Cahiers de Science politique

Université de Liège



since 05 February 2011 :
View(s): 6940 (30 ULiège)
Download(s): 0 (0 ULiège)
Min Reuchamps

Referendum as a Tool for Building European Identity. The Case of Belgium, Luxembourg, and The Netherlands

(Cahier n°6)
Open Access


Hitherto, forty-one national referendums have been conducted on European integration in over twenty countries. While it is clear that this process has not led to a widely shared common identity, this paper explores the hypothesis that national referendums on European issues do in fact contribute to the emergence of a European identity. I will propose three mechanisms by which this is likely to happen. Firstly, although referendums and especially the debate preceding them may sharpen the divide between Europhiles and Euroskeptics, they bear the potential to bring into the public sphere awareness of European issues and therefore to spark a learning process. Secondly, the debates triggered by referendums give rise to the emergence of a European public sphere, even in member-states which do not conduct referendums. Thirdly, they can contribute to narrowing the gap between political elites and ordinary citizens, thereby reducing the democratic deficit.

In addition to the burgeoning literature on the topic, which however focuses on the relationship between referendums and support for European integration, Eurobarometers provide substantive and quantitative data on the issues –namely European referendums and identities– raised in this paper. The study will focus on the twenty-seven referendums of the post-Treaty of Maastricht era. The discussion of the concept of European identity opens the essay. In the following sections, I present briefly the political and historical context in regard to referendums upon which I attempt to demonstrate the three specific mechanisms (learning process, European public sphere, and reduction of the democratic deficit) in the emergence of a European identity, using Eurobarometers data which show the evolution of national and European identification before and after the organization of the referendums. The conclusion discusses the findings and explores other elements that matter in constructing a European identity.


1On June 1, 2005, three days after the French Non,2 a majority of Dutch (61.6%) rejected the Treaty establishing a Constitution for Europe. A month later, Luxembourg held a referendum3 on the same subject. The “Yes” won with 56.5% of the votes. In contrast to her neighbors, Belgium ratified the constitutional treaty without popular consultation. In both Luxembourg and The Netherlands, holding a referendum on a European treaty constituted a new experience; hitherto, Belgium has never held a popular consultation on a European issue. Since the beginning of the European integration in the 1950’s, the population of these three founding members has usually strongly supported the European enterprise. While it is clear that the fifty years long process of Europeanization has not led to a widely shared common identity, this paper explores the hypothesis that national referendums on European issues do in fact contribute to the emergence of a European identity. Using John Stuart Mill’s method of agreement and method of difference (Mill 1843/2002), I shall compare the cases of Belgium, Luxembourg, and The Netherlands; whereas the former is not likely to hold any European referendums in the near future (for historical and political reasons), the latter two have recently consulted their population for the first time but with an opposite outcome.4

2My argument is threefold. First, although referendums may sharpen the divide between Europhiles and Euroskeptics, they bear the potential to bring into the public sphere awareness of European issues and therefore to spark discussions about them (a learning process), regardless of the outcomes. This learning process can foster the development of a European identity. Second, in addition to this process, the debates triggered by referendums resound in a European public sphere across the countries, even in member-states which do not conduct a consultation on the issue. Third, they can contribute to narrow the gap between political elites and ordinary citizens; thereby European integration could resonate more among European populations –facilitating citizens’ identification with the process. Above all, the construction of a collective identity, in this case European, is a complex matter where popular consultations can be a useful tool –but not a necessary one, as the case of Belgium will show– along with other forms of participation (civil and political associations as well as EU-wide referendums). Nonetheless, a paradox lies at the heart of European direct democracy: while referendums can contribute to the emergence of a European identity, Europeans can ultimately reject the process with which they start to identify.

3This paper is primarily based on two sources. On the one hand, the growing literature –combining quantitative and qualitative research– on the subject, which focuses on the relationship between referendums and support for European integration, gives us the theoretical and empirical framework. On the other hand, this paper is based on the analysis of Eurobarometers, qualitative studies, and other European surveys prepared principally by the European Commission. Both sources have provided substantive and quantitative data on the issues –namely European referendums and identities– raised in this paper. The discussion of the concept of European identity opens the essay. In this section, I argue that a common European identity, or better European spirit, will differ widely from the existing national identities and above all, will not exclude but rather complement the latter. In the following sections, I present briefly the political and historical context in regard to popular consultations in the three case studies upon which I attempt to answer the question whether referendum can be a tool for building a European identity. The conclusion discusses the findings and explores other forms of participation that facilitate the emergence of a European identity.

European Identity – European Spirit

4Notwithstanding fifty years of tremendous achievements in many aspects of the European political, economic, social, and environmental life, the European Union (EU) has not a common identity. This état de fait raises the question whether the EU can have a common identity. The European Treaties proclaim both “an ever closer union among the peoples of Europe” (art. 1, Treaty on European Union) and the Union’s respect for “the national identities of its Member States” (art. 6§3, Treaty on European Union). Above all, since 1992, “every person holding the nationality of a Member State shall be a citizen of the Union.” But “citizenship of the Union shall complement and not replace national citizenship.” This European citizenship confers several rights: among others, to move, reside, vote, to enjoy diplomatic protection, to petition (Part Two, Treaty establishing the European Community). Nonetheless, Eurobarometers as well as the low-out for European elections show, on the one hand, a strong national identity amongst every people of Europe and, on the other hand, little evidence of a European common identity (see Appendix 1 and 2).

5Collective identity is made of both a self-definition and a sense of belonging. Social groups share “a set of ideas to which members can relate positively” (Marcusen et alii 1999: 615); thereby, on the basis of this set of ideas, which shapes a common worldview, the members of the groups form an “imagined communities.”5 Next to the sense of belonging, the identity of a group is also shaped by its self-definition: who are we and who is/are “the other(s).” Thus, a collective identity finds its roots in several aspects of the group’s life: culture and history, symbols, leadership, public policies, education, among others. These various facets contribute to the construction of a common identity. Moreover, and especially important in the discussion of a common identity for the E.U., “individuals are members of several social groups, which may or may not overlap. In other words, they hold multiple identities” (Marcusen et alii 1999: 616, emphasis in the original).

