EUROSCEPTICISM: A CONCEPTUAL ANALYSIS AND A LONGITUDINAL, CROSS-COUNTRY EXAMINATION OF PUBLIC SCEPTICISM TOWARDS THE EUROPEAN UNION PhD Thesis by Catharina Srensen Danish Institute for International Studies/Institute of Political Science, University of Copenhagen Submitted April 30, 2007 Main supervisor: Professor Morten Kelstrup Institute of Political Science, University of Copenhagen Co-supervisor: Department leader Gorm Rye Olsen Department of Society and Globalisation, Roskilde University Acknowledgements This thesis would not have been possible without inspiration from fellow scholars and the PhD grant from the Danish Institute for International Studies. In particular, I would like to thank my supervisors, Morten Kelstrup and Gorm Rye Olsen, for their invaluable support. I am grateful to Cristiano Vezzoni, Jrgen Elklit, Hans Jrgen Nielsen, Mogens Dam, Morten Lihn Jrgensen and Kasper Mller Hansen for assistance with SPSS and comments on the statistical analyses, and to members of the `EU PhD network' (Rebecca Adler-Nissen, Sevasti Chatzopoulou, Ulrik Pram Gad, Caroline Grn, Rasmus Leander Nielsen, Rasmus Brun Pedersen, and Maja Mller Sousa) for helpful and thorough remarks to several chapters. Comments, support, reflections, and/or assistance with access to data material from Martin Marcussen, Palle Svensson, and Ian Manners, as well as from Simon Atkinson, David Baker, Erik Beukel, Thomas Christiansen, Dominic Cummings, Christopher Flood, Dieter Fuchs, Mark Gill, Stefano Guzzini, Ulla Holm, Liesbet Hooghe, David Larsen, Andrew Moravscik, Roger Mortimore, Hans Mouritzen, Cas Mudde, Julie Pruzan-Jrgensen, Louise Srensen, Yvonne Srensen and Marlene Wind, are also acknowledged with gratitude. Responsibility for the thesis' contents is of course fully my own. Finally, I would like to thank the Institute of Political Science at the University of Copenhagen, and my colleagues at the Danish Institute for International Studies, for providing me with a friendly and inspirational research environment. 1 Table of contents: Acknowledgements ....................................................................................................................... 1 List of tables, figures and graphs .................................................................................................. 5 Abstract.......................................................................................................................................... 7 Part One: Introducing euroscepticism.......................................................................................... 9 Chapter 1. The emerging focus on euroscepticism...........................................................................9 1.1 Background....................................................................................................................11 1.2 Motivation and research questions.................................................................................15 1.3 Terminology...................................................................................................................21 1.3.1 European integration and the European Union ............................................................21 1.3.2 Public opinion................................................................................................................23 1.4 Outline of the thesis .......................................................................................................25 Chapter 2. Methodological framework...........................................................................................28 2.1 Conceptualising euroscepticism.....................................................................................28 2.1.1 What is a concept?.........................................................................................................28 2.1.2 Concepts and the dependent variable problem..............................................................30 2.1.3 On solving the dependent variable problem ..................................................................32 2.2 The three-level structure ................................................................................................35 2.2.1 The basic level................................................................................................................35 2.2.2 Secondary level types.....................................................................................................37 2.2.3 The indicator level: concept operationalisation............................................................38 2.2.4 Method of aggregation...................................................................................................39 2.3 Comparing euroscepticism: the method of structured, focused comparison .................42 2.3.1 Country selection ...........................................................................................................43 2.3.2 Context specificity..........................................................................................................48 2.3.3 Time perspective ............................................................................................................49 2.4 Epistemological and ontological assumptions...............................................................50 2.4.1 Explanation and theory..................................................................................................50 2.4.2 A critical rationalistic framework..................................................................................51 2.4.3 On deducing constitutive types of euroscepticism and parsimony ................................54 Part Two: Constituting euroscepticism ...................................................................................... 56 Chapter 3. Defining euroscepticism: the basic level.......................................................................56 3.1 A critique of the state of the art in defining euroscepticism..........................................56 3.2 A definition ....................................................................................................................61 Chapter 4. Types of euroscepticism: the secondary level...............................................................65 4.1 Euroscepticism theory....................................................................................................65 4.2 Euroscepticism and EU legitimacy................................................................................69 4.2.1 Approaches to EU legitimacy ........................................................................................70 4.2.2 Legitimacy deficits and euroscepticism .........................................................................74 4.2.3 Vertical and horizontal euroscepticism .........................................................................75 4.3 Utilitarianism .................................................................................................................79 4.3.1 Contribution...................................................................................................................87 4.4 The Silent Revolution and ideological euroscepticism..................................................88 2 4.4.1 Contribution...................................................................................................................91 4.5 National integrity ...........................................................................................................91 4.5.1 Exclusionary populism...................................................................................................97 4.5.2 Contribution.................................................................................................................100 4.6 The national context.....................................................................................................101 4.6.1 Contribution.................................................................................................................107 4.7 The democratic deficit .................................................................................................108 4.7.1 Versions of the deficit...................................................................................................109 4.7.2 Defence of the deficit....................................................................................................111 4.7.3 Scepticism about the deficit..........................................................................................112 4.7.4 Contribution.................................................................................................................112 4.8 Political contestation....................................................................................................113 4.8.1 Contribution.................................................................................................................118 4.9 Constitutive types: a summary.....................................................................................119 Part Three: Analysing euroscepticism ....................................................................................... 121 Chapter 5. Operationalisation: the indicator level ........................................................................121 5.1 The data-set: advantages and shortcomings.................................................................122 5.2 Indicator selection and handling ..................................................................................127 5.2.1 Relativity ......................................................................................................................129 5.2.2 Durability.....................................................................................................................131 5.2.3 Data validity.................................................................................................................131 5.3 Establishing, and dealing with, a lack of corroboration...............................................135 5.4 Summary of operationalisation and observables .........................................................137 5.5 Statistical tests and results............................................................................................141 5.5.1 Intra-type associations (coherence).............................................................................141 5.5.2 Status............................................................................................................................147 5.5.3 Inter-type associations (independence) .......................................................................147 5.5.4 Statistical conclusions..................................................................................................152 Chapter 6. Measuring euroscepticism in Denmark, France and the United Kingdom .................153 6.1 Utilitarian euroscepticism: economic benefit ..............................................................153 6.1.1 Summary.......................................................................................................................161 6.2 Sovereignty-based euroscepticism...............................................................................163 6.2.1 Summary.......................................................................................................................171 6.3 Democratic euroscepticism..........................................................................................173 6.3.1 Summary.......................................................................................................................178 6.4 Social euroscepticism...................................................................................................181 6.4.1 Summary.......................................................................................................................187 6.5 The level of hard euroscepticism .................................................................................189 6.5.1 Summary.......................................................................................................................193 6.6 Degree of hard euroscepticism within the four types ..................................................195 6.6.1 Discussion....................................................................................................................198 6.7 Comparative findings...................................................................................................199 6.7.1 Euroscepticism, an overview........................................................................................200 6.7.2 Euroscepticism walkthrough........................................................................................205 6.7.3 Analysing the development of euroscepticism .............................................................211 3 Part Four: Consequences and Conclusions ...............................................................................219 Chapter 7. Euroscepticism revisited..............................................................................................219 7.1 Revisiting the dependent variable problem..................................................................220 7.2 Bearings on research into public EU attitudes.............................................................224 7.3 Multifaceted and contradictory: consequences of contemporary euroscepticism .......231 7.3.1 Euroscepticism and EU visions ...................................................................................231 7.3.2 Euroscepticism and Eastern Enlargement...................................................................234 7.3.3 Euroscepticism and initiatives during the Reflection Period.......................................238 7.4 Evaluation and suggestions for further research..........................................................242 7.4.1 Evaluation....................................................................................................................243 7.4.2 Suggestions for further research..................................................................................245 Bibliography ...............................................................................................................................249 Appendix: One-dimensional Eurobarometer indicators ...........................................................272 4 List of tables, figures and graphs No. Figures Page Tables Page Graphs Page 1 Research Foci 19 Main research concerns 26 Feeling of no benefit 153 2 The Ogden- Richards Triangle 29 Examples of the dependent variables engaged in existing literature 30 Meaning of the EU: Waste of money 156 3 The three-level structure of the concept of euroscepticism 41 How many cases? 44 Bad effect in specific areas 157 4 The research cycle 53 Kopecky and Mudde's typology of euroscepticism 60 The economic situation: negative role 160 5 Possible configurations of types of euroscepticism 55 Flood and Usherwooďs labelling of positions towards the EU or EU developments 61 Against an EU government 163 6 Vertical and horizontal euroscepticism 76 Five theoretical expectations as to the nature of euroscepticism 119 The EU does not need a government 164 7 The concept of euroscepticism 201 Operationalisation 137 The EU Constitution: against because of loss of national sovereignty 167 8 Beetham and Lorďs concept of EU legitimacy 224 Contingency table (identity threat by feeling of no benefit) 149 Integration is a threat to national identity 168 9 Contingency table (no EU government by membership bad thing) 196 Decision-making level: national 1999 169 10 Overview of relative euroscepticism 202 Decision-making level: national 2001 170 11 Overview of the development of euroscepticism 204 Dissatisfaction with EU democracy 173 12 Movement in EU attitudes 215- 216 The EU Constitution: not democratic enough 175 5 13 Euroscepticism and Eastern Enlargement 237 The Community is not democratic 176 14 The European Parliament does not protect interests 177 15 EU priorities: more social Europe 182 16 The EU Constitution: not social enough 183 17 Fear of losing social benefits 185 18 More cooperation in social matters 187 19 EU membership is a bad thing 189 20 The EU Constitution: against because of rejections of the EU 191 21 Feeling the EU inspires: rejecting it 192 22 EU referendum: leave the EU 193 23 Dynamics of euroscepticism: EU 207 24 Dynamics of euroscepticism: DK 208 25 Dynamics of euroscepticism: FR 209 26 Dynamics of euroscepticism: UK 210 27 Attitudes prior to Eastern Enlargement 235 6 Abstract This PhD thesis examines public euroscepticism, the new veto player in the European integration process, whose seeming multifarious nature thus far has escaped systematic conceptualisation. It argues that existent literature is scattered and contradictory because of the `dependent variable problem', namely the extensive conceptual ambiguities as to what euroscepticism is. By solving this fundamental problem, the thesis aims to facilitate the accumulation of knowledge on euroscepticism and improve our comprehension of public reactions to EU developments and events. The thesis examines the extent to which, and how, euroscepticism is multifaceted. It does so by conducting a concept analysis with three levels. A definitional level builds on existing definitions to define the term `public euroscepticism' as `sentiments of disapproval--reaching a certain degree and durability--directed towards the EU in its entirety or towards particular policy areas or developments'. Next, a constitutive level identifies and deduces the ensemble of characteristics associated with public euroscepticism in existent literature, in order to establish the platform for developing the thesis' conceptualisation. Third, an indicator level specifies each theoretical type of euroscepticism in such detail that data can be gathered and analysed. To confirm or reject the types, statistical tests of the coherence and independence of each type are run. The result of this endeavour is a four-fold typology: Euroscepticism can be `economic', `sovereignty-baseď, `democratic' and/or `sociaľ of nature. These four types of euroscepticism are subsequently examined through a longitudinal comparative analysis of three case countries, namely Denmark, France and the United Kingdom, in comparison with the EU average. The aim is to demonstrate the extent to which, and how, euroscepticism differs across member states, as well as in what ways contemporary patterns of euroscepticism are different from past manifestations. The diverse portraits and patterns that result from this endeavour are argued to present the EU with a `win-lose dilemma of euroscepticism', which contributes to an explanation of how initiatives that the Union has undertaken in recent years have at one and the same time increased scepticism in some member states and reduced it in others. Indeed, precisely that area where one population wishes the EU to focus, risks being that area where another population fears its influence. It will depend on the type(s) of euroscepticism characterising a country. Having confirmed the typology on this basis of theoretical, statistical and empirical evidence, the thesis applies it to the existing literature in order to explain why it has failed to capture the nature and dynamics of euroscepticism. It is shown how the typology improves the cumulability of empirical and theoretical contributions to the study of EU attitudes, and how it explains the popular setting with 7 regard to events on the Union's agenda. In conclusion, the consequences of the new framework for the field of euroscepticism research, and for the EU's endeavour to be close to its citizens, are evaluated, and fertile strategies for future research are identified. It has been easier to deal with elites and elite attitudes. But what has Europe meant to the electorate? Alan Milward (1997: 17) `Europe began as an elitist project in which it was believed that all was required was to convince the decision-makers. That phase of benign despotism is over'. Jacques Delors (in Leonard 1998: 6) Citizens' attitudes towards European integration and the EU's institutions and policies are increasingly important. Simon Hix (2005: 147) Democratic legitimacy also means a Europe which listens to the expectations of its citizens and addresses their concerns through adequate policies. For any of its policies, including enlargement, the EU has to win the support of its citizens. European Commission, Enlargement report (2006: 23) 8 Part One: Introducing euroscepticism 1 The emerging focus on euroscepticism By the eve of 2nd June 1992, it was all too clear to the pro-EU Danish government and the European community at large that euroscepticism, a hitherto largely unrecognised phenomenon, was a powerful force to be reckoned with. The Danes had sensationally voted `no' to the Maastricht Treaty and made it obvious that limitations to the project and prospect of further European integration did no longer pivot around political negotiations at high-level intergovernmental summits (for instance Hix 1998; Weiler 1999; Medrano 2003; Fllesdal 2004).1 While the latter had long been at the centre of the European agenda with numerous studies devoted to its resolution, few researchers and politicians had attributed much importance to public opinion (Sinnott 1995; Milner 2000). Still today, the literature is broadly characterised by highly theoretical studies of the integration process at `elite leveľ on the one side and highly empirical studies of public attitudes on the other side. The problem with this is, of course, that many theoretical accounts have a tendency to rush away from empirical data with the result that their findings are inadequate at accounting for public attitudes towards integration, and conversely that many empirical studies into public opinion have difficultly in capturing the larger perspective. Examining one account without the other is likely to paint an incomplete portrait of a phenomenon and concept,2 which is both durable and multifaceted. Taking its point of departure in this context, the present thesis seeks to complement a thorough theoretical investigation of the nature of euroscepticism with an analysis of sceptical citizen attitudes, as these are expressed in public opinion polls. It will do so by first contextualising and consequently conceptualising the phenomenon: it starts by offering an account of the background and geography of euroscepticism, and proceeds with a theoretical inquiry into its constitutive types--an inquiry, which is subsequently subjected to systematic empirical scrutiny.3 1 When talking about the `limitations' or `constraints' posed by euroscepticism on the EU, the thesis does not wish to imply that euroscepticism is only of a constraining nature--indeed, this would be misleading, as euroscepticism may also inspire renewed impetus in the EU--nor, of course, does it wish to pass a normative judgement on euroscepticism as a negative (or positive) phenomenon. The thesis adopts the word because its interest lies with studying the segments of public opinion that show a desire to change or stop, and thereby constrain, the current, actual formulation of integration or particular developments on the EU's agenda, and thus not in what potential policy consequences this may inspire. 2 The thesis defines `phenomenon' as an observable event or process susceptible of scientific description or explanation. `Concepť is defined as a data storage container that represents similarities or common characteristics in phenomena (see Fortune and Reid 1999); Section 2.1.1 returns to the discussion of what a concept is). 