complex geopolitical process. Britain’s entry and exit from the EU must be understood as responses to fundamental political changes in the European continent and globally. Brexit campaigners won victory by reacting against the bounds that these changes have placed on national sovereignty. Their campaign slogan appealed to traditional British pride with the slogan ‘take back control from Brussels’ and made use of the Union Jack and its colours. The campaign for remaining in the EU did not fly the EU flag; instead, it promoted fear about the cash cost to future economic growth by leaving. However, retrospective evaluation shows that initial estimates were wide of the mark and Brexiters cite this as grounds for dismissing all expert warnings of future costs. Negotiating terms for Britain’s withdrawal from the EU and future relations with Europe necessarily involves reaching agreement with foreigners. However, British politicians have tried to treat leaving the EU as

x

PREFACE: SPECTRES HAUNTING EUROPE

a domestic political issue. Failure to understand the bounds imposed by Brussels on what she wanted Brexit to mean cost Theresa May her job as prime minister. The failure of Parliament to agree either a Brexit policy or a policy to stop Brexit has shown that it can check but not supplant the executive branch of British government. Boris Johnson ruthlessly gained the backing of both the electorate and Parliament to make the United Kingdom an ex-EU member state as of midnight Brussels time and 11 pm British time on 31 January 2020. Beyond Brexit worlds of challenges face both the United Kingdom and the EU, starting with the goal of reaching a formal Political Agreement about future UK-EU economic relations no later than 31 December 2020. Johnson has set out conditions that seek to maximize the extent of the break with Brussels: no need to conform to EU rules and regulations that apply to countries in the Single Europe Marker, including the free movement of people across national borders; and rejection of European courts enforcing the terms of any Agreement. Positively, Johnson wants trade with the EU to be free of tariffs and quota restrictions with a minimum of friction arising from the compliance costs for British exporters to the EU. The EU’s starting point is protecting its political authority by demanding that the more trade benefits Britain wants the more it must accept EU rules covering its 27 member states. The EU has already realised the political benefit of removing from its policymaking process the leading political opponent of EU proposals for further integration. Taking referendums seriously no more makes one an advocate of referendums as a cure-all for the political ills of Europe than taking China seriously makes one an advocate of authoritarian capitalism. The more one is familiar with referendums, the more their limitations and faults become evident. Likewise, taking Britain’s withdrawal from the United Kingdom seriously is not to condemn it as an act of political self-harm. It should be treated as a symptom of how politics in England works today rather than when I wrote a book by that title more than half a century ago. The concluding chapter draws lessons from Brexit relevant to national governments and to the European Union in a globalising world. When comparing and evaluating political institutions, this study relates abstract concepts to practical politics. Instead of comparing the practice of referendums and the theoretical justification of representative democracy, it examines the faults and justifications of both types of elections. It thus gives two cheers for referendum democracy and two cheers for representative democracy. Nor is this a book that assumes the challenges

PREFACE: SPECTRES HAUNTING EUROPE

xi

are signs that democracy is dying. Citizens are using their ballots to send demands for action. The democratic governors of Europe can decide how they respond, and citizens are free to give them their democratic desserts. The use of key terms in this book follows conventional inconsistencies. The term Europe can refer to an ill-defined geographical region; a set of states within it; or the peoples who, by the fiat of an EU treaty, are European Union citizens. The European Union is both a legal concept and a set of political institutions. For the sake of euphony, the terms European Union, EU and European institutions are used interchangeably. The terms Britain, the United Kingdom and the UK are similarly linked. Since the 2016 referendum asked voters to choose between the United Kingdom leaving or remaining in the European Union, political groups can be referred to as Remainers or Leavers, according to how they voted in that ballot. Brexit refers to a lengthy process that extends from the launch of a British political movement to promote euroscepticism a quarter-century ago to when Britain will be beyond Brexit and challenged to find its place in a post-Brexit world. The authors of this book do not fit into Theresa May’s Manichaean division of people into citizens of somewhere or of nowhere. As specialists in comparative politics, we belong to many ‘wheres’, having lived, studied and worked in a variety of European countries. These experiences give us a more rounded perspective on the challenges facing Britain and Europe than that of a British prime minister whose outlook on the world, including the United Kingdom, has been shaped by a life spent in the Thames Valley. The lead author has been studying comparative politics and elections across Europe since the days when Winston Churchill and Konrad Adenauer were heads of government. This book owes its inception to a seminar at the Robert Schuman Centre of the European University Institute Florence in January 2018. It examined the spate of referendums that were challenging Europeans to rethink the idea that representative democracy was the only form of democratic government. The support then and subsequently of the Director of the Centre, Professor Brigid Laffan, and her staff is greatly appreciated. The book also follows on an earlier book on Representing Europeans: a Pragmatic Approach, first published in 2013, which described the EU as ‘democracy lite’ and two years of writing about the difficulties Westminster politicians faced before and immediately after the 2016 referendum.

xii

PREFACE: SPECTRES HAUNTING EUROPE

Karen Anderson and Natalie Wilson were very helpful in dealing with the manuscript as it grew and changed under the pressure of political events. Glasgow, UK February 2020

Professor Richard Rose

About This Book

To understand the challenges facing European democracy today we need to think in three-dimensional terms. It is necessary to understand how and why referendums challenge conventional ideas of parliamentary democracy; how and why the European Union has a democratic deficit that cannot be reformed; and how and why people who are both national and European citizens cast national votes that Brussels policymakers do not like but cannot ignore. The expansion of the EU’s impact on national politics reflects national governments deciding that problems of the economy, immigration, climate change and much else are best dealt with collectively at the EU level, because their causes cross national and continental boundaries. When EU actions have unpopular consequences that are visible nationally, prime ministers are pulled two ways. They are accountable to their national electorate but also committed to EU decisions that they have approved when participating in policymaking in Brussels. The United Kingdom is an object lesson that is no less valuable for illustrating how not to balance pressures of a national electorate and EU treaty obligations. Like his predecessors, David Cameron was only halfhearted in efforts to convince the British people of the desirability of EU membership, The consequence was Brexit, the 2016 referendum majority for Britain leaving the EU. On becoming prime minister Boris Johnson quickly won a general election that gave him the power to confirm Britain’s EU withdrawal. Beyond Brexit Britain is being tested to advance

xiii

xiv

ABOUT THIS BOOK

its policies by going it alone in a world of interdependence. Simultaneously, the EU is being tested to achieve collective policies for the peoples of the European continent by pressures from across Mediterranean, the Near and Far East, and Washington.

Contents

1

1

How Democratic Elections Differ Richard Rose

2

How Referendums Differ from Each Other Richard Rose

19

3

Europe’s Democratic Deficit and Democratic Surplus Richard Rose

37

4

A Paradigm Shift in National Referendums on Europe Fernando Mendez

59

5

When Institutions and Issues Change, Voting Changes Ian McAllister and Richard Rose

77

6

How Politicians Ought to Talk About Europe: Lessons Learned from Experimental Evidence Konstantin Vössing

101

Comparing Voting in National Referendums on EU Issues Alexia Katsanidou and Slaven Živkovi´c

117

7

xv

xvi

CONTENTS

8

Britain: Still Searching for a Role in the World Tim Oliver

135

9

Forecasting the Economic Consequences of Brexit Paul Whiteley and Harold D. Clarke

153

10

Policy-Making in a Bounded Democracy Richard Rose

171

11

The Failure of Parliamentary Government Richard Rose

191

12

A New Prime Minister Meets Old Constraints Richard Rose

209

13

Beyond Brexit in a World of Interdependence Richard Rose

225

List of Contributors

Harold D. Clarke University of Texas, Dallas, TX, USA Alexia Katsanidou GESIS–Leibniz Institute for the Social Sciences, Mannheim, Germany; University of Cologne, Cologne, Germany Ian McAllister Australian National University, Canberra, ACT, Australia Fernando Mendez University of Zurich, Zurich, Switzerland Tim Oliver Loughborough University, Loughborough, UK Richard Rose Centre for the Study of Public Policy, University of Strathclyde, Glasgow, UK; Robert Schuman Centre, European University Institute, Florence, Italy Konstantin Vössing City University of London, London, UK Paul Whiteley University of Essex, Colchester, UK Slaven Živkovi´c GESIS–Leibniz Institute for the Social Sciences, Mannheim, Germany; University of Cologne, Cologne, Germany

xvii

List of Figures

Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig.

1.1 1.2 3.1 4.1 4.2 4.3 6.1 6.2

Fig. 6.3 Fig. 7.1 Fig. Fig. Fig. Fig. Fig.

7.2 9.1 9.2 9.3 11.1

Extent Prime Minister represents national electorate Big majority of Europeans see referendums as desirable National referendums show no consensus on EU Distribution of referendums per type (percent) Failure rate of referendums (by Decade and Type) Types of referendums over time EU justifications and support for the European Union Public policy justifications and support for the European Union Political parties and support for the European Union Perceived referendum consequences—Greece, UK and the Netherlands Evaluations of the economy and the anti-EU vote Forecasts of growth in GDP in real terms, 2008–2022 Forecasts of consumer price index 2010–2022 Forecasts of unemployment, 2010–2022 Most and least popular choices for implementing Brexit

7 14 48 62 63 63 101 105 113 122 130 159 161 162 201

xix

List of Tables

Table Table Table Table Table Table Table Table Table Table

1.1 2.1 3.1 5.1 5.2 5.3 5.4 6.1 7.1 9.1

Table 9.2 Table Table Table Table Table Table

10.1 11.1 11.2 11.3 12.1 12.2

Many citizens see their representatives as unrepresentative Outcomes of think-again EU referendums Party groups in the European Parliament, 2019 Contrasting party votes in general and EP elections Class and culture as influences in three types of elections Influences change when institutions and issues change List of BES variables, definitions and means Policy justifications and support for the European Union Logit models of anti-EU referendum vote Treasury estimates: short-term impact of Brexit, fiscal year 2016/17 to 2017/18 Effects of Brexit on growth, inflation and the change in unemployment EU not most helpful ally for big problems House of Commons rejects 15 Brexit proposals Public opinion about Europe fluctuates UK vote for European Parliament 2019 Votes and seats of pro- and anti-Brexit parties Leave and Remain voters’ party choice

13 23 53 80 87 92 97 107 127 157 167 174 196 198 206 218 219

xxi

CHAPTER 1

How Democratic Elections Differ Richard Rose

When democracy is defined as government by the people, for the people and of the people, a great deal of ambiguity is packed into eleven words. Government by the people is literally impossible in any political system larger than a village. A representative democracy can be described as government for the people because members of the parliament are accountable to citizens in a free and fair election. In contemporary Europe, the most common form of democracy is representative democracy. However, if representatives are very disproportionately male and middle-class and belong to one race in a multiracial society, their claim to be of the people is incorrect in the sociological sense. A referendum is of the people and by the people because all citizens can participate in directly deciding whether a policy should be adopted or rejected. Neither representative elections nor referendums represent the will of the people in Rousseau’s sense of being a single opinion held by everyone in a society (Weale 2018). In a parliamentary election, the ballot offers voters a choice between parties, and in a referendum, there is a choice between policies. Because a free and fair election offers a choice of alternatives, the outcome inevitably shows that the people are not of a single mind but divided in their political views. Referendums divide people into an absolute majority and a losing minority, and the difference between the two groups can be as small as it is sharp.

© The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_1

1

2

R. ROSE

The British experience of trying to resolve membership in the European Union by a referendum and then a general election illustrates differences between the two institutions. The 2016 referendum on European Union membership offered a simple choice between remaining in and leaving the EU. By a narrow margin, there was an absolute majority in favour of leaving, 51.9%. By contrast, the ballot in a parliamentary election offers a choice of up to a dozen or more parties. The 2019 British general election result gave MPs from ten different parties a chance to voice the views of a portion of the British people. The Conservative party came first with a plurality of votes, and thanks to the first-past-the-post electoral system, it won a House of Commons majority big enough to secure it a five-year term in government. Elections have been held for centuries in political systems that are not democratic. In the extreme case of the British Parliament, it was undemocratic for seven centuries before it was subject to election by a majority of its citizens. Referendums have a history that can be traced back to early sixteenth-century Europe (Qvortrup 2018: 12ff.; Mendez and Germann 2018). They were initially plebiscites in which referendum institutions were used to give the veneer of popular endorsement to whatever result the ruler wanted. In 1795 the Assembly of the French Revolution pioneered plebiscites, and Napoleon Bonaparte and Napoleon III followed suit. In the first eight French plebiscites, the median vote in favour of the authorities was 99%. In Nazi Germany, Adolf Hitler called three plebiscites to give the appearance of legitimacy to the presidential powers he claimed for himself; this accounts for the avoidance of referendums in post-war democratic Germany. Although communist regimes held undemocratic elections to national parliaments, they did not hold referendums. Communist ideology made this unnecessary since the partystate represented the will of the people. A variety of Latin American countries have used referendums as a political weapon to consolidate the power of undemocratic rulers (Wheatley 2017). In an era in which democracy has become a positive symbol, almost every country in the United Nations now has an elected parliament, a president or both, and referendums are widespread too. In consequence, undemocratic elections are now as common as free elections, since authoritarian regimes call votes to give their power the appearance of representing the voice of the people (Freedom House 2019). Likewise, the formal procedures of a referendum can be turned into an undemocratic plebiscite in which the outcome is pre-determined (Ulieri 2000).

1

HOW DEMOCRATIC ELECTIONS DIFFER

3

The symbolic appeal of democracy results in the term being invoked to justify all kinds of political systems and practices by adding an adjective. More than 500 different forms of democracy with adjectives have been documented (Collier and Levitsky 1997). Within the European Union, most countries have governments that describe themselves as liberal democracies, but in Hungary Viktor Orban proclaims that his government is an illiberal democracy. Until the Soviet Union collapsed, a dozen communist regimes described themselves as workers’ democracies. Democratic elections must meet four conditions: the ballot offers a choice between alternatives; competitors are able to campaign freely; votes are counted accurately; and the outcome is uncertain (Przeworski 1995). Uncertainty is particularly important to distinguish a democratic referendum from an undemocratic plebiscite. Uncertainty does not mean that competing sides have an equal chance of victory: it is sufficient that the result is not a foregone conclusion. If voters endorse the position of the government in a referendum, this is not evidence that the result was fixed in advance, but that the government’s position was widely supported and it ran a more effective campaign. By these conditions, the British referendum majority for leaving the European Union was democratic. Notwithstanding the backing of most elite representatives of public opinion, the prime minister’s cause lost after running a campaign that was a textbook example of how to lose a referendum (Farrell and Goldsmith 2017). Both parliamentary ballots and referendums are democratic elections; however, they differ in form and in purpose. Parliamentary elections decide who governs or, in countries with multiparty coalitions, which parties are eligible to haggle for a place in government. By contrast, referendums decide a specific issue. The coexistence of parliamentary elections and referendums in established democracies suggests they are complementary. Since neither form of election is free of faults, there is a trade-off of costs and benefits whichever approach is used.

1.1

Direct and Representative Democracy Compared

Both referendums and representative government can claim legitimacy for their decisions because they are the result of democratic elections. However, they differ in six significant ways: whether they are necessary

4

R. ROSE

for democratic governance; their place in the policy-making process; ballot forms; how winning is defined; the coalitions they represent; and the accountability of decision-makers. Electing representatives is necessary because the volume of decisions that a contemporary government must take each year is far greater than could be handled by referendums that only take a decision about a single policy. Parliamentary representatives entrust managing the everyday tasks of governing to an executive branch headed by a prime minister who represents the party or coalition of parties that can deliver parliamentary support for government decisions. Periodic elections give citizens the opportunity to hold their representatives accountable for their activities (Schumpeter 1952). Referendums are not necessary for democratic government; neither Germany nor the United States holds national referendums. Their timeconsuming procedures and cost in money and political attention mean that referendums cannot be used to decide the mass of issues facing a government. Nonetheless, more than nine-tenths of European countries make provision for referendums, and from time to time, such votes are held (see Qvortrup 2018a: 291–298). When referendums are held and the outcome endorses the decision of a government, it adds legitimacy to a decision taken by elected representatives. If a referendum rejects the position of elected representatives, it shows the significance of directly asking people to confirm or reject the assumption that the actions of elected representatives always reflect the preferences of those they represent. The use of referendums must be infrequent, because holding a national referendum takes time, money and a substantial investment of political capital by contending politicians. Months or even years may be needed to move from the initial debate about whether a vote should be held on an issue to the declaration of a referendum result. Since contemporary governments are responsible for enacting dozens of laws annually and administering thousands of laws, there is limited time in the political calendar to call a referendum. Even in Switzerland, where national referendums are most frequently held, only eight to ten issues are put to a vote each year, many on the same day (Serdült 2018: 79). Because few issues are of high importance on the political agenda at any one time, insofar as a referendum is good medicine for the ills of representative democracy, it should be taken in small doses. Voting for representatives is a policy input. The electoral system converts the votes that individuals cast for parties into an allocation of seats

1

HOW DEMOCRATIC ELECTIONS DIFFER

5

between parties in parliament. This results in the formation of a government. Within the black box of government, decisions about policies are in the hands of elected governors, non-elected civil servants and expert advisers. Even if the manifesto of the prime minister’s party contains a promise to adopt a particular policy, this promise is subject to modification or even abandonment by civil servants pointing out legal difficulties and by the finance ministry emphasising the cost of implementing a promise. By contrast, a referendum result is a policy output. The decision of a majority of voters to adopt or reject a law approved by their representatives in parliament can be immediately implemented. If the vote is about a principle, there is much scope for the government to decide how the principle should be implemented. In the case of a referendum about abortion or speed limits for motorists, the national government has the power to implement a referendum decision. However, if the outcome requires the agreement of other political systems, then the majority vote becomes an input to intergovernmental politics. This became spectacularly evident after a majority of Britons voted to leave the European Union. Both the government and Parliament have learned that they cannot impose the terms of withdrawal unilaterally. To agree on the terms of withdrawal, they have had to take into account the position of EU officials and 27 member states. The ballot choice in a parliamentary election is about which party governs. When campaigning for office, political parties need not emphasise their policies. If the party in office has been unsuccessful in managing the economy and much else, the opposition can offer to replace it with a government that is competent. Tony Blair led the British Labour party to election victories by avoiding reference to contentious economic and foreign policies. Instead, he asked voters to trust him to do ‘what works’, while leaving open what that meant. The claim of an election winner that it has a popular mandate to enact every policy in its election manifesto is unwarranted. Very few voters read the manifesto of the party they vote for. At most, they will be familiar with a few major policies that it contains. Given the tendency of voters to support the same party from one election to the next, an individual may vote for a party without regard to policy. A vote can reflect longterm party identification or trust in the party’s leader (Fisher et al. 2018: Chaps. 10–16). Even if a vote is cast because of agreement on a single important policy, voters can disagree with a party’s position on other

6

R. ROSE

policies, for example, agreeing on economic grounds but disagreeing on cultural matters or vice versa. A governing party cannot claim that gaining office is evidence that its own voters, let alone a majority of the electorate, endorse all the policies that it packages in a take-it-or-leave-it manifesto. A referendum offers a choice about an issue rather than about who governs. If the issue is one on which a voter disagrees with his or her party’s policy, for example a cultural rather than an economic issue, he or she can endorse the policy they favour without risking their party’s control of government. Whereas manifesto promises are contingent statements about what a party might do if it wins control of government, a referendum can decide what a government must do. Winning a parliamentary election requires a plurality of the vote rather than an absolute majority. In proportional representation elections, the winner’s plurality almost always falls below an absolute majority. Of the 27 prime ministers of EU member states (Fig. 1.1), almost two-thirds lead parties that won less than one-third of the national votes. Only the prime minister of Malta represents an absolute majority of voters. The inability of a prime minister to win the endorsement of a big majority of national voters creates a conflict with the prime minister’s role in representing the country in the European Council in Brussels. Since EU treaties are agreements between states, they assume that each prime minister represents the whole of the country and has the authority to commit all of its population to whatever decision the multi-national Council takes (see Chapter 3). Winning a referendum requires an absolute majority, a requirement beyond the reach of almost all national prime ministers. In a democratic referendum, the winners cannot claim that the outcome represents the views of a populist general will, since the vote is inevitably divided. Most referendums are won by a relatively slim majority, for if there is a popular consensus in favour of an issue there is no need for a referendum. However, in a highly competitive referendum, a national result can go either way. Thus, in 1992 a French referendum approved the adoption of the Maastricht Treaty to expand the EU’s powers greatly with a slim majority of 51%, while in a 2004 referendum on a Constitution for Europe 55% of French voters rejected it. Narrow victories are immediately decisive but produce a large minority of losers. Nonetheless, the winning side in a referendum can claim it represents most voters and should therefore be accepted by a government on the losing side.

1

Malta Hungary Romania Poland Greece Austria Cyprus Portugal Croatia Germany Bulgaria Czech Republic Spain Sweden Slovakia Ireland France Estonia Lithuania Netherlands Finland Denmark Luxembourg Slovenia Belgium Latvia Italy Vote, PM’s party

HOW DEMOCRATIC ELECTIONS DIFFER

7

55% 49% 46% 44% 40% 37% 36% 36% 35% 33% 32% 30% 28% 28% 28% 26% 24% 23% 22% 21% 21% 19% 17% 13% 8% 7% 0%

0%

50%

100%

Fig. 1.1 Extent Prime Minister represents national electorate (Note In France and Cyprus the entry is for the first-round presidential vote. In Italy the prime minister is an independent. Data as of 1 February 2020)

There is minimal scope for choice in a referendum ballot that only offers the alternative of voting for or against a single issue or choosing between two alternatives. It can be a yes/no choice between adopting a policy such as legalising same-sex marriage or it can be a choice between two mutually exclusive alternatives such as remaining in or leaving the European Union. The consequence of a binary choice is that it leaves no room for considering a third alternative or a compromise proposal. The stronger

8

R. ROSE

the feeling on both sides, as in the case of the Brexit referendum, the more likely it is that a dichotomous choice will polarise politics. Many policy alternatives are canvassed in a parliamentary debate or party conference before a government decides on a policy that can combine different views in a single Act of Parliament. Likewise, public opinion polls often give people a choice between three or more alternatives, and citizens’ assemblies tend to be even more wide-ranging. In the discussion of the future of Scotland, three broad views are invariably presented: maintaining the status quo; giving Scotland’s devolved parliament more powers; and independence. However, the 2014 Scottish referendum simply asked: ‘Should Scotland be an independent country?’ This forced people who wanted significant changes short of independence to make a radical choice between the status quo and independence. Coalition building is encouraged by the need for a parliamentary majority to maintain a stable government; it is also needed to create the absolute majority required to win a referendum. The practice of coalition government in most countries of Europe qualifies the ability of voters to decide who governs. In most European countries, the government is a coalition formed after weeks or months of negotiation. To create a coalition government, each party usually compromises some of its manifesto pledges and thereby fails to represent fully the views of a significant portion of its voters on compromised issues. Some coalition governments have a majority of MPs while others may be a minority government. To gain the French presidency in 2017, Emmanuel Macron needed only 24% of the vote in the initial ballot in order to be the sole alternative to the radical right-wing Marine Le Pen in the second-round run-off vote. Three-fifths of Macron’s second-round majority came from people who had voted for other candidates in the first round. A referendum coalition is different. Winning requires combining support from voters who back different parties or none but agree about a particular issue. Instead of only mobilising supporters of the governing party, to win an absolute majority campaigners must put together a crossparty majority. This is done by stressing multiple themes. In the 2016 referendum, the winning Brexit coalition was composed of people who supported different parties at a general election while agreeing in their attitudes towards cultural issues such as immigration and English nationalism (see Chapter 5). Parties competing for government are institutionally accountable, because they fight repeated elections; this enables voters to make choices

1

HOW DEMOCRATIC ELECTIONS DIFFER

9

based on how well they are seen as representing the people who vote for the party. Between elections, their actions are held to account by parliament, by 24/7 media and by public opinion polls. This makes anticipation of the next election a continuous pressure on representatives wanting to get re-elected. Referendum victors cannot be held institutionally accountable for implementing the policy they campaigned for because they are an ad hoc coalition. Once the result is announced, the purpose of a campaign organisation ends, their funding dries up, and their staff members disperse to other jobs. Parties that may co-operate with each other in a referendum campaign will immediately revert to competing with each other in preparation for the next parliamentary election. Responsibility for implementing a referendum decision reverts to the government of the day, and the task of monitoring how it is implemented reverts to elected representatives in the parliament.

1.2

Are Referendums a Good Thing? Politicians Mix Principles and Pragmatism

The classic argument in defence of representative government as compared to government policy-making dictated by voters was made in 1774 by Edmund Burke in an address to his electorate in Bristol: It ought to be the happiness and glory of a representative to live in the strictest union, the closest correspondence, and the most unreserved communication with his constituents. Their wishes ought to have great weight with him; their opinion high respect; their business unremitted attention. It is his duty to sacrifice his repose, his pleasures, his satisfactions, to theirs; and above all, ever, and in all cases, to prefer their interest to his own. But his unbiased opinion, his mature judgment, his enlightened conscience, he ought not to sacrifice to you, to any man, or to any set of men living. These he does not derive from your pleasure; no, nor from the law and the constitution. They are a trust from Providence, for the abuse of which he is deeply answerable. Your representative owes you, not his industry only, but his judgment; and he betrays, instead of serving you, if he sacrifices it to your opinion.

Even though Burke was speaking in a pre-democratic era, his principle is frequently invoked as an argument against giving a referendum superior

10

R. ROSE

authority to that of MPs. A Google search for Burke Brexit on 7 January 2000 returned more than three million results. Article 56 of the contemporary Danish constitution echoes Burke’s rhetoric: ‘The Members of the Folketing shall be bound solely by their own consciences and not by any directions given by their electors’. However, referendums are authorised in article 42, the longest in the Danish constitution, and 17 referendums have been held since the constitution was last revised in 1953. In the world of politics, the desirability of referendums is debated not only in terms of abstract principles but also according to the specific issue on the ballot. A referendum confirming a decision of elected representatives can be dismissed as a waste of time. If a referendum stands a good chance of overturning a government decision, that is a political incentive for it to avoid a ballot. However, if a referendum is held and a government policy is rejected, this is not proof that the electorate has made a bad decision; it may be evidence of bad government. If the governing party sees a referendum as in its interest, this is sufficient to justify calling it. Two centuries after Burke’s statement, in 1975 Prime Minister Harold Wilson abandoned the convention that referendums are un-British and called a referendum to confirm the UK’s membership in the EU. His motive was that there was no consensus about membership within the Labour Cabinet. Pragmatic reasons likewise led Prime Minister David Cameron to call Britain’s 2016 referendum on EU membership in order to appease anti-EU Tory MPs. Cameron did so confident of victory; in the event, he miscalculated and resigned. If there is a constitutional obligation to hold a referendum, the immediate pragmatic question is: How can it be won? If the government is very confident of victory because there is a consensus among major parties, a referendum may be called to add popular legitimacy to parliamentary legitimacy. In the case of East European national referendums on joining the European Union, the confidence was not misplaced. In 2003, a total of 90% of Slovenes and 84% of Hungarians voted to join the EU. In countries in which a referendum can be triggered by a popular initiative, the only choice that governors have is whether to take sides or, if it is a moral issue that divides its supporters, to refrain from doing so. For groups that have never been in government, such as the gilets jaunes in France, demanding a referendum is a means of satisfying supporters, getting media attention and embarrassing a government that has adopted unpopular policies. Macron’s government has rejected their demand for

1

HOW DEMOCRATIC ELECTIONS DIFFER

11

his government’s policies to be subject to national referendums. However, it has approved their use for policies adopted by regional and local authorities. Political Scientists Divide: Indifferent, Against and For The characteristic response of political scientists is to ignore referendums. In a major handbook on elections, referendums are the subject of only one of its 41 chapters (Marsh 2018). The classic American theorist of democracy Robert Dahl concentrates exclusively on representative democracy; the index of his major study contains no reference to referendums (1989). A lengthy volume entitled Representation, Elections and Beyond has no chapter about referendums (Nagel and Smith 2013). When political scientists engage with the subject, they usually focus on the defects of referendums. This is particularly the case among writers who regard the ideal form of decision-making as taking place through deliberation in which a variety of alternatives are discussed, resulting in a consensus policy incorporating multiple interests (Habermas 1996: 135ff.; Lijphart 2012). Decision-making by choosing between two alternatives is criticised as undesirable because it polarises opinion and rules out compromises that offer something for almost everybody. The winnertake-all character of a referendum decision is also said to be a potential threat to minority rights that are not protected constitutionally. To argue that referendums should be held only when voters are in possession of the facts assumes that facts exist independently of their political interpretation, thereby making it possible to distinguish false facts from true ones. Non-partisan fact-checking institutions monitor referendums and provide evidence questioning the validity of statements disseminated by opposing sides through new social media as well as traditional media. Yet even if opponents agree about specific facts, they can disagree about how to interpret evidence when deciding how to vote in a referendum. The British Independent Electoral Commission endorses measures that promote greater transparency about the sources of campaign information, especially online. However, it rejects increased government control of the content of campaigners’ statements on the grounds: ‘While the truth is vital, it is also contested’ (ICE 2018: 170). It is philosophically naive to assert that referendums should be held only on issues for which the future consequences are known. This assumes that the future impact of a policy is known before it is put into effect.

12

R. ROSE

This is not the case. Careful forecasters attach the qualification ‘all other conditions remaining equal’; however, this condition cannot be met in the real world. Campaigners on both sides of an issue are free to weaponise forecasts and offer unicorn visions of future consequences. However, the first- and second-order consequences of a referendum only become visible years after a referendum decision is put into effect. Advocates of participatory democracy see referendums as good for citizens and good for governance. They make democracy stronger by giving ordinary people a chance to participate in politics and make decisions about issues of public policy (Pateman 1970; Barber 2004). Representative democracy is described as ‘thin democracy’ because citizens have only one chance every few years to elect parliamentary representatives. A major comparative review of empirical evidence about the effect of referendums on participation tends to endorse their positive effect (Talpin 2018: 406ff.). The increase in referendums has encouraged survey research on referendum voters. The electorate divides into two unequal groups: a well-educated, largely middle-class minority that is informed and reasonably knowledgeable about politics, and a mass of citizens with limited knowledge and interest in politics (cf. Aachen and Bartels 2016). The median voter is, in Sartori’s terms, ‘cognitively incompetent’ (1987: 120; cf. Colombo 2018). A study of American voters by Hibbing and Theiss-Morse (2002) concluded that, given limited political interest and knowledge, many voters would prefer their representatives to decide major policies rather than having to do so themselves. Such evidence is cited as justification for keeping policy-making exclusively in the hands of political elites. Lack of knowledge is particularly relevant in referendums on European Union issues. Individuals are asked to decide a complicated policy that is usually remote because it involves countries spread across Europe. For this reason, referendum voters can be influenced by simple appeals that selectively pick out a particular cost or benefit (Marsh 2018). In the words of a former president of the European Commission, José Manuel Barroso, referendums should be avoided because they ‘undermine the Europe we are trying to build by simplifying important and complex subjects’ (Rose 2015: 94). Criticisms of direct democracy often overlook the fact that many defects ascribed to referendums can equally be applied to the election of parliamentary representatives (see, e.g., Rosenbluth and Shapiro 2018).

1

13

HOW DEMOCRATIC ELECTIONS DIFFER

Table 1.1 Many citizens see their representatives as unrepresentative

Government doesn’t care what people like me think Politicians only out for themselves Can’t trust politicians to do what’s right

Agree %

Don’t know %

Disagree %

55

21

24

50 40

28 32

22 28

Source International Social Survey Programme 2014 Citizenship study in 17 European countries with 24,872 respondents (www.GESIS.org/issp/modules)

The political intelligence of citizens is much the same whether they are electing MPs or voting in a referendum. The binary division of the vote is the normal way in which MPs register confidence or lack of confidence in the government of the day and decide whether a bill becomes a law. If the rights of minorities are not protected by the constitution and the courts, they are at risk from parliamentary legislation as well. When a referendum is turned into an undemocratic plebiscite in which the government pre-determines the result, a practice familiar in Latin America (Ruth et al. 2017), this is not due to weaknesses in the institution of referendums. It is because leaders of an authoritarian political system want to exploit the symbolic value of democracy by holding undemocratic presidential, parliamentary and referendum elections.

1.3

Citizens Favour Referendums

If politicians were always responsive to the concerns of the people they represent, then referendums would be redundant, for every ballot would simply confirm a decision taken by the government the people elect. However, many citizens do not have confidence that those they elect can be trusted to pay attention to their concerns (Zmerli and van der Meer 2017). The Citizenship Survey of the International Social Survey Programme (ISSP) provides striking evidence of the extent to which Europeans see politicians as unrepresentative. A majority believe that government does not care what they think and that politicians are only out for themselves and a plurality do not think they can be trusted to do what is right (Table 1.1). When representatives are seen as unrepresentative, referendums offer voters a chance to prevent having their views being ignored. Moreover,

14

R. ROSE

if a majority of citizens favour the principle of holding referendums, this gives political weight to the demand that a vote should be held on a major issue. It particularly encourages an opposition party to call for a vote on a policy that it supports, since advocating a popular referendum choice will win far more votes than a party can win in a parliamentary election. The prime campaigner for a referendum on leaving the EU, the United Kingdom Independence Party (UKIP), won only 12.6% of the 2015 British parliamentary election vote and was unable to win even one seat in Parliament. However, in a referendum the following year, four times as many people voted for Brexit. The Citizenship Survey asked respondents to evaluate the use of referendums about important issues without specifying topics. Doing so avoids confounding attitudes towards the principle of referendums with attitudes on a specific issue such as the European Union. A total of 61% favoured referendums. The remainder was divided between one-quarter with no opinion and one-eighth who believed referendums were undesirable (Fig. 1.2). There was an absolute majority in favour of the principle of referendums in 15 countries. Support was highest in Switzerland, 86%. Reflecting the UK’s tradition of parliamentary democracy, referendum support in Britain was below average, 54%. When the ISSP asked Q. Referendums are a good way to decide important poltical questions (Agree or Disagree) Agree

No opinion

Disagree

12%

27% 61%

Fig. 1.2 Big majority of Europeans see referendums as desirable (Source 2014 Citizenship Survey of the International Social Survey Programme. Interviews with 24,872 people in 17 European democracies [see http://www.issp.org])

1

HOW DEMOCRATIC ELECTIONS DIFFER

15

the same question in 2004 in 15 European countries, an almost identical proportion of respondents, 62%, gave the same answer as a decade later. Moreover, there was an extremely high correlation, 0.91, between the percentages of national respondents favouring referendums ten years apart. This shows that popular support for direct as well as representative democracy is rooted in long-standing values. Theories of populism and civic values offer contrasting explanations of why there is a popular majority in favour of referendums. The populist explanation is that support comes from those who think that the ‘real people’ should make decisions; who have extreme ideological views; and who feel ignored by politicians because of their low socio-economic status (Mudde and Kaltwasser 2017; Müller 2016). An alternative view is that referendums are supported by people who have civic values that favour political participation and welcome referendums because they offer an additional democratic opportunity to participate in the political process. The Citizenship Survey has lots of questions that can be analysed by multivariate statistical analysis in order to identify which of 17 populist and civic indicators actually influence referendum support. Five measures increase by at least half the odds that a person will favour referendums by comparison with a random respondent (for full statistical details of the logistic regression, see Rose and Wessels 2020). Civic values account for four of the five strongest influences on referendum support. Discussing politics with friends is the simplest way in which people can participate in politics: it does not require attending meetings, giving time and money or identifying with a party. Moreover, the topics discussed are free from the agenda-setting power of political parties and the media, where political elites select topics. After controlling for all other indicators, discussing politics has the strongest impact: individuals who do so are almost twice as likely to favour referendums as the average European. In contemporary democracies, an individual’s vote can have little influence because it is only one among millions of votes, yet substantial majorities of citizens vote in parliamentary elections. This is done from a sense of civic duty rather than from a belief in one’s personal power to affect the outcome. Among ISSP respondents, 48% strongly believe that good citizens should always vote. If people see voting as important for citizenship, they should support referendums, and this is the case. The likelihood of supporting referendums is three-fifths higher among people with a normative commitment to voting.

16

R. ROSE

If people see democratic institutions working as they should work, this encourages support for voting in referendums as well as in parliamentary elections. Among ISSP respondents, more than half-believed democracy was working well in their country, raising the likelihood of support for referendums by more than half. A necessary condition of democratic institutions working well is that parties offer a choice of policies to reflect differences of opinion within the electorate and just under half see this as the case in their country. Since referendums do offer a choice, those who see no real alternative in parliamentary elections are three-fifths more likely to support referendums. The anti-referendum argument that representative democracy is sufficient to make major decisions has an empirical flaw. It assumes that the majority of politicians behave in fact as they should in theory, as trustworthy agents of their voters. However, many voters see politicians as unrepresentative (Table 1.1). Of the three indicators of politicians failing to show concern for their voters, the biggest impact is the perception that politicians are only out for themselves. This boosts by three-fifths the likelihood of people supporting referendums so that they have the chance to make judgments on public policies independently of distrusted representatives. Because referendums replace decision-making by a minority of welleducated professional politicians by allowing the great mass of citizens to vote, populist theories assume that they appeal to people who feel left out of politics because they lack such socio-economic advantages. However, multivariate regression analysis shows that this is not the case. Neither people with low income nor those of low education are significantly more likely to favour referendums, and age also has no effect. Nor does support for referendums reflect the views of the small proportion of ISSP respondents who describe themselves as strongly right wing or left wing. Most people who favour referendums place themselves in the ideological centre of politics. The evidence thus fails to justify the claim that popular support for referendums only comes from people with unflattering sociological or ideological characteristics. Referendums are not only democratic in institutional form but also supported by people who hold civic values. Endorsement comes primarily from people who think it is important to participate in politics and do so. Whereas populist leaders want to replace existing representative institutions with those that centralise power in the hands of populist leaders, civic democrats want to reform institutions to strengthen the capacity of

1

HOW DEMOCRATIC ELECTIONS DIFFER

17

citizens to influence public policy. Holding referendums on major issues is one such reform. Popular support for referendums does not mean that the institution is fault-free and elected representatives and political scientists are quick to emphasise these faults. However, representative institutions have their faults too. The failure of British politicians to arrive at any decision about what Brexit means three years after it was endorsed by a referendum is an outstanding example. While the institutions of representative and direct democracy differ from each other, they share a common feature of political institutions: in practice, they are imperfect realisations of democratic ideals.

References Aachen, C., and L. Bartels. 2016. Democracy for Realists: Why Elections Do Not Produce Responsive Government. Princeton: Princeton University Press. Barber, B. 2004. Strong Democracy: Participatory Politics for a New Age, 2nd ed. Berkeley: University of California Press. Collier, D., and S. Levitsky. 1997. Democracy with Adjectives. World Politics 49 (3): 430–451. Colombo, C. 2018. Justifications and Citizen Competence in Direct Democracy: A Multilevel Analysis. British Journal of Political Science 48 (3): 787–806. Dahl, R. 1989. Democracy and Its Critics. New Haven: Yale University Press. Farrell, Jason, and Paul Goldsmith. 2017. How to Lose a Referendum: The Definitive Story of Why the UK Voted for Brexit. London: Biteback. Fisher, J., E. Fieldhouse, M.N. Franklin, R. Gibson, M. Cantijoch, and C. Wlezien (eds.). 2018. The Routledge Handbook of Elections, Voting Behavior and Public Opinion. London: Routledge. Freedom House. 2019. https://freedomhouse.org/report/. Habermas, J. 1996. Between Fact and Norms: Contributions to a Discourse Theory of Law and Democracy. Cambridge, MA: MIT Press. Hibbing, J., and E. Theiss-Morse. 2002. Stealth Democracy: Americans’ Beliefs About How Democracy Should Work. New York: Cambridge University Press. ICE: Independent Commission on Referendums. 2018. Report of the Independent Commission on Referendums. London: University College London Constitution Unit. Lijphart, A. 2012. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, 2nd ed. New Haven: Yale University Press. Marsh, M. 2018. Voting Behaviour in Referendums. In Fisher et al. 2018, 256– 266.

18

R. ROSE

Mendez, F., and M. Germann. 2018. Contested Sovereignty: Mapping Referendums on Sovereignty over Time and Space. British Journal of Political Science 48 (1): 141–165. Mudde, C., and C. Kaltwasser. 2017. Populism: A Very Short Introduction. Oxford: Oxford University Press. Müller, J.-W. 2016. What Is Populism? London: Penguin Books. Nagel, J., and R. Smith (eds.). 2013. Representation, Elections and Beyond. Philadelphia: University of Pennsylvania Press. Pateman, C. 1970. Participation and Democratic Theory. Cambridge: Cambridge University Press. Przeworski, Adam. 1995. Sustainable Democracy. New York: Cambridge University Press Przeworski. Qvortrup, M. (ed.). 2018a. Referendums Around the World. London: Palgrave Macmillan. Qvortrup, M. 2018b. The History of Referendums and Direct Democracy. In The Routledge Handbook to Referendums and Direct Democracy, ed. L. Morel and M. Qvortrup, 11–26. London: Routledge. Rose, R. 2015. Representing Europeans: A Pragmatic Approach, 2nd ed. Oxford: Oxford University Press. Rose, R., and B. Wessels. 2020. Do Populist or Civic Values Drive Support for Referendums in Europe? European Journal of Political Research, in press. Rosenbluth, F., and I. Shapiro. 2018. Responsible Parties: Saving Democracy from Itself. New Haven: Yale University Press. Ruth, S.P., Y. Welp, and L. Whitehead (eds.). 2017. Let the People Rule. Colchester: ECPR Press. Sartori, G. 1987. The Theory of Democracy Revisited. Chatham, NJ: Chatham House. Schumpeter, J. 1952. Capitalism, Socialism and Democracy, 4th ed. London: George Allen & Unwin. Serdült, U. 2018. Switzerland. In Referendums Around the World, ed. M. Qvortrup, 47–112. London: Palgrave Macmillan. Talpin, J. 2018. Do Referendums Make Better Citizens? In The Routledge Handbook to Referendums and Direct Democracy, ed. L. Morel and M. Qvortrup, 405–418. London: Routledge. Ulieri, P. 2000. Plebiscites and Plebiscitary Politics. In International Encyclopedia of Elections, ed. R. Rose, 199–202. Washington, DC: CQ Press. Weale, A. 2018. The Will of the People: A Modern Myth. Oxford: Polity Press. Wheatley, J. 2017. A Problem or a Solution? In Ruth et al. 2017, 41–61. Zmerli, S., and T. van der Meer (eds.). 2017. Handbook on Political Trust. Cheltenham: Edward Elgar.

CHAPTER 2

How Referendums Differ from Each Other Richard Rose

The frequency with which countries hold referendums differs greatly. Within Europe, Switzerland holds hundreds of referendums at the federal, cantonal and local levels. Even though a referendum has never been held at the national level in the United States, such ballots are very frequently held at the state and local government level. Since 1990 Italy has held 56 votes; Ireland, 27; and Slovenia, 25. While three-fifths of European countries only make infrequent use of their power to call referendums, when they do so it is usually a major national or even European event (Anckar 2018: 110). The frequency with which referendums are held is increasing, whether the measure is the number of countries holding ballots or the number of referendums called (ICE 2018: 19). More than 20 European countries have held referendums in the past 30 years, and the total number has increased from 58 in the three decades from 1960 to more than 240 referendums in the most recent three decades. Before 2000 more countries around the world held ballots that were plebiscites in which the dictator calling a vote was certain of the outcome. Since then, democratic referendums have increased in number while plebiscites have decreased. More than half a dozen democracies annually hold a referendum (Qvortrup 2018: 263ff.). The UK is a newcomer to the use of referendums. For more than a century after democratisation began, referendums were dismissed as an institution that was alien in a country that was the mother of © The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_2

19

20

R. ROSE

parliaments. Since 1973 three UK referendums have been held, plus ten referendums in Northern Ireland, Scotland or Wales, and some English cities have held ballots about whether to have an elected mayor. The results have been democratic: the government side has sometimes won and sometimes lost. In addition, there have been pledges to call referendums on major issues. Tony Blair pledged that the UK would not join the Eurozone nor would it approve the draft European Constitution without first calling a referendum. In opposition David Cameron promised to hold a referendum on the Lisbon Treaty but did not do so because it was endorsed by the Labour government before he took office in 2010. Legislation makes referendums necessary in Scotland and Wales before their devolved systems of government could be abolished and in Northern Ireland before it could become part of the Republic of Ireland.

2.1

Why Referendums Are Held

A majority of democracies have constitutional requirements making a referendum obligatory for a limited number of policies before they can become law. A vote may be mandated on laws that alter the territory and boundaries of the state, the rights of citizens or international treaties. Denmark is distinctive in requiring a referendum in order to lower the voting age. The Irish constitution stipulates that there must be a postlegislative referendum to confirm any constitutional amendment approved by the Irish parliament. The logic for doing so is that, if government derives its powers from the people, it should go back to the people to change its powers. In the words of an Irish Supreme Court judge: ‘The State is the creation of the People and is to be governed in accordance with the provisions of the Constitution which was enacted by the People and which can be amended by the People only, and the sovereign authority is the People’ (quoted in ICE 2018: 59). A referendum can become de facto obligatory if there is a consensus among the political elite that a particular policy is so important that it would benefit by gaining legitimacy from popular approval (Mendez et al. 2014: 75ff.). A national constitution may prohibit holding a referendum on a specific issue. When the nationalist government of the region of Catalonia called a referendum in 2017 on whether it should become an independent state, the Spanish Supreme Court declared the ballot illegal and its organisers have been arrested and indicted for sedition and rebellion. The Italian constitution forbids holding referendums on major taxing and

2

HOW REFERENDUMS DIFFER FROM EACH OTHER

21

spending policies, amnesties and pardons, and the ratification of international treaties, even if an international treaty effectively changes the country’s national constitution. In the absence of a written constitution, the British government has responded inconsistently to proposals for referendums on Scotland becoming independent. The Conservative government under David Cameron accepted the initiative of the Scottish parliament to hold an independence referendum in 2014, whereas in 2019 Boris Johnson pledged that his Conservative government would not allow another referendum on Scottish independence. When there is discretion in deciding whether a referendum should be called, political calculations are of major importance. The rational choice theory of politics implies that governors should call a referendum only when it is virtually certain of that the electorate will endorse its position. If a government is a multiparty coalition, it may resolve internal differences by calling a referendum (Vospernik 2018). A government can call a referendum if it expects that doing so will mobilise more electoral support for itself and divide the opposition. If both the government and the opposition are internally divided, a referendum can be called with multiparty endorsement. The bill authorising the 2016 Brexit referendum was carried by an all-party House of Commons vote of 544 to 53. An opposition party can call for a referendum from conviction or to embarrass a government carrying out an unpopular policy. However, by definition it lacks the votes to have its demand adopted. This has been the experience of the British Liberal Democrat party, which endorses the principle of referendums but has almost invariably lacked enough MPs to get a bill authorising a referendum approved by Parliament. The popular initiative gives citizens the statutory power to call a referendum to adopt a new policy or repeal a policy in force. It thus prevents elected representatives from having a monopoly of the right to decide what the mass electorate can vote on. That is why a majority of European countries do not give their citizens the right to initiate a referendum. To launch an initiative, money and organisation are needed to collect the signatures required to put a proposal to a popular vote. Whatever the number of signatures called for, it is usually a very small proportion of the electorate. In Italy, 500,000 signatures are needed to initiate a referendum; this is about 1% of the electorate. In Switzerland, the 100,000 signatures needed to initiate a vote on a constitutional amendment constitute about 2% of the electorate. Ironically, the success of petitioners in putting their favoured policy to a popular vote can reveal its lack of

22

R. ROSE

widespread support. Less than one-third of referendums called by petition are approved by voters. A variety of theories have been put forward to explain why countries differ in the frequency with which referendums are held, such as small countries being more likely to hold referendums than large countries or their occurrence being more frequent in ethnically divided countries. The failure of many putative causes to achieve statistical significance when tested suggests that policy-makers call a ballot when the particulars of a policy and political interests favour a ballot. The only statistically significant influence is that referendums are more frequently called in countries in which the popular initiative can be used to call a ballot (Qvortrup 2018: 51). This suggests that in countries that do not allow for a popular initiative there is an unrealised demand for popular votes that is not met because elected representatives do not want to share their power to decide policies with those who elect them.

2.2

Repeating a Referendum

The basic logic of a democratic election is that competitors accept its rules and procedures and that losers as well as winners consent to the winner-take-all result. If the vote is to confirm a legislative act, the referendum immediately puts the result of the vote into effect. However, if the referendum is a pre-legislative vote about a principle, it is up to the government and parliament to decide the terms on which the principle is to be implemented. Losers are sometimes attracted by the idea of a think-again referendum that could reverse a previous majority, but getting approval of a second referendum on the same subject faces obstacles of principle and of politics. In the short period of time between a referendum result and it being put into effect, there is unlikely to be fresh evidence of changing circumstances. Today, the eighteenth-century Burkean doctrine that the government knows best what is in the national interest carries less weight with MPs than the argument that a referendum majority has superior legitimacy. If the government of the day believes a defeated policy is exceptionally important, it may promote a repeat referendum in the hope that the second time around voters will give the ‘right’ result, that is, the result it wants. One justification for doing so is that times have changed. Public opinion on matters important to individuals tends to change slowly. If

2

Table 2.1 Outcomes of think-again EU referendums

HOW REFERENDUMS DIFFER FROM EACH OTHER

Turnout

23

Pro-EU vote

Norway reject joining Europe twice 1972 EEC 79 1997 EU 89 Change +10 Switzerland reverses European association 1972 Join EFTA 53 1997 Join EU 35 2001 Join EU 53 Change 0 Denmark reverses Maastricht Treaty vote 1992 83 1993 86 Change +3 United Kingdom reverses European membership 1975 EEC 64 2016 EU 72 Change +8 Ireland reverses rejection of EU treaties twice 2001 Nice 35 2002 49 Change +14 2008 Lisbon 53 2009 59 Change +6

46 47 +1 72 26 23 −49 49 56 +7 67 48 −19 46 63 +17 47 67 +20

Source Qvortrup (2018: 291–298)

there is agreement before holding a referendum that it is an experiment with uncertain consequences, a sunset clause can be added requiring a second vote in seven to ten years. Sweden has a think-again requirement for constitutional amendments. They must be approved by two different parliaments, the second called after a general election. This would have made no difference for Brexit, since in the general election following the 2016 referendum 85% of the vote went to parties pledged to implement withdrawal. A democratic government cannot dissolve the electorate, but in five European countries voters have been asked to think again after rejecting a government-backed policy (Table 2.1). When the rejected measure is an EU treaty requiring unanimous approval by all member states, the pressure to approve is strong to prevent the country’s government being

24

R. ROSE

blamed for a referendum vetoing a treaty accepted by up to 27 other member states. Norwegians have twice been asked by their government to vote on joining Europe. In 1972 the referendum was triggered by the prospect of closely linked neighbours, Denmark and Britain, becoming EU members. Norwegian political parties were divided about EU membership. On a high turnout, 54% of votes were cast against membership, and the prime minister resigned. By 1994 the EU had expanded its powers and importance. When the proposal was once again put to a referendum, it was again rejected (see Saglie 2000). The Norwegian government has not made a third attempt to join the EU. Instead, it has been a prime mover in developing an association with the EU that gives the country economic rights and obligations without the political rights of membership. Switzerland has never been a member of the European Union but it has had formal treaty relationships with it since 1972, when almost threequarters of Swiss referendum voters endorsed joining the European Free Trade Area (EFTA). When it became evident in the 1990s that all the countries bordering Switzerland would become EU members, the Swiss government opened formal discussions with Brussels about joining. However, in 1997 almost three-quarters of Swiss voters rejected negotiating membership. In 2001 a popular initiative triggered a ballot asking the government to apply for membership but three-quarters of Swiss voters again rejected doing so. In 2016 the Swiss government formally withdrew its request for EU membership. Danish voters were the first to show that a national referendum could put a stop to further European Union integration when 50.7% rejected the Maastricht Treaty in 1992. The very high turnout left little scope for reversing the result by mobilising support from people who had not voted. To help the Danish government secure enough votes to reverse the result, the EU agreed to legally binding opt-outs for Denmark in four areas of EU policy. The pro-EU government used this as a means of reducing the salience of political features of the Maastricht Treaty and increasing the positive relevance of economic benefits (Slune and Svensson 1994). In a second referendum, the initial result was turned around. The 56% majority for the Maastricht Treaty meant that it could be put into effect and European integration could proceed apace. Irish referendums have approved five EU treaties since 1992; twice it took two ballots before the Irish electorate would do so. The initial vote on the Nice Treaty in 2001 was lost by 8 percentage points on a very low

2

HOW REFERENDUMS DIFFER FROM EACH OTHER

25

turnout. The Irish government responded by getting the EU to confirm that Ireland would not be required to abandon its historic neutrality policy. Much more important, the Irish government realised that it could not count on voters doing what they should without an active campaign to mobilise support. Doing so gave a big boost to turnout and swamped the anti-EU vote (cf. Hobolt 2009: 189ff.). During the first ballot on the Lisbon Treaty in 2008, the No side anticipated a second referendum by arguing that, if people voted to reject the treaty, this would strengthen the Irish government’s ability to extract concessions for use in a second ballot. A small majority voted No, and the Irish government secured two treaty amendments relevant to Ireland’s priorities. Given these concessions, a significant number of No voters switched sides, and the Lisbon Treaty carried the second time around. Between the first and second British referendums on EU membership, enormous changes had occurred in the European Economic Community (EEC) that British voters had approved in 1975. The EEC had been transformed from a very imperfect common market into a European Union making economic, political and social policies with a visible impact on the citizens of member states. The number of EU member states trebled and they had histories very different from Britain. The historic argument for European integration—preventing a depression like the 1930s and a Third World War among European states—had little resonance with a British population, most of whom had been born after the 1975 referendum. In June 2016 a new electorate reversed the old result: Britons voted to leave the EU. The 2017 election manifestos of both the Conservative and Labour parties made commitments to implement Brexit, but there was no agreement about the terms of implementing withdrawal within the Conservative government, among the Labour opposition, or in a hung Parliament (Chapter 11). The more the Conservative government pursued a ‘hard Brexit’ policy or was inclined to leave the EU without a deal, the more uncomfortable pro-European MPs became about accepting the government’s policy of leaving the EU whatever the cost. A think-again referendum offered opponents of the government’s approach to Brexit a way to justify their very firm commitment to Britain’s long-term EU membership. It also showed their resigned commitment to accepting the superiority of a referendum vote over a vote in Parliament. The People’s Vote campaign was launched to fight fire with fire. Instead of lobbying MPs to use their Burkean best judgement to reject

26

R. ROSE

the Brexit verdict, it called for a think-again referendum (https://www. peoples-vote.uk). The reason for seeking a think-again referendum so soon after the first ballot was that negotiations for leaving the EU had revealed major difficulties in implementing withdrawal that were ignored or not known when the referendum was held in 2016. The underlying motive of British campaigners for a second vote was the belief that it would reverse the first vote and Britain would remain in the European Union. Very extensive public opinion polling was conducted to identify favourable signs of change. The electorate had changed as some older voters who were disproportionately anti-EU had died and another cohort of youthful, pro-EU voters had joined the electorate. To secure a second referendum required an Act of Parliament. Conservative MPs were virtually united in opposition to a second referendum while opposition MPs were hesitant about reversing the consent they had given to the referendum majority. However, in a hung parliament there was the potential to create a cross-party coalition. In votes on a second referendum in March 2019, an amendment for a second vote was defeated, and there were many abstentions (Table 11.1). The cross-party coalition then changed tactics. It advocated a referendum offering a choice between confirming the government’s Brexit policy, whether no deal or a hard Brexit or remaining in the EU. On 19 October a cross-party alliance of MPs addressed a pro-referendum demonstration in London with hundreds of thousands attending. Within ten days Boris Johnson offered MPs an alternative strategy to get Brexit done: holding a general election and opposition party leaders accepted this challenge. Johnson framed the election as a conflict of the People (i.e. the referendum majority) versus the anti-Brexit Parliament (Chapter 12). The election of a Conservative government with an absolute majority put an end to the People’s Vote campaign.