6Indeed, in some segments of the European population, a European identity complements national identities: certainly amongst E.U. civil servants and leaders, increasingly amongst students. The personnel of the Commission sensu lato, mainly based in Brussels, is uniquely characterized by “its own ethos and a strong esprit de corps” (Shore 2000:127). In fact, E.U. fonctionnaires share a European common identity; i.e. the pursuit of the same European interest and ideal as well as a common tongue (a Euro-speak made of jargon, acronyms, and a mix of French and English). Similarly, though to a lesser extent, European leaders show a common identity. Thus, there is an increasing gulf between those who share a common identity and those who don’t. To be sure, “changing minds on questions of identity is no a small task” (Savage 2004: 45).

7In the context of the European integration, the neo-functionalist approach suggests that eventually doing (i.e. functions) leads to being (i.e. identity): one function creates the need for another function in a different area by way of spill over and finally, theses functions will engender a European identity because they have an increasingly direct effect on European’s lives (see, e.g., Streeck and Schmitter 1991; Stone Sweet and Sandholtz 1998).6 In contrast to the neo-functionalist perspective, the supranational approach contends that the integration of the economies and markets as well as the policies will lead to a common identity. A striking example is the principle of mutual recognition by which any product that meets the country of production’s standards can be sold in any other countries (see, e.g., Cassis de Dijon 1979).

8These lines of arguments lead to the conclusion that there is a relationship between someone’s support for European integration (or the membership of his country in the EU) and his European identity. Thomas Risse shows that the stronger an individual’s European identity, the greater his support for the EU (2002). Using three alternative conceptualizations of national identity, Sean Carey reaches a similar conclusion (2002). Therefore, at the core of a European identity lies the support for European integration. However, the two concepts do not match completely. On the one hand, an individual can be supportive of the European integration but without having a European identity. For instance, the reason for his support might rest only on economic factors –this is the political economy theory for explaining support for European integration (see, e.g., Carey 2002; McLaren 2002). On the other hand, an individual could hold a European identity but not be supportive of the actual direction of the European integration. Rather, he might support, as Jacques Derrida puts it, “une autre Europe, une Europe plus sociale et moins marchande” (2004).

9Finally, “new visions of political order need to resonate with pre-existing collective identities embedded in political institutions and cultures in order to constitute a legitimate political discourse” (Marcusen et alii 1999: 615, emphasis in the original). Chiefly, a european identity is not mutually exclusive to national identities, rather it complements (and resonates with) national identities. The E.U. should not be seen as a super state where all identities and states will be subsumed for the European good. On the contrary, the states are the raison d’être of the Union. As such, it respects and even promotes cultural and linguistic diversity.7 So, in addition to a national identity (and sometimes a regional identity), Europeans could hold a European identity, which could be better called a European spirit (“esprit européen”) in order to stress the difference with national identity. This spirit would reflect the pursuit of a common European well-being (economic, environment, education-wise). This European spirit mixes both cosmopolitanism and multiculturalism. European spirit combines the cultural diversity defended by multiculturalists with a directly understandable lingua franca promoted by cosmopolitans.8

10This said, in the rest of the paper, I analyze how national referendums can foster the emergence of such a European spirit (a combination of both national and European identification) in the case of Belgium, Luxembourg, and The Netherlands.

European Referendums. To Referendum or Not to Referendum

11Apart from Switzerland, European countries do not frequently include their citizens in the law- and decision-making process. However, since the 1970’s, referendums on the European integration have occurred in several member and non-member states. Three different types of referendums can be distinguished: accession referendums,9 treaty-ratification referendums,10 and European issues related referendums.11 Forty-one referendums have been conducted on European issues in almost twenty countries.12 Nonetheless, since the origin of the European construction, no referendum has ever been organized in Belgium; Luxembourg and The Netherlands had their first European referendum this year. Before exploring the use of referendums as a tool for building a European identity in these three countries, the next section presents their respective political and historical context in relation with referendums.


12Since the popular consultation on the Question Royale in March 1950, when a majority of Dutch-speaking Belgians voted for the return of King Leopold III but a majority of French-speaking Belgians opposed it, Belgians in general and Belgian politicians in particular have feared direct democracy. Indeed, the consultation was followed by serious riots and civil war was only avoided because of the abdication of the King (see, e.g., Dumont 1993; Luykx and Platel 1985; Mabille 2000; Wils 1996; Witte and Craeybeckx 1987). Foremost, the Constitution proclaims that “all powers emanate from the Nation and are exerted in the manner established by the Constitution” (art. 33 Const.); in short, in Belgium, the principle of representation does not allow direct democracy, i.e. the organization of referendums. In this regard, the non-binding 1950 referendum was referred as a –popular– consultation. Moreover, citizens are also excluded form the treaty-ratification process where the normal parliamentary procedure applies (art. 53 and 167 Const.). Hence, Belgians have never been consulted on the question of the European integration.

13However, the fear of direct democracy has gradually diminished –the ever-present internecine linguistic conflict notwithstanding– partly because of the growing number of referendums organized throughout the world (Butler and Ranney 1994), partly because of the increasing gap between political leaders and citizens –to which referendums are considered to be a solution (Rigo 2004).13 In the last decade, many propositions and arguments have flourished for more civic participation in the res publica via the organization of consultative referendums. Nonetheless, despite the new possibilities to hold local and provincial consultations,14 no consensus has hitherto been reached at the national level and it seems that both political and constitutional factors might hinder the use of direct democracy instruments in Belgium.

14Thus, although Belgians show a high support for European integration in general, and for the European Constitution in particular, they have not been given the chance to state it through a formal procedure.15 Noteworthy, despite the absence of a referendum, echoing the constitutional debate in both France and The Netherlands, the media as well as the politicians discussed the topic of the European Constitution. Eventually, a large majority in every of the six parliaments ratified the proposed text.16


15Similar to Belgium, Luxembourg held its first referendum on a royal question. In 1919, Luxembourgers were asked whether they wished to live in a Republic or in a Monarchy; the latter was chosen by more than 80% of the voters. The second and third referendums were organized on socio-economic issues during the inter-war period. Since then, instruments of direct democracy had not been employed in the Grand Duchy on national issues, let alone on European ones.

