The Palgrave Handbook of European Referendums

Are justifications for democracy more likely to be delivered with referendums or without them? Are referendums more likely to satisfy some democratic standards than others? Why do people disagree so much about the desirability of referendums? This chapter discusses what referendums may mean for aggregative, deliberative and epistemic aspects of democracy, as well as concerns that referendums risk manipulation, political irresponsibility and majority domination. Yet, the chapter also doubts that any democratic system can manage without some provision for referendums. Securing the benefits of referendums without too many of their shortcomings then requires embedding them in the very practices of democratic constitutionalism and representative democracy that are often understood as opposites to referendums.

This chapter locates referendums within a history of European democracy, understood in terms of Paul Nolte’s three Leitmotifs: search, fulfilment and crisis. The idea of the referendum is first presented as a search for something to complement representative democracy. The second part of the chapter explores how to fulfil direct democracy’s promise, a dilemma that revolves around considerations of how effectively referendums live up to their promise when used. Thirdly, and finally, the analysis looks at the history of the memory, or legacy, of referendums. Judging by the historical record, there is no apparent reason to abandon direct democracy for fear of its consequences: crises provoked by referendums can spur the democratic search for fulfilling the potential of citizen empowerment.

Referendums, like any large-scale processes of democratic decision-making, need to take place within a framework of rules. This chapter examines such rules. It surveys how referendums are regulated in Europe today, focusing on five key dimensions: when referendums take place; how the vote is structured; the status of the result; the conduct of the campaign; and how the rules are enforced. It also assesses how referendums ought to be regulated. It finds that there are many important gaps in current regulatory frameworks: much more could be done to ensure that referendums play the most positive role they could in the overall democratic system.

Some nations provide that the decision to hold a referendum rests with an authority other than the legislature. Where legislatures have the power to provide for referendums, it has tended to be least employed in parliamentary systems, especially Westminster-type systems, and in northern Europe, broadly defined, compared to southern Europe. The distinctions are becoming less pronounced, with more parliaments legislating for referendums. Four types of relationship of legislatures to referendums are identified: subordination, non-decision making, rejection and utilisation. In recent years, there has been a move to greater utilisation of referendums to resolve contested issues, especially constitutional issues, not resolvable through traditional party conflict. Enthusiasm for referendums may be diminishing, with governments that have sought them not necessarily achieving the results they expected. Although governments usually take the initiative in proposing referendums, legislatures may come to the fore at a second-order level in determining the rules for holding them.

Referendums on European integration present a puzzle for academics, policymakers and the general public. In the recent past, some of these referendums witnessed a considerable shift in public opinion, going from a favourable attitude to outright rejection. On several occasions, voters changed their verdict entirely in a second referendum on the same topic. What is more, EU referendums have a reputation for being more advantageous for the anti-EU side. The answer to these puzzling aspects lies in the peculiar dynamics of referendum politics and the power of certain kinds of political arguments. The analysis in this chapter relies on extensive field research in France, Spain, the Netherlands and Luxembourg in 2008 on the European Constitution referendums, in Denmark and Ireland in 2011 on the Nice, Maastricht and Lisbon Treaty referendums, in Ireland again in 2015 on the Fiscal Compact referendum and in the UK in 2016–2017 on the Brexit referendum.

Since the first nationwide referendum was held in France, on a new constitution in 1793, literally hundreds of issues gave been submitted to the voters. The chapter analyses when and why provisions for referendums have been introduced, when they have been held, won or lost—and what impact they have on public policy and political culture. It is concluded that the referendum mainly has been used to ratify constitutional changes. The referendum is aimed at being a constitutional safeguard and has generally performed this function well.

The aim of this chapter is to reflect on the role referendums might play when it comes to national and European legitimacy and decision-making processes, by looking at these issues through the unusual prism of the only referendum ever held in Belgium, on King’s Leopold III right to rule in 1950 (the question royale). Belgium is founded upon a consensual political system which allows, with some caveats, to draw parallels with the EU, a polity that is held together by the consensus built among the Member States. The way in which the question royale played out in an ethnically divided Belgium seems to offer some insights into the polarising effects of EU-related referendums. Moreover, the unconstitutional nature of the Belgian 1950 referendum allows to highlight what has been referred to as the legitimacy clash between exercising direct democracy at a Member State level and the Union’s consensual decision-making processes.