3 Inquiries into the route by which individuals have reached their stance are thus not undertaken by the thesis. For a discussion of the role of the media in attitude formation about the EU, see for instance de Vreese and Boomgaarden 2006. 9 To guide the remaining sections in Part One, the thesis offers here (in a self-consciously premature manner) the semantic definition of the term public euroscepticism introduced and discussed by the thesis in Chapter 3. Euroscepticism can be defined as `sentiments of disapproval--reaching a certain degree and durability--directed towards the European Union (EU) in its entirety or towards particular policy areas or developments in the Union'. As the thesis further elaborates upon in Chapter 3, it is important to note that it includes under the label `euroscepticism' sentiments that are sceptical towards part of the `EU of the day', whilst still being supportive towards the issue of membership and/or a strong European Union. The thesis thus rejects the distinction by some scholars and commentators between euroscepticism on the one side and `constructive criticism' on the other. Indeed, the thesis does not find it relevant to reserve a study of euroscepticism only to the segment of the population who reject the very idea of membership or European co-operation. Not only do opinion polls show this segment to be rather small, such a distinction would furthermore exclude a considerable group of `no' voters in EU referenda from the label--and as the thesis holds that some form of scepticism towards some aspect of integration is reflected in the act of rejecting, say, an EU treaty, it accordingly defines euroscepticism in a broad, inclusive manner.4 While it could be argued that euroscepticism had been embryonic for decades at least,5 its birth in June 1992 came as a surprise to the generally pro-EU governments across the member states. Prompting an instantaneous wave of self-conscious attempts at justifying the merits of the integration process, it also necessitated an acute awareness of the large abyss that had seemingly existed between `Brussels' and the broader populations. `One might wonder', as Schmitt and Thomassen have done, `whether the governments and politicians responsible for the Maastricht Treaty were living in the same European world as the people they were supposed to represenť (Schmitt and Thomassen 1999: 4). A widespread reaction by pro-EU politicians to this emerging insight was that `communicating the EU to its citizens' would be the key to counterbalancing their scepticism--a task which by now has become institutionalized within the European Commission.6 It is noteworthy that this task has often been construed in terms of the need to reconnect citizens to the EU, without substantiation of when in the EU's history citizens had, in fact, been `connecteď, what this `connection' implied, and why it became broken.7 4 As Chapter 3 develops, the term euroscepticism is problematic. The thesis' inclusive definition is adopted because it is seen as the most appropriate in light of the ambition to examine what type(s) of public disapproval the EU encounters in its member states. 5 Eurobarometer polls account from their outset in 1973 of public scepticism in several member states. 6 In the shape of a Commissioner dedicated to this task. Communication issues have moreover emerged at the top of the entire Commission's agenda. 7 References of the need to reconnect citizens abound; especially in policy statements: for instance Commissioner Siim Kallas (2005): Speech/06/280; European Commission, COM(2005)551. See also Collignon 2006. 10 The success of these communicative efforts in combating euroscepticism have at best been doubtful, and persisting scepticism has become an increasing concern in the EU, with direct bearing on the process of further integration. In the early summer of 2005, for instance, the public euroscepticism voiced at the French and the Dutch referenda on the EU's Constitutional Treaty prompted the Union's leaders to call for a lengthy `period of reflection', which still characterises the climate of 2007.8 Despite the potency of euroscepticism, it has remained an elusive concept--even in academic circles, where tautologies are just below the surface: if, for instance, the feeling of being poorly represented in the European Parliament (EP) is understood to breed euroscepticism (for instance Rohrschneider 2002), we may ask ourselves whether this very concern in itself is not euroscepticism. The elusiveness surrounding euroscepticism is sustained and intensified by often contradictory findings in the literature as to what, when and why it is. We may think of the disagreements as to whether a favourable domestic economic climate in fact breeds EU­support (Andersen and Kaltenthaler 1996) or scepticism (Rohrschneider 2002); whether post-materialism is positively (Inglehart 1971) or negatively (Andersen and Reichert 1996) related to supportive EU attitudes; as well as in what sense sociodemographic variables, such as gender and age, are related to scepticism (Andersen and Reichert 1996, for instance, hold that these variables are not related to scepticism; Siune 1993 and de Vreese 2004 hold that they are). Rectifying the confusion regarding the nature of euroscepticism is a primary motivation behind the thesis' research questions. However, prior to explicating these, a further note on the contextual backdrop of euroscepticism is offered. Background Although the margin between the `yes' and `no' votes had been small in the Danish 1992 referendum-- about one percentage point--photographs of the small, celebrating country that had `dared to say no to Brussels' were abundant in all member states. From one day to the other, or so it seemed, the `permissive consensus' (Lindberg and Scheingold 1970, developing an 1961 argument from V.O. Key) amongst Europe's populations in favour of integration had collapsed (for instance Dinan 1994; Hix 1998, 2005; Gabel 1998c; Medrano 2003; Fllesdal 2004).9 Whereas European leaders, at least on the 8 Marlene Wind points out that the period of reflection was initiated by the European Council on 16th June 2005, without clear indication as to its purpose (Wind 2006: 13-14). 9 This hypothesised end of the permissive consensus is usually seen as the trigger in academic circles for `bringing public opinion back in'; however, some scholars prefer to emphasise the events of 1989 as the original catalyst (Sinnott 1995: Chapter 2). 11 Continent,10 in the 1960s, 1970s and 1980s were able to rely on what was seen as positive prevailing attitudes towards integration, or simply on a lack of interest allowing developments to take place without much opposition, such options were no longer available to their colleagues in the 1990s.11 In a self-confident gesture, then French President, Francois Mitterrand, offered to re-assure European leaders of public dedication to the EU, by letting the French demonstrate their support to the Treaty in a referendum in the autumn of 1992 (for instance Nicoll and Salmon 2001: 416). The narrow `yes' that emerged here (51 percent), however, served only to cement the political realisation that euroscepticism was both more prevalent and powerful than had hitherto been realised. Ominously, euroscepticism sparked an interrogation into the possibility that public support for the EU had reached a ceiling--a natural limit--even in countries, which had traditionally played a central role in furthering European integration (Milner 2000: 6). These developments followed a series of momentous events in European history. With the fall of the Berlin Wall in 1989, the ensuing unification of East and West Germany, and the dissolution of the Soviet Union, the foundations and perspectives for further integration of the EU changed fundamentally. Enlargement with a considerable number of newly independent states had become a real possibility, and necessitated major reforms of the Community's internal make-up. The transition from Community to Union has to be seen in this light, with the new Treaty on European Union (the Maastricht Treaty) introducing far-reaching components particularly strengthening the supranational aspects of co-operation. The EU after Maastricht was to be united by a single currency and its citizens to be holders of a European citizenship. The number of policy areas under EU competence was increased and co-decision introduced, innovating the roles of the Commission (as sole initiator) and the European Parliament (as co-decider with the Council of Ministers), and making qualified majority voting the norm.12 In reaction to the public rejection of this Treaty, Danish politicians drew up four opt-outs that centred on the EU's supranational, `state-like' aspects: the single currency, defence, justice and home affairs, and citizenship. This could reflect a view of euroscepticism as a largely sovereignty-based 10 Denmark and the United Kingdom arguably shared a reputation for being eurosceptic since their entry into the EC in 1973. By rejecting accession in a referendum in 1972, the Norwegian population also demonstrated early euroscepticism, as Greenland has done, by in fact withdrawing from the Union in 1985. 11 Ole Wver speaks of a shift of referent object from `state' to `nation': today's constraints on integration are no longer exclusively set at state level. They are increasingly also set by societies--or at the nation level. Wver argues that it is arguments about the survival of the nation that are mobilised and speaks of `security identities'--a discussion the thesis returns to in Part Two, Section 4.5 (Wver 1996: 111-112). 12 See http://europa.eu/scadplus/treaties/maastricht_en.htm for an introduction to the Maastricht Treaty. 12 phenomenon;13 however, it seems fair to argue that the four opt-outs were drafted in an improvised manner. The `yes' parties and the Danish Socialist People's Party (whose shift from advising a `no' to campaigning for a `yes' was seen as crucial to winning a new referendum on the so-called Edinburgh Agreement14 ) had to relate to public dissatisfaction (for instance Christophersen,15 also Danish Institute of International Affairs 2000), without any substantial empirical analyses available on the topic. There is little evidence from opinion polls to confirm whether or not the substance of the opt-outs was in fact taking Danish public concerns into account.16 Interestingly, this `sovereignty view' on euroscepticism did not come to dominate the subsequent debates amongst EU leaders--nor academics--on public support. Indeed, also provoked by the German Constitutional Court ruling on the Maastricht Treaty,17 these largely centred on diagnosing and rectifying what was understood to be two critical deficiencies of the Union: its `democratic deficiť and its `information deficiť. As will be developed in Section 4.7, the former carries with it the assumption that the EU has democratic credentials--which it falls short of fulfilling--while the latter refers not so much to a structural deficit, but to the perceived failure of the political system to adequately inform the public of the merits of its policy making (see especially Section 4.3). A brief and non-exhaustive list of the political consequences of this understanding includes the gradual strengthening of the European Parliament to improve the EU's democratic standing; the emergence of communication issues at the top of the Commission's priorities; the attempted (re)invigoration of concepts such as subsidiarity and transparency; the increasing role of the European Ombudsman; the invention of the Convention method; and the increasing use of referenda--virtually introducing the public as a veto player on certain issues. Recently, a `D-plan'18 was proposed as a strategy of winning the hearts of the European populations, D standing for democracy, debate and 13 Indeed, a widespread understanding of the Danish Maastricht campaign seemed to be that `yes' voters focused on economic arguments, while `no' voters focused on sovereignty arguments. The thesis will show that this dichotomous understanding is misleading. First, it is clearly possible to be eurosceptic on economic grounds; second, other types of euroscepticism interact with economic and sovereignty-based concerns. 14 See for instance the EU Information Centre of the Danish Parliament: http://www.eu- oplysningen.dk/dokumenter/traktat/eu/edinburgh/. 15 Speech at the `Europa-Konference' (the European Commission's Representation in Denmark), 27th May 2005. 16 See also Worre 1995: 253, arguing that there would have been a majority in favour of full CFSP-participation in both 1992 and 1993. 17 Bundesverfassungsgericht, Judgement of 12th October 1993. 18 Originally formulated by Commissioner for `communicating the EU to its citizens', Margot Wallström. Advocated also by Commission President José Barroso, and mentioned as a central initiative by then acting president of the EU, Luxembourg's Prime Minister Jean-Claude Juncker, at the press conference following the EU summit in mid-June 2005 on how to proceed with the ratification process of the Constitutional Treaty after the French and Dutch no's. The thesis returns to discuss Plan D in Chapter 7. 13 dialogue.19 Central to these initiatives and strategies, as the thesis returns to below, lies the far from thoroughly analysed belief that democratic innovations and information campaigns about the EU are the main foundation for public support. Another consequence of increased political awareness of euroscepticism has been that national heads of state or government use public opinion as a bargaining chip at intergovernmental conferences. References to a `euroscepticaľ public are, in other words, evoked in order to justify and secure national red lines. This was for instance the case of British negotiators in the late stages of negotiations on the Constitutional Treaty, and, in the months leading up to the French referendum, the understanding that led French President Jacques Chirac to succeed in taking the service directive off the EU's imminent agenda.20 In December 2006, with regard to a proposal by the then Finnish EU Presidency to abolish the veto in co-operation on Justice and Home Affairs, Swedish Prime Minister Fredrik Reinfeldt (The Moderate Party) justified his rejection of this move to easing decision-making by stating: `We do not have the support of the public to go through with this' (quoted in EUobserver, 15th December 2006). Have these endeavours to bridge the frequently declared EU-citizen gap helped? A number of occurrences suggest otherwise. According to one study, only a few of the arguments used by political parties and movements up to the Danish referendum on the introduction of the Euro in 2000--which resulted in a solid `no'--were deemed trustworthy by the electorate (Andersen 2003). In Ireland, the referendum on the Nice Treaty in June 2001 attracted only 35 percent of the electorate (less than a million people and by far the lowest turnout in the history of Irish EC/EU referenda--seven in total between 1972 and 2002), hardly signalling that EU issues had captured the hearts of citizens. That vote also saw a solid `no'. Sweden decided to submit its opt-out on the single currency to a referendum held in 2003, thus after the Euro had become a physical reality for millions of Europeans. Again, a `no'. With the French and the Dutch `no' votes in 2005, strong public euroscepticism once again shocked a largely unprepared European Union. In total, more than half of the referenda on EU issues held in the 15 old member states in this decade have suggested the persisting prevalence, and perhaps even increase, of public euroscepticism.21 19 See Adler-Nissen and Knudsen (2005) for a study of the EU's Constitutional Treaty from the perspective of theories of democracy. 20 See for instance EUobserver: `Chirac reopens attack on `unacceptable' services directive', 16th March 2005, and EUobserver: `Paris and Berlin to present alternative to controversial services law', 27th April 2005. 21 The thesis sees `no' votes as expressions of euroscepticism. Eight referenda on EU issues (not including accession referenda) were held in the 15 old member states between 2000 and 2006. Five resulted in a `no' vote (underlined): 2000: Denmark voted `no' to the introduction of the single currency; 2001: Ireland voted `no' to the Nice Treaty; 2002: Ireland voted `yes' to the Nice Treaty; 2003: Sweden voted `no' to the introduction of the single currency; 2005: Spain voted `yes' to the Constitutional Treaty; 2005: France voted `no' to the Constitutional Treaty; 2005: The Netherlands voted `no' to the 14 With hindsight, it was pointed out that the pro-Nice parties in Ireland had been too complacent about the 2001 referendum, taking for granted that the population would endorse it, and thus not prioritising the `yes' campaign (Grabbe 2001), while a well organised anti-EU campaign successfully campaigned about the Treaty's alleged negative effect on Irish sovereignty (Miller 2001: 14). Also with hindsight, it was pointed out that Jacques Chirac was unlikely to have ever secured a `yes' vote in France given his then personal unpopularity with the French population and the manner in which he was re-elected as President in April 2002.22 Indeed, as we shall see in Chapter 4, the theoretical inquiry, a powerful interpretation of euroscepticism has been the `protest thesis', holding that the popularity of governments and their time since entry into power are crucial determinants at EU elections. It does, however, appear premature to immediately cast aside `no' votes as a result of the neglect of information about the EU or a consequence of particular domestic concerns. In fact, doing so testifies to the lack of a comprehensive political and academic understanding of euroscepticism--a lack which arguably contributes to, and indeed sustains, the existence of a gap between governments and mass attitudes towards the EU.23 This would explain why so many referenda on European integration have gone astray for EU leaders; why new mass-based anti-EU protest movements have emerged; and why citizen engagement is still difficult to mobilise at times of European Parliamentary elections (Hix 2005a: 166). Motivation and research questions Many indicators suggest that euroscepticism is a diverse and erratic phenomenon. Diversity was recently demonstrated through French and Spanish public attitudes toward the Constitutional Treaty (at the Spanish referendum 17 percent voted `no'--at the French referendum 55 percent voted `no'24 ), which moreover pointed to the changeability of public opinion with marked scepticism emerging in the Netherlands, a member state population previously characterised as EU cheerleaders. Changes in Constitutional Treaty; 2005: Luxembourg voted `yes' to the Constitutional Treaty. This is in contrast to the results of the eight referenda prior to 2000, where only one resulted in a `no' vote (underlined). 1986: Danes accepted the Single European Act; 1987: the Act was endorsed by the Irish; 1992: Danes rejected the Maastricht Treaty; 1992: the Irish voted `yes' to the Maastricht Treaty; 1992: the French accept the Treaty; 1993: Denmark votes `yes' to the Edinburgh compromise; 1998: Denmark votes `yes' to the Amsterdam Treaty; 1998: the Irish accept the Amsterdam Treaty. Source: the Danish EU Information Office (www.euo.dk). 22 His opponent in the final round of the election was unexpectedly the leader of the extreme right party Front National, Jean-Marie Le Pen, which `forceď many socialist voters to demonstrate their lack of support for Le Pen through a vote for Chirac. Prime Minister at the time--Jean-Pierre Raffarin--was also acutely unpopular in national opinion polls. 23 It may be useful to also view this as a seemingly growing gap between what politicians are traditionally expected to live up to and their actual ability to fulfil these expectations in today's `polycentric' societies (where the EU assumes a defining role)--leading not just to a discontent with individual leaders but also to a possibly more profound structural discontent with politics (see Kelstrup 1996: 10, 12). 15 popular attitudes to the EU are also visible in Denmark, where an otherwise eurosceptical reputation has more recently come under pressure from widespread public endorsement of the issue of membership as well as of several other EU issues.25 Such developments have exposed the inadequacy of existing research on public euroscepticism in coming to terms with the phenomenon. Despite many and obvious manifestations--and implications--most vividly and dramatically demonstrated through national referenda, the contours and borderlines of euroscepticism remain elusive both politically and academically (Flood 2002: 2; de Vreese 2004; Ray 2007). To begin with the former, years of extensive and expensive information campaigns and efforts to democratise the EU have had questionable success in countering euroscepticism. Indeed, there are few signs that public opinion throughout the European Union generally is becoming more positively inclined, neither towards the EU as a whole nor toward specific developments. Public opinion, as the thesis returns to below, in fact seems to be moving in the opposite direction, with support for EU membership, for instance, decreasing in most member states. In the view of several observers, commenting on the French and the Dutch referenda (Whitman 2005: 673; Dybkjr26 ), it is `puzzling' and `ironic' that public euroscepticism reached a peak in connection with the birth of the Constitutional Treaty; a document drafted according to a new, more open method (the European Convention), signalling increased attention to the endeavour of democratising the EU. Indeed, as Andreas Fllesdal writes, the Convention method was precisely intended as a pre-emptive response to growing fears amongst EU politicians `that Europeans might refuse to accept future steps toward deeper integration' (Fllesdal 2004: 5). Academically, research into euroscepticism has over the past decades increased in volume while remaining patchy and contradictory, and as such no coherent theory exists that details what it is, or why, when and how it occurs and develops. As Liesbet Hooghe and Gary Marks observe, a direct consequence of this shortcoming was the failure by the academic community to `predict the rise of euroscepticism' (Hooghe and Marks forthcoming 2007) that was arguably witnessed by the French and the Dutch referenda. 24 EU Information Centre of the Danish Parliament: http://www.euo.dk/spsv/off/alle/afstemning/. 25 Recent Eurobarometer polls show that since 1994, the Danes and the Spanish are the only populations to have increasingly endorsed the issue of membership of the EU. Over the past few years, Danes have consistently been favourable to giving up the Danish opt-outs, and prior to the no's in France and the Netherlands, there was marked public support for the Constitutional Treaty in Denmark. Moreover, Danes were consistently among the biggest supporters of the EU's Eastern Enlargement (Eurobarometer trends; Eurobarometer 60). Naturally, the thesis goes more into depths with this possible development below. 26 Speech at the `Europa-Konference' (the European Commission's Representation in Denmark), 27th May 2005. 