16In his 2003 state-of-the-nation speech, Prime Minister Claude Juncker announced a referendum on the European Constitution stressing the importance of the issue (Juncker 2003). Moreover, even though only consultative referendums may be organized, he stated that the choice of the citizens would be respected –while, however, he pledged to resign should the treaty be rejected (Glesener 2005). On July 10, 2005, following six months of rotating presidency and a few weeks after the French and Dutch “No,” Luxembourgers went to the polling booths and a majority of them endorsed the constitutional text (56.5%),17 which was formally ratified by the parliament two weeks later.

17Indeed, the victory of the “Yes” camp was more difficult than expected, especially knowing Luxembourg’s exceptional Europhilia with 82% of her citizens in favor of being a member of the EU (Eurobarometer 64, Luxembourg, 2005: 3; and Eurobarometer 173, “Post-Referendum in Luxembourg” 2005). The campaign both produced one of the most heated debates about Europeans politics to date and revealed the existence of a discrepancy between the attitude toward Europe among the elites and the general public (Hausemer 2005). This said, the heat of the campaign was also a consequence of the French and Dutch “No.” Finally, the campaign and the results of the referendum illustrate the evaporation of the “permissive consensus” (Inglehart 1977) which dominated the European (non-)debate until then in Luxembourg.

18The Netherlands

19The experience of direct democracy via a referendum constituted a unique event in Dutch modern history. The Netherlands was the only European countries which never held a nationwide referendum (Rigo 2005: 2). The Dutch constitution organizes a political system chiefly based on the principle of representation (art. 50 and art. 67 Const.), even in the case of the transfer of sovereignty towards international institutions (art. 91 and 92 Const.). However, proponents and opponents of direct democracy have argued for the last century.18 Since the 1990’s, bills introducing the possibility of referendums have been discussed at the national, provincial, and communal levels. Finally, a specific authorization to consult the population over the European constitution was passed by the House in October 2004 and by the Senate in January 2005.

20On June 1, 2005, the Netherlands organized her first nationwide referendum, which was non-binding (i.e. consultative) and non-compulsory. Almost two-thirds of Dutch electors participated in the consultation; 61.6% of them rejected the constitutional treaty (Eurobarometer 172, “Post-Referendum in The Netherlands” 2005). Although the Dutch have usually strongly supported the European integration (see appendix 1), the victory of the “No” was not unexpected. Both national and European factors seem to explain the result of the referendum.

21On the one hand, for the last decade, a widespread disenchantment with the political establishment has haunted the Dutch polity. This disenchantment was notably illustrated by the capture of 17% of the votes in the 2002 national elections by the List Pim Fortyun –despite the absence of the Pim Fortuyn himself. Moreover, since all main parties campaigned for the “Ja,” voting “Nee” was an occasion to protest against the political establishment (Crum 2005) as well as to manifest one’s concern about issues of identity (for instance, immigration). On the other hand, the Dutch public perceived the process of Europeanization as “a blind train which is running far too fast” (Cuperus 2005: 1). Above all, the rejection of the European constitution indicated, like in Luxembourg, the end of the “permissive consensus” over the European enterprise and the ever-increasing gulf between the political elites and the citizens.

22How can referendums contribute to the emergence of a European identity in these three countries is the interrogation that the rest of the paper attempts to answer.

Referendums on European Integration

23The first national referendum on European integration was held in France in 1972, since then, as mentioned above, forty-one referendums have been conducted on European issues in almost twenty countries. I briefly present the burgeoning literature on this subject in order to inform the study of the relationship between referendums and European identity.

The Voters

24Since the 1970’s, an increasing amount of scholarship has addressed the question of how voters decide in European referendums. Two competing approaches have endeavored to explain voting behavior. On the one hand, the “attitude” school “focuses on individuals’ values and beliefs and argues that voting in EU referendums reflects people’s underlying, broad attitudes towards European integration” (Hobolt 2006: 155; see, e.g., Siune et alii 1994 or Svenson 2002). Thus, voting in a referendum is similar to voting in an ordinary election and voters make reasoned decisions about the future of the EU. On the other hand, inspired by the second-order theory of elections, the alternative school maintains that “national issues tend to dominate the campaigns and voters are thus expected to use their vote as a means of signaling their satisfaction or dissatisfaction with the government or to follow the recommendations of national parties” (Hobolt 2006: 155; see, e.g., Franklin et alii 1994). Hence, the outcomes of referendums constitute either plebiscites or punishments for the performance of national governments.

25Chiefly, the explanation for voting behavior lies probably in-between, varying from one referendum to the other, as well as from one voter to the other. In other words, it is a question whether voters “really address the issues and involve themselves actively in the policy-making process in a vital issue or do they merely vote for or against the current government” (Svenson 2002: 733). Noteworthy, both Franklin (2002) and Hobolt (2005) suggest that when voters have a great knowledge of European affairs, they are less likely to treat the referendum as a second-order election.19

The Political Elites

26The chief characteristic of referendums is the participation of the electorate in policy-making. However, the role of the political elites should not be minimized. Indeed, very few referendums are required by law;20 most of them are “the product of strategic elites considerations” (Hobolt 2006: 157). Morel presents different reasons why governments have initiated referendums: certainty of success, solution to an intra-government, intra-party or broader political disagreement, and de facto obligatory referendums (2001). Yet, when the process is launched, “referenda increase the salience of European issues and, perhaps more importantly, limit the capacity of political parties and their leaderships to control debate. Unlike general elections, referenda highlight conflicts within, as well as among political parties” (Hooghe and Marks 2003: 7-8). Whereas general elections see political parties (controlled by the political elites) seek to win the right to govern, referendums, in contrast, constitute unrestrained expressions and therefore can spin out of the elite control (Leduc 2001). Thus, the relationship between support for European integration and national identity should vary between the countries which have experienced referendums on European issues and those which have not.