The chapter starts with a sketch out of the referendum history in France since the Revolution. While it seemed destined for an ambitious future, the device has raised many fears and its Bonapartist use led to its rejection under the Third and Fourth Republics. It was not until the Fifth Republic that the referendum regained in importance in a top-down presidential version. But the evolution of the regime led to a loss of utility and efficiency causing its relative decline after de Gaulle. The chapter focuses on the three referendums on Europe held since 1958 (accession of new countries, Maastricht treaty and Constitutional treaty), with particular attention to the reasons for the decision to organise them, campaign dynamics, electoral behaviour and the consequences on party divisions. The last two sections review the present provisions and legal framework of nationwide referendums and the current debate on the device in French politics.

In this chapter, the reader will gain a concise and practice-oriented overview of the Swiss referendum history, use, and functionality. Over the centuries referendums turned into the dominant characteristic of the Swiss political system. Political parties had to learn how to live with them are using them as tools for power maintenance as well. Direct democratic mechanisms such as the initiative and the optional referendum but also less known instruments such as the finance or tax referendum are exemplified. Detailed long term referendum frequencies demonstrate how dominant ballot votes are for Swiss politics. After a referendum vote is before the next one, maybe only three months later. Referendum practice examples given in the chapter include the initiative to abolish the Swiss army, the minaret initiative as well as some local and regional level illustrations.

Referendums were rarely employed as a means of decision-making in British politics up to the end of the twentieth century. Their introduction into the British system came as late as the 1970s, as one way to tackle difficult and heavily contested issues regarding the structure of the system of government. In previous decades the notion of parliamentary sovereignty had reigned supreme, as a rather pure form of representative democracy, with no room for direct-democratic elements. Thus, in the UK there were—and still are—no absolute constitutional or legal requirements to conduct referendums on any political issue. Decisions on all aspects of referendums remain the prerogative of Parliament. For the same reason, referendum results are not absolutely binding and it is Parliament’s responsibility to decide on the implementation of any outcome through appropriate legislation.

The Agreement that emerged from the multi-party talks in Belfast on Good Friday 1998 was put to referendums on 22 May 1998 in both Northern Ireland and the Republic of Ireland. It was approved by 71% in Northern Ireland and 94% in the Republic of Ireland. This chapter considers the reasoning behind the simultaneous referendums and what was at stake in the 1998 Agreement being voted on. It also compares the different experiences of the two referendums, including in the structures, funding and fallout of the different campaigns. It concludes by considering the fact that the 1998 Agreement includes provision for further concurrent referendums in the future, namely on whether Northern Ireland should become part of a united Ireland.

On 5 May 2011, voters across the United Kingdom (UK) took part in a referendum that posed the question: ‘At present, the UK uses the “first past the post” system to elect MPs to the House of Commons. Should the “alternative vote” system be used instead?’ Of a total electorate of 45,684,501, 19,165,730 took part, a turnout of 42%. They were required to answer ‘Yes’ or ‘No’. The ‘No’ option won, by 19,165,730 (67.9% of those voting) to 13,013,123 (32.1%) (Mortimore and Blick 2018: 448). This episode was significant for a number of reasons. It was important from the point of view of the establishment and operation of the Conservative-Liberal Democrat coalition formed the previous May (Adonis in 5 Days in May: The Coalition and Beyond, Biteback, London, 2013; Bogdanor in The People and the Party System: The Referendum and Electoral Reform in British Politics, Cambridge University Press, Cambridge, 2011; Seldon and Snowden in Cameron at 10: The Verdict, William Collins, London, 2016).

The discovery of North Sea oil in the 1960s was followed by significant support for the pro-independence SNP. A devolved Scottish Parliament was established in 1999, which unionists hoped would undermine support for independence. However, in 2011 the SNP secured an overall majority, and the UK government accepted that an independence referendum should be held. While there were some disagreements about the rules, there were few grounds on which to question the legitimacy of a contest whose outcome both sides were committed to respecting and which secured the highest turnout since the advent of the mass franchise. Although a majority voted to remain in the UK, support for independence was much higher than expected when the referendum was called. The result triggered an increase in the powers of the devolved parliament, saw the SNP come to dominate Scotland’s Commons representation, and failed to end the debate about independence.