2.3

Conditions for Holding a Referendum Vary

Scepticism about the responsiveness of representatives justifies popular endorsement of referendums as a means of compensating for the behaviour of unrepresentative representatives (see Chapter 1). However, approving referendums in the abstract leaves open a lot of choices about how a referendum is conducted. There are many steps in the process of moving from a demand for a referendum to announcing the result of

2

HOW REFERENDUMS DIFFER FROM EACH OTHER

27

the vote (Morel and Qvortrup 2018). Choices must be made about what issues should be put to voters; what the question should be; and the size of the vote required to win. Because referendums concern issues on which there are political conflicts, procedural choices are open to political challenge on the grounds of favouring one side or the other. Selection of Issues A heterogeneous range of issues are the subject of the 300-plus referendums held since 1990. The largest group, 39%, is concerned with constitutional amendments. This reflects the fact that in four democracies— Australia, Denmark, Ireland and Switzerland—referendums are required to alter the national constitution and in many countries this may be considered desirable even if not formally required. One-seventh of ballots ask voters to approve the country participating in an international organisation or accepting an EU treaty. Votes have been held on such international issues as membership of NATO and free trade. A small number of referendums have involved territorial divisions, such as regional autonomy or a nationalist party seeking endorsement of independence, successfully or otherwise. Socio-economic, environmental and public order issues have been the subject of two-fifths of referendums. Moral issues such as abortion, divorce and same-sex relationships have been the topic of about one in eight ballots. They create divisions within as well as between parties, since they are independent of conventional left/right divisions in the electorate. A few states, such as Italy, explicitly ban holding referendums on such issues as taxing and spending, to prevent voters creating fiscal havoc by approving tax cuts while simultaneously mandating increases in public expenditure. Countries where referendums are more frequently held are likely to put all kinds of issues to a popular vote. Danish referendums have concerned issues as diverse as legalising female succession to the throne, which was approved, and joining the euro, which was rejected. In the space of two years, the Irish have voted on constitutional amendments to abolish the upper chamber of parliament, which failed, and to legalise gay marriage, which was approved. In Italy votes have been held to repeal laws on everything from the rights of hunters to authorising the construction of nuclear power stations.

28

R. ROSE

Critics of referendums argue that issues that are technically complex, such as medical matters and environmental regulations, should not be put to a referendum for decision because they are beyond the capacity of the great majority of voters to understand. The argument can also be used to disqualify decision-making by a majority of MPs and Cabinet ministers. Insofar as this is done, the result is technocratic government by experts with professional qualifications in specialist fields such as medicine, public health and environmental engineering. The insulation of decisions from electoral scrutiny on the grounds that they are technical is open to challenge because what experts regard as a technical matter is often viewed by citizens as an issue of political principle. For example, while economists may treat openness to immigration as a technical matter of labour supply, for citizens immigration is a normative political issue. While liberals tend to favour the multicultural consequences of immigration, conservatives want to protect their national culture from such a radical change. A referendum vote offers a means of making a political decision. In the Brexit campaign, the case for remaining was argued largely in terms of economic costs and benefits, while the case for leave was argued in terms of the priceless value of taking back political control in order to protect the national culture. Importance of an Issue At any one time only a few issues are of major political importance. The more frequently referendums are held, the more likely they are to concern issues of limited importance. In Switzerland, where 238 federal referendums have been held in the past quarter-century, the use of the initiative enables groups with narrow interests to put an issue to a vote that is important to them but not to the general public. The dozen referendums held in Switzerland in 2014 ranged from a vote that approved a policy placing limitations on immigration to a ballot rejecting harmonising the value-added tax on the sale of food by restaurants and take-away shops (Serdült 2018: Table 3.5). When referendums are about issues the public considers unimportant, the turnout is much lower. Referendum issues related to national cultural values tend to be more important to ordinary citizens than issues related to political institutions and regulations. When they occur, citizens can view the issue from their local experience and not just rely on cues from party leaders and elite media. The national interest as viewed from the grassroots of society can

2

HOW REFERENDUMS DIFFER FROM EACH OTHER

29

look different from the view of cosmopolitan governors at the top. If an issue is considered important, ordinary people are likely to have some knowledge about it as well as interest and, therefore, be more likely to vote independently of their party ties (Chapter 5). In the 2016 Brexit referendum, turnout was higher than at any British parliamentary election since 1966. The turnout at the 2014 Scottish independence referendum was higher than at any parliamentary election in Scotland since the introduction of universal suffrage. If a vote is called on an issue that people consider unimportant, the rational response of uninterested citizens is not to bother turning out to vote (Dahl 1970: 40ff.). Turnout tends to vary within a country depending on the issue on the ballot. In Denmark, 87% voted in the referendum on whether the country should adopt the euro as currency, while turnout for the vote on succession to the throne was 30% points lower. In Switzerland, the popular initiative makes it possible for minority interests to put an issue on the ballot, while voters can show their indifference to it by abstaining. In 2012 only 27.6% bothered to vote on an amendment to the law on epizootic diseases. A Vote on a Principle or an Act of Parliament A referendum can ask for a vote about a pre-legislative statement of principle or a vote about endorsing a law already approved by the parliament. A pre-legislative referendum enables voters to decide an issue of principle while giving elected representatives the power to choose how the principle should be implemented. If the proposal wins a majority and is also clear, such as the legalisation of abortion, it leaves elected representatives to determine the specific conditions in which abortions may be held. If there is a majority against change, then no legislation follows. If a referendum is held to approve or reject an Act of Parliament, the choice is very explicit. A post-legislative referendum is not about an abstraction, but about implementing a policy after it has been subject to critical scrutiny by elected representatives. Voters thus have before them the full details of a policy. However, few voters will have the skills or inclination to read a lengthy Act of Parliament. Therefore, campaigners in favour will invoke generalised appeals, and opponents will cherry-pick unpopular clauses that detract from it, such as the cost of putting a principle into effect.

30

R. ROSE

The British government has held both pre- and post-legislative referendums on the devolution of powers to Scotland and Wales. After more than four years of parliamentary debate leading to devolution acts, in 1979 the legislation was then put for approval in Scottish and Welsh referendums. In Wales four-fifths rejected the act and even though a majority of Scots favoured devolution, the act failed because there was also a turnout threshold. In 1997 the Labour government asked for prelegislative approval of the principle of establishing a Scottish parliament, which a substantial majority of Scots gave. It asked Welsh voters if they wanted a Welsh Assembly as recommended by the British government, and 50.3% approved. The votes disarmed critics, and devolution bills were then prepared in Whitehall and approved by the Westminster Parliament without the loss of time and effort invested previously. The 2016 Brexit referendum endorsed the principle of leaving the European Union. The Leave campaigners painted an attractive picture of how it would make Britain great again but had no plan about how to do so. Theresa May was likewise clueless about the challenges of implementing the referendum result when she gained the prime ministership by pledging ‘Brexit means Brexit’. It took more than three years of debate, two prime ministers, and the recurring rejection of Downing Street proposals by the EU and by Parliament before an Act of Parliament could be approved that made clear what Brexit meant in practice. If a vote is called on a principle without attention to how it can be turned into an Act of Parliament, the outcome can be inconclusive. A 1993 Italian popular initiative resulted in a 92% vote to repeal a major electoral law; it could not state what should be put in place of the old law because an Italian initiative cannot propose new legislation. Elected representatives have since spent decades inconclusively trying to achieve a consensus about an alternative (Passarelli 2018: 855ff.). An Unbiased Question If a referendum is to be democratic, the question put to the electorate should not be loaded in favour of the government. Since referendum issues are politically disputed, groups that do not control the wording of the question can make allegations of bias. If opinion polls show that the government-worded question is likely to win, this increases the incentive for losers to complain, but challenges of question-wording approved by parliament are rarely successful.

2

HOW REFERENDUMS DIFFER FROM EACH OTHER

31

National practices vary in the extent to which a referendum question can be assessed for bias independently of the government that proposes it. In the UK, the Electoral Commission has a statutory duty to report on the wording of the question contained in a draft referendum bill. It evaluates a question on the basis of whether it is unambiguous, easy to understand, to the point and not slanted in favour of one or another alternative. Before publishing comments, it can test the proposed text of a question in focus groups with ordinary voters. The UK government bill authorising the 2016 referendum proposed as the question: ‘Should the United Kingdom remain a member of the European Union?’ The Commission recommended removing the response bias in favour of the government’s preferred position and replacing it with an even-handed choice between remaining in or leaving the EU. A Conservative government divided between Remain and Leave supporters accepted the recommendation. A Dichotomous or a Multiple Choice The great majority of referendum questions offer a binary choice. A postlegislative referendum can require a simple yes/no decision about whether the proposed act should be put into effect. A referendum can also offer a choice between two alternatives. The 2016 UK referendum question did this in 16 words: Should the United Kingdom remain a member of the European Union or leave the European Union?

Public opinion surveys often give respondents a choice between multiple alternatives; this gives voters the chance to endorse views excided by a dichotomous choice. The standard Eurobarometer question for measuring popular evaluation of the EU asks: ‘Do you think our country’s membership of the European Union is a good thing, a bad thing, neither a good thing nor a bad thing, or don’t you know?’ Academics measure support for European integration by asking: ‘Some say European integration should be pushed further. Others say it has already gone too far?’ Opinions are then recorded on an 11-point scale giving people the in-between alternative, point 5, maintaining things as they are (Rose and Borz 2016). A multiplicity of alternatives also means that the leading choice in a poll

32

R. ROSE

usually reflects the view of a plurality rather than an absolute majority of respondents. It is technically possible for a referendum to offer multiple alternatives. The simplest way is to make the winner the alternative with the most votes, even if it is not an absolute majority. However, this can cast doubt on the authority of the plurality choice. When the 1980 Swedish referendum on maintaining the use of nuclear power offered three choices, there was a gap of only 0.4% between the two leading alternatives, each of which had less than two-fifths of the vote (Widfeldt 2010: 1863). As of today, most Swedish nuclear reactors are still in use. An absolute majority can be created from a multi-choice ballot by a system of preferential voting in which electors must number alternatives from most to least favoured. The second choices of least-favoured alternatives are then re-assigned until one alternative has an absolute majority, which is composed of first-, second- and sometimes third-choice preferences. The system is good at preventing an alternative that a majority opposes from winning with a plurality (see Chapter 11). New Zealand is unusual in having held two referendums on the New Zealand flag: the first offered a multiplicity of alternatives and the second a binary choice. In December 2015 voters ranked five different designs; after three transfers of votes an absolute majority was given to one alternative. A second ballot was held three months later with a binary choice between the new design and the status quo; the status quo received an absolute majority. Another type of two-ballot system was used to decide New Zealand’s electoral system. In 1992 voters chose between the status quo and four different alternatives. Five-sixths wanted change, and a large majority favoured a mixed-member proportional system. A year later a second ballot offered a binary choice between a fully specified mixedmember system and the status quo; the new system received an absolute majority (see ICE 2018: 105f.). Setting a Turnout Threshold Although the winner must have an absolute majority of votes, this is usually less than half the registered electorate. When the turnout is below 50%, by definition an absolute majority is only a plurality of the electorate. Even if turnout is high, it is unusual for the winner of a strongly contested referendum to come close to getting the vote of half the electorate, unless the vote is held in a country such as Australia, where voting is compulsory

2

HOW REFERENDUMS DIFFER FROM EACH OTHER

33

in referendums as well as parliamentary elections. In the Brexit election, the turnout of 72.2% was higher than at any election since 1992; it was nonetheless only 37.4% of the registered electorate. The definition of democracy as majority rule makes it difficult to require a super-majority to win a referendum, since such a rule could give a minority of voters victory if that level was not reached. To reduce the risk of a limited minority of the electorate deciding a policy in a lowturnout referendum, states such as Italy, the Netherlands and Portugal have set a turnout threshold that must be met for votes on specified issues to be binding. It is typically 50% of the registered electorate; this makes it possible for a referendum to be won with little more than one-quarter of the vote. To guard against this, on some major issues Denmark requires the winning side to have the support of 30–40% of the electorate. Since up to one-third or more of registered electors do not vote in a national election, a turnout threshold gives opponents of an issue an advantage. If it expects to lose, a group can organise abstention among its supporters in order to defeat a proposal by preventing the threshold being reached. British legislation stipulates that Northern Ireland can become part of the Republic of Ireland only if this is approved by a majority of the voters. Irish Republicans, a minority there, reject this view: they advance the absolute claim for the whole of the island to be part of a Dublinbased republic. Thus, when a referendum was held on Northern Ireland’s future, Sinn Fein and the Irish Republican Army told Catholics to boycott the vote and Unionists did the opposite. This resulted in more than 98% of votes being cast in favour of remaining part of the UK. Given the potential abuse of a turnout threshold by a minority, the Council of Europe’s Electoral Commission has recommended against its use because ‘it assimilates those who abstain to those who vote no’ (quoted in ICE 2018: 116). Legally Binding or Advisory A referendum is legally binding if there is an explicit statutory requirement to this effect. A binding vote can be readily implemented if the question is about adopting or repealing an existing Act of Parliament or a European Union treaty. However, most referendums are held without the legal compulsion to implement the result, and a vote on an issue of principle necessarily requires further action by the parliament in order to put the principle into effect.

34

R. ROSE

The advice that referendum voters give is politically binding insofar as governors accept that they ought to carry out the policy endorsed by an absolute majority of voters. Even if Cabinet ministers have doubts about the wisdom of a majority vote, they are likely to implement it in one form or another rather than dismiss it on the grounds that they know better than the people who elect them. The 2016 UK referendum was not legally binding but the government leaflet distributed to all registered electors said: ‘This is your decision. The government will implement what you decide’. In a post-referendum judgement, the UK Supreme Court ruled that the promise to accept the outcome was not legally binding. However, the court did not question the political authority of the government to carry out its promise to implement withdrawal. Even though a majority of MPs favoured remaining in the EU, 498 MPs endorsed formal notification to Brussels of the UK’s intent to withdraw from the European Union. Only 114 MPs voted against. Uncertainty of Outcome The democratic status of a referendum is most clearly demonstrated when a majority of voters reject a government recommendation, an outcome that would not be possible in an authoritarian regime. When the government of the day calls a referendum it is normally confident of victory, but sometimes the result shows that it was overconfident. Even though Charles de Gaulle showed charismatic leadership in becoming president of France, when 53% of referendum voters rejected his plan for regional reform, he resigned as president. When David Cameron’s decision to gain referendum endorsement for Britain’s EU membership backfired, he resigned within hours. When Italian Prime Minister Matteo Renzi failed to convince referendum voters to reform the country’s constitution, he tried to hang on to office but was gone within months. In most referendums in Europe in the past quarter-century, the government’s position is endorsed. Governments win most referendums because they are a biased sample of the major decisions that it takes. If there is a consensus among party leaders that the policy on the ballot is in the broad interest of the country as a whole, this makes approval likely. Actions that would be most at risk of rejection by a popular vote are either not taken or postponed. For example, European Union leaders would like to draft a new treaty giving it greater power to deal with its growing responsibilities. However, since a treaty requires unanimous

2

HOW REFERENDUMS DIFFER FROM EACH OTHER

35

approval by all member states, this has not been done because of the fear that a single national referendum would veto it (see Chapter 3). A variety of theories offer explanations of why governments sometimes lose a referendum and there is statistical evidence that four influences significantly influence whether the government’s position is endorsed by voters (Qvortrup 2018: 34ff.). The chances of approval are reduced the longer a government has been in office. If there is no elite consensus, this will mobilise a bigger vote against the government. Since control of government is not at stake, its normal supporters are free to ignore their party’s endorsement if they disagree with it. Unconstrained by office, opposition parties and movements are free to make simplified and emotive campaign appeals. Since these multiple influences can push in opposite directions, it is the net effect of opposing influences that determines whether the government is on the winning or losing side in a referendum. Complex problems, disputes about facts and uncertainty about consequences create problems for decision-making by elected representatives as well as by referendum voters. However, they are not arguments against an elected government making major decisions. The critical political question is not whether representative or direct democracy is better in the abstract. It is whether a specific issue is better decided by a majority of MPs or by holding a referendum in which it is decided by the citizens who elect them.

References Anckar, D. 2018. Why Referendums? On Appearances and Absences. In The Routledge Handbook to Referendums and Direct Democracy, ed. L. Morel and M. Qvortrup, 107–122. London: Routledge. Dahl, R. 1970. After the Revolution? New Haven: Yale University Press. Hobolt, S.B. 2009. Europe in Question: Referendums on European Integration. Oxford: Oxford University Press. ICE: Independent Commission on Referendums. 2018. Report of the Independent Commission on Referendums. London: University College London Constitution Unit. Mendez, F., M. Mendez, and V. Triga. 2014. Referendums and the European Union. Cambridge: Cambridge University Press. Morel, L., and M. Qvortrup. 2018. The Routledge Handbook to Referendums and Direct Democracy. London: Routledge.

36

R. ROSE

Passarelli, G. 2018. Electoral Systems in Context. In The Oxford Handbook of Electoral Systems, ed. E.S. Herron, R.J. Pekkanen, and M.S. Shugart, 851– 870. Oxford: Oxford University Press. Qvortrup, M. (ed.). 2018. Referendums Around the World. London: Palgrave Macmillan. Rose, R., and G. Borz. 2016. Static and Dynamic Views of European Integration. Journal of Common Market Studies 54 (2): 370–387. Saglie, J. 2000. Values, Perceptions and European Integration: The Case of the Norwegian 1994 Referendum. European Union Politics 1 (2): 227–249. Serdült, U. 2018. Switzerland. In Qvortrup 2018, 47–112. Slune, K., and P. Svensson. 1994. The Danes Said No in 1992 but Yes in 1993: How and Why? Electoral Studies 13 (2): 107–116. Vospernik, S. 2018. Effects of Referendums on Party Cohesion and Cleavages. In The Routledge Handbook to Referendums and Direct Democracy, ed. L. Morel and M. Qvortrup, 433–447. London: Routledge. Widfeldt, A. 2010. Sweden. In Elections in Europe, ed. D. Nohlen and P. Stöver, 1841–1878. Baden-Baden: Nomos.

CHAPTER 3

Europe’s Democratic Deficit and Democratic Surplus Richard Rose

The European Union’s founders learned about politics in the late nineteenth century, long before democracy became the touchstone for legitimacy. They relied on the traditional authority of the state to negotiate agreements with other countries. Robert Schuman, Konrad Adenauer and Alcide de Gaspari saw themselves as trustees of the common interests of Europeans. European institutions were launched for the people, not by the people. Reliance on traditional authority meant that elites saw no need for extensive public debate or a referendum to justify national governments signing the treaties that created European institutions (Haller 2008). In the words of Jean Monnet, the behind-the-scenes organiser of the treaties, ‘I thought it wrong to consult the peoples of Europe’ since they had no practical experience of diplomatic negotiations and politics (1978: 367). European institutions agreed between national governments were legitimated on instrumental grounds: preventing another European war and promoting economic growth. The treaties gave legal legitimacy to the many small-scale policies promoted by technocrats. The cumulative advance of European integration led to the creation of a popularly elected European Parliament in 1979. However, its power to hold policy-makers accountable is less than that of the national parliaments of member states. This has resulted in claims that the EU lacks the legitimacy to impose © The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_3

37

38

R. ROSE

policies on member states because it has a democratic deficit (see, e.g., Follesdaal and Hix 2006; cf. Moravcsik 2002). The European Union paid tribute to the idea of democracy in the 1992 Maastricht Treaty. It conferred European citizenship on all citizens of member states, and Article 10.3 gives a qualified commitment to the rights of European citizens, declaring: ‘Every citizen shall have the right to participate in the democratic life of the Union. Decisions shall be taken as openly and as closely as possible to the citizen’. In response to applications for membership from more than a dozen post-communist countries, the EU’s Copenhagen criteria have set out democratic principles that applicant countries should meet to qualify for EU membership. The legitimacy of government in the twenty-first century depends on it being democratically elected. National governments are democratic because they are accountable to parliaments elected by universal suffrage. Citizens thus have multiple means of showing their dislike about what is done in their name. They can vote to remove from office a national government that has endorsed an EU policy that is nationally unpopular and they can elect Eurosceptic MEPs. Citizens can also vote for referendum measures that challenge the EU’s authority. It is impossible for the European Union to have the same democratic legitimacy as that of member states, because the obstacles to a democratic EU are structural. In the election of the European Parliament, the EU cannot apply the basic democratic principle of one person, one vote, one value, because of gross inequalities between the population of member states. In any case, the EU is not a creation of the people of Europe. It is a league of member states, all of which are equal in law regardless of their population. The treaties that serve as its constitution are created by deliberation and bargaining between leaders of member states. Policies that the multi-national Council of Ministers and Parliament approve are binding on all European citizens, whether or not they have given a mandate to their nationally elected government and MEPs to approve EU actions. EU policies cannot be altered by the outcome of an election. The EU doctrine of the acquis communautaire means that once a policy or treaty is adopted it is not subject to repeal, unlike a national constitution or national Act of Parliament. British prime ministers have learned this the hard way. David Cameron promised Conservative MPs that he would secure the reduction of EU policies affecting the UK; he failed. Theresa May lost office because she could not get the EU to concede that Brexit

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

39

should mean what Conservative MPs wanted it to mean: getting rid of many EU regulations and constraints affecting British policies. The European Union’s political system is best described as ‘democracy lite’ (Rose 2015: Chap. 4). Given structural obstacles to full democratic representation of its citizens, the European Union is not a political system that is democratising; it is a system with a built-in democratic deficit.

3.1

The EU’s Structural Deficit

A state requires more than free elections to be fully democratic. Its elected government must also respect the rule of law and be subject to checks and balances between its political institutions. EU treaties grant equal legal rights to every member state and give smaller states disproportional voting strength in policy-making. Smaller EU member states are collectively insistent on enforcing procedures and laws that favour them against abuse by the largest member states. EU permanent officials prefer to be called bureaucratic because strict adherence to EU laws is their best protection from undue influence by national governments wanting to bend laws to their national advantage. In cases of major dispute, the Court of Justice of the European Union has the power to make and enforce decisions to prevent the violation of EU laws. Within the European Union system, there are a substantial number of horizontal checks and balances. The adoption of policies requires the approval of both the Council of Ministers of national governments and the European Parliament. Within each institution, there are checks to prevent dominance by one or a few populous countries. The political heads and supranational civil servants of the European Commission propose and administer EU laws, but their actions are subject to monitoring by other EU institutions. The checks are so numerous that ambitious EU officials claim that they delay or prevent necessary decisions being made. Horizontal checks and balances pioneered in seventeenth-century England to restrain the power of a hereditary monarch are insufficient to qualify a political system as democratic. The ideal model of democratisation is that a political system has institutionalised horizontal accountability and the rule of law before making its government vertically accountable to universal suffrage. This was the case in older democracies of Scandinavia and the Anglo-American world. To be fully democratic, a political system needs the vertical check provided by elections held on the basis of one person, one vote, one

40

R. ROSE

value. The EU cannot hold such elections because it is a union of states extremely unequal in population. Germany has more than 150 times the population of Malta and almost ten times the population of states with the median population, Portugal and Sweden. Smaller states are the norm; more than two-thirds have national populations below the EU arithmetic average of 16.5 million. Smaller states are very unequal in population too. They range from Belgium, the ‘biggest’ of the below-average states, to eight countries each of which has 1% or less of the EU’s population. Four states—Germany, France, Italy and Spain—collectively have a majority of the EU’s population. To protect the majority of member states from domination by a few countries, EU treaties confer many rights of representation equally on each member state regardless of its population. In the European Council, nationally elected heads of governments collectively discuss EU policies of highest importance. Each state has one seat which is taken by its prime minister. In parallel, the Council of Ministers deals with issues of political concern to a specific government department. The government is represented there by its appropriate national minister. When the subject is more technical, informed national civil servants represent each member state. Gross inequalities in population between member states mean that the Council does not take decisions by a majority vote of states nor are votes weighted by population. Instead, endorsement of a policy depends on a super-majority weighted by both the number of states and by population. In a 27-member EU, a minimum of 15 states must give their approval. This means that two-thirds of the votes cast to approve a policy must come from states ranging in population from average to tiny. In addition, approval requires endorsement by states that collectively have at least 65% of the EU’s population. Thus, a policy can be adopted only if it is supported by the biggest states and by many small states. Unanimous approval is required for a small number of major issues, such as treaties and the admission of a new member state. The unequal representation of citizens and the requirement of unanimity are inconsistent with the democratic practice within member states. The significance of inequalities in Council votes is mitigated by the great majority of decisions being taken by a consensus arrived at without a vote. Before putting a proposal to the Council for approval, Commission staff sound out country representatives most likely to be affected in order to modify features that they would object to. Upwards of fourfifths of Commission proposals have so little political impact that they are

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

41

agreed by committees of national civil servants without reference to their national ministers. If a policy has features that are politically disputed, national ministers are brought together to see if contentious points can be resolved. When Council members discuss major policies, participants are well aware of the scale of resources that big states have and of the seriousness of a particular issue to some small states. The upshot is that a bargain usually gives almost all states something but not everything that they want. If there is no consensus, then Council members can recommend that a proposal be withdrawn to avoid a bitter conflict (Thomson 2012). The European Commission is the approximate equivalent of the executive branch of a national government. Each member state has the right to nominate one Commissioner to head a directorate that is the equivalent of a Cabinet ministry in a national government. Each directorate is staffed by a multi-national civil service committed to European integration and recruited through rigorous competitive examinations (see Kassim et al. 2013). The Commission is responsible for preparing major policies consistent with the guidelines laid down by the Council and subject to approval by the European Parliament. Four-fifths of the EU’s laws and regulations are technical and limited in impact; they are approved by discussions between the Commission and national civil servants. Cumulatively, the spillover of lots of little regulations advances European integration (Niemann and Schmitter 2009). For this reason, British advocates of leaving the EU have attacked the Commission as an undemocratic institution imposing laws that make Britons vassals of Brussels. The European Parliament (EP) was initially created as a multi-national assembly of national MPs, but since 1979 members of the European Parliament (MEPs) have been elected in 27 national constituencies. For MEPs to represent the views of voters, a significant majority of the electorate ought to vote. At the first EP election, the nine member states saw a turnout of 62%. Turnout then declined, falling below half the electorate in 1999 and further with the enlargement of the EU to its present size. In 2014 it hit its lowest level, 42.6%, before rising to 50.6% at the 2019 election, a figure still well below national levels. Until the past decade, low turnout has been interpreted as indicating passive and permissive consensus to EU policies (Lindberg and Scheingold 1970: 41). Although the European Parliament is popularly elected, the formula for allocating MEPs to countries is a structural obstacle to the equal representation of European citizens. Seats are allocated by a system of degressive

42

R. ROSE

proportionality that turns proportional representation upside down (Rose 2015: 102ff.). The smaller the population of a member state, the fewer the number of electors required to obtain a seat; moreover, the least populous states are each guaranteed a minimum of six MEPs. The result is disproportional representation. In the 2019 European Parliament election, Malta had one MEP per 70,000 people while Germany had one MEP for 860,000 people. Twenty-two member states had more MEPs than would have been the case if seats were allocated by the proportional representation of European citizens. Within each national constituency, the EU requirement of proportional representation fragments the representation of national interests among up to half a dozen parties. The combination of proportional representation within national constituencies and disproportional representation between countries results in upwards of 200 different national parties being represented in the European Parliament. However, only seven parties have won as much as 33% of their national vote, and only four parties have returned at least two dozen members. The two largest national parties after the 2019 election, each with 29 MEPs, were the German Christian Democrats and the British Brexit Party, which left the EP when the UK withdrew from the EU. In order to gain recognition and resources, nationally elected MEPs must join a multi-national Party Group that has at least 25 MEPs. The requirement that a Party Group has MEPs from at least one-quarter of member states reinforces the multi-national outlook of MEPs. The three largest Party Groups, the European People’s Party, the Socialists and Democrats and the Liberal Renew Europe Group, each have MEPs from more than 20 countries, whereas the Eurosceptic Groups have members from half the member states or less. The policies of Party Groups are created by aggregating the views of MEPs who have won office by campaigning on a variety of national programmes. Within a Party Group, the national manifestos of their MEPs tend to show agreement in support of or opposition to further European integration, but limited agreement on other major dimensions of policy. This is particularly the case for the EP’s largest group, the People’s Party. For all Groups, immigration and social welfare are the issues on which the national manifestos of MEPs are most likely to disagree with the policy of their Party Group (Rose 2015: 126).

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

43

When there is a conflict between national public opinion and the policy of their Party Group, the MEPs in a Group tend to vote as a multinational bloc rather than divide along national lines. The re-election of individual MEPs depends less on how they vote in the European Parliament than on the domestic popularity of their national party. The incentive to give priority to the Party Group’s policy is strong, since individual MEPs benefit from belonging to a Group that gives them the potential to leverage the influence of their national party. Even though MEPs campaign by speaking to their voters in their national language, the great majority represent their constituents by working in a foreign language in a foreign country and voting for policies reflecting the consensus of a Party Group dominated by foreign parties. Adding Weight to Democracy Lite? Evaluating the democratic character of the EU by the standards of national governments, let alone those of normative political theories, justifies its description as light on democracy. This judgement encourages proposals to introduce reforms to give citizens more opportunities to make inputs to the EU’s policy process. Since 1973 the European Commission has sponsored the Eurobarometer, a multi-national survey of public opinion about EU institutions and activities (http://ec.europa.eu/commfrontoffice/publicopinion/). Its purpose is practical market research to identify means for promoting commitment to the European Union and its policies. Surveys thus ask questions about popular demand for EU services and how popular the EU is. Answers tend to show a plurality of positive responses; criticism is diluted because of the large number of respondents with no opinion about the European Union. Unlike academic surveys, the Eurobarometer is not allowed to ask questions about respondents’ party preferences. In response to pressures to give European citizens a direct say on EU policies, in 2012 the EU introduced what it called the Citizens’ Initiative. Citizens can petition the European Commission to consider an issue as long as it is within its existing powers. If the request is endorsed by a million citizens in at least seven member states, the Commission is obliged to publish a reasoned reply to the petition. The EU initiative cannot propose fresh policies nor can it initiate a referendum, as is the case in many member states. To date it has not been much used: only four petitions have been successfully filed and to little effect. In the Commission’s words,

44

R. ROSE

‘The Citizens’ Initiative is an agenda-setting initiative which obliges the Commission to give serious consideration to requests made by citizens, but it is not obliged to act on them’ (http://ec.europa.eu/citizensinitiative/public/faq). The subtitle of a major academic review of this EU initiative is The Myth of a Citizens’ Union (Blockmans and Rusack 2019). Party Groups sought to increase engagement with the EP electorate in 2014 by each nominating a Spitzenkandidat (German for leading candidate) for the presidency of the European Commission, a power formally within the hands of the European Council but subject to EP confirmation (Christiansen 2016). Given linguistic and geographical obstacles, there was no campaign comparable to that for electing a national president. After the election, the Party Group that won the most seats, the middleof-the-road European People’s Party, nominated as its candidate for the presidency Jean-Claude Juncker. After hesitating, the Council formally nominated Juncker, who had sat in the Council for 18 years as prime minister of Luxembourg. In 2019 the EPP was reduced to less than onequarter of MEPs, but it remained the largest party. Its Spitzenkandidat, Manfred Weber, a German, had no experience of the Council as he had never held a ministerial post in his national government. The Council rejected him and nominated a German woman, Ursula von der Leyen, who had held ministerial posts in the German government. Making the whole of the EU a single constituency so that European citizens could vote in common for a transnational list of several dozen candidates has been mooted for decades (Duff 2018). The list would be required to have candidates and seek votes from a variety of countries, unlike the nationally elected MEPs that currently constitute the whole of the EP’s members. In an EP election campaign, Party Groups would have an incentive to campaign for a transnational list in their own name across Europe. When raised by an EP committee in 2018, the idea of MEPs elected by a pan-European rather than a national constituency was rejected by a majority. Formally linking national parliaments with the European Parliament has been proposed as another way to reduce the EP’s democratic deficit. However, most national parliaments already have committees to scrutinise policies and regulations coming from Brussels, and MPs can apply pressure on their national government to object to a particular policy in the European Council. To give national parliaments power to vote on EP measures raises the same problem as allocating seats: How would the

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

45

votes of national parliaments be weighted? By analogy with federal systems, adding a second chamber to the European Parliament to represent each member state would face the same structural problems as the existing unicameral chamber (Rose 2012). In response to the EU almost doubling in size by admitting countries from Eastern Europe, in 2000 Romano Prodi, then the president of the European Commission, proposed measures to remove the ‘paralysis’ that he said was preventing EU decision-making, especially the right of a single member state to veto a proposal. Changes advanced included deciding the great majority of issues by majority voting and capping the number of EU Commissioners and Directorates. Since Prodi’s proposals would have greatly limited the influence of member states, both large and small, they were rejected. Prescriptions to strengthen the role of citizens in the EU policy process assume that its political problems are due to a lack of policy inputs rather than caused by its policy outputs (cf. Scharpf 2009). If decisions about EU policy decisions are viewed as technically complex, then giving power to technocrats rather than voters is a logical alternative. This has been done in the European Central Bank, where policy formation is in the hands of economists. However, making the EU more technocratic risks trouble if the promised benefits of expert policy-making are not delivered. Proposals to add weight to the democracy-lite institutions of the EU face both structural and political obstacles. Gross disparities in population and economy between member states make it impossible to give European citizens the equality of representation that they enjoy in their national parliament. Population inequalities also prevent granting each member state a single vote in the European Council. Political opposition to reform comes from existing holders of power.

3.2

A Surplus of Elections Dozens of National Elections

To participate in EU policy-making, national politicians must first win a national election. Whereas European Parliament elections are held once every five years, a half-dozen or more national elections are held each year in member states, and many countries have significant regional elections too. Since national politicians focus on a forthcoming election well before an official campaign starts, the behaviour of voters casts a long shadow

46

R. ROSE

on national governments. Around-the-clock monitoring by old and new social media increases the intensity of national electoral pressures. Since the party winning control of the premiership has an ex officio seat in the European Council and the Council of Ministers, national elections influence what the EU does. The effect may favour more integration, as in the election of an actively pro-European Emmanuel Macron as president of France in 2017. It may be ambiguous, as in the Netherlands, where anti-EU parties can affect the formation of a coalition government. The 2018 Italian election placed government in the hands of two populist parties that challenge the EU’s authority for their domestic electoral advantage. In Britain, national electoral pressures have paved the path to Brexit. National Referendums on EU Issues Because a referendum decision is the result of a majority vote about a binary ballot choice, it differs from making an EU policy. An EU policy is adopted by consensus after bargaining between national governments in deliberations that get little public attention or understanding. Many governments can claim to be winners thanks to this process and few are labelled losers. By contrast, victory or defeat in a winner-take-all referendum comes after a contested and very public campaign. Consistent with the European Union being a union of states, there is no provision in its treaties for a referendum in which European citizens collectively decide a policy. The right to call a referendum on EU as well as domestic affairs is a subsidiary power of member states. Since all treaties require the unanimous approval of member states, this gives citizens of a single state a veto. It also enables a national government to threaten to call a referendum to veto. When Turkish membership of the EU was being actively discussed, the French government introduced a constitutional clause requiring a national referendum on the admission of any new EU member state. The well-understood implication was that such a vote would veto Turkish membership. In 2001, the Laeken Declaration of the European Council declared that ‘citizens undoubtedly support the EU’s broad aims’, but there was a need to become more democratic in order to bring citizens closer to the European design of an ever closer Union. Towards this end, it authorised a Convention to draft a Constitution for the European Union. European citizens did not elect the members of the Convention. Instead, they

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

47

were appointees representing national governments and parliaments along with EU institutions. The text was drafted by a small executive committee chaired by the forceful Valéry Giscard d’Estaing, the former president of France (Castiglione et al. 2007). The first draft of the Constitution’s preamble echoed Rousseau and prefigured twenty-first-century populist rhetoric, ‘Power is in the hands not of a minority but of the whole people’. When a small number of Convention members suggested asking European citizens to endorse the Constitution by a referendum of the whole of Europe’s people, the proposal was summarily rejected. National governments took the opposite view. Most authorised referendums giving their citizens a chance to vote on the Constitution drafted in their name. This created the paradoxical situation of the major stakeholder in the outcome, the European Union, being a bystander in balloting on its future, since the principle of subsidiarity bars EU institutions from participating in national elections. National referendums showed that European citizens were divided about accepting the European Constitution (Fig. 3.1). Referendums held in Spain and Luxembourg produced majorities in its favour. However, the reverse was the case in France, 55% against, and in the Netherlands 62% of Dutch voters rejected the Constitution. The remaining countries that had planned to call a referendum abandoned doing so since the Constitution was politically dead. There have been 14 national referendums on EU policies since 2005 (Fig. 3.1) and most have been of major political significance. In addition to the unprecedented British decision to leave the EU, there have been two Irish votes on the Lisbon Treaty in 2008/9; Greek and Irish votes on fiscal austerity measures; and a Hungarian referendum that rejected an EU immigration policy. In nine ballots an absolute majority of national voters have rejected an EU measure, while in five the EU position has been endorsed. The outcomes cannot be dismissed as unrepresentative because of a low turnout. The percentage voting in national referendums on EU issues averages more than 12 points higher than the turnout in the country’s European Parliament vote. When national referendums give citizens a vote on a Brussels policy described as being for the good of all, there is no consensus. The median referendum result is a division with 52% for the anti-EU alternative and 48% for the EU position. EU defenders can point out that the vote in a single national referendum represents only one member state in which up to 99% of EU citizens are unable to vote (Rose 2015: 93). Whatever the

48

R. ROSE

% anti-EU vote 98%

100%

62%

61%

55%

53%

50%

50%

53% 64%

52%

38%

44% 33%

40%

23%

0%

Fig. 3.1 National referendums show no consensus on EU (Source Author’s compilation from national sources)

size of the national electorate or the outcome, the division of the votes in national referendums shows that there is no permissive consensus in favour of major EU policies.

3.3

Coping with Conflicting Legitimacies

Interdependence has eroded the old distinction between first-order national politics and second-order European politics. Major problems facing political institutions can no longer be neatly divided into national and European issues. They are ‘Euromestic’ when problems have both domestic and European causes. The outcome of an interdependent policy depends not only on decisions taken in a national capital, be it Paris or Copenhagen, but also on decisions taken by EU authorities in Brussels and by institutions further afield. The location of influences changes with the issue: energy prices are affected by what happens in the Middle East, and revenue from tourism on the number of tourists and the exchange

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

49

value of the euro. Policies that depend solely on domestic decisions, such as marriage and divorce laws, are a decreasing proportion of the major issues facing national governments in Europe today. In dealing with interdependence, all national democracies are bounded democracies (Rose 2014). Interdependent problems require institutions for collective action that complement national policies. This was a fundamental assumption of the founding states of the European Union, a view rejected by Britain’s Brexit government. The leaders of small states acknowledge this as a fact of life and welcome opportunities to band together as the Benelux countries, the Nordic Council and the Baltic states often do. They know from experience that whatever is decided nationally will be indecisive unless it takes into account or is co-ordinated with what other countries do. The attraction of the European Union is that it offers every member state a chance to participate in EU policymaking. Euromestic policy-making involving EU institutions that have a democratic deficit and those national governments that have a democratic surplus create conditions in which conflicts of legitimacy are possible but not inevitable. If national citizens elect a parliament that accepts the EU’s treaty authority and national public opinion of an EU policy is favourable, there is no conflict. However, if EU policies are viewed unfavourably, citizens can elect Eurosceptic MEPs as a protest against EU influence on their nationally elected government. Eurosceptic parties have a political incentive to create conflict with parties that have participated in making decisions in Brussels that are domestically unpopular. The EU Strategy: Conflict Avoidance EU policy-makers have sought to pursue a policy of conflict avoidance. For decades this was anything but a ‘do nothing’ policy, as European institutions advanced integration by stealth, adopting small-scale, politically invisible policies that cumulatively have had a substantial effect on national societies (Haas 1958). Policy-makers assumed that the unawareness of national voters gave them permissive consensus to do so. European integration has increased the possibility of conflicts between national and European institutions by increasing the national impact and visibility of EU decisions. Major agreements between national governments in the 1980s created a single European market and the free movement of people across national borders. These steps also created conflict

50

R. ROSE

among European elites about the EU’s future. Social democrats such as Jacques Delors, then the European Commission president, saw the single market justifying an increase in the EU’s influence to ensure that it was a single market. Margaret Thatcher argued the opposite in a speech at Bruges. The creation of the Eurozone in 1999 gave the technocratic European Central Bank control of the currency of most member states, while control of public expenditure and taxation has remained in the hands of national governments reliant on the euro. The economic crisis of 2008 has created popular dissatisfaction with Eurozone policies. The free movement of peoples across national borders is a major EU responsibility, especially when immigrants come from Middle Eastern and African cultures. Popular rejection of the proposed European Constitution reinforced the EU view that referendums are undesirable. One way to do this is to avoid taking decisions that a single national referendum could veto because unanimity is required. Another way is for a policy to be adopted by a coalition of the willing, resulting in policy-making by ‘differentiated integration’. EU rules allow a minimum of 11 member states that want to adopt a new policy to do so. National governments that do not want to participate can opt out but cannot block the move (Leuffen et al. 2013). While differentiated integration may be used for many policies, including the Eurozone, it cannot be used for treaties that require unanimous approval. To finesse the referendum rejection of the draft European Union Constitution, policy-makers consulted with national governments about what changes were needed to avoid another round of national referendums on a treaty expanding the EU’s powers (Oppermann 2013). Many changes requested were cosmetic, for example, dropping symbols of European statehood such as an official flag and anthem, and calling the new document a treaty rather than a Constitution. The powers conferred on the EU remained much the same. This resulted in all but one national government signing what became the Lisbon Treaty without calling a referendum. Only Ireland was constitutionally obligated to hold a referendum. The government’s confidence that Irish voters would follow their lead was punctured by defeat in a June 2008 ballot but approved in a second ballot after the Irish government negotiated a few treaty changes. Referendum avoidance has subsequently enabled more steps towards integration to occur by making broad use of new powers granted by the Lisbon Treaty.

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

51

The EU cannot avoid responding to national referendums that explicitly challenge the EU’s authority (see Rose 2019). It does so by invoking the legal authority of its treaties to show a national government the negative consequences for a country of fully implementing a referendum decision. The Swiss government was told that if it implemented a referendum vote to curb immigration this would result in the repudiation of a number of bilateral Swiss–EU agreements. The Swiss government backed down. In response to Brexit, the EU presented the UK government with the choice of withdrawing on terms set in Brussels or withdrawing with no deal. The 2016 Hungarian referendum rejecting an EU policy on immigration has been a persisting case of conflict avoidance. The EU initially did nothing rather than precipitate a conflict between its treaty legitimacy and the electoral legitimacy claimed by the Eurosceptic government of Viktor Orban. As the Hungarian government has persisted in challenging the EU, different EU institutions have challenged what has been done nationally as appearing to violate the EU’s democratic standards. Containing Eurosceptics Depending on the measure used, Eurosceptics are between one-fifth and one-third of Europe’s citizens (cf. Hawkins et al. 2019). While this makes them a minority, Eurosceptics are upwards of 100 million or more European voters. Thus, Eurosceptic politicians hold seats in the European Parliament and the European Council, where they can voice their opposition to advancing European integration or demand that it be reduced. EU policy-makers do not want to engage in a continuing debate about the direction the EU should take. They want to neutralise Eurosceptics in order to adopt policies that deal with challenges of interdependence. Their majority in the EU’s decision-making bodies enables them to do so. To gain a voice in the European Council, Eurosceptics first need to win control of their national government. In the past decade, Eurosceptic prime ministers have sat in the Council representing Greece, Hungary, Poland and Italy. A single member can veto the adoption of treaty. Since the Lisbon Treaty came into effect in 2009 this has stopped a new round of treaty-making. However, since reducing the existing powers of the European Union would require treaty reform, Eurosceptics cannot roll back the current state of European integration. The presence of more than one Eurosceptic prime minister in the Council protects them from

52

R. ROSE

sanctions for breaking EU rules that require virtual unanimity. However, Eurosceptic governments would need to have a substantial minority of places on the Council to stop super-majorities approving the adoption of pro-integration measures. For MEPs to defeat a specific proposal put forward by the European Commission would require a substantial number of pro-integration MEPs to join with Eurosceptics in order to create the absolute majority needed to oppose a proposal. If this happened it would reflect not so much the strength of Euroscepticism but Commission staff showing political misjudgement by ignoring widespread opposition to a particular measure. Unlike the British Brexit Party, the majority of Eurosceptic MEPs are ‘soft’ Eurosceptics, that is, they want European institutions to have less influence on their country rather than advocating withdrawal from the EU or the Eurozone (Brack 2019). Since they have places on all the EP committees considering new legislation, individual Eurosceptic MEPs can try to influence legislation by promoting amendments that may be adopted if they are seen as improving rather than wrecking a proposed policy. Because control of national government does not depend on an EP election, Franklin and Nielsen (2017) have argued that this makes it a first-rate expression of how people feel about Euromestic issues. Citizens can cast a vote for opposition parties that raise issues that have been ignored by governing parties that share responsibility for EU policies. Doing so is not wasting a vote but sending a warning shot to national and EU policy-makers to pay attention to Euromestic issues such as unemployment and immigration. The 2019 European Parliament election came at a time when Euromestic issues were particularly salient. In consequence, turnout rose to 51%, the highest level in a quarter of a century and in half the member states a majority of registered voters cast a ballot. The collective outcome has reduced the dominance of the pro-integration European People’s Party and the Socialists and Democrats, which together had had an absolute majority of seats for four decades. While the parties remain first and second in size, because each lost more than 30 seats they no longer command the absolute majority needed to approve legislation (Table 3.1). To a significant extent, their loss of MEPs is compensated for by an increase in the strength of the other pro-integration parties, the Liberal Renew Europe Group and the Greens/Free Alliance. Altogether, the four pro-EU Party Groups have more than two-thirds of the seats, but four-party negotiation is extremely time-consuming. All

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

53

Table 3.1 Party groups in the European Parliament, 2019 Seats N Pro-EU People’s Party Socialist and Democrats Renew Europe/Liberals Greens/Free Alliance Total Eurosceptics Identity and Democracy European Conservatives and Reform United Left and Nordic Green Not attached Total

Biggest EP Party, vote %

182 154 108 74 518

24.2 20.5 14.4 9.8 69.0

Ger: CDU-CSU 29% Spain: PSOE 33% Fr: RM:Macron 21% German: Green 21%

73 62 41 57 233

9.7 8.3 5.5 7.6 31

Italy: La Lega 34% Poland: PSI 45% Greece: Syriza 24% UK: Brexit 31%

Source www.europarl.europa.eu/about-parliament/en/organisation-and-rules/organisation/politicalgroups, 19 November 2019

but the Socialists and Democrats are internally divided on many major issues. The largest Group, the EPP, includes socially conservative parties with a Catholic foundation and secular Nordic parties with a more market-oriented outlook. Moreover, the EPP has tolerated Fidesz as a member, even though its Hungarian leader, Viktor Orban, advocates what he describes as a programme of illiberal democracy. The readiness of Party Groups to support further integration has created a political backlash. Collectively, Eurosceptic parties won 31% of the EP’s seats, a slight gain from the 2014 election (Table 3.1). However, this total is made up of more than 40 national parties that share little except a readiness to campaign by attacking the EU as an institution and the source of unpopular Euromestic policies. Many of them are described as populist, itself a broad term that can be subdivided into the populist right, the populist left and other populists (Rooduijn 2019; Fella et al. 2019: 66ff.). Eurosceptic parties also differ: some participate in national government coalitions and some are treated as pariahs because of their undemocratic views, such as the Alternative for Germany (AFD) and the party of Marine Le Pen in France. This, plus clashes of personalities and ideological differences unrelated to Europe, divide Eurosceptic parties into three Groups. There is the Identity and Democracy Party, led by the Italian La

54

R. ROSE

Lega and including pariah parties from France and Germany; the Conservative & Reform Group for national parties to the right of the People’s Group; and the United Left and Nordic Greens, a Group to the left of mainstream Socialists. The Euroscepticism of almost one-third of MEPs qualifies the claim that EU institutions represent all its citizens. An important feature of the 2019 election was that a substantial majority of Eurosceptic parties endorsed a ‘soft’ Eurosceptic position compared to a small number favouring ‘hard’ Euroscepticism. Only nine MEPs in favour of leaving the EU were returned in five member states plus the 29 UK Brexit Party MEPs. The rhetoric of soft Eurosceptics endorses reforming the EU from within (Brack 2019). Specific reforms include changing Euromestic policies on immigration and austerity and leaving national governments more scope for making Euromestic policies for which the EU has concurrent jurisdiction. Soft Eurosceptics returned 185 MEPs. Containing Eurosceptic parties at the EU level does not get rid of the fundamental conflict between MEPs and the European electorate. More than two-thirds of MEPs favour further integration but only one-quarter of European voters do so. The median position, taken by one-third of respondents, favours keeping things as they are and almost one-third want less integration. Moreover, less than half of voters agree with the policy on EU integration of the party for which they vote in an EP election (Rose and Borz 2016: Figure 2). Prime Ministers Face the Goldoni Problem The European Union treats the need to secure national consent to its policies as a subsidiary responsibility of national governments. Consistent with international law, the assent to a treaty by a national government of the day is seen as binding the whole of a country’s population, including all those who voted for the government’s opponents. Furthermore, the EU’s acquis communautaire binds a member state’s current government to treaties signed by a prime minister who left office up to half a century ago. National prime ministers cannot treat the views of their voters as subsidiary. Nor do they have the right that voters have of backing a single choice. Instead, they are like the central character in the eighteenthcentury Italian play, The Servant of Two Masters by Carlo Goldoni. In

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

55

their national office prime ministers are accountable to their national electorate and in a coalition government to other parties as well. As a European Council member, a national prime minister must serve the collective interest of prime ministers of 26 other member states too. This can be done by participating in a multi-national coalition. To ignore this responsibility risks marginalising the country’s influence on interdependent policies that no national government can resolve on its own (Rose 2015: Chap. 8). Yet to ignore accountability to national voters risks losing the office that makes him or her eligible to participate in collective EU decisions. As Jean-Claude Juncker, former prime minister of Luxembourg, is reputed to have said about EU policy-making, ‘We all know what we ought to do. We just don’t know how to get re-elected after we’ve done it’. Prime ministers must be nimble in dealing with interdependent policies involving Brussels as well as their national society. As long as an EU policy is not unpopular nationally, a prime minister can readily serve two masters. However, interdependence is increasing the national visibility of collective decisions taken in Brussels. While a prime minister can claim exclusive credit for a popular policy, she or he finds it hard to escape blame for policies that are domestically unpopular. The pressure of having two masters is particularly strong for Eurozone policies affecting the national economy and EU policies that result in immigration from nonEuropean cultures increasing. In smaller European states, the problem of serving two masters is taken for granted, since their prime ministers have long been decision-takers rather than decision-makers. A politician who promotes a European policy without a significant national political base risks losing office. This happened to Mario Monti, an Italian academic economist and former European Commissioner, who became Italy’s prime minister in 2011 without ever having held an elected office. The appointment carried with it responsibility for implementing EU austerity policies, a task that elected Italian politicians had shunned. In the national election held after Monti had done so, his party secured only 10.5% of the vote and Monti subsequently left office. Referendums in Norway and Switzerland have shown which master is immediately stronger by voters rejecting EU association on terms endorsed by their national government. However, this has not got rid of the Goldoni problem, because the economies of the two countries are integrally related to other European countries. Hence, the Norwegian and Swiss national governments have made treaties with the EU

56

R. ROSE

to advance their national interest by having access to the Single Europe Market. However, instead of participating in making EU policies through democracy-lite institutions, Norway and Switzerland are excluded from EU institutions making policies they are bound to accept. For British politicians raised on the nineteenth-century doctrine of parliamentary sovereignty, the idea of sharing policy-making powers is alien. It was the reason Winston Churchill resisted becoming a founding partner in European institutions in the 1950s. Subsequent prime ministers have invested little political capital in promoting the political benefits of collective policy-making for Euromestic issues. However, the experience of Downing Street in negotiating withdrawal terms has not got rid of the Goldoni problem. The failure to serve two masters costs both Cameron and May their job as prime minister. Boris Johnson has dealt with the problem by first accepting EU conditions for leaving the EU and then telling British MPs that these terms represented what Brexit had meant even though this was far from the case. Subsidiarity gives citizens of a member state the unilateral power to hold their national government accountable through a national referendum on a Euromestic issue that their prime minister has agreed in their name in Brussels. Calling such a referendum faces a prime minister with the risk of becoming a piece of rope in a tug of war between a majority of national citizens and the democracy-lite institutions of the European Union.