27Moreover, both Dutch and Luxembourgish referendums have signaled the end of the permissive consensus on European integration; i.e. the citizens desire to have a say in the direction and the speed of the European process. It is also a call for the reduction of the gap between political elites and ordinary citizens. Finally, and chiefly, discourses and values do matter. In the context of the reform of the welfare state, Vivien Schmidt stresses the role of values and legitimizing discourses in policy change (2000). Her emphasis on political discourse can be aptly applied in the context of referendums on European integration. During the campaign, Dutch political leaders from mainstream parties did not address the content of the Constitution, rather they urged the voters to be reasonable as “voting no was considered to be dangerous and stupid” (Cupérus 2005: 5). In Luxembourg, the highly popular Prime Minister Jean-Claude Juncker, a cheerful supporter of the European Constitution, threatened to resign, should the citizens reject the constitutional text. Yet, a survey showed that 45% of the respondents thought that the Prime Minister’s attitude amounted to “blackmailing voters.”21 In both cases, political discourses did not resonate much among Dutch and Luxembourgese populations. Therefore, it seems clear that, in future referendum campaigns, should political leaders fail to offer an appropriate discourse or worse use an arrogant tone, the electorate would inevitably use the poll to sanction them rather than position itself on the issue at stake.

Support for European integration: national identity and referendums.

28Relying on the large-scale quasi-experiment of European integration, Christin and Hug test whether referendums lead to more supportive voters with data covering more than 20 years. They find suggestive evidence that referendums lead to citizens’ being more supportive of European integration. Moreover, similarly to Eichenberg and Dalton (1993), their results suggest that voters more strongly support European integration immediately after a referendum vote, on average by more than 4% (Christin and Hug 2002). However, some caution is necessary in interpreting these results. First, “simply holding a referendum on European integration does not necessarily increase” support for European integration (ibidem: 606). Second, as suggested above, it is likely that voting behaviour is shaped not only by the issue at stake but also, more or less, by other issues. Third, quantitative designs assume that the votes are made under complete information.

29 Foremost, although it seems counter-intuitive, “the greater the citizens’ pride in and attachment to their nation, the greater is their support for European integration” (Hooghe and Marks 2003: 19) and therefore, the same authors conclude, “cautiously but emphatically, that national attachment is not mutually exclusive with either European attachment or support for European integration” (ibidem: 19, emphasis in the original). For instance, like Spaniards, Belgians have multiple identities in the same time (Billiet, Doutrelepont, and Vandekeere 2000). Thus, “even in an era in which perceptions of the European Union as successful seemed to decline, the tendency to identify with both nation and Europe increased” (Citrin and Sides forthcoming: 8-9). Yet, the content of a European identity may vary profoundly across territorial contexts (Risse 2002). To sum up, “whether one conceives of the relationship as embedded, nested, or enmeshed, there is broad consensus that a strong sense of national identity is consistent with European identity and support for European integration” (Hooghe and Marks 2003: 20).

Referendum as a Tool for Building European Identity

30Notwithstanding both the absence of referendum in Belgium and the opposite outcomes in Luxembourg and The Netherlands, the analysis of the three case-studies suggests that three effects generated by referendums can foster the emergence of a European identity. Although the three effects are related and definitely reinforce one another, they still can be distinguished as a learning process, a European public sphere, and the narrowing of the political gap.

A learning-process

31In The Netherlands and Luxembourg, the campaign for the referendum gradually generated a complex learning-process (Crum 2005: 3). This is important to note since in November 2004 over three-quarters of the voters in the two countries had not personally heard about the draft of the European Constitution or knew very little about its contents. In Belgium, it was the case for nine-tenths of the respondents.22 For this reason, Leduc maintains that in a referendum the campaign matters much more than in ordinary elections (2005: 5). Thus, the question arises of the level of knowledge of the voters: “one of the criticisms often leveled against the use of referendums is that ordinary citizens lack sufficient knowledge to vote on complex policy issues” (Hobolt 2006: 156). On the other hand, the organization of a public consultation can contribute to increasing the level of knowledge via the information offered by the media as well as by the political leaders.23

32Above all, this learning-process is not only one-way: from those who have the knowledge to those who don’t. A referendum offers the opportunity for a broad and public deliberation: before, during, and after the actual voting day, regardless of the outcome. Such a deliberation brings to the agenda important issues related to the question of the referendum itself. Indeed, during both Dutch and Luxembourgish campaigns, the debates focused not only on the content of the Constitution but also on European integration (as well as on non-European issues). The benefits and the costs of the EU were discussed; the question of future enlargement was also part of the debate. And even if, in the short term, the voters cannot expect to renegotiate every compromise brokered between the twenty-five member-states, a public deliberation has the potential to bring about, in the long run at least, a process of Europeanization which would resonate more among the populations, and therefore, they could identify –more– with the process.

33Finally, in Belgium, although no referendum was conducted, a learning process did occur. Indeed, the announcements that a referendum would be held on the constitutional treaty in France and in The Netherlands triggered a learning process in Belgium. For instance, before the referendum the public radio organized debates between supporters and opponents to the Constitution. Newspapers discussed in length the content, as well as the arguments for and against the constitutional text. (Web)blogs were also created for the occasion: quite a few journalists –sometimes based in Paris– reported on a daily basis the evolution of the campaign, commented on the text, and gave their opinions.24 Most of these blogs also provided forums where (blog)readers could express their views. After the referendum, the media tried to explain what the results meant and what was going to happen for the EU. This trans-national attention reflects the development of a European public sphere to which I turn now.

A European public sphere

34This effect resembles and diverges from the first effect –the learning process. The emergence of a European public sphere resembles the learning process because it occurs not only in the country where the referendum is conducted but also in the neighboring states. It diverges because many more issues than the learning process are discussed in the burgeoning European public sphere. Moreover, a referendum alone does not generate a trans-national public sphere, but it can contribute to the development of such a European public sphere. Nonetheless, this so-called public sphere is fragmented rather than united, plural rather than unique. Neighboring and similar countries, such as Belgium, Luxembourg, and The Netherlands, are more likely to share a public sphere. Furthermore, the public sphere is not frozen once and for all and public spheres overlap one another.