The chapter provides an analysis of the controversial 2017 vote organised by the Catalan government. Following a tour d’horizon of Catalan history and the often overlooked previous referendums in the territory, the chapter presents an overview of, respectively, the history and the legal and constitutional rules pertaining to independence votes. It is concluded that the referendum was a result of short-term political considerations by actors who had no incentive to reach a compromise. Unlike other independence referendums, such as the ones in Montenegro and Scotland, the referendum was not conducive to deliberative policymaking.

Ireland is among a small group of countries which make increasing use of referendums. The state has an extensive written constitution which can only be changed by a national referendum vote. In its original draft, the constitution contained detailed clauses on the legal and political architecture of the state, social policies informed by conservative Catholic social thinking, a notional claim to the territory of Northern Ireland and an expansive interpretation of sovereignty. As elements of the constitution became outdated, referendums were necessitated to modernise the document; in addition, Ireland’s membership of the European Union is especially noteworthy in requiring a large number of referendums since 1973 (see chapters by Svensson, Laffan and Mendez and Mendez elsewhere in this volume). The focus of this chapter is on constitutional referendums dealing with aspects of social policy. It utilises two recent referendums to highlight key issues and implications in the conduct of referendum votes. The marriage referendum introduced marriage equality in 2015 and in 2018 the abortion referendum liberalised the provision of abortion services.

The chapter deals with the referendum of 4 December 2016 on the constitutional revision approved by the Italian Parliament in April 2016 and rejected by the voters. The 2016 constitutional referendum was a ‘non-arrival’ point in a long and grueling political debates and confrontation which remained open and unsolved after almost forty years. Parallel and closely related to the debates on institutional reforms were changes to the electoral laws. The 2016 referendum is a constituent part of political processes that marked Italian politics with the 2013 and 2018 elections. The 2016 constitutional referendum was neither a ‘plebiscite’ nor an expression of ‘direct democracy’. The use of the verbal expression ‘direct democracy’ with reference to the referenda is mainstream: it is a misleading expression for the analysis and empirical theory of the referenda phenomenon. Referenda may be part of the representative government but do not establish any type of ‘direct democracy’.

On 7 June 2015, the Grand Duchy of Luxembourg held a consultative referendum on extending voting rights to non-citizen residents (the so-called Auslännerwahlrecht). In a country otherwise known for its stability, the referendum result resembled a political earthquake as nearly 80 per cent of the Luxembourgish electorate said ‘No’ to enfranchising non-citizens. This chapter discusses the reasons behind the referendum and analyses the campaigns leading up to the vote. The conclusion reflects on the wider implications of the referendum. We argue that the referendum proved divisive in the sense that it stirred up nationalist sentiments and ultimately failed to resolve the lingering democratic deficit it was meant to address.

Turkey has had seven referendums since 1961, originally for seeking popular consent for the new constitutions of 1961 and 1982. The last five referendums functioned as tools for promoting the political agenda of the ruling political parties and their leaders, as these parties and leaders seemed to have aimed at changing the rules of the political game while attempting to eliminate the challenges and resistance from the opposition. Although the 1987 and 1988 referendums failed to aid the governing parties, the last three referendums resulted in changing the regime of the country from democratic semi-parliamentarism to semi-presidentialism first (in 2007) and eventually to an unchecked authoritarian presidentialism. The outcome of the 16 April 2017 Referendum was determined by partisanship in a highly conflictual polarized political milieu, which occurred under the influence of emergency measures, curtailed opposition, scuppered media and produced a regime that resembled nothing less than neo-patrimonial sultanism.

The political background for the referendums in Ireland and Denmark in 1972 was quite similar. The referendums were caused by the successful result of four countries’ application of membership of the European Communities in 1970–71, and for both countries the UK’s central position in the applications process was important, both politically and economically. However, the referendums were regulated by different constitutional rules in Ireland and Denmark, as EC membership required an amendment of the Irish constitution—which required a referendum—whereas the Danish referendum was called because a five-sixths majority for the delegation of sovereignty was not achieved. Nevertheless, similar, intense public debates took place in both countries and the result was a similarly high ‘Yes’-vote as well as a high turnout. It goes without saying that few policy decisions have had such profound impact on life in the two countries as the EC/EU membership, both with regard to their economic and social development. Furthermore, the 1972 referendums in the two countries marked the first step into a new era with an element of direct democracy about European integration.