References Blockmans, S., and S. Rusack (eds.). 2019. Direct Democracy in the EU: The Myth of a Citizens’ Union. London: Rowman & Littlefield. Brack, N. 2019. Eurosceptic Parties at the 2019 Election: A Relative Success. In Euroflections, ed. N. Bolin, K. Falasca, M. Grusell, and L. Nord, 64–65. Sundsvall, Sweden: Mittunivesiteter. Castiglione, D., J. Schönlau, C. Longman, E. Lombardo, N.P.-S. Borragan, and M. Aziz. 2007. Constitutional Politics of the European Union. Basingstoke: Palgrave Macmillan. Christiansen, T. 2016. After the Spitzenkandidaten: Fundamental Change in the EU’s Political System? West European Politics 39 (5): 992–1010. Duff, A. 2018. How to Govern Europe Better: Reflections on Reform of the European Parliament, Commission and Council. European Policy Centre Discussion Paper, Brussels.

3

EUROPE’S DEMOCRATIC DEFICIT AND DEMOCRATIC SURPLUS

57

Fella, S., E. Uberoi, and R. Cracknell. 2019. European Parliament Elections 2019. House of Commons Library Briefing Paper 8600, London. Follesdaal, A., and S. Hix. 2006. Why There Is a Democratic Deficit in the EU. Journal of Common Market Studies 44 (3): 533–562. Franklin, M.N., and J. Nielsen. 2017. The 2014 Elections as a Lens on Euroscepticism. In The Eurosceptic European Parliament Elections 2014, ed. J. Nielsen and M.N. Franklin, 239–254. London: Palgrave Macmillan. Haas, E.B. 1958. The Uniting of Europe: Political, Social, and Economic Forces, 1950–1957. Stanford: Stanford University Press. Haller, M. 2008. European Integration as an Elite Process. London: Routledge. Hawkins, K., R. Carlin, L. Littvay, and C.R. Kaltwasser (eds.). 2019. The Ideational Approach to Populism. London: Routledge. Kassim, H., J. Peterson, M.W. Bauer, S. Connolly, R. Dehousse, L. Hooghe, and A. Thompson. 2013. The European Commission of the Twenty-First Century. Oxford: Oxford University Press. Leuffen, D., B. Rittberger, and F. Schimmelfennig. 2013. Differentiated Integration. Basingstoke: Palgrave Macmillan. Lindberg, Leon, and Stuart Scheingold. 1970. Europe’s Would Be Polity. Englewood Cliffs, NJ: Prentice-Hall. Monnet, J. 1978. Memoirs. New York: Doubleday. Moravcsik, A. 2002. Reassessing Legitimacy in the European Union. Journal of Common Market Studies 40 (4): 603–624. Niemann, A., and P. C. Schmitter. 2009. Neofunctionalism. In European Integration Theory, ed. A. Wiener and T. Diez, 2nd ed., 45–66. Oxford: Oxford University Press. Oppermann, K. 2013. The Politics of Discretionary Government Commitments to European Integration Referendums. Journal of European Public Policy 20 (5): 694–701. Rooduijn, M. 2019. How to Study Populism and Adjacent Topics. European Journal for Political Research 58 (1): 362–372. Rose, R. 2012. Representation in Parliamentary Democracies. In Multilayered Representation in the European Union, ed. T. Evas, U. Liebert, and C. Lord, 73–90. Baden-Baden: Nomos. Rose, R. 2014. Responsible Party Government in a World of Interdependence. West European Politics 37 (2): 253–269. Rose, R. 2015. Representing Europeans: A Pragmatic Approach, 2nd ed. Oxford: Oxford University Press. Rose, R. 2019. Referendum Challenges to the EU’s Policy Legitimacy—And How the EU Responds. Journal of European Public Policy 26 (2): 207–225. Rose, R., and G. Borz. 2016. Static and Dynamic Views of European Integration. Journal of Common Market Studies 54 (2): 370–387.

58

R. ROSE

Scharpf, Fritz W. 2009. Legitimacy in the Multilevel European Polity. European Political Science Review 1 (2): 173–204. Thomson, R. 2012. Resolving Controversy in the European Union. Cambridge: Cambridge University Press.

CHAPTER 4

A Paradigm Shift in National Referendums on Europe Fernando Mendez

This chapter reviews some of the broad patterns of referendum activity on EU matters, from the first referendum in 1972 to the most recent ones in 2016. This activity has been a distinctly national affair; European institutions lack the authority to hold Europe-wide referendums, although proposals for such a procedure have been floated periodically (Habermas 2001; Cheneval 2007; Rose 2015). This means that the EU referendum experience is shaped by the interplay of two forces: the supranational politics structuring the broader EU integration process and the domestic politics of the referendum state. Explaining how these two dynamics have at times reinforced each other and at others collided is the object of this chapter.

4.1

Types of EU-Related Referendums

EU-related referendums come in a variety of types that can have a significant effect on referendum outcomes. This has led scholars to frequently distinguish between types of referendum (Acer 2005; Beach 2018; Hobolt 2009; Lacey 2017). The fourfold typology presented here will be used to explore the shifting dynamics of EU referendum activity.

© The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_4

59

60

F. MENDEZ

(1) Membership referendums: this type can take two distinct forms, an accession referendum or a withdrawal referendum. The former is undertaken by candidate states after a typically lengthy negotiation process. The candidate state votes on the negotiated accession treaty. In very rare instances, a subnational entity may also hold a separate accession referendum (as did the Finnish Åland Islands). This definition precludes the inclusion of referendums on the domestic ‘policy’ of initiating the formal process of joining the EU by a non-candidate state. The second category of referendum in the membership type relates to withdrawal from the EU. (2) EU Treaty revision referendums: these are referendums held by member states as part of the process of ratifying an additional EU treaty. Thus far there have been six rounds of major treaty revision, each generating at least one referendum in the course of the ratification process. The existence of these referendums, and the complications to which they give rise, stems from the fact that for a new treaty to enter into force the EU’s rules require unanimous approval by processes that each country is free to choose. (3) Policy referendums: these referendums are undertaken by an EU member state on a specific policy field, such as monetary policy, fiscal policy or the EU’s foreign policy. This type of referendum is a somewhat mixed bag, best defined in terms of what it is not: i.e. a member state referendum related to EU matters that cannot be accommodated by the precise definitions of types (1) and (2) above. Policy referendums take two forms: they either relate to (i) the direction of an EU policy (such as EU enlargement or an EU trade agreement) or to (ii) a member state’s domestic policy (such as whether the member state joins the Eurozone or opts into the EU’s justice and home affairs provisions). This distinction is important since the consequences of the former concern the whole of the EU while the latter are mostly restricted to the domestic arena of the member state. (4) Third-country referendums: these are referendums held on the topic of EU integration by a third country that is not an EU member state or has not acquired ‘candidate status’. Third-country referendums typically come in two forms: they are about ‘joining’ the EU or about a state’s bilateral relationship with the EU. A number of states have held referendums on the ‘policy’ of joining the EU. The most recent example is San Marino’s 2013 referendum

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

61

on accession negotiations with the EU—which failed the quorum. The second type is not about the policy of ‘joining’ the EU but about a country’s relationship with the EU. Third-country referendums have been only held thus far by countries neighbouring the EU. As for the distribution of referendum types, the first thing to note is that third-country referendums are actually quite popular, accounting for one-fifth of EU-related referendums. Twelve such referendums have been held. Yet, while third-country referendums have been held by Liechtenstein, San Marino and Switzerland, in practice it is Switzerland that accounts for most of this activity as a result of its repeated referendums on EU matters. Indeed, no current EU member state has held more EUrelated referendums than Switzerland. For the remainder of the analysis, I shall exclude third-country referendums since they are mostly accounted for by a single country. The reason why the EU has generated this particular configuration of referendums lies in its multi-layered territorial structure. The fact that the polity that has evolved into the EU has triggered three very specific types of referendums is not surprising. Indeed, the three functional types of referendums would be familiar to any scholar of comparative federalism. They relate to: (1) how a polity decides on the question of territorial boundaries; (2) how a polity alters its constitution; and (3) how a polity decides upon important matters of policy. Mendez et al. (2014) show that all the classic federations (Australia, Canada, Germany, Switzerland and the United States) have been confronted at one time or other with challenges regarding how to accommodate (or prohibit) the use of referendums within a federalised polity. While no federal model is the same, each offers insight regarding the inevitable dilemmas of squaring two often conflicting principles: the democratic principle of ‘one person, one vote’ with the federal principle of ‘one state, one vote’. The clash of these principles is at the heart of many of the EU’s problems in accommodating direct democracy (see Chapter 1). Figure 4.1 shows the proportion of EU-related referendums by type. Of the forty-eight referendums covered, almost half have been on membership while exactly one-third have been on treaty ratifications. The smallest proportion to date has been on policy referendums (only ten referendums).

62

F. MENDEZ

Fig. 4.1 Distribution of referendums per type (percent) (Source Author’s database)

An obvious concern when dealing with EU referendums is their success rate. By success rate, I mean the approval of a referendum in which voters endorse measures that entail an increase in European integration. In a democratic referendum, failure to achieve this is always possible. In terms of consequences, the negative outcome of a referendum can on occasion be negligible, such as when the Danes decided in 2015 not to replace their full opt-out regarding justice and home affairs issues with a case-by-case opt-out. At the other extreme, the immediate political impact of a failed referendum at the member-state level is the toppling of a sitting prime minister, as when David Cameron resigned the day after the result of the Brexit vote. The consequences of a failed referendum can be systemic and affect the whole EU polity. This was the case when France, followed by the Netherlands, rejected the European Constitutional Treaty in 2005 and in doing so brought the EU’s flagship constitutional project to a halt. As noted by Rose (2019), the EU and the member states have developed strategies to deal with referendum rejection of a pro-EU measure. These include what he calls legal coercion and a variety of concessions, including substantive ones and symbolic ones, as well as risk avoidance through adopting policies via differentiated forms of integration, which do not require the participation of all member states. Figure 4.2 shows the distribution of referendum failures by type and over time. Two central points are immediately apparent in Fig. 4.2: membership referendums rarely fail while most policy referendums do fail. When looking at the proportion of failures by decade there is a notable

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

63

Fig. 4.2 Failure rate of referendums (by Decade and Type) (Source Author’s database)

spike as we move into the present decade. Prior to the 1990s, referendum failure was an isolated event (e.g. Norway’s failed membership and Greenland’s withdrawal). Although Fig. 4.2 shows a sharp increase in the proportion of failures in the new millennium, the absolute number of referendum failures (five) is the same for each decade. In Fig. 4.3, the distribution of referendums by type is plotted over time. Distinctive patterns are clearly visible for all three referendum types. Why such patterns exist is the focus of the next two sections, beginning

Fig. 4.3 Types of referendums over time (Source Author’s database)

64

F. MENDEZ

with the dynamics surrounding the membership and treaty revision referendums. There has been at least one membership referendum every decade (Fig. 4.3). There were none before the 1970s because the EU’s foundational moment in the 1950s was not legitimated via a referendum ratification process at the constituent level. Unlike some of the federations mentioned above, none of the original six founding members needed, constitutionally speaking, to put membership in the European Economic Community (EEC) to a national referendum. By the 1970s, a series of European Court decisions such as the direct effect and supremacy of EU law (Weiler 1991) left the prospective candidate countries in no doubt that membership of this new political organisation implied limitations on national sovereignty. Thus, the referendum emerged at the first feasible juncture, which was in the early 1970s. The issue that triggered the first referendum in France was the organisation’s imminent enlargement. The first enlargement round led to referendums in all candidate states (Denmark, Ireland and Norway) with one notable exception, the UK. It was evident that Denmark and Ireland would have to hold referendums because of transfer-of-power clauses in their respective constitutions. In the case of Norway, although a referendum was not constitutionally required, the issue of accession was politically polarising and the ruling parties were internally split. Under such conditions, the referendum option was effectively inevitable. To this day, Norway is the only country to have rejected an accession treaty to join the EU, and it has done so twice. Despite calls for a referendum (Norton 2011), the UK avoided the referendum route to accession and joined what was then the EEC via a legislative act. Yet, a change in government led the Labour administration of Harold Wilson to call a popular vote on maintaining the country’s new membership barely two years after joining. Starting with the first Atlantic round of accessions in the 1970s, most EU enlargement rounds have triggered national referendums. The one notable exception is the Mediterranean enlargement of the 1980s in which Greece, Portugal and Spain joined without consulting their electorates. For these countries transitioning from authoritarian military rule, membership signalled an important and symbolic re-integration into the Western democratic fold. Furthermore, accession referendums were not constitutionally required. This is not the case for the enlargement round in the 1990s in which all candidate countries held referendums.

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

65

A membership referendum was deemed either constitutionally mandatory (Austria and Sweden) or politically mandatory (Finland and Norway). With the collapse of the Berlin Wall in 1989 came the largest of the EU accession waves, the eastern enlargement (Szczerbiak and Taggart 2004). A domino wave of membership referendums was held as Central and East European states used previously enacted constitutional provisions (or created them) to provide for popular legitimation of accession. Apart from Bulgaria, all eleven post-communist states that have joined the EU held referendums on accession. With very few exceptions, the accession referendum has become the constitutional norm for endorsing EU membership. However, the membership referendum is likely to be little used in future as the EU has consolidated its territorial boundaries and there are very few accessions on the horizon. The second major impetus of referendum activity, successive rounds of treaty revisions, is at least partly a consequence of enlargement. Following the 1970s enlargement, the EU incorporated two countries, Ireland and Denmark, that had transfer-of-powers clauses in their constitutions, thus ensuring that referendums to deliver domestic ratification of any proposed revision to the treaties would be likely. This is why Ireland and Denmark account for the highest number of EU-related referendums, nine and eight, respectively. Figure 4.3 shows how treaty revision rounds have clustered around critical junctures, in particular the 1990s and the 2000s. The reason reflects the constitutional constraints that condition how the EU polity can realise its political aspirations. Similar to Switzerland and the United States, the EU is a system of delegated competencies in which powers not expressly delegated to the upper level remain the prerogative of the constituent units. Until the 1980s the political aspirations of the then EC could be firmly accommodated within the framework of the EEC Treaty. That framework was used in imaginative ways to adapt to new challenges and empower the Union to exercise functions in areas with a rather tenuous link with the original objectives of the treaty (Weatherill 2011). By the 1980s an enlarging Union, composed of more member states and confronting new policy challenges, wanted to amend its foundational constitutional document in order to advance European integration. A new constitutional settlement would be required to provide the Union with the competencies to act in new policy areas (e.g. foreign policy or monetary policy) and to do so more effectively (e.g. with qualified majority

66

F. MENDEZ

voting). And so began a period of six successive large-scale treaty revision rounds that lasted from the mid-1980s to the late 2000s. Throughout this constitutional transformation, the treaty revision referendum has played a crucial constraining role (Dehousse 2006; Cheneval 2007; Hooghe and Marks 2009). The potency of this particular referendum is the result of a stringent treaty revision procedure that requires unanimity among heads of state to sign a new treaty as well as unanimity during a ratification process in which each member state has to deliver ratification according to its own constitutional provisions (be they a popular vote, a simple legislative majority or a qualified majority). This ‘double unanimity lock’ is remarkably stringent by any comparative standard (Mendez and Mendez 2017). Yet it has not prevented significant constitutional transformation. Partly, this ‘success’ is the consequence of European leaders’ ability over the years generally to avoid the referendum route, as in the case of the Lisbon Treaty. As a result, there has actually been a rather limited use of the referendum by individual countries for ratification purposes. More than two-thirds of such referendums have been held by just two member states, Denmark and Ireland. Just as the membership referendum arose at the first possible juncture, so too did the treaty revision referendum. The 1987 Single European Act (SEA) introduced qualified majority voting to many areas of EU decisionmaking. Both Ireland and Denmark held referendums on the SEA, in the former case after court ruling instructing the government to do so. While the 1987 SEA enjoyed a smooth ratification process, the same cannot be said for successive rounds of revision. The landmark Maastricht Treaty signed in 1992 provided an early glimpse of the ratification problems that would henceforth afflict the treaty revision route. The Maastricht Treaty delivered political union and the route to monetary union, but was rejected in Denmark and barely endorsed by a major founding member state, France. For some, that symbolic moment marked the end of what scholars had referred to as the ‘permissive consensus’, a period in which Europe’s governing parties constructed the EU’s institutional edifice largely unencumbered by domestic public opinion. Further treaty ratification problems were encountered a decade later with the Nice Treaty. This time the referendum bottleneck was caused by Ireland. Solutions were ultimately found in both the Danish (Maastricht) and the Irish (Nice) cases to allow for a second referendum on essentially the same treaty. While Denmark secured a series of opt-outs, Ireland voted again

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

67

on the basis of a declaration noting that its military neutrality would not be threatened. These earlier referendum skirmishes were a harbinger of a much bigger crisis that afflicted the Constitutional Treaty when treaty ratification referendums in 2005 derailed this landmark project in France and the Netherlands. The European Commission, with leadership from Germany, then brokered a new treaty that removed clauses that might trigger a referendum leading to its failure. This was signed by heads of state in Lisbon, and only one country, Ireland, was constitutionally required to hold a referendum. Irish voters rejected the treaty in 2008. The Irish rejection triggered a renewed constitutional crisis for the EU. At this point, EU leaders did what they had done previously in such cases. They added some legal clarifications as well as an important change to the treaty text—the maintenance of the principle of ‘one state, one commissioner’—before asking the Irish to vote again. Under intense political pressure, the Irish voted a second time and duly delivered ratification of the Lisbon Treaty in 2009. Referendum ratification problems since the 2000s were a symptom of a broader sea change in the politicisation of the EU. Referred to by Hooghe and Marks (2009) as the ‘constraining dissensus’, this new phase in the European integration process offers a marked contrast with the ‘permissive consensus’ golden age in which EU elites were relatively free from public opinion constraints. Referendum failures have signalled a growing disconnect between leaders and their domestic electorates. Referendum ratification problems neatly highlight the clash between the demand for supranational policies to address new challenges and domestic politics constraining member states’ ability to supply popular legitimacy for the project. The institutional machinery provided by Lisbon has equipped the EU to deal with most policy challenges without recourse to the risky treaty revision route. This explains the break in the trend line of Fig. 4.3 for treaty revision referendums. For instance, in the midst of the Eurozone crisis, EU leaders implemented, without a referendum, a narrow treaty revision (Article 136 amendment) to authorise a permanent stability mechanism (see Fabbrini 2016). More generally, under the Lisbon framework the EU has adopted an array of measures to transform the constitutional architecture of European economic governance without recourse to treaty revision. Treaty revision of the sort that occurred between 1993 and 2009 has not been a concrete issue on the policy agenda for a decade.

68

F. MENDEZ

In the current climate of Euroscepticism, treaty avoidance is the guiding principle.

4.2

Rise of the Policy Referendum

The first-ever popular vote held by a member state occurred when France staged a policy referendum in 1972 on the topic of the EEC’s Atlantic enlargement. The policy referendum then disappeared for 30 years before making its reappearance in the year 2000 when Denmark held a referendum on whether it should adopt the Euro. These two referendums neatly illustrate a key difference within the mixed-bag category of policy referendums. A policy referendum can be on implementing an EU-level policy, like the French 1972 referendum on enlargement. It has an extraterritorial impact beyond its national borders because it can prevent the EU from moving forward in a specific policy domain. On the other hand, the decision to adopt—or not—the Euro currency, the topic of the Danish policy referendum in 2000, primarily affected the domestic constituency holding the vote. It does not pose a veto threat on an EU policy. As shown in Fig. 4.3, there is a hiatus of roughly three decades between these two different types of policy referendums. The pertinent question is why, after such a long time gap, the policy referendum re-emerges to become the most frequent type of referendum in the current decade. Even though there have subsequently been six enlargement rounds in which the extraterritorial policy referendum could have been deployed, the pioneering French move has not been repeated. Because the dangers of a unilateral veto on enlargement were realised early on, European political elites have by and large been successful in ‘deconstitutionalising’ enlargement. Shaw (2005) has emphasised that although enlargement is a constitutional question it has never been treated as such. The end result is that enlargement has avoided the difficult ratification processes that afflicted most major treaty revision rounds. During the three-decade period between the first French policy referendum and the Danish ballot that marked the new millennium, the policy ambitions of the EU were bundled into constitutional packages that took the form of new treaties. But there was a price to pay for making it easier for member states to deliver treaty ratification. The special opt-out was a strategy pioneered during the Maastricht Treaty’s ratification process to pave the way for a second referendum in Denmark. The Danes secured an opt-out from adopting the Euro (as well as justice and home affairs,

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

69

the subject of its most recent referendum). It was thus inconceivable that when the government decided it wanted Denmark to join the Euro it could do so without consulting the electorate. Denmark’s Euro referendum duly failed, as did that held by neighbouring Sweden on the same topic shortly afterwards. The opt-out, or more generally ‘differentiated integration’ (Leuffen et al. 2012), is the main driver explaining half of the policy referendums since 2000. In an increasingly differentiated EU, the desire by member states for more policy integration in areas that are politically salient (such as adopting the Euro or further integration in areas such as justice and policing) can easily fuel demands that lead to policy referendums. This particular type of policy referendum, however, tends not to have an impact beyond the member state holding it. The logic of opting in to a policy domain can also be reversed. A member state can hold a referendum whose consequence, if implemented, is to lead to an opt-out in a given policy field. The most notorious example of this was the Greek bailout policy referendum of 2015. That particular referendum was framed in manifold ways, but for those voting ‘Yes’ (to an EU bailout package) the issue was one of remaining in the Eurozone (Triga and Manavopoulos 2017). In the event, Prime Minister Alexis Tsipras ignored the referendum result he had campaigned for and proceeded to agree to the bailout conditions imposed on Greece. Another referendum that is functionally equivalent to a (unilateral) opting out of a policy domain is the migrant quota policy referendum staged by the Hungarian prime minister, Viktor Orbán, in 2016. Another type of policy referendum that has made a comeback is the extraterritorial policy referendum on an EU-wide topic. Successfully avoided for four decades, this French-pioneered referendum re-emerged in the 2010s after the EU signed an Association Agreement in 2014 with Ukraine (as well as Georgia and Moldova). As with a revision to the EU’s treaty, the Association Agreements with these third countries required domestic ratification by all EU member states. This occurred against a backdrop of growing Russian assertiveness, leading to its annexation of Crimea, a region of Ukraine. While negotiations for the Association Agreement were taking place with the three successor states of the Soviet Union, the Dutch passed the Advisory Referendum Act. The new law provided for the activation of a referendum in certain areas (including international treaties) if a threshold

70

F. MENDEZ

of 300,000 signatures was reached. Once the law providing for citizeninitiated referendums came into effect, Dutch Eurosceptics challenged the first series of EU-related laws they could (van den Akker 2017). Their strategy met with success in the case of the Ukraine Association Agreement, where the required number of signatures was easily achieved. The EU’s Association Agreement with Ukraine was rejected by the Dutch electorate on the lowest recorded turnout of any EU-related referendum, 32%. Nonetheless, it narrowly crossed the 30% turnout quorum for ensuring the referendum’s validity. The Dutch government accepted a reinterpretation of the Association Agreement as the basis for ratifying the agreement, the referendum vote notwithstanding. Having seen the uses to which its referendum law was put, Dutch political elites have since repealed the law allowing citizen initiatives. There is one way in which the extraterritorial policy referendum on an EU-wide matter can be neutralised. This bypassing strategy was used in relation to the 2012 Fiscal Compact Treaty when member states agreed an intergovernmental treaty under international law rather than through the EU. The 2012 Fiscal Compact Treaty involved all the member states except the UK, which would have vetoed it had it been dealt with through the EU process. Since its primary purpose was to achieve fiscal discipline by Eurozone states, the Fiscal Compact Treaty was intimately connected to EU law and even made use of EU institutions and staff. However, by structuring it as an intergovernmental treaty, EU leaders were able to achieve an important step forwards in European integration without having it subject to a British veto. Ireland was required by its national constitution to hold a referendum. If the Irish had rejected signing the treaty, it would still have come into force because it had achieved sufficient signatories.

4.3

Referendums on Disintegration

A distinctly new feature of the referendum landscape in the post-2010 phase of European integration is the appearance of the ‘disintegrative referendum’. The concept of a disintegrative referendum draws on the framework of Mendez and Germann (2018) for analysing sovereignty referendums. It is an especially dangerous development for the continued existence of the EU.

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

71

A referendum typically implies a change in the status quo towards a new state of affairs. The proposed change can take two distinct logics— integrative or disintegrative. Drawing on Haas (1958: 16), an integrative logic refers to the dynamic ‘whereby political actors in one or more political (sub-)systems are persuaded to shift their loyalties, expectations and political activities toward a new center’ that then acquires new powers. Both treaty ratification referendums and the policy referendum are ultimately about the balance of competencies within the multi-layered polity we call the European Union. On the other hand, membership referendums address the boundary question of who is in or out of the political organisation. A disintegrative logic, as described by Wood (1981: 111), operates in the opposite direction ‘whereby political actors in one or more subsystems withdraw their loyalties, expectations and political activities from a jurisdictional centre to focus them on a centre of their own’. The integrative logic has dominated the history of European integration and the referendums it has engendered. A similar integrative logic has predominated in federal systems that have a referendum tradition (e.g. Australia and Switzerland). Since all powers not delegated to the EU level rest with the constituent units, it is thus hardly surprising that almost half of the EU-related referendums are about the allocation of functional competencies to the EU level and take the form of an integrative referendum. At the same time, it is hardly surprising that, given the EU’s aspiration to embrace a continent, the remaining half of EU-related referendums are membership referendums that follow an integrative logic. The United States has held many more integrative referendums than the EU in its 200-year-long continental expansion (Mendez and Germann 2018). A referendum on membership can logically be about leaving rather than joining a political body. Such a referendum would follow a disintegrative logic. One of the most common disintegrative referendums is a secession or independence referendum. Historically, disintegrative referendums have occurred in federations: Canada has experienced two, Australia one, and a number of disintegrative referendums were held in the United States during the buildup to the American Civil War. Unlike most federations, the EU system actually provides for withdrawal, the Article 50 procedure. An EU withdrawal referendum thus follows the logic of a disintegrative boundary referendum. There have been three such referendums during the EU’s short history.

72

F. MENDEZ

The UK’s first ‘withdrawal’ referendum was held in 1975, less than two years after it formally joined the EEC. The question put to the people was: ‘Do you think that the United Kingdom should stay in the European Community (the Common Market)?’ However, this referendum is much closer to an accession referendum endorsing membership of the then European Community than it is to a withdrawal referendum. The ballot was the result of a divided Labour party—which was in opposition when the Conservative prime minister took Britain into the EU— making a 1974 manifesto promise to hold a referendum on continued membership. A different type of disintegrative referendum vote was about whether a subnational territory could leave the European Union. Greenland is an overseas territory of Denmark that enjoys a special autonomous status. Its vote followed a territorial re-configuration within Denmark and dissatisfaction in Greenland with EU policy encouraging multi-national access to fishing rights in what had historically been exclusive national territory. In 1982, 53% of voters approved Greenland leaving the EU. An unusual compound occurs when a nationalist party within a region of a member state promotes a disintegrative referendum to advance towards its goal of independence, while simultaneously wanting to keep the status of belonging to the EU. The possibility of this happening arises from developments in Scotland and Catalonia. Before the 2014 Scottish referendum on independence, the UK government said it would accept a majority vote for withdrawal but also made clear that an independent Scotland could not count on the UK, then a member state, giving its application for membership automatic support. The Scottish National Party in charge of the devolved government of Scotland is planning a second independence referendum on the grounds that Brexit creates a fundamentally new situation. The goals of Catalan nationalists are independence from the Kingdom of Spain along with EU membership. However, the Spanish government and courts have refused to recognise the legitimacy of a Catalan independence referendum on the grounds that the Spanish Constitution stipulates that Spain is an indivisible state. Moreover, as it is the Madrid-based national government that sits in the European Council, it is strategically placed to stop any attempt by a Catalonia government to mobilise support from EU institutions. Turning to specific policy fields, rather than membership of the EU, there are two examples of a disintegrative policy referendum in which the

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

73

majority vote is to nullify the national implementation of an EU policy. Both the 2015 Greek bailout and the 2016 Hungarian migrant quotas entailed a change in the status quo which, if implemented, implied unilateral repatriation of a power (i.e. Greece leaving the Eurozone) or opting out of an EU decision (i.e. the migrant quota allocation for Hungary). In both cases, the implied change in the status quo was overwhelmingly approved. The Greek bailout referendum in 2015 was the climax of five years of friction as a consequence of the Greek government’s spiralling deficit causing it to turn to a Troika of the European Commission, the European Central Bank and the International Monetary Fund for financial assistance. It received e289 billion in loans on condition that it reform its finances and cut spending. The resulting austerity caused unemployment to skyrocket and the economy to contract greatly. In June 2015, the Greek government broke off negotiations with the Troika about the conditions of further financial support and called a referendum to be held in ten days’ time. The government asked for popular support for rejecting the conditions, with the potential consequence of Greece leaving the Eurozone. Its policy was endorsed by 61% of the voters. Within a week of the referendum vote, Prime Minister Tsipras accepted the bailout package offered at the July 2015 Euro summit by heads of state of the Eurozone and the Troika. The 2016 Hungarian policy referendum was called in response to the European Council’s qualified majority approval of a mandatory refugee relocation quota. Hungary was one of four member states opposing the measure. Prime Minister Orbán responded by calling a referendum on whether the EU should have the power to impose a migrant quota on Hungary. The result was an almost unanimous ‘Yes’ vote, since the main opposition parties recommended boycotting the referendum, and it failed to achieve the requisite quorum of participants (Pállinger 2017). The Hungarian migrant quota referendum is arguably the most dangerous referendum outcome because it challenges the very legitimacy of authoritative EU decision-making in the Council. To use a unilaterally staged referendum to overturn a binding EU decision-making procedure would, if replicated by other states, lead to EU policy gridlock. The EU’s referendum landscape between the 1970s and the 2010s has been shaped by the interplay of two structural features: the supranational politics of a polity that has evolved from a common market to an overtly political organisation, and constraints arising from the domestic politics of

74

F. MENDEZ

the member states that make up the EU. How these two forces interact within the EU institutional matrix explains the patterns of EU-related referendums we have witnessed in the past. Two specific developments have generated the most EU referendum activity: first, the expansion in the number of candidate states seeking to join the EU and, second, the functional pressure to equip the EU with the necessary powers to deal with the ever growing demands placed on it. However, the paradigm shift in the use of referendums in the past halfcentury is a reminder that their use is subject to change when the political environment of Europe changes. Insofar as membership and treaty revision issues are less frequent in the future, the policy referendum will gain in importance. Three important changes are already altering the contours of the future referendum landscape. First, the EU is consolidating its territorial boundaries, thereby reducing the likelihood of many more accession referendums. Second, domestic political constraints among and within member states have taken off the EU’s policy agenda large-scale treaty change that can be vetoed by a national referendum in one of more than two dozen member states. The Lisbon Treaty’s entry into force just before the current EU backlash provided the EU with most of the powers it needs to deal with current policy pressures while largely avoiding the need for formal treaty revision. Third, the trend towards more differentiated forms of European integration provides a means to strike a balance between member states where there is political support for further measures of European integration and member states with the right to call referendums to reject such steps (Rose 2015: 153ff.). Any member state where there is domestic controversy about participating in a policy does not have to wield a veto to stop all states adopting it, nor does it have to be forced to go along with a policy that only a minority oppose. It has the right to call a national referendum that gives its citizens the power to decide whether to opt in or opt-out of a major policy change without the outcome vetoing a policy favoured by the majority.

References Acer, A. 2005. The Constitutional Scheme of Federalism. Journal of European Public Policy 12 (3): 419–431. Beach, D. 2018. Referendums in the European Union. In Oxford Research Encyclopedias of Politics, ed. W.R. Thompson. Oxford: Oxford University Press.

4

A PARADIGM SHIFT IN NATIONAL REFERENDUMS …

75

Cheneval, F. 2007. “Caminante, no hay camino, se hace camino al andar”: EU Citizenship, Direct Democracy and Treaty Ratification. European Law Journal 13 (5): 647–663. Dehousse, R. 2006. The Unmaking of a Constitution: Lessons from the European Referenda. Constellations 13 (2): 151–164. Fabbrini, F. 2016. Economic Governance in Europe: Comparative Paradoxes, Constitutional Challenges. Oxford: Oxford University Press. Haas, E.B. 1958. The Uniting of Europe: Political, Social, and Economic Forces, 1950–1957. Stanford: Stanford University Press. Habermas, J. 2001. A Constitution for Europe? New Left Review 11 (5): 1. Hobolt, S.B. 2009. Europe in Question: Referendums on European Integration. Oxford: Oxford University Press. Hooghe, L., and G. Marks. 2009. A Postfunctionalist Theory of European Integration: From Permissive Consensus to Constraining Dissensus. British Journal of Political Science 39 (1): 1–23. Lacey, J. 2017. National Autonomy and Democratic Standardization: Should Popular Votes on European Integration Be Regulated by the European Union? European Law Journal 23 (6): 523–535. Leuffen, D., B. Rittberger, and F. Schimmelfennig. 2012. Differentiated Integration: Explaining Variation in the European Union. Basingstoke: Palgrave Macmillan. Mendez, F., and M. Mendez. 2017. The Promise and Perils of Direct Democracy for the European Union. Cambridge Yearbook of European Legal Studies 19: 48–85. Mendez, F., and M. Germann. 2018. Contested Sovereignty: Mapping Referendums on Sovereignty over Time and Space. British Journal of Political Science 48 (1): 141–165. Mendez, F., M. Mendez, and V. Triga. 2014. Referendums and the European Union: A Comparative Inquiry. Cambridge: Cambridge University Press. Norton, P. 2011. Divided Loyalties: The European Communities Act 1972. Parliamentary History 30 (1): 53–64. Pállinger, Z. 2017. Hungary. In Referendums on EU Matters, ed. F. Mendez and M. Mendez. European Parliament. www.europarl.europa.eu/supportinganalyses. Rose, R. 2015. Representing Europeans: A Pragmatic Approach, 2nd ed. Oxford: Oxford University Press. Rose, R. 2019. Referendum Challenges to the EU’s Policy Legitimacy—And How the EU Responds. Journal of European Public Policy 26 (2): 207–225. Shaw, J. 2005. Europe’s Constitutional Future. Public Law 1: 132–151. Szczerbiak, A., and P. Taggart. 2004. The Politics of European Referendum Outcomes and Turnout: Two Models. West European Politics 27 (4): 557– 583.

76

F. MENDEZ

Triga, V., and V. Manavopoulos. 2017. Greece. In Referendums on EU Matters, ed. F. Mendez and M. Mendez. European Parliament. www.europarl.europa. eu/supporting-analyses. van den Akker, J. 2017. The Netherlands. In Referendums on EU Matters, ed. F. Mendez and M. Mendez. European Parliament. www.europarl.europa.eu/ supporting-analyses. Weatherill, S. 2011. The Limits of Legislative Harmonization Ten Years After Tobacco Advertising: How the Court’s Case Law Has Become a “Drafting Guide”. German Law Journal 12 (3): 827–864. Weiler, J. 1991. The Community System: The Dual Character of Supranationalism. Yearbook of European Law 1 (1): 267–306. Wood, J.R. 1981. Secession: A Comparative Analytical Framework. Canadian Journal of Political Science/Revue canadienne de science politique 14 (1): 107– 134.

CHAPTER 5

When Institutions and Issues Change, Voting Changes Ian McAllister and Richard Rose

Voting behaviour is about how individuals cast ballots, but elections differ in the political issues that are at stake: control of government or making a decision about a major issue of policy. Electoral institutions also differ in how they translate votes into victory for one side or the other (Rose 1974: 9). The institutional differences between a referendum and a parliamentary election are categorical. In a referendum, an individual is asked to make a choice about an issue, and the outcome is a politically binding decision in favour of the policy that wins an absolute majority. By contrast, in a British parliamentary election, an individual votes for a party candidate; the outcome is then translated into MPs’ seats, and the outcome is that the party with the most seats gains control of government. The close link in time between the 2016 Brexit referendum and the 2017 general election offers an unusual opportunity to test whether changing the institutional choice between voting for a party or an issue changes the influences on how people vote. The May 2019 election of British members of the European Parliament added yet another combination, for voters were asked to choose between parties in a proportional representation ballot. However, the issue of Brexit had more influence than party labels. The December 2019 general election was a hybrid; Prime Minister Boris Johnson sought to turn it into a referendum on

© The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_5

77

78

I. MCALLISTER AND R. ROSE

Brexit while the Labour party sought to make it a vote on class and economic issues. In this chapter, we analyse British Election Study surveys to see whether the same influences affect voters whatever their format or whether changing electoral institutions and changing issues affect how people vote.

5.1

Same Voters, Different Institutions and Issues

The Brexit referendum on 23 June 2016 was the culmination of decades of Eurosceptic criticism of the European Union and a tepid defence of EU membership by successive Conservative and Labour governments (Young 1998; Waldegrave 2019). The Conservative party was very publicly split, with senior Cabinet ministers such as Boris Johnson taking the opposite side from Prime Ministers David Cameron and Theresa May. The Labour party took refuge in evasion in efforts to hide its internal divisions. An additional source of the blurring of party lines was the vocal activity of Nigel Farage, a leading campaigner for Brexit and nominal leader of the UK Independence Party (UKIP). The media publicised divisions on the issue of Brexit rather than party leaders and manifestos, since the referendum was about an issue rather than which party should control British government. The referendum ballot offered voters an unqualified choice: ‘Should the United Kingdom remain a member of the European Union or leave the European Union?’ The turnout, higher than at any parliamentary election in almost two decades, showed the importance of the issue. The referendum’s institutional character guaranteed an absolute majority for the winning side, something no British party has achieved in a parliamentary election since 1935. The Leave alternative won with a majority of 1,269,501 votes over the alternative. However, given that more than 33 million votes were cast, the margin of victory was relatively narrow: 51.9% for leaving the EU and 48.1% for remaining (Uberoi 2016). In the expectation of the electoral system converting a plurality of Conservative votes into a much stronger majority for her Conservative government, Prime Minister Theresa May unexpectedly called a parliamentary election for 8 June 2017. This was only weeks after the Conservative government had formally notified Brussels that the UK would withdraw from the European Union in two years. However, the election was not about an issue but about who governs Britain. Brexit was not a

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

79

major issue as both parties had voted in favour of notifying the EU of the UK’s withdrawal, and both manifestos pledged to deliver Brexit in one form or another. In the 2017 parliamentary election, Jeremy Corbyn proved surprisingly effective at mobilising electoral support for the Labour party’s most left-wing economic programme in a third of a century. By contrast, Theresa May made a botched appeal for support on economic grounds and showed herself an ineffective campaigner (Seldon 2019). The Conservatives won 42% of the vote and Labour 40%. Consistent with this being an election to decide who governs, the vote for the two parties contesting control of government was the highest since 1970. However, the electoral system did not convert the Conservatives’ lead in votes into a majority of parliamentary seats. Theresa May remained prime minister responsible for delivering Brexit without a parliamentary majority. The election of British members of the European Parliament (EP) on 23 May 2019 was institutionally different. The ballot was not meant to decide who governs the European Union (see Chapter 3) but to give Britons a chance to express their preference on issues. The EP election occurred when there was maximum uncertainty about whether or how the Brexit referendum decision should be implemented, and the allocation of EP seats by proportional representation ensured that no votes were wasted, thereby giving equal opportunity to parties that had no chance of winning control of government under the first-past-the-post electoral system used in British parliamentary elections. Neither the Conservative nor the Labour party wanted to intensify its internal problems by actively campaigning for EP seats that would cease to exist once Brexit was accomplished. The most keenly pro- and anti-Brexit parties, the Liberal Democrats and Nigel Farage’s Brexit Party, which he instantly created for this election, campaigned by stressing their opposing positions on the referendum result. Thus, the issue context was similar to the Brexit referendum, but the ballot offered a choice between parties rather than between policies and the seats were allocated by proportional representation rather than as in a British general election. The election of British MEPs by proportional representation produced a unique result. The Brexit Party came first with 31% of the vote and the Liberal Democrats second with 20%. Labour came third with 14% of the vote and the Conservatives fifth, three points behind the Green party, which won 12% (see Table 5.1). The two parties competing for

80

I. MCALLISTER AND R. ROSE

Table 5.1 Contrasting party votes in general and EP elections 2017 General (%) Pro-Brexit Conservatives UKIP/Brexit parties Total Ambivalent Labour Pro-EU Liberal Democrats Green party Change UK Nationalists Total Others

2019 EP (%)

Change (%)

42.3 1.8 (44.1)

9.1 34.9 (44.0)

−33.2 +33.1 (−0.1)

(40.0)

(14.1)

(−25.9)

7.4 1.6 0 4.8 (13.8) 2.1

20.3 12.1 3.3 5.0 (40.7) 1.2

+12.9 +10.5 +3.3 +0.2 (+26.9) −0.9

Source Authors’ classifications from official sources

British government together collected only 24% of the vote in the European Parliament election. In the December 2019 parliamentary election, the institutional arrangements were unchanged from two years previously, but the issue was different. Prime Minister Boris Johnson treated the election as a rerun of the 2016 referendum, a ballot to ‘get Brexit done’. The message hammered home to the electorate was that only by voting for a Conservative government could the referendum decision be implemented. The argument sidelined the Brexit Party because it had no MPs and no chance of forming a government. The Labour party’s manifesto gave priority to left-wing social and economic issues. It promised to negotiate a soft Brexit withdrawal agreement with Brussels and then hold a referendum offering voters the choice of a soft Brexit or remaining in the EU. Jeremy Corbyn endorsed the hard Brexit view of the superior legitimacy of referendums by saying he would remain neutral in a second referendum so that the people could decide the issue. The Liberal Democrats’ policy was clear—to revoke EU withdrawal without the need for a referendum—but it was also incredible, because it assumed the party would win control of Downing Street. By making the Conservatives the chief defender of Brexit, Johnson won a big parliamentary majority, while Labour’s share of the vote was its lowest since 1935.

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

81

Each election in this quadrathlon of choices differed in institutions, issues or both. The referendum required an absolute majority to win. The two parliamentary elections required only a plurality of votes to qualify as the winner. This was radically different from the multi-national European Parliament election since British voters were not choosing a European government but representatives to express their views on European affairs. Even when institutions are the same, the issue can change. In the 2017 British general election, Theresa May tried to campaign on social issues and only won enough MPs to form a minority government. Two years later, the same system gave Boris Johnson a big majority after he ran a referendum-style campaign to ‘get Brexit done’. Although the elections differed, the electorate consisted of virtually the same people over the three-year period. Before British voters could cast a ballot in the 2016 Brexit referendum, the median elector in her or his forties had already voted in half a dozen elections and had developed political attitudes, social interests and party preferences that were standing influences on their voting behaviour. The influences reflect income, education and left- and right-wing ideologies associated with socio-economic characteristics. These familiar and long-standing commitments imply: Hypothesis 1 If individual characteristics influence how people vote, influences should remain the same regardless of electoral institutions and context. Cultural values about national identity, traditions and trust or distrust in foreigners start being formed well before a person is old enough to vote. These attitudes have had little influence on voting behaviour in parliamentary elections. However, the 2016 referendum on membership of the European Union made attachment to traditional or cosmopolitan values immediately salient. The referendum not only involved a change in electoral institutions but also a shift from class to cultural issues, thereby changing the influences that determine how people vote: Hypothesis 2 If electoral institutions and context influence how people vote, influences on individuals should vary between types of elections. To test these hypotheses, the following pages report multivariate statistical analyses of British Election Study surveys of 2016, 2017 and 2019 elections, and compare the results to see whether the same individual characteristics are consistently influential or whether influences change when institutions and issues change.

82

I. MCALLISTER AND R. ROSE

5.2

Theories of Individual Voting Behaviour

Voting in parliamentary elections is the focal point of theories of voting behaviour (Fisher et al. 2018). It is also the starting point for developing theories about voting in referendums (Hobolt 2009; Sciarini 2018; Marsh 2018). For example, six of the seven sets of influences tested by Clarke et al. (2017: Chap. 7) in their intensive study of the Brexit referendum come from the literature of voting in parliamentary elections. However, this focus tends to ignore the fact that Britons vote in three types of elections differing in institutions and issues. If individual characteristics and attitudes are of pervasive importance, then the influences listed below should consistently influence their voting behaviour regardless of differences in electoral institutions and issues at stake. However, if institutions and issues matter for how a person votes, then a voter will rely on different influences when the electoral context changes. Class Britain’s social structure and electoral system have made it conventional to divide the electorate into Labour-voting manual workers and middleclass Conservative voters. In Peter Pulzer’s (1967: 98) words: ‘Class in the basis of British politics. All else is embellishment and detail’. However, the extent of class voting has tended to be exaggerated, and major changes in social and economic structure have increasingly questioned the traditional division of voters into two classes (Rose 1968; Evans and Tilley 2012; Savage et al. 2013). Income and education as well as subjective class identification are indicators of class that influence voting, but their effect varies across time and across national boundaries (Lewis-Beck et al. 2013). Left- and Right-Wing Ideologies Political ideologies incorporate political ideas and values that can be used to evaluate political parties and issues. Left- and right-wing ideologies are commonly employed in voting studies to summarise a host of political, social and economic issues in a single variable. Ideological beliefs result in some people with higher education adopting left-wing views and some low-income people being right-wing in their values. Since ideology

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

83

provides a comprehensive guide to making political choices, its influence should be the same whatever the institutional form of an election. Europeanisation and globalisation have divided people who value national traditions from individuals with a cosmopolitan outlook (see Mau and Verwiebe 2010). This challenges theories of the persistence of class as the principal source of electoral cleavages (Lipset and Rokkan 1967). In contemporary circumstances, an epigram of a Bill Clinton election strategist, ‘It’s the economy, stupid’, can be paraphrased to emphasise, ‘It’s culture, stupid’ (Hutter et al. 2016). National Culture Traditional cultural values use the past to provide a standard for evaluating present-day politics. The past can be summed up in terms of a homogeneous national culture, whether described as English or British. It rejects the idea that a multi-cultural society created by immigration is desirable (Eatwell and Goodwin 2018; Kaufman 2018). Speaking to the Conservative Party Conference in 2015, Theresa May attacked cosmopolitans: ‘You believe you are a citizen of the world; you are a citizen of nowhere’. She identified the Conservative party as the party from somewhere. Traditions glorify such achievements as Britain under Winston Churchill’s leadership standing alone against the threat of invasion by a German army that had conquered Europe. Staunch advocates of Brexit proclaimed that multinational EU institutions threatened to make Britons their vassals. Cosmopolitan Culture Cosmopolitan individuals have a positive attitude towards people of different nationalities and cultures and are therefore comfortable being European as well as British citizens. Theories describe cosmopolitans as the winners from Europeanisation because they are more educated and younger. Individuals who have been to university are accustomed to mixing with people from different national backgrounds there, at work and in leisure. Whatever their education, young people are socialised into a multiracial Britain that was part of the European Union long before they were born (cf. de Vries 2018). In voting on the referendum issue, cosmopolitans should be more likely to favour remaining in the EU.

84

I. MCALLISTER AND R. ROSE

Partisanship and Anti-partisanship The theory of party identification describes it as providing a simpler shortcut than ideology to making a decision about how to vote (Butler and Stokes 1970). An individual who identifies with a political party does not need to be informed about issues; the only thing required is that the ballot carries the name of the party with which one identifies. This logic applies to referendum voting too, subject to the condition that parties have a clear position on what is at stake. However, if parties are internally divided on a major issue, as has been the case with Brexit, the influence of party identification becomes problematic. Because a referendum is about a single issue, it offers voters who disagree with their party’s position a chance to vote against their party without jeopardising its status in government and if a party is divided on the issue at stake, party guidance is ambiguous. Today there is a complementary phenomenon of anti-partisanship, a strong collective dissatisfaction with established parties. Individuals who think that politicians do not care about ordinary people and are only out for themselves can protest by voting against the party in government or for a protest party such as the Brexit Party and reject the commitment of their leaders to the European Union. In our model, the behaviour of individual voters (BIV) can be as a function of socio-economic class (SEC), ideology (I), national culture (NC), cosmopolitan culture (CC) and partisanship (P). BIV ( f ) SEC, I, NC, CC, P Quantitative studies invariably find that multiple potential influences divide into three groups: those that are significant and substantial; influences that are statistically significant but have a limited effect; and those that lack significance.