35Van Gunsteren argues that the Dutch “No” vote has the potential to foster the development of a European identity. The Dutch referendum (similar to the French referendum) and more precisely its outcome have made plainly visible the growing existence of a European public space: the “No” of the Dutch voters “resounded in a European public space. The reactions of the authorities in Brussels –the ratification process must continue– were broadcast on television in many European countries. International journalists sped towards the Netherlands and reported about anxiety, emotions, and gut feelings. Such a European public space did not really exist before, or only in a very limited sense. It was the domain of professionals and Europeans, not of the citizens. … In acting by way of referendum, voters have finally created a really European public space, without which European democracy and citizenship could never materialize” (Van Gunsteren 2005: 407).

36Although no referendum was to be conducted in their country, Belgians followed both French and Dutch campaigns (however, for linguistic reasons, Francophone Belgians paid more attention to France, and Flemish Belgians focused mainly on the Dutch campaign). Francophone and Flemish media echoed the French and Dutch constitutional debate. These developments can be seen “as intimating both the need for and the possibility of a new public sphere: an agora for our day that both compromises and transforms the public spheres of existing states so as to provide communicative direction to a new Eurodemocracy” (Cohen and Sabel 2003: 360).25 In these new developments, television, more than any forms of media, plays a crucial role. Television programs reach a large audience and cable television (which is very widespread in Belgium and Luxembourg)26 offers national channels as well as foreign channels. What’s more, in addition to the national channels Francophone and Flemish Belgians watch respectively French and Dutch channels, while Luxembourgers watch German and Belgian programs.

37To sum up, the emergence of a European identity is likely to follow the burgeoning of a European public sphere, at least in closely-related countries such as Belgium, Luxembourg, and The Netherlands. Holding a referendum on a common issue can spark –or reinforce– the development of such a European public sphere.

The narrowing of the political gap

38Both Dutch and Luxembourgish referendum signaled the end of the permissive consensus on the European integration –which was the business of the elites but not of the citizens who trusted the former to take good care of it. They also blatantly revealed –although there were signs of it before– the gap between the political class and the ordinary citizens. In The Netherlands particularly, the rejection of the treaty took place in a context of political disenchantment (Crum 2005; Van Gunsteren 2005). Above all, it seems that there is a democratic deficit both at the national and the EU level. As of the latter level, European citizens feel that the EU lacks democracy, indeed. Representation at the EU level is in fact very different from what is known at the national level. In fact, only the Members of the European Parliament (E.P.) are directly elected though in second-order elections (very low turn-over) and they are seen as second-order politicians. Ministers sitting in the Councils are not elected in first place as “European” ministers but above all as national ministers and therefore on national (and not European) issues.

39This said, some scholars argued that representative democracy is not the only form of democracy. Vivien Schmidt contends that “as such, the E.U. level taken on its own is not democratic in the sense of governance by and of the people, although it can be democratically legitimated in other ways, by its combination of governance with the people through semi-pluralistic policymaking processes and governance for the people through its quasi-federal institutional structures” (Schmidt 2006: 1; emphasis in the original). In this perspective, Stijn Smisman sees a European civil society (which regroups not only traditional interest groups, but also non-governmental organizations, associations from several levels –neighborhood, city, region, state– and communities) acting as governance with the people (Smisman 2003). Associative democracy seems to be better fitted to the European context than classic representative democracy.

40In this perspective, instruments of direct democracy (namely referendums, regardless of their binding effects or not; or if they are required or not) can play an important role in narrowing the gap between the political elite and the voters, and thus set the stage for a governance with the people. Mayer and Palmowski add that “it has become commonplace to assert the disjuncture presented by the accelerated drive for integration and the manifest absence of a popular European ‘will’, but in essence this is a question of identity” (2004: 574). Thus, before dealing with hard issues (such as taxation, or social and economic policies), soft issues (such as the democratic deficit) need to be resolved. Policies can at most create and sustain the conditions under which a European identity may emerge. Such an identity can only become a reality through the actions of citizens themselves. Referendums on European integration constitute such a condition for European identity to emerge. Herman Van Gunsteren concludes that “the citizens have seized this opportunity and acted. Thus, they have in one move realized European citizenship and a real European public/political space” (Van Gunsteren 2005: 411).

41In short, referendums not only shed light on the gap between the political elites and the voters but also have the potential to narrow it by integrating the citizens in the decision about the content, the direction, and the speed of the European enterprise, at least partially and in combination with the two other effects. By so doing, the process of Europeanization resonates more among Europeans and hence, increasing their identification with it.


42A paradox accompanies necessarily the implementation of direct democracy on European integration. While referendums can contribute to the emergence of a European identity, Europeans may ultimately reject the process with which they start to identify. It is worth reflecting on this paradox. First, by definition, direct democracy bears the potential to bring a “positive” or a “negative” outcome –whose meaning may vary from one person to the other and from one issue to the other. In any case, this outcome would be democratic, even though the voters might have actually not voted on the specific issue. A second comment logically follows: referendums must be practically organized as to minimize second-order effects (i.e. that referendum be an opinion poll on the work of the government or on the state of the economy). In the case of the European integration, this is particularly crucial since the outcome of a negative vote seems not to matter much (i.e. voters feel they have a freebie). Thus, the wording of the question should be precise, clear, and specific (as to yield a yes or no answer). Third, the turnout for EU elections being very low, referendums on a European issue may draw very few voters, too.27 Therefore, in the absence of laws, the government should a priori set the conditions (minimum turnout rate as well as the minimum percentage of positive votes) under which it will follow the outcome of the referendum.28 Fourth, referendums, which differ widely from ordinary elections, have some particular characteristics : highly volatile electorate with dramatic change during the campaign, highly uncertain outcomes, context and timing are important, risk of second order effects, and the political advantage rests with the “No” camp (Leduc 2005: 12).