16 Puzzlement arises, which the thesis aims to address: why is it that existent research falls short in coming to terms with public euroscepticism? Why has euroscepticism escaped theoretical and empirical delineation? Is it a consequence of there being as many reasons to be eurosceptic as there are members of the public? The thesis suggests, rather, that a main reason behind the inadequacy of much existent research resides in the sheer number and diversity of conceptualisations of the dependent variable of euroscepticism (or at the other side of the continuum of EU attitudes, citizen support). Throughout the thesis, this will be referred to as the `dependent variable problem'. Its implications are significant and explain many of the shortcomings of existing approaches. Indeed, imprecise definitions of the dependent variable mean that most studies have inadvertently surveyed different phenomena altogether; a fact which to a considerable degree explains their sometimes contradictory findings. Conceptual disagreements, in other words, have hindered the accumulation of knowledge (see Sartori 1984; Weyland 2001): if we do not grasp the nature of the study object, we are hardly ready or able to engage in explanatory, causal analysis. In response, this thesis offers a meticulous concept analysis of euroscepticism. By establishing theoretically-derived and statistically tested indicators that gauge various independent types of euroscepticism, we gain not merely increased cumulability, 27 but essential foundational work for future research into the underlying causes of disapproving public EU attitudes. From this summarised background, which elucidates the motivation behind the present research into aggregate level public euroscepticism, four main research questions may be formulated: * To what extent, and how, is euroscepticism a multifaceted phenomenon? * How do patterns of euroscepticism differ between member states? And in what ways do contemporary patterns of euroscepticism differ from past manifestations? * Why has existing research failed to capture the nature and dynamics of euroscepticism? * What consequences does a refined understanding of euroscepticism have for the research field, and for the European integration process? 27 Below the thesis will refer to this as the `cumulability' of research. To cumulate means to increase or heap together. By cumulability the thesis refers to the ability of new studies to depart from, and build on, existing studies on the same phenomenon in order to increase our knowledge about it. 17 This `explanandum' is approached through the in-depth study of the aggregate attitudes of citizens in the member states of the EU, as expressed in public opinion polls. As such, the thesis' research questions inspire four analytical foci. A first, explorative, is concerned with narrowing down contradictions and inadequacies in existent understandings of euroscepticism--as such, a large scale examination of theoretical literature on the topic is undertaken to fill the gap in our existent knowledge. A second focus reflects a conceptual ambition:28 It is concerned with identifying the contours of different facets of euroscepticism in the literature, and developing these into coherent types in order to propose a typology29 of euroscepticism. Embedded in this endeavour is the task of constructing multiple indicators of each type of euroscepticism in order to create statistical measures for testing the validity and solidity of the typology. The thesis develops, in other words, the statistical means with which we can confirm or reject theoretical expectations as to the nature of euroscepticism. It should thus already be emphasised here that the thesis departs from the hypothesis that euroscepticism is multifaceted--most complex concepts are--wherefore the confirmation of this hypothesis (although presently not thoroughly documented) is not what is really novel. What is novel is the discovery of in what ways, and how, euroscepticism is multifaceted. The third focus is comparative, as it aims to empirically establish the prevalence of the confirmed typology of euroscepticism in a number of member states, and further compare the phenomenon geographically as well as historically. The thesis thus analyses the complete euroscepticism situation of selected countries from the 1970s to today, and builds on this analysis to explain the popular setting for recent EU endeavours, such as Eastern Enlargement and Plan D. Finally, a fourth focus is evaluative--aiming to discern the consequences of the thesis' new understanding of euroscepticism to theoretical debates on public EU attitudes. Indeed, the suggestion of a solution to the dependent variable problem gives us a framework for re-evaluating the contributions of existent theoretical and empirical research on euroscepticism,30 as well as a platform for proposing how future research into euroscepticism can be pursued. In Chapter 2, following clarification of the thesis' methodological approach and epistemological assumptions, a detailed research design is presented. Below is presented an introductory illustration of its four research foci: 28 As the thesis will argue in Section 2.4.1 on epistemological assumptions, this conceptual ambition arguably carries with it also an explanatory aim, similar to Alexander Wendťs idea of `explanation by concepť (Wendt 1998). 29 Typology is understood as the study or systematic classification of types. 18 1. 2. 3. Explorative Conceptual Comparative Figure 1: Research foci Examination of existing `euroscepticism theory' Deduction and development of types of euroscepticism CONTRIBUTION Typology of euroscepticism Application of the typology in case countries. Observations Tests Confirms / rejects Survey data 4. Evaluative. Review and reconsideration of existing literature. Application of typology to explain public opinion towards recent EU-developments. Suggestions for further research This agenda locates the thesis in a number of different research fields. Thematically, euroscepticism emerges in light of the gap between the EU's objective to be close to its citizens and actual public opinion. As a theoretical problem, it raises a number of issues that are dealt with by European integration theories, theories of EU legitimacy and theories of voting behaviour. Methodologically, it approaches the research questions through a concept analysis of euroscepticism and through the method of structured, focused comparisons of the types of euroscepticism in selected case countries. Importantly, the study is `holistic' (see for instance Gerring 2001: 60): its research questions can only be meaningfully answered on the basis of a joint theoretical and empirical agenda. A contribution of the thesis' conceptual analysis to euroscepticism research consists of laying a path for increased cumulability, which is an essential feature of social science research (for instance Sartori 1974, 1984; Eichenberg 1999; Gerring 2001; Hooghe 2001; George and Bennett 2005). Cumulative studies-- dependent variable of euroscepticism. This thus especially refers to quantitative studies of causal chains. 30 Importantly, the thesis aims to explain the shortcomings of such studies only with regard to their understanding of the 19 studies comparable in terms of study object--contribute incrementally to the development of coherent theories and knowledge by discovering new aspects that may be consistent with, or complementary to, existent findings.31 As the thesis maintains that literature on euroscepticism is characterised by a low degree of cumulability, due to the dependent variable problem briefly introduced above, it seeks to develop a typology, which can not only focus this literature, but also help design more targeted inquiries in the future. A study of euroscepticism, however, is important for several reasons. With opinion polls across the member states revealing a mounting scepticism toward the EU over the past decade,32 in spite of increasing efforts at European and national levels to convey the merits of further integration, an indepth account of the motivations that together make up euroscepticism is long overdue. While isolated analyses of voting behaviour at particular voting times are not missing in existent literature (there are thorough studies of referenda results and European Parliament elections33 ), a systematic account of the `everyday' euroscepticism that constitutes the alleged gap between citizens and the political elite on EU issues is non-existent. Research is needed not so much on when a population is sceptical enough to vote `no' in a referendum, but on the day-to-day issues that inspire its sceptical EU attitudes and sustain the oft-acclaimed gap between `citizens and Brussels'.34 Had such an account already existed prior to 2005, the French and the Dutch `no' votes would have been less surprising.35 A conceptualisation of euroscepticism, however, also provides an addition to the broader study of the EU. More precise knowledge of public scepticism can strengthen endeavours to deal with the general spurs and constraints, and, on a more normative note, the potential inadequacies of the ongoing integration process. In a time where treaty changes--along with major events such as Turkey's EU accession--to an increasing extent are submitted to referenda, euroscepticism is arguably a dominant 31 This view on `contribution' is advocated inter alia by King, Keohane and Verba (1994), who underline the importance that research projects `make a specific contribution to an identifiable scholarly literature by increasing our collective ability to construct verified scientific explanations or some aspect of the worlď (p. 15). Sartori writes with specific reference to the need to enhance cumulability that `The need for reconstruction [of concepts] results from destruction, from the fact that our disciplines have increasingly lost all "discipline." Amidst the resulting state of noncumulability, collective ambiguity, and increasing incommunicability, it is imperative to restore or attempt to restore the conceptual foundations of the edifice' (Sartori 1984: 50). For a rare article on cumulation with regard to public EU support, see Eichenberg 1999. 32 At least a look at two abstract indicators often used as the dependent variable in studies of EU support--support for membership and the feeling of benefit from the EU--suggest rising scepticism in the Union as a whole (the thesis returns to discuss these indicators in Chapter 6. See Eurobarometer Interactive Search System for an overview). 33 To name but a few: Marthaler 2005; Ricard-Nihoul 2005; Ivaldi 2006 (French referendum on the Constitutional Treaty); Siune et al. 1992, 1994; Svensson 2002; Andersen 1998, 2000 (several Danish referenda); Grabbe 2001; Miller 2001; Hayward 2002; Garry et al. 2005 (Irish Nice referenda). See Section 4.6 for a pursuance of findings from voting analysis. 34 A main difference between the analysis of `everyday' opinion and voting analysis is that voting analysis has to take into account a strategic dimension, as voters, for instance, may use an election to punish their government, or further a goal that is not directly related to the issue at stake. 20 player, if not a veto player, in the Union. Moreover, an account of the phenomenon would be advantageous in maximising the efficiency of EU related campaigns: both `yes' and `no' sides, as well as general information activities, would increase their utility if they were able to direct themselves at more discernible target groups. Euroscepticism has both delayed and ultimately modified institutional reform of the EU, most directly through national referenda, and also influences EU politics through opinion polls and more traditional channels of citizen politics such as lobbying, protests and general elections. Potentially, a study of euroscepticism also has a wider relevance than EU politics: As Aleks of support for political institutions and political elites and so our T alisation of what constitutes popular Terminology Szczerbiak and Paul Taggart have noted: `euroscepticism is one manifestation of a lack understanding of it helps to demonstrate the way new issues are entangled, embedded and implicated in wider political concerns. It is a potential bell-weather for understanding the tenor of politics in a climate of sceptical or distrusting mass public sentimenť (Szczerbiak and Taggart 2003: 22). hus, even outside the realm of EU studies, the conceptu scepticism towards a political system is relevant from a general political science perspective.36 nion37 It is no impl s (for instance Fossum 2006). While one may European integration and the European U s e endeavour to narrow down what the EU i arrive at an exhaustive and even uncontroversial description of its various institutions and bodies, theoretical disagreement continues as to what label to attach to the whole. Arguably, a meticulous elaboration of the thesis' own conceptualisation of the EU's nature is not central to its focus: what matters is discovering various types of euroscepticism that reflect public perceptions of what the EU is. Public opinion, as will be returned to below, may both be misled and mistaken. At least, it is hardly wrong to assume that the majority of public perceptions of what the EU is are detached from theoretical debacles on whether a political system (for instance Hix 1999), an institution sui generis (Jachtenfuchs and Kohler-Koch 1996), a new form of governance (for instance Hooghe and Marks 35 Indeed, to return to Dybkjr and Whitman's aforementioned surprise that euroscepticism seemingly peaks in connection with the Constitutional Treaty, their surprise rests on the assumption that existent understandings of euroscepticism were in fact correct in their portrayal of public euroscepticism as a largely democratic critique. 36 It remains an open question whether this contribution to the study of public scepticism can in fact be applied outside the European Union. It would certainly be interesting, although outside the scope of the thesis, to examine whether the broad types of scepticism that the thesis develops could for instance encapsulate sceptical attitudes of citizens in a federation like the United States towards the central administration in Washington (I owe this reflection to Andrew Moravcsik). 37 For simplicity, `European Union' (EU) is sometimes used to also cover developments that took place prior to the actual establishment with the Maastricht Treaty in 1992 of the Union. 21 2001) or a regulatory state (Majone 1996) is the most adequate conceptualisation, as well as from detailed scientific observations of the EU's actual functioning. Nevertheless, an underlying assumption of the thesis is that (a supportive) public opinion is important for the well functioning of the EU, and this carries with it a basic assumption about the nature of the Union--namely that it can loosely be of the term, requiring some degree of common identity and/or homogeneity ut see Scheuer 1999). approached as a political system forming some kind of political unity. In David Easton's famous conception of the political system as those interactions through which values are authoritatively allocated for a society (Easton 1965: 21), a distinction is drawn between three components: the political community, the political regime and the political authorities (for instance Easton 1965). While it seems fair to speak about the existence of some kind of EU regime as well as of EU authorities, one could perhaps question the empirical existence of an EU political community. The thesis argues that this depends on whether or not one adopts a thick or a thin conception of the political community. The thesis adopts a broad definition when referring to the EU as some kind of political system, with `community' referring to a set of people with some shared element, and `politicaľ to those shared elements that arise as a consequence of membership in the EU. It may in fact be more accurate to speak of this as social community in the sense that Karl Deutsch arguably uses when he refers to the need for a `sense of community' in the EU (Deutsch 1957). At least, it is questionable whether the populations of the member states of the EU can be said to form a political community in a more narrow definition (b `Integration' refers in a generic sense to the bringing together of two or more things. Unless otherwise specified, `European integration', 'the integration process', etc. refers in the thesis to the wider and deeper, positive and negative integration that takes place among the member states as a result of cooperation in the European Union. `Process' underlines that integration is an on-going development, which--depending on what type of integration is in focus at a particular time--may inspire changing public attitudes. Deeper integration refers to an increase in the number of political areas that are decided and discussed at EU level; while wider integration implies that more countries become members of the EU (Nedergaard 2004: 106). Broadly speaking, positive integration refers to the addition of new policy areas under EU competence, while negative integration suggests the removal of barriers. The terms `European integration' and `the integration process' are thus only used in an abstract context: the mentioning of, for instance, an `opposition to integration' does not allow inferences about whether the opposition reflects citizens who are opposed to, say, enlargement, or citizens who are 22 opposed to, say, tax harmonisation. `Hard eurosceptics', however, as Chapter 3 returns to discuss, are citizens ho t e any integration, and thus the entire rationale of the EU. the term public opinion, it is useful to clarify its use of the ion polls in the EU, namely `the population of the respective nationalities of the ttitudes when they are manifested as actions or articulated as opinions (Christenson and and actions of individuals, groups and vernmental policy'.40 As already mentioned, aggregate public opinion will in the thesis for practical reasons be analysed through recourse to quantitative opinion polls.41 w ypically oppos Public opinion To delimit the thesis' employment of components `public' and `opinion' separately as well as together, and to pursue an inquiry into the potential rationality of public opinion. The thesis defines the `public' as `a particular aggregation of people, resident in a country.' However, it pragmatically narrows down the term to include members of the sample of the population surveyed by typical public opin European Union Member States, resident in each of the Member States and aged 15 years and over' (Eurobarometer 63, technical specifications). `Opinion' is sometimes used interchangeably with `attitude'. This would, however, ignore the subtle but vital distinction between the two terms. Technically, an attitude is simply a posture or position or stance; the posture the human organism adopts in relation to its environment (Christenson and McWilliams 1962: 2). To denote more precisely what is referred to by the term, Christenson and McWilliams rely on a social psychologist definition: `An attitude is an organized and consistent manner of thinking, feeling, and reacting with regard to people, groups, social issues, or, more generally, any event in one's environment (ibid38 ). In short, thus, it may be seen as a predisposition. We learn the nature of an individuaľs a McWilliams 1962). `Opinion' is thus viewed by the thesis as an expression, or externalisation, of an `attitude'.39 As for `public opinion', it may be defined as `the aggregate of individual opinions held by the adult population' and that it is relevant with regard to issues of general public interest when it by its `intensity and/or constancy may support, oppose or influence the behaviour go 38 Christenson and McWilliams quote Lambert and Lambert, Social Psychology, Prentice-Hall, Inc. (1964). 39 Opinions may also be expressed in a non-verbal fashion; however, the thesis is concerned with examining opinions as they are formulated in quantitative polls. 40 These definitions build on the works of public opinion scholars such as Allport (Allport, G. 1937: `The functional autonomy of motives', American Journal of Psychology, Vol. 50, pp. 141-156; and Davison, W. 1968: `Public opinion ­ introduction', in D. L. Sills (ed.): International Encyclopedia of the Social Sciences, New York: MacMillan, pp. 188-197). 41 Taking recourse to polls does not mean that the thesis reduces public opinion to the results of polls. It recognises that nuances, such as a changing climate of discussion and debate, may not be measurable by polls (see for instance Bogart 1989: 55). 23 The thesis' interest in public euroscepticism naturally builds on the assumption that public opinion exists and constitutes a worthy topic for investigation. Ontologically, two camps have widely opposed each other with regard to whether or not public opinion at all exists (and thus matters). James Bryce (British politician, lawyer and historian, eventually Viscount and member of House of Lords; cf. Fishkin 1995) argued in 1888 that public opinion was not a recent, democratic invention but rather that it `has really been the chief and ultimate power in nearly all nations at nearly all times' (Bryce in Christenson and McWilliams 1962: 6). He moreover was of the opinion that public opinion was the `key' that would `unlock every door' (ibid: 264). Taking an opposite stance, French sociologist Pierre Bourdieu famously stated that `public opinion does not exisť (Bourdieu 1973: title). Little less controversially, Scottish historian Thomas Carlyle once called popular opinion `the greatest lie in the worlď (in Christenson and McWilliams 1962: 13). More pragmatic, but along the same line, integration theorist Ernst B. Haas, with specific regard to the process of European integration, stated that `it is as impracticable as it is unnecessary to have rom democratic theory that that the people are the nly legitimate source of power (Obradovic 1996).42 recourse to general public opinion and attitude surveys' (Haas 1958: 16). Few people would arguably denounce the importance of public opinion for the integration process today. Instead, public opinion has been recognised by scholars working within a variety of different approaches as becoming increasingly important in the design of future integration scenarios (for instance Kritzinger 2003: 220; Hix 2005a). Public opinion is seen as necessary for the establishment and maintaining of the EU, based on the assertion f o Accepting the existence and relevance of public opinion, however, does not necessarily mean accepting the assumption that public opinion is rational. Few, even among advocates of increasing the weight of public opinion on the political process, have assumed the existence of an environment of perfect information when citizens `make up their minds'. And few would deny the amorphous and fluid quality of opinions. Bryce, for instance, did not claim the solidity or rationality of public opinion, recognising instead the slim basis on which public opinion might be formed (see Fishkin 1995: 74). Nevertheless, as Benjamin Page and Robert Shapiro have also argued, while individual citizens may know little about an issue, and while individual opinions may fluctuate widely, when taking society as a whole, a different 42 However recent this actual acknowledgement of the centrality of public opinion may be at EU level, it could be mentioned that it is in fact nothing new in International Relations literature. Indeed, an early writer, Nicolo Machiavelli, strongly warned leaders (princes) to take good note of the public's views: `I conclude, therefore, that when a prince has the goodwill of the people he should not worry about conspiracies; but when the people are hostile and regard him with hatred he should go in fear of everything and everyone. Well-organized states and wise princes have always taken great pains not to exasperate the nobles, and to satisfy the people and keep them content; this is one of the most important tasks a prince must undertake' (Machiavelli (1514) 1961: 105). 