5.3

Institutions and Issues Matter

Whereas the first hypothesis assumes that the influence of individual characteristics should be constant across elections, the institutional hypothesis predicts the opposite. Categorical differences between referendums, an election to decide which party governs, and electing MEPs should change the chief influences on how individuals vote. For example, class may be of primary importance when voting for a party government, but cultural

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

85

values when voting on a referendum issue, and both types of influence may be equally important in voting for MEPs. To test the two hypotheses, we ran separate logit regressions of multiple influences on BES voting surveys at the 2016 referendum, the 2017 general election and the 2019 European Parliament vote, and compared the results with available survey data from the December 2019 election. Institutional differences create differences in the ballot choice of parties and issues. For the 2017 and 2019 parliamentary elections, the choice was whether to vote Conservative or Labour; for the referendum, to Remain in or Leave the European Union; and for the EP proportional representation election, which of several pro-Brexit or pro-Remain parties to vote for. The individual characteristics that influence voters, such as class, ideology, culture and education, are likely to remain the same over the three-year span covered by the elections. When electoral institutions and issues change, people do not change their class, ideology or cultural values, but what may change is the importance that an individual ascribes to these characteristics when deciding how to vote. Insofar as the regressions show that the same influences are similarly significant in their impact regardless of the type of election, this will support the first hypothesis. However, insofar as influences change when institutions and issues differ, this will support the second hypothesis. Because the 2019 BES sample was very large, 5693 respondents who voted in the election, the statistical significance level is set at p < .000, while for the BES sample of the two preceding elections, 1191 and 807 respondents who voted, respectively, it is p < .00. The impact of a significant variable is shown by the Relative Likelihood Estimate (RLE). It indicates the extent of influence on a scale of probabilities ranging from zero to one; an estimate above a random 50:50 division shows a positive influence on voting Conservative, for a Brexit party or for leaving the EU, while a number below .50 shows a negative influence. To ensure comparability in comparison, all independent variables are coded on 0 to 1 scales (for details, see Appendix, Table 5.4). The BES asks about many topics that could potentially influence voters, but analysis invariably finds that many potential influences have no significant effect. Hence, we initially tested two dozen variables for their potential influence on voting in one or another type of election. To avoid overloading the presentation of results, we excluded from the final analysis those that consistently failed to achieve statistical significance: gender, living in Scotland or the North of England, the cost of living, trade

86

I. MCALLISTER AND R. ROSE

union membership, tolerance towards others, moral values, white ethnicity, media attention and environmental attitudes. Class and Ideology Consistent with generations of studies of parliamentary elections, socioeconomic class had a significant and substantial effect at the 2017 parliamentary election (Table 5.2). Among those most affected by income inequality, there was a 76% likelihood of voting Conservative among individuals in the top sixth of incomes, and a 25% likelihood among voters with the lowest income. Similarly, in 2017 the likelihood of voting Conservative is reduced by 20 percentage points among people who saw themselves as working class. However, there is a stark contrast between voting on an issue and for a choice of government. Neither income nor class has a statistically significant effect on the referendum vote or on an EP election that was in effect a referendum between pro- and anti-Brexit parties. Ideology has a similarly strong influence on an individual’s choice of party government. After controlling for the effect of other variables, among those with extreme right-wing views the probability of voting for a Conservative government in 2017 was 77% and there was a 68% likelihood of voting for the Brexit or Conservative parties in the 2019 EP election. Among people with extreme left-wing views, the likelihood of voting Conservative was 18% and of voting for Brexit it was 34%. However, the impact on election outcomes of being on the left or the right is limited, since only one in seven respondents place themselves on the extreme right and only one in ten on the extreme left. Since the abstract nature of ideological beliefs has little meaning to many voters, attitudes towards economic policies provide an additional indicator of ideology. In the 2017 BES survey, 30% endorsed without qualification the view that private enterprise is the best way to solve Britain’s economic problems, while 29% favoured having major industries and public services in state ownership. After controlling for other influences, the probability of voting Conservative was 74% among those who favoured private enterprise as against 24% for those favouring state ownership. Even though much of the referendum debate stressed the impact of EU membership on the economy, neither economic preference had a significant effect on referendum voting, while in the EP election a preference for state ownership significantly increased the likelihood of voting for

5

Table 5.2

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

87

Class and culture as influences in three types of elections 2017 parliament Con vs Lab

2016 referendum Leave vs Remain

2019 Euro Parl Leave vs Remain

p value

Con prob %

p value

Leave prob %

p value

Leave prob %

76 25 0 30

.077 .072 .073 .076

40 60 40 42

.668 .006 .203 .975

49 42 46 0

77 18 74

.033 .009 .571

64 34 47

.000 .372 .017

68 47 57

24

.082

77

.000

34

56 62

.001 .027

66 60

.000 .000

67 68

78 42

.734 .803

48 51

.000 .649

67 52

82

.000

88

.000

78

69

–

–

.000

7

49 43

.450 .316

45 45

.000 .019

36 44

–

–

–

.000

37

– – 35

.649 .019 .011

53 38 60

– – .058

– – 45

Socio-economic High income .000 Low income .000 Middle-class ID .959 Working-class .001 ID Ideology Right-wing .003 Left-wing .001 Private .000 ownership best State .000 ownership best National culture English ID .399 Respect .059 tradition Old 65 plus .000 Minimum .308 education Too many .000 immigrants Voted leave .001 Cosmopolitan culture Young 18–25 .871 University .241 education London – Partisanship Conservative ID – Labour ID – Politicians don’t .011 care R2 .549 (N ) (807)

.426 (1191)

.661 (5693)

88

I. MCALLISTER AND R. ROSE

pro-Remain parties, which included Labour. By contrast, this variable had no influence on voting pro-Brexit in the EP election, since most in this group endorsed the cross-class Brexit Party rather than the Conservatives. To test whether anti-EU voting was motivated by commitment to a populist ideology (Mudde and Kaltwasser 2017; Müller 2016), in a preliminary analysis we tested the influence on voters of populist ideological principles such as intolerance of minorities. After controlling for all other variables, populist indicators had no significant effect on referendum choice or party vote and were therefore dropped from the final analysis (see also Eatwell and Goodwin 2018: Chap. 1). However, Boris Johnson’s description of the 2019 general election as about the People versus Parliament did frame the choice in populist terms. The analysis of class and ideology shows strong support for the second hypothesis: when institutions and issues change, their influence on voting behaviour changes. None of the eight measures had a significant effect in the referendum vote on the issue of Britain’s EU membership nor did class have any significant effect in the 2019 parliamentary election held as a referendum about getting Brexit done (see below). While two ideological indicators did have a significant influence on voting for Brexit parties at the European Parliament election, in both cases the size of the effect was less than in the parliamentary election two years previously. This conclusion raises the question: What did influence voting when the choice was about the issue of Brexit rather than who governs? Culture is an umbrella concept that refers to values that are particularly relevant for identification with groups, whether national or trans-national. Unlike citizenship, cultural identities and attitudes are not determined by law but by social choices. Table 5.2 contains five measures of attachment to national culture that are likely to favour voting Leave and three indicators of cosmopolitan characteristics disposing people to favour the EU and pro-Remain parties. Brexit campaigners have consistently emphasised taking back control of government from Brussels and have done so with an English accent. The 2017 BES survey offered people the choice of describing themselves as English, British or a mixture of the two. A total of 24% said they were English not British, or more English than British. Only four of 2194 respondents described themselves as first of all Europeans. English identity had the predicted effect on voting. Among those who saw themselves as English, there was a 66% probability of voting for Brexit in the referendum and a 67% probability of voting for a Brexit party in the 2019

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

89

EP election. While the impact at the individual level was not among the highest, because England has such a disproportionate share of the population of the UK, the 1.9 million majority for Brexit in England in the 2016 referendum could overwhelm the majority for Remain in Scotland and Northern Ireland. Multi-culturalism is the opposite of a homogeneous national culture. Before the UK entered the European Union, immigrants from the West Indies, the Indian subcontinent and Africa were beginning to make England multi-cultural. Significant immigration from other member states of the European Union became visible only after East Europeans came in significant numbers after their countries joined the EU. Cumulatively, these trends have turned Britain into a multi-cultural society. Because the European Union is associated in the public mind with the free movement of people, immigration confounds two contrasting sources—immigration from within the European Union and from distant continents and cultures. Immigration has stimulated a negative public reaction: 66% of BES respondents think that too many immigrants have been let into Britain. Pride in Britain’s political traditions contrasts with the wartime experience of the continental countries that founded European institutions to prevent a recurrence of a shameful political past (Almond and Verba 1963; Rose 1985). By the time of the 2016 Brexit referendum, pride in resisting military defeat was part of the distant past. Nonetheless, the Brexit campaign played up national pride and promised that leaving the EU would return Britain to its former global glory. When the BES asked whether young people had enough respect for traditional British values, 60% said that they did not. Respect for tradition gave a limited boost to Brexit voting. It raised the likelihood of doing so to 68% in the European Parliament election and fell just short of statistical significance in the referendum. However, it had no significant effect when the choice was which party should govern Britain. Even though most older Britons have been born in a post-war and post-imperial age, those over 65 when the 2016 referendum was held had been socialised into politics before Britain joined the European Union. Among older voters, there was a 78% likelihood of voting Conservative at the 2017 general election and a 67% likelihood of voting for a proBrexit party in 2019. After controlling for low income, class and other correlates of low education, it had no significant influence on party or on Brexit voting.

90

I. MCALLISTER AND R. ROSE

Popular attitudes to immigration are distinctive in having a similarly large impact on voting for parties and for Brexit. Among those who think Britain has had too many immigrants there was an 88% likelihood of voting for Brexit in the 2016 referendum and a 78% likelihood of voting for a pro-Brexit party in the European Parliament election. Even though the Conservative party has failed to meet its promise to reduce immigration by hundreds of thousands, there was likewise an 82% likelihood that people who thought immigration had gone too far would vote Conservative in the 2017 general election. Being cosmopolitan is the opposite of favouring a traditional English way of life. Having a higher education and being young are each linked with a cosmopolitan lifestyle involving food, having friends in other countries, and understanding foreign languages and non-British cultures. However, after controlling for other influences, having a higher education has no significant effect on voting for remaining in the European Union or for the Conservatives. Young people, regardless of their education, have been socialised into a more open and culturally diverse society. Likewise, after controlling for all other influences youth had no significant influence on the choice of which party governs or for leaving the EU. Young people were significantly more likely to vote for Remain parties in the 2019 EP election, but the effect was limited. Although London is by far the largest city in England and the capital of the UK, because of its global attractiveness it is the most multi-cultural and least English of Britain’s cities. The diversity of peoples in Greater London resulted in it being the only English region to favour remaining in the EU; 59.9% voted to remain in the 2016 Brexit referendum. Moreover, in the EP election 70.4% of London’s voters endorsed pro-Remain parties. This did not reflect a common London identity. After taking into account such important influences as immigration, income and Englishness, living in London had no significant influence on party voting or referendum choice, and a limited influence on voting for a Remain party in the European Parliament election. Partisanship is inherent in parliamentary elections: for this reason describing party identification as a cause of party choice risks confusing correlation with causation (cf. Dalton 2017). The nominal distinction between party identification and vote is a distinction with little empirical difference. A total of 83% of BES respondents who identified themselves as Conservative voted for the party in the 2017 parliamentary election, as did 87% of Labour identifiers.

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

91

Since the great majority of Britons formed their party identification long before the 2016 Brexit referendum, it could influence referendum voting even though party labels were not on the ballot. However, both parties were internally divided. Conservatives showed their divisions by loudly banging on about Europe, while Labour was mute about Brexit since most Labour MPs disagreed with the position of their party leader, Jeremy Corbyn. The consequence of these divisions is that party identification had no significant influence on either referendum voting or voting for MEPs at the 2019 European Parliament election. The intensity of Brexit as an issue has resulted in more Britons identifying with the Remain or Leave side than with a political party. Following the referendum, 89% identified themselves with their fellow Leave or Remain compared to 64% identifying with a political party (UKICE 2019: 6–24). Moreover, there is a stronger feeling about identifying with the issue than with a political party. In parliamentary elections offering a choice of parties, individual identification with their referendum choice was substantial. The likelihood of those who identified with Leave voting Conservative was 69% in 2017 and rose to 93% in favour of Brexit parties in the 2019 EP election. The effect of favouring Leave remained high in the December 2019 general election because Boris Johnson turned it into a referendum on getting Brexit done. Partisanship need not be restricted to positive identification with a political party: people can also be anti-party. Throughout Europe mainstream parties are under attack because they ignore the views of the people they are supposed to represent. A total of 46% of 2017 BES respondents agreed that politicians don’t care what people like themselves think. After controlling for all other variables, being anti-party fell short of being a statistically significant influence on voting in all three elections. This suggests that voting against the EU was not so much a protest vote against British politicians being uncaring as a rejection of Britain remaining part of what Leave voters see as an un-English and pro-immigrant institution. Good Fit But Different Combinations of Influence Combining economic and cultural influences in a multivariate logistic regression produces a good overall account of what influences individual voters in elections that differ institutionally. The analysis of voting in the Brexit referendum accounts for 43% of the variation in the vote; in the parliamentary election that followed it accounts for 55% of the variation;

92

I. MCALLISTER AND R. ROSE

and in the 2019 European Parliament vote the figure is 66%. However, the chief influences on how individuals vote differ between electoral institutions and issues. Table 5.3 displays the impact of statistically significant influences on voting in three types of election: a winner-take-all referendum about an issue; choosing a party to govern in a parliamentary ballot; and choosing members of the European Parliament by proportional representation. The evidence supports the second hypothesis: when institutions and issues change, influences on the behaviour of individual voters change. Only one of 14 measures consistently influences voting across all three types of elections. People who think Britain has too many immigrants were more likely to vote to leave the EU, to favour the Conservatives in the 2017 parliamentary election, and to vote for Brexit parties in the EP ballot. The impact of changing institutions is sharpest for class and ideology. Seven of its eight indicators had a significant influence on voting for party Table 5.3 Influences change when institutions and issues change

Probability of voting

Class, ideology High income Low income Working-class identity Right-wing Left-wing Private ownership best State ownership best National culture English identifier Respect tradition Old 65 and over Too many immigrants Voted Leave Cosmopolitan culture Young 18–25 London

Con 2017

Leave 2016

Leave 2019

76 25 30 77 18 74 24

n.s. n.s. n.s. n.s. 34 n.s. n.s.

n.s. n.s. n.s. 68 n.s. n.s. 34

n.s. n.s. 78 82 69

66 n.s. n.s. 88 n.a

67 68 67 78 7

n.s. n.s.

n.s. n.s.

36 37

Significance level 2016, 2017, p < .00; 2019 EU election p < .000; n.s. indicates not significant Source Logit regressions in Table 5.2. No partisanship variable tested had a significant influence on voting in any election

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

93

government in the 2017 election. Individuals with a high income and a right-wing ideology, without a working-class identification, and preferring private rather than state ownership were significantly more likely to vote Conservative. However, only one of these eight variables influenced voting in the referendum. While two measures of ideology were significant in the EP election, the Brexit Party rather than the Conservatives benefited most as the front-running pro-Leave party. The impact of these three variables on voting was less than that on support for a Conservative or Labour government in 2017. Although the 2019 parliamentary election was the same in form as its predecessors, the prime minister chose to conduct it like a referendum to get Brexit done (see Chapter 12). By contrast with the 2017 general election when class was important for party choice, it had no significant influence on voting in the 2019 election. Among voters in the lowest income bracket, 45% supported the Conservative party, giving it an advantage of 11 percentage points over Labour. By contrast, in 2015 when Brexit was hardly an issue, the Labour party had a 7% lead over the Conservatives. Moreover, in the 2019 referendum the Conservatives had a lead in all social classes. Among middle-class respondents, 10% more supported the Conservatives rather than Labour and among working-class respondents the Conservatives had an advantage of 15% over Labour. Whereas in 2015 the Labour party had a lead of 8% over the Conservatives among semiskilled and unskilled workers, in 2019 the Conservatives had a lead of 13 percentage points in this group (Chorley 2019; Cowley and Kavanagh 2016: Table 14.2). Cultural influences rather than class and ideology had the biggest impact on referendum voting. This is strikingly evident when Britons used the occasion of the European Parliament election to express their views on Brexit, placing Nigel Farage’s party and the pro-Remain Liberal Democrats first and second (see Table 5.1). The class and ideological variables that so influenced party choice in the 2017 parliamentary election had virtually no effect on voting in a European election. Seven measures of national and cosmopolitan culture had a significant effect: identification with England, respect for traditions, being older and wanting fewer immigrants boosted support for Brexit parties, while younger voters and Londoners were more likely to favour Remain parties. The impact on voters of changing the issue can be tested by comparing two elections when voters were asked to choose between competing parties. While this was the case in the 2017 general election and the 2019 EP

94

I. MCALLISTER AND R. ROSE

vote, different issues were dominant. The former ballot decided whether the Conservative or Labour party would govern Britain, whereas the latter decided which parties would express the views of British voters in a multi-national parliament. The use of the first-past-the-post electoral system to allocate seats to parties favoured the concentration of votes on governing party and the biggest opposition party in the general election. By contrast, the EP’s use of proportional representation gave individuals the opportunity to express their support for the party that came closest to their views on a major issue such as immigration without affecting the choice of government (Neilsen and Franklin 2017). Changing institutions and issues radically altered the behaviour of voters. In the 2017 election to decide who governs, the Conservative and Labour parties together won 82% of the vote. By contrast, in the election that allowed voters to express their views about Europe when uncertainty about the outcome of the Brexit referendum was highest, the two parties of government secured only 23% of the vote. Labour finished third and the Conservatives fifth. The front-running parties were those with the most clearly defined views for or against leaving the European Union. The Brexit Party and the remnants of UKIP came first, with almost five times the share of the vote as the Conservatives. The Liberal Democrats came second and along with the Green party took almost one-third of the EP vote. Collectively these three groups won 67.3% of the EP vote, compared to 10.8% when the election issue was whether the Conservative or Labour party should control British government. There is a correlation between party identification and EP choices in the EP election, but it is opposite to what theories of partisan voting would predict. Among people who identified with the Conservatives, 59% voted for the Brexit Party and only 21% voted Tory. Among Labour identifiers, 47% voted for pro-Remain parties and 14% cast their votes for Nigel Farage’s pro-Brexit parties. Only three in eight of Labour identifiers actually voted Labour. The December 2019 British general election that followed shortly after the EP contest had the same parties competing. However, Boris Johnson framed it as a hybrid contest that combined the institutional focus on who governs Britain with the referendum focus on getting Brexit done. Since Labour’s EU policy was ambivalent, as in the 2017 general election its manifesto emphasised left-wing class and economic issues. Opinion polls showed that voters agreed with Johnson’s approach, naming Brexit as the most important election issue.

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

95

The change in the method of voting and the outcome of choice produced radical changes between the two 2019 elections. Because the general election was to choose which party would govern, the combined vote for the Conservative and Labour parties rose from 23% in May to almost 76% in December. For pro-Brexit voters, the change in institutions shifted the issue from which party best expresses your views about Brexit to which party is most likely to form a government that can deliver Brexit. While the total pro-Brexit vote went up by only 2%, its distribution between parties was transformed. The focus on government boosted the Conservative vote from 9.1 to 43.6% while the Brexit parties, which had no MPs, plummeted from 34.9 to 2.0%. The change in institutional focus created cross-pressures among voters who favoured a soft Brexit or no Brexit, because the only way to replace Johnson in Downing Street was to vote for a Labour party that had an ambivalent policy towards Brexit. This was enough to boost Labour’s vote from 14.1% at the EP election in May to 32.1% in December. However, the greater salience of Brexit as against class issues in 2019 meant that Labour lost one-fifth of its vote and parliamentary seats by comparison with the 2017 general election. Because they had no chance of taking control of British government, the collective vote for the parties that most clearly expressed pro-EU views, the Liberal Democrats and the Green party, fell from 32.4% in the EP election to 14.3%. Moreover, the change from a proportional representation to a first-past-the-post electoral system gave the two parties less than 2% of MPs in the British parliament compared to winning one-third of British MEPs, seats they were doomed to lose by the Conservatives delivering Brexit in January 2020.

5.4

Voters Are Discriminating

When elections are held under different institutional rules, voters are immediately aware that things have changed, because they must mark their ballot differently than in a conventional parliamentary election. Instead of being asked to make a familiar choice between parties competing for government, in the 2016 referendum voters were asked to make an unprecedented choice about leaving or remaining in the European Union. In the following year’s parliamentary election, the ballot reverted to the

96

I. MCALLISTER AND R. ROSE

familiar alternative of offering a choice between party candidates competing for control of government. In the 2019 EP election, voters were confronted with a third type of ballot requiring them to cast a vote for a party list seeking to win seats through proportional representation. Although the form of the December 2019 ballot was the same as in 2017 parliamentary election, the result showed that the Conservatives had succeeded in implanting in the minds of many voters that the effective choice was about getting Brexit done or leaving that, and much else, up to Jeremy Corbyn. In theory, institutional differences could cause voter confusion, evidenced by lower turnout or an increase in invalid ballot papers. However, this was not the case. Turnout at the 2016 referendum on EU membership was 72.2%, higher than at any parliamentary election since 1992 and higher too than at the 2017 and 2019 parliamentary contests that followed. The number of invalid referendum ballots, 25,359, was onequarter that at the 2015 parliamentary election. The turnout at the European Parliament election was 36.9%, a bit more than half that of a parliamentary election and an increase on turnout at the two preceding EP elections. Although electoral institutions and issues change, individuals remain the same, particularly over the three years when Britons were asked to vote in four different electoral contexts. Thus, instead of relying on their party identification, class and ideology, they must learn to discriminate between what is and is not relevant in the new situation created by an issue overriding traditional partisan divisions and by a ballot that must be marked in a different way. Our analysis shows that British voters are able to respond discriminatingly, and this is true among people with a minimum of education as well as among those with a higher education. The influences that affect their behaviour vary when institutions and issues change. When data become available for multivariate regression analysis of the December 2019 election, it is likely to reinforce this finding, since there is already clear evidence that class had no influence on party choice and that referendum divisions did. Ironically, getting Brexit done will force the pro-Brexit and proRemain parties to change the issues they offer voters. By 2024, when the Conservative government’s term of office expires, Britain’s departure from the EU will be a fact. Even though the institutional framework for choosing which party governs will be the same, voters will face a choice: whether to rely on class and ideology to influence their vote or whether to make a choice that reflects their national or cosmopolitan culture.

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

97

Appendix See Table 5.4. Table 5.4 List of BES variables, definitions and means Variables

Definition in BES 2017 (all scored 0 or 1)

Mean BES 2017

Mean BES 2019

High income

£60,000 or more per year Less than £16,000 per year Belongs to middle class Belongs to working class 8, 9 or 10 on 0–10 point scale 0, 1 or 2 on 0–10 point scale Strongly agree or agree Strongly agree or agree English not British or more English than British Strongly agree or agree Aged 65 or more No qualifications Too many immigrants let into country Voted leave Aged 18–25 Graduate or postgraduate qualification Feels close or a little closer to Conservative party Feels close or a little closer to Labour party Strongly agree or agree

.192

.178

.153

.168

.242 .348

.352 .414

.130

.103

.132

.167

.347 .427 .225

.263 .497 .343

.574 .303 .194 .656

.567 .236 .077 .437

.499 .125 .394

.526 .107 .510

.439

.272

.440

.275

.462

.720

Low income Middle-class identity Working-class identity Right-wing Left-wing Private enterprise best State ownership best English identifier

No respect tradition Old 65 or over Minimum education Too many immigrants Voted leave Young, 18–25 University education

Conservative identifier

Labour identifier Politicians don’t care

Source 2017 British Election Study (N = 2914), 2019 British Election Study Wave 16 (N = 37,951)

98

I. MCALLISTER AND R. ROSE

References Almond, G.A., and S. Verba. 1963. The Civic Culture. Princeton: Princeton University Press. Butler, D., and D. Stokes. 1970. Political Change in Britain. London: Macmillan. Chorley, M. 2019. Working Class Switched to Tories. The Times, 17 December. Clarke, H.D., M. Goodwin, and P. Whiteley. 2017. Brexit: Why Britain Voted to Leave the European Union. Cambridge: Cambridge University Press. Cowley, P., and D. Kavanagh. 2016. The British General Election of 2015. Basingstoke: Palgrave Macmillan. Dalton, R. J. 2017. Party Identification and Its Implications. Oxford Research Encyclopedia. 10.1093/acrefore.9780190228637.013.72. de Vries, C. 2018. The Cosmopolitan-Parochial Divide: Changing Patterns of Party and Electoral Competition in the Netherlands and Beyond. Journal of European Public Policy 25 (11): 1541–1565. Eatwell, R., and M. Goodwin. 2018. National Populism: The Revolt Against Liberal Democracy. London: Pelican. Evans, G., and J. Tilley. 2012. The Depoliticization of Inequality and Redistribution: Explaining the Decline of Class Voting. Journal of Politics 74 (5): 963–976. Fisher, J., E. Fieldhouse, M.N. Franklin, R. Gibson, M. Cantijoch, and C. Wlezien (eds.). 2018. The Routledge Handbook of Elections, Voting Behavior and Public Opinion. London: Routledge. Hobolt, S.B. 2009. Europe in Question: Referendums on European Integration. Oxford: Oxford University Press. Hutter, S., E. Grande, and H. Kriesi (eds.). 2016. Politicising Europe: Mass Politics and Integration. Cambridge: Cambridge University Press. Kaufman, E. 2018. White Shift: Populism, Immigration and the Future of White Majorities. London: Allen Lane. Lewis-Beck, M.S., R. Nadeau, and M. Foucault. 2013. The Compleat Economic Voter. British Journal of Political Science 43 (2): 241–261. Lipset, S.M., and S. Rokkan (eds.). 1967. Party Systems and Voter Alignments. New York: Free Press. Marsh, M. 2018. Voting Behavior in Referendums. In Routledge Handbook of Elections, Voting Behaviour and Public Opinion, ed. J. Fisher et al., 256–266. London: Taylor & Francis. Mau, S., and R. Verwiebe. 2010. European Societies: Mapping Structure and Change. Bristol: Policy Press. Mudde, C., and C.R. Kaltwasser. 2017. Populism: A Very Short Introduction. Oxford: Oxford University Press. Müller, J.-W. 2016. What Is Populism? London: Penguin Books.

5

WHEN INSTITUTIONS AND ISSUES CHANGE, VOTING CHANGES

99

Neilsen, J.H., and M. Franklin (eds.). 2017. The Eurosceptic 2014 European Parliament Elections: Second Order or Second Rate? London: Palgrave Macmillan. Pulzer, P. 1967. Political Representation and Elections in Britain. London: Allen & Unwin. Rose, R. 1968. Class and Party Divisions: Britain as a Test Case. Sociology 2 (2): 129–162. Rose, R. (ed.). 1974. Electoral Behavior: A Comparative Analysis. New York: Free Press. Rose, R. 1985. National Pride in Cross-National Perspective. International Social Science Journal 37 (1): 85–96. Savage, M., Fiona Devine, Niall Cunningham, Mark Taylor, Yaojun Li, Johs Hjellbrekke, Brigitte Le Roux, Sam Friedman, and Andrew Miles. 2013. A New Model of Social Class? Findings from the BBC’s Great British Class Survey Experiment. Sociology 47 (2): 219–250. Sciarini, P. 2018. Voting Behaviour in Direct Democracy Votes. In The Routledge Handbook to Referendums and Direct Democracy, ed. L. Morel and M. Qvortrup, 289–306. London: Routledge. Seldon, A. 2019. May at Ten. London: Biteback. Uberoi, E. 2016. European Union Referendum 2016. London: House of Commons Briefing Paper 7639. UKICE. 2019. Brexit and Public Opinion 2019. London: United Kingdom in a Changing Europe. Waldegrave, W. 2019. Three Circles into One. London: Mensch Publishers. Young, H. 1998. This Blessed Plot: Britain and Europe from Churchill to Blair. London: Macmillan.

CHAPTER 6

How Politicians Ought to Talk About Europe: Lessons Learned from Experimental Evidence Konstantin Vössing

Referendums about the political authority of the European Union offer campaigners the opportunity to establish issue links between Europe and other topics, which can be more or less closely related. Because of this, and also as the result of a long tradition of second-ordering European issues to national affairs (Reif and Schmitt 1980), the choices of voters in referendums about Europe are never entirely about the issue at hand (Hobolt 2005). This chapter focuses on the justifications that politicians use to explain their policy positions in referendums about Europe. Policy justifications establish issue links between a policy and a political goal by claiming that the policy is instrumental for achieving the goal in question (see Fig. 6.1). This can entail specific material benefits (e.g. economic growth) as well as abstract political goals (e.g. social justice). Alternatively, politicians may justify their position by pointing out norms that

advances Policy

Goal justifies

Fig. 6.1 EU justifications and support for the European Union

© The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_6

101

102

K. VÖSSING

motivate them, such as party loyalty or religious conviction. In an entirely different explanatory strategy, politicians can also defend their position by using excuses (Bennett 1980; McGraw 2002). Theresa May ended up doing this by arguing that the terms for Brexit were less than ideal but that there was no alternative. Issue links between policies and goals in policy justifications stem from strategic political action as well as habit and ideology. However, irrespective of politicians’ motives, the issue links they deploy are a powerful tool for winning voters’ hearts and minds. The Leave campaign, for instance, successfully issue-linked the policy of leaving the EU to the goal of controlling the influx of immigrants by proposing that Brexit would allow immigration to be curbed. Policy justifications and the issue links they establish are a case of ‘persuasive communication’ (McGuire 1985). They have certain features that make them more or less effective for winning hearts and minds, that is, for changing or reinforcing people’s attitudes. This chapter sets out five lessons that pro-Europe campaigners could follow when attempting to win hearts and minds. The evidence I use to analyse political rhetoric and develop these lessons comes from laboratory experiments conducted between 2014 and 2018 with groups of people who represent the entire diversity of the German voting-age population. While opinions expressed in these experiments are not necessarily representative, the mechanisms of opinion formation and the lessons that can be derived are broadly applicable. The focus on Europhile politicians is not a normative statement. It is simply the result of experimental design choices. To keep the experiments as straightforward as possible, they all feature politicians justifying the extension rather than the reduction of European political authority.

6.1

Stop Believing in the Existence of a Rhetorical Silver Bullet

Referendum campaigns are about specific policy alternatives, typically broken down to simple dichotomous choices, such as ‘Should the United Kingdom remain a member of the European Union or leave the European Union?’ (Remain versus Leave). Given the nature of the contest, campaigners will have no choice but to align themselves with one of two opposing alternatives, even if they privately prefer an in-between choice. This restriction is critical, as calibrating policy preferences in between opposite poles is a popular technique to enhance the persuasiveness of

6

HOW POLITICIANS OUGHT TO TALK ABOUT EUROPE …

103

political rhetoric. However, this is not an option in referendum campaigns. Effective links between the pre-determined policy alternatives and other political issues are therefore crucial for political rhetoric to be persuasive in referendum campaigns. Politicians might be tempted to believe, out of incompetence, habit or conviction, that there is such a thing as an overarching winning formula— an issue link that can become the rhetorical silver bullet of the campaign, overriding all other policy justifications. Some campaigners stick to their presumed silver bullet because they are unwilling or unable to listen to their constituents or analyse their preferences. Others are over-socialised into certain habits and scripts of political communication they cannot easily escape. A third group of campaigners are ideologues so convinced of a certain issue link that they are oblivious to the more sceptical views of the world around them. Avoiding these mistakes is critical for boosting the persuasive effectiveness of pro-Europe campaigners. The unquestioned belief in a rhetorical silver bullet prevents an analytical and strategic approach to political communication. Most importantly, in the end it will fail to deliver. The ‘silver bullet’ lesson is supported by experimental evidence. The experiments discussed in this chapter were conducted with groups of diverse participants that are representative of the range of the German voting-age population on all relevant demographic, social, political and economic factors. Completing each experiment took about 20 minutes, and the protection of participants, including appropriate debriefing, is guaranteed according to the guidelines prescribed by an Institutional Review Board (IRB). The experiments were conducted in the experimental laboratory at the Institute of Social and Political Sciences at Humboldt University Berlin. In one of the experimental studies, I randomly assigned participants to different versions of a mock newspaper article, which discusses a vote in the European Parliament (EP) on a resolution demanding the introduction of a Euro-tax. The article mentions Herbert Brueckner, a fictional member of the European Parliament (MEP), as one of the legislators who had voted in favour of the resolution. The MEP explains his decision using a justification. He says that he voted in favour of introducing the Euro-tax because he believes that it will help to achieve an important political goal. The experimental treatment consisted of randomly assigning participants to one of five different goals said to be achieved by the Euro-tax policy: European identity, free markets, social justice, national

104

K. VÖSSING

economic benefits or European economic benefits. These five goals were selected to cover a wide range of linkable issues, and to comprise intangible norms (the first three) as well as specific benefits (the final two). The mock article describes the Euro-tax as a measure that would not produce higher levels of taxation or greater government intervention in the economy, but instead shift fiscal authority from the national level to the European Union.1 It is thus indicative and representative of the kind of question that voters are presented with in the typical referendum about European integration: Do you support or reject the European Union having more or less political authority? The participants in the experiment were asked to indicate how much they support the European Union (EU) on a feeling thermometer after reading the manipulated newspaper article. The question reads as follows: ‘We would like to know your overall view of the European Union (EU). We would like you to rate the EU on a “feeling thermometer”. Numbers between 50 and 100 mean that you have a favourable opinion or “have warm feelings” about the EU. The higher the number, the more positive your feelings. Numbers between 0 and 50 mean that your opinion is unfavourable or that you “have cold feelings” about the EU. The lower the number, the more negative your feelings’. Analysis of responses showed there is no rhetorical silver bullet because none of the five justifications for more European integration dominates the others in its ability to change opinions. As shown in Fig. 6.2, support for the EU does vary between the groups of participants that received different policy justifications, but the differences are small and not statistically significant. An ANOVA test of the effect of the categorical variable comprising the five policy justifications on EU support yields an F-value of only 1.27 (p = 0.28), and post hoc tests show that none of the differences in EU support between any combination of two justifications approaches standard levels of statistical significance. Even widely varying types of justifications, which suggest very different arguments about the consequences of European integration, do not automatically translate into varying levels of persuasive effectiveness.

1 Comparable policies used in similar experimental studies include, for instance, the delegation of political authority over employment policy to the European Union (Vössing and Weber 2016, 2017) or, alternatively, general assessments of the status and ramifications of European integration (Vössing 2015).

6

HOW POLITICIANS OUGHT TO TALK ABOUT EUROPE …

105

Fig. 6.2 Public policy justifications and support for the European Union (Notes Graph shows mean value of support for the EU on a 0–100 feeling thermometer for different justifications used in the treatment article. The F statistic is 1.27 [p = 0.28]. None of the differences between the five justifications is statistically significant [p < 0.05] in a post hoc test [Bonferroni-adjusted for multiple comparisons])

6.2 Know How Much Room for Manoeuvre You Have People already have ideas and opinions about Europe before politicians start talking to them in referendum campaigns. Whether they like or dislike the European Union before the beginning of a campaign should be a powerful predictor of their views about the EU at the end of the campaign and their subsequent referendum vote. The extent to which pre-existing views of Europe will determine vote choices thus sets the boundaries of persuasion that campaigners have at their disposal. The size of the corridor and as a result the room for manoeuvre afforded to campaigners vary between specific referendum campaigns. The more the final vote depends

106

K. VÖSSING

on prior levels of support for the EU, the less impact a campaign has had on winning (new) hearts and minds. The exact amount of manoeuvring space—that is, the size of the effect of pre-existing views about Europe on the eventual referendum votes— can be determined only after the campaign is over. Moreover, the room for manoeuvre that defines a particular referendum is endogenous to the campaign. However, pro-European campaigners can still make educated guesses (or estimates informed by social scientific analysis) about the room for manoeuvre they can expect. Doing so makes it possible for campaigners to develop realistic strategies that take into account the malleability of varying groups of voters when they design and deploy rhetorical strategies. In another study using the experimental procedures described above, I asked participants to express their views of the European Union before reading the manipulated newspaper article. Because this pre-treatment measure of EU support needs to be equivalent in substance but different in implementation compared to the post-treatment thermometer rating of the EU, I used a semantic association test adapted from Castano et al. (2003) to estimate people’s pre-existing EU support. Participants were asked to select five words that ‘best describe their thoughts about the European Union’ from a list of words that are positive (enthusiasm, satisfaction, trust, appreciation, approval), neutral (disinterest, indifference, detachment, aloofness, neutrality) and negative (uneasiness, irritation, distrust, anger, rejection). Every positive word listed by participants receives a score of +1, the neutral ones 0 and the negative ones −1. An estimate of people’s pre-existing support for the EU is constructed by summing the scores and rescaling them to a scale from 0 (lowest level of pre-existing support for the European Union) to 1 (greatest level of support). The average value for the 277 participants is 0.6 (with a standard deviation of 0.25). The first model displayed in Table 6.1 shows that the pre-existing level of support for the EU has a large and statistically significant effect on the eventual level of EU support, which is recorded after the treatment is administered. Forty-seven per cent of the variance in support for the EU is explained by what people thought about the EU before they received the new information and the policy justification of the MEP that was embedded into the mock newspaper article. The other models shown in Table 6.1 include a range of additional variables related to the content of the policy justification and the party affiliation of the MEP. The effect of

6

HOW POLITICIANS OUGHT TO TALK ABOUT EUROPE …

107

Table 6.1 Policy justifications and support for the European Union Model

(1)

(2)

(3)

(4)

(5)

(6)

Pre-existing EU support Support for justifying goal Validity of justification Party EU-issue ownership Party goal-issue ownership Party identification Constant

52.9 (3.5)**

52.0 (3.5)** 7.8 (3.2)**

50.4 (3.1)** 6.2 (3.4)*

51.8 (3.6)**

52.0 (3.6)**

51.9 (3.6)**

R2 N

9.2 (4.0)** 3.2 (4.2) −0.36 (3.6) 4.9 (3.5) 36.9 (2.3)** 0.47 264

33.1 (2.7)** 0.48 264

29.4 (3.1)** 0.49 260

35.4 (3.6)** 0.45 260

37.7 (3.2)** 0.46 259

35.0 (2.7)** 0.47 262

Marginally significant *p < 0.10; Statistically significant **p < 0.05 Note Cell entries are non-standardised OLS regression coefficients and standard errors for independent variables on a scale of 0–1

pre-existing support for the EU remains stable in all these model specifications. The effect sizes observed in the analyses are not necessarily good estimates of the specific room for manoeuvre politicians have in any given referendum campaign. Nonetheless, the results do show that there is such a thing as a confined corridor of persuasion, and that it is very robust and unaffected by new information and arguments.

6.3 Issue-Linking European Integration to Popular Goals One key mechanism of persuasion that campaigners have at their disposal, issue-linking, does have the capacity to win hearts and minds. Referendum campaigners can choose from a wide range of political goals that could be achieved by the policy they support and issue-link European integration with popular goals. For instance, campaigners can argue that remaining in the EU has economic benefits. They can suggest that a constitution for

108

K. VÖSSING

Europe secures human rights, that the Maastricht Treaty allows people to travel seamlessly across Europe or that the Treaty of Nice will facilitate the enlargement of the EU. Given the constraints on persuasion in referendum campaigns, successful issue links between European integration and popular political goals are of critical importance for campaigners. In my second experiment, I asked participants how they felt about the various goals invoked by the fictional MEP to justify support for European integration in different renditions of the treatment article. Participants answered these questions before reading the article to avoid an undue effect of the experimental treatment on this explanatory variable. Support for each justifying goal was measured by exposing participants to three or four thematically relevant statements. Using standard five-point Likert scales, participants were asked to express their level of agreement with these statements. The variables identifying support for European identity, social justice, free markets, European economic benefits and German economic benefits were constructed by averaging the scores for the relevant Likert-style questions and rescaling the values to a scale of 0 (the lowest possible level of support) to 1 (the highest level). For instance, ‘we would all be better off if the government did not interfere with the economy’ is one of the four statements that was used to measure people’s support for the goal of free markets.2 The variable identifying participants’ support for a justifying goal used in the empirical analysis is determined in reference to the particular goal that was invoked in his or her treatment condition. The second model displayed in Table 6.1 shows that the extent to which people like the goal invoked in a policy justification has a significant and sizable effect on their acceptance of a justification for the advancement of European integration. The more people like a goal that politicians issue-link to European integration, the more they will favour European integration itself. The step from the lowest to the highest possible level of support for different goals used in justifications for European integration increases support for the European Union by 7.8 points on a scale from 0 to 100. This shows that campaigners can use issue links to transfer positive sentiments about goals from other issue domains to the domain of European integration.

2 A complete list of statements and other methodological details is available in an online appendix deposited on the author’s website at https://sites.google.com/site/ konstantinvossing/ and on his dataverse page at https://dataverse.harvard.edu/dataverse/ vossing.

6

HOW POLITICIANS OUGHT TO TALK ABOUT EUROPE …

109

Have politicians campaigning for Europe already learned the lesson of issue-linking European integration to popular political goals? Most of the protagonists of the Remain campaign in the recent Brexit referendum did not learn it, in stark contrast to their opponents. Leave campaigners claimed that voting to leave the European Union would allow the UK to re-establish full control over the country’s borders and immigration policies. The issue link between European integration and immigration was a successful rhetorical strategy, as shown by evidence of the British Election Study survey analysed in Chapter 5. The Leave campaigners also justified their position by arguing that ‘voting to leave the EU will improve the quality of the National Health Service’. The claim that saving money by no longer paying into the EU budget would allow the UK government to invest more in the NHS was the second well-chosen issue link of the Leave campaign. The Remain campaign, by contrast, did not deploy a comparable popular issue link. Instead of invoking positive goals that would be achieved by remaining in the EU, the Remain campaign emphasised the negative consequences of leaving. In addition to the Brexit case study, there is further evidence that proEuropean politicians have a hard time issue-linking Europe to popular goals. Evidence comes from a quantitative text analysis of the speeches of the members of the European Convention during the debates about the drafting of a European constitution and speeches of members of the European Parliament during the plenary debates about open borders and the Euro crisis. The text analysis I conducted identifies more than 3000 justifications used by politicians: 38% of these justifications invoke goals allegedly achieved by European integration that qualify as intangible norms, such as democracy, human rights and European values; 25% invoke specific and measurable policy outputs; and 34% suggest that more European integration is needed to improve the performance of EU institutions. I also conducted a survey experiment with 1376 participants who are representative of the German voting-age population to see how popular the three different types of goals are. Justifications that invoke intangible norms feature an average level of support among German voters of 73 points on a 0- to 100-point thermometer scale. The reference to specific policy outputs receives an average rating of 59 points, and the goal of improving the performance of EU institutions has a popularity rating of

110

K. VÖSSING

56 points. Thus, only 38 per cent of all justifications for European integration invoke the most popular type of goal. Most justifications rely on significantly less popular goals to justify support for policies of European integration. Europhiles in the representative institutions of the European Union could do a better job at issue-linking European integration to popular political goals.

6.4

Use Valid Issue Links

The third component of a policy justification, in addition to the justified policy and the justifying goal, is the causal issue link between them. Politicians need to consider whether their causal claim about the effect of a policy on a political goal is perceived as valid by voters, in order to influence public opinion. The third model shown in Table 6.1 adds participants’ views about the validity of the policy justification offered by the MEP to the two variables included previously. In order to evaluate the validity of the policy justification proposed by the MEP in the manipulated newspaper article, they were asked: ‘regardless of whether you share this opinion or not, how valid did you find the justification offered by Herbert Brueckner for his decision (to support European integration)?’ Participants recorded their judgements on a 7-point scale ranging from ‘not valid at all’ to ‘entirely valid’. Their responses were then adjusted to a scale ranging from 0 (lowest validity judgement) to 1 (highest validity judgement). The perceived validity of the justification has a sizable and statistically significant effect on support for European integration. Plausible justifications that establish valid causal connections between an advertised policy of European integration and an issue-linked political goal increase support for the European Union. Implausible justifications, by contrast, diminish EU support. The change from the lowest to the highest judgement of validity accounts for an increase of 9.2 points in support for the European Union on a scale from 0 to 100. This effect occurs while pre-existing views about the EU and individual levels of support for the goal are held constant. Individual judgements of the validity of the issue link are independent of the popularity of the goal and the policy between which the issue link is established. Interestingly, model 3 also shows that including the validity variable diminishes the effect of goal support. The transfer of a justifying goal’s popularity into positive judgements of the justified policy depends to some extent on the perceived validity of the issue link between policy and

6

HOW POLITICIANS OUGHT TO TALK ABOUT EUROPE …

111

goal. The failure of Europhile referendum campaigners in the UK and elsewhere to deploy successful policy justifications is arguably rooted in the difficulty of devising justifications for policies of European integration that simultaneously issue-link European integration with popular goals and devise issue links between justifying goals and the justified policy of European integration that voters perceive as valid. It is hard to find popular goals that can be plausibly issue-linked to policies of European integration. On the one hand, there is a wide range of political goals that are clearly achieved by policies of European integration, such as unified product standards and the efficient management of external trade relations. Yet these goals are technical in nature, and they are examples of measurable policy outputs, which receive low popularity ratings. The political goals that fall into the category of intangible norms, such as democracy, human rights and European values, are highly popular. But it is much more difficult to show unequivocally that these goals are truly advanced by specific policies of European integration.

6.5

Don’t Expect Too Much of Your Party Label

In addition to talking about their positions, referendum campaigners also mention their party label, sometimes inadvertently and sometimes strategically, to sway voters. When parties make choices or express preferences, voters can evaluate these actions based on what they think about the party. The party thus offers cues that voters can follow, for example when they form their own opinions about whether to like the European Union or when they decide how to vote in a referendum. A large body of research has shown that party cues are an effective vehicle of preference formation in various issue domains (Cohen 2003; Petersen et al. 2013; Rahn 1993). Party cues allow voters an easy shortcut to forming opinions and making decisions that absolves them of the need to engage in extensive thinking about the issue.3 They can simply rely on the party label and whatever they associate with it, allowing the party label to guide them towards their opinions and choices.

3 Sometimes, as Petersen et al. (2013) have shown, voters will end up thinking intensely about the issue anyway, even after forming an opinion based on a party cue, especially when they feel the need to reconcile their partisanship with contradictory information about the issue.

112

K. VÖSSING

However, campaigners should not expect too much of invoking their party label in a referendum campaign. Notwithstanding differences in the clarity and strength of the signals parties in different systems provide to voters about European integration, mainstream parties all over Europe have traditionally avoided clear positioning on European integration due to pronounced within-party variation of opinions over the issue (Hix and Lord 1997; van der Eijk and Franklin 2004). This is why perceptions of competing party positions can be experimentally manipulated (Tilley and Wlezien 2008; Vössing and Weber 2016, 2017). Because voters are uncertain about mainstream party positions regarding European integration, they can hardly rely on party cues when voting in a referendum about a European issue (van der Brug and van der Eijk 1999; see also Chapter 5, Table 5.1, in this book). Participants in my experiments were given randomly assigned information about the party affiliation of the MEP justifying his positive view of European integration. The assigned party label varies between the four parties represented in German parliament at the time. Figure 6.3 shows that none of the four parties is inherently better at producing support for the European Union. The F-statistic identifying the effect of the categorical variable ‘party affiliation of the MEP’ (F = 0.43; p = 0.73) is clearly low, and none of the between-group comparisons of any two party categories reveals statistically significant differences. Three additional empirical tests displayed in Table 6.1 provide further evidence that voters do not rely on party cues. To begin with, models 4 and 5 test whether parties have issue-related reputations that voters rely on when forming their opinions. Model 4 includes a variable identifying the extent to which people believe that a party ‘owns’ the issue of European integration. For this variable, participants were asked, before reading the treatment article, to indicate whether the party assigned to them supports European integration on a scale from 0 (does not support European integration at all) to 10 (supports it a lot), and replies were rescaled to a range of 0–1. Petrocik (1996) and Tresch et al. (2015) suggest that issue ownership identifies parties as authentic and credible representatives of claims. One might expect that perceptions of greater issue ownership would make voters more trusting of the party’s views, including its position on European integration. However, no such effect occurs in the experiments. Model 4 finds that, while there is a correlation between a party’s ownership of the European integration issue and participants’ levels of EU support, it is not statistically significant.

6

HOW POLITICIANS OUGHT TO TALK ABOUT EUROPE …

113

Fig. 6.3 Political parties and support for the European Union (Notes Graph shows mean value of support for the EU on a 0–100 feeling thermometer for different justifications used in the treatment article. The F statistic is 0.43 [p = 0.73]. None of the differences between the five justifications is statistically significant [p < 0.05] in a post hoc test [Bonferroni-adjusted for multiple comparisons])

Model 5 tests the effect of a different type of issue ownership. It includes a variable identifying the extent to which people believe that a party ‘owns’ the issue domain of the issue-linked goal invoked by the MEP. The conclusion here is even clearer than for ownership of the integration issue. Variation in parties’ ownership of the issue domain of the goal with which they are aligned in a policy justification does not have a significant meaningful effect on opinions about European integration. The most powerful party cue is rooted in the extent to which a voter has an identification with a political party that can be relied on when forming opinions about political issues (Cohen 2003; Petersen et al. 2013; Rahn 1993). From this point of view, greater identification with the party

114

K. VÖSSING

of the MEP justifying his position should increase support for the European Union. To test this expectation, I asked participants before reading the treatment article how much they liked or disliked the political party to which they were about to be exposed in the experiment. Model 6 shows that, while identification with the party of the politician justifying a decision has a larger effect on support for the EU than the other party-related judgements, it still fails to achieve statistical significance when pre-existing attitudes of voters about the EU are held constant. By contrast, partisanship is significant when pre-existing attitudes are removed from the model (B = 11.2, SE = 4.6, p < 0.05). These results show that, while there is an alignment between people’s views of the EU and their party identification, it is not due to a cueing effect. For that to be the case, participants would have to adjust their opinions about Europe in line with their partisanship irrespective of the strength of their previous levels of support for the EU, and they do not do this. This conclusion highlights the role that Jeremy Corbyn, leader of the UK Labour party, performed in the Brexit controversy. Ardent Remainers attributed the outcome to Corbyn’s merely lukewarm support for the European Union and the Remain campaign. However, the analysis of model 6 suggests that parties have only a very limited capacity to cue their voters on the issue of European integration. Insofar as this holds true outside the experimental laboratory, even clear support by the leader of the Labour party for a Remain vote would not have led Labour identifiers to vote Remain in larger numbers. Political parties might have little influence to cue voters once a campaign has begun, but they can contribute in the long run. Had Corbyn and the Labour party provided clearer and stronger arguments in favour of European integration before the referendum campaign began, this might have convinced more of its supporters to vote Remain. This chapter has used a series of laboratory experiments with diverse samples of participants to generate lessons for how politicians ought to talk about Europe to win hearts and minds in referendum campaigns. The experiments show that referendum votes about the European Union depend, more than anything else, on the opinions about Europe that voters already hold before the beginning of the campaign. Instead of seeking a silver bullet, referendum campaigners should be keen to listen to their constituents or use social scientific methods to understand the arguments they find convincing instead of relying uncritically on traditional but ineffective justifications. The experimental studies also caution that the room

6

HOW POLITICIANS OUGHT TO TALK ABOUT EUROPE …

115

for manoeuvre by campaigners is restricted by the extent to which the preferences of voters are malleable. Calibrating policy positions is not an option in referendum campaigns, given that they feature dichotomous choice alternatives. This is why the two most important mechanisms of persuading voters in referendum campaigns are related to the goals that campaigners invoke to justify their preferred referendum vote. My experimental studies emphasise the importance of invoking goals that are viewed as desirable by voters rather than concentrating on those that the campaigners find congenial. For a justification to be effective, voters need to believe that campaigners are advocating a good policy that can credibly have a positive effect on a desirable political goal. Following these lessons individually is easier said than done, but a real dilemma emerges when campaigners try to follow several at the same time. Many goals that are plausibly achieved by European integration are not popular, and for many popular goals it is hard to show convincingly that European integration contributes to advancing them. The final lesson learned from the experimental evidence analysed in this chapter is not to expect much from a party label; it has very little power to sway people’s opinions on European integration.

References Bennett, W.L. 1980. The Paradox of Public Discourse: A Framework for the Analysis of Political Accounts. Journal of Politics 42 (3): 792–817. Castano, E., V. Yzerbyt, and D. Bourguignon. 2003. We Are One and I Like It: The Impact of Ingroup Entitativity on Ingroup Identification. European Journal of Social Psychology 33: 735–754. Cohen, G. 2003. Party over Policy: The Dominating Impact of Group Influence on Political Beliefs. Journal of Personality and Social Psychology 85 (5): 808– 822. Hix, S., and C. Lord. 1997. Political Parties in the European Union. New York: Macmillan. Hobolt, S. 2005. When Europe Matters: The Impact of Political Information on Voting Behaviour in EU Referendums. Journal of Elections, Public Opinion and Parties 15 (1): 85–109. McGraw, K. 2002. Manipulating Public Opinion. In Understanding Public Opinion, ed. B. Norrander and C. Wilcox, 265–280. Washington, DC: CQ Press. McGuire, W. 1985. Attitudes and Attitude Change. In Handbook of Social Psychology, ed. G. Lindzey and E. Aronson, 233–346. New York: Random House.