43Mark Leonard, who strongly advocates the use of referendums, provides a “democratic functionalist” solution to the paradox of referendums. After a moratorium on new treaties –during which the EU should focus on the vrais problèmes des gens, citizens should be included in the Europeanization process. Since they might disagree on how to integrate more, the EU will transform progressively into a multiple EU. On the common background of the 800,000 page acquis communautaire, a few willing members could further the integration by way of coopération rapprochée (as it was the case for the Schengen agreement or the Euro-zone). Finally, the new treaties (after the moratorium) should be narrower (thus, it is easier to hold referendum on the issue) and the referendums should be coordinated in order to avoid to end the series of referendums with Euro-skeptic countries.29 Nonetheless, using extensively a differentiated integration will eventually lead to a multiple EU and therefore, to the loss of the essence of the European construction.


44Since its very beginning, Belgians, Luxembourgers, and Dutch have shared a strong support for the European integration. Nonetheless, after fifty years of Europeanization, they still don’t share a common European identity. The inexistence of a widely shared common European identity started the reflection of this paper. Moreover, these three similar countries diverged drastically on the issue of the European Constitution. Whereas Belgium did not hold any referendum on the constitutional treaty, both Luxembourg and The Netherlands conducted their first public consultation on a European issue, albeit with an opposite outcome. On this background, I raised the question whether referendums could contribute to the emergence of a European identity.

45The exploration of the three case-studies suggests a positive –but cautious– answer. Three positive effects have been identified. Referendums can trigger a learning process, the emergence –or reinforcement– of a European public sphere, and the narrowing of the political gap between the elites and the ordinary citizens: the combination of these effects could lead to the emergence of a European identity –or European spirit, in order to stress the difference from national identities. Nevertheless, holding referendums on European integration can paradoxically lead to stop the process of integration. Indeed, in the case of the development of a European identity, while referendums can contribute to the emergence of such an identity, Europeans can ultimately reject the process with which they start to identify. To this paradox, there is no perfect solution. However, the practical organization of referendums can contribute to minimize second-orders effects. The wording should be specific and clear; the timing should be thought of.

46Above all, last year’s referendum was a total novelty for both Dutch and Luxembourgers. The exercise of direct democracy has to be learnt, especially in the case of European integration where voters feel that they have a freebie since Europe seems not to matter. The development of a collective, in this case European, identity is a long process. Referendums do in fact contribute to the emergence of such a collective identity. Yet, referendums are neither the only instruments nor the solution to the problem of participation at the EU level. To conclude this paper, I elaborate briefly on two other forms of public participation at the EU level: EU-wide referendums and European conventions.

47EU-wide referendums, like national referendums on European integration, have the potential to foster a European identity. Such referendums would trigger the three effects discussed in the essay; and probably in a more striking fashion since all Europeans would be part of the process. One can easily imagine the impact on the development of a European public sphere. Nevertheless, EU-wide referendums would face the same pitfalls as national referendums (even perhaps in a more dramatic fashion). What’s more, the very idea of conducting EU-wide referendums is likely to be fought harshly for such referendums would give to much weight to big countries or they would be practically difficult to be held, let alone the absence of political will.

48European conventions, reproductions grandeur nature of the constitutional convention, could be organized on European issues. There –it might also be virtual conventions, European citizens could gather and discuss thoroughly, in the same fashion the members of the convention did, European questions. These conventions could pave the way for the formation of a Habermasian “constitutional patriotism.” A European identity could emerge from these European gatherings and working-together-for-the-common-good events. However, the practical organization of such gatherings may bare this theoretical suggestion to ever become reality.

49May 2006


Appendix 1: Support to country’s membership to the European Community/Union: 1973-2004.

50Question: “Generally speaking, do you think that (your country's) membership of the European Community (Common Market) is ...?”

51Source: Eurobarometer Interactive Search System,

52, accessed April 1, 2006.

Country: Belgium Period: From September 1973 (EB1) To April 2004 (EB61)

Country: Luxembourg Period: From September 1973 (EB1) To April 2004 (EB61)

Country: The Netherlands Period: From September 1973 (EB1) To April 2004 (EB61)

Appendix 2: National identity and European identity: 1992-2004.

53Question: In the near future do you see yourself as...?

54Source: Eurobarometer Interactive Search System,

55, accessed April 1, 2006.

Country: Belgium Period: From April 1992 (EB37.0) To April 2004 (EB61)

Country: Luxembourg Period: From April 1992 (EB37.0) To April 2004 (EB61)

Country: The Netherlands Period: From April 1992 (EB37.0) To April 2004 (EB61)


56Anderson B., 1983, Imagined Communities: Reflections on the Origin and Spread of Nationalism, New York, Verso.

57Archibugi D., 2004, “The Language of Democracy: Vernacular or Esperanto? A Comparison between the Multiculturalist and Cosmopolitan Perspectives”, Paper presented at Harvard University.

58Billiet J., Doutrelepont R., and Vandekeere M., 2000, Types van sociale identiteiten in België: convergenties and divergenties, Mimeo.

59Butler D. and Ranney A. (ed.), 1994, Referendums around the World, London, Macmillan.

60Carey S., 2002, “Undivided Loyalties. Is National Identity an Obstacle to European Integration?” in European Union Politics, Vol. 3, No. 4, pp. 387-413.

61Chirac J., 2000, “Our Europe,” Speech Given at the Bundestag, June 25, Berlin.

62Citrin J. and Sides J., Forthcoming, “Can There be Europe without Europeans? Problem of Identity in a Multinational Community.” In Identities in Europe and the Institutions of the European Union, edited by Richard Herrmann, Marilynn Brewer, and Thomas Risse. Lanham, Md.: Rowman and Littlefield.

63Christin T. and Hug S., 2002, “Referendums and Citizen Support for European Integration,” in Comparative Political Studies 35:586-617.

64Crum B., 2005, “EU Referendum Tests the Dutch Political Establishment,” Brussels, Center for European Policy Studies, Working Paper.

65Cuperus R., 2005, “Why the Dutch Voted No. Anatomy of the New Euroscepticism in Old Europe,” Wiardi Beckman Foundation, Working Paper.

66Derrida J., 2004, “Une Europe de l’espoir,” in Le Monde Diplomatique, November.

67Dumont G.-H., 1993, La Belgique : hier et aujourd'hui. Paris, P.U.F.