This chapter focuses on Norway’s two EU referendums. EU membership and treaty change are among the most frequent triggers of referendums in Europe. Norway is no exception. Both referendums saw high levels of participation and political contestation. Nevertheless, membership rejection has not prevented Norway from becoming very closely affiliated with the EU. In this chapter, we start with some theoretical reflections on referendums in relation to democracy. Thereafter we outline the legal framework on referendums in Norway. Referendums are not constitutionally required, although EU referendums may be an exception. The subsequent section outlines the rules guiding each referendum. We then turn to the practice of referendums and referendum debates. The subsequent section considers effects, with emphasis on the cleavage structure, executive–legislative changes, and Norway’s distinct Europeanisation paradox. The chapter demonstrates the complexities of translating the referendum results into a viable policy stance, which requires attention to how the referendum instrument is adapted to the other decision procedures in place.

Harold Wilson’s decision to hold a referendum on staying in the European Economic Community (EEC) in 1975 was unprecedented in the UK and unparalleled elsewhere in Europe. Four decades on, the UK became the first Member State to vote to leave what had now become the European Union (EU). What had changed in the intervening years in European and domestic politics? Why did a country that voted two to one to stay in 1975 narrowly favour leaving four decades later? This chapter looks at the way the use of referendums had changed in the UK in the intervening period and considers how far socio-economic and political transformations in the UK as well as institutional change in Europe affected voters’ choices, alongside the changing media landscape and a fragmenting party system. Finally, it considers the challenges that direct democracy can bring to parliamentary sovereignty.

This chapter reviews the course of events in Austria, Finland and Sweden, where the voters came out in favour of joining the EU in 1994. In these countries, the referendums constituted the final chapter in a debate that had been ongoing for 35 years. In the course of this period, their relations with the EU had slowly developed from foreign policy cooperation to deeper economic integration without full Community membership. The experience gained in the course of this process shaped their policies towards the EU and the domestic debates and campaigns in the run-up to the referendums. Ultimately, these three countries all wanted more than the EEA framework provided for. As seen from the perspective of both government and voter, full membership offered better prospects for economic growth, for security, for the nurturing of a western identity and, crucially, a seat at the table in Brussels as an equal partner.

This chapter explores the most common type of EU-related referendum, the accession referendum, which has been the subject of surprisingly little scholarship. Arising in the context of an EU enlargement, the accession referendum has become the constitutional norm for legitimating membership of the EU. Unlike other types of EU referendums, it very rarely fails to deliver a pro-EU outcome with only one country out of 17 having ever rejected accession. Apart from generating higher ‘Yes’ vote shares than other types of EU referendums, the accession referendum also generally has higher levels of participation and lower levels of polarisation among political parties. Yet its very success in delivering legitimation and the expansion of the EU is likely to give way to its demise as the boundaries of the EU crystallise and its raison d’être disappears.

Denmark holds the distinction of being the member state in which most EU referendums have been rejected (1992, 2000, 2015). The Danish electorate has been asked to vote in referendums related to accession to the EU, treaties amending the EU (e.g. the Treaty of Maastricht) and to remove Danish opt-outs (e.g. euro, Justice and Home Affairs). In all, Danes have voted No in three out of the eight held EU referendums. What can explain No votes when the propositions were supported by a large majority of political parties, newspapers and interest organizations? The chapter presents a brief history of Danish EU referendums and the rules governing them, followed by an analysis of what can account for the No and Yes votes on basically the same proposition in the two Treaty of Maastricht referendums in 1992 and 1993. This is then followed by an examination of why Danes have rejected removing several of the Maastricht opt-outs in subsequent referendums, using the most recent No vote (Justice and Home Affairs) in December 2015 as a case study.

On 29 November 2002, the leaders of the parties in the Swedish Parliament (Riksdag) announced an agreement that a national referendum on ‘the introduction of the euro as a currency’ would be held on the third Sunday of September the following year. They had also decided on the wording of the question that would take the Swedish voters to the polls in September 2003: ‘Do you think that Sweden should introduce the euro as its official currency?’ In early 2003, the Riksdag enacted the law on a national referendum about the euro currency (Government proposition 2002/03:46).