116

K. VÖSSING

Petersen, M., M. Skov, S. Serritzlew, and T. Ramsøy. 2013. Motivated Reasoning and Political Parties: Evidence for Increased Processing in the Face of Party Cues. Political Behavior 35 (4): 831–854. Petrocik, J. 1996. Issue Ownership in Presidential Elections, with a 1980 Case Study. American Journal of Political Science 40 (3): 825–850. Rahn, W. 1993. The Role of Partisan Stereotypes in Information Processing About Political Candidates. American Journal of Political Science 37 (2): 472– 496. Reif, K., and H. Schmitt. 1980. Nine Second-Order National Elections: A Conceptual Framework for the Analysis of European Election Results. European Journal of Political Research 81: 3–45. Tilley, J., and C. Wlezien. 2008. Does Political Information Matter? An Experimental Test Relating to Party Positions on Europe. Political Studies 56: 192– 214. Tresch, A., J. Lefevere, and S. Walgrave. 2015. “Steal Me if You Can!” The Impact of Campaign Messages on Associative Issue Ownership. Party Politics 21 (2): 198–208. van der Brug, W., and C. van der Eijk. 1999. The Cognitive Basis of Voting. In Political Representation and Legitimacy in the European Union, ed. H. Schmitt and J. Thomassen, 129–160. Oxford: Oxford University Press. van der Eijk, C., and M. Franklin. 2004. Potential for Contestation on European Matters at National Elections in Europe. In European Integration and Political Conflict, ed. G. Marks and M. Steenbergen, 32–50. Cambridge: Cambridge University Press. Vössing, K. 2015. Transforming Public Opinion About European Integration: Elite Influence and Its Limits. European Union Politics 16 (2): 157–175. Vössing, K., and T. Weber. 2016. The Company Makes the Feast: Party Conflict and Issue Voting in Multi-party Systems. In Voting Experiments, ed. A. Blais, J.-F. Laslier, and K. van der Straeten, 43–66. New York: Springer. Vössing, K., and T. Weber. 2017. Information Behavior and Political Preferences. British Journal of Political Science 49 (2): 533–556.

CHAPTER 7

Comparing Voting in National Referendums on EU Issues Alexia Katsanidou and Slaven Živkovi´c

Following the unwritten rules of EU-related referendums, parties in government at the time of the referendum advocated the pro-EU option in the Netherlands and in Britain. However, even though the British Conservatives, led by Prime Minister David Cameron, campaigned for the Remain option, there were some strong voices in the party that wanted the country to leave the European Union, most notably Boris Johnson. There were some unique, context-specific developments in other cases as well. In Greece, it was the scope and depth of the economic crisis along with the broader implications for the EU economy and the Eurozone. In the Netherlands, on the other hand, the whole nature of the referendum offered a unique context as there were various groups pushing for higher turnouts and campaigning on one side or the other, while the citizens were mainly uninterested. Nonetheless, all three referendums posed many similar dilemmas to citizens: opting for or against incumbents, a full acceptance or rejection of the EU, a positive or negative evaluation of the economy, and most interestingly acceptance or rejection of specific policies such as immigration. In this chapter, we explore common features that shaped citizens’ votes as well as explaining differences and similarities.

© The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_7

117

118

´ A. KATSANIDOU AND S. ŽIVKOVIC

7.1

Context and Framing of the Referendums Britain

The referendum on British exit from the European Union was held on 23 June 2016 and remains one of the most significant turmoils that has happened in Europe in the twenty-first century (Hobolt 2016). The UK became the first country to invoke Article 50 of the Treaty of Lisbon, as—contrary to all expectations and to most of the polls’ predictions— the referendum resulted in the country’s decision to leave the EU. The decision to hold a referendum was strongly driven by two main factors: (a) the increasing popularity of the UK Independence Party (UKIP) and (b) the growing number of MPs from the Conservative party putting pressure on David Cameron, who was prime minister at the time. He committed to keeping the party manifesto promise made during the 2015 election to organise a referendum on the UK’s membership of the EU, but himself vigorously campaigned for the UK to remain in the EU. The question that appeared on the ballot was: ‘Should the United Kingdom remain a member of the European Union or leave the European Union?’ Voters had to choose between two options: ‘Remain a member of the European Union’ or ‘Leave the European Union’. The Brexit referendum resulted in a very high turnout, 72%, the highest since the 1992 elections, and the highest recorded in a nationwide vote in many years (Goodwin and Heath 2016). The Leave option won 51.9% of votes, making the UK the first EU member state to invoke Article 50 to implement withdrawal from the European Union. The Netherlands Dutch citizens were called to vote on 6 April 2016 in a referendum on the Ukraine–European Union Association Agreement, the new legal framework that aims to establish closer co-operation between the parties (the European Union, Euratom, the twenty-eight EU member states and Ukraine) in the areas of politics and economics. The previous agreement between Ukraine and the EU (a Partnership and Cooperation Agreement) was widely perceived as an outdated document (Petrov and Van Elsuwege 2016). Several media reports from around the world stated that the real goal of the new agreement was for the EU’s exports not to be threatened by instability in the region, and for Ukraine to enjoy the benefits of free

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

119

trade with the EU and attract more investment from the West (van der Brug et al. 2018). The Dutch organisation GeenPeil (No Poll), which went from being a political initiative to a political party in 2017, collected the required number of signatures to initiate the referendum in the country. Under the Dutch Advisory Referendum Act, most laws can be subject to suspensory, non-binding referendums if requested by 300,000 citizens. The referendum is considered successful if the turnout is higher than 30% of eligible voters. After the Dutch Electoral Council verified the 472,849 signatures, the referendum was scheduled for 6 April 2016. The question on the ballot was: ‘Are you for or against the Approval Act of the Association Agreement between the European Union and Ukraine?’ The referendum had a rather small campaign. The government actively promoted the ‘Yes’ option but without mobilising a large-scale campaign. Several opposition parties (Party for Freedom, Socialist Party, Party for the Animals and VoorNederlanden) led the ‘No’ campaign. The GeenPeil organisation said they would lead a neutral campaign with their primary goal being a high turnout. On a turnout of 32.3%, the EU– Ukraine treaty was rejected, with 62% of voters casting ballots against, and 38% for the treaty. The result meant that the Netherlands could not ratify the agreement and that the whole agreement was stopped from entering into force. Moving forward from that rejection required additional acts and clarification to be added to the treaty, which the Netherlands negotiated with other parties. After these addendums were approved, the ratification was submitted to the Dutch parliament; it was approved in both houses, which allowed the Netherlands to deposit an instrument of ratification. All other parties did the same, and the agreement officially entered into force on 1 September 2017. Greece The Greek bailout referendum took place on 5 July 2015 and was another high point of the negotiations rollercoaster (Tsebelis 2016) between the Greek government and the so-called Troika (the European Commission, European Central Bank and International Monetary Fund). It was particularly interesting as it was simultaneously hailed as a highlight for democracy and as a populist firework undermining liberal democracy.

120

´ A. KATSANIDOU AND S. ŽIVKOVIC

Greek citizens were not used to referendums, as the previous one had taken place in 1974 with a clear constitutional question on the abolition of the monarchy. Thus this call for a referendum as well as the type of the question was considered confusing for the average citizen. The referendum asked Greeks to accept or reject two documents available only in English with the titles ‘Reforms for the Completion of the Current Programme and Beyond’ and ‘Preliminary Debt Sustainability Analysis’, the two possible answers being ‘Not approved/No’ and ‘Accepted/Yes’. These documents were basically proposals for new taxes, which did not add to their clarity, as the only choice was to either accept or to reject them both. The referendum was conducted a mere eight days after its initial proposal and caused great polarisation. It was deemed necessary as the government believed that the agreement with the Troika would not stand a chance of passing through the parliament. To add to the polarisation caused by the short time between the decision to hold a referendum and the date of the vote, the European Central Bank cut off the already limited liquidity to the Greek banks, leading to banks being shuttered and strict capital controls, while Greece defaulted on its IMF obligations. What is more, the referendum itself was on the verge of being unconstitutional, as the Greek constitution does not allow referendums on fiscal matters. The government claimed, however, that this was not simply a fiscal matter but a national emergency. Another confusing issue was that by the time the referendum took place the Troika’s documents were no longer valid proposals, as the financial conditions had already changed due to the capital controls. The turnout was 62%. The result gave a clear win to the No camp with 61%, while the Yes side managed 39%. The government campaigned against its own negotiations for No, while Europhile opposition groups such as the New Democracy centre-right party campaigned for Yes. The result of the referendum was clear. What was not clear was its meaning, so it was left to Prime Minister Alexis Tsipras to interpret it. He was then blamed for transforming the glorious No into a humiliating Yes with worse conditions than those offered in the initial rejected proposal. The three referendums described above have many differences. One posed the fundamental question of the country’s membership of the European Union (Britain), while the other two put into public deliberation the acceptance or the rejection of a specific proposal made by the EU

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

121

but under very different circumstances. What they have in common, however, is the fact that they do not offer two clear alternatives. The pro-EU side of the referendum is clear, as it offers the continuation of the existing status quo. For Britain the known alternative was to remain a member of the European Union; it meant no change in existing policy was necessary. For the Netherlands, accepting the Ukraine–European Union Association Agreement meant that the agreement would pass without further deliberation. For Greece, voting Yes in the referendum would be a clear indication that the Greeks would approve the negotiated agreement with the Troika and would continue the austerity course. The anti-EU options, however, were formulated as the rejection of the pro-EU path. They were not clearly formulated alternatives, but empty signifiers. The anti-EU alternatives in the referendums offered a clean slate. They were not only a clear-cut rejection of the status quo but they did something almost magic: they allowed every individual to imagine their own utopia. Citizens and elites could use the anti-EU option in order to construct a better or worse place, depending on their beliefs. Each camp projected their own understanding about what it would mean to vote for the anti-EU option. This was possible only because the public lacked specific information about the meaning and function of the European Union. In the case of the Brexit referendum, the British public lacked a basic understanding of what functions the EU was performing for them, and what would it mean, financially and in terms of governance, if it stopped doing so. Walter et al. (2018) in their work comparing several EU issue referendums demonstrated that there was significant divergence in the understanding of the meaning of an anti-EU vote (Fig. 7.1). The perceived consequences of an anti-EU outcome of each referendum are strikingly different among the camps. In Greece people who considered voting for the pro-EU option viewed the anti-EU alternative as a choice with grave consequences. Almost 60% of the Yes camp associated a potential No vote with Grexit. This percentage was around 5% in the No camp. There, the main interpretation (of almost 90% of No voters) was that a No vote would result in new negotiations with the EU. Similarly in Britain, more than 60% of Remain voters agreed that an anti-EU vote would mean Britain losing full access to the single market. A similar percentage in the Leave camp disagreed, believing that Britain would not lose full access to the single market. An analysis of data from the Dutch referendum shows a similar pattern in the differences between the pro- and the anti-agreement

122

´ A. KATSANIDOU AND S. ŽIVKOVIC

2015 GREEK BAILOUT REFERENDUM What do you think will be a consequence of a No vote? 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% Yes

No Grexit

Undecided

New negotiations

Don’t know / NA

2016 UK BREXIT REFERENDUM If Britain votes to leave the EU, the UK will lose full access to the EU’s Single Market. 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% Remain

Leave Agree

Disagree

Don't know

Don’t know

2016 DUTCH REFERENDUM: UKRAINE-EU ASSOCIATION AGREEMENT The Association Agreement will lead to EU membership for Ukraine 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% For Agreement Agree

Against Agreement Neither agree nor disagree

Non-voters Disagree

Fig. 7.1 Perceived referendum consequences—Greece, UK and the Netherlands (Source UK: BES, N: 5030. Netherlands: LISS panel, N: 1519; Greece Aristeia, N: 797)

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

123

camps about whether the agreement would lead to EU membership for Ukraine. More than 60% of the pro-agreement camp disagreed with that statement, while the reverse was true for the anti-agreement camp. All three examples demonstrate the same thing: the unknown option, the non-status quo, is a blank canvas on which citizens and elites can project their wishes and fears. The anti-EU campaigns capitalised on this and projected fear of further immigration and desire for protection of the national borders on the anti-EU choice in Britain and the Netherlands, while in Greece they used the anti-EU option to show their disagreement with the austerity measures imposed by the Troika. This is in line with the literature showing that positions in favour of or against the European Union are associated in the crisis-ridden countries with economic positions while in the northern countries, which are less affected by the crisis, they are associated with positions on immigration (Otjes and Katsanidou 2017). The pro-EU and the anti-EU camps offered their own visions of the future consequences of a pro- or anti-EU vote, and their supporters believed them. This is not very different from the effects of partisanship in framing reality in a parliamentary election, except that referendums directly influence a policy outcome.

7.2

Hypotheses and Data

Social and demographic characteristics have long been pillars of voting behaviour research (Lipset 1960). This topic has mostly been researched via the relationship between these characteristics and vote choice in national elections. Hypothesis 1 is that demographics, given that they are fixed characteristics of individual voters, will influence voting behaviour in referendums. Party identification is a well-established social-psychological influence on voting (Campbell et al. 1960). Partisanship is a somewhat different concept when applied to EU-related referendums. Citizens who identify with a party, but do not have the opportunity to vote for that particular party, are expected to vote for the alternative that their party endorses. Usually parties that are in government advocate the pro-EU option, while one or more opposition parties take the contrary view. Hypothesis 2 assumes that individuals who support a governing party will be more likely to vote for the pro-EU option.

124

´ A. KATSANIDOU AND S. ŽIVKOVIC

Even though the questions at stake at these three referendums were very different, it was obvious which was the option preferred by the EU. Hypothesis 3 is that the higher the voter’s opinion of the EU, the more likely she or he is to vote for the pro-EU option. The EU is perceived not only as an institution but also in terms of its policies and issues. These can sometimes overlap with national issues. Thus, the EU can be blamed for national problems, and citizens may see the EU as helpful in dealing with a major national problem. We include two key issues that have been most salient in recent years—the economy and immigration. We hypothesise that people who perceive the economy to be doing well, or that it will do well in the future (in case of Greece), will choose the more pro-EU option (Hypothesis 4a), as this is often the preferred option by the governing party. However, the increase in immigration is often perceived as a threat by those that have a more close-minded way of thinking, feel their culture is threatened or believe that European integration brings immigrants to their country. People who are against immigration, and think migrants pose a threat to their country (whether to its economy or its culture or both), will be more likely to vote against the EU-preferred option in EU-related referendums (Hypothesis 4b). Referendums, and particularly EU referendums, can be highly salient (as in the UK or Greek cases) or have low salience (in the Netherlands in our analysis). They can be contested on the battle lines between the government and the opposition or can have divisions within the governing party (Conservatives in the UK). Given differences in the national contexts of referendums on EU issues, we may find only partial support for the above hypotheses, such predictors performing better or differently in one country than in another. Hence, our final Hypothesis (5) is that cross-national comparability will be very limited. In order to test these ideas, we use three national sample surveys fielded not long after a national referendum. For the UK we used Wave 10 of the British Election Study (BES) Internet Panel with 30,319 respondents. We follow the recommendation of the BES and analyse respondents from the November–December 2016 round of the panel. For the Netherlands, we used the Election Survey Ukraine Referendum which was a three-wave part of broader LISS panel. The survey was fielded three times, twice during the campaign and once after the referendum. We have used Wave 3, the post-referendum data set, which was conducted in April 2016, with fieldwork starting the day after the referendum. It had 2502 completed

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

125

interviews. For Greece, we used the data set from the project ‘Collective Action of Indignant Citizens in Greece’ from September 2015, a few months after the referendum. This was the third wave of the survey and the only one that captured data on the referendum. The data set contains 1000 respondents of a multi-stage sample using telephone interviews. We tried to identify the questions that are most similar in the three questionnaires. Of course this was not always possible, but we did our best to give a global idea of the mechanisms at play and to compare them. Our dependent variable is the voting outcome for each individual. The variable is coded as 1 for all respondents who voted for the option that is not preferred by the EU. In the British context, this was voting Leave in the Brexit referendum; in Greece, it was voting against the austerity proposals; in the Netherlands, it was opposing the EU–Ukraine Association Agreement. All others are coded 0. In the models, we operationalised an independent variable about voter support for the incumbent government at the time of the referendum (the Conservatives in the UK; the People’s Party for Freedom and Democracy (VVD) and the Christian Democratic Appeal in the Netherlands; Syriza and the Independent Greeks (ANEL) in Greece) as 1, with 0 for supporters of other parties. Non-voters are coded as missing. Political trust is identified through similar questions in the three data sets. Confidence in the EU is operationalised for the UK through the question of satisfaction with the way democracy works in the EU. For the Netherlands, the question was about trusting the European Union. In Britain, positions on immigration played a very important role both in the campaign and in vote choice. We included a variable about the perceived effect of immigrants on domestic culture. The question asked in the British Election Study was ‘Do you think that immigration undermines or enriches Britain’s cultural life?’; respondents answered on a scale of 1–7. In the Netherlands, the question was ‘Do you think that, overall, cultural life in the Netherlands is undermined or enriched by people from other countries that have come to live here?’, with a 10-point scale, from 1 meaning undermined to 10 meaning enriched. There was no such question in Greece. However, it is less important, as the referendum was not framed in this light, and this was not a salient issue at the time. Apart from cultural side of the story, immigration is often perceived negatively by citizens because they fear immigrants will hurt their economy—‘steal our jobs’, ‘lower our wages’, etc. The UK study asked: ‘Do you think immigration is good or bad for Britain’s economy?’ with a scale

126

´ A. KATSANIDOU AND S. ŽIVKOVIC

ranging from 1 (‘Bad for the economy’) to 7 (‘Good for the economy’). In the Netherlands, the question was ‘Do you think that, overall, it is bad or good for Dutch economy that people from other countries come to live here?’ on a scale from 1 (‘Bad for the economy’) to 10 (‘Good for the economy’). For the reasons mentioned above, there was no such question in Greece. In the descriptive section above, we showed that people voting in favour of or against the EU option had different evaluations of the economy. We control for the perceived role of the economy, given that this often captures the general sentiment towards the government. In both the UK and the Netherlands, the question was: ‘Do you think that, in the past 12 months, the UK/Dutch economy improved, remained the same or deteriorated?’ For both studies, the reference category is ‘improved’. For Greece we included the perception of the future of economy, as there was no variation in the retrospective measure; almost all respondents thought that the Greek economy had worsened in the past. In the Greek case, we added two more variables that are relevant: whether the respondent believed the Greece should remain in the EU and whether the EU was actually to blame for the crisis Greece was experiencing. Finally, we included a set of demographic variables as controls, since they traditionally explain some part of the vote in elections and referendums. Education is coded as a categorical variable and age as a continuous variable in all models.

7.3

Comparative Results

Overall, each of the three models has a decent fit, with an R 2 between 0.20 and 0.44 (Table 7.1). Two things need to be taken into consideration when interpreting the results in Table 7.1. First, for ease of comparability across the three countries, we present odds ratios rather then logits. In odds ratios, values above 1.00 increase the chances of voting for the anti-EU option while values below 1.00 mean that predictor increases the chances of favouring the EU option. Second, differences in sample sizes influence statistical significance. Sizes of the three samples vary from 5030 in the UK to 797 respondents for Greece. We find very limited support for the first hypothesis claiming that socio-demographics will influence vote choice. Education was indicated in the literature as a factor that divides citizens between winners and losers of globalisation. We find indications that this conflict appeared in the British

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

Table 7.1 Logit models of anti-EU referendum vote Vote anti-EU option

UK

Odds ratios; Standard errors in parentheses Hypothesis 1 Demographics Male 0.895 (0.0745) Secondary school 0.690*** (0.0776) University 0.562*** (0.0551) Age: 35–64 1.310** (0.132) Age: 65+ 1.964*** (0.256) Hypothesis 2 Partisanship Vote party in government 1.526*** (0.132) Hypothesis 3 European Union Satisfaction EU democracy 0.153*** (0.017) EU—membership – EU—crisis – Hypothesis 4 Issues Economy stayed same Economy worse Immigrants harm economy Immigrants harm culture Constant Pseudo R 2

0.462*** (0.0619) 0.125*** (0.0166) 1.373*** (0.046) 1.428*** (0.043) 60.52*** (12.73) 0.436

NL

GRE

1.107 (0.143) 0.997 (0.287) 0.813 (0.245) 0.895 (0.186) 0.890 (0.193)

1.044 (0.188) 0.787 (0.274) 0.417* (0.147) 0.732 (0.199) 0.618 (0.207)

0.847 (0.127)

8.206*** (1.488)

0.735*** (0.0232)

– – 0.960*** (0.0337) 1.057 (0.350)

– –

1.386 (0.235) 1.947*** (0.369) 1.142 (0.059) 1.220*** (0.002) 5.276*** (2.275) 0.204

0.699 (0.208) 0.564* (0.143) – – 1.192 (0.588) 0.288

Statistical significance * p < 0.05, ** p < 0.01, *** p < 0.001 Source UK: BES, N: 5030. Netherlands: LISS panel, N: 1519; Greece Aristeia, N: 797

127

128

´ A. KATSANIDOU AND S. ŽIVKOVIC

Brexit referendum and to a lesser extent in the Greek referendum. People with secondary education were less likely to vote for Brexit, and this effect intensifies for people with university education. We found a similar effect in the case of Greece; people with university education appear to reject the anti-EU option in favour of the pro-EU camp. In the Netherlands, we find no effects of education. This most likely has to do with the importance of the referendum. In Britain and Greece, the referendum polarised people in favour or against the EU while in the Netherlands the issue was mainly focused only a single aspect of the EU. The effect of age is purely a UK characteristic. We did not find that older citizens voted differently from younger ones in either Greece or the Netherlands. When compared to the base category of younger voters (under 35 years old), both middle-aged and older Britons were more likely to vote for Brexit. Older generations remember a perceived glorious past before the EU became central to everyday politics. For the second hypothesis about party identification, identifying with the governing party is a good guide to referendum voting on Europe. In Britain, David Cameron, then the prime minister, campaigned for Remain, but many Conservative personalities such as Boris Johnson campaigned for Leave, as did then UKIP leader Nigel Farage, whose arguments appealed to many Conservatives. Thus, the views of the party leader and prime minister were disregarded by those who had voted for him a year before. In Greece, however, the Syriza and ANEL government that had negotiated the agreement also campaigned against it. The coalition of these two unlikely partners—Syriza, until recently a party of the radical left, and ANEL, a nationalist populist party—both rejected austerity measures. They led a very polarising campaign in favour of ‘No’, the anti-EU option. Party identification was by far the strongest influence on voting in the Greek referendum. It was not significant in the low-turnout Dutch referendum, called by an initiative petition rather than by the government. Satisfaction with the way democracy works in the European Union is a good reason to want to favour it in a referendum. In Britain, this had a very strong effect, and a significant effect in the Netherlands too. In the case of Greece we did not have a similar variable and used a generic variable on Greece’s EU membership. Supporting the country’s membership in the EU increases the likelihood of voting for the pro-EU option. The European Union promotes a single European market and a more open economy. That can be good for exports but it can also mean higher vulnerability to external shocks. Additionally, the EU is associated with

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

129

the influx of structural funds in the weaker economies and the potential for a wider market for the stronger economies. In times of crisis, like the latest financial and sovereign debt crisis, people may turn their backs on the EU, forcing support for it to new lows (Gomez 2015). There are two possible reasons for this. First, the economy is doing badly and citizens blame the EU for the bad times they have been experiencing. Therefore, they decide to vote in a referendum in favour of the anti-EU option. Second, citizens evaluate the economy of their country positively, being very optimistic about its economic survival even outside the Eurozone (Greece) or the European Union itself (Britain) or not fearing that their anti-EU choice might cause a possible crisis (Netherlands). Figure 7.2 shows how likely an anti-EU vote was in the case of a good or bad evaluation of the economy. In Britain, the probability of voting for the anti-EU option, in this case for Brexit, is very high (0.65) when respondents evaluate the British economy as doing better in the previous two years than it had done earlier. This probability drops dramatically when the respondents evaluate the economy as doing similarly or worse than before; the likelihood of voting for Brexit drops below 0.2. It appears that Brexit voters saw the British economy as thriving and therefore were convinced that Britain could continue to flourish outside the European Union. Based on this we can also speculate that they had taken notice of the recent crisis in the EU and considered that Britain, being economically strong, would not benefit from membership in a crisis-ridden Europe. In the Netherlands we find the exact opposite effect. When respondents evaluate the Dutch economy as better than in the years before, they are less likely to cast an anti-EU vote. The likelihood increases if the respondents evaluate the economy as worse than in the past. Voters opting for the anti-EU option feared for the Dutch economy (Fig. 7.2). Their interpretation of the acceptance of the Association Agreement was that Ukraine might join the EU as a member state and harm the economy of existing member states such as the Netherlands. This belief, in combination with a poor evaluation of the Dutch economy, might explain the anti-EU/anti-agreement vote. Finally, the case of Greece is based on evaluations of the economy in the future. In the Greek case, the likelihood differences are not as marked as in the British and Dutch cases. Believing that the economy will improve in the near future has a likelihood of voting for the anti-EU option of about 0.68. This likelihood drops as optimism about Greece’s economic future drops. Believing that the economy would improve regardless of the

´ A. KATSANIDOU AND S. ŽIVKOVIC

Probability to vote for anƟ-EU opƟon

130

United Kingdom 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Worse

Same

BeƩer

Probability to vote for anƟ-EU opƟon

Netherlands 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Probability to vote for anƟ-EU opƟon

Worse

Same

BeƩer

Greece 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Will worsen

Will stay the same

Will improve

Fig. 7.2 Evaluations of the economy and the anti-EU vote (Source BNES; LISS panel; ARISTEIA)

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

131

government’s policies could allow citizens to cast an anti-austerity ballot as they deemed further austerity unnecessary. The contrary is true for those who believed the economy would worsen. A significant point of difference between the three referendums is the impact of economic evaluations. We see a very similar pattern in Greece and in Britain. Negative evaluations of the economy (retrospectively in Britain and prospectively in Greece) clearly have an inhibiting effect for an anti-EU vote. People who negatively evaluate their economy are less likely to opt for Brexit or for the ‘No’ choice against the Troika-imposed austerity measures. In the Netherlands, in contrast, we observe the exact opposite effect. Negative evaluations of the economy increase the likelihood of voting for the anti-EU option in the Dutch referendum. The assumption we made in the descriptive section holds. Dutch voters fearing a worsening economy did not want to risk the potential opening of economic borders to a weak Ukraine, while British and Greek voters who were optimistic about their economy believed that their country could thrive outside the EU (Britain) or that their country did not need further austerity measures (Greece). We tested the impact of perceptions of immigration in the cultural and economic life of the country only in Britain and the Netherlands, as in Greece there was no such variable, and immigration was irrelevant for the referendum. As expected, respondents who positively evaluate the impact of immigrants in their country both on the economy and for the culture they bring are less likely to vote for the anti-EU option, while those who have negative opinions about immigration are more likely to make that same vote choice. These effects are strong both in Britain and in the Netherlands. The connection of anti-EU feelings and anti-immigration sentiment is long established in the literature, so this comes as no surprise for both countries. A commonality among the three referendums is that the pro- and antiEU sides did not evaluate the economy in the same way. In Britain and Greece those who chose the anti-EU option were more positive in evaluating their country’s economy, while in the Netherlands the exact opposite was true. In all three cases, the relationship between economic evaluations and vote choice was strong and significant. The British and the Greek referendums have two further things in common. For both we identified a strong influence on vote choice of support for the government party. In Britain this pushed in the opposite direction from Prime Minister Cameron’s recommendation but in Greece the majority of voters

132

´ A. KATSANIDOU AND S. ŽIVKOVIC

followed Prime Minister Tsipras’ lead. The other common characteristic is the impact of education. In both countries, we found a clear indication of the pro-integration versus national protection divide, leaving the less educated—as the losers in globalisation—voting against the EU. The British and the Dutch referendums share a common context. They both took place in countries that are concerned with immigration. In both countries, there are fears among the electorate that the economic and cultural consequences of immigration are not only present, but are also not discussed openly. These fears found their expression in the two referendums where citizens chose the anti-EU option to show their objections to immigration.

References Campbell, A., P.E. Converse, W.E. Miller, and D.E. Stokes. 1960. The American Voter. Oxford: Wiley. Gomez, R. 2015. The Economy Strikes Back: Support for the EU During the Great Recession. JCMS: Journal of Common Market Studies 53 (3): 577–592. Goodwin, M.J., and O. Heath. 2016. The 2016 Referendum, Brexit and the Left Behind: An Aggregate-Level Analysis of the Result. Political Quarterly 87 (3): 323–332. Hobolt, S.B. 2016. The Brexit Vote: A Divided Nation, a Divided Continent. Journal of European Public Policy 23 (9): 1259–1277. Lipset, S.M. 1960. Political Man. New York: Doubleday. LISS Panel Data Archive. n.d. Election Survey Ukraine Referendum Measurement 3. Retrieved December 25, 2018, from https://www.dataarchive. lissdata.nl/study_units/view/651. Otjes, S., and A. Katsanidou. 2017. Beyond Kriesiland: EU Integration as a Super Issue After the Eurocrisis. European Journal of Political Research 56 (2): 301–319. Petrov, R., and P. Van Elsuwege. 2016. What Does the Association Agreement Mean for Ukraine, the EU and Its Member States? A Legal Appraisal. SSRN Scholarly Paper No. ID 2779920, Social Science Research Network, Rochester, NY. Retrieved from https://papers.ssrn.com/abstract=2779920. Tsebelis, G. 2016. Lessons from the Greek Crisis. Journal of European Public Policy 23 (1): 25–41. van der Brug, W., T. van der Meer, and D. van der Pas. 2018. Voting in the Dutch “Ukraine-Referendum”: A Panel Study on the Dynamics of Party Preference, EU-Attitudes, and Referendum-Specific Considerations. Acta Politica 53 (4): 496–516.

7

COMPARING VOTING IN NATIONAL REFERENDUMS …

133

Walter, S., E. Dinas, I. Jurado, and N. Konstantinidis. 2018. Non-Cooperation by Popular Vote: Expectations, Foreign Intervention, and the Vote in the 2015 Greek Bailout Referendum. International Organization 72 (4): 969– 994.

CHAPTER 8

Britain: Still Searching for a Role in the World Tim Oliver

In 1962, Dean Acheson, the former US secretary of state, gave a speech at the US Military Academy at West Point in which he said: Great Britain has lost an Empire and has not yet found a role. The attempt to play a separate power role apart from Europe, a role based on a special relationship with the United States, and on being head of a commonwealth which has no political structure, or unity, or strength, this role is about played out. (Acheson 1962)

Acheson, an Anglophile, was taken aback by the negative reactions to the speech in Britain. Britain was no longer the power it had been when it had emerged as one of the Big Three victors in the Second World War. Victory had hidden the extent of Britain’s exhaustion as a global power, something that had been under pressure since well before the outbreak of war in 1939. A series of events such as the Suez crisis of 1956 and ongoing economic problems had demonstrated how overstretched Britain was. Coming to terms with this, however, was politically difficult. British decision-makers were accustomed to sitting at the top tables of international relations. The British public appeared to expect this, with some parts of British identity, demographics, and culture connecting the country extensively to a global empire. Britain still chose to uphold expensive commitments, such as large overseas garrisons, that were deemed necessary not only to British interests but also to the wider Western order. © The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_8

135

136

T. OLIVER

Britain also had obligations as a permanent member of the UN Security Council and a major member of the North Atlantic Treaty Organization (NATO). British strategic thinking since 1945 has repeatedly struggled to balance desired ends and limited means (Sanders and Houghton 2016; Daddow 2019). The struggle has involved the relative priority to be given to the country’s long-standing economic, cultural and security links with the United States and Commonwealth countries. When European institutions were established in the 1950s, they were regarded as unsuited to British membership. For decades after joining in 1973, Britain struggled to fit Europe into its older but weakening ties with the United States and the Commonwealth. Brexit has disengaged the UK from the European Union, but it has not answered the question, ‘What is Britain’s role in the world?’

8.1

Looking Back

This chapter frames its analysis of Britain’s attempts to find a role through an examination of the foreign policy agendas defined by post-war British prime ministers. We start with Winston Churchill, whose ideas were formed by experience in the empire of Queen Victoria and expanded by nurturing the Anglo-American alliance in the Second World War. They have set the framework for thinking about Britain’s role in the world up to the present. Advocates of leaving the European Union have held out a vision of a restoration of Britain’s traditional global role, while advocates of remaining have offered a vision adapted to Britain’s reduced role in a world in which bigger powers now contend globally. In identifying periods with prime ministers, we must avoid expecting too much of prime ministerial influence, which is constrained both domestically and internationally. To describe the prime minister’s role in foreign affairs as presidential (Foley 2000) overlooks the significant constraints on the power of presidents in dealing with foreign governments and transnational actors in a global system. It also overlooks the constraints on 10 Downing Street within Westminster, which have been very evident in forcing the Brexit referendum and on implementing it. At best a prime minister within UK politics can be predominant: ‘Predominance enables the Prime Minister to lead, but not command, the executive, to direct, not control, its policy development, and to manage,

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

137

but not wholly dominate, the legislature’ (Heffernan 2003: 350). In foreign policy a prime minister normally has more room for manoeuvre than in domestic matters, but is always subject to significant constraints. Rose (2001) describes the prime minister as being subject to an increasing variety of influences from new powers and international economic forces in what has becoming a shrinking world. Churchill Of all British prime ministers, it is Sir Winston Churchill who casts the greatest shadow over his successors. It has been said that British foreign policy since 1945 has been footnotes to Churchill (Garton Ash 2004: 36). Ideas about his leadership and ‘bulldog’ spirit have inspired many politicians, as have his ideas about Britain’s place in the world. They have also been confused. Churchill, like any senior statesman, was a complex politician whose outlook reflected a range of historical, ideological, emotional and strategic ideas that did not always cohere. The veteran leader Churchill identified three circles that were shaping Britain’s destiny in a speech to the 1948 Conservative party conference in Llandudno: The first circle for us is naturally the British Commonwealth and Empire, with all that comprises. Then there is also the English-speaking World in which we, Canada, and the other British Dominions and the United States play so important a part. And finally there is United Europe. These three majestic circles are coexistent and if they are linked together there is no force or combination which could overthrow them or even challenge them. Now if you think of the three inter-linked circles you will see that we are the only country which has a great part in every one of them. (Churchill 1948)

The idea that Britain is at the heart of three overlapping circles provides a framework that is relevant for assessing the country’s multiple engagements. Those using it normally do so with optimistic assumptions about Britain’s continuing influence. They overlook the fact that Dean Acheson also used this framework to emphasise Britain’s declining position in the world. The Commonwealth and empire, to which Churchill especially held a strong attachment, can form an end in themselves, but the empire was

138

T. OLIVER

also a means to an end of strengthening the British policy of holding the balance of power in Europe against rivals such as France and Germany (Simms 2016). After the fall of France in 1940, Britain was able to draw on military support from as far afield as Australia. American ties were strengthened when President Franklin D. Roosevelt introduced LendLease military aid and, after Hitler brought the United States into the war in Europe, Churchill’s Britain became the partner with the United States in the liberation of occupied Europe. The Labour government of 1945–1951, led by Clement Attlee with Ernest Bevin as foreign secretary, began the process of cutting back on Britain’s imperial role and strengthening Britain’s ties with the United States. It accepted the independence of India and Pakistan, which Churchill had always opposed. It also withdrew unilaterally from administering Palestine. It gave priority to integrating an American commitment to defend Britain against the Soviet Union by taking the lead as the European partner for a trans-Atlantic partnership. In economics, it was the leader in organising the delivery of Marshall Aid to sixteen European countries, and Britain itself received the biggest share. Withdrawal of Britain’s security presence from Greece and Turkey was complemented by the administration of Harry S. Truman taking over Britain’s role. This was followed by the United States deciding to guarantee the security of Western Europe through NATO, in which Britain was the leading European partner. When the United States sent troops to defend South Korea against invasion from North Korea in 1950, Britain sent a contingent of troops in a show of support for American-led collective security. Simultaneously, the Labour government decided to maintain Britain’s role as an independent atomic power. The relative decline of Britain from the end of the nineteenth century onwards slowly weakened London’s ability to maintain the British Empire. Under pressure from older Dominions such as Australia and from independence movements in colonies to grant them greater autonomy, the empire was transformed into a commonwealth of independent nations. As Acheson pointedly noted, it had no ‘political structure, or unity, or strength’ (1962). Nonetheless, the legacies that the Commonwealth has inherited from its imperial origins have created many common ties from which Britain has benefited. To this day, Britain is a country whose identity has been shaped by the idea of being a global nation with links across the world. The English language, common law, the Crown, trading, educational and security links, and a history of shared rule and

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

139

misrule have been reinforced by contemporary immigration (Bennett 2004). With the empire in decline after 1945, British relations with Europe gained in prominence through co-operation in mobilising both hard power via NATO to counter the Soviet Union and the soft power of cultural, linguistic and personal ties, which for Churchill reached from his American mother to wartime dealings with President Roosevelt. American politicians felt much more comfortable and secure when dealing with London rather than with France and West Germany, each weakened and discredited by its wartime history. The partnership between unequals was summed up by Harold Macmillan thus: ‘These Americans represent the new Roman empire and we Britons, like the Greeks of old, must teach them how to make it go’. In co-operating with the United States, British prime ministers hoped to maintain the country’s global role and gain influence in Europe as a broker between Washington and European countries. What started out as a British image of partnership has inevitably declined with the modern usage focusing on the idea of the UK–US relationship being ‘special’ (House of Commons Foreign Affairs Committee 2010). Nevertheless, the three underlying British goals have remained stable: to ensure the United States did not retreat into isolationism as it had in the 1930s; to keep the United States committed to the security of Europe; and to allow British decision-makers to shape wider US foreign policy at a global level. This outlook for UK decision-makers was reinforced by the Suez crisis of 1956, when Britain, France and Israel conspired to occupy the Suez Canal, which had recently been nationalised by the Egyptian government. President Dwight D. Eisenhower, leader of the Anglo-American military alliance in the Second World War, was furious at what he saw as an act of outdated imperialism. US economic pressure on Britain forced a withdrawal, a clear sign of imperial overreach. It was a painful lesson that drove France to look more towards European co-operation and Britain towards staying close to the United States. As for Europe, Churchill gave rhetorical support to a united Europe before its first institutions were formed. In a speech in Zurich in 1946, he called for a ‘United States of Europe’. When the French Foreign Secretary Robert Schuman took the lead in 1950 in institutionalising co-operation through the European Coal and Steel Community, Britain stood aside (Young 1998). This reflected not only British ideas of power and standing in the world, but also a post-war strategic outlook shaped during the

140

T. OLIVER

Second World War by extra-European alliances with the United States and Commonwealth countries. This view also found support among Labour leaders who had served in the wartime coalition government. Having nationalised the coal industry, the Labour party was not minded to share control with France and West Germany. As Deputy Prime Minister Herbert Morrison argued, ‘It’s no damn good—the Durham miners won’t wear it’ (Forster 2002: 299). In 1962 the Labour leader Hugh Gaitskell opposed joining Europe on the grounds that ‘it would mean the end of Britain as an independent nation-state’, ‘the end of a thousand years of history’ and ‘the end of the Commonwealth’. Macmillan to Heath Having become prime minister following the 1956 Suez debacle, Harold Macmillan commissioned the Study of Future Policy 1960–1970 (Cabinet Office 1959). The report was both wide-ranging, covering everything from the future of the Commonwealth to Britain’s economic outlook, and candid in its assessments about Britain’s future. It was so candid— and in retrospect accurate—in its assessment of Britain’s relative decline that Macmillan blocked discussion of it in Cabinet. It was classified and banished to the archives. Nevertheless, whether it was the Conservative government of Harold Macmillan or the Labour governments of Harold Wilson, a series of decisions, such as those on defence reviews (Cornish and Dorman 2010), dealt with Britain’s relative decline and its prioritisation of the North Atlantic and Western Europe. Only four years after the European Economic Community (EEC) was established with Britain staying aloof, Macmillan decided to pursue British membership. The success of the EEC had quickly made it the predominant organisation in West European politics and economics. The alternative European Free Trade Association (EFTA), which Britain had led the establishment of, lacked the EEC’s supranational powers, but had not lived up to British hopes. Its seven members were referred to as the ‘outer seven’, compared to the ‘inner six’ of the EEC. Anxiety about Britain’s inability to shape the EEC from the outside made membership a priority. That it was backed by the United States provided a way to justify membership as it would also enhance relations with the United States. Although Britain had recovered from the war with a growing economy, its economic growth had been erratic compared to that achieved by West Germany and other EEC countries. It was hoped that membership of the

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

141

EEC would help replicate their success and reflect Britain’s trade patterns, which had been shifting from the Commonwealth towards Europe. Joining the EEC unleashed divisions that continue to this day within the Conservative party and more widely in the country. However, the extent to which the issue excited the British people is debatable. Until the 2016 referendum, concern about UK–EU relations was generally a leading topic of concern only within the political elite. It did raise the issue of whether Britain could remain a global power and become a European power (Simms 2016). The experience of negotiating with the EEC showed that as an applicant seeking membership Britain was the weaker party; there was a structural imbalance between Britain and the combined weight of the six EEC member states. There was support for British membership, especially as it would help balance France, which under President Charles de Gaulle had been a disruptive and awkward partner. However, the need for unanimous approval gave de Gaulle the right to veto Britain’s application, which he did. His argument that the UK was not sufficiently European and was committed more to its ties with the United States was not without justification. It was also an act to protect French leadership of the EEC. A second application in 1967 by Labour Prime Minister Harold Wilson ran into the same problems. Edward Heath’s strong commitment to European integration successfully secured British membership in 1973, after overcoming opposition from within the Conservative party. Preferences for trading with Commonwealth states such as New Zealand were one objection. Another was the importance of maintaining the priority of basing Britain’s military security on the American-led NATO alliance. The House of Commons voted in 1971 by a majority of 112 in favour of membership. The votes to pass the European Communities Act 1972 were much closer, with the ballot on the third reading passing by only 17 votes (Norton 2011: 59). It was a demonstration of how important but difficult it would be to sustain parliamentary support for British membership. Hesitancy about Europe continued after Wilson returned to Downing Street. Labour’s internal divisions led Wilson to call a referendum on continued membership, the first nationwide referendum in Britain’s history, using renegotiation of terms of membership to justify campaigning to remain a part of European integration (Glencross 2016). Wilson’s renegotiation produced little substantial change, but it did help generate a level of support that led Britain to vote 67 to 33% in favour of staying in

142

T. OLIVER

the EEC (Butler and Kitzinger 1976). The referendum, however, proved only a temporary solution. By the early 1980s Labour had split, and public opinion had shifted strongly towards opposing British membership (Ipsos Mori 2016). Threatening to leave the EEC revealed Britain’s isolation. Britain’s attempts at membership had been accompanied by applications from three other EFTA members with strong links to the UK: Denmark, Ireland and Norway. Their close relations with the UK meant they had withdrawn their applications in 1963 and 1967 when that of Britain was vetoed. In 1975, however, both Denmark and Ireland made clear that if Britain did withdraw, then they would not do the same. For them the EEC now served as an important partner, one which framed their relations with Britain. Thatcher to Blair Britain’s role in the EU since membership has often been described as being ‘an awkward partner’ (George 1998). This reflects Britain’s different experience of the Second World War, late membership and problems joining, a majoritarian as opposed to consensus-based political system and close relations with the United States and Commonwealth countries on other continents. Unlike other member states, British politicians rarely embraced or advocated EU membership as an end itself. Instead, it was portrayed as a necessary means to an end of economic benefit. This was reinforced by the belief among policy-makers and the public that the UK could actively promote its special relationship with the United States and with the Commonwealth without regard to European ties. Britain’s size and links beyond Europe made this possible to an extent denied to many small EU member states for which EU participation was their international role. It ignored the resource constraints on Britain achieving its high ambitions for influence and the growth of new powers on other continents, especially Asia. The premierships of Thatcher, Major, Blair and Brown, especially Thatcher and Blair, point to how Britain’s role in the EU could be very Janus-faced: an awkward partner that could also be a constructive, if quiet, European (Daddow and Oliver 2016). Margaret Thatcher upturned Britain’s post-war domestic consensus, replacing Keynesian economics and a mixed economy with an approach that today would be termed neoliberal: monetarism, deregulation and privatisation. She took an equally determined approach to Britain’s place

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

143

in the world, not least towards trying to reverse both Britain’s relative decline and a declinist attitude amongst policy-makers. Her combative attitude put her in conflict with some of her consensus-oriented ministers. Early in her tenure she also battled with the EEC to reduce Britain’s budgetary contributions. On joining, Britain had become one of the largest contributors to the EEC budget, despite at the time being one of its less prosperous member states. This was a result of joining after the rules for the budget had been set in favour of subsidies to agricultural producers, which particularly benefited France. Thatcher’s success in securing a rebate for Britain is often pointed to as evidence that a British prime minister can succeed in negotiating with the EU. While her hard work and detailed knowledge certainly helped secure the rebate, what is overlooked is that the EEC relented only when it was in a weak position. Thatcher’s attitude to Europe is today framed in no small part by the role UK–EU relations played in her downfall in 1990 and the Euroscepticism that defined her politics after she left office. Yet the EU today resembles many of Thatcher’s preferences. Under her leadership the British government took the lead in the creation of the European Single Market that swept away many barriers to trade across Europe that had been accepted by the original founding countries of the EEC. Some in British politics on the right and on the left regard this as giving the European Commission the power to run the British economy from Brussels. The liberal market measures introduced to reduce state control made continental critics view the EU as a promoter of an Anglo-Saxon market economy. In France, this view played a part in France’s 2005 rejection of the European Constitution. Thatcher’s vision of a deregulated European free market set out in her 1988 Bruges speech, often seen as a defining text of Euroscepticism, has left a legacy affecting the EU’s political economy today (Rohac 2016). Britain’s links with the United States, however, limited the degree to which UK–EU relations could become the central role for Britain in the world. The positive personal relationship between Thatcher and President Ronald Reagan was paralleled by common ideological beliefs about the importance of markets and of military power. The ‘special relationship’ today has become one in which the levels of co-operation in nuclear weapons, intelligence and special forces between officials in both governments reflect mutual trust in ways not normally found in relations with other military allies of the United States. It took practical form in American military assistance to the UK during the Falklands War. While

144

T. OLIVER

British officials and prime ministers have continued to give priority to the United States, the United States, which is both an Atlantic- and a Pacificfacing political and economic power, has been giving increasing attention to Japan and China. John Major inherited Thatcher’s legacy of a Conservative party beset by infighting and splits over Europe. This played out notably in three areas. Major had to struggle to secure Parliament’s approval of the Maastricht Treaty, where the government had a narrow majority and he faced Eurosceptic opposition from what he described as ‘bastards’ in the Cabinet. Britain and EU regulators were at loggerheads over the ban on the export of British beef following the outbreak of a deadly infection. Speculative pressure on an overvalued pound led to pressure in international markets that caused Britain’s humiliating ejection from the European Exchange Rate Mechanism. When Tony Blair took office in 1997, he declared that his Labour government was to be one ‘that gives this country strength and confidence in leadership both at home and abroad, particularly in respect of Europe’. Blair was the most pro-European prime minister since Edward Heath. But his leadership on European matters was often criticised, with him being described not so much as a pro-European as an ‘anti-antiEuropean’ because he was willing to contest criticism from Eurosceptics (Donnelley 2005: 2). However, he did so without articulating a strong pro-European message as was desired by an increasingly marginal group of advocates of an active British role in the EU. Blair promised to call referendums on two divisive policies of the European Union: a Constitution for the European Union and Britain joining the Eurozone. The first ballot was not called because referendums in France and the Netherlands vetoed the Constitution. Britain did not adopt the euro because Chancellor Gordon Brown vetoed doing so. Relations with the United States initially flourished as Blair and President Bill Clinton bonded personally and shared a centrist, Third Way political philosophy. However, when a Republican, George W. Bush, became president, these ties were an obstacle to Blair basking in the aura of a ‘special relationship’ with the White House. In pursuit of maintaining this relationship, Blair’s chief of staff, Jonathan Powell, told Sir Christopher Meyer, the British ambassador to the United States, to ‘get up the arse of the White House and stay there’ (Meyer 2006). As proof of his commitment to a special relationship under American leadership, Blair ordered British troops to war in Afghanistan and also to support

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

145

the United States when it invaded Iraq. The latter caused the resignation of Foreign Secretary Robin Cook and subsequently led to Parliament attempting to place constraints on the war-making powers of prime ministers. Towards the end of Blair’s time in office, the United States and the EU were more unpopular among the British people than ever before (Wallace and Oliver 2005). As a long-term Chancellor of the Exchequer, Blair’s successor, Gordon Brown, saw foreign policy through the prism of international economic policy, and he looked to Harvard and MIT economists for ideas rather than to the European Union (Daddow 2011). Despite a secure majority in the House of Commons, Brown’s government faced problems approving the Lisbon Treaty. It was a harbinger of things to come. Not only did the pro-European Liberal Democrats experience opposition from some of their MPs to the party’s position of not supporting a referendum on the treaty. The Conservative opposition also voted against and its opposition amendment to put the treaty to a referendum was rejected, even though the Foreign Affairs Committee of Parliament concluded that the treaty did not materially differ from the European Constitution, which Tony Blair had promised to put to a referendum for approval.

8.2

Looking in All Directions

A major priority of David Cameron’s goal of modernising the Conservative party was to stop ‘banging on about Europe’ (BBC 2006). That his wish spectacularly failed to come true reflects not only the realities of Britain’s position in the EU but also his own miscalculations about how to manage UK–EU relations. Early signs of this were to be found in his decision as leader of the opposition to withdraw the Conservatives from the European People’s Party, the main centre-right political group in the European Parliament. It detached the Conservatives from some of their closest allies in the EU, including the German chancellor’s Christian Democratic party, and isolated them in a Eurosceptic group with little influence in that parliament. On becoming prime minister in 2010 Cameron found himself leading a three-group coalition on Europe: the Liberal Democrat pro-Europeans, liberal and pragmatic Conservatives such as Cameron himself, and the Eurosceptic right of the Tory party. Growing demands from the right of his party coupled with the electoral pressure from the rise of the United

146

T. OLIVER

Kingdom Independence Party (UKIP) helped drive him into several policy positions that shaped Britain’s EU future. The 2015 Conservative election manifesto promised a referendum on EU membership. It ignored the conclusion of a lengthy Balance of Competences Review that the powers the EU held were consistent with British policy-making powers (Oliver 2018). Daily relations with the rest of the EU saw Britain tending to sit on the sidelines when crises arose, as in Ukraine. Cameron’s attempt to prevent the adoption of an EU fiscal treaty was humiliatingly rejected. Whereas Britain had traditionally been able to play something of a balancing role between France and Germany, its function now appeared to be a diminishing and passive one (Krotz and Maher 2016). Cameron faced problems for which he was ill prepared when he unexpectedly won an absolute majority in the 2015 general election and had to deliver on his manifesto pledge to renegotiate the UK’s membership and put the results to the British people in an Remain or Leave referendum. For many other EU governments, the renegotiation was an illtimed sideshow to more important matters. The governments of Central and East European member states were annoyed when Cameron used the 2015 Eastern Partnership Summit in Riga to raise British renegotiation aims at a meeting intended to address the issue of security vis-à-vis Russia. British ministers and officials struggled to define what they wanted to change. The concessions that Cameron secured, while difficult to achieve and not entirely insubstantial, failed to convince either the British people or, more importantly, large numbers of Conservative MPs. They were quickly abandoned as an issue in the referendum campaign that soon followed. Britain’s role and place in the world were raised in the referendum debate, but were framed by three factors. First, Cameron struggled to counter a sense of hypocrisy and insincerity in the message in seeking a vote to remain because of the importance of the EU to Britain’s place in the world. He went from saying that he would prefer Britain to leave the EU unless he secured major concessions on membership to claiming that Britain’s membership was a vital national interest important to the security of Europe. Even government reports had played down the importance of the EU. The 2015 National Security Strategy and the Strategic Defence and Security Review (HM Government 2015), both published before the referendum, said next to nothing about the importance of the EU to the UK’s place in the world. With few exceptions, advocates of Remain in all parties or none faced difficulties in arguing that the EU was essential for

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

147

Britain’s place in the world after they and their predecessors had shown indifference or even outright hostility to the EU. Second, Britain’s role was framed in a transactional way based largely on economic benefits compared to economic costs. Any positive message about European co-operation being a positive political end in itself was peripheral. The economic argument from the Remain campaign about losing benefits from trade with Europe had to compete with globally focused promises of the Leave campaigns which invoked Britain’s ties with the United States and the Commonwealth. The future of relations with emerging markets such as China and Commonwealth states such as India and Australia contrasted with the less than positive future that seemed to lie ahead for the troubled European Union. Accusations that Leave campaigners were ‘little Englanders’ seeking to isolate Britain were rejected with a message that leaving would avoid Britain being held back by ‘little Europeans’. This gave Leave campaigners a more optimistic message about Britain asserting itself in the world instead of resigning itself to being just one among twenty-eight European Union member states. Third, an important component of the Leave victory was the ability of the two main Leave campaign groups to make flanking attacks on the official Remain campaign (Shipman 2016; Clarke et al. 2017). They did this by putting out different and sometimes contradictory messages about Britain’s place in the world. That allowed the Leave campaign to appeal on both protectionist and free trade grounds and as both nativist and internationalist on immigration. President Barack Obama’s intervention to support Remain was rejected by those on the right as unwarranted interference, while on the left Britain’s continued EU membership threatened to involve it in what was seen as the unacceptable neoliberal project of the Transatlantic Trade and Investment Partnership. Both Leave campaigns focused heavily on immigration. Britain had been a strong backer of EU enlargement, seeing it as a way to anchor Central and East European countries into the wider European and trans-Atlantic framework. It had also been hoped that a wider European Union would make integration more difficult. This now came back to haunt Britain, with Britain’s long-standing support for membership of such countries as Turkey used by the Leave campaigns to add to already strong feelings about the cultural, economic and social costs of immigration. This fed into a feeling of distance from London-based metropolitan elite that benefited disproportionately from Britain’s place in the world.