68Franklin M., Marsh M. and McLaren L., 1994, ‘Uncorking the bottle: popular opposition to European unification in the wake of Maastricht’, in Journal of Common Market Studies, Vol. 32, pp. 455–72.

69Fuchs D. and Klingemann H.D., 2002, “Eastward Enlargement of the European Union and the Identity of Europe”, in Mair P. and Zielonka J. (ed.), The Enlarged European Union: Diversity and Adaptation, London, Frank Cass & Co, pp. 19-54.

70Geddes A. and Guiraudon V., 2004, “Britain, France, and EU Anti-Discrimination Policy: The Emergence of an EU Paradigm”, in West European Politics, Vol. 27, pp. 334-353.

71Glesner M., 2005, “Es gibt keine Alternative zu Europa,” in Wort, March 19.

72Gunsteren H. van, 2005, “The Birth of the European Citizenship Out of the Dutch No Vote,” in European Constitutional Law Review, pp. 406-411.

73Hausemer P., 2005, “Luxembourg’s Referendum on the European Constitutional Treaty,” European Parties Elections and Referendums Network, Referendum Briefing Paper, No. 14.

74Hobolt S.B., 2006, “Direct Democracy and European Integration,” in Journal of European Public Policy, Vol. 13, pp. 153-166.

75Howorth J., 2000, “Being and Doing in Europe since 1945: Contrasting Dichotomies of Identity and Effciency”, in Adrew J., Crook M., and Waller M. (ed.), Why Europe? Problems of Culture and Identity, New York, St. Martin’s, pp. 85-97.

76Inglehart R., 1977, The Silent Revolution, Princeton, Princeton University Press.

77Juncker J.-C., 2003, “Speech on the State of the Nation,” Luxembourg,

78 accessed on April 10, 2006.

79Laitin D., 2002, “Culture and National Identity: ‘The East’ and European Integration”, in Mair P. and Zielonka J. (ed.), The Enlarged European Union: Diversity and Adaptation, London, Frank Cass & Co, pp. 55-80.

80Leduc L., 2001, “Referendums and Elections: Do Voters Behave Differently? When and How?,” Paper presented at the Workshops of the European Consortium for Political Research, Grenoble.

81Leduc L., 2005, “The European Constitution Referendums: What We Already Know,” Working Paper,, accessed on April 10, 2006.

82Luykx T. and Platel M., 1985, Politieke geschiedenis van België, Anvers, Kluwer.

83Mabille X., 2000, Histoire politique de la Belgique : facteurs et acteurs de changement, Bruxelles, CRISP.

84Marcussen M. et alii, 1999, “Constructing Europe? The Evolution of French, British and German Nation State Identities”, in Journal of European Public Policy 6:4 Special Issue, pp. 614-633.

85McLaren L., 2002, “Public Support for the European Union: Cost/Benefit Analysis or Perceived Cultural Threat?”, in The Journal of Politics, Vol. 64, pp. 551-566.

86Hooghe L. and Marks G., 2003, “National Identity and Support for European Integration,” Wissenschaftszentrum Berlin für Sozialforschung (WZB), Working Paper.

87Miles L., 2004, “The Paradox of a Popular Europe,” in Journal of Common Market Studies, Vol. 42, pp. 1-8.

88Mill J. S., 1843/2002, A System of Logic, Honolulu, University Press of the Pacific.

89Morel L. ,2001, “The rise of government-initiated referendums in consolidated democracies’, in M. Mendelsohn and A. Parkin (eds), Referendum Democracy: Citizens, Elites and Deliberation in Referendum Campaigns, New York: Palgrave.

90Rigo A., 2004, “The Ratification of the European Constitution in Belgium,” Working Paper,,%20A._2005%20Belgium.pdf, accessed on April 9, 2006.

91Rigo A., 2005, “Why the Dutch Voted No: Anatomy of the New Euroscepticism in Old Europe,” Working Paper,

92,%20A._2005%20Netherlands.pdf, accessed on April 9, 2006.

93Risse T., 2002, “Nationalism and Collective Identities. Europe versus the Nation-State?,” in Heywood P, Jones E., Rhodes M. (eds.), Developments in West European Politics, New York, Palgrave, pp. 72-93.

94Savage T.M., 2004, “Europe and Islam: crescent Waxing, Cultures Clashing”, in The Washington Quarterly, Summer 2004.

95Shore C., 2000, Building Europe: The Cultural Politics of European Integration, London, Routledge.

96Siune K., Svensson P. and Tonsgaard O.,1994, “The EU: the Danes said “no” in 1992, but “yes” in 1993: how and why?”, in Electoral Studies, Vol. 13, pp. 107–16.

97Smisman S., 2003, “European Civil Society: Shaped by Discourses and Institutional Interests”, in European Law Journal, Vol. 9, No. 4, September, pp. 473-495.

98Stone S., Sandholtz A., and Sandholtz W. (eds.), 1998, European Integration and Supranational Governance, New York, Oxford University Press.

99Svensson P., 2002, “Five Danish referendums on the European Community and European Union: a critical assessment of the Franklin thesis,” in European Journal of Political Research, Vol. 41, pp. 733–50.

100Streeck W. and Schmitter P.C., 1991, “From National Corporatism to Transnational Pluralism: Organized Interests in the Single European Market,” in Politics and Society, Vol. 19, No. 2, 133-164.

101Schmidt V.A., 2000, “Values and Discourse in Adjustment,” in Scharpf F.W. and Schmidt V.A., Welafre and Work in the Open Economy, From Vulnerability to Competitiveness, Vol. 1, Oxford, Oxford University Press.

102Schmidt V.A., forthcoming-2006, Democracy in Europe: The EU and National Polities, Oxford, Oxford University Press.

103Wils L., 1996, Histoire des nations belges, Ottignies, Quorum.

104Witte E. and Craeybeckx J., 1987, La Belgique politique de 1830 à nos jours : les tensions d'une démocratie bourgeoise, Bruxelles, Labor.