Changes to the Irish Constitution require a referendum and all major new EU treaties have been put to a popular vote in Ireland. Eight times since the mid-1980s, the Irish electorate has engaged in an intense, albeit not always informed debate, on a European treaty. Twice Irish voters rejected an EU treaty and twice Irish governments, supported by the main opposition parties, went back to the electorate to ask ‘Is this your final answer?’. It is important to analyse the Irish case of revisiting the results of referendums to understand the dynamics involved. A number of clear conclusions may be drawn. First, the Irish government backed by the Oireachtas managed the process. Second, returning to the electorate will only work if there is a broad domestic consensus that it is legitimate to revisit the outcome. Third, the EU collectively and the other member states make concessions, short of reopening the treaty, to support a second referendum.

This chapter investigates the referendums of May and June 2005 in which the Dutch and French electorates rejected the ratification of the draft Treaty Establishing a Constitution for Europe. The chapter discusses what happened in France and The Netherlands at the time, set against the two votes’ broader political and discursive backgrounds. It reviews the roles and dynamics of the referendum campaigns, and what we know about why people voted the way they did. In the process, the question is raised of how we might know what people’s electoral choices may actually have meant to them.

Switzerland has had more referendums related to Europe than any other country. These matter qualitatively because EU relations are sensitive and central to Swiss politics and also raise questions about direct democracy. They also have implications for Europe but have rarely been studied. Remedying this requires examining the evolution of Swiss EU policy since referendums have played a crucial part in its changing evolution. 17 referendums are then analysed, covering the three forms of constitutionally recognized votations. Their subjects include matters directly related to EU relations, other issues indirectly affecting European policy and rules on direct democracy. All are examined for their effects, origins, results and significance, along with the arguments used in the closely fought campaigns. They divide into four phases, reflecting the evolution of Swiss–EU relations and reveal changing patterns of influence with Euroscepticism now being challenged by newer forces. Critical questions remain to be decided.

In the Greek referendum on bailouts of 2015 the only winner was populism, with democracy and the institution of referendum itself being the big losers, proving that the end result of populist responses to what is perceived as a ‘democratic deficit’ can only be a trivialisation of democracy itself. Not only did the SYRIZA-ANEL government exhibit no prudence in the use of referendums, despite their allegedly positive stance towards them but, by spoiling and trivialising the institution as a whole, they gave the opponents of direct democratic instruments the best arguments against them. In other words, if one wanted to design a mock referendum, to undermine the institution as a whole, one would create one exactly like the Greek referendum of 2015, which is now often presented as a caricature and as a tool in the hands of demagogues.

Sending a message to Brussels was the main declared objective of Hungary’s Fidesz government, which posed the following referendum question to voters on 2 October 2016: ‘Do you want the European Union to be entitled to prescribe the mandatory settlement of non-Hungarian citizens in Hungary without the consent of parliament?’ Set against the background of the EU’s refugee crisis reaching its peak in 2015, the answer, from virtually everyone who cast a valid ballot, was ‘No’. However, turnout failed to reach the required 50% for a valid result. The ‘quota referendum’, as it is known, is to date the only EU-related vote in the Central and Eastern European Member States since they joined the EU in 2004, and a rare example of a referendum fought on a Eurosceptic platform by an internally united ruling party.

In 2016 The Netherlands held an advisory referendum on the ratification bill on the Association Agreement of 2014 between Ukraine, the EU and its Member States. This advisory popular veto was triggered by 400,000 signatures, based on the new 2015 Advisory Referendum Act. 61% of the voters rejected the agreement, with a turnout of 32.3%, just passing the 30% turnout threshold. Rendering a valid outcome, it was not binding, however. This chapter explores this first nationwide advisory referendum in The Netherlands. It first briefly outlines the legal framework of referendums and rules in place at the time of the vote and the Dutch experience with preceding referendums. The main part is dedicated to the analysis of the ‘Ukraine referendum’ and its aftermath, discussed both from the domestic and the European perspective.

The Danish referendums on the Unified Patent Court in 2014 and on the Opt-Out on Justice and Home Affairs in 2015 were characterized by issue-voting, but the 2015 vote more so than the 2014 vote. Second-order effects were few and small. Very few used the opportunity to cast a vote against a government that they disliked. A large number of voters made up their mind well before the campaign started, and those who became doubtful because of the arguments advanced during the campaign ended up following the cues given by their preferred party, but in case of conflict with those cues relied on their own views and attitudes to the EU.