148

T. OLIVER

8.3

What Role for a Post-Brexit Britain?

At heart Brexit is about choices regarding Britain’s place in a world that has changed since Churchill defined it as being central in three different circles (Oliver 2018). Five different choices that are repeatedly offered are set out below as ideal-types (Kitchen and Oliver 2017). Each raises significant questions about whether it meets Britain’s need for economic trade and national security; whether Britain has the resources to meet each alternative; and how the United States, the European Union, and Commonwealth and Asian nations would respond. Switzerland with Nukes In this isolationist option, Britain would retreat from both European and global security, economic and political commitments. The focus of Britain’s defence would be entirely on the British Isles with few if any military commitments abroad. Immigration and economic policies would be hostile to open markets and free movement, favouring protectionism and therefore designed to please domestic audiences seeing threats to Britain’s security and economy from immigration and globalisation. Such a course carries the risk of being deeply unpopular with allies, lowering Britain’s soft power, and, if the promised benefits do not materialise and if the costs of going it alone rise, unpopular domestically too. Pivot Away from Europe Britain would build foreign and security relations with the world beyond Europe, especially working closely with the United States in areas such as East Asia. However, pivots are slow and difficult, as the United States itself has found, and Britain almost certainly lacks the necessary networks and defence resources to extend commitments substantially beyond Europe. Europe remains, as Churchill put it, ‘where the weather comes from’. Disengaging from Europe to pursue ambitions elsewhere in the world that do not connect back to support Britain’s European relations would put the country out on a strategic limb.

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

149

An EU–UK Special Relationship In this role Britain would seek a post-Brexit partnership with the EU more special than that with the United States or the Commonwealth. It would keep the UK as deeply connected to the EU as possible without rejoining—and rejoining the EU would be a long-term possibility. Such a role would raise difficult questions about the role of NATO and also require Britain to reverse a policy that has over the past two decades seen it shift diplomatic and military resources towards other continents. Moreover, it would be dependent on the rest of the EU reciprocating in the creation of new institutions and an agreement or treaty that defines British obligations as well as rights and benefits. The negotiations about withdrawal have shown that there is no assurance that the EU would accept what Britain wants as being in the EU’s interest too. To re-engage with the European Union in the near future would be challenging domestically, given the deep divisions the referendum vote has raised about Britain’s relationship with Europe. The most that might be hoped for would be some form of rapprochement with the EU, followed by a piecemeal rebuilding of ties where there is a clear case of mutual benefit. A Global European Balancer This option would combine the positive aims of a pivot from Europe with those of an EU–UK special relationship. In defence matters, it would entail having a wide spectrum of defence capabilities in order to engage in a variety of conflicts. In addition to trade agreements with the United States and EU, the UK would pursue such agreements with other emerging powers. It would be the most ambitious option set out in terms of scope and cost. Since these goals would invite significant overstretch, they would therefore be regarded sceptically by allies and competitors. Adrift and Lost at Sea In the absence of a strategic goal, this would continue Theresa May’s practice of having a foreign policy framed in terms of British parliamentary politics—and internal party politics especially—whether or not it would be accepted outside Westminster. Rather than a preferred strategy, it would be one that would emerge if successive British governments did not settle on a clear positive strategy. It would see Britain following a ‘muddling

150

T. OLIVER

through’ scenario in which it would attempt to cope with events through an ad hoc mix of committing to European security, pursuing a global posture including close relations with the United States, and sometimes adopting an isolationist policy. It might sometimes mimic the impulsive foreign policy language of Donald Trump and lead to confusion and doubts among allies and the British public. It would mean British foreign policy continuing along the path it has followed since the Brexit referendum, when its all-consuming nature left Britain largely absent or passive in international affairs. The Brexit debate has invariably focused more on the means of withdrawal rather than goals. Instead of realistic goals and plans, soundbites have been offered. When goals are stated, they are often unicorn visions that ignore the constraints imposed by the interests of other countries and the lack of British resources. The referendum and post-referendum debates have not defined a role for Britain in a globalising world. Instead, public opinion, Parliament and the government have been debating how or even whether to withdraw from the European Union without plans for where this will leave Britain in a post-Brexit world.

References Acheson, D. 1962. Our Atlantic Alliance: The Political and Economic Strands’. Speech at the United States Military Academy, West Point, 5 December. Quoted in D. Brinkley. 1990. Dean Acheson and the “Special Relationship”: The West Point Speech of December 1962. Historical Journal 33 (3): 599– 608. BBC. 2006. Cameron Places Emphasis on Optimism. BBC News, 1 October. http://news.bbc.co.uk/1/hi/uk_politics/5396358.stm. Bennett, J.C. 2004. The Anglosphere Challenge: Why the English-Speaking Nations Will Lead the Way in the Twenty-First Century. Lanham, MD: Rowman & Littlefield. Butler, D.E., and U. Kitzinger. 1976. The 1975 Referendum. Basingstoke: Macmillan. Office, Cabinet. 1959. Study of Future Policy 1960–1970. London: Cabinet Office. Churchill, W. 1948. Speech at a Conservative Mass Meeting, Llandudno, 9 October 1948. In Europe Unite—Speeches: 1947 and 1948 by Winston S. Churchill, ed. R.S. Churchill, 1950. London: Cassell. Clarke, H., M. Goodwin, and P. Whiteley. 2017. Brexit: Why Britain Voted to Leave the European Union. Cambridge: Cambridge University Press.

8

BRITAIN: STILL SEARCHING FOR A ROLE IN THE WORLD

151

Cornish, P., and A. Dorman. 2010. Breaking the Mould: The United Kingdom Strategic Defence Review 2010. International Affairs 86 (2): 395–410. Daddow, O. 2011. New Labour and the European Union: Blair and Brown’s Logic of History. Manchester: Manchester University Press. Daddow, O. 2019. Global Britain™: The Discursive Construction of Britain’s Post-Brexit World Role. Global Affairs 5 (1): 5–22. Daddow, O., and T. Oliver. 2016. A Not So Awkward Partner: The UK Has Been a Champion of Many Causes in the EU. LSE Brexit blog, 15 April. http://blogs.lse.ac.uk/brexit/2016/04/15/a-not-soawkward-partner-the-uk-hasbeen-a-champion-of-many-causes-in-the-eu/. Donnelley, B. 2005. The Euro and British Politics. European Policy Brief, 15, September. Foley, M. 2000. The British Presidency. Manchester: Manchester University Press. Forster, A. 2002. Euroscepticism in Contemporary British Politics: Opposition to Europe in the British Conservative and Labour Parties Since 1945. London: Routledge. Garton Ash, T. 2004. Free World: America, Europe and the Surprising Future of the West. New York: Random House. George, S. 1988. An Awkward Partner: Britain in the European Community, 3rd ed. Oxford: Oxford University Press. Glencross, A. 2016. Why the UK Voted for Brexit: David Cameron’s Great Miscalculation. Basingstoke: Palgrave. Heffernan, R. 2003. Prime Ministerial Predominance? Core Executive Politics in the UK. British Journal of Politics and International Relations 5 (3). HM Government. 2015. National Security Strategy and Strategic Defence and Security Review 2015: A Secure and Prosperous United Kingdom. CM 9161. House of Commons Foreign Affairs Committee. 2010. Global Security: UK–US Relations, HC114. London: House of Commons. Ipsos Mori. 2016. European Union Membership—Trends, 15 June. https://www. ipsos.com/ipsos-mori/en-uk/european-unionmembership-trends. Kitchen, N., and T. Oliver. 2017. Written Evidence to the House of Commons Inquiry, ‘The Indispensable Ally? US, NATO and UK Defence Relations’. House of Commons Defence Committee, 30 March. http://data.parliament. uk/writtenevidence/committeeevidence.svc/evidencedocument/defencecommittee/the-indispensable-ally-us-nato-and-uk-defence-relations/written/ 48569.pdf. Krotz, U., and R. Maher. 2016. Europe’s Crisis and the EU’s “Big Three”. West European Politics 39 (5): 1053–1072. Meyer, C. 2006. DC Confidential. London: Phoenix. Norton, P. 2011. Divided Loyalties: The European Communities Act 1972. Parliamentary History 30: 53–64.

152

T. OLIVER

Oliver, T. 2018. Understanding Brexit: A Concise Introduction. Bristol: Policy Press. Rohac, D. 2016. Towards an Imperfect Union: A Conservative Case for the European Union. Lanham, MD: Rowman & Littlefield. Rose, R. 2001. The Prime Minister in a Shrinking World. Oxford and Boston: Polity Press. Sanders, D., and D.P. Houghton. 2016. Losing an Empire, Finding a Role. Basingstoke: Palgrave. Shipman, T. 2016. All Out War: The Full Story of How Brexit Sank Britain’s Political Class. London: William Collins. Simms, B. 2016. Britain’s Europe: A Thousand Years of Conflict and Cooperation. London: Allen Lane. Wallace, W., and T. Oliver. 2005. A Bridge Too Far: The UK and the Transatlantic Relationship. In The Atlantic Alliance Under Stress, ed. D. Andrews, 152–176. Cambridge: Cambridge University Press. Young, H. 1998. This Blessed Plot: Britain and Europe from Churchill to Blair. London: Macmillan.

CHAPTER 9

Forecasting the Economic Consequences of Brexit Paul Whiteley and Harold D. Clarke

During the EU referendum campaign in June 2016, the UK Treasury published forecasts of what would happen to the economy by the year 2030 if Britain left the European Union (HM Government 2016a). They examined three different Brexit scenarios, and these were uniformly pessimistic about the outcomes. In our book on the EU referendum vote, we criticised these forecasts in some detail (Clarke et al. 2017). The main points were that uncertainty about the future, problems with the underlying methodological approach in the modelling and weaknesses in the theory make it impossible to credibly forecast the performance of the UK economy fifteen years into the future (Clarke et al. 2017: 175–180). We simply cannot know what will happen to the economy that far ahead using current econometric methodology. This means that pessimistic (or optimistic) predictions about the long-run effects of Brexit are little more than guesswork. Shortly after the publication of this long-term analysis, the Treasury produced a second report, this time on the short-term effects of the UK leaving the EU over a two-year period (HM Government 2016b). Shortterm forecasting is potentially much more accurate than long-term forecasting because the closer the forecasts are to the measured behaviour of economic variables the more likely they are to hit the targets. There is a

© The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_9

153

154

P. WHITELEY AND H. D. CLARKE

thriving community of agencies, both in the private sector and in government, which undertake these kinds of forecasts and, if policy-makers or private corporations did not find these to be useful, it is unlikely that this community would exist. In October 2018, the Office for Budget Responsibility (OBR), which is now responsible for most economic forecasting in the UK, published a discussion paper on the economic consequences of Brexit (Office for Budget Responsibility 2018). This paper updated the earlier Treasury analyses and reviewed a number of forecasts from various different sources relating to Britain’s economic performance in the post-Brexit period. These were made in the period since the referendum took place in June 2016. The purpose of this chapter paper is to review the accuracy of the Treasury short-term forecasts in the light of the performance of the economy in the two years following the referendum vote. We will also investigate the accuracy of the OBR discussion paper forecasts. In the case of the Treasury report, we can make direct comparisons between forecasts and outcomes since the projections were made up to 2017–2018. In the OBR case, we focus more on examining the accuracy of the organisation’s economic forecasts over the eight-year period since it was first set up by the Conservative–Liberal Democrat coalition government in 2010. Overall, the aim is to determine if these largely pessimistic forecasts of the postBrexit future are credible. As we will show, both the Treasury and the OBR have performed poorly in relation to short-term forecasting of the economy over time. Consequently, these findings cast considerable doubt on the assertion that Brexit will be unequivocally a bad thing for the UK economy. After discussing the effectiveness of the forecasts, we undertake a modelling exercise to see if trends in three key variables—economic growth, unemployment and inflation—were in fact changed by the Brexit vote in 2016. This is a ‘backcasting’ rather than a forecasting exercise, and the results show that voting to leave the European Union has had no discernible effect on trends in these variables since the referendum took place. Of course, it can be argued that the effects of Brexit will not be experienced until Britain actually leaves the EU, which is a valid point. On the other hand, given that economic performance is strongly influenced by expectations, a topic which is discussed more fully below, the fact that we find no Brexit effects two years after the referendum casts substantial doubt on the pessimistic forecasts.

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

155

The paper starts by reviewing the Treasury and OBR reports and evaluating some of the claims they make in the light of their subsequent performance. Then, we examine some explanations of their failure to forecast the economy before testing autoregressive-moving average models (ARIMA) of the three variables. These types of models were developed in the 1970s (see Box and Jenkins 1976) and are designed to capture temporal dynamics in variables after statistical ‘noise’ and any shocks in time series caused by specific events have been taken into account. The long-run effects of the decision to leave the European Union will not be known for many years, but at this stage we can say something about the immediate effects arising from an analysis of what has happened since the referendum. We begin by reviewing the Treasury short-term forecasts.

9.1 Treasury Forecasts of the Short-Term Impact of Brexit The Treasury forecasts of the immediate effects of Brexit were made in April 2016, just prior to the referendum vote, and were designed to examine what would happen in the two-year period following the referendum. The report did not mince words about what its authors thought were likely to be the consequences of Brexit: A vote to leave would cause an immediate and profound economic shock creating instability and uncertainty which would be compounded by the complex and interdependent negotiations that would follow. The central conclusion of the analysis is that the effect of this profound shock would be to push the UK into recession and lead to a sharp rise in unemployment. (HM Government 2016b: 5)

The methodological approach in the modelling was to estimate a vector autoregressive model (VAR) of several measures of economic uncertainty thought to be affected by Brexit and then to feed these into a more general econometric model known as NiGEM, to calculate their effects on the wider economy. The latter model was developed at the National Institute for Economic and Social Research and is widely used by governments and other forecasting agencies (https://nimodel.niesr.ac.uk). The hypothesised effects of Brexit on economic uncertainty were classified into three types:

156

P. WHITELEY AND H. D. CLARKE

• A transition effect due to the impact of the UK becoming less open to trade and investment after leaving the European Union; • An uncertainty effect due to increasing uncertainty following the referendum on decisions made by economic actors; • A financial conditions effect created by the extra volatility in financial markets arising from the Brexit decision. The transition effect was associated with the assumption that ‘Businesses would start to reduce investment spending and cut jobs in the short term, consistent with lower external demand and investment in the future. This transition effect would also lead to lower incomes, reducing household spending’ (HM Government 2016b: 6). The report drew on the findings of the earlier report on the long-term consequences of Brexit to reach this conclusion while at the same time suggesting that the effects would kick in immediately after the Brexit vote. The uncertainty effect was thought to have a rather similar impact as the transition effect. The report suggests that: ‘Businesses and households would respond to this by putting off spending decisions until the nature of new arrangements with the EU became clearer. This uncertainty effect would also lower overall demand in the economy in the immediate aftermath of a vote to leave’ (HM Government 2016b: 6). It conceded that these effects would depend on the success or otherwise of the negotiations with the EU in the event of a vote to leave, but it nonetheless thought that they would be substantial. Finally, the financial conditions effect related largely to the impact of the vote on the City of London and on financial assets more generally. It suggested that: ‘In the immediate aftermath of a vote to leave, financial markets would start to reassess the UK’s economic prospects. The UK would be viewed as a bigger risk to overseas investors, which would immediately lead to an increase in the premium for lending to UK businesses and households’ (HM Government 2016b: 7). It suggested that these effects had already become apparent in April 2016 before the referendum took place, citing a fall in the pound, a similar effect in the housing market and also in business investment as examples of these anticipated effects. The Treasury’s econometric analysis produced some specific predictions about the financial effects of a vote to leave the EU arising from models analysed over the two-year period. Two scenarios were examined, one of which was described as a ‘shock scenario’ and the second as a

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

157

‘severe shock scenario’. The first of these assumed that the Brexit vote shock would be rather similar in magnitude to the effects of the recession in 1990–1991. The ‘severe shock scenario’ was assumed to be 50% worse than the ‘shock scenario’ in its effects on the variables specified in the modelling. The specific predictions of the effects of the vote on the two scenarios appear in Table 9.1. As shown, the forecasts are uniformly negative, anticipating a reduction in GDP of 3.6% in the ‘shock scenario’ and 6.0% in the ‘severe shock scenario’. In addition, the modelling predicted significant increases in inflation and rising unemployment by just over half a million people in the first scenario and more than three-quarters of a million in the second. It also predicted substantial reductions in real wages, house prices and the sterling exchange rate. Finally, it forecast that the Public Sector Borrowing Requirement, the difference between government income and expenditure, would increase by at least £24 billion over the two-year period. The actual outcomes up to the third quarter of 2018 appear in the last column of Table 9.1, and they differ rather dramatically from the forecasts. In every case, the forecasts were wrong by significant margins. Thus, gross domestic product rose by 3.4% rather than fell; inflation was positive but it grew at a lower rate than predicted; unemployment declined by more than a quarter of a million rather than increased; average wages and house prices rose rather than fell; the sterling exchange rate declined, Table 9.1 Treasury estimates: short-term impact of Brexit, fiscal year 2016/17 to 2017/18 Shock scenario Gross domestic product Consumer price index Unemployment rate Unemployment level Average real wages House prices Sterling exchange rate index Public sector net borrowing

Severe shock scenario

Actual outcomes

−3.6% +2.3 +1.6 +520,000 −2.8% −10% −12%

−6.0% +2.7 +2.4 +820,000 −4.0% −18% −15%

+3.4% +2.0 −0.9 −280,000 +4.8% +4.9% −6.3%

+£24 billion

+£39 billion

−£7.2 billion

Source HM Treasury (2016b) and office of national statistics

158

P. WHITELEY AND H. D. CLARKE

but by much less than predicted; and finally the Public Sector Borrowing Requirement was less than a third of that predicted in the ‘shock scenario’ and lower still for the ‘severe shock scenario’. Overall, the forecasts from the Treasury modelling were seriously misleading in the light of what subsequently happened. As mentioned earlier, further estimates of the economic effects of the referendum were provided by the OBR, and we examine these next.

9.2

Office for Budget Responsibility Forecasts

The OBR was created in 2010 ‘to provide independent and authoritative analyses of the UK public finances and the impact that government policy has on them’ (Office for Budget Responsibility 2018: 1). The organisation produces two medium-term forecasts each year, one at the time of the Budget and the second during the spring/autumn statements. At the time of publication of the discussion paper in October 2018, there was no final agreement between the UK and EU on the terms of the divorce or about their relationship after Brexit. So the report stresses that these uncertainties make forecasting difficult. The paper started by pointing out that UK economic growth was at the top end of the range of G7 countries in 2015, but by the third quarter of 2018 it had fallen to the bottom end of the range. The implication was that, in part, this was due to the Brexit vote. This point was reinforced by a quote from studies by the Centre for European Reform which attempted to estimate how fast the UK might have grown if the referendum had not taken place. This exercise involved creating an ‘artificial’ or ‘doppelganger’ UK economy, based on a weighted average of the growth rates of countries that are similar in size and economic performance to the UK. On this basis, Born et al. (2017) argued that the shortfall in growth was 2.0 percentage points between the second quarters of 2016 and 2018. This approach to economic forecasting is clearly dependent on which ‘doppelganger’ countries are chosen, and the time period over which their performance is measured. The latter is particularly important given that countries are at different stages of the economic cycle at any one point of time. The OBR made a forecast in November 2016 which suggested that real growth would be reduced by 2.4 percentage points by the year 2021 (Office for Budget Responsibility 2018: 6). The October 2018 discussion paper points out that: ‘We have not updated these judgements since

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

159

November 2016, while we await the results of the negotiations’ (Office for Budget Responsibility 2018: 7). This highlights the difficulties of reaching any conclusions about the performance of the economy in the absence of a final deal between the UK and the EU. That said, it is still surprising that two additional years of data available since the original forecast have been ignored. Given what we observe in Table 9.1, if the OBR had taken these data into account, then the growth forecasts would have been significantly more positive. Figure 9.1 gives us insight into the OBR growth forecasts since it was founded in 2010. The figure is known as a ‘porcupine chart’ since it compares various forecasts of economic growth made by the OBR with the actual growth outcomes. The ‘quills’ in the porcupine chart are successive forecasts starting in June 2010 and continuing up to 2022. The heavy solid black line represents the actual outcomes, and this ends in March 2018. We can see how accurate these forecasts are by comparing the black line with the ‘quills’.

Real GDP Growth - Per Cent

4

2

0

-2

-4

-6 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 June 2010

Successive Forecasts

Outturn Data

March 2018

Fig. 9.1 Forecasts of growth in GDP in real terms, 2008–2022 (Source Office of Budget Responsibility Historical Database)

160

P. WHITELEY AND H. D. CLARKE

As is well known, forecasts are subject to statistical errors, which grow in magnitude as they move further away from their point of origin. This means that we would expect the porcupine quills to fan out more the further they are removed from the actual data. An optimal forecast would place the outcome line at the centre of this fan, thereby avoiding the twin problems of being either too optimistic or too pessimistic. However, a glance at Fig. 9.1 reveals that the growth forecasts were consistently pessimistic, undershooting actual outcome by large margins up to 2015. After that, the pattern goes into reverse with the forecasts being far more optimistic than the outcomes. If we look at the outcomes between the date of the referendum in 2016 and the last observation of growth available in 2018, then there was a slight decline in growth over this period. However, the decline was negligible compared with the period between 2008 and 2009 during the financial crisis and ‘Great Recession’. It was also smaller than the decline in growth that occurred between 2013 and 2015 during the coalition government’s austerity years just prior to the referendum vote. If we compare the forecasts with the outcomes, then it can be argued that growth slowed significantly after 2016, but this was only because the forecasts were so over-optimistic to begin with. If we focus only on the outcomes up to 2018, it would be misleading to claim that a significant decline in growth occurred after 2016 because of the results of the referendum. To be fair, this exercise consists of ‘eyeballing’ the data rather than modelling trends, but it casts considerable doubt on the argument that the referendum produced a significant reduction in economic growth. A similar pattern of poor short-term forecasts exists in the case of the OBR’s predictions of inflation, as measured by changes in the Consumer Price Index in Fig. 9.2. In this case, the forecasts consistently undershot outcomes between 2010 and 2012, and then, the opposite occurred between 2013 and 2015. Finally, after 2016, the pattern of the forecasts underestimating the rate of inflation returned. In the case of unemployment in Fig. 9.3, it is clear that the OBR’s forecasts have consistently and quite seriously overestimated trends during the entire period since 2010. These data are a direct challenge to any interpretation, which claims that the Brexit vote produced a slowdown in economic growth. Actually, a significant reduction in growth during this period would have produced rising unemployment, not the fall in unemployment observed in Fig. 9.3.

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

161

CPI Inflation - Per Cent

5

4

3

2

1

0 10

11

12

13

14

15

16

17

18

19

20

21

June 2010

Successive Forecasts

March 2018

Outturn Data

22

Fig. 9.2 Forecasts of consumer price index 2010–2022 (Source Office of Budget Responsibility Historical Database)

Similarly, the short-run relationship between unemployment and inflation, traditionally known as the Phillips curve, would have produced a fall in inflation, not the fairly rapid rise observed in Fig. 9.2 after 2016.1 The Treasury handed over much of the responsibility for short-term forecasting to the OBR when this was established in 2010. The OBR came into existence because the Chancellor of the Exchequer, George Osborne, agreed with the widespread criticism that Treasury forecasting had become over-optimistic to the point of bias as a result of political

1 The Phillips curve involves an inverse relationship between unemployment and money wages, and it implies that low unemployment produces higher inflation (Phillips 1958). The existence of this relationship has been much debated over the years but it still plays an important role in monetary policy, particularly in influencing short-run inflationary expectations. See Janet Yellen’s 2017 speech on monetary policy and inflation retrieved 29/10/2018, at https://www.federalreserve.gov/newsevents/speech/yellen20170926a. htm.

162

P. WHITELEY AND H. D. CLARKE

Unemployment Rate - Per Cent

9

8

7

6

5

4 10

11

12

13

14

15

16

17

18

19

20

June 2010

Successive Forecasts

March 2018

Outturn Data

21

22

Fig. 9.3 Forecasts of unemployment, 2010–2022 (Source Office of Budget Responsibility Historical Database)

interference.2 The above figures suggest that the OBR cannot be accused of falling prey to this particular accusation. This is because a chancellor seeking to put a positive spin on his performance in government would want optimistic forecasts involving high growth, low unemployment and low rates of inflation. Instead, the figures actually show that a series of over-optimistic forecasts tend to be followed by over-pessimistic ones, and in the case of unemployment, they have been consistently excessively pessimistic throughout the whole period. Overall, the figures show that the track record of the OBR in forecasting the key macroeconomic variables of real growth, inflation and unemployment is poor, much as in the case of the Treasury. Clearly, if political interference is not to blame for this state of affairs, there must be other factors at work, which we examine next.

2 See http://news.bbc.co.uk/1/hi/uk_politics/8685989.stm.

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

163

9.3 Long-Term Forecasting of the Economic Impact of Brexit In our book on the Brexit referendum (Clarke et al. 2017), we criticised the methodological basis of the long-term Treasury forecasts. Part of the criticism related to ‘gravity models’ of trade relationships is used in the analysis. These models assume that countries with close economic, geographical and cultural ties are more likely to trade with each other than countries which lack these relationships (HM Treasury 2016a: 158). Subsequently, considerable doubt was cast on these models in a report published by the Centre for Business Research at Cambridge University. This report suggests that the Treasury estimates are unstable and very likely exaggerate the impact of Brexit on trade relationships (Gudgin et al. 2017). A subsequent working paper produced by the same team goes further and argues that the consensus among many economists that the referendum produced a decline in economic performance is wrong (Coutts et al. 2018). The gravity models are only one aspect of the problems associated with these forecasting models. Another weakness is the confusion between uncertainty and risk. The economist Frank Knight (1921) first introduced the distinction between these concepts (see also Taleb 2007). Risk is a state of affairs in which the probability distribution of outcomes is known, but specific outcomes are not. For example, we can attach a probability to motorists of a certain age and gender with a given number of years of driving experience and other characteristics being involved in a road accident. This can be determined from historical data, which are the basis of calculating insurance premiums. However, we cannot know if a particular driver is going to have an accident on a given day. Risks of this type can be incorporated into the modelling by examining average outcomes over time. In contrast, uncertainty is very different concept from risk. In this case, there is no probability distribution of possible outcomes available to model effects. As the former governor of the Bank of England, Mervyn King, explains: At the heart of modern macroeconomics is the same illusion that uncertainty can be confined to the mathematical manipulation of known probabilities. To understand and weather booms and slumps requires a different approach to thinking about uncertainty. (King 2017: 121)

164

P. WHITELEY AND H. D. CLARKE

In other words, one of the reasons why these models fail to provide accurate forecasts is because they try to deal with uncertainty by redefining it as risk. This means that not only are forecasts of fifteen years into the future likely to be of limited value, but short-run forecasts are likely to be problematic too, particularly at times of uncertainty of the type generated by economic shocks like the Great Recession. This massively consequential event was foreseen by very few economists, so it is clearly inappropriate to think of it in terms of run-of-the-mill risk. A third problem is even more fundamental and calls into question the validity of neoclassical theory, which underpins the econometric modelling. The unorthodox economist Hyman Minsky (1982, 1986), writing well before the financial crisis, was highly critical of this theory. As his biographer points out, Minsky explains the failure of orthodox theories in the following terms: The neoclassical approach that provides the foundation for mainstream macroeconomics is applicable only to an imaginary world, an economy focused on market exchange based on a barter paradigm. Money and finance are added to the model as an afterthought – they really do not matter. Because an invisible hand supposedly guides rational individuals who have perfect foresight towards an equilibrium in which all resources are efficiently allocated, there is little role for government to play. The current crisis has shown this approach to be irrelevant for the analysis of the economy in which we live. (Wray 2016: 60)

Minsky’s argument is that economies are unstable because periods of prosperity create the conditions for subsequent recessions. During an upswing in the business cycle, firms and banks become more and more optimistic and so take on riskier financial commitments. They commit larger portions of their expected revenue to debt service and at the same time, as lenders, they accept smaller downpayments and poorer quality collateral for their loans. These developments increase financial fragility, and when, inevitably, returns turn out to be lower than expected and interest rates rise, they react by curtailing loans. Reacting to this, debtors cut back on spending and in some cases are forced into a fire sale of assets in order to meet their obligations. This produces a financial crisis, which then can trigger a recession in the wider economy. The Great Recession started in 2008 after the financial crash, and it was made much worse by years of deregulation and the creation of

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

165

very risky financial assets such as Credit Default Swaps (CDSs) and Collateralised Debt Obligations (CDOs) by banks and other financial institutions (Tett 2009). The fact that many of these were held offshore meant they were subject to no effective regulation. This meant that when the crash came financial institutions had little in the way of adequate reserves to stabilise the situation. Governments had to step in and save them from collapse to prevent the entire financial system from a catastrophic meltdown. The dominant view in the economics profession is that the economy is fundamentally stable and self-correcting and so will be resistant to shocks of the kind which produced the Great Recession, a view which is grounded in ‘general equilibrium’ theory. This theory has its origins in the nineteenth century, and it underpins the NiGEM model and pervades contemporary economic reasoning in the form of a belief in the ‘invisible hand’, an idea which goes back to Adam Smith. It proposes that the economy is fundamentally stable because there is a unique equilibrium which maximises the welfare of all participants and that rational actors will choose (Arrow and Debreu 1954). However, theoretical work shows how stringent and unrealistic are the assumptions required to make general equilibrium work in practice. They include a complete absence of uncertainty of the type we have examined, and also an assumption that actors should have a full knowledge of states of the world when making decisions. Ironically, this formalisation of Adam Smith’s ‘invisible hand’ ended up demonstrating how impossible it is to achieve in practice. As Offer and Söderberg put it: ‘The invisible hand is magical thinking’ (2016: 15). Despite this, the belief in the hidden stability of the economy pervades economic modelling, and this is one of the sources of the failure of forecasting we have examined.

9.4

Did the Brexit Vote Influence the Economy?

Clearly, the evidence suggests that the record of forecasting the consequences of Brexit in both the short run and the long run is poor. So what can be done to say something about the likely consequences of Brexit for the future of the economy? Essentially economists have to be much more modest about what can be done and not overreach themselves in making judgements about future events. One way of approaching this is to do ‘backcasting’ rather than forecasting, i.e. to model the relationship

166

P. WHITELEY AND H. D. CLARKE

between events such as the Brexit vote and trends in the economy in the past to see if they have had any influence on outcomes. This exercise can be undertaken with the use of ARIMA models of key economic variables. These are univariate models, which extract information in a time series and separate it from random noise in order to model how it behaves over time. This technique is used for short-term forecasting without requiring any theoretically defined relationships to exist between different time series, since it focuses purely on extracting the information in one series to identify trends. We apply it to the task of identifying if growth in real terms, inflation and unemployment were affected by the Brexit vote in the short run. We use quarterly data from the start of 1990 to the third quarter of 2018 to investigate the trends in these variables, which gives 114 observations altogether. The data show that the growth of gross domestic product in real terms, inflation and changes in the rate of unemployment were all stationary during this period.3 This relatively long time series makes it possible to estimate the relationships between successive observations in an efficient way. It is evident that the Great Recession which started in 2008 had a significant impact on growth during this period, and so the effects of this are controlled in the modelling by including a dummy variable which scores 1 from the third quarter of 2007 to the second quarter of 2009, and 0 otherwise. The Brexit effect is identified with a variable, which scores 1 after the second quarter of 2016, and is 0 otherwise, taking in a total of eight quarters altogether. This coincides exactly with the time horizon of two years for the Treasury short-term forecasts discussed earlier. Essentially, the modelling exercise identifies if the Great Recession and/or the referendum vote changed the way each of the series behaved over time. Table 9.2 contains the results of the ARIMA modelling exercise applied to real growth, inflation and unemployment. In the case of real growth, the analysis shows that growth in the previous quarter strongly predicts the current rate, but the relationship is not one to one. Rather a weighting factor of about 0.50 applied to the previous growth rate efficiently predicts the current rate. Not surprisingly, the Great Recession variable had a negative impact on growth trends which fell by more than three-quarters

3 Stationary means that they fluctuated around a constant mean and did not increase or decrease over time. This is a requirement for ARIMA modelling.

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

167

Table 9.2 Effects of Brexit on growth, inflation and the change in unemployment Predictors ARIMA model Lagged dependent variable Moving average shock variable Great recession Brexit Wald Chi-Square AIC

Growth

Inflation

Change in unemployment

0.49*** −

0.97*** −0.84***

0.83*** −0.56***

−0.78*** −0.07 91.2*** 156.25

0.02 0.21 2006.2*** 236.4

0.28** 0.08 515.3*** 36.9

Source Federal Reserve Bank of St. Louis economic database (*** p < 0.01; ** p < 0.05)

during this period. In contrast, the Brexit variable had no statistically significant impact on growth, which continued on trend. The Wald statistic shows that the model was a highly significant predictor of growth during this period and the Akaike Information Criterion (AIC) shows that the Great Recession variable made a significant contribution to improving the goodness of fit of the model. In contrast, the Brexit variable had no effect on the trends in economic growth in real terms. The inflation variable in Table 9.2 had a slightly different specification from the growth model. In this case, inflation in the previous period had a stronger impact on current inflation than was true of growth. One way of interpreting this is to say that inflation was subject to greater inertia than economic growth. In addition, efficient estimation of inflationary trends required that successive error terms or random shocks to the system should also be modelled. This is referred to as specifying a moving average error term in the model, and the coefficient is statistically significant in this case. This means that errors from the previous period should be modelled since they help to forecast subsequent dynamics, so this variable controls for additional ‘noise’ in the series. It is noteworthy that neither the Great Recession variable nor the Brexit variable had a significant impact on inflationary trends. After briefly spiking at the start of the 1990s, inflation fell to relatively low levels by historical standards during this period, so that the deflationary effects of the recession did not affect it. The model of change in unemployment in the final column of Table 9.2 again included autoregressive component that is the product of inertia over time, and also a moving average component associated

168

P. WHITELEY AND H. D. CLARKE

with the error term. In this analysis, the Great Recession had a significant impact on trends with unemployment rising by nearly 30% according to the estimates. Once again, however, the Brexit variable had no impact on trends in the rate of unemployment. Thus, this modelling exercise simply fails to find any evidence of a post-referendum effect on the economy up to 2018.

9.5

Conclusions

Our analysis of public perceptions of the Brexit referendum campaign in 2016 showed that a third of the public thought that the Leave campaign was negative whereas almost four out of ten thought this about the Remain campaign (Clarke et al. 2017: 214). The Remain campaign was described as ‘Project Fear’ by its critics who supported Brexit. This was given some credence by George Osborne, the Chancellor of the Exchequer, who argued in a 15 June referendum campaign speech: ‘Quitting the EU would mean less money, billions less. It’s a lose–lose situation for British families and we shouldn’t risk it’. At the time, the official shortterm and long-term forecasts of the consequences of Brexit from the Treasury certainly supported this assertion. But in the event, as the analyses in this paper show, the forecasts were wrong, and not merely slightly wrong but seriously wrong to the point of being misleading. When the UK joined the European Union in 1973, the hopes expressed at the time that this would give a boost to economic growth and prosperity did not actually materialise since membership had no impact on growth (Clarke et al. 2017: 175–190). The evidence in the present paper suggests that there is a real possibility that leaving the EU will have no appreciable direct effect in the long run on UK economic growth, which is primarily driven by other factors. If so, it is also likely that Brexit will have little influence on inflation or unemployment. These conclusions can be confirmed only by backcasting several years after decisions are taken about Britain’s departure from the EU. Nonetheless, the overall results of the analyses suggest that the forecast negative impact on the economy of the UK leaving the European Union was much exaggerated.

9

FORECASTING THE ECONOMIC CONSEQUENCES OF BREXIT

169

References Arrow, K., and G. Debreu. 1954. Existence of an Equilibrium for a Competitive Economy. Econometrica 22: 265–290. Born, B., G.J. Mueller, M. Schularick, and P. Sedlacek. 2017. The Economic Consequences of the Brexit Vote, Centre for Macroeconomics Discussion Paper No. 1738, November. Box, George E.P., and Gwilym M. Jenkins. 1976. Time Series Analysis: Forecasting and Control. San Francisco: Holden-Day. Clarke, H.D., M. Goodwin, and P. Whiteley. 2017. Brexit: Why Britain Voted to Leave the European Union. Cambridge: Cambridge University Press. Coutts, K., G. Gudgin, and J. Buchanan. 2018. How the Economics Profession Got It Wrong on Brexit. Working Paper 493, Cambridge Centre for Business Research. Gudgin, G., K. Coutts, and N. Gibson. 2017. The Macro-Economic Impact of Brexit Using the CBR Macro-Economic Model of the UK Economy. Working Paper 483, Cambridge Centre for Business Research. HM Government. 2016a. HM Treasury Analysis: The Long-Term Economic Impact of EU Membership and the Alternatives. CMND 9250, April, London, https://www.gov.uk/government/publications. HM Government. 2016b. HM Treasury Analysis: The Immediate Economic Impact of Leaving the EU. CMND 9292, May, London, https://www.gov. uk/government/publications. King, M. 2017. The End of Alchemy: Money, Banking and the Future of the Global Economy. London: Abacus. Knight, F.H. 1921. Risk, Uncertainty and Profits. Boston, MA: Hart, Schaffner & Marx and Houghton Mifflin Company. Minsky, H.P. 1982. Can ‘It’ Happen Again?. Armonk, NY: M. E. Sharpe. Minsky, H.P. 1986. Stabilizing an Unstable Economy. New Haven, CT: Yale University Press. Offer, A., and G. Söderberg. 2016. The Nobel Factor: The Prize in Economics, Social Democracy and the Market Turn. Princeton, NJ: Princeton University Press. Office for Budget Responsibility. 2018. Brexit and the OBR’s Forecasts. Discussion Paper No. 3. Phillips, A.W. 1958. The Relation Between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom. Economica 25 (100) (November): 283–299.

170

P. WHITELEY AND H. D. CLARKE

Taleb, N. 2007. The Black Swan: The Impact of the Highly Improbable. London: Penguin. Tett, G. 2009. Fool’s Gold: How Unrestrained Greed Corrupted a Dream, Shattered Global Market and Unleashed a Catastrophe. London: Little Brown. Wray, L.R. 2016. Why Minsky Matters: An Introduction to the Work of a Maverick Economist. Princeton, NJ: Princeton University Press.

CHAPTER 10

Policy-Making in a Bounded Democracy Richard Rose

The British doctrine of parliamentary sovereignty is set in a closed political system. A general election gives the government of the day the power to take whatever decision it wants as long as it is supported by the majority party in Parliament. The rhetoric of prime ministers from Churchill to Blair saw foreign policy as a process in which the British government could get what it wanted by drawing on the country’s traditional standing and superior diplomatic skills to ‘punch above its weight’ in dealing with other countries. David Cameron, Theresa May and Boris Johnson each promised the British people that the EU would give a British prime minister whatever he or she decided was in Britain’s national political interest. The referendum majority for leaving the EU to take back control of public policy reinforced the belief in Britain’s unbounded sovereignty. British politicians have ignored the bounds that interdependence places on a democratic political system. Withdrawing EU membership is, in Eurospeak terms, a subsidiary right that any member state can decide on its own. However, the European Union is also free to decide on its own how to deal with the consequences of Brexit for its future. While the Brexit referendum was an exercise in national democracy, negotiating withdrawal is an exercise in bounded democracy. Any settlement must be within red lines set in Brussels as well as within lines drawn in Westminster. Confronted with an unprecedented and unexpected referendum result to leave the EU, the Conservative government accepted that a majority of © The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_10

171

172

R. ROSE

voters had the right to overrule the majority of MPs, who had favoured remaining in the EU. While the referendum vote was clear in principle, it left open the direction in which the country should go. Prime Minister Theresa May’s sound bite—‘Brexit means Brexit’—likewise left open how Brexit could and should be implemented. Consistent with the idea that Parliament was sovereign, lots of ideas were put forward in Westminster about what Brexit should mean. However, the proposals took no account of the need to get the EU’s agreement to the terms of withdrawal. By ignoring the bounds set by interdependence, the British government has been pursuing a foreign policy without foreigners.

10.1

A Vote for Change

The single sentence on the ballot for the 2016 referendum offered voters the opportunity to make a clear and unqualified choice between principles: ‘Should the United Kingdom remain a member of the European Union or leave the European Union?’ The referendum question avoided setting out any details about what would be involved in remaining, and it did not refer to what would be involved in withdrawing from the EU. Given a choice between maintaining the status quo and a change in principle, Nobel Prize economists postulate that most people prefer the status quo because, however rosy the scenario for change appears, there are inevitably uncertainties and risks in abandoning the known for the unknown (Kahnemann and Tversky 1979). In referendums on EU membership, the status quo option is to vote against joining, as the Norwegians and the Swiss have done. However, the EU has grown to more than two dozen member states because in most cases a majority of referendum voters have accepted bounds on their country’s sovereignty by joining the European Union. The dynamics of politics blurs the distinction between the status quo and change. In 1975, Britons voted to confirm the decision to join what was then the European Economic Community (EEC). Leaders in both parties treated EU membership as having the passive consent of an indifferent public. Those MPs and think tanks that did talk about it talked to each other. Within two decades, the undynamic EEC that Britain had joined was part of the past. The 1992 Maastricht Treaty gave the institutions of the renamed European Union the dynamic authority to move towards an ever closer union. This made problematic the very idea of a EU status quo.

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

173

British Eurosceptics saw the advance in European integration as undermining British sovereignty and began mobilising opposition to the EU two decades before the Brexit referendum. The United Kingdom Independence Party (UKIP) fought its first parliamentary election in 1997 and by 2001 was nominating candidates in a big majority of constituencies. After Nigel Farage was elected one of three UKIP members of the European Parliament in 1999, he led the party as it progressed from being fourth in the number of British MEPs to being the biggest British party in the European Parliament in 2014 and 2019. Voters See the World as Risky The 2016 referendum asked British voters to choose between two sets of risks: remaining in the EU as it would become in the decades ahead or learning what the UK would become after finding a post-Brexit role in a world different from that envisioned by Winston Churchill and Tony Blair. Insofar as voters see today’s world as full of political risks, the critical question is: Which institutions are best able to deal with the risks? National governments are responsible for responding to all kinds of challenges. Insofar as the risks are bounded nationally, such as road accidents or violent crime, it is logical to rely on the national government to act. Interdependence creates challenges that go beyond a nation’s boundaries. Britain began participating in such multi-national institutions as the United Nations and NATO long before joining the EU. Belonging to a multiplicity of multi-national institutions makes it possible to turn to different institutions for different problems, for example, NATO for military defence and the EU for trade. To assess how the British public view risks in today’s world, I commissioned a national survey asking about the risks facing Britain and which intergovernmental institution could best help deal with these risks. Many Britons realise that there is no drawbridge that can be pulled up to isolate the country from problems arising beyond the country’s shores. Terrorism was seen as a big problem by an absolute majority of respondents, and immigration was seen as important by 47%. By contrast, only 22% thought problems in the global economy a big risk. Fear of military threats was much lower than fear of terrorism. Only one in eight saw this as a big threat (Table 10.1).

174

R. ROSE

Table 10.1 EU not most helpful ally for big problems Q1 Big problem

Terrorism Immigration Global economy Military threats

Q2 Who can help? %

UN %

EU %

USA %

Look after selves %

55 47 22 12

23 6 8 26

11 19 18 8

14 1 6 20

34 59 46 26

Q1 How much risk does Britain face from the global economy, military threats, immigration and terrorism? Big, fair, not much, none, don’t know? Q2 Who can best help Britain deal with this problem—the UN, the EU, the United States or do we need to look after ourselves? Source BMG Research online survey of a stratified sample of 1517 respondents, 17–23 February 2016. Don’t knows not reported above; on average 19% of respondents

To deal with a problem, a government has three broad choices: to be self-reliant; to work with allies on an ad hoc basis; or to join a multinational institution. To determine which, if any, intergovernmental institutions Britons believed would be best to work with, respondents were offered four alternatives: the United Nations, the European Union, the United States or Britain needs to look after itself. The most frequently endorsed way to deal with interdependent problems, favoured by an average of 41% of respondents, was that Britain should look after itself. In addition, an average of almost one-fifth did not know if any institution could help Britain deal with problems. These replies appear less an expression of bulldog English nationalism than of uncertainty about whether the United Nations, the American government or the European Union would offer prompt and effective help that was in Britain’s interest. Consistently, less than one-fifth thought the European Union best suited to help Britain deal with major problems. While 19% saw the EU as a source of help in dealing with immigration, the Leave position—Britain must look after immigration on its own—was endorsed by 59%. Similarly, while 18% thought the EU could help deal with problems of the global economy, a key argument of the Remain campaigners, 46% thought that Britain should look after its international economic interests alone. Military threats and terrorism are the risks that Britons see as least suited to be handled exclusively by the national government. However, instead

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

175

of turning to the EU for help, the United Nations is the most favoured institution for help with these problems, followed by the United States. A Campaign of Fear and Hope Because both major parties were internally divided on EU policy, the EU referendum campaign could not be run like a general election campaign. The Electoral Commission recognised an official group to campaign on each side with public funding, and additional groups were registered on both sides. The Leave side was divided by party personalities, with Boris Johnson speaking for the official group and Nigel Farage promoting Brexit on behalf of a second pro-Brexit group. The Remain groups emphasised the negative consequences of going it alone while pro-Brexit groups offered a hopeful vision of what Britain could become free of constraints from Brussels. The Remain campaign sought to make voters fear the economic cost of leaving the European Union, since the EU has been Britain’s biggest trading partner. Forecasts came from independent sources such as the Bank of England as well as from campaign groups. Assumptions drawn from economic theories and data about past economic relations with the EU were fed into sophisticated econometric models to produce numerical estimates of the damage of departure. Forecasts showed a reduced rate of future economic growth rather than an absolute contraction in the economy. Headline figures were published down to a tenth of a percentage point without sensitivity to the margin of error inherent in economic forecasts (see Chapter 9). When multiplied to represent a four-person household and increased again to cover a five- or ten-year period, the forecasts showed that leaving the EU threatened British households with the loss of thousands of pounds in extra income. Concentration on the economy reflected the instrumental approach of the campaigners. However, it did not play on the major anxieties of voters; barely one in five Britons saw the global economy as posing a major threat to Britain. The Leave campaigns stressed the hope that departing from the EU would enable the British government to take back control of immigration, which was much more often seen as a big problem than the economy. Even though the Conservative government, and Theresa May as home secretary, had conspicuously failed to bring down net immigration since 2010, the promise

176

R. ROSE

of curbing immigration had a big impact on referendum voters of all parties (see Chapter 5). To counter pocketbook anxieties raised by Remain forecasts, the Leave side offered an economic vision of Britain, free of EU regulations, once again becoming a powerful global force. Few details were offered about how this could be done in a world that had changed greatly since the UK was the world’s biggest trading nation. The Leave side also claimed that ending Britain’s financial contribution to the EU would offer an additional £350 million a week to spend on the National Health Service, a claim that Boris Johnson subsequently described as an understatement. Government statisticians described the estimate as an overstatement; after taking into account payments from the EU, the net saving from not paying EU dues was calculated as £161 million weekly. The two sides were not debating with each other about alternatives for achieving a common goal. Instead, they were talking past each other, using arguments that were intended to mobilise their own supporters without engaging with what their opponents were saying (Shipman 2016; Farrell and Goldsmith 2017). Remain campaigners were scornful of what were described as the Leave campaign’s unicorn vision of how Britain would be a global power by getting rid of the bounds imposed by EU membership. Leave campaigners derided the fears of the Remain campaign as showing a lack of faith that Britain could regain its past sovereign greatness. Each side branded the forecasts of the other side as false facts (https://fullfact.org/europe/false-claims-forecasts-eu-referendums). This was a category error. Strictly speaking, forecasts and promises about the future are not the same as falsifiable statements about the historic past and present. They are speculative statements that may appear more or less credible, depending on the reasoning and rhetorical force behind them and the predisposition of voters to find them agreeable. By definition, the truth of forecasts made during a referendum campaign can be known only after the event. To ascertain how voters evaluated the campaign, the Electoral Commission sponsored a post-election survey (BMG 2016). When asked about their knowledge, 79% felt they had enough information to cast their vote on the principle of Brexit. Given that the EU represented the status quo, 65% felt they had at least a fair amount of knowledge of the EU. When asked how they voted, 48% said they had voted to remain, while only 35% reported voting to leave and 17% preferred not to say

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

177

what they did. The unwillingness of up to one-third of Leave voters to say how they voted suggests that many may have felt social pressures to vote Remain but were unwilling to do so. A Politically Binding Result The dichotomous nature of the referendum ballot produced a clear division of the vote. The 51.9% majority for leaving the EU was different in kind from the plurality endorsement required to win control of government at a general election. It was also more than four times the UKIP vote at the 2015 general election, showing that supporters of Leave were not hard-core populists but represented a much broader cross-section of society. The vote to leave the EU was much larger than any British governing party had achieved in more than half a century. Likewise, the 16.1 million votes cast for remaining in the EU was more than enough to win a parliamentary election, but in a referendum it meant defeat (Uberoi 2016). The turnout of 72.2% added to the authority of the Leave majority. It was more than double the UK turnout at the 2014 European Parliament election and six percentage points higher than the turnout at the 2015 parliamentary election. A BMG survey (2016) found that the main reasons given for voting were a sense of civic duty and wanting to have a say on the issue, while most of the people who didn’t vote said it was because of personal circumstances. Only 4% didn’t vote because the option they favoured was not on the dichotomous-choice ballot. Because of divisions within both major parties, the referendum was a vote about an issue. Both the Leave and the Remain sides were coalitions formed by supporters of competing parties or of no party (see Chapter 5). Thus, the vote reflected divisions within parties more than between them. The split in the Conservative party in Parliament was paralleled at the grass roots. Among those who had voted to return a Conservative government at the 2015 general election, 58% rejected the recommendation of David Cameron and voted to leave the European Union. The Labour party was divided about the leadership of Jeremy Corbyn; in June 2016, a big majority of Labour MPs voted no confidence in his leadership. Corbyn sought to straddle both sides. Through clenched teeth, he paid lip service to the party’s official policy of remaining in the EU while simultaneously emphasising that he would not allow marketoriented EU policies to get in the way of creating a socialist economy in Britain. Labour voters were divided. The 63% who voted for remaining

178

R. ROSE

tended to be in London and big cities, while in many safe Labour seats in the North of England there were big majorities for leaving the EU. The referendum result showed that the United Kingdom was disunited about membership in the European Union. Majorities in favour of leaving were returned in England (53.4%) and Wales (52.5%) while proEU majorities were returned in Northern Ireland (55.8%) and Scotland (62.0%). England itself was divided. Although London is the capital of the UK, it is least English in the cultural diversity of its population. London’s 60% vote to remain in the EU was closer to that of Scotland than to the clear majority for leaving that piled up elsewhere in England. Since the United Kingdom is a unitary state, England’s vote was sufficient to take the UK out of Europe. The Act of Parliament authorising the referendum had no clause making the result binding, but politically the result was treated as such. Having campaigned for the UK to remain in the EU, David Cameron resigned as prime minister within hours of his position being rejected. A majority of MPs who voted to remain because they believed it was in Britain’s best interest found themselves on the losing side. Instead of adopting the view of Edmund Burke that representatives should follow their convictions, most have given their consent to the referendum decision to leave the EU. In the general election that followed in June 2017, both the Conservative and Labour party manifestos gave a commitment to deliver Brexit.

10.2

No Agreement About Brexit in Practice

There was no conflict between the British government and the European Union about the right of the UK to withdraw from the European Union. By contrast with national constitutions, Article 50 of the Treaty on European Union explicitly recognises the right of any member state to withdraw. It also specifies procedures controlling the process of withdrawal. While giving notice of withdrawal was easy, negotiating the conditions of withdrawal is not. The unexpected majority for leaving the European Union caught the British government unprepared. Cameron and his associates had dismissed the anti-EU victors as ‘fruit cakes, closet racists and mad, swiveleyed loons’ (Walters 2013; Kirkup 2019). Since the government expected the referendum to endorse remaining in the EU, there was no need to prepare plans for change. Whitehall officials accustomed to integrating

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

179

the EU dimension into British policies were unprepared for the lengthy and complicated task of reviewing and removing references to the EU from hundreds of laws and regulations in which they were embedded. The shock result demoralised the highly educated and London-based political class, which had patronised anti-EU campaigners as ‘Not one of us’. The losers were divided about what to do. Some apathetically gave grudging consent to the verdict of a majority of voters; some urged the Conservative government to emulate Norway by maintaining close economic and political links with the EU as a non-member state; some wanted a ‘think again’ referendum so that voters could reverse their mistake in voting to leave. Unlike a party seeking office to implement manifesto policies, proBrexit groups were formed for the sole purpose of campaigning to win a majority vote to leave the EU. They believed that, once the referendum was won, withdrawal could be carried out easily. After the referendum, the ad hoc Leave groups broke up and individuals went to positions in Parliament, the media, public relations and Eurosceptic think tanks. No plans were left behind for dealing with the political, economic and legal issues that had to be resolved in order to implement their victory. MPs who had pushed for years to get Britain out of the EU had little experience in government, and those holding Conservative government posts were unable to prepare their department for a Brexit victory against the prime minister’s policy. Given that the object of the Brexit campaign was to cut Brussels’ bounds on Britain, no attention was given to the prospect of having to negotiate the terms of leaving with the agreement of the EU. Conflicting Views About What Brexit Ought to Mean Competition for the job of becoming the Conservative leader responsible for implementing Brexit revealed conflicts of policy, personality and calculation. The electoral process gave Conservative MPs the opportunity to identify the two candidates whom they most favoured. Their two names were then voted on by the party membership, who by choosing a party leader also chose the prime minister. Because of his prominent contribution to the Leave campaign, Boris Johnson was a favourite among MPs who had supported Brexit. However, his belated conversion to the Brexit cause raised doubts about whether he could be relied on to take Britain out of the EU. His decision to take the anti-EU side in the referendum came only after considering

180

R. ROSE

David Cameron’s offer to give him a top Cabinet job that would place him in a good position to become prime minister whenever Cameron retired. After wavering, Johnson decided to campaign for Brexit. At the last minute, Michael Gove, a long-standing Eurosceptic, filed nomination papers for the leadership to offer an alternative to MPs who distrusted Johnson. At a press conference called to launch his candidacy for the party leadership, Johnson announced he was withdrawing because he could not unify the party. The home secretary, Theresa May, was the favoured candidate of MPs who had voted to remain in the EU. As a Cabinet minister, she had followed the lead of David Cameron and voted to remain. Because she had never been a prominent supporter of the EU, May was ready to accept that losers should consent to the decision of a referendum majority. To publicise her change of sides, May committed herself to implement the referendum result with the vague slogan ‘Brexit means Brexit’. The phrase avoided giving offence to Conservatives with competing beliefs about what form Brexit should take. MPs could read whatever meaning they wanted into her gnomic slogan. In the first-round ballot of MPs, May won just over half the vote, and in the second round, her share of the vote increased. This prompted the runner-up, Andrea Leadsom, to concede her victory by withdrawing without the need for a vote by party members. On entering Downing Street, May became the leader of a party in which Conservative MPs divided into three main groups about how to implement the referendum vote to leave the EU. Hard-line Brexiters were the most vocal and best organised, although not a majority of the party’s MPs. They gave absolute unqualified priority to withdrawal from the European Union. The problems arising from withdrawal were ignored or dismissed as inconsistent with their vision of Britain becoming an unbounded sovereign state. The median group of Conservative MPs were ‘soft’ Brexiters who had cast a referendum vote to remain, but accepted the binding authority of the referendum decision. Many wanted to keep major economic benefits of membership and were prepared to accept the EU’s conditions for doing so. Hard Brexiters labelled this BRINO (Brexit in name only). The third and smallest group consisted of Conservative MPs who regarded leaving the EU as a mistake and tried to find a way to reverse the decision, if necessary by co-operating with opposition MPs.

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

181

Theresa May filled her Cabinet with ministers who had opposing views about Brexit. She believed that the doctrine of joint Cabinet responsibility would silence criticism when the time came to put down in black and white what Brexit really meant. Ex officio responsibility for preparing Brexit policies was given to Boris Johnson as foreign secretary; David Davis as secretary of state for a newly created Department for Exiting the European Union (DEXEU); and Liam Fox as secretary for international trade. Brexit ministers maintained their campaign enthusiasm for making speeches about the benefits of Brexit without grappling with the nuts-and-bolts problems of implementing the referendum decision. Foreign Secretary Johnson declared that his policy for relations with Europe was ‘having our cake and eating it’. His response to being told by senior officials about the realities of Brexit was to cover his ears and hum the national anthem. During the referendum campaign, Davis had shown ignorance of EU policy by advocating trade deals with each EU member state, a practice explicitly forbidden by EU treaties. On being appointed international trade secretary, Fox blithely declared that maintaining frictionless trade with the EU would be ‘one of the easiest deals in history’ (Shipman 2017). The political style of Theresa May was an obstacle to gaining allies among ministers and MPs. She followed her Home Office practice of limiting discussion to a small number of trusted advisers and was cautious in revealing her preferences. She took advantage of the need for centralising communication between Brussels and Her Majesty’s Government by creating a Downing Street team to deal with negotiations. This compensated for the greater priority that pro-Brexit ministers and MPs were giving to mobilising support within the Conservative party for their untested claims of a quick withdrawal with lots of cake to enjoy on the day and thereafter (Seldon 2019). In response to internal party criticism of her unclear strategy, in January 2017 Theresa May committed herself to the hard Brexit policy of ‘a clean break’ with the European Union. Her red lines were that Britain would no longer participate in a customs union or the Single Europe Market; be free of the jurisdiction of the European Court of Justice; and take back complete control of immigration. If the EU did not accept these conditions, then Britain would withdraw unilaterally: ‘No deal for Britain is better than a bad deal for Britain’. This was a signal to hard-line Brexiters that she would not compromise on her red-line conditions. She gave this

182

R. ROSE

pledge speaking at Lancaster House under a portrait of George III, the monarch who presided over the loss of the American colonies. Hard-line advocates of Brexit became impatient when they learned that EU treaties meant the UK could not leave the European Union immediately. Under Article 50 of the Treaty on European Union, the British government needed to give two years’ notice in order to allow time for negotiating with the EU how to deal with the consequences of withdrawal and to discuss arrangements for future political relations between the EU and the UK as a non-member state. In addition, the government needed time to review thousands of clauses in British laws that made reference to EU legislation. Impatient Brexiters, confident that Britain could get what it wanted quickly, put pressure on the prime minister to withdrawal promptly. Senior Whitehall officials such as Ivan Rogers argued for delay until the government had a well-prepared strategy for withdrawal to ‘avoid being screwed’ by well-prepared EU negotiators (Shipman 2017: 135; Rogers 2019). The prime minister bowed to party pressures rather than civil service advice. To quell doubts among Conservative MPs that she was not fully committed to Brexit, May put forward a government bill to authorise withdrawal. On 1 February 2017, MPs voted 498 to 114 in favour. The massive majority was due to Conservative unity and a three-line Labour whip to respect the referendum majority. Scottish Nationalists and 47 Labour MPs were the chief opponents of the measure. It set 29 March 2019 as the date that Brexiters could celebrate as Britain’s Independence Day. In an attempt to establish her personal authority by leading the Conservative party to a big election victory implied by opinion polls, Theresa May called a surprise election in May 2017. The party manifesto listed many policies that would be taken back from Brussels. It also optimistically promised ‘a smooth and orderly departure from the European Union’ and ‘a deep and special partnership with our friends and allies across Europe’. The strategy backfired. Instead of gaining dozens of MPs as promised by pre-election polls, Theresa May lost the Conservative majority in the House of Commons (see Chapter 5). Many who had defected to UKIP at the previous election returned to the Conservatives. However May’s mismanagement of the campaign and the unexpected appeal of Labour under Jeremy Corbyn’s leadership resulted in the disappearance of the party’s opinion poll lead. In order to have a parliamentary majority, the Conservative leader had to rely on the votes of 10 Northern Ireland Democratic Unionist MPs, who much preferred Brexit

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

183

to Brussels. The very public demonstration that she was not an electoral asset reduced Theresa May’s nominal authority as a party leader. Being in opposition and internally divided on Europe, the Labour party adopted the position endorsed by Sir Winston Churchill: the opposition’s job is to oppose. Corbyn and his left-wing supporters were ideologically against the EU, which was seen as part of an international capitalist network. He proposed a unicorn policy of a new relationship with the EU that would protect British exports while leaving a Labour government free to impose socialist economic subsidies that would run afoul of the EU’s single market regulations. Most Labour MPs had cast their referendum vote to remain in the EU, but were divided about how to respond to defeat. The irrelevance of Corbyn’s policy provided cover for many Labour MPs while those who were committed to the EU emphasised the desirability of holding a second referendum giving voters a choice between endorsing the Conservative government’s withdrawal agreement with Brussels or remaining in the EU. The two parliamentary parties that favoured remaining in the EU were politically marginal. The Liberal Democrat party was a pro-referendum party as well as being pro-EU, but with only a dozen MPs its voice was barely audible. The Scottish National Party, the third-largest party in the Commons, was pro-European and invoked the Scottish majority to remain in the EU as justification for a unicorn policy of Scotland remaining part of the EU while England withdrew. This policy was unacceptable both in Westminster and in Brussels. It could only be achieved if the SNP won a referendum on independence and then joined the European Union.

10.3

A Domestic Foreign Policy

A domestic foreign policy is a zen foreign policy; it is the sound of one hand clapping, because it ignores the hand that foreigners have in politics that is literally international. An opposition party can have a foreign policy that meets pressures within the party and ignores foreign constraints because it does not have to deal with non-British institutions such as the European Union (Rose 1960). David Cameron could appease anti-EU MPs while Conservative party leader in opposition because his statements had no immediate effect. Likewise, Jeremy Corbyn could advance an EU policy that would not be tested by contact with foreign governments.

184

R. ROSE

Winning the 2016 referendum on withdrawing from the European Union meant that hard-line Brexiters could no longer make foreign policy without foreigners. While the British government could unilaterally notify withdrawal from the EU, the consequence has been years of negotiating with foreigners in Brussels. The radical change from debating a domestic foreign policy to being a bounded democracy negotiating with foreigners exacerbated conflicts within the Conservative party. Theresa May initially sought to avoid conflict with Brexiters by endorsing hard-line conditions for Brexit, including the extreme alternative of leaving the EU with no withdrawal deal. This did not end conflict; instead, it created a conflict between Downing Street and Brussels. When May resolved that conflict by agreeing a deal with Brussels, she stoked the domestic conflict within the Conservative party. Outcome Not in British Hands Even though Brexit was described by its advocates as taking back control from Brussels, to implement the referendum vote British policymakers have had to spend more than three years in Brussels negotiating terms of withdrawal and a Political Agreement about future relations (see Chapter 12). Negotiations have sought to minimise the disturbance of long-established economic and political links between the UK and Europe and to identify the extent to which there is mutual agreement about a broad or narrow range of policies where future relations may be little altered. Both the UK and the EU have been partners in what a senior EU official described as ‘lose, lose’ negotiations in which the immediate and visible costs of withdrawal have been greater than speculative long-term benefits. However, they have not been equal partners. The combined population of EU member states is seven times that of Britain. The EU is the world’s second-biggest economy—six times the size of the British economy. Withdrawal has left Britain with the challenge of maintaining exports to a single Europe market that buys almost half its exports. Even if the absolute costs of withdrawal were equal for both sides, because of differences in size, the relative impact would not be equal. The challenges of Brexit are central to the British economy while marginal to the European economy. The EU’s leaders began to react within hours of the referendum result being known (van Middelaar 2019). By noon of the day after the British

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

185

vote, the leaders of the European Commission, Council and Parliament issued a joint statement intended to protect the authority of the EU at a time when its policies for the euro and for immigration were under threat. The president of the Commission, Donald Tusk, stated as the goal, ‘That which does not kill us makes us stronger’ (Laffan 2019). The EU demonstrated a unity of policy by establishing a Brexit negotiating team led by Michel Barnier, a sophisticated politician very experienced in both EU and French politics. Concentrating responsibility in one place frustrated the efforts of Downing Street to pursue a divideand-win strategy of negotiating concessions in bilateral discussions with leaders of major member states, especially the German chancellor, Angela Merkel. Like her predecessors, Merkel saw Germany’s overriding priority as maintaining the political and economic unity of the EU rather than trying to help a national prime minister who could not even maintain the unity of the British government. Article 50 of the Treaty on European Union provided the legal authority for Britain unilaterally to withdraw from the EU and outlined procedures for doing so. The EU’s negotiating team identified alternative ways in which Britain’s future relationship could be mutually beneficial while respecting obligations of member states. It announced that it would start negotiations with the British government only once formal notice of withdrawal had been given. At this point, the clock started ticking: if no agreement could be reached within 24 months, then the UK faced the challenge of abruptly leaving the EU with no deal. Once Theresa May became prime minister in July 2016, she was hesitant about negotiating Brexit. The slogan that had helped her into Downing Street—Brexit means Brexit—had become a handicap. Its vagueness gave Whitehall civil servants no guidance about whether they should prepare briefs and tactics in pursuit of a hard Brexit or a soft Brexit. The vagueness was intentional. There was no agreement in the Cabinet about what form Brexit ought to take. Moreover, the ministers responsible for negotiating Brexit—Boris Johnson, David Davis and Liam Fox—preferred to repeat campaign rhetoric about the putative benefits of Brexit rather than face up to the disillusionment inevitable in calculating the cost of retaining benefits of EU membership. When Downing Street made the decision to trigger Brexit in March 2017, there was no British negotiating team in place nor was there a clear statement of what kind of Brexit the prime minister wanted. By default, the EU was free to establish an agenda for negotiating the terms of withdrawal.

186

R. ROSE

For the EU, the first priority was to get a legally binding agreement on immediate consequences of withdrawal before discussing a Political Agreement about future EU–UK relations. The three immediate issues were the size of the UK’s divorce bill to meet pre-existing financial commitments; the rights of EU citizens in the UK and British citizens in EU states; and the unique circumstances of Northern Ireland’s border with the Republic of Ireland. All three points in the Brussels agenda involved EU benefits and British costs. The Zaubertorte (magic cake) that Boris Johnson had promised—Britain free of the EU’s bounds while enjoying the benefits of membership à la carte—was not on the menu. In December 2017, a Joint Report gave British assent to the agenda of Brussels. This meant that British negotiators could not trade concessions on EU exports to the UK in return for EU concessions on UK priorities. May Decides Any Deal Is Better Than No Deal Negotiations about withdrawal confronted Theresa May with the Goldoni problem: the need to serve two masters (see Chapter 3). To secure a deal required not only approval by hard-line Conservative MPs committed to a domestic foreign policy but also approval by the EU negotiators. Both sets of negotiators shared the same goal: self-protection. However, the ‘self’ that demanded protection was different for the EU and the prime minister. For the EU, its primary concern was to protect its authority against disruption by the consequences of Brexit. For May, the aim was to find or fudge a deal that had sufficient support among Conservative MPs to maintain her position as prime minister. Theresa May presented her draft proposal for Brexit at an all-day Cabinet meeting at Chequers on 6 July 2018. Because the proposal included acceptance of a free trade zone, David Davis resigned as chief Brexit minister. He charged this would prevent realising the purpose of Brexit, Britain regaining control of all of its economic activities. Davis’s action prompted Boris Johnson’s resignation as foreign secretary a day later, arguing that May’s softening of hard Brexit terms would make the UK a colony of Brussels. Negotiating withdrawal required British politicians to learn that agreeing a deal within the EU bounds is different from promoting an unbounded domestic foreign policy. Since Theresa May’s red lines were inconsistent with the EU’s red lines, she had to water down pledges to her party to reach a deal. May decided that a withdrawal deal that the EU

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

187

would accept was better, or at least less unsatisfactory, than leaving the EU with no deal. In November 2018, the prime minister and the EU initialled a 599page Withdrawal Agreement. In it, the UK agreed to pay the EU £39 billion to meet its financial commitments; to recognise rights of UK and EU citizens living outside their country; and to keep an open border between Northern Ireland and the Republic of Ireland, guaranteed by a backstop arrangement (O’Rourke 2018). The conditions proposed for the border between Northern Ireland and the Republic left the prime minister exposed to attack from all sides (see Walker 2019). Getting an agreement approved by the prime minister and Brussels did not make it binding. This required securing endorsement by the British Parliament too. Opposition by Conservative MPs to the exceptional arrangements linking one part of the United Kingdom with the Republic of Ireland made parliamentary approval problematic. To soft Brexiters and even more to Remainers, the backstop could be made unnecessary by maintaining maximum economic links with the EU after withdrawal. But belonging to a customs union and the single Europe market would require the UK accepting EU laws and the jurisdiction of the EU’s court; this was anathema to hard-line Brexiters and to May herself. Conservative MPs, aggrieved by the concessions given to Brussels, demanded a vote of no confidence in Theresa May as party leader. Cabinet members on both sides of the hard/soft Brexit debate declared their support for the prime minister. To win support, May pledged that once Brexit was achieved she would not lead the party at the general election scheduled for 2022. The MPs’ vote favoured May, 200 to 117, but it was a pyrrhic victory. Half of the backbench Conservative MPs voted no confidence in their leader, and many with government posts gave only lukewarm support. When the whole House of Commons voted on Theresa May’s Withdrawal Agreement in January 2019, it suffered the biggest government defeat in modern political history: 432 MPs against and 202 MPs in favour. The defeat occurred because the Conservative party was badly split: more than one-third of its MPs voted against the deal. They were joined by parties that supported a soft Brexit, such as the Liberal Democrats and the SNP; the Northern Ireland Democratic Unionist Party (DUP), which wanted a hard Brexit; and a Labour party united in opposition to the Conservative prime minister. The Labour party immediately tabled a vote of no confidence in the government,

188

R. ROSE

which if successful would have resulted in a general election. May won another pyrrhic victory, 325 to 306, as all the Conservative MPs who had voted no confidence in her Brexit deal and the DUP gave formal endorsement to her government in order to avoid the risk of a general election returning a Labour government led by Jeremy Corbyn. After making cosmetic amendments, May again tried to secure parliamentary approval of her Withdrawal Agreement. On 12 March, the deal was rejected by a vote of 391 to 242, with 75 Conservative MPs voting to reject the deal. Nine days before the date on which the UK was due to leave the EU without any deal, May requested a three-month extension of the deadline. However, senior EU officials were losing patience with UK negotiators. Frans Timmermans and Guy Verhofstadt described British policy-makers as ‘running around like idiots’ and pursuing an ‘insane’ strategy of grandstanding to a domestic British audience rather than having a plan to win a deal with Brussels (Moore and Blackburn 2019). The EU offered two dates for an extension: leaving the EU on 12 April if the Withdrawal Agreement was again rejected by Parliament or a deadline of 22 May if the prime minister could use the extra time to gain parliamentary approval of her deal with Brussels. With her parliamentary support drained, Theresa May adopted a populist tactic: she made a television appeal to the British people to put pressure on MPs to back her deal. She declared, ‘You, the public, have had enough. You want this stage of the Brexit process to be over and done with. I agree. I am on your side’. She blamed MPs for the delay in getting Brexit done. MPs in all parties were furious with her for blaming delays on Parliament. Angry Conservative MPs called for her resignation. In a third vote on 29 March, MPs again rejected the deal by a vote of 344 to 286, because of the opposition of hard-line Conservative Brexiters. Having failed to win the support of her own party, in desperation May issued an early April call for all-party talks on implementing Brexit. However, there was no basis for all-party agreement. The Labour party had no wish to rescue the Conservative party from the hole it was in. Labour’s priority was to avoid blame for the problem of implementing Brexit. The Liberal Democrats and the Scottish National Party wanted a referendum on what should be done. The talks got nowhere. To avoid the UK leaving the EU by default without a deal, May requested another extension of the deadline for withdrawal to 30 June. The EU granted an extension to 31 October. This was not a vote of confidence in her ability to master the challenge of serving both

10

POLICY-MAKING IN A BOUNDED DEMOCRACY

189

hard-line MPs and EU negotiators. The longer extension gave time to the Conservative party to choose a new prime minister who might have the authority to secure a parliamentary majority for whatever could be agreed with Brussels. On 21 May, the prime minister announced a fresh 10-point plan for withdrawal. It was immediately rejected by leading Cabinet ministers as well as by opposition MPs. Two days before the European Parliament election result showed the Conservatives finishing in fifth place (see Table 11.2), Theresa May resigned as prime minister.

References BMG. 2016. Electoral Commission Post-EU Referendum Survey. BMG, 3,5533 Respondents, Birmingham. Farrell, J., and P. Goldsmith. 2017. How to Lose a Referendum. London: Biteback. Kahneman, D., and A. Tversky. 1979. Prospect Theory: An Analysis of Decision Under Risk. Econometrica 47: 263–292. Kirkup, J. 2019. The Truth About David Cameron and “the Mad, Swivel-Eyed Loons”. The Spectator, 17 May. Laffan, B. 2019. How the EU27 Came to Be. Journal of Common Market Studies 57 (Annual Review): 12–27. Moore, M., and J. Blackburn. 2019. Brussels Team Ridicules No. 10’s “Insane” Strategy. The Times, 9 May. O’Rourke, K. 2018. A Short History of Brexit: From Brentry to Backstop. London: Pelican. Rogers, I. 2019. 9 Lessons in Brexit. London: Short Books. Rose, C.R. 1960. The Relation of Socialist Principles to British Labour Foreign Policy, 1945–1951. DPhil, Faculty of Social Studies, Oxford. Seldon, A. 2019. May at 10. London: Biteback. Shipman, T. 2016. All Out War: The Full Story of Brexit. London: William Collins. Shipman, T. 2017. Fall Out: A Year of Political Mayhem. London: William Collins. Uberoi, E. 2016. European Union Referendum 2016. House of Commons Briefing Paper 7639, London. van Middelaar, L. 2019. Alarums and Excursions: Improvising Politics on the European Stage. Newcastle upon Tyne: Agenda Publishing. Walker, N. 2019. Time Line EU. House of Commons Briefing Paper 7639, London. Walters, S. 2013. Now New Claims Suggest Cameron Used “Swivel-Eyed Loons” Jibe. Daily Mail, 19 May.

CHAPTER 11

The Failure of Parliamentary Government Richard Rose

Membership of the EU tested the authority of parliamentary government in the UK. In the traditional model, the government of the day has wide discretion to make policies as long as it has a majority in Parliament. This is usually possible by the first-past-the-post electoral system awarding the party that has a plurality of votes an absolute majority of MPs. Voters can hold the government of the day to account in periodic parliamentary elections that offer a binary choice of who governs; the governing party decides what government does. Generations of politicians and political scientists have praised the British system as an ideal combination of democratic representation and effective government (see, e.g., Schumpeter 1952). The nineteenth-century theory of Parliament as the supreme authority has gradually become part of the dignified theory of the Constitution. Party loyalties in Parliament have ceded authority to party government, in which the prime minister and Cabinet are the effective government of the country. Since the Second World War the idea of British government as a closed system not bound by what happens outside the UK has increasingly been undermined (Rose 2001). Global economic pressures have forced major changes in economic policy starting with the 1949 devaluation of the pound. American pressure stopped Britain’s use of independent military action in the 1956 Suez War. Disputes over European Union membership have made the voice of the people the supreme source of legitimacy rather than Parliament or © The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_11

191

192

R. ROSE

party. Party government first broke down when Edward Heath sought approval for Britain joining the EU. Parliamentary approval was only secured by a number of Conservative MPs voting against membership being offset by a larger number of Labour MPs defecting from the Labour whip to vote for joining the EU. When Labour returned to office in 1974, Harold Wilson shifted the final say on EU membership from Parliament to a referendum in order to resolve the Labour party’s split. The UK’s first-ever referendum in 1975 approved Britain belonging to the European Union by a margin of two to one (Westlake, 2019: 17ff). It also set a precedent for subsequent calls for referendums on devolution and local government issues. Eurosceptics subsequently seized on the precedent to demand referendums before giving the EU additional powers that placed bounds on British policy-making. On 20 April 2004, Prime Minister Tony Blair told the House of Commons that it was only right to ‘let the people have the final say’ in a referendum on a proposed constitution for Europe. Blair also pledged to hold a referendum on Britain replacing the pound with the euro if five economic tests were met for joining the Eurozone. In opposition from 1997 to 2010, the Conservative and Liberal Democrat party made manifesto demands for referendums on joining the Eurozone and on treaties increasing the powers of the EU to put bounds on national policy-making. Political developments meant that none of these pledges was put into effect. By the time David Cameron became leader of the Conservative party in 2005 all parties had acknowledged that a referendum was appropriate before any further powers were transferred to the European Union. In opposition, Cameron gave a ‘cast-iron’ pledge’ to hold a vote on the Lisbon Treaty, and opinion polls showed his view had majority support. Prime Minister Gordon Brown dismissed the call, stating that Parliament was the proper institution for deciding whether to approve the treaty, and it did so. On entering Downing Street, Cameron rejected demands from Eurosceptic Conservatives for a post hoc referendum that would have threatened a crisis of legitimacy with the EU, which makes no provision for a national government to withdraw approval of a treaty. Since the EU does make provision for a member state to withdraw from the European Union, Eurosceptics shifted their efforts to demand a referendum on whether the UK should remain in or leave the EU. Given the precedents for such a ballot, Cameron, like Wilson, sought peace within his party by accepting their demand. The 2016 Brexit referendum was the result.

11

THE FAILURE OF PARLIAMENTARY GOVERNMENT

193

The 2017 general election gave Parliament the power to decide how or even whether Brexit should be implemented by collectively giving opposition parties the majority in the House of Commons and making the Conservatives a minority government. But divisions on Europe on both sides of the House of Commons resulted in a breakdown of party government. The Conservatives could not count on their Eurosceptic and Brexit-sceptic MPs both voting unanimously for whatever Downing Street proposed. Nor could Labour, the principal opposition party, produce disciplined support among its own MPs for an alternative policy. The failure of parliamentary government was confirmed when MPs used their collective majority to defeat a variety of policies for dealing with Brexit, including two measures that Prime Ministers Theresa May and Boris Johnson had negotiated with the EU. Public opinion polls suggested that voters were more strongly identifying themselves as belonging to Leave and Remain parties rather than as Conservative or Labour supporters. Parliament asserted its negative authority by rejecting proposals to hold a second referendum with the alternatives approving terms for implementing Brexit or remaining in the EU. Voters used the European Parliament election to demonstrate the breakdown of party government by giving Leave and Remain parties a majority of votes and giving the government and official opposition parties only a quarter of the vote. Paradoxically, the impasse was broken only when Boris Johnson succeeded in forcing a general election and winning by treating it as a referendum on getting Brexit done (see Chapter 12).

11.1

Parliament Asserts Its Authority Negatively

To treat minority governments as the absence of majority government is misleading, in that there is always a majority of MPs on one or the other side of the benches in the House of Commons. The critical question is whether the majority is on the government side of the House or on the opposition benches. If an absolute majority of MPs belong to the governing party, the most important opposition comes from within the prime minister’s party, as David Cameron found out after his party won a parliamentary majority at the 2015 election. After the 2017 election, Theresa May faced a majority of MPs sitting opposite her. She had the doubly daunting task of trying to keep the united support of Conservative MPs who disagreed about implementing Brexit and winning enough support from opposition MPs to get a majority vote in favour of her Brexit policy.

194

R. ROSE

Boris Johnson did not try to put together a coalition majority from both sides of the Commons. Instead, Johnson successfully turned the opposition defeat of his Brexit deal with Brussels into the justification for the December 2019 election that gave the Conservative government the parliamentary majority needed to deliver Brexit. Negative Numbers While it is numerically correct to refer to the opposition having a majority of MPs, it is politically misleading because the opposition is not a single party but a collection of parties that disagree with each other on policies and compete with each other for votes. In the House of Commons elected in June 2017, the 333 opposition MPs collectively had a majority of 15 over Conservative MPs. However, they belonged to seven different opposition parties. The support of 10 Northern Ireland MPs of the Democratic Unionist Party (DUP) added just enough votes to give the Conservative government the majority needed to win a vote of confidence. Although the 2017 parliamentary election returned 589 MPs representing parties nominally committed to leaving the EU, there was no agreement within or between parties about how this should be done or whether a second referendum should be called. The official opposition party, Labour, was internally divided between an anti-EU group led by Jeremy Corbyn, pro-EU MPs, and MPs prepared to accept a soft Brexit. Although the opposition parties collectively had enough MPs to vote no confidence in the Conservative party, they did not want to do so as long as this would result in Corbyn becoming prime minister. Thus, the parliamentary session launched in June 2017 became the longest session since the seventeenth-century English Civil War. It ended in November 2019 after consistently rejecting Brexit proposals put forward by the government and by opposition MPs. The government’s lack of a parliamentary majority enabled MPs to demand more information about the government’s negotiations with Brussels and impose conditions that gave parliament opportunities to vote on the government’s actions. An all-party select committee on Exiting the European Union was set up in October 2016, chaired by a Labour MP, Hilary Benn. Its 20 members were divided evenly between Conservative and opposition MPs. In the referendum, 13 of these MPs had voted to remain in the EU and 7 had voted for Brexit. Their activities were

11

THE FAILURE OF PARLIAMENTARY GOVERNMENT

195

abetted by the Speaker of the House of Commons, Sir John Bercow. In unprecedented circumstances, Bercow made unprecedented decisions that increased the capacity of MPs to put pressure on the government. MPs spent three years engaging in activities that provided lots of information about a great variety of Brexit-related policy choices. Sir William Cash, a long-time critic of the EU, described the harrying of the government by MPs as breaking constitutional conventions because it paved the way for ‘government by Parliament’ (Cash 2019). After rejecting the prime minister’s EU deal by the biggest negative vote in parliamentary history on 15 January 2019, MPs tried to take policy-making into their own hands. In a sense, anything a majority of MPs proposed could only be a domestic foreign policy, since the EU would only recognise the prime minister as the representative of the UK. Domestically, the prime minister could hardly ignore a positive recommendation indicating the position of a majority of MPs about Brexit. In total, MPs rejected 15 proposals concerned with Brexit (Table 11.1). A cross-party initiative of former Cabinet ministers led votes on eight proposals put forward by Conservative, Labour and Scottish Nationalist MPs on 27 March. Two endorsed no deal, two favoured remaining in the EU and four favoured some form of soft Brexit. While none was binding in the legal sense, all were politically significant. All eight measures were rejected by margins ranging from 6 to 313 votes. A soft Brexit motion put forward by a former Conservative Cabinet minister came closest to success. It called for a permanent and comprehensive UK-wide customs union to be part of any Brexit deal. A Labour MP’s motion asking for a second referendum received stronger support than the alternative plan put forward by Labour leader Jeremy Corbyn; however, both were rejected. Proposals for no deal were backed by less than one-quarter of MPs. Two days later, when Theresa May’s withdrawal agreement was put to the Commons for the third time, it again fell short of a majority. Given the impasse created by the House of Commons rejecting both the prime minister’s proposal for Brexit and all eight alternatives put forward by its own members, the Speaker called a second round of indicative votes on April Fool’s Day. Two proposals carried the possibility of remaining in the EU and two were soft Brexit options. All were again rejected, albeit by slightly lower margins than had been the case a week earlier (Table 11.1). After becoming Theresa May’s replacement as prime minister, Boris Johnson was able to negotiate a new withdrawal agreement with Brussels

196

R. ROSE

Table 11.1 House of Commons rejects 15 Brexit proposals

Theresa May’s deal 15 January 2019 12 March 29 March Commons motions 27 March Customs union Second referendum Labour alternative Common market Revoke article 50 No deal Managed no deal EFTA membership Commons motions 1 April Second referendum Common market 2 Revoke article 50 Boris Johnson’s deal 19 October 2019

For

Against

Margin of rejection

202 242 286

432 391 344

230 149 58

265 268 237 189 184 160 139 64

271 295 307 283 293 400 422 377

6 27 70 94 109 240 283 313

280 261 191

292 282 292

12 21 101

306

322

16

Source Calculated by the author from Nigel Walker 2019. Brexit Timeline. London: House of Commons Briefing Paper 7960

only a few weeks before the UK was scheduled to leave the European Union. In an emergency Saturday sitting of Parliament, the House of Commons showed that it was policy, not personality, that had led to the impasse between Parliament and prime minister. MPs refused approval of Johnson’s deal by a vote of 322 to 306. The repeated rejection of Downing Street proposals demonstrated, albeit negatively, the critical role of party in maintaining parliamentary government. When Parliament was asked to cast a collective vote in favour of what Brexit should mean, there was no majority party to support the government position. The opposition’s collective majority was not based on a party manifesto backed by the whip of a dominant party. It was an ad hoc alliance of MPs from six different parties who could cast up to 432 votes against what the prime minister declared Brexit meant. Without a party majority in the Commons, May and Johnson could choose but not govern.

11

11.2

THE FAILURE OF PARLIAMENTARY GOVERNMENT

197

Opinion Polls No Substitute for Elections

For voters to have any opportunity to update their preferences, there needs to be continuous feedback between governors and governed. Elections give the winning party the power to make public policies that should reflect the preferences and interests of their voters. However, ballots are blunt instruments. Winning an election is not proof that everything in the winning party’s manifesto is supported by public opinion. Moreover, in the years between elections a government faces many issues that were not debated at the latest election. Between elections as well as during campaigns, public opinion polls provide a flow of up-to-date information about the views of voters. The results are well publicised in the media, and political strategists assiduously study unpublished party polls for clues about what to do to increase public support. The evidence of polls is different from what journalists write based on their introspective insights or vox pop quotes obtained by occasional forays outside London. It also differs from what can be collected from social media sites, which can reflect extreme views. Even if a tweet attracts 100,000 likes, this is less than 1% of the electorate. Because surveys invariably show the public is divided about major issues, the evidence differs from assertions of politicians that assume everybody thinks as they do. What Opinion Polls Have Shown Leavers who made withdrawal from the European Union in 2020 an absolute obligation treated the opinion of those who voted for Brexit in 2016 as set in stone. Likewise, Remainers argued that the turnover of generations in the electorate as pro-Europe younger voters replaced an ageing cohort of Eurosceptic voters would produce a majority in favour of the European Union. They likewise assumed that voters have opinions fixed for life. However, neither assumption is correct. The fundamental characteristic of public opinion about the EU over 60 years is that it has fluctuated up and down in the short term and in the long term (Table 11.2). Opinion polls about Europe first became politically relevant in the 1960s when Harold Macmillan and then Harold Wilson put in applications to join European institutions. The standard question asked was whether the respondent approved or disapproved of joining what was

198

R. ROSE

then the European Economic Community. There was a 35 percentage point range between the highest and lowest level of support for membership. It was as low as 36% in June 1962 and again in 1968. In between these two troughs, there was a high of 71% approving applying for EU membership in June 1966. When Prime Minister Edward Heath initiated Britain’s third and successful attempt to join the what was then the European Economic Community public opinion was very negative: 56% disapproved of doing so and only 21% approved in August 1970. At no point did support rise above two-fifths of respondents, and in all 16 surveys a plurality or an absolute majority was against seeking EU membership. Heath was not for turning and the UK became a member state in 1973. Once the 1975 referendum was called to decide whether Britain should remain in Europe, polls began asking whether people would vote to leave or remain in the EU. As most of Harold Wilson’s Labour cabinet favoured membership and Margaret Thatcher, leader of the Conservative opposition, maintained Heath’s pro-EU commitment, polls showed a high and relatively stable level of EU backing throughout the 1975 referendum campaign. The two-thirds vote in favour was consistent with the results of polls (Table 11.2; Butler and Kitzinger 1976). Table 11.2 Public opinion about Europe fluctuates Period

1960–1967 1970–1973 Referendum 1975 1977–1996 1997–2007 2010–2016 Referendum 2016 2016–2020

High

71 41 73 63 53 61 55 48

Low Pro EU % 36 21 58 26 39 41 39 40

Range

Don’t know %

35 20 15 37 14 20 16 8

42 26 n.a. 17 22 13 16 16

Source 1960–2015 compiled from Westlake 2019: 182–188, where texts of questions and sources are given. The 1970–1973 data combine Tables 7.4–7.6. Post-2015 data from www.whatUKthinks.org reporting YouGov Eurotrack results, July 2016–January 2020

11

THE FAILURE OF PARLIAMENTARY GOVERNMENT

199

Once EU membership was endorsed by both Parliament and a referendum, polls only infrequently asked about the topic. In the wake of the economic crisis of the late 1970s public opinion turned against the EU, with a high of 61% favouring leaving and only 26% remaining in March 1980. After the economy recovered in the 1980s and Margaret Thatcher had secured new policies in Britain’s interest, attitudes reversed. By June 1991, there was a 63% majority in favour of EU membership. In the decade of his premiership, Tony Blair made tactical commitments to hold referendums on a Constitution for Europe and Britain joining the euro. Rejection of the Constitution in French and Dutch referendums made a British vote unnecessary, and opposition to joining the Eurozone by Chancellor of the Exchequer Gordon Brown meant there was no occasion for a referendum. Foreign policy attention shifted to Blair’s use of military force in Iraq and Afghanistan. The number of don’t knows rose to the highest level since the early 1970s, and there was usually a majority or plurality saying that if a referendum were called they would vote for remaining in the EU. By the time David Cameron became prime minister in 2010, fluctuations in public opinion were tending to move between a plurality for leaving the EU or a majority for remaining. While the Conservative government was steering a bill through Parliament to call an in/out referendum in June 2016, polls showed a majority in favour of remaining in the EU. During the referendum campaign, polls fluctuated between showing a plurality for remaining in the EU or for leaving the EU. Of the seven final pre-election polls, three correctly described their results as showing the outcome was too close to call because the Remain lead was within the margin of sampling error. In the event, the 51.9% vote for leaving was consistent with sampling error, but politically damaging for the polls, since six reported Remain in the lead while audiences thirsty for certainty about the future ignored small-print qualifications about sampling error (NCRM 2017). The pattern of fluctuations continued while the possibility of a second referendum was being debated in Parliament. Since the Johnson government treated its December 2019 election as a referendum on getting Brexit done, any poll question about EU membership since then would have to use more speculative language than in the 1960s. It would have to assume that there was a government that would apply to join the European Union and that it would call a referendum seeking endorsement of such a choice.

200

R. ROSE

Polls about Britain being in or out of European institutions show, as the 2016 referendum did, not only that British people are divided in about EU membership, but also that divisions have persisted for well over half a century. The peak level of support was before and during the 1975 referendum that confirmed Britain’s membership. In the two decades that followed there were big swings between a low of 26% approval of EU membership in 1980 and a peak of 63% in 1991. Since then the peaks and troughs have eased, and support for belonging to the EU has normally been between 40 and 60%, depending on political circumstances. To some extent, fluctuations are amplified because a substantial fraction of people interviewed have little interest or knowledge about European affairs. Rather than admit ignorance or indifference, some will give a superficial, spur-of-the-moment response that is not a stable position but a non-attitude; it nonetheless counts as an opinion in reports of survey results (Converse 1970). In the many surveys over the decades in which there was no majority support for or against the EU, the median respondent had no opinion. However, when a referendum is held, the votes they cast are just as valid as the choices of the well informed. As political developments have made the European Union more salient to Britain, the proportion of self-confessed don’t knows has dropped from a high of 42% before British joined the EU to a sixth or less in the past decade, and the proportion of opinions that are weakly held non-attitudes has fallen too. Strictly speaking, polls are about hypothetical events. They ask people how they would vote if a referendum or a general election were held today, when a vote is days, weeks or years ahead. Like economists, pollsters acknowledge that today’s results can only be projected into the future on the assumption that all other conditions remain equal. The assumption of total stability ignores the fact that the purpose of a political campaign is to create conditions that change the behaviour of enough of the electorate to give the campaigners victory. Britain joined the EU in 1973 because Edward Heath looked to his wartime experiences rather than public opinion polls when risking his career in disregard of public opinion. Britain left the EU in 2020 because Prime Minister David Cameron was reassured by opinion polls that if he called a EU referendum he would win it.

11

THE FAILURE OF PARLIAMENTARY GOVERNMENT

201

The Lack of an Institutional Link Opinion polls cannot substitute for elections when there is no explicit link between a survey question and a specific government decision. For example, the Eurobarometer of the European Commission periodically evaluates support for the European Union by asking people whether they think the European Union is a good thing or a bad thing. When Britons were asked this question in April 2019, a positive reply was given by 43%, almost twice the proportion thinking the EU a bad thing (Eurobarometer 2019: 65ff.). While a positive evaluation of the EU is always desirable, it has no specific meaning for EU policy. By contrast, the 2016 Brexit referendum question produced a politically binding result. Many opinion polls went beyond asking which side people took in the Leave vs Remain debate. Four alternatives could be readily linked with votes in the House of Commons: leaving Europe without a deal, a hard Brexit, a soft Brexit or remaining in the EU (see Fig. 11.1). This invariably produced a plurality rather than an absolute majority in favour of the leading option.

Fig. 11.1 Most and least popular choices for implementing Brexit (Source Christina Pagel and Christabel Cooper. London. University College London, Survey of 5307 respondents conducted by YouGov, 22–26 March 2019)

202

R. ROSE

Whereas rules about counting referendum votes to determine the winner are given, the identification of the winner in a poll can depend on how the results are interpreted. Pagel and Cooper (2019) demonstrated this by asking people to rank four alternatives for implementing Brexit from the most to the least favoured. Offering four options resulted in no single alternative having an absolute majority. The first preference of 45% was in favour of remaining in the EU. The first preferences of those in favour of leaving the EU divided three ways. A total of 27% preferred leaving with no deal; 14% wanted a softer Brexit; and 14% endorsed Theresa May’s EU deal (Fig. 11.1). While the only option for remaining in the EU came first, leaving the EU by one means or another had the collective support of an absolute majority. An absolute majority can be arrived at in two different ways. The alternative vote system does so by eliminating, in order, choices with the least first-preference support and redistributing their second and if need be third preferences until only two alternatives are left. At this point, there is certain to be an absolute majority in favour of one of the two alternatives. This method is used to elect the mayor of London. Applying it to Brexit preferences redistributes the 27% who initially endorsed May’s deal or a soft Brexit. The no-deal option gains three-quarters of these preferences. However, because remaining in the EU had so much initial support its much smaller gain from redistribution is sufficient to make it the winner, with 52% favouring it. Another way of summing preferences is to start by eliminating the most unacceptable alternative. When this is done, the outcome is reversed. At the first stage, remaining in the EU is eliminated, because it is put last by an overwhelmingly proportion of the majority in favour of one or another form of Brexit (Fig. 11.1). At the second stage, the transfer of preferences vetoes no deal, a very unpopular alternative among Remainers. Having eliminated the two most unpopular alternatives, the final choice is between a soft Brexit and May’s deal. There is a 62% majority for a soft Brexit as the least worst alternative, even though it is also the option with the least initial support. In an effort to secure a second referendum, the People’s Vote campaign linked polling about Brexit with lobbying opposition MPs to use their hung parliament majority to call a second referendum in which remaining in the EU would be an option on the ballot. This was particularly relevant for Labour MPs. Many felt cross-pressured; even though they personally favoured remaining in the EU, they did not want to reverse the 2016

11

THE FAILURE OF PARLIAMENTARY GOVERNMENT

203

referendum vote in which a majority of their constituents had endorsed leaving the EU. Extensive survey data often indicated that the pro-Brexit vote in a Labour constituency was often due to a minority of Labour voters joining pro-Brexit supporters of other parties, while most Labour supporters voted for remain. Although the People’s Vote campaign could not claim that a second referendum was certain to reverse the 2016 result, surveys suggested there was a good chance of achieving this goal with a well-run campaign (https://yougov.co.uk/topics/brexit/survey-results). To succeed, the People’s Vote first had to win a House of Commons majority to hold another referendum. Doing so faced multiple obstacles of principle and practice. The primary division among Conservative MPs was about how withdrawal should be implemented, not whether it should happen. A second referendum was strongly opposed by almost all Conservative MPs because it would jeopardise the achievement of Brexit. There was substantial support among many Labour MPs for a second referendum, but hesitancy to defend it against the charge that they were seeking to undermine the voice of the people expressed in 2016 and endorsed by the Labour party’s election manifesto the following year. Before Easter 2019, the House of Commons twice voted on motions calling for a second referendum. In both cases, a big majority of Labour MPs voted in favour of holding a second referendum while a virtually united Conservative bloc voted against, and each time a second referendum was rejected. More than 75 MPs abstained in each round rather than challenge a decision of their voters (Table 11.1). After Boris Johnson became prime minister, he successfully argued that the vote to test support should be a general election rather than a second referendum, and it gave him the parliamentary majority he sought to get Brexit done forthwith.

11.3

Electing MEPs: The Referendum Re-run

An unintended and unwanted consequence of the delay in withdrawal from the EU was that in May 2019 the Conservative government faced two elections with Brexit not yet done: local government ballots and the election of British members of the European Parliament (MEPs). The Conservative government’s handling of the implementation of Brexit had shown the party as both divided and incapable of implementing Brexit as promised, and Labour was unable to come up with a realistic solution.

204

R. ROSE

The results of both ballots showed that voters had no confidence in either the governing party or the official opposition. In local government elections in England and Wales on 2 May, the Conservatives lost more than 1300 English council seats they were defending. Instead of making big gains from the Tories, Labour lost more than 300 seats. The big winners were the Liberal Democrats, who gained more than 650 seats, and councillors running independent of any party. Conservative activists blamed Theresa May and her government for their massive losses, and the Labour leadership was reminded that when the government’s vote goes down there is no assurance that the vote of the opposition will increase. An unwanted by-product of the extension of the deadline for Britain leaving the EU was that the EU insisted that, as it was still a member state, the UK should participate in the European Parliament election on 23 May. British participation had little significance for European institutions, but a big impact on British politics because it created what was, in effect, a second referendum on Europe. Instead of offering a referendumtype choice between Remain and Leave, it offered a choice between a multiplicity of parties claiming votes on two very different grounds. Parties with unambiguous pro- or anti-EU policies welcomed the chance to campaign for votes by appealing to those who had voted leave or remain in the 2016 Brexit referendum. Without agreed policies on Europe, the Conservative and Labour parties made little effort to get out the vote for their party’s candidates. At the time of the EP election, the most important issue—Brexit— was intermestic, that is, relevant to decisions taken in Westminster and in Brussels. Therefore, the repudiation of parties of British government could not be dismissed as irrelevant, even though MEPs sat across the English Channel from Westminster. However, as Franklin and Nielsen have argued (2017), the EP election gave voters the opportunity to express their views about what they regarded as important without the risk that a party they disliked would take control of government. As in previous European Parliament elections, a majority of Britons expressed their lack of interest in engaging with the EU by not bothering to vote in the EP election. Turnout of 37% was two-fifths below the turnout for the election of Westminster MPs. Low turnout makes EP voters less representative of the demographic characteristics of the population than a sample survey. Nonetheless, the results did more than register

11

THE FAILURE OF PARLIAMENTARY GOVERNMENT

205

preferences: they showed the competitive electoral strength of parties with contrasting views of Brexit (see Chapter 5). The European Parliament requires the use of proportional representation to allocate MEPs to parties, and this gave Nigel Farage a chance to mobilise anti-EU sentiment. Because UKIP, which Farage had led to first place in the 2014 EP election, was full of internal conflicts, he launched a new single-issue party aptly called the Brexit Party. Its symbol on the ballot paper was a horizontal arrow, the exit symbol in public places everywhere. The Brexit Party attacked the Conservative government for failing to meet its self-imposed deadline for leaving the EU. It asked for votes to ensure Britain left the EU promptly and free of the conditions that Theresa May had agreed with Brussels. This strategy succeeded: 64% of those who had voted for leaving the EU in the 2016 referendum backed the Brexit Party, as did a majority of voters who had voted Conservative at the 2017 general election. The EP result turned the system of party government upside down. The Conservative and Labour parties together got only 23% of the UK vote. Labour’s 14% share was the lowest since it began contesting elections nationally in 1918. The Conservative vote of 9% was its lowest in almost two hundred years of contesting elections. The result undermined Labour’s strategy of winning a parliamentary majority by the collapse of the Conservative vote. Among those who had voted Labour in 2017, a majority switched their support to other parties: 43% turned to proRemain parties and 13% switched to the Brexit Party (Ashcroft 2019). Proportional representation gave parties with few or no seats at Westminster more than two-thirds of the popular vote and three-quarters of the UK’s seats in the European Parliament. The newly created Brexit Party came first with 31% of the vote, followed by two parties advocating remaining in the EU, the Liberal Democrats and the Green Party (Table 11.3). Voting was polarised along Leave/Remain rather than left/right lines. Only 10% of those who voted to leave the EU in the referendum voted for a Remain party and only 14% who voted Remain voted for Leave parties. Among parties supporting ties with the EU, there was a four-way contest. The Liberal Democrats came out best, winning half the pro-EU vote and the Green Party won seven seats in the European Parliament, compared to only one seat in the British Parliament. The electoral failure of the Change UK party, formed by a handful of Labour and Conservative MPs who could not take their party’s policy on Brexit, confirmed the

206

R. ROSE

Table 11.3 UK vote for European Parliament 2019

Pro-Brexit Brexit Party Conservative UKIP/DUP Ambivalent Labour Other Pro-EU Liberal Democrat Green Change UK Nationalists

48% 34% 10% 4% 17% 16% 1% 35% 15% 11% 5% 4%

Source Author’s classification from data in Fella, S., Uberoi, E., and Cracknell, R. 2019. European Parliament Elections 2019. House of Commons Library Briefing Paper 8600

Liberal Democrats as the leading party for Remain voters. In Wales and Scotland, Plaid Cymru and the Scottish National Party won more votes than either the Conservative or Labour party. Voting for British parties in the EP election was along referendum rather than party lines. The EP contest produced a 48% plurality of votes for pro-Brexit parties, but not an absolute majority. The median position between the two poles was occupied by the Labour party. However, in a polarised political context the party’s facing-both-ways position was unattractive and its attempt to make the EP vote about economic issues appeared irrelevant. The majority of Labour’s general election supporters the year before defected to the Liberal Democrats or the Greens. The British vote had little impact on the European Parliament, where the UK has accounted for less than one-tenth of the EP’s seats and Britain’s governing parties had shown little interest in the multi-national politics of the EP. Moreover, in anticipation of the UK leaving the EU, the European Parliament had redistributed British seats among 14 member states, and their voters elected candidates ready to take the place of British MEPs when Brexit occurred. The consequences of the EP result for British politics were immediate and severe. The first post-election YouGov poll asking how people would vote in a general election placed the Liberal Democrats first and the Brexit Party second at 24% and 22%, respectively. Labour was third with a sixth of the vote and the Conservatives got only 10%. Theresa May could not escape responsibility for giving Nigel Farage the opportunity to launch a party that threatened the Conservatives with a disastrous parliamentary election. Her resignation as party leader and prime minister opened up a

11

THE FAILURE OF PARLIAMENTARY GOVERNMENT

207

race for Downing Street among candidates promising to stop the haemorrhage of Conservative votes to a party that unequivocally wanted to get Brexit delivered.

References Ashcroft, Lord. 2019. How Britain Voted and Why. lordashcroftpolls.com/ category/elections, 12 December. Butler, D.E., and U. Kitzinger. 1976. The 1975 Referendum. London: Macmillan. Cash, Sir W. 2019. Hansard, vol. 661, Column 697. London: House of Commons, 12 June. Converse, P.E. 1970. Attitudes and Non-Attitudes. In The Quantitative Analysis of Social Problems, ed. E.R. Tufte, 168–189. Reading, MA: Addison-Wesley. Eurobarometer. 2019. Closer to the Citizens, Closer to the Ballot. Brussels: European Commission. Fella, S., E. Uberoi, and R. Cracknell. 2019. European Parliament Elections 2019. London: House of Commons Library Briefing Paper 8600. Franklin, M.N., and J. Nielsen. 2017. The 2014 Elections as a Lens on Euroscepticism. In The Eurosceptic European Parliament Elections 2014, ed. J. Nielsen and M.N. Franklin, 239–254. London: Palgrave Macmillan. NCRM. 2017. Opinion Polling in the EU Referendum: Challenges and Lessons. Southampton: National Centre for Research Methods. Pagel, C., and C. Cooper. 2019. A Country of Purists. Politics.co.uk, 14 June (London). Rose, R. 2001. The Prime Minister in a Shrinking World. Oxford and Boston: Polity Press. Schumpeter, J. 1952. Capitalism, Socialism and Democracy, 4th ed. London: Allen & Unwin. Walker, N. 2019. Brexit Timeline: Events Leading to the UK’s Exit from the European Union. London: House of Commons Research Briefing Paper 7960. Westlake. 2019. Slipping Loose: The UK’s Long Drift Away from the European Union. Newcastle upon Tyne: Agenda Publishing.

CHAPTER 12

A New Prime Minister Meets Old Constraints Richard Rose

When Boris Johnson achieved his long-standing ambition of becoming prime minister in July 2019, he inherited a government in political disarray and without a parliamentary majority. His predecessor Theresa May had badly split Conservative MPs with her definition of Brexit, which had encouraged MPs to vote against the government in sufficient numbers to defeat Downing Street proposals (see Table 11.1). Johnson’s career as a provocative journalist made him a leading campaigner for leaving the EU in the 2016 referendum. However, it also showed he valued a headlinecatching story about the EU over factual accuracy. As an MP, his rhetoric about Brexit has mixed statements that fact-checkers could readily disprove and claims that were disputable. His brief period as foreign secretary was marked by diplomatic gaffes, while resigning freed him from responsibility for May’s failures. On achieving his ambition to become prime minister, Johnson faced old constraints that go with the job at number 10. After delivering Brexit, he faces a new constraint: to deliver the brighter future that he promised would result. Boris Johnson was elected Conservative party leader in the hope that he would deliver a hard Brexit. In his first six months in Downing Street, he achieved striking results. He struck a withdrawal deal with Brussels that was little different from that negotiated by Theresa May. When it was rejected in Parliament, he turned aggressively populist, forcing a general election in which he framed the election as a referendum with the choice: the People vs Parliament. The slogan—‘Get Brexit done’—combined an © The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_12

209

210

R. ROSE

appeal to confirmed Eurosceptics and people who, whatever their choice in 2016, felt it was time to move on to more important things. The morning after winning by sweeping dozens of former Labour seats he went to Tony Blair’s old constituency of Sedgefield and proclaimed, ‘We are the people’s government. We are not the masters, we are the servants now and our job is to serve the people of this country and deliver on our priorities’ (www.bbc.co.uk/news/politics, 13 December 2019). Up to a point, Johnson immediately delivered on his slogan. Less than eight weeks after his election victory, the UK ceased to be a member state of the European Union. Johnson celebrated departure as making Britain a global power independent of EU constraints. However, this did not complete the Brexit process. It simply started a transition in which the UK remains de facto part of the EU economy while Brussels and Downing Street negotiate a new Political Agreement to make Brexit work. Johnson promises to achieve an agreement leaving the UK free of its past EU bounds. EU officials have emphasised that the old constraints of interdependence are still in place. The extent to which UK retains economic benefits of membership depends on the extent to which it accepts EU obligations. In other words, interdependence faces Johnson with the same choice as Theresa May: to accept a deal that satisfies Brussels or to end up with no deal.

12.1

Winning Downing Street

Taking Over the Conservative Party As soon as Theresa May resigned, Boris Johnson was quick to announce he was standing for the party leadership, giving him the key to Downing Street. One source of strength was his ability to win votes, as demonstrated by twice being elected mayor of Greater London. At a time when the Conservatives were trailing in the opinion polls his campaigning skills made him appear a potential election winner when pitted against Jeremy Corbyn and Nigel Farage. In the three years after Johnson’s shambolic first attempt to become prime minister, his opposition to May’s negotiations had established him as a committed adherent of Brexit. Johnson campaigned for the leadership with a ‘do or die’ commitment to take the UK out of the European Union. Echoing Farage’s pledge to repair a broken political system, Johnson claimed that respecting the will of the referendum majority was ‘fundamental to trust in democracy’. In

12

A NEW PRIME MINISTER MEETS OLD CONSTRAINTS

211

doing so, he created a bidding war in which his opponents were pressed to be just as hard or harder in their position on Brexit. His chief opponent, Foreign Secretary Jeremy Hunt, a late convert to Brexit, declared, ‘If we don’t do what the people tell us to do, we’re not a democracy’ (Sylvester 2019). In the two-stage process for electing the Conservative leader there were seven candidates. In the initial round, Johnson came first with the votes of 114 of 313 Conservative MPs. After other candidates were progressively eliminated in the fifth and final round, he won the backing of 160 MPs, an absolute majority. In the contest against Hunt for the support of Conservative party members, Johnson won 66.4% of the 139,318 votes. This made him not only party leader but also prime minister. No Majority to Govern The legacy of Theresa May’s handling of Brexit left the new prime minister facing formidable tasks. Internal disagreements in the Conservative party meant that not only was there no withdrawal agreement acceptable to Brussels but also there was not even a domestic foreign policy, that is, an agreed position among ministers and backbench Conservative MPs. Internal party divisions exacerbated the effect of being a minority government. The government’s standing with public opinion was even worse. At the start of January 2019, the YouGov poll gave the Conservatives 41% support, six points more than Jeremy Corbyn’s Labour party. After multiple defeats in the House of Commons and the European Parliament election, when Theresa May left office YouGov showed the Conservatives with only 19%, tied for third with Labour and behind the Liberal Democrats and the newly formed Brexit Party. By the time that a new leader could be installed in Downing Street there would be only a few months before the UK was officially due to leave the European Union. In keeping with his image as an English eccentric, Johnson was happy to be associated with cricket, but once in Downing Street he made the name of the game hard ball. Instead of including MPs of diverse views in his Cabinet as Theresa May had done, he sacked 17 ministers and gave top jobs to MPs who were committed to Brexit. When 21 Conservative MPs, including former Cabinet ministers, broke ranks to support a measure that imposed constraints on his dealings with Brussels, Johnson withdrew the party whip. One of those pushed out, May’s Chancellor of the Exchequer, wrote, ‘I no longer recognise this party of radicals’ (Hammond 2019). In

212

R. ROSE

his first appearance in the House of Commons as prime minister, Johnson uncompromisingly pledged to fulfil his promise to the people to come out of the EU by 31 October with or without a deal. Whereas Theresa May gave Parliament its power over Brexit negotiations by losing a general election, Boris Johnson was determined to take back control for Downing Street by calling a general election and winning an absolute majority of MPs. As long as he could convince Brexit supporters to vote Conservative, and the Remain vote was divided, the firstpast-the-post electoral system would convert a plurality of votes into the absolute majority he sought. Opinion polls were encouraging. By the time Parliament met in September after the summer recess, polls showed the Conservatives with a lead of 10 percentage points or more over Labour and the Brexit Party relegated to fourth place. A cross-party majority of MPs was fearful that in eagerness to meet his self-imposed deadline for withdrawal by 31 October Johnson would accept Britain leaving the EU without a deal. To prevent this happening, Parliament approved an Act requiring the prime minister to ask the EU for an extension of the date of withdrawal to 31 January 2020 if Parliament had not approved a withdrawal deal by Johnson’s ‘do or die’ deadline (cf. Walker 2019). Johnson called this measure a ‘Surrender Act’ because he saw the threat to leave without a deal as putting pressure on Brussels to make concessions rather than a powder keg that could blow up and damage the British economy. In an attempt to prevent the House of Commons from imposing constraints on Downing Street’s dealings with Brussels, Johnson promptly prorogued Parliament so that it would not meet for six weeks. The reason he gave was that the new government needed time to prepare measures to introduce in the new session of Parliament. Opponents of Johnson’s Leave strategy filed suits in Scottish and UK courts to annul his action. On 24 September, the UK Supreme Court unanimously ruled that Johnson’s decision was unlawful. A retired Supreme Court justice explained the Court’s upholding the supreme legitimacy of Parliament against Johnson’s ‘constitutional vandalism’. The Court’s decision did treat the referendum as a source of legitimacy (Sumption 2019).

12

A NEW PRIME MINISTER MEETS OLD CONSTRAINTS

213

The Bounds of Brussels Boris Johnson sought to escape the constraint that Parliament had imposed on his goal of leaving the EU in October by securing a lastminute compromise deal with Brussels. He hoped to sell it both to softBrexit MPs by saying it met their priority of avoiding leaving the EU without any deal and to hard-Brexit MPs by claiming it met their demand to take Britain out of the EU. The change of British prime ministers left the policy of Brussels unchanged. Jean-Claude Juncker, the president of the European Commission, declared that there was no scope for renegotiating the deal that Brussels had agreed with Theresa May and the British Parliament had rejected three times. He called it the best and only agreement possible. Michel Barnier, the EU’s chief negotiator for Brexit, dismissed Johnson’s fresh proposals for withdrawal as a combative and unacceptable demand. The Irish Taoiseach Leo Varadkar described new proposals for a border between Northern Ireland and the Republic as not in the real world. Leaked Whitehall files showed that if Britain left the EU without a deal at the date of Johnson’s self-imposed deadline there would be substantial and immediate disruptions to trade, industry and the health service. EU leaders assumed that Johnson would live up to his reputation for flexibility by agreeing a deal within the EU’s red lines and then sell it to Parliament. The Withdrawal Agreement that Johnson reached with the EU on 17 October differed little from the three points in the ill-fated agreement with Theresa May. The UK still had to pay tens of billions of pounds to meet its pre-existing financial commitments to the Union and there was little alteration in provisions for protecting the rights of EU citizens in Britain and British citizens living in EU states. The agreement removed a backstop ensuring the border for trade between Northern Ireland and the Republic of Ireland was kept open on conditions set by the EU. Johnson had branded the backstop as ‘anti-democratic’ and presented its removal as a triumph to pacify hard-Brexit MPs. However, the border arrangement he agreed to—the introduction of a check on internal trade between Great Britain and Northern Ireland—had previously been described by May as something no British prime minister could accept and by Johnson as inconsistent with the sovereignty of the UK. The series of votes that followed in the House of Commons led to an impasse. On 19 October, in the first Saturday sitting of the Commons in more than a third of a century, the Commons approved by 322 votes

214

R. ROSE

to 306 a motion by Conservative MP Oliver Letwin to delay approval of his new agreement and force the prime minister to fulfil his obligation under the Benn Act to request a further delay in withdrawal until 31 January. Johnson did so by sending Brussels the Act of Parliament requesting delay on a plain sheet of paper without his signature, and simultaneously sending a signed letter stating why he thought delay undesirable. The following Tuesday Johnson won an initial vote approving his deal in principle. However, MPs showed their distrust of the prime minister’s plan to refuse Parliament time to scrutinise it in detail by rejecting his timetable for rushing enactment to meet his end of October deadline. Johnson sought to overcome MPs’ repeated rejection of his proposals by calling an election in the belief that voters would return a pro-Brexit majority in the new Parliament. However, the Fixed-term Parliaments Act of 2011 frustrated Johnson’s proposal to call an early election before 2022. Each time the government proposed an early contest, it failed to meet the Act’s requirement that calling an early election required the positive support of two-thirds of all MPs. The Labour party was unwilling to co-operate in triggering an election, arguing that Brexit should be settled before a fresh election was held. Moreover, opinion polls indicated that the Conservatives had a fair chance of winning an absolute majority over Labour. To get what he wanted, Johnson introduced a fresh bill to hold an election on 12 December. Because no Parliament can bind its successor, it overrode the earlier Fixed-term Parliaments Act and only required a majority of MPs’ votes to secure adoption. This was achieved by division among the opposition parties. While Labour was fearful of a popular vote, the bill was supported by the Scottish National Party, which rightly saw this as an opportunity to gain seats and by the leader of the chief proRemain party, the Liberal Democrats, who wrongly saw an election as an opportunity to gain enough seats to eject Johnson from Downing Street and install a government that would reverse the referendum decision. The election bill was fast-tracked through the Commons and approved on 30 October, the day before Johnson was due to die in a metaphorical ditch, in his own words, because of failing to deliver Brexit by then. It was also the day on which a YouGov opinion poll reported the Conservatives enjoying a 15% lead over Labour.

12

12.2

A NEW PRIME MINISTER MEETS OLD CONSTRAINTS

215

Winning a Referendum Against Parliament Johnson Plays to His Strengths

Running a general election with the slogan ‘get Brexit done’ replicated the simple emotional appeal that his chief adviser, Dominic Cummings, had used to win the Brexit referendum. It also played to Johnson’s journalistic ability to frame issues in simple headline terms. It was designed not only to appeal to those who had voted to leave the EU three years earlier but also to others who accepted the authority of a referendum and were tired of endless indecision in Westminster. Labour’s ambivalent policy could only be framed in terms of ‘Yes, but we want to leave on much better terms’ or ‘No, but if second referendum produces another majority for leaving the EU we will respect it’. Getting Brexit done appealed to a significant bloc of voters who saw a hung Parliament as ‘worrying, depressing or despairing’. Moreover, by an 11-percentage-point margin voters saw a Labour government led by Jeremy Corbyn as a worse outcome than leaving the European Union (Ashcroft 2019). The Brexit slogan was used to counterattack the big spending policies that Labour promised to end what it described as nine years of economic austerity. When social policies were raised with Johnson, he would typically argue that after Brexit the UK would no longer be contributing billions of pounds a year to the EU budget and this money could be spent on improving health care, education and other popular services. As a form of political rhetoric, it linked the party’s pro-Brexit stance with support for popular social policies. Johnson was correct to emphasise that after paying a £39 billion divorce bill to the EU the UK would no longer have to make a £13 billion annual contribution as well. But, from a public finance point of view, the statement omitted many relevant facts, such as the billions of pounds that EU programmes paid to British institutions and farmers. Framing the election in terms of the People vs Parliament was a contradiction in terms of the theory of representative democracy, because they depend on each other (see Chapter 1). Members of Parliament need popular votes to get elected and re-elected. Likewise, the people need MPs to decide which party forms a government and to monitor the activities of government on their behalf. Ironically, Eurosceptic Conservative MPs had used their position in Parliament to force the Brexit referendum on a prime minister who saw no popular demand for such a vote.

216

R. ROSE

Claiming to speak for the People with a capital P against the political elite is a textbook strategy of populist party leaders; they charge members of Parliament and government ministers with being a self-interested class who do not care for what ordinary people think (Mudde and Kaltwasser 2017). There is substantial empirical evidence that the political elite are seen as unresponsive by the people they represent (Table 1.1). Britons tend to have a low level of trust in institutions of representative government such as parties and Parliament, and 74% see a majority of politicians as ready to say one thing to get elected and then do the opposite once in office (Rose and Wessels 2019: Table 1). For an Eton- and Oxford-educated classical scholar to speak for the people against the elite appears paradoxical. But from Johnson’s point of view, it is a means of intimidating a Parliament that had imposed checks on what he wants his government to do and on judges said to be making decisions that ‘conduct politics by another means’. The Conservative Party Manifesto (2019: 48) promised to set up ‘a Constitution, Democracy & Rights Commission that will examine these issues in depth, and come up with proposals to restore trust in our institutions and in how our democracy operates’. By contrast with the tightly controlled and focused campaign led by Boris Johnson, the opposition was divided into multiple parties. They differed about whether to settle for a soft Brexit or to demand a second referendum that could reverse the decision of the narrow majority to leave the EU in the first referendum. The Labour party wanted to avoid the issue of Brexit because its Labour MPs in the London area tended to represent constituencies that had voted to remain in the EU, while Labour MPs in the North of England had voted to leave. Jeremy Corbyn viewed the EU as a capitalist institution imposing austerity on Europe. A majority of Labour MPs lacked confidence in Corbyn’s fitness to be prime minister. In a Parliamentary Labour party vote of confidence after the 2016 referendum, 197 Labour MPs voted no confidence in Corbyn’s leadership and only 40 supported him. Johnson’s personal reputation for untruthfulness was offset by Corbyn’s disdain for traditional English values and reliance on support from a Marxist and Trotskyite coterie associated with undemocratic regimes and left-wing anti-Semites. Conservative campaigners saw the Labour leader Jeremy Corbyn as an electoral asset second only to Brexit. YouGov surveys supported this view. In a pre-election poll, 59% considered Corbyn unsuitable to be a

12

A NEW PRIME MINISTER MEETS OLD CONSTRAINTS

217

party leader. While Boris Johnson also had a negative rating among the electorate, by comparison with the Labour leader he was the lesser evil. When Lord Ashcroft’s final pre-election poll asked about the worst election result, 48% said it would be a Labour government under Jeremy Corbyn as against 38% saying leaving the European Union would be worse. The logic of the first-past-the-post electoral system recommended that opposition parties make electoral pacts so that there would only be one anti-Brexit candidate in each constituency. By successfully putting pressure on Nigel Farage to withdraw his party’s candidates from Conservative-held seats to get Brexit done, the Conservatives assured themselves of the pro-Brexit vote. However, major political and personal differences between parties and their leaders prevented electoral pacts being agreed that would consolidate the anti-Brexit vote in a way that might have prevented Johnson gaining a parliamentary majority. Opponents of Brexit had no inhibition about voting tactically, and campaigners for a second referendum to reverse Brexit placed loyalty to their referendum cause above party loyalties. They set up a number of websites offering advice about which anti-Brexit Party had the best chance of unseating a sitting Conservative MPs. Recommendations were based on a mixture of sources such as the constituency voting patterns, census data and surveys. Given differences in data and interpretation, tactical voting websites sometimes disagreed about which party had the best chance of winning a seat for the anti-Brexit side. A Majority for Getting Brexit Done While the Conservative party was consistently ahead of Labour in opinion polls during the campaign, there was uncertainty about whether Johnson would win an absolute majority or end up heading the biggest party in a hung parliament. Two of the three preceding Conservative victories had involved a hung parliament and this had stopped Theresa May from delivering Brexit on schedule (see Chapter 11). If Boris Johnson had won a plurality rather than a majority of seats he would have been challenged to form a cross-party alliance to deliver Brexit since every other party had rejected his Brexit deal. Nor were other parties inclined to back a minority Labour government under Jeremy Corbyn to resolve the Brexit deadlock. Without a government in place Britain was scheduled to leave the EU without a deal on 31 January 2020.

218

R. ROSE

The first-past-the-post electoral system disposed of the threat of a hung Parliament; it turned the Conservatives’ plurality of 43.6% of the popular vote into 56.1% of seats in the House of Commons (Table 12.1). Winning 365 seats gave the Conservatives a majority of 80 seats over opposition parties, which were divided on many issues. This was the Conservatives’ biggest parliamentary majority since 1987. Since Labour won only 202 seats, the Conservative government now enjoys a 163-seat lead over the official opposition. Voting at the constituency and regional levels tended to follow a division between places where a majority had voted to leave or remain in the EU in the 2016 referendum. Thus, the Conservative party did worst in the two strongholds of Remain voters. In Scotland, it lost 3.1% of its previous share of the vote and seven seats, and in London it lost 1.1% of its previous vote and one seat to Labour. The party did best in the industrial Midlands; its vote went up 4.2% and it gained 16 seats. In the North Table 12.1 Votes and seats of pro- and anti-Brexit parties

Seats

Pro-Brexit Conservatives Brexit Party Democratic Unionist Total Ambivalent Labour Others Speaker Total Pro-EU Liberal Democrats Scottish National N. Ireland parties Plaid Cymru Green Party Total

Votes

N

%

%

365 0 8

56.1 0 1.2

43.6 2.1 (with UKIP) 0.8

373

57.3

46.5%

202 0 1 203

31.1 0 0 21.1

32.1 1.3 0.0 32.1%

11 48 10 4 1 74

1.7 7.4 1.5 0.6 0.2 11.4

11.5 3.8 1.5 0.5 2.8 20.1%

Source Author’s calculations from C. Baker, R. Cracknell and E. Uberoi General Election 2019: Results and Analysis. London House of Commons Briefing Paper CBP 8749

12

A NEW PRIME MINISTER MEETS OLD CONSTRAINTS

219

East, helped by the Brexit Party taking most of its votes from Labour supporters, the Conservatives won seven seats that had not gone Tory for generations or ever, including Tony Blair’s old constituency. As in the 2016 British referendum, these regional differences showed that while London had a more cosmopolitan and prosperous electorate, the North of England had more votes and MPs (see Chapter 5). Collectively, the opposition’s share of the vote was 12 percentage points greater than that of the Conservatives, but its vote was split (Table 12.1). The Labour party’s claim to be the only party capable of displacing Boris Johnson from Downing Street was offset by being ambivalent about whether it would deliver a softer Brexit or a referendum that would keep Britain in the EU. Jeremy Corbyn’s policy was to negotiate a soft Brexit but to remain neutral and let voters decide whether to accept it or vote to remain in the EU after all. The seven parties unambiguously committed to the EU—the Scottish Nationalists, the Liberal Democrats, Plaid Cymru, the Greens and three Northern Ireland parties—won only 22% of the vote and 74 seats. While the Brexit Party had won more than three times the Conservative vote in the EP election seven months earlier, Johnson’s ‘get Brexit done’ policy was critical to the Conservative party’s rapid recovery of the Brexit vote (see Table 5.3). Because the December election was about who governs, the Conservatives were the only choice for people who had voted to leave the EU in 2016 and wanted to get Brexit done. Among that group, the Conservatives gained 73% of the vote, and only 4% went to the Brexit Party (Table 12.2).

Table 12.2 Leave and Remain voters’ party choice

2016 referendum vote 2019 party vote

Leave %

Remain %

Conservative Brexit Labour Liberal Democrat SNP Green Other

73 4 16 3 2 2 1

20 0 47 21 6 4 1

Source Election day survey of more than 13,000 voters. Lord Ashcroft, How Britain Voted and Why (lordashcroftpolls.com/ category/elections)

220

R. ROSE

The Remain vote fragmented among parties offering a variety of alternatives for dealing with the referendum vote for Brexit, and no party received a majority of the Remain vote. Labour’s policy of renegotiating a withdrawal with Brussels and then calling a second referendum that might reverse Brexit gained it less than half the vote of the Remain supporters. If it had been able to get as big a share of the Remain vote as the Conservatives did of the Leave vote, this would have deprived Boris Johnson of a majority and created a hung parliament. Boris Johnson’s strategic gamble has given him a majority of Conservative MPs committed to getting Brexit done on his terms. Neither David Cameron, John Major nor Edward Heath had ever managed to command an unquestioning majority of Tories in favour of Britain’s participation in the European Union. Johnson’s expulsion of 21 MPs who voted against no deal removed Conservative MPs favouring a soft or no Brexit. None of the rebels who sought re-election as independents or Liberal Democrats gained a seat in the new House, a cautionary warning to re-elected Conservative MPs. The 109 new Conservative MPs were committed followers. Johnson had taken the unusual precaution of having every candidate pledge to support a Downing Street deal with Brussels. Their lack of parliamentary experience makes them particularly dependent on advice from Johnson’s whips. The prime minister’s power to make appointments to more than 100 government posts offers an incentive not to question the government’s EU policy.

12.3

Getting Brexit Done--Up to a Point

Boris Johnson moved quickly to reward the Conservatives who made him party leader and newly elected MPs and voters who gave him a substantial parliamentary majority. In the first week of the new Parliament, he introduced a revised withdrawal bill similar to what the previous Parliament had rejected eight weeks before. Conciliatory language was removed giving MPs more say over the Brexit process. A new clause was added fixing 31 December 2020 as the deadline for ending the transition period for negotiating future British–EU relations. The new House of Commons approved the bill. Johnson declared in presenting the withdrawal bill that its passage would mean that the sorry story of the last three years was at an end. Brexit will be done. To symbolise this, Downing Street banned the use of the word Brexit in official communications and the Department

12

A NEW PRIME MINISTER MEETS OLD CONSTRAINTS

221

for Exiting the European Union was abolished. Nonetheless, a Taskforce Europe was established in the Cabinet Office to negotiate a Political Agreement with Brussels. Its head, the experienced diplomat David Frost, has the confidence of the prime minister. However, public opinion is sceptical about Johnson’s claim that Brexit is finished. A YouGov survey a month before the official departure date found that 46% thought politicians would need to spend more time focusing on Brexit, and another 23% thought the amount of attention needed would be the same. Leaving the EU Does Not Get Brexit Done The formal departure of the UK from the European Union is the end of the beginning. It also marks the start of negotiations for a Political Agreement between the UK and the EU that covers what the withdrawal bill leaves out about Britain’s future relationship with the European Union. Paradoxically, withdrawal from the EU gives the UK the freedom to negotiate new relations with the rest of the world while simultaneously requiring it reach a mutual agreement with Brussels governing its new relationship with the EU. As French President Emmanuel Macron commented on the day Britain left the EU, ‘You may be leaving the EU but you are not leaving Europe’. EU negotiators have had years of experience of dealing with Britain as a member state, while the public statements of David Cameron, Theresa May and Boris Johnson have shown ignorance of how the EU works and ignored informed advice from British officials about how a national government can influence the EU’s multi-national policy-making process (Rogers 2019). The disparity in knowledge is most extreme in trade matters, since Whitehall has had a minimum of trade officials since responsibility for trade was transferred to Brussels when the UK joined the EU in 1973. British public opinion agrees with EU negotiators in the evaluation of Brexit negotiations. Whether the prime minister was Boris Johnson or Theresa May, in the year leading up to the UK’s departure from the EU, from 70 to 87% of YouGov respondents thought the government was handling Brexit badly. The symbolism of national sovereignty has been an end in itself for Brexit campaigners. It gives Westminster the power to diverge from EU laws and regulations that affect national policies on trade, goods and services, working conditions, aid to industry and immigration. The UK regains the freedom to make new trade deals with the United States,

222

R. ROSE

China and other countries to replace existing trade agreements that the EU has negotiated with these countries on behalf of Britain when it was an EU member state. However, new trade agreements can only be arrived at by the UK compromising some of its conditions in exchange for the benefits the agreement should bring. The future terms of UK-EU trade are central to the politics of the Political Agreement. For the Conservative government, maintaining access to EU markets for manufactured goods produced the North of England, where the party gained many MPs from Labour in 2019, is a major priority. Maintaining EU access for the financial services of the City of London is important because of its disproportionate contribution to the country’s gross domestic product and revenue. Five continental countries want to maintain their access to offshore UK fishing rights. British exports to the EU account for almost half of the UK’s total exports, whereas EU exports to Britain account for a small share of the EU’s global exports. Brexit is almost certain to reduce trade. Negotiations will set the rules that determine how large or small that may be. The Treasury has estimated the Conservative government’s goal of restricting an agreement to avoid tariffs and quotas on goods could lower economic growth about 5% in 15 years, that is, about one-quarter of one percent per annum (cf. Chapter 9). The EU wants future trade negotiations with Britain to maintain a level playing field, that is, British goods and services should meet the same regulatory standards that the UK met when it was an EU member state. Since EU regulations are currently evolving in such fields such as environmental protection and data handling, British businesses should comply with future regulations too. To ensure compliance, the British government would need to agree to the EU regulations being judicially enforced. From a Westminster point of view, aligning the British economy with EU regulations and accepting European courts would undermine the purpose of Brexit: taking back control of British laws. The Johnson government wants the EU to accept the mutual equivalence of regulations that would lead to same outcome even if they are not identical in content. This would allow gradual divergence between Britain and EU economies in ways that Westminster believes would increase British competitiveness and that the EU fears would do so by undercutting EU standards. It also wants to avoid giving the EU an effective means of enforcing equivalence. This would allow the EU to

12

A NEW PRIME MINISTER MEETS OLD CONSTRAINTS

223

make a unilateral judgement that British regulations were not equivalent to EU standards, thereby reducing EU imports from Britain. In order to prevent the departure of the UK from harming integration, the EU does not want to allow a post-Brexit Britain to ‘cherry-pick’, that is, enjoy the same relationship as it had as a member state while being free of obligations. A senior EU policy-maker has described the British position as: ‘Before they were in the EU with lots of opt-outs; now they are out and want a lot of opt-ins’ (quoted in Fleming et al. 2019). The president of the European Commission, Ursula von der Leyen (2019) told the European Parliament, ‘It’s the choice of the UK how far they want to align with the EU or diverge’. Boris Johnson shortened the time to complete a Political Agreement that regulates what comes after Brexit by making another do-or-die pledge to end the process by 31 December 2020. If this deadline is not extended and no agreement can be reached, Brexit ends with no deal, thereby maximising the effect of breaking links developed in the 47 years in which the UK was a member of the European Union. In accepting the British deadline, EU officials have cautioned that it makes likely a ‘barebones’ Political Agreement that minimises the extent to which benefits can be secured. The coronavirus crisis, which has frustrated negotiations, gives him reason to seek an extension. Given Johnson’s Houdini-like skill in dealing with promises, he could alter his position to take into account Brussels’ red lines. Alternatively, to meet his self-imposed deadline he could emulate Candide and tell Parliament that whatever agreement he reaches with Brussels will deliver the best of all possible Brexits in the best of all possible worlds. A third option would be to invoke the spirit of Dunkirk and claim that no deal is better than surrender to the demands of a united European Union.

References Ashcroft, Lord. 2019. My Final Election Dashboard. London: Lord Ashcroft Polls. Survey of 4,046 Respondents, December 10. Conservative Party. 2019. Get Brexit Done: Unleash Britain’s Potential. London: Conservative & Unionist Party Manifesto. Fleming, S., J. Brunsden, and G. Parker. 2019. Johnson Comes Under Fire from EU for Dearth of Brexit Detail. Financial Times, September 17. Hammond, P. 2019. I No Longer Recognise This Party of Radicals. The Times, September 28.

224

R. ROSE

Mudde, C., and C.R. Kaltwasser. 2017. Populism: A Very Short Introduction. Oxford: Oxford University Press. Rogers, Sir I. 2019. 9 Lessons in Brexit. London: Short Books. Rose, R., and B. Wessels. 2019. Money, Sex and Broken Promises: Politicians Bad Behaviour Encourages Distrust. Parliamentary Affairs 72 (3): 481–500. Sumption, Lord. 2019. Constitutional Vandalism. The Times, September 25. Sylvester, R. 2019. Tory Contest Has Become a Brexit Bidding War. The Times, July 2. Von der Leyen, U. 2019. A Union that Strives for More: My Agenda for Europe. Brussels: European Commission. Walker, N. 2019. Brexit Timeline: Events Leading to the UK’s Exit from the European Union. London: House of Commons Research Briefing 7960.

CHAPTER 13

Beyond Brexit in a World of Interdependence Richard Rose

The bounds that interdependence places on the UK are not the result of democratic choice. They are the unintended consequence of decisions taken by generations of Britain’s governors as they adapted to changes in the world since Queen Victoria’s time. By taking back powers it had ceded to the European Union, the British Parliament can claim it has regained sovereignty in the Victorian sense of the term. But the government’s experience in negotiating Brexit illustrates that constraints of interdependence have hollowed out the meaning of national sovereignty. Moreover, the British government faces fresh constraints as it seeks to build new ties to replace those with Europe that are reduced by Brexit. The European Union was founded by heads of national government to create bounds of interdependence that would prevent their governments from conducting military and trade wars with each other. National parliaments have ratified a series of treaties that grant the European Union an increasing range of powers that constrain their own policy-making capacity. As EU policies concerned with climate change, digital technologies and immigration become more important, the EU is increasingly bound by global interdependence.

© The Author(s) 2020 R. Rose, How Referendums Challenge European Democracy, Palgrave Studies in European Union Politics, https://doi.org/10.1007/978-3-030-44117-3_13

225

226

R. ROSE

13.1 A Global Britain Is Still a Bounded Democracy Globalisation presents similar challenges to Britain and the European Union, but Britain’s international position is distinctive. It is a member of both the IMF and the World Bank and has the resources to make its presence felt in both institutions. Unlike major EU economies, it also has its own international currency and, whatever Brexit’s impact on London as an international financial centre, there is no EU city that competes with it. The major source of British immigrants is not other EU member states or Turkey, but New Commonwealth countries, especially from the Indian sub-continent. When Britons emigrate, they are much more likely to go to Australia, Canada or the United States than to Europe. Britain’s security is managed differently because it can command the military force that the EU lacks and it has a trusting relationship with American intelligence services. When the transition process ends, Britain will be beyond Brexit. In the words of strong advocates of Brexit such as Jacob Rees-Mogg, it will no longer be a vassal of the European Union. With its national sovereignty restored, the British government can seek economic and political agreements that will define its role in a world of global interdependencies. However, achieving the status of an ex-EU member state does not in itself make the UK a global power. It forces Britain to find a role in a world that has become less British-oriented than when former American Secretary of State Dean Acheson noted in 1962 that its standing among the world’s leading powers was about played out (see Chapter 8). After leaving the EU, Britain retains soft power relevant to a global role. English is now the global language of communication, and learning English gives pupils a familiarity with a variety of English institutions and customs. English rather than American is the preferred foreign language of Europeans and the lingua franca of the European Union too. The BBC and elite print media have an international audience. The Commonwealth gives Britain long-established links with countries on every continent. Winning a parliamentary majority has changed the political arithmetic for getting Brexit done, but it has not disposed of the legacy of conflicts that Brexit has created or exacerbated. Brexit has battered the political system’s reputation for good government. Successive prime ministers have demonstrated ignorance and incompetence in dealing with Brussels and

13

BEYOND BREXIT IN A WORLD OF INTERDEPENDENCE

227

refused to make use of knowledgeable civil servants. Government leaders have attacked the courts for making judgements that they describe as political. Parliamentary debate and social media communications to MPs have taken a toxic turn, including the need for protection against violent attack. Identities rather than issues have been emphasised: Remainers are called traitors and Leavers are called racists and stupid (Rutter 2020). Public opinion has been soured too. The 72% who think the British political system needs a great deal or quite a lot of improvement is the highest in the Hansard Society’s 15-year series of Audits of British politics (2019). For the first time since the economic crisis of the 1970s, more than half the public is dissatisfied with the state of democracy in Britain (Foa et al. 2020: 21). Many problems that have dogged the British government for years have not been caused by EU membership nor will they be resolved by Brexit. These include the limited skill of many workers, slow growth in productivity, limited investment and the high cost of housing. Cultural problems arising from immigration have been less due to immigration by Europeans; the greater cause has been the clash between British values and the non-Western values that some New Commonwealth immigrants and refugees hold. If Britain is to attain the global role that Brexiters have promised, there is an unprecedented need to strengthen ties with governments from Washington to Beijing. None of these goals can be unilaterally achieved by Westminster; each depends as well on what other countries choose to decide. The promotion of greater interdependence with countries on other continents places fresh bounds on Britain’s national democracy. Conflicting Readings of an Unwritten Constitution The Brexit referendum has raised the stark question: Where does legitimacy lie—with the people or with Parliament? This is a novel issue because referendums have been exceptional events in British politics, and previous ballots have supported the position of MPs and Parliament. Politicians reason backwards when defining the principle of legitimacy. Those in favour of Brexit have argued that since democracy means rule by the people the decision of a majority of referendum voters must be implemented without question. Politicians most strongly in favour of EU membership have questioned whether the popular vote for Brexit should be accepted and, if so, how it should be implemented. Those who wanted

228

R. ROSE

a ‘think again’ second referendum on EU membership tacitly acknowledged the superior legitimacy of a referendum compared to a vote of MPs. The debate on the future of the UK as a multi-national state raises a second question of legitimacy: Which people and which Parliament have the right to decide whether one part can leave the multi-national UK of Great Britain and Northern Ireland? The Westminster position is inconsistent. Before the 2014 Scottish referendum on independence was called by the Scottish Parliament, David Cameron’s government accepted that if a Scottish referendum endorsed independence the Westminster Parliament would accept the result. Since 55% of voters favoured remaining in the UK, no conflict of legitimacy arose. There is now a conflict of legitimacy between Westminster and the Scottish National Party government about calling a second referendum on independence. The notional cause for thinking again is that the English majority for Brexit in the 2016 UK referendum contradicted the 62% Scottish majority for remaining in the European Union. Only by leaving the UK can Scotland become an independent country eligible for admission to the EU. The Scottish Parliament in Edinburgh has passed an Act authorising a second referendum, while Boris Johnson has refused to authorise an independence vote. The SNP government can challenge the action of the prime minister in the courts. If it loses there, it will emulate Johnson’s strategy by turning the May 2021 election of a Scottish Parliament into a referendum on Scottish independence. The UK government has long recognised the right of a Northern Ireland referendum to decide whether it should remain in the UK or become part of the Republic of Ireland. In 1973 the UK government held such a ballot, knowing that it would endorse the status quo because the majority of the electorate was Unionist. Irish Republicans boycotted the ballot, deeming it illegitimate because Irish unity was an absolute right. The Northern Ireland Act of 1998 authorises the British government to call a referendum on Irish unification as and when it appeared likely that a majority of voters would endorse unification. The Act also commits the UK government to implement a referendum majority for leaving the UK on terms agreed between it and the Republic of Ireland.

13

BEYOND BREXIT IN A WORLD OF INTERDEPENDENCE

229

What Role in the World for a Post-Brexit Britain? As a country giving priority to sea-power, Britain has had global political interests for more than five centuries. Since the end of the Second World War, Britain has been searching for a new role to replace its late nineteenth-century pre-eminence. The leading doctrine, formulated by Winston Churchill and accepted by successive prime ministers including Tony Blair and Boris Johnson, places Britain in three circles of influence: the Commonwealth, a special relationship with the United States, and Europe (see Chapter 8; Westlake 2019). Because international influence is relative, Britain’s role in the world has declined as new economic and military powers have emerged. In withdrawing from the EU, the British government hopes to become an independent global power punching above its weight. However, weight is relative. While Britain is a big European country in population, globally it is a medium-sized country ranking twenty-second in population between Thailand and South Africa. In Asia, Britain cannot claim to be a major power compared to China, Japan or the United States. Former colonies such as Hong Kong and Singapore do not look to London to advance their international status. British political and military advisers to Middle Eastern rulers have disappeared and oil-rich Middle Eastern states now exercise military and economic power in their own right. What was once the British Empire and then the British Commonwealth has become the Commonwealth of Nations. The special relationship with the United States places bounds on Britain’s independent use of military power. President Dwight D. Eisenhower forced a halt to the Anglo-French invasion of Suez, and Margaret Thatcher was not consulted by Ronald Reagan before the United States sent troops to change the regime in Grenada, a Commonwealth country. Tony Blair’s attempt to re-establish Britain’s influence in Washington crashed after he made Britain a junior partner in the American-led invasion of Iraq. Barack Obama’s tilt towards Asia diminished the importance not only of Britain in Washington but also that of the whole of Europe. Since Donald Trump adopted an unpredictable America First policy, Boris Johnson has become the object of tweets from Trump. Britain increased its influence in Europe by joining the European Union. Within the EU it was a major power along with France and Germany and could sometimes hold the balance of power. Margaret

230

R. ROSE

Thatcher successfully pushed for the creation of a Single Europe Market and won a substantial rebate on Britain’s contribution to EU finances because it received less benefit from major EU subsidies. When continental countries with a social market philosophy started expanding EU regulation of the Single Market, Thatcher began the process of British resistance to further European integration. The government used its influence to put the brake on laws and regulations increasing European integration and opted out from participating in the Eurozone. Brexiters have promised great benefits from Britain leaving the EU, but the British public has not been convinced. YouGov polls in the years following the 2016 referendum have consistently found British people sceptical about the putative benefits of Brexit. When Boris Johnson became prime minister in July 2019, 40% thought leaving was the right thing to do and 47% thought it wrong. A week before Britain left the following January, public opinion was exactly the same. By a substantial margin more people think that leaving the EU will make the country worse off rather than better off economically and the country’s influence in the world will be less. In both instances, there is a large group in the middle who think that leaving the EU will make no difference to Britain in future or don’t know what its effect is likely to be (https://yougov.co. uk/topics/political-trackers/survey-results). If the perceived benefits of leaving the European Union turn out to be a mirage while the costs are large and visible, the British government could apply to again be a member state. Before this could happen a major British party would have to make joining the EU its priority. To apply for EU membership, the party would have to win enough MPs to have a majority secure against anti-EU rebels. By this time the EU would no longer be the institution that it is today. Free of British opposition, the EU is likely to continue to deepen and broaden its policies’ impact on member states. Its future attractiveness will also depend on how well or badly it responds to current problems of interdependence affecting the Eurozone, immigration and unexpected events. Although the European Union has a procedure for leaving the EU, it has no procedure for a country to rejoin. Should such a request come from London it would apply the procedure for evaluating an application for accession set out in Article 49 of the Treaty on European Union. This is a process in which the EU ensures that an applicant country will accept all the existing obligations of member states as a condition of receiving its benefits. Thus, it is moot whether the UK could regain opt-outs that

13

BEYOND BREXIT IN A WORLD OF INTERDEPENDENCE

231

it gave up on withdrawal such as not being a member of the Eurozone. Applicant countries are also expected to implement EU laws and directives. This would require changing many British Acts of Parliament to revert to the situation before Britain removed many references to EU directives from British Acts of Parliament. It would also require giving up post-Brexit policies that diverge from EU standards. As a relatively prosperous economy, Britain would have to make a substantial financial contribution to the EU’s annual budget. At any stage in the process, a single member state could veto a British application. After taking several years to arrive at an agreement with the EU about membership, a future British government would have to get the approval of Parliament. In addition, on past precedents a referendum would be required about whether the UK should join the EU or remain a nonmember state. Altogether, the process of seeking EU membership would take a Parliament or two after the impact of Brexit became evident. Boris Johnson and his government have immediately launched a voyage to learn what Britain’s role in the world can be. Johnson has shown skill in selling the benefits of a Global Britain cruise but losers in the referendum have sketched maps showing the world beyond Brexit as full of dangerous obstacles. The prime minister’s desire to get Brexit done speedily has not allowed time for him to chart a course. The British people now find themselves in the position they were in almost half a century ago. When Britain was about to join what was then the European Economic Community, the BBC’s Reith lecturer, Andrew Shonfield (1973), described this as A Journey to an Unknown Destination.

13.2

The EU Is Bounded by Interdependence Too

Britain’s withdrawal from the European Union has been welcomed in Washington by President Donald Trump, but there is no sign of any other EU member state wanting to emulate the UK. Whereas the British government acted in the belief that withdrawal would get rid of the bounds of interdependence, the EU was founded to impose such bounds. Even though member states disagree about many specific EU policies, they see it as in their national interest to be bound by treaties to a Union that can make policies collectively and speak for common European interests in an era of global economic interdependence.

232

R. ROSE

Multi-national Treaties Place Bounds on the Legitimacy of National Elections Because it consists of states very unequal in size, the EU cannot claim that its actions are legitimated by democratic elections as is the case in member states (see Chapter 3). Instead, the EU’s legitimacy rests on legal powers conferred by treaties unanimously agreed by member states. In the half-century since the signing of the Treaty of Rome in 1957, successive treaties have broadened and deepened the EU’s policy-making powers. Unlike many international organisations, the EU has the effective authority to make policies that bind national governments and enforce them through EU courts. Contrasting forms of legitimacy create the potential for conflicts between member states and EU institutions. This is most evident when a national referendum gives citizens a vote on an issue of EU significance. If a referendum confers democratic legitimacy on an EU policy with legal legitimacy, there is no conflict. If national citizens reject EU membership, as has happened in Norway twice, that is not an EU problem but a problem of the national government on the losing side. When national referendums have rejected an EU policy, the EU has sought to de-escalate conflict by tactics ranging from accepting the verdict or making minor concessions to pursuing its aim by other means (see Chapter 3). National governments have occasionally used the legitimacy of being popularly elected to adopt policies that EU authorities have regarded as inconsistent with the broad but vague democratic values set out in EU treaties. In 2015, a left-wing Greek government faced a domestic financial crisis that threatened the stability of the Eurozone. To support its refusal to accept bail-out conditions set by EU institutions, the Greek government called and won a referendum. After discussion in both Athens and Berlin of Greece leaving the Eurozone with uncertain consequences for all involved, the conflict was avoided by compromise on both sides. Under Article 7 of the Treaty of the European Union its institutions have the power to suspend a member state if its nationally elected government has acted in ways deemed inconsistent with European values. Its use has been considered but not invoked against government actions in Austria, France and Romania. The European Court of Justice has issued an opinion describing actions of the Polish Law and Justice government as going against European values, and the European Parliament has also expressed disapproval of Hungarian laws promoted by Viktor Orban in

13

BEYOND BREXIT IN A WORLD OF INTERDEPENDENCE

233

the name of ‘illiberal democracy’. Effective action to enforce EU values against a nationally elected government requires unanimous support from all member states except the accused, and that has always been lacking. Moreover, EU treaties do not give its institution the power to expel a member state. Small states welcome the EU’s bounds because they tie the hands of all member states, large as well as small. The great majority of EU members are small states that see the EU as increasing their capacity to advance their interests. Instead of being bound to accept decisions taken by a big and powerful neighbour without consultation, each state has the right to participate in the EU’s policy-making institutions and use the EU’s complex decision-making rules to build multi-national coalitions to promote policies that will incorporate specific measures in their national interest. For example, the Irish government has advanced the unification of the island of Ireland by getting the EU to include in the Brexit withdrawal agreement special provisions for Northern Ireland that the British government was loath to accept. Given their histories, big states accept EU bounds too. The European Coal and Steel Community was founded to place bounds on the national use of essential war materials. Three large member states— Germany, France and Italy—had previously relied on large national armies to advance their interests while three small member states were militarily defenceless. All six saw a European institution as offering better security against war than had national armaments. Even though Germany has a bigger population and economy than Britain, its leaders accept the bounds of interdependence. In the words of Chancellor Angela Merkel, ‘I see the European Union as our life insurance. Germany is far too small to exert geopolitical influence on its own’ (quoted in Barber and Chazan 2020). Globalisation Places Bounds on the Exercise of EU Powers The European Union’s powers are categorically different when it moves from making policies for its member states to global policy-making. Globalisation affects the EU because many major policies are influenced not only by what happens within Europe but also by what happens on other continents, because the EU is part of a global network of interacting institutions and processes (Rose 2014). Major policy priorities of the current president of the European Commission, Ursula von der Leyen (2019), involve interdependence

234

R. ROSE

with countries outside the authority of Brussels. These policies include climate change, new digital technologies, the international use of the euro, protecting the European way of life with strong borders and a fresh start on migration. The value of the euro is affected by its foreign exchange rate with the dollar and the Chinese renminbi. The extent of immigration into the EU depends on the push to flee from poor and strife-torn states as well on EU immigration policy. The size of the EU is relative to the states with which it interacts and the policy at issue (Rose 2015: 139ff.). As a league of 27 countries, EU institutions necessarily exercise authority over far more people than any one European state. Collectively, European citizens are up to six times the number of citizens of Germany, France or Italy. In economic terms, the EU’s gross domestic product and the Eurozone make it a major participant in the global economy. However, the Single Europe Market is only a fraction of the global market for goods and services, and the euro is not as important a global currency as the dollar. In military terms, European forces are dwarfed by comparison with the United States and China. Within the broadly defined continent of Europe, the EU is not the only intergovernmental institution concerned with major policies. NATO has its European headquarters in Brussels but is controlled from Washington. Europe is subject to trans-national terrorism and trans-national criminal activity addressed by institutions such as the Organization for Security and Co-operation in Europe and Interpol, whose member states include countries exporting crime and terrorism. Legal and illegal refugees and immigrants use Turkey and Libya as major entry points for gaining freedom of movement within the EU. The EU offers these countries money to finance activities to staunch the flow of migrants, but the outcomes are not determined by EU directives. The EU looms large in relation to states in its near abroad, a fuzzy term that describes countries in Eastern Europe and the Mediterranean that are historically and geographically linked with EU member states (Christiansen et al. 2000; Mahncke and Gstöhl 2008). Porous borders between EU states and countries in the near abroad create interdependence in dealing with smuggling, illegal immigration and trans-national crime. Six small Balkan countries are challenging the EU by seeking membership or closer association: Albania, Bosnia-Herzegovina, Kosovo, Montenegro, North Macedonia and Serbia. Turkey’s geographical location makes it a buffer state for refugees since it borders three EU member states as well as Syria, Iraq and Iran. It has a larger military force than any

13

BEYOND BREXIT IN A WORLD OF INTERDEPENDENCE

235

EU member and a readiness to use force. The current territory of Ukraine has historical ties with EU member states as well as Russia. Moscow sees Ukraine as part of its own near abroad and has used military action to seize parts of Ukraine’s territory to discourage EU ties. The EU is prepared to admit any European country that meets its political, economic and administrative standards and that it feels capable of integrating. None of the eight countries in the near abroad currently meets these standards nor is EU financial aid capable of resolving severe domestic problems. Moreover, Turkey and Ukraine must balance EU links with interdependence with the Middle East and Russia. While the EU officially welcomes applications from countries in its near abroad, their internal problems of governance, ethnic tensions and violence have resulted in the EU doubting its own capacity to absorb them. The EU’s ability to influence global interdependence is greatest in international trade. EU treaties give it the power to regulate trade between member states, to represent all member states in the World Trade Organization and to make bilateral agreements with other countries. It can thus set standards that other countries must meet to gain access to the Single Europe Market (Bradford 2020). It is the dominant partner in trade deals with Norway and Canada and an equal in dealing with the United States. It can also impose sanctions on major multi-national companies. The EU has fined Google e8.2bn and Microsoft more than e1.4bn for breaching EU competition rules and similarly fined major European manufacturers. Because the euro is a major global currency, the monetary policy of the European Central Bank is of major significance in global financial markets. Some high-priority concerns of the EU, such as global pollution or the flight of refugees, are not controlled by any international institution. In many important international organisations European states are members, and national governments of the biggest states can play a prominent role, while the European Union only has observer status. At the United Nations, two European countries, France and the UK, are among the five permanent members of the Security Council. The great majority of European countries are members of the International Monetary Fund but not the EU. The IMF monitors the economies of European countries and has made conditional loans to four Eurozone states and the World Bank has financed projects in four EU member states. Both institutions maintain offices in Brussels to liaise with EU institutions about common interests, but the EU is not a member of either. NATO is led by

236

R. ROSE

the United States, and two non-EU countries, the UK and Turkey, have had major roles within NATO. The EU is not a member nor does it have a substantial military force of its own. Globalisation has placed bounds on the authority that national governments have gained from being democratically elected and on the authority the EU has gained from treaties unanimously approved by democratic national governments. Both the UK and the EU have been caught up in networks of interdependence that they can influence but not control. The post-Brexit UK is unusual in Europe in the belief that by acting alone it can best advance its interests in a world of interdependence. By contrast, 27 nationally elected European governments seek to advance their interests within the bounds of EU institutions. They do so in the belief that collective action, for all its complications, is the better way to achieve influence in a world of interdependence.

References Barber, L., and G. Chazan. 2020. FT Interview: Angela Merkel. Financial Times, 16 January. Bradford, A. 2020. The Brussels Effect: How the European Union Rules the World. Oxford: Oxford University Press. Christiansen, T., F. Petito, and B. Tonra. 2000. Fuzzy Politics Around Fuzzy Borders: The European Union’s “Near Abroad”. Co-operation and Conflict 35 (4): 389–415. Foa, R.S., A. Claystone, M. Slade, A. Rand, and R. Williams. 2020. The Global Satisfaction with Democracy Report 2020. Cambridge: Centre for the Future of Democracy. Hansard Society. 2019. Audit of Political Engagement 16. London: Hansard Society. Mahncke, D., and S. Gstöhl (eds.). 2008. Europe’s Near Abroad. Bern, Switzerland: Peter Lang. Rose, R. 2014. Responsible Party Government in a World of Interdependence. West European Politics 37 (2): 253–269. Rose, R. 2015. Representing Europeans: A Pragmatic Approach, 2nd ed. Oxford: Oxford University Press. Rutter, J. 2020. How Brexit Has Battered Britain’s Reputation for Good Government. The Guardian, 7 January. Shonfield, Andrew. 1973. Europe: Journey to an Unknown Destination. London: Allen Lane. von der Leyen, U. 2019. A Union that Strives for More: My Agenda for Europe. Brussels: European Commission. Westlake, M. 2019. Slipping Loose: The UK’s Long Drift Away from the European Union. Newcastle upon Tyne: Agenda Publishing.