1 Paper prepared for presentation in the seminar Social Europe (IR551), Pr. Vivien A. Schmidt, Boston University, Spring 2006. The author was at the time a fellow of the Belgian American Educational Foundation (B.A.E.F.) and a Fellow of the Francqui Foundation.
2 On May 29, 2005, French voters rejected the constitutional treaty with 54.7% in favor of the “No.”
3 I use the term referendum to designate all kind of referendums and consultations (binding and non-binding; required, non-required active and non-required passive).
4 France could have been included in this study. However, it was excluded for several reasons : the three BeNeLux countries forms a more or less homogenous set, they are considered as “small countries” in European terms and as “low countries” in the political science jargon, they have usually shared a federal vision of the European integration, and European referendums constitute a new experience.
5 Benededict Anderson refers to nation as imagined communities: “imagined as both inherently limited and sovereign. […] It is imagined because members […] will never know most of their fellow-members […], yet in the minds of each lives the image of their communion. It is limited because […] the nation has finite, if elastic boundaries, beyond which lie other nations. […] It is sovereign because it came to maturity at a stage of human history when freedom was a rare and precious ideal. […] Finally, it is imagined as a community because […] it is conceived as a deep, horizontal comradeship” (Anderson 1983: 5-7).
6 The founders of the European Union hoped to deepen European identity as a by-product of integration that could be led by elites with the implicit consent of publics (Marks and Hooghe 2003: 29). Likewise, Jolyon Howorth argues that the E.U. is characterized by its doing; however, it lacks a being. Therefore, the Europeans do not share a political culture which is the harmonization between being and doing. The question is, then, “could Europe ever become a place of being, a place of identity?” Howorth suggests that “units of human collectiveness which begin life naturally as units of doing, normally evolve into units of being, and hence of identity,” yet he acknowledges, unlike at the E.U. level, at the national level, the state played a chief role in the development of the nation-sate identity (Howorth 2000: 93).
7 For instance, notwithstanding the cost and the practical impact, any official language of a country becomes an official language of the EU
8 This view of a European spirit fits David Laitin’s 21 cultural configuration, which is consolidating itself in Europe (Laitin 2002). On top of their national culture, Europeans embrace a more or less common European culture and sometimes, they can also embrace a regional culture. As of languages for instance, in addition to English (an all-European lingua franca), each state conserves its own language. Moreover, in some regions, private -and sometimes public- life is operated in a third language (Catalan in the North-East of Spain, or Russian in Latvia).
9 In Denmark, Ireland, and Norway (1972); in the United Kingdom (1975); in Austria, Finland, Norway, and Sweden (1994); in the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia, Slovenia (2003).
10 In Denmark (1986) and Ireland (1987) for the Single European Act ; in Denmark, France, Ireland (1992) and in Denmark again (1993) for the Maastricht Treaty ; in Denmark and Ireland (1998) for the Amsterdam Treaty ; in France, Luxembourg, The Netherlands, and Spain (2005) for the European Constitution.
11 In France (1972) on the enlargement of the European Community ; in Italy on the mandates for members of the European parliament (1989); in Liechtenstein and Switzerland on the European Economic Area treaty (1992); in Denmark  (2000) and Sweden (2003) on the European currency.  
12 This figure is correct as of April 2006 and excludes referendum in Greenland, the principality of Liechtenstein and the Åland Islands (Finland).
13 Interestingly, the special Eurobarometer 199 “Citizenship and Sense of Belonging” reports that with 62% of Belgians supporting the statement “Citizens should participate more actively in politics in our country” Belgium has the lower score amongst EU-15 countries (Greece and Sweden with respectively 89% and 85% have the higher percentage)(2004: 26). However, it’s impossible to know what citizens understand by “participate more actively in politics.”
14 The Constitution was first modified in 1997 in order to allow the possibility for local consultative referendums, and then in 1998 to extent the possibility to the provincial level.
15 In Autumn 2005, on the one hand, 78% of Belgians thought a European Constitution was necessary to ensure that the European institutions work well but on the other hand, a majority suggested that the Constitution should be renegotiated (60%)(Eurobarometer 64, Executive Report, Belgium, 2005: 2-7).
16 The Federal Parliament (Chamber and Senate) and the Assemblies of both Communities and Regions approved the treaty (April 2005-February 2006).
17 Under Luxembourgese law, voting is compulsory, which explains
18 In 1903, the Parliament discussed the possibility of introducing referendums in the Dutch polity (Rigo 2005: 3).
19 The topics of opinion change, the campaign, the electorate and turnout, the vote decision, and the results have been discussed elsewhere (see, e.g., LeDuc 2005).
20 With the exception of Ireland, and the partial exception of Austria and Denmark, referendums on European integration have taken place at the discretion of governments.  Most of the new constitutions in Central and Eastern Europe also contain provisions for referendums.
21 See ILRES web site,, accessed April 19, 2006.
22 Special Eurobarometer 214 “The Future Constitutional Treaty” (2005).
23 Hobolt observes that “while voters have very little knowledge on the specific ballot proposal, they can still make informed choices by relying on heuristic shortcuts, such as elite endorsements and campaign cues” (Hobolt 2006: 156).
24 For example, the blog by journalists at La Libre Belgique:, accessed on April 14, 2006 or by journalists at De Standaard:, accessed on April 14, 2006.
25 For background, see Jürgen Habermas’s influential writing about the public sphere (Habermas 1989) and for a more recent application to the EU (see Habermas 1996).
26 See the statistics for Belgium at:, accessed on April 22, 2006.
27 In Luxembourg (even in the case of referendums) and Belgium, voting is compulsory and since 1979 the turnout is more or less 90%. In The Netherlands, it was 39.3% for the elections of 2004 and 62.8% for the referendum.
28 For instance, the Dutch government set the two following conditions: a turn-out superior to 30% and an unambiguous outcome (i.e. either 60% in favor of the “Yes” or 60% in favor of the “No”).
29 Mark Leonard, “Why Europe Will Run the 21st Century?,” Boston University, March 14, 2005.

To cite this article

Min Reuchamps, «Referendum as a Tool for Building European Identity. The Case of Belgium, Luxembourg, and The Netherlands», Cahiers de Science politique [En ligne], Cahier n°6, URL :

About: Min Reuchamps

Aspirant F.R.S.-FNRS, Département de sciences politiques, Université de Liège