24 kind of public opinion is revealed (Page and Shapiro 1992: 17-23). Collectively, as random fluctuations in individual views will tend to cancel each other out, public opinion can thus end up being more rational and stable than what could be expected from examining individual opinions (Fishkin 1995: 87). One argument for construing public opinion as rational is thus the claim that although individual citizens may hold ill-informed opinions, the collective public's opinion achieves a greater rationality-- that is, when all the opinions are put together (Fishkin 1995: 86). This line of reasoning underlies the thesis' approach to aggregate public opinion, echoing also James Stimson's ideas about the difference b e empirical leverage, also in is' broader aim of elucidating a possible gap between U leaders and the broader public on EU issues. etween the political importance of individual and mass opinion: `Useful, and therefore consequential, opinion is aggregate. Politicians care about the views of states, districts, areas, cities, what-have-you. Individual opinion is useful only as an indicator of the aggregate. For a politician to pay attention to individual views is to miss the main game ... The politician must, as a matter of image, appear to b concerned about individuals, but aggregate opinion is what matters' (Stimson43 , quoted in Anderson 2002: 1). As Christopher J. Anderson argues, recent evidence suggests that Stimson's ideas about the importance of aggregate public opinion for shaping elite action hold significant connection with the process of European integration (Anderson 2002: 1). The thesis accepts the view on aggregate public opinion that it may, and may not, be rational. It admits to having little to say about the wisdom with which the public holds its views. Again, this rather open definition is accepted on the basis of the thes E Outline of the thesis The remaining chapter in Part One (Chapter 2) introduces the methodological and epistemological framework of the thesis. Thus, it considers its approach to conceptualisation, before commenting on the criteria used for the selection of case countries, the time perspective, and the role of explanation and theory. In Part Two, a theoretical inquiry into euroscepticism, Chapter 3 provides a critique of existing literature on defining euroscepticism before introducing and discussing the definition employed by the thesis. In Chapter 4, expectations about euroscepticism are identified and deduced from existing literature, and from this inquiry constitutive types of euroscepticism are developed. The thesis argues that our understanding of aggregate public EU attitudes is enhanced by the examination of a general approach offering an inquiry into euroscepticism as a legitimacy problem for the EU, as well as six more specific approaches, giving us substantive clues as to the diverse nature of sceptical attitudes. Part 25 Two concludes with the advancing of a typology of euroscepticism. This typology is then subjected to empirical scrutiny in Part Three. Chapter 5 operationalises it in the form of statements, or theoretical propositions in testable form. The data-sets relied upon to measure euroscepticism across countries and over time are the European Commission's Eurobarometer (EB) polls. Advantages and disadvantages of the use of Eurobarometer are discussed, as is how indicators from the surveys are identified, recoded and tested in order to be applicable as substitutes for the various types of euroscepticism. In order to corroborate the typology, gamma and alpha tests are run to assure that the theoretical propositions can indeed be approached as coherent and independent types of euroscepticism. In Chapter 6, the tested typology is examined in three case countries: Denmark, France and the United Kingdom. Following a summary of the main findings from this analysis, Part Four concludes by evaluating the conceptualisation and the theoretical framework, and discussing the theoretical and practical The table below sums up the main research concerns of each part of the thesis: ain research concerns implications of the findings (Chapter 7). Table 1: M Part One * Identifies and specifies the research topic * Introduces the conceptual, methodological and epistemological framework Part Two * e `state of the arť in the field. Generates an overview of existingOffers a critique of th `euroscepticism theory' * Deduces and develops constitutive types of euroscepticism * Proposes a typology of euroscepticism Part Three * Operationalises and tests the typology * Analyses the presence and intensity of each type of euroscepticism in the case countries * Analyses the development and interplay of the types over time * Compares euroscepticism in the case countries Part Four * Evaluates the overall conceptualisation of euroscepticism in light of the analyses: 43 Stimson, James A. (1991): Public Opinion in America: Moods, Cycles, and Swings, Boulder, CO: Westview Press. 26 - e conceptualisation by concluding on the `status'In case of corroboration, the thesis refines th and dynamics of the types of euroscepticism - In case the analyses did not corroborate the conceptualisation, the thesis rejects it in its current form and suggest possible amendments, pursuable in future research, in light of the newly acquired findings * Explains inadequacies of existing literature in terms of the thesis' analytical framework * Discusses the consequences of its contribution to euroscepticism research * Evaluates implications of euroscepticism for the European integration process 27 2 Methodological framework The methodological framework of the thesis rests on two broad pillars: a concept analysis and a comparison of euroscepticism across selected member states. The examination of the concept of euroscepticism is approached through a three-level structure developed for the social sciences (Goertz 2005), and the empirical analysis is carried out according to the method of structured, focused comparison (in particular George 1979). In the first part of this chapter, the inquiry into euroscepticism is taken one step further by an introduction to the field of concept building and analysis, as well as, more specifically, to the conceptual framework used to structure the thesis' analysis. In line with Giovanni Sartori, the argument is that concept formation stands prior to data analysis (Sartori 1970: 1038): the reverse order, or a lack of emphasis on conceptualisation, would leave the simple question `monitoring of what?' impossible to answer in a satisfactory way. This realisation is of special significance here given the present definitional and conceptual ambiguities surrounding the concept of euroscepticism. Without a thorough conceptualisation of its constitutive types, it is the argument of the thesis that the term euroscepticism will continue to evoke a multitude of muddled associations and thus remain inappropriate as a basis for theoretical and empirical examinations. 2.1 Conceptualising euroscepticism 2.1.1 What is a concept? The `rationale' behind concepts is simple. As Earl Babbie proposes, many of our observations in life seem to have something in common, alluding that they represent something more general than the simple content of a single observation. It is inconvenient to keep describing all the specific observations whenever we want to communicate about the general concept they seem to have in common, so we give a name to the general concept--to stand for whatever it is the specific observations have in common (Babbie 1998). There is, of course, more to concepts than their name: conventionally, as John Gerring points out: `a concept refers to an alignment among three intertwined components: the term (...), the phenomena to be defined (...), and the properties or attributes that define those phenomena (...)'. (Gerring 2001: 39). This definition of a concept in terms of its term, meanings and referents has come to be known as the Ogden-Richards Triangle (based on a book by Ogden and Richards from 1946 that respectively Sartori 1984 and Gerring 2001 develops), and is illustrated below: 28 A Term is `a linguistic label comprised of one or a few words' (Gerring 2001: 39) and refers to the `word allocated to a concepť (Sartori 1984: 24). Meaning refers to `the definition, intension, or connotation of a concepť (Gerring 2001: 39)--in Sartori's words to `the ensemble of characteristics and/or properties associated with, or included in, a given term' (Sartori 1984: 24). Referents refer to `whatever is out there before or beyond mental and linguistic apprehension' (ibid)--to the `extension, or denotation of a concepť (Gerring 2001: 39). These three intertwined components are returned to below, when the thesis specifies the various steps involved with its approach to conceptualising euroscepticism. Figure 2: What is a concept? The OgdenRichards Triangle (building on Sartori 1984: 22; Gerring 2001: 41) Term Meaning Referents As mentioned above, the thesis' approach builds on the concept structure developed by Gary Goertz for the social sciences. To Goertz, focusing on concepts is to think about the nature of the phenomenon being conceptualised--concepts, in other words, are theories about ontology; about the fundamental constitutive elements of a phenomenon (Goertz 2005: 4, 12).44 He stresses that these ontological characteristics may play a role in causal hypotheses, and that an empirical analysis of a phenomenon is required for an adequate conceptualisation. Goertz' view of concepts, in short, is an ontological, causal and realist one (in particular Goertz 2005: Chapters 1 and 2). Goertz develops a three-level framework for social science concepts (2005), which is adopted by the thesis. Euroscepticism, as most complex concepts, has a three-level character, consisting of a basic (or cognitive/semantic) level, a secondary level of constitutive types, and an indicator (or data) level. While the thesis relies on Goertz' approach to concept analysis for structure, it should be mentioned that the underlying idea of combining a semantic definition with an engagement of more substantive inquiries into existing conceptualisations of a term is widely employed (see for instance Gerring 1997, 2001; Sartori 1997; Weyland 2001). Before introducing these three levels, however, it is important to rehearse the rationale of concept analysis in the social sciences. 44 The term `ontology' is used by Goertz to designate the core characteristics of a phenomenon and their interrelationships (Goertz 2005: 4). 29 2.1.2 Concepts and the dependent variable problem Many fashionable words recur: to suggest but a number of contemporary examples from the field of political science, words such as `legitimacy', `terrorism', `globalisation', `europeanisation' and `governance' flourish in the literature with more or less extensive definitions. There is, as Sartori states, nothing inherently problematic with fashionable words--except, perhaps, that they risk becoming `abused, distorted, and trivializeď (Sartori 1997: 58). Indeed, with specific reference to the term `pluralism', Sartori argues that its very explosion in prominence in the 1960s rendered it a `noble and ennobling worď with little if any substance (ibid: 61). Kurt Weyland, in a similar vein, demonstrates with reference to the term `populism' that as a result of conceptual disagreements, a too wide variety of governments, parties, movements, leaders, and policies has today misleadingly come to be labelled populist (Weyland 2001). David A. Baldwin's book Economic Statecraft is another example underlining the importance of taking concept analysis seriously: Through a thorough identification of various forms of statecraft and existent understandings of these, Baldwin establishes an analytical framework through which he is able to review several classic studies. Having established clarity regarding the subject under scrutiny, he authoritatively concludes that much conventional wisdom has been misleading or wrong in drawing inferences (Baldwin 1985). However, not only may a term be clouded by vagueness and errors as a result of lacking definitions: in the absence of conceptual clarity, terms--such as pluralism or euroscepticism--may also come to mean so many different and even contradictory things that authors applying the same term in fact `talk past each other', thereby hindering the cumulation of knowledge (Weyland 2001: 1; see also Gerring 1997: 957). More discernibly, differing definitions and conceptualisations of a term are likely to produce differing operationalisations, which can lead to problems of data incommensurability. It is inter alia such problems that lead John Gerring to suggest that the absence of a thorough conceptualisation of `ideology' has implied that `the debate over the ideological proclivities of the mass public does not seem much closer to resolution today than it did in the 1960s' (Gerring 1997: 959). As the introduction to this thesis hinted, this situation indeed appears to be characterising research into public euroscepticism. The below table is a non-exhaustive list showing the dependent variable(s) relied upon by oft-cited euroscepticism researchers (listed chronologically, except when an author is listed twice). Table 2: Examples of dependent variables engaged in existing research on public EU attitudes Author/Year Study object Dependent variable(s) Christopher Andersen Public opinion - Support for European unification 30 and Karl Kaltenthaler (1996) - Support for membership - Regret if the EU was scrapped Christopher Andersen and M. Shawn Reichert (1996) Utilitarian concerns - Support for membership Matthew Gabel (1998c) Public support - Support for membership - Support for European unification Richard Eichenberg (1999) Utilitarian concerns - Support for membership - Feeling of benefit from membership Richard Eichenberg and Russell Dalton (forthcoming 2007) Public support - Support for membership Lauren McLaren (2002) Public support - Support for membership - Feeling of benefit from membership45 Lauren McLaren (forthcoming 2007) Euroscepticism - Support for supranational decision-making Robert Rohrschneider (2002) Public support - Support for an EU government - Satisfaction with EU democracy Sylvia Kritzinger (2003) Public support - Support for West European unification Claes de Vreese (2004) Public support (small member states) - Support for membership - Integration is pushed too fast - EU is a threat to smaller countries - Willingness to make sacrifice for less strong country - EU has more disadvantages than advantages for people like me Liesbet Hooghe and Gary Marks (2004, Public opinion - Support for membership - Desired speed of integration 45 Lauren McLaren (2002) states that she would have liked to adopt the dependent variable employed by Matthew Gabel (1998c), but since the question of support for European unification was absent in the Eurobarometer poll for her yearunder-scrutiny, she substitutes it with the question of feeling of benefit. The sole justification given is that the two questions `arguably gauge general attitudes towards the EU' (McLaren 2002: 556). The thesis does not see the two questions as substitutable, and McLaren's otherwise timely and interesting study becomes to me a clear example of too little emphasis on the dependent variable, which reduces the confidence with which we know what we are actually learning from the study. 31 2005) - Desired direction of integration Kees van Kersbergen and Catherine Netjes (forthcoming 2007) Public support - Support for membership - Regret if the EU was scrapped - Image of the EU - Feeling about the EU Liesbet Hooghe and Gary Marks find that the Eurobarometer question on satisfaction with EU membership is the most frequently used measure for EU support and scepticism (Hooghe and Marks 2005: 426-430), as the above table also suggests. However, the table shows that the membership question is often used in combination with other variables, and, moreover, that rather different variables are at times employed to measure public EU attitudes. We may note that even in the same journal volume, three articles employ rather different dependent variables for EU support (Eichenberg and Dalton; McLaren; and Kersbergen and Netjes: forthcoming 2007). This makes it difficult to draw overall conclusions about public support from them. Considering such differences in the operationalisation of what is actually being surveyed, it is perhaps less surprising that the confusion about euroscepticism mentioned in the introduction has remained throughout the history of European integration, and that seemingly contradictory causal mechanisms have been identified (see introduction to Chapter 1). In the concluding chapter (Chapter 7), the thesis attempts to revisit and reconsider a number of existing studies to propose how the thesis' conceptualisation can narrow down their contribution and thus assist the cumulability of euroscepticism research. 2.1.3 On solving the dependent variable problem As the thesis relies on existing literature to deduce theoretical expectations as to the nature of euroscepticism, a further note on its review, and use, of this literature is due. Indeed, even sympathetical readers might be perplexed by the thesis' strong emphasis on the need for undertaking a meticulous reading of an extensive body of literature in order to deduce and develop, and subsequently test, a number of expectations about the nature of euroscepticism--some of which may appear to merely replicate findings of existing studies. However, this perplexity would overlook the very issues that, to the thesis, contribute to the dependent variable problem and thus to today's persistent confusions and contradictions as to what euroscepticism is. Indeed, first, a barely discernible number of studies have in fact concluded on the substantive nature of different types of euroscepticism--i.e. the study object of this thesis. As mentioned above, euroscepticism is in most existing literature simply operationalised in terms of support for membership, 32 and efforts are instead spent on identifying and testing the explanatory power of a number of independent variables--that then logically cannot in themselves be considered to be euroscepticism.46 Say, for instance, that a study wants to investigate whether the fear of losing national sovereignty has an impact on euroscepticism and tests this by checking to what extent the indicator polling the `wish to keep the national veto' correlates with the dependent variable of membership support. This simplified (but in its essence typical) study might conclude that to a high extent this fear is relevant for our understanding of euroscepticism. Looking at actual figures it becomes clear that the size of the group under scrutiny only represents the attitude of about 15 percent of the population in the European Union (indeed, this was the EU average finding membership to be a `bad thing' in 2005; Eurobarometer 64). We have, in other words, not learned much substantive about the general nature of euroscepticism from the study. The thesis argues that what we have learned is `merely' the extent to which one possible constitutive type of euroscepticism, in this instance sovereignty concerns, leads to citizens holding the view that the very idea of membership is a `bad thing': in other words, amongst this relatively small group of people, there is a high number who are also afraid of losing national sovereignty. This finding is sound and relevant; however, it tells us little about the dynamics at play in connection with attitudes towards, for instance, the Constitutional Treaty, the idea of a more social Europe or other issues characterising the day-to-day politics of the Union. Neither have we had confirmation as to whether or not sovereignty concerns in themselves can be approached as a coherent and independent type of euroscepticism (the thesis returns in the conclusion with more substantive reviews of existing studies). This at least partly explains Liesbet Hooghe and Gary Marks' aforementioned finding that existent research has failed to predict today's alleged rise of euroscepticism. Adopting the analogy of an attempt to construct a big table without knowledge of how an ideal table is constructed, we can say that by only expecting euroscepticism to be the negative evaluation of membership--or in more elaborate studies, the evaluation of membership in combination with, say, the evaluation of benefit--we have come up with a one- or two-legged table (a partial theory), which does perhaps form the contours of a solid table, but which is not stable and may not stand upright in all its corners (i.e. it might fail to explain the dynamics of euroscepticism in particular member states). In this case, adding one leg to the table would render it considerably more stable, whilst most engineers probably would end up with the conclusion that four legs are required for a solid table. 46 Studies that combine several indicators to form their dependent variable may be more refined, but as they rarely substantiate their choice of dependent variable, they run the risk of being tautological: if, for instance, the benefit question is engaged as (part of) the dependent variable--and the indicator `meaning of the EU: A waste of money' is engaged as (part 33 It is fair to say that there is yet no coherent overview of what euroscepticism is. Indeed, while most readers may agree with the thesis' underlying assumption that euroscepticism is multifaceted, there is no agreement or clarity as to how many coherent and independent facets this umbrella term consists of. Continuing the use of analogy to further these points, we can think of the famous Indian story of the blind man who had to identify what he was faced with when coming across an elephant-- having never seen one before. Today, existing literature gives us thorough, but largely independent, accounts of what we may later confirm is a trunk, ears and a tail. In fact, to be blunt, we have yet no clear idea of whether elephants have 16 legs.47 What is due is the overview of the entire elephant and thus the creation of synergy between the scattered partial accounts. The thesis attempts to provide such overview and synergy so that future studies into euroscepticism do not have to search east and west before pursuing specific investigations into the causes and dynamics of various types of euroscepticism. An overview is provided by Chapter 4's extensive reading of existent accounts of euroscepticism and its causes, while synergy is attempted achieved by the thesis' overarching aim of solving the dependent variable problem. Coherence and independence will be controlled for statistically (see Chapter 5). Importantly, the literature examination in Chapter 4's does not only attempt to provide overview. Indeed, the thesis also proposes to engage differently with the existent literature. In recognition of the present situation of no exhaustive overview of what euroscepticism is, it employs, to reiterate from the introduction, a constitutive as opposed to a causal perspective. In the above simplified example of a study testing the explanatory power of sovereignty-concerns on support for EU membership, this means that if the thesis on the basis of its overall reading of the literature finds reason to expect such concerns to be relevant for the study of euroscepticism, it will try to isolate this as a type of scepticism-- and thus not as a cause of scepticism. It will, however, only accept it as a type if statistical tests prove its coherence and independence on the basis of multiple indicators. In Chapter 4 below, the thesis will therefore look both to independent and dependent variables of existent studies for clues as to what scepticism is. Indeed, in line with the above-presented arguments by Giovanni Sartori, concept analysis stands prior to causal analysis. of) the independent variable--what we may have is two indicators of one type of euroscepticism, and thus a constitutive relationship, rather than a causal relationship. 47 The elephant story is famously used by Donald Puchala in his analogy of describing what the EU is. I here include Puchala's version of the story, as it underlines the thesis' argument in a somewhat different manner: `Several blind men approached an elephant and each touched the animal in an effort to discover what the beast looked like. Each blind man, however, touched a different part of the large animal, and each concluded that the elephant had the appearance of the part he had touched. Hence, the blind man who felt the animaľs trunk concluded that an elephant must be tall and slender, while the fellow who touched the beasťs ear concluded that an elephant must be oblong and flat. Others of course reached different conclusions. The total result was that no man arrived at a very accurate description of the elephant. Yet each man had gained enough evidence from his own experience to disbelieve his fellows and to maintain a lively debate about the nature of the beasť (Puchala 1972). 34 To sum up, the thesis holds that the bulk of existent literature has in fact not provided substantive answers as to what euroscepticism is. Indeed, euroscepticism, to most studies, is the membership question, and we are `merely' introduced to a number of variables, which are claimed to explain variation in the number, and patterns, of replies to this question. A famous thesis of euroscepticism, such as the Democratic Deficit thesis (see Section 4.7 below), is therefore rarely conceived of as being euroscepticism, but as being the cause of euroscepticism. It should be clear from this line of argumentation that the thesis rejects the disentanglement of the concept of euroscepticism from many of the independent variables examined in existent studies and instead proposes to undertake an investigation into whether, in this case, democratic concerns can in themselves be conceived as an independent, coherent type of euroscepticism. The below chapters engage in the substantiation and test of these lines of thought. 2.2 The three-level structure48 2.2.1 The basic level Returning to the presentation of the three-level concept structure adopted by the thesis, what Goertz refers to as the `basic leveľ is related to Sartori's idea of `high level categorizations' (Sartori 1970: 1041). Euroscepticism--the label or term designated to the concept under scrutiny--is, in itself, an empirical universal, vaguely indicating the occurrence of some kind of scepticism towards the EU, but revealing little more than `the main entr[y] of a filing system' (Sartori 1970: 1043). Indeed, as Babbie puts it, the terms associated with concepts are merely devices created for purposes of filing and communication (Babbie 1998). This, however, is not to say that there is nothing more to the basic level of a concept than semantics: indeed, important conceptual questions pose themselves at each of the three levels of Goertz' approach. At the basic level, the question of the negative of a concept, or the `non-concepť, arises. This inquiry is useful as it sharpens the analysis of the positive pole. Without negative identification, 48 This walkthrough, as mentioned above, builds primarily on Goertz (2005); however, it could be mentioned that Robert Adcock and David Collier (2001) adopt a similar approach to conceptualisation. As they distinguish between indicators and the scores that these generate, their structure is based on four levels. From the broadest to the most specific level, these are termed (i) the Background Concept (`the broad constellation of meanings and understandings associated with a given concepť) akin to Goertz' basic level; (ii) the Systematized Concept (`a specific formulation of a concept used by a given scholar or group of scholars') akin to Goertz' level of constitutive dimensions; (iii) the Indicator Level (also referred to as `measures' and `operationalisations'); and (iv) the Scores for Cases. I like the simplistic three-level model proposed by Goertz (2005) and have found his efforts to clarify the respective tasks of the three levels the most useful for the thesis. 35 concepts have no specified termination or boundary, which limits the clarity and value of theory as well as empirical analyses (Sartori 1970: 1042, 1043).49 The `negative' of the concept of euroscepticism is not restricted to the phenomenon of `Europhilia', since this would exclude the range of opinion that is indifferent to the EU. Indifference, it should be obvious, is not scepticism. Euroscepticism, as elaborated upon in Chapter 3, is taken to be opinions hesitant towards the idea of the EU or developments on the EU agenda; whether or not these in fact further or constrain integration is not significant for the definition. Euroscepticism is politically interesting partly because of the role it may play in influencing the agenda or `setting' of the existent EU. The term for the non-concept of euroscepticism, i.e. the absence of euroscepticism, may be borrowed from Leon Lindberg and Stuart Scheingold, who identified a `permissive consensus' in their largely neo-functionalist analysis of public attitudes towards the EC in the early years of its existence (Lindberg and Scheingold 1970). Borrowing the term for the non-concept of euroscepticism from Lindberg and Scheingold does not imply ascribing in a minute fashion to their conceptualisation of a permissive consensus, which assumes that the EU is taken for granted and accepted, whilst not being an issue of high political salience (ibid, see also Inglehart 1971). Indeed, this conceptualisation moreover implies that citizens are ready to take onboard further integrative steps if only politicians can agree on these (for instance Panebianco 1996). It should, however, be noted that contrary to what is sometimes casually observed, Lindberg and Scheingold were under no illusion that citizens shared a robust support for the EU that was not vulnerable to periods of crisis. Indeed, elsewhere in their book they note the unlikelihood of the European Community possessing a strong sense of `diffuse supporť in the member states. Quoting David Easton's famous definition of diffuse support (see also Chapter 4 below), they write that if the EC at all contains the `reservoir of favourable attitudes' that helps citizens tolerate unpopular outputs, it is `most certainly a shallow one' (Lindberg and Scheingold 1970: 274). With its employment of the term `permissive consensus', the thesis does not assume that non-eurosceptics are ready to uncritically pursue any type of integration. Rather, it conceptualises a permissive consensus as including enthusiastic, critically approving and passively supportive attitudes towards the EU or EU developments. Included under the term may thus be `europhiles' (approving most integrative developments); `pragmatics' (approving EU developments and issues following a case-to-case evaluation of their merits); and citizens who may be uninterested in, but passively compliant towards, 49 While it is possible to classify responses to a public opinion survey as eurosceptic or not-eurosceptic, it is by no means assumed that this is a static positioning, or that the continuum between the two poles is completely dichotomous. There is unlikely to be such a clear cut limit between the positive and negative concept. The aim of classification is to reflect the broad trend characterising a population at a given moment. 36 the EU. In terms of the purpose of this thesis, the crucial common denominator of these groups is that they are not in any explicit way constraining EU developments.50 In mathematical terms, the negative concept can be operationalised as zero or neutral on all the indicators substituting the secondary level types (Goertz 2005: 23). The negative pole in the project, the permissive consensus, is thus the absence of the positive pole, euroscepticism. 2.2.2 Secondary level types Euroscepticism is, like most important concepts, multidimensional. Its multidimensionality appears at the secondary level in Goertz' concept structure. It is in other words at this level, or a step down the `ladder of abstraction' from the high, universal level to the `fat slice of the medium level categories', to continue the parallel to Sartori's influential work on concepts (Sartori 1970: 1043), that we find the constitutive dimensions (Goertz 2005: 5)--or types--of euroscepticism. For example, when we say that euroscepticism can be sovereignty-based or economic in nature, we are giving some of the constitutive, or secondary level, types of this phenomenon. These types are part of the theoretical edifice--indeed, in line with the critical rationalist perspective offered by Section 2.3.2 below, they are derived from theoretical literature--but they are concrete enough to be operationalised at the indicator level. In conceptual analyses, the basic and secondary levels are really the theory of the concept, while the indicator level is the connection to measures and data collection (Goertz 2005: 7), which allows for empirical scrutiny of the adequacy of the conceptualisation. The constitutive, as opposed to causal, nature of the secondary level types must be fully appreciated: these types do not cause the phenomenon, they constitute what the phenomenon is (Goertz 2005: 11). What constitutes euroscepticism? As mentioned above, the theoretical investigation will deduce and develop with a number of different constitutive types of the phenomenon. Each of these types is theorised as being public scepticism or hostility towards the EU (i.e. euroscepticism). The various types do not cause the phenomenon, they are the phenomenon in the same manner that, as an example, states providing certain goods and services are welfare states (Goertz 2005: 4). Nevertheless, the types may play a central role in (`post-conceptuaľ) causal mechanisms: `One cannot neatly separate the ontology of a concept from the role it plays in causal theories and explanations' (Goertz 2005: 20; see also Alan Zuckerman, who underlines the `theoretical importance' of concepts, which `has to do with the utility of a concept in the development of statements of wide explanatory and predictive power'; Zuckerman 1975: 231). The duality of 50 Of course, if people completely uninterested in the EU refrain from voting at an EU referendum, and they are so substantial in number that they either make the referendum invalid (in several countries, for instance, at least 50 percent of the electorate must vote for a referendum to be valid) or in combination succeed in sending such a strong signal to the EU 37 simultaneously having an ontological theory of a phenomenon and at the same time choosing secondary level types in part because of their causal powers at the basic level is inescapable (Goertz 2005: 49). Finally, it should be emphasised with regard to the secondary level types that they represent the thesis' expectation as to what constitutes aggregate level public euroscepticism, based on comprehensive theoretical accounts about the phenomenon. 51 This expectation is subjected to empirical scrutiny to demonstrate whether the types are, after all, prevalent in the case countries. The secondary level types thus allow for the ambitions of the thesis (see Section 1.2) to explore, conceptualise and compare euroscepticism, and also to evaluate the adequacy of the existing literature in capturing the phenomenon. 2.2.3 The indicator level: concept operationalisation The indicator level is where the concept is specified in such detail that data can be gathered (Goertz 2005: 5). This means that it is at the indicator level that it becomes possible to actually identify whether or not citizens in a member state are eurosceptic. This level thus brings in actual empirical data, which in the present case consist of public opinion polls. Sartori's `low level categories' are characterised by a configurative conceptualisation, where denotation is sacrificed to accuracy of connotation--sometimes to the extent where definitions become contextual (Sartori 1970: 1041). This is similar to the indicator level, which, consequently, rarely comes into play in subsequent descriptions of causal mechanisms or in the theoretical framework of concepts (Goertz 2005: 195). In Sartori's differentiation among three levels of abstraction, high, medium and low levels of abstraction, the high level represents universality, the medium level represents generalisation, and the lower level represents uniqueness. A high level (or basic level, to return to Goertz' vocabulary) universal is likely to contain multiple medium level generalisations (secondary level types), which again are likely to have multiple low level configurative categories (indicators). A synthesis of these three levels, concepts provide guidelines of interpretation and observation (Sartori 1970: 1040). To Sartori, the concepts of any social science are not only the elements of a theoretical system; they are equally and just as much, data containers: Indeed, data is information that is distributed in, and processed by, `conceptual containers' (Sartori 1970: 1049). establishment that it has a visible constraining effect on the proposed developments, one may speak of these groups as forming part of the positive concept of popular euroscepticism, and not of the negative concept (the permissive consensus). 51 This assertion is stated from a contemporary perspective: It is possible that ten years from now, depending on the way in which the EU evolves, other types, not conceived of today, may become prevalent. Moreover, as the thesis returns to discuss in the concluding chapter, it is possible that individual-level qualitative studies of public attitudes (for instance through discourse analysis) would conjure up with somewhat different expectations. 38 2.2.4 Method of aggregation What remains to be discussed with regard to framing the three-level conceptualisation of euroscepticism is the manner in which the multiple indicators at the indicator level `produce' the secondary level types and, equally, how the secondary level types are structured or combined to arrive at the basic level concept. Throughout his book, Goertz (2005) contrasts two different structural principles: the widely used structure of `necessary and sufficient conditions' and the alternative `family resemblance' concept structure. The structure of necessary and sufficient conditions defines a concept by stating the conditions that a case must (necessarily) satisfy in order to belong to the basic level concept (to be a democracy, free elections and freedom of speech are necessary conditions), as well as whether or not these conditions are enough (sufficient) in order for a case to belong to the concept (for instance Goertz 2005: 7, 27; in the above example, free elections and freedom of speech are necessary but not sufficient conditions for democracy). The family resemblance structure has no necessary conditions: what is required is enough resemblance on the secondary level types--`sufficiency with no necessary condition requirements' (Goertz 2005: 26). The basic level concept of euroscepticism in the thesis is constructed with family resemblance secondary level types, which again are operationalised by family resemblance multiple indicators. The structure is one of family resemblance because the prevalence of, for instance, the concern of a loss of national sovereignty in a country may suffice to distinguish that country as eurosceptic, without the type of, for instance, economic dissatisfaction with membership being prevalent. The relationship between the secondary level and the basic level is one of ontology, while the relationship between the indicator level and the secondary level is one of substitutability. It should follow from the above discussion that this is so because the secondary level (family resemblance) types of euroscepticism constitute the phenomenon, while the (family resemblance) indicators are operational substitutes for (i.e. they stand instead of) the secondary level types (see Figure 3). Once there are multiple types and/or multiple indicators, the question of weighting arises, i.e. the question of how to establish the individual weight of each sufficient type of euroscepticism. In the thesis, weighting the influence of the secondary level types is simply done on the basis of empirical observations. Thus, weights (or scores) can only be ascribed following the empirical analysis. No two types of scepticism are expected to have equal significance in a case country, neither are the same types expected to be equally significant across the member states. Which types are important in a country thus depends on how large a percentage of its population supports the indicators standing in for that 39 40 type. As pursued in greater detail by Chapter 5, one direct indicator is sufficient to create the secondary level type of which it is a substitute; however, it goes without saying that if all the indicators identified as standing in for a given type reveal the presence of eurosceptical sentiments, this points to a weighty secondary level type. Though different types of scepticism are likely to have different political implications (addressing a, say, democratic scepticism is different from addressing a sovereignty-based scepticism, and some types may be more `severe' than others in that they are more conducive to the wish of leaving the EU altogether), no type is in itself more representative of the phenomenon of euroscepticism than another. This explains the mathematical `OR' between the types in Figure 3 below. The empirical analysis may reveal that one type is more prevalent than another, but, as has been stressed above, all types are in and by themselves sufficient.52 Again, Chapter 5 on operationalisation supplies more details on the weighting and measurement of euroscepticism at the indicator level. To sum up, euroscepticism is a phenomenon with a number of family resemblance types, each of which is sufficient for the phenomenon to manifest. The absence of one type of scepticism does not exclude a member state from the phenomenon, but a member state may be characterised by the presence of more than one type. All types are perceptions held by the population of a member state of the European Union (aggregate national level as expressed through public opinion polls). The absence or neutrality of all types in a member state amounts to a permissive consensus amongst that population towards the EU and/or EU developments. Leaving the substantiation of the levels for subsequent chapters, Figure 3 illustrates the broad three-level structure of the concept of euroscepticism. Substantiation of the basic level awaits Chapter 3; the secondary level types are deduced and developed in Chapter 4, while a completed structure of the figure is presented following the identification and test of empirical indicators in Chapter 5. 52 The empirical analysis may reveal the complete lack of prevalence in all the case countries of one (theoretically derived) dimension of popular euroscepticism. This possibility cannot be ruled out, as many existing assumptions and hypotheses about the phenomenon in the theoretical literature have not been validated statistically or by empirical data. Builds on Goertz 2005. Legend: Substitutability; Ontological; OR is the logical term for family resemblance Euroscepticism Indicator level Secondary level Basic level Figure 3: The three-level structure of the concept of euroscepticism. Non-substantiated illustrative model OR OR 41 OR OR OR OR OR OR Poll question Poll question Poll question Poll question Poll question Poll question Poll question Poll question Constitutive type 1 Constitutive type 2 Constitutive type 3 2.3 Comparing euroscepticism: the method of structured, focused comparison Subjecting the thesis' conceptualisation of euroscepticism to empirical scrutiny is an integral part of its conceptual approach: it logically follows the definitional and theoretical inquiries and is, so to say, the thesis' third leg. What is to be analysed, in other words the thesis' unit of analysis, is the populations of the member states of the EU. The thesis is aware that this perspective to some extent reifies nationstates and risks ignoring important intra-population variances that will exist. However, it is important to note that the thesis' conceptualisation of euroscepticism--i.e. the typology it develops to capture its various types--is not country exclusive. It is, in other words, not only applicable to country level analysis, but intended to be able to capture--outside the confines of this thesis--public euroscepticism at all its various levels (individual, regional, and so forth). The thesis chooses to illustrate the conceptualisation of euroscepticism at country level for both theoretical and pragmatic reasons, namely (i) since this is the level of analysis most frequently engaged by existing studies of euroscepticism, whose cumulability the thesis seeks to enhance; and (ii) since the member states continue as crucial actors in the integration process when it comes to decisions `affecteď by euroscepticism (euroscepticism, in other words, is in the overall perspective of the European integration process arguably most relevant at country level). It is beyond the scope of this study to engage an empirical analysis of the populations in all 27 member states of the Union. Instead, a systematic analysis of a small number of cases (see for instance Lijphart 1971) is undertaken. To the thesis, cases represent `an opportunity to learn more about the complexity of the problem studied, to develop further the existing explanatory framework, and to refine and elaborate the initially available theory employed by the investigator in order to provide an explanation of the particular case examineď (George 1979: 52). A comparison of aggregate eurosceptical opinion in a number of member states will for instance demonstrate whether or not euroscepticism converges along similar lines, and whether or not there are similarities in the development of euroscepticism over time. Applying the theoretical types of euroscepticism to cases is thus also a means to identify strengths and shortcomings in the overall conceptualisation. Cases are in other words necessary building blocks in the attempt to provide an adequate conceptualisation of euroscepticism. It is recognised that the `costs' of a small n focus--for instance the risk of losing the generality that might be achieved with a wider range of cases (ideally all member states of the EU)--are an inevitable part of the trade-offs inherent within any research design (for instance Brady and Collier 2004: 9). In the present context, with the given practical limitations, the focus on a small n best facilitates the illustration and analysis of the various types of euroscepticism by allowing for a finegrained, contextually sensitive study (Brady and Collier 2004: 7-10, 14). 42 The method of structured, focused comparison (George 1979; Bennett and George 1997) guides the analysis. It stipulates clear methodological requirements for conducting and managing a series of cases. The method is structured in that it borrows from statistical research models by asking a set of standardised, general questions of each case; and focused because it deals selectively with only certain aspects of the cases, i.e. a selective theoretical focus guides the analysis (George 1979: 62; Bennett and George 1997: 2). In Chapter 5, indicators for each type of euroscepticism are identified, and these indicators are in the subsequent chapter examined for each of the cases. To maintain a focused analysis, Chapter 5 likewise draws up a list of empirical observables on the basis of the findings of the theoretical inquiry of Part Two, which then constitutes the overall focus of the empirical analysis. In this way, the conceptualisation is designed to be applicable to all member states of the EU. Euroscepticism in any member state can in other words be `looked up' and classified in accordance with the criteria set out in the thesis. 2.3.1 Country selection The thesis engages a comparative analysis in order to apply its conceptualisation of euroscepticism. It follows from the above discussions that this empirical analysis should be seen as an inherent part of the endeavour to conceptualise. The following lines of reasoning have to be considered with regard to case selection. To borrow an analogy from John Gerring (Gerring 2004: 342), cases are conceptualised as the rows in a rectangular representation of a dataset, where observations are the cells and the types of euroscepticism the columns. Ideally, the thesis is interested in selecting the smallest number of member states that, fulfilling the below criteria, provide an exhaustive illustration of euroscepticism--or, in line with the above analogy, it wants the maximum variation in the content of the cells obtained by the smallest number of rows. There should thus be enough rows (i.e. cases) to make sure that there is no column with no `eurosceptic celľ, while rows that only repeat the cell content of a previous row are superfluous. To illustrate this, Case 4 in the below idealised model with three types of euroscepticism is superfluous, while Cases 1 to 3 are necessary for an exhaustive conceptualisation. 43 Table 3: How many cases? Euroscepticism? Type 1 Type 2 Type 3 Case 1 Yes No No Case 2 No Yes No Case 3 No No Yes Case 4 Yes No No The broad aim of the cases is to examine the prevalence and dynamics of the various types of euroscepticism (see Chapter 5). Countries are expected to differ in their euroscepticism, and cases are thus chosen to reflect maximum variation within the concept, implying a mixed type of comparison: one case may be most-similar to another and both may be least similar to a third case. To the thesis, the central issue for case selection is whether the addition of a case might allow inferences on additional types of the phenomenon (Bennett and George 1997; also George 1979: note 27). Apparent differences among the cases open the possibility of illustrating the widest possible range of euroscepticism, while apparent similarities among the cases may, conversely, help to unveil unexpected variation that supports the assumption that the phenomenon is more complex and less uniform than often assumed. It follows from the above that the cases useful for evaluating the theoretical conceptualisation of euroscepticism are cases where we may reasonably expect some kind of widespread euroscepticism. This selection criterion, illegitimate to classic causal studies, is necessary for the non-causal, constitutive analysis envisaged by the thesis. The purpose is to examine whether all the theoretically derived types of euroscepticism are relevant empirically--this purpose would not be fulfilled by investigating member states in which a reasonable expectation is that hardly any euroscepticism exist. Certainly, no country may be characterised by zero euroscepticism, wherefore relative scepticism in relation to the EU average is required in order for us to have a meaningful analytic concept (this point is developed in Chapter 5). In summary, the following criteria determine the choice of countries: * Individually, the country should be an old member of the EU (the longer the better for the thesis' ability to trace and compare euroscepticism over time, as well as for assuring a `settleď public opinion towards the EU).53 53 Indeed, the thesis holds that great care should be taken before undertaking comparisons between old and new member states during the first years after accession. Examining Eurobarometer polls, previous accessions have revealed rather large fluctuations in support and scepticism levels of newly acceded countries. This instability of public opinion in the early years following accession may be due to a number of diverse reasons. Apart from the issue of having to get used to life in the EU, 44 ˇ Individually, the country should have an indisputable presence of euroscepticism (established by a `no' vote in an EU referendum or the presence of opt-outs from central areas of co-operation). It is unrealistic to assume that all member states are relatively eurosceptic, wherefore, in Arend Lijpharťs terms, cases should be `most similar' when it comes to a eurosceptical reputation (see for instance Lijphart 1971). * However, cases should be `most differenť when it comes to their experience with the EU. More concretely, the countries should be as diverse as possible in terms of the factors often relied on in studies on EU attitudes: size; time since entry; geographical position; gross national product (GNP per capita as a broad indicator of affluence); and World War II experience. The broad theoretical reasoning is as follows: smaller countries are more favourable towards membership, because their influence in relative terms is enhanced through the EU's principle of equality among member states; date of entry is important as familiarity with the EU could be thought to breed support; countries with borders to several other EU member states may gain more from co-operation; low domestic affluence often means high economic benefits from the EU's structural funds, which is thought to lead to more favourable public opinion; finally, impetus for closer European integration has been largely recognised as related to the experience of a country under the Second World War (both in terms of the number of deaths and psychologically) (for instance Janssen 1991; Anderson and Kaltenthaler 1996; Anderson and Reichert 1996). On the basis of these criteria, the cases selected for analysis are the Danes, the French and the population of the United Kingdom (henceforth, for simplicity (though inaccurately) referred to as the British). Each country is briefly introduced below. However, even the casual observer should have noted that there could be expected to be both similarities and differences among the three cases with regard to their type(s) of public euroscepticism: Denmark and Britain have often been singled out in the literature as the `eurosceptical couple' in the EU, while France and Britain are sometimes seen as complete opposites with regards to EU preferences. It was repeatedly argued in connection to the ratification process of the Constitutional Treaty that virtually all `yes' arguments in France would be `no' arguments in the United Kingdom. As a check on the thesis' conceptualisation of euroscepticism, and the fact that many promises or threats raised by prior campaigns in those states holding a national referendum will have to stand the test of time, previous newcomers have frequently been economically weaker states (Ireland, Spain, Portugal, Greece, and not least the 12 new member states), and, moreover, relatively new democracies (Spain, Portugal, Greece and the ten new Central and East European member states). These factors may contribute to making public opinion in recently acceded member states difficult to compare with older members. It should be mentioned that in the thesis, the data analyses sometimes presents the EU average of 25 member states--if this is the only average given by Eurobarometer (Bulgaria and Romania are not included in any calculation of the average). 45 it is relevant to investigate if these countries concur in terms of their type(s) of euroscepticism--a finding which would undermine the thesis' expectation of euroscepticism being different from country to country, or if they fail to coherently capture an expected type of euroscepticism--a finding which would question the adequacy of the theoretical inquiry in capturing euroscepticism. Denmark spurred political and academic awareness of euroscepticism when the Danes rejected the Maastricht Treaty in a referendum in June 1992. Denmark soon gained a reputation for being one of the most eurosceptical member states in the Union, a reputation which was seemingly confirmed when Danes also rejected the single currency in a referendum in September 2000. Danish membership of the European Communities dates back to 1973. With some 43.1 thousand square kilometres and 5.4 million inhabitants (respectively 2001 and 2004 figures from the European Commission), it is reasonable to classify Denmark as a small member state. A GDP per capita, expressed in Purchasing Power Standards (PPS), of 26200 (2003 figures, European Commission) places Denmark among the top-3 in the EU-15 (ibid). Denmark was occupied under the Second World War, while, it seems fair to say, not counting among the most severely affected countries.54 Denmark is part of Scandinavia and has a close geographical proximity to several EU member states, but its only land border is with Germany. Denmark has close ties to the Nordic countries, of which Norway and Iceland are non-EU members. `Nordic Union' has repeatedly been (and continues to be) stressed by EU opponents in Denmark as a viable and more attractive alternative to EU membership (for instance by the People's Movement Against the EU). As touched upon in the introduction, Denmark has four EU opt-outs, namely in the domains of Justice and Home Affairs (supranational co-operation), the third phase of Economic and Monetary Union (the Euro), Union citizenship and Defence policy (see Branner and Kelstrup 2000 for an introduction to Denmark and the EU). France, a founding member of the EU, surprised many when its population almost rejected the Maastricht Treaty in a referendum in 1992. French politicians are reputed to share grand ambitions with regard to Europe; ambitions, which, it would seem, are not shared by many French. Throughout the first decade of integration, France had not been portrayed as a markedly eurosceptical member state, but its eurosceptical credentials were confirmed when the French rejected the Constitutional Treaty in May 2005. France is a big member state, counting 59.9 million citizens and 544 thousand square kilometres (respectively 2004 and 2001 figures from the European Commission). France shares a border with five other member states of the EU (Belgium, Germany, Italy, Luxembourg and Spain). 46 France's experience with war, occupation and liberation under the Second World War left deep scars and divided the political establishment (for instance Gowland, O'Neill and Dunphy 2000: 119). French GDP per capita (23700 in PPS; 2003 figures from the European Commission) is below the EU-15 average (25033 cf. the cited 2003 figures) (see for instance Drake 2005 and Holm 2006 for introductions to French relations with Europe and the EU). Few are in doubt with regard to the presence of some kind of euroscepticism in the United Kingdom, although the British have never voted `no' in an EU referendum. Indeed, often distinguished in public opinion polls as strongly opposed to EU developments, British scepticism appears to cover most facets. The United Kingdom--England, Northern Ireland, Scotland and Wales--makes up an area of 243.8 thousand square kilometres and a population of 59.7 million citizens (respectively 2001 and 2004 figures from the European Commission). The UK joined the EU in January 1973, alongside Denmark and Ireland. The UK was never occupied under the Second World War and shares close ties both across the Atlantic and to the Commonwealth. British GNP per capita (25300 in PPS, 2003 figures from the European Commission) is very close to the EU average, and thus slightly higher to that of France, although big regional differences characterise the country. The United Kingdom has a number of special agreements with the EU: in addition to its opt-out from the third phase of EMU (the Euro), it has a protocol regarding Schengen co-operation, and an opt-in possibility with regard to Justice and Home Affairs (see Baker and Seawright 1998 for an introduction to Britain in the EU). In combination, the thesis expects these three countries to reflect both similarities and variances in their versions of euroscepticism. Denmark and France both appear to have experienced a change in public support for the EU, with public opinion polls recently indicating growing support for several new EU developments in Denmark and the reverse situation in France. On these issues, the seemingly strong euroscepticism in the United Kingdom has appeared stable. It could be added that all three countries have had--or have at least attempted to have--largescale public debate on the issue of European integration. Denmark has held six referenda on European issues, France two and Great Britain one. There is extensive analysis of public opinion towards the EU in the three countries (see bibliography), and scholarly focus on this issue is likely to continue to be strong in the years to come following recent events in connection with the ratification of the Constitutional Treaty. 54 Former long-term Minister of Foreign Affairs in Denmark, Uffe Ellemann-Jensen (Liberal Party) even sees Denmark's relatively mild encounter with the War as a main explanation behind Denmark's historically rather pragmatist (and 47 2.3.2 Context specificity With regard to both pillars in the methodological framework engaged by the thesis--concept analysis and comparison--the issue of context specificity poses important questions. Indeed, while appropriate measures of concepts should have construct validity--that is, they should be `good predictor[s] of phenomena that are widely hypothesized to be associated with [them]'55 (Elkins 2000: 294), the researcher should also aim for measurement validity (Adcock and Collier 2001: 534). However, it is crucial to simultaneously recognise the virtual impossibility of constructing single, all-purpose definitions of complex terms such as euroscepticism (the basic level), or the terms associated with the individual constitutive types (the secondary level), usable for all times, places, and purposes (Gerring 1997: 983). This limitation, needless to say, is especially acute for (though certainly not limited to) researchers engaged in cross-national analysis or comparisons across distinct historical periods:56 indeed, as a result of history- or culture-specific conditions, a term may be imbued with such diverse associations that careless employment threatens the validity of measurement. While context specific definitions cannot be easily avoided, it is important that researchers recognise their situatedness in an as self-conscious way as possible (Gerring 1997: 983; Adcock and Collier 2001: 534). As regards the term and concept of euroscepticism, the recognition by the thesis that it is multifaceted goes a long way in accommodating the fact that it may mean different things in different countries. In addition, it may even travel with more ease across geographic and theoretical boundaries (Gerring 1997: 983; i.e. be less context dependent) than most concepts in comparative politics, as it by definition is restricted to European countries and as it bears reference to a delineated entity (the European Union). Nevertheless, as regards the terms employed to give meaning to the individual constitutive types of euroscepticism, the issue of context specificity will pose acute questions. One such term, as we shall see, is `sovereignty'. It goes for all three case countries (and indeed most member states of the EU) that they are old nation-states with a long and treasured history of national sovereignty. The particular associations invoked by the term `sovereignty' are likely to be different from the Danes to the French to the British. Potentially, this could cloud the ability of the thesis to label one of these countries more or less concerned about the EU's impact on national sovereignty than the rest. This study will nonetheless not enter a conceptual analysis of the meaning of sovereignty, nor of the other terms engaged by the thesis as labels to the constitutive types of euroscepticism. In the attempt to achieve measurement equivalence across the case countries, however, the thesis will in the delimitation economic) approach to co-operation in the EU. 55 In the case of this thesis, an example of such a phenomenon could be a `no' vote in an EU referendum. 56 Indeed, as Robert Adcock and David Collier point out, subgroups in a population (in terms of religion, gender, class...) may differ systematically in their response style to survey questions (Adcock and Collier 2001: 534). 48 of the various types, as well as in their operationalisation (the indicator level), specify as explicitly as possible the reference object of the term. Expressed differently, with regard to the above example of sovereignty the thesis will be careful to define and operationalise it specifically in terms of the reluctance to increase the supranational competencies of the European Union. With the pragmatic additional remark that the interpretation of Eurobarometer surveys amongst academics as well as EU politicians is often based on the assumption of measurement equivalence, the thesis deems these lines of reasoning sufficient for its purpose and aims.57 2.3.3 Time perspective The data material (Eurobarometer polls) permits the thesis to adopt a longitudinal focus stretching from 1974 to today (2006 is the latest year included in the analysis, allowing a 32 year focus). A broad time perspective allows for the fullest evaluation of the nature of euroscepticism. It makes possible the monitoring of the dynamics of its various types, such as for instance the gradual disappearance of the relevance of one type within a country or the shift in relevance between two types. Longitudinal studies generally boast of a number of advantages compared to one-off crosssectional studies. While the latter can point to the prevalence of a particular situation at a particular point in time, longitudinal studies importantly allow researchers to identify trends--for instance the occurrence of change or stability. This is central to the thesis, given (i) the existing uncertainty as to what euroscepticism is; (ii) the fair assumption that it is not static; as well as (iii) the well known volatility of public opinion. Political or socio-economic events may dominant mass sentiments at certain points in time, and result in highly skew responses to surveys (for instance a one-off low in scepticism levels)--momentary upheavals that cross-sectional studies have no possibility of capturing, but which longitudinal studies are able to place in context. The longitudinal perspective may also provide an interesting addition to the evaluation of the adequacy of the theoretical framework in capturing euroscepticism. Indeed, it may shed light on whether or not some existing ideas about the nature of euroscepticism are better equipped than others in accounting for the contemporary dynamics of euroscepticism, as well as allowing for qualified guesses about its future development. It should be stressed, however, that when evaluating the fit of the conceptualisation and discussing its potential implications in Part Three, emphasis will be on contemporary types of euroscepticism, i.e. the euroscepticism of the mid 2000s. It should moreover be noted that depending on the suitability of the available Eurobarometer data, it may not be possible to present an even portrait of the more than three decades of monitored scepticism. 49 2.4 Epistemological and ontological assumptions 2.4.1 Explanation and theory As an important aim of the thesis is to conceptualise euroscepticism, it is important to linger a little on how it approaches this type of knowledge. The thesis, one could say, aims to explain euroscepticism in a non-causal manner. Recalling the above discussion, the motivation for this endeavour was formulated in response to the lack of an existing coherent conceptualisation of euroscepticism and the underlying assumption that concept formation must stand prior to data analysis: answers to `why' questions, in other words, require answers to `whať questions, which are an important end in themselves (Wendt 1998: 108; 1999: 83, 86-87). The idea of `constitutive explanation' is rather uncommon in the social sciences, where `explanation' is often understood purely in causal terms (it should be clear that this is not the type of explanation that the thesis aims to provide). Its epistemological foundation is discussed at some length by for instance Alexander Wendt, who predominantly works within the constructivist tradition. Constitutive explanations however, are not only of interest to scholars working within a constructivist perspective,58 nor do they pertain exclusively to the social sciences. Their rationale rests on the claim that `whať and `how-possible' questions are not necessarily engaged in pure description but may also fulfil an explanatory role, which calls for a broader conceptualisation of `explanation' than the traditionally positivist equation between explanation and causality (Wendt 1998, 1999). In Daniel Barbiero's words, to provide an explanation, in the constitutive sense of the term, is to show how a phenomenon can be understood in terms of its relevant constitutive factors (Barbiero 1996). Importantly, constitutive explanations should not be seen as alternatives or competitors to causal accounts or `classic explanations', rather they should be approached as complementing and preceding causal studies, providing explanations into how certain events or developments are possible, and even why (Schwellnus 2001: 15-16, quoting a 1999 article by Barnett and Finnemore).59 The example of studies into the nature of the European Union (Wendt 1998) is a relevant one. Answers to the oft-asked research question `what is the European Union' have included `an emerging federation', `a system of multi-level governance', `a supranational state', and so forth--partly descriptive concepts, which nevertheless seem to seek an explanatory role as well (Wendt 1998: 111). Indeed, the claim that the European Union is an emerging federation includes the assumption that the EU will have 57 The thesis discusses the use of Eurobarometer polls in Section 5.1. 58 Wendt even suggests that constitutive explanations, or `explanations by concepť, are the dominant mode of explanation in history and pervade the social sciences, even economics (Wendt 1998: 111). 50 a propensity to behave in a certain way under certain conditions. For instance that it will have a disposition to centralise authority (ibid). This kind of explanation of the EU's dispositions is constitutive as opposed to causal. While causal stories would have the explanation of change as their central objective, constitutive stories are occupied with accounting for the properties of things (which may well be static) (Wendt 1999: 105). Wesley Salmon distinguished between `etiological explanations', explaining a fact by showing how it came to be as a result of antecedent events, processes or conditions, and `constitutive explanations', which do not explain in terms of antecedents, but by showing that the fact-to-beexplained is constituted by underlying mechanisms--an `exhibition of [its] internal causal structure' (Salmon 1984: 270). Other scholars speak of constitutive explanations as looking inside (for instance Marko Barendregt, a biologist60 ), or downward from the empirical level (`level 0') to the underlying level of organisation (`level -1'). Above, in Section 2.2 of the thesis, the relationship between the various types constituting the concept of euroscepticism was described as one of ontology: When one or more of the constitutive types come into being, so does, logically, euroscepticism (cf. Wendt 1998; Goertz 2005). As already touched upon earlier in this chapter, the constitutive view on explanation has implications for the role of theory in the thesis. Adequate answers to constitutive questions (i.e. `whať and `how-possible' questions) must satisfy different truth conditions than answers to causal questions (such as `why' questions; Wendt 1998: 105).61 In the thesis, theory is brought in to be able to discern and derive the various types of euroscepticism. Theory, in other words, is the platform from which the thesis develops the secondary level in the three-level concept structure, and thereby advances a `constitutive explanation'. There is thus a close connection between theory and concept formation, which is reflected by the term `explanation by concepť, used interchangeably with constitutive explanation by Wendt (Wendt 1998). 2.4.2 A critical rationalistic framework As hinted at by Figure 4 below, the projecťs logic rests on the critical rationalist approach to science in so far as its interest lies in (adapted from Hay 2002: 81): 59 In the concluding chapter, the thesis discusses a number of ways in which its constitutive explanation of euroscepticism facilitates the conducting of causal studies. 60 See: http://logica.rug.ac.be/censs2002/abstracts/Barendregt.htm. 61 Constitutive theories are subject to the same correspondence tests of truth as causal theories, and this endeavour need not be more problematic with constitutive theories. As Alexander Wendt argues, while all observation is theory-laden, and this means that we can never test our theories directly against the world, but only indirectly via other competing theories, this is equally true of causal and constitutive theories (Wendt 1999: 106). The `scientific solution' for both is to rely on publicly available observations of the world, which critics can access and assess for accuracy, relevance, etc. (ibid). 51 1) Drawing out and developing constitutive types of euroscepticism from existing literature (`euroscepticism theory') 2) Developing, deductively, operational statements from the resulting types of euroscepticism (propositions in testable form) 3) Statistical tests. Observing empirical evidence (public opinion polls) 4) In case of corroboration: refining the conceptualisation 5) In case of inconsistency: rejecting the conceptualisation in its current form, suggesting possible amendments in light of newly acquired facts Robert Adcock and David Collier (2001: 530-531) propose a variant of these research steps with specific regard to conceptualisations. Once a `background concepť--i.e. the broad constellation of meanings and understandings associated with a given concept--has been identified, the researcher (1) needs to formulate a `systematised concepť through reasoning about the background concept, in light of the goals of research. Next (2), she needs to operationalise the systematised concept by developing one or more indicators for scoring/classifying cases. Third (3), scores must be applied to the indicators for the cases under scrutiny. Following the analysis (4), the researcher may find herself involved with modifying indicators, or potentially creating new indicators, in light of observed scores (indeed, the iterative nature of research is explicitly emphasised by Adcock and Collier, and taken to include the possibility of introducing new conceptual ideas during the validation process; something which is referred to as `friendly amendments'; ibid: 533). Likewise, task (5) involves the process of fine-tuning the systematised concept, or possibly extensively revising it, in light of insights about scores and indicators. Finally, task (6) revisits the background concept, which had formed the point of departure of the conceptualisation, by exploring broader issues emerging from the research process. These considerations can be illustrated by the basic research cycle below. The shape of the circle underlines that the various steps are interdependent and should be seen as a flow. 52 2. Theory Theoretical inquiry: identifying and developing secondarylevel types from existing literature 4. Data Statistical tests. Confirmation/rejection of conceptualisation. Observation and analysis of empirical data Concluding evaluation: 5. refinement or 6. suggestions for amendment 3. Operationalisation Deducing operational statements from the conceptualisation Figure 4: The research cycle 1. Definition Defining the broad understandings associated with euroscepticism The critical rationalist approach to science was formulated as a critique of logical positivism, which claimed that empirical induction formed the foundation of science, and thus that by observing single phenomena, conclusions of universal validity could be drawn (see for instance Gilje and Grimen 1993: Chapter 3). To logical positivists, scientific statements were meaningful if they could be empirically verified by observation or logical analysis. Critical rationalism centred on the claim that logical positivism could not fulfil its demand that all scientific statements be verifiable (problem of correspondence): indeed, its central proponent Karl Popper famously claimed that universal statements are by principle not verifiable. Falsifiability, to Popper, is instead what constitutes the scientific part of a statement or a theory. Induction may point to probability, but not to certainty, and would need to be qualified to form the basis for scientific realisations (for instance Gilje and Grimen 1993: Chapter 4). The inductive method pursued by logical positivists was thus modified by the stress by critical rationalists of the hypothetic-deductive method: First, a hypothesis is chosen from the researcher's assumptions about the nature of the world; second, the researcher deduces the observable implications of the hypothesis and only then are empirical data brought in to test whether or not the deductions hold water. If they do not conform to what the theory (or in the case of this thesis: the secondary level types, which are developed from theory) would have expected, the theory is discarded and the researcher may try to conjure up with a different theory. To Karl Popper, a good theory, i.e. one that has not yet been falsified, may eventually result from first jumping from an observation statement to 53 any theory, then continuously testing this by repeatedly applying the critical method, eliminating many bad theories and improving other ones along the way (ibid). Accordingly, as the illustration of the research cycle alluded to, the contours of constitutive types of euroscepticism are in the thesis deduced from a finite body of theoretical literature, which is demarcated by an accumulated stock of knowledge about the main research questions and closely related ones. From these contours, the thesis develops its typology of euroscepticism, from which it is possible to generate a series of operational statements, the appropriateness of which are subjected to empirical scrutiny (data to function as attempted refutations of the statements, cf. Popper). If the analyses corroborate the theoretical expectations, the proposed explanation of euroscepticism is accepted in its current form and its potential political implications will be discussed. If the analyses do not corroborate the theoretical expectations, the explanation is discarded in its current form and an evaluative discussion of the adequacy of the theoretical framework is pursued. This idea of granting indicators a role as `falsifiable claims' as regards the validity of a conceptualisation is also advocated by Adcock and Collier (2001). They suggest that: `Because error is a pervasive threat to measurement, it is essential to view the interpretation of scores in relation to systematized concepts as falsifiable claims (...). Scholars should treat these claims just as they would any causal hypothesis, that is, as tentative statements that require supporting evidence' (Adcock and Collier 2001: 532). Section 5.3 explicates this discussion by conjuring up a list of possible cases of empirical non-corroboration. These include the finding that a theoretically-deduced type of euroscepticism (i.e. an expectation as to what euroscepticism is that the thesis has developed on the basis of the literature) proves to be unfounded following statistical tests. 2.4.3 On deducing constitutive types of euroscepticism and parsimony A final point merits emphasis: the thesis, as stated above, deduces and develops constitutive types of euroscepticism from existing literature on public EU attitudes. It will become clear from Chapter 4 that this inquiry involves the reading of a wide array of often diverse studies and the grouping of these under broad headings following the thesis' evaluation of what their common denominators are. This process necessarily involves a trade-off between parsimony and accuracy, as the finding of the most adequate common denominator connecting various complex approaches to euroscepticism implies compromising on the subtle nuances that will exist between two studies. This trade-off is accepted not only in recognition of the necessity of simplifying for economical reasons, but also in light of the thesis' assumption that just as there are not as many reasons to vote `no' in an EU referendum as there are `no-voters,' it is indeed judicious to speak of broad structuring types of academic approaches to euroscepticism. 54 Giovanni Sartori's illustration of possible configurations of concept characteristics (Sartori 1984: 46-50) is a helpful tool in specifying this endeavour. Adapting Sartori's ideas to the thesis' purposes, the box encompassing the circles in the below figure represents the pool of all possible connotations of the concept of euroscepticism--in other words, it represents in this study the existing research on public euroscepticism (or EU support) at the aggregate level. Each circle represents a cluster of characteristics that are internally congruent, whilst the core of the concentric circles represents the common denominator under which various similar approaches can be summarised. Figure 5: Possible configurations of types of euroscepticism (building on Sartori 1984: 47). It should be clear that individual circles and clusters of closely connected circles represent theoretical expectations to coherent and individual types of euroscepticism. To reiterate with the help of this figure, it is to the task of discovering the number and nature of these circles or clusters of circles that Part Two of this thesis now turns. 55 Part Two: Constituting euroscepticism As Part One explicated, theory-based inquiries into euroscepticism are the thesis' requisite for proposing a constitutive explanation of the phenomenon. It is in other words by examining and developing the assumptions, hypotheses and statements about euroscepticism in the literature that the thesis derives its constitutive types. This examination is the aim of Part Two. Chapter 3 defines euroscepticism--the basic level of the concept structure--on the basis of a critical review of existing literature (the concept structure was explicated in Chapter 2). Chapter 4 then introduces the theoretical framework for understanding the constitutive nature of euroscepticism. Its point of departure is `euroscepticism theory', an emerging body of literature dealing with (sceptical) EU attitudes. The thesis distinguishes between general and specific approaches to understanding euroscepticism, and uses a discussion of these approaches (Sections 4.2--4.8) as the basis for developing the `operationalisable' types of euroscepticism. Section 4.9 is a summary. 3 Defining euroscepticism: the basic level 3.1 A critique of the state of the art in defining euroscepticism Lack of satisfaction with the European Union has assumed many labels, which are rarely specified in great detail--eurocriticism, euroscepticism, EU-scepticism, eurorealism, europessimism to mention but a few. These are all potentially problematic terms, which the below paragraphs, explicating the thesis' employment of the term euroscepticism, testify.62 Euroscepticism, indeed, is a problematic term. With its increasing employment today in sweeping statements by the press, by politicians and among academia, it has become an alluring catchword. However, the term euroscepticism may have come to mean so many things that, by itself, it means nothing. The one really radical remedy--namely, that we should all cease talking about euroscepticism--is rather certain not to be adopted, and the thesis shall pragmatically resort to the task of providing a more adequate definition and conceptualisation (see Gerring 1997: 960-1; Weyland 2001: 62 A Google search carried out by my office computer on 14th December 2006, came up with approximately 109.000 results for euroscepticism, 9.400 results for europessimism, 1.100 results for eurorealism, and 51 results for eurocriticism (Google does not search on the hyphen, wherefore results for, for instance, EU-scepticism would be conflated). 56 1, for a discussion of the temptation of simply doing away with `semantic troublemakers'--their conclusion being that often it is not practical to propose their abolition). Beginning with some of the more notorious ways in which euroscepticism has been employed politically, a starting point could be the Danish Foreign Minister, Per Stig Mller (Conservative Party), to whom scepticism assumes philosophical importance. Borrowing from René Descartes, he stated in speeches leading up to the European Parliament election of June 2004 that `I doubt, therefore I am a European' (see also Mller 1992).63 Doubt is something truly European according to Mller, something that underpins the philosophical, cultural and scientific history of the Continent. A less philosophical usage was that of British Prime Minister Tony Blair (Labour Party), who was cited for calling euroscepticism `dumb'64 --and a more proud one that of Polish President and twin brother of the Polish Prime Minister, Lech Kaczynski, who stated in 2006: `I am more eurosceptic than my brother' (EUobserver 13th November 2006). British EU Trade Commissioner Peter Mandelson adopted a more pragmatic stance in an article in The Guardian, relying on a dictionary definition to locate scepticism as being `open to persuasion'.65 Euroscepticism is featured in The Oxford English Dictionary as: `a person, esp. a politician, who is sceptical about the supposed benefits to Britain of increasing co-operation with fellow members of the European Union, esp. one who strongly opposes greater political or economic integration' (for instance quoted in Szczerbiak and Taggart 2003: 6). Assuming that the definition refers equally to the greater public as it does to politicians, as well as to nationals of other member states than Britain, it is one highly centred on an economic understanding of the phenomenon. One is sceptic if one doubts the benefits of integration. The word `increasing' is interestingly included in this connection, apparently suggesting an existing consensus around the status quo. The thesis returns to this question in due course. The thesis proposes that the definitional problems surrounding the label euroscepticism appear at all of its three parts: `euro', `sceptic', and `ism'. Narrowing down their individual semantic meaning is important as differences in the intension of a term--i.e. its different attributes or defining characteristics--produce differences in extension--i.e. the ensemble of things (or objects) to which the term applies (Sartori 1984: Chapter 1; Weyland 2001: 1). Regarding the first part, `euro', problems are many. From its popular usage in the press, it is clear that the phenomenon is not restricted to scepticism towards the Euro (the single currency), nor, of 63 The statement has been included in several speeches by the Foreign Minister; see, for instance, speech to a European Parliament election conference on 30th April 2004. 64 Watt and Black (2004): `Blair draws battle lines on Europe', The Guardian, 27th March. 65 Branigan, Tania (2005): `Straw kicks off the great EU debate', The Guardian, 27th January. 57 course, to anything that has to do with `euro'; however, it is not helpful to replace the expression with the slightly more concrete version, `EU-scepticism'. While sometimes denoting sceptics of the EU as a whole, the term is frequently used in relation to specific areas of concern, such as the Common Agricultural Policy, or the Constitutional Treaty. Confusion, however, especially surrounds the `sceptic' part of the term. According to dictionaries, and as we saw Peter Mandelson has been keen to point out, being sceptic means being open to persuasion, but many `sceptics' are far more sincere in their opinions than this definition would imply (a non-negligible part of public opinion included within most usages of the label euroscepticism, of course, aims for a complete breakdown of the Union). Indeed, as the thesis substantiates below, we are likely to deal with a continuum of public scepticism. `Ism' is the suffix attached to most ideology labels, lending the term to being incorporated into the domain of political belief systems (Flood 2002: 3). However, whether or not euroscepticism can be seen as an ideology in its own right is a topic of on-going debate (see Flood and Usherwood 2005. Ben Crum and Harmen Binnema distinguish between thick and thin euroscepticism, where the thick dimension is conceptualised as a full-fledged ideology; Crum and Binnema 2006).66 Through a review of existing definitions in the literature, the remaining part of this section lingers on the question of definition in the semantic sense of the term. Robert Harmsen defines euroscepticism exclusively as a fundamental opposition to, or scepticism towards, the `European projecť, directly implying the European Union itself. He thereby rejects labelling opposition to for instance the Common Agricultural Policy as euroscepticism. This, in Harmsen's view, should be perceived as part of normal political debates (discussion with author, May 2005). Harmsen also traces the first use of the term to Britain, where he argues it refers to actual opposition to co-operation and constitutes a very different and more intensive phenomenon than in the rest of the EU (Harmsen 2005). While Harmsen is right to stress that some distinction must be made in order for euroscepticism not to become an all-inclusive term, reflecting also momentary and ephemeral perceptions, his analysis is explicitly geared towards organised, party-based euroscepticism. The thesis argues that the restrictions he applies are not relevant for an understanding of the scepticism of the wider public. Indeed, recalling that the interest here is not to identify a threshold for when a population is sceptical enough to vote `no' in an EU referendum (or for a particular candidate at a European Parliament 66 Some languages permit the use of the word scepticism without the -ism ending, such as the Danish `euroskepsis', which may be a more adequate term. 58 election), but rather to examine the various types of scepticism that exist in the EU, it is indeed useful to be aware of any large-scale disapproval of a particular facet within the integration process. Paul Taggart noted in 1998 that euroscepticism included `the idea of contingent or qualified opposition, as well as incorporating outright and unqualified opposition to the process of European integration' (Taggart 1998: 365). Together with Aleks Szczerbiak, Taggart has found it useful to break the definition into two; hence their influential working typology of hard and soft euroscepticism (especially Szczerbiak and Taggart 2002, 2003), which to date is perhaps the most widely used definition of euroscepticism in the literature. The soft variant of euroscepticism is defined as the contingent or qualified opposition, while the hard variant denotes the outright principled rejection of the process of European integration (ibid). Again the definition is directed towards party-based, or organised, euroscepticism; however, the thesis sees no reason as to why the fundamental distinction between contingent and principled scepticism should not be directly transferable to the public level. Though Szczerbiak and Taggarťs simple typology has on several occasions been criticised for being too inclusive (for instance Kopecky and Mudde 2002), a criticism which is well-founded and in essence also accepted by the authors themselves (Szczerbiak and Taggart 2003), alternative typologies have also encountered terminological problems. Petr Kopecky and Cas Mudde (2002) have argued for a distinction between on the one hand attitudes towards the principle of integration in the EU, and, on the other hand, support or opposition towards further integration in the EU. This has led them to advance a fourfold typology of pro- as well as anti-EU (party) positions, consisting of (i) euroenthusiasts, committed to the idea of European integration and optimistic about further integrative plans; (ii) eurosceptics, committed also to the idea of European integration but pessimistic about further integrative plans; (iii) europragmatists, who do not appreciate the idea or ideals of integration, but broadly accept what they see as potential gains from the practical development of the EU; and (iv) eurorejects, who neither support the idea of integration nor see any possible benefits from further developments in the EU. 59 Table 4: Petr Kopecky and Cas Mudde's typology of euroscepticism Perspective on further integration Optimistic I: Committed/Optimistic `Enthusiasts' II: Not committed/Optimistic `Pragmatists' Committed III: Committed/Pessimistic `Sceptics' IV: Not committed/Pessimistic `Rejects' NotCommitted Pessimistic IdeaofEUintegration Kopecky and Mudde 2002 In Kopecky and Mudde's typology, `eurorejects' would correspond to the `hard euroscepticism' category in Taggard and Szczerbiak's conceptualisation, invoking a principled opposition to the EU and European integration, while the other three categories allow for a more contingent, or soft scepticism. Kopecky and Mudde's categories, nevertheless, can also be reproached for being too inclusive (see for instance a reply by Taggart and Szczerbiak 2003). Christopher Flood and Simon Usherwood moreover point out that Kopecky and Mudde do not fully address the problematic nature of especially the term euroscepticism (Flood and Usherwood 2005: 4): the title of their paper `The two sides of Euroscepticism' and their concomitant employment of the term euroscepticism as one of the four categories in their typology, leaves the reader confused as to when euroscepticism is employed in a generic sense or in a particular sense, as well as to its applicability (the EU or Europe as a whole). Flood and Usherwood (2005) offer an alternative approach to classifying euroscepticism. It is, once again, directed toward the organised (and presumably strategic) party-level, but could arguably also be workable at broader public level. Indeed, although the average citizen is presumably less interested in formulating an overall coherent or strategic position on the issue than a group or a party is, broad attitudinal positioning towards the EU, such as the classification proposed by Flood and Usherwood, is hardly limited to organised groups. The two authors propose a classificatory schema to encompass the range of alignments on Europe, which interestingly avoids the term scepticism altogether. Leaving the content issue aside, they suggest six categories (carrying no prefix) into which positions towards the EU's development, either as a totality or in some particular policy area(s) can be summarised, from most positive to most negative 60 (Flood and Usherwood 2005, building on Flood 2002). Their schema, presented in a slightly modified version below, may in this sense be seen as an `intensity' table of alignments. Table 5: Flood and Usherwooďs labelling of positions towards the EU or EU developments Maximalists... wish to push integration as far and as fast as possible Reformists... endorse the advance of integration, subject to remedying the deficiencies of what has already been achieved Gradualists... accept the advances of integration as long as it is slow and piecemeal Minimalists... accept the status quo, but want to limit further integration as far as possible Revisionists... want to return to an earlier state, usually before a treaty revision Rejectionists... refuse integration and oppose participation From Flood 2002; Flood and Usherwood 2005 It is not an entirely unproblematic classification, which in particular poses a number of operational difficulties. Parties or people rarely elaborate their policies or attitudes in such detail that they may adequately fit the fine-grained schema proposed by the authors (for instance Szczerbiak and Taggart 2003: 10). Moreover, the categories are not mutually exclusive--a party may be revisionist with regard to one policy, rejectionist with regard to another, but reformist in its overall posture (Flood and Usherwood 2005: 6). This raises questions concerning weighting, which are lacking from the paper. Before specifying the thesis' own definition of euroscepticism, we may in light of these on-going debates about the attributes of the term euroscepticism bear in mind John Gerring's advice that `when attempting to cope with semantic plenitude in a concept,' a sound strategy is to settle on a `minimaľ definition-- ideally a single attribute that is universally agreed upon (Gerring 1997: 979-980). In the case of this thesis, such a definition is hopefully able to offer a sense of coherence to the multifaceted term and concept of euroscepticism. 3.2 A definition Based on the above reflections, the thesis employs the following semantic perspective on the term euroscepticism: 61 Euro... includes the possibility of euroscepticism being directed against European cooperation as a whole, as well as towards specific formulations of co-operation proposed by the EU. Scepticism... is taken to be variable and include outright opposition. Most citizens are assumed to be eurosceptic to some extent, which allocates defining importance in terms of measurement and comparison to questions of intensity and durability.67 The term scepticism in itself is understood as the opposite of praise or agreement: as sentiments of disapproval. Euroscepticism... thus, is a sentiment of disapproval--reaching a certain degree and durability-- directed towards the EU in its entirety or towards particular policy areas or developments. `Eurosceptic public opinion' refers to citizens perceiving faults or shortcomings with regard to the EU-of-the-day.68 In line with most studies on defining euroscepticism, it is important to recognise that it may assume a hard, or principled, component as well as a soft, or contingent, component: Hard euroscepticism is the principled rejection of co-operation--the existence of the EU itself is contested--while Soft euroscepticism is the contingent scepticism towards particular aspects within the co- operation. Before the thesis turns to the examination of various types of eurosceptic sentiments, three additional points merit emphasis. First, the thesis assumes that the structure of hard and soft euroscepticism is similar: instead they are foremost distinguished by their intensity. Thus, hard euroscepticism is distinct not so much in terms of the various prior considerations that have led citizens to assume it--as the thesis returns to discuss below, these considerations are likely not to be distinct from more contingent scepticism--but in the extent of its `rejectionism' (wherefore, also, general levels of this component of scepticism are not 67 Intensity is reflected in the relative percentage of citizens sharing a particular opinion, while durability, where possible to measure, is defined as a poll tendency persisting more than two years. Insisting on a certain durability of a trend should minimise the problem of basing conclusions on momentary upheavals that, for one reason or the other, may characterise public opinion. See Chapter 5 for a more detailed operationalisation of intensity and durability. 68 This definition is in line with Liesbet Hooghe and Gary Mark's brief conception that `Scepticism has come to mean "an attitude of doubt or a disposition of disbelief". Euroscepticism refers simply to scepticism about some aspect of Europe or European integration' (Hooghe and Marks forthcoming 2007: 2). 62 expected to be high).69 While soft euroscepticism in theory can be changed into a permissive consensus if the particular object of scepticism--be it financial, ideological, an unpopular prime minister or the fear of losing national identity--is `rectifieď, hard euroscepticism is much more persistent, and, indeed, hard eurosceptics may constitute an ontologically distinct group of citizens. Some types of scepticism may, certainly, be more readily conducive to hard euroscepticism than others, something which Chapter 6 turns to measure statistically. Nevertheless, hard euroscepticism essentially occurs on the same grounds as soft euroscepticism. Second, it should be clear from the above that, contrary to virtually all existing attempts to categorise it, the present thesis includes under the definition of euroscepticism what some scholars have sought to differentiate from it, namely `constructive' scepticism aiming to `improve the EU in its own interest by pointing out what is wrong with it and proposing better alternatives' (Flood 2002: 4). This attitude is explicitly exempted from the definition of euroscepticism by for instance Flood (2002), Szczerbiak and Taggart (2002) and Harmsen (2005). The thesis argues that the upholding of this distinction easily tips over to an implicit view of euroscepticism as an unconstructive and inherently negative pathology--as opposed to an integral, and normal--if not in some parts healthy, characteristic of democratic political systems. 70 Moreover as perceptions of inadequacies, and propositions of `better alternatives', are imbued with a scepticism towards the manner in which the EU-of-the-day runs, it is indeed not convincing why such attitudes should be exempted from the phenomenon of euroscepticism. Moreover, to give a specific example, the distinction would exclude from the classification a significant number--perhaps even the majority--of `no' voters in the French referendum on the Constitutional Treaty of May 2005, as consistent analyses of that vote showed large segments of French `no' voters to be clearly supportive of the EU as a whole (an example is the campaign slogan `J'aime ľEurope ­ Je vote non'71 ). Voting `no' in an EU referendum on a treaty is not necessarily an expression of opposition to European integration as such--yet, the thesis holds that `no' votes are an expression of (some kind of) euroscepticism. A durable, widespread lack of satisfaction with a development in, or aspect of, the EU is important to attempts to bring the Union closer to its citizens, also when formulated in a constructive or Europhile language. The thesis' conceptualisation consequentially includes both kinds of scepticism. 69 As we shall see in Chapter 5 on operationalisation, to measure whether citizens merely criticise an aspect of EU cooperation or if scepticism surmounts to a complete rejection of (membership of) the EU, separate indicators of hard euroscepticism are required. 70 Susan Milner assumes a critical view on whether euroscepticism is `healthy' in her article on the Danish referendum on the Maastricht Treaty (Milner 2000). 71 This was the slogan on several French campaign posters, for instance by Attac (www.france.attac.org), and the title of a book by French Republican politician Nicola Dupont-Aignan (see: www.nda2007.fr/J-aime-l-Europe-je-vote-NON.html). 63 Finally, we may note that hard euroscepticism is often imbued by a generational logic. To Ronald Ingleharťs theory of the Silent Revolution (for instance 1971), hard euroscepticism should become less prevalent with time: As pre-war and war generations are being replaced by younger generations, or as people get used to the necessity and advantages of European integration (perhaps through information campaigns, increased transactions, the successful construction of affective attitudes, or even, as Jürgen Habermas would put it: patriotism towards the EU), they abandon principled opposition to European integration. The generational logic, however, has been challenged on empirical grounds (for instance Wessels 1995), as for instance EU referenda in various member states, as well as general public opinion polls, have provided contradictory results as to whether it is the older or the younger generations who are most likely to vote `no' or share eurosceptical sentiments.72 Bernard Wessels, however, shares the assumption that euroscepticism generally decreases over time. He sees this as an instance of his `dissemination thesis', which holds that growing familiarity with the EU reduces socio-demographic differences in attitude (Wessels 1995; see Section 4.4 below). A different, non-demographic perspective on the development of hard euroscepticism is that as European integration develops in both depth and width, it is bound to move closer towards the wishes of some and further away from the wishes of others. According to this line of thought, hard euroscepticism is not age contingent but rooted in public perceptions of the `EU-of-the-day'. The thesis returns to this discussion in Part Four, although it should be emphasised that it does not engage in a socio-demographic analysis of what groups of citizens are most likely to be eurosceptic. 72 A few figures highlight the unclear picture: In Spain, the Netherlands and Luxembourg, for instance, 18-24 year olds represented the highest share of `no' votes in the referenda on the European Constitutional Treaty. The picture was less clear in France. To the general question of support for membership, 18-24 year olds constitute the biggest group of supporters age-wise in Denmark and the United Kingdom (Flash Eurobarometer 168, 171, 172, 173; Standard Eurobarometer 65, Denmark and United Kingdom reports). 64 4 Types of euroscepticism: the secondary level The purpose of this chapter is to deduce and develop constitutive types of public euroscepticism from existing literature. As mentioned in Chapter 2, this way of deriving constituent components of a concept is widely applied in the field of concept analysis (see for instance Gerring 1997; Sartori 1997; Weyland 2001). Section 4.1 introduces the theoretical framework, which the thesis provisionally summarizes under the heading of euroscepticism theory. This heading invariably remains provisional, as the thesis' aim precisely is to create the fundamental categories that are required before a true theory of euroscepticism, able to stipulate the conditions under which, given x or y assumptions, euroscepticism comes into being or assumes this or that type.73 Within this broad field, two crude bodies of literature assist our understanding of aggregate public EU attitudes: a general approach offers an inquiry into euroscepticism as a legitimacy problem for the EU (Section 4.2), and six more specific approaches offer clues as to the diverse nature of sceptical attitudes (Sections 4.3 to 4.8). Each section ends with an evaluation of the approach's contribution to the thesis' conceptualisation of euroscepticism. Section 4.9 is a summary. 4.1 Euroscepticism theory `Euroscepticism theory' is a term introduced by the thesis as an attempt to denote and group literature that is mainly occupied with explaining (sceptical) public attitudes towards the EU. Euroscepticism theory may depart from a variety of different disciplines and adopt a variety of different foci and levels of analysis--one prominent area deals with voting analysis. Although the term in itself may be not familiar, it thus merely indicates the grouping together of diverse approaches dealing with the same object of study in order to aid our conceptualisation of euroscepticism. Albeit a research area of growing interest (see for instance Hooghe and Marks 2005; Ray 2007), existing work on euroscepticism is still scarce and often highly empirical. Indeed, it is in this connection important to draw a distinction between studies of euroscepticism and studies of public attitudes towards the EU more broadly. While the latter abound, it is not straightforward to make inferences about the former: different variables, and not just the value of the prefix, may be in play. At least, proEU arguments such as `the EU guarantees lasting peace in Europe' or `the EU protects European 73 To King, Keohane and Verba, a `social science theory is a reasoned and precise speculation about the answer to a research question, including a statement about why the proposed answer is correct. Theories usually imply several more specific descriptive or causal hypotheses' (1994: 19). It is clear that no such theory is readily at hand for the thesis' endeavour to understanding euroscepticism. This is not, however, the same as to say that the theoretical discussions in the thesis aiming to tease out the constitutive types of 65 values against US dominance' offer little insight into the nature of scepticism. Because of the scarcity of literature on euroscepticism, it is fruitful to complement existing studies with other approaches dealing with the nature of public opinion in a supranational community in order to put forward a thorough conceptualisation. Heeding the recommendation of Christopher Flood, suggesting precisely this, work on euroscepticism is in the sections below reviewed in connection with the already extensive literature on European integration, democracy, legitimacy, and questions of collective identity (Flood 2002). As mentioned in the introduction to this chapter, approaches to understanding (sceptical) public EU attitudes can be broadly divided into two groups. Within one, we find studies engaged with general inquiries into the legitimacy of the political system of the European Union (for instance Beetham and Lord 1998; Banchoff and Smith 1999; Scharpf 1999; Bellamy and Castiglione 2001; Schmitter 2001). Inspired to a considerable extent by the ratification crisis surrounding the Maastricht Treaty in 1992, this literature queries the justification with which the Union can exercise its influence on the everyday lives of citizens. In particular, it perceives euroscepticism as a serious pointer that the EU may be lacking normative justifiability. The other group of literature is concerned with identifying specific types of euroscepticism (Inglehart 1971; Feld and Wildgen 1976; Hewstone 1986; Eichenberg and Dalton 1993; Franklin et al. 1994; Niedermayer and Sinnott 1995; Anderson 1998b; Gabel 1998a, b, c; McLaren 2002, 2005; Rohrschneider 2002; Hooghe and Marks 2004, 2005; and so forth). These works are marked by tremendous diversity (Ray 2007), both in terms of object of study (which ranges from criticism of specific policies and/or institutions to scepticism towards the EU as a whole), level of analysis (micro(the individual), meso- (regional), and macro- (country/European) level) and methodology (case studies, qualitative studies, quantitative studies, election analyses, use of secondary data, etc.). As with the more general approach to public opinion in the EU, studies of specific types of euroscepticism exploded in the aftermath of the Maastricht ratification crisis. However, this is not to say that inquiries into the nature of public attitudes were absent prior to 1992. As will be developed below, despite their reputation of an exclusivist elite focus, the endeavour of classic theories of European integration to explain the how's and why's of the European integration process (Wiener and Diez 2004) inherently necessitated some consideration of the role and opinions of the broad public. To date no thorough conceptualisation of different types of public euroscepticism exist. As will be clear euroscepticism should not be informed by the same standards of falsifiability, generation of observations and truthfulness that `true' theories should (ibid: 19-20). 66 from the remaining sections of this chapter, this absence is not equivalent to a lack of studies into EU public opinion, but to the scattered and inconsistent accumulated knowledge that these studies have so far provided. The task of the thesis has therefore been to group studies according to their basic common denominator. Following the process of establishing an overview of the relevant literature, the thesis finds that it is possible to sub-divide contributions on the specific nature of aggregate level euroscepticism in terms of six main approaches, which exhibit an almost chronological evolution. They are (1) a utilitarian approach; (2) a post-materialist understanding of EU attitudes; (3) a sovereignty-based approach, inspired by theories of nationalism; (4) a national context approach; (5) a democratic deficit approach; and (6) a political contestation approach. Indeed, in the first years of co-operation within the framework of the European Communities, focus was predominantly on the demand for creating an internal market, which would bring about benefits to society. In as far as public scepticism was singled out, it was perceived of as judging the EU in terms of its ability to actually provide these tangible benefits, or as the inclination of almost extremist, `anti-groups' in society, for instance individuals with strong communist or anti-market ideologies. The thesis finds two common denominators of this utilitarian understanding of public opinion (Section 4.3), namely that there is both an economic (money-based) and a performance-based aspect to public euroscepticism. In the 1970s, more substantive explanations as to why `anti-groups' were persistently eurosceptic emerged. Most widespread was Ronald Ingleharťs generational logic that materialistic-oriented war generations would be apprehensive of integration, which was the informed strategy of cosmopolitan, cognitively mobilised post-materialists (see for instance Inglehart 1971). Indeed, the basic assumption of these lines of thought is subsumed in Ingleharťs idea of the Silent Revolution, discussed in Section 4.4. However, with the gradual acknowledgement that large groups of so-called post-materialist citizens might also be hesitant about an ever closer union, despite its potential utility, scholars turned to theories of nationalism to understand euroscepticism. The importance of these accounts was in fact recognised already in the late 1960s (for instance Deutsch 1967: 250-251), but experienced a revival following the Maastricht Treaty ratification debacle in 1992. The Danish `no' suggested that public concerns about national integrity might pose limits as to how far EU integration could go. A common denominator for a growing number of studies into EU public opinion thus became sovereignty-based concerns (Section 4.5). Perhaps because this conclusion posed a somewhat disappointing perspective to pro-EU 67 politicians and academics, scholarly focus on euroscepticism following 1992 centred not so much on concerns about national sovereignty as on concerns about the democratic standing of the Union. The `democratic deficiť thesis became a buzz-word. Indeed, as Fritz Scharpf notes, given today's widespread concern about the EU's democratic standing, it is remarkable that through most of the history of European integration this was not an issue at all (Scharpf 2003: 2). The thesis groups literature construing euroscepticism in terms of dissatisfaction with the particular structure of the EU or in terms of more specific identifications of an alleged deficit under the heading of `democratic euroscepticism' (Section 4.7). A rather different approach took issue with the growing popularity of the democratic deficit thesis. Building almost exclusively on analyses of EU elections, theories of protest voting posited themselves as another influential explanation of euroscepticism, which actually had little to do with the EU itself, and more to do with the lack of popularity of the incumbent member state governments--or of a general dissatisfaction with politics. Perhaps, the EU was to a large extent an innocent recipient of `no' votes given at the Danish, and later the Irish and Swedish, negative referendums, and neither the actual issue of contention at disappointing elections to the European Parliament. Below, studies sharing this assumption are grouped under the heading of `the national contexť (Section 4.6). The French and the Dutch no's to the Constitutional Treaty in 2005 were, however, largely unpredicted by the existent academic literature on euroscepticism: neither democratic, nor sovereignty, concerns--nor protest votes--seemed to go all the way in explaining the new rejections from founding members of the Community. Partly in response to this alleged failure, new studies into euroscepticism seem to converge around a `contestation' understanding of EU attitudes. They build mainly on an axis of contestation well known from the political system of the member states, namely the left-right (or `social-liberaľ) schism of politics. However, the thesis finds that the novel and distinguishing contribution of this approach is its recognition that disputes surrounding EU politics may be conceivable as issues of `normal political contestation' also in the eyes of the broader publics (Section 4.8). The chapter proceeds by examining each of these approaches to euroscepticism. Thus, it discusses the relationship between euroscepticism and legitimacy deficits before it divides existent studies on public euroscepticism along the six groups outlined above, according to the thesis' assessment of their common denominator(s). This examination serves a twofold purpose. It demonstrates the `state of the arť of existing knowledge on aggregate level euroscepticism, while providing an operational platform, at the end of each section, for identifying the contribution of the individual approaches towards the 68 thesis' own conceptualisation of euroscepticism. Such a discussion of existing theoretical understandings of euroscepticism invariably reflects diverse and sometimes competing theories. In Part Three of this thesis, an empirical analysis seeks to clarify whether or not all the deduced types of euroscepticism are in fact prevalent in the member states, and which are dominant. This coupling with empirical data opens up opportunities for theory evaluation in the concluding chapter of the thesis. Indeed, since the different approaches to euroscepticism offer competing views on the development of the European integration process and the nature of public opinion, empirical analysis allows us to evaluate their relevancy in the face of actual manifestations of euroscepticism. 4.2 Euroscepticism and EU legitimacy Philippe Schmitter has observed that only when a regime is manifestly challenged by its citizens do political scientists invoke `lack of legitimacy' as a cause for the crisis (Schmitter 2001: 1). This was evident in discussions of the scepticism that were voiced by the Danish population towards the Maastricht Treaty in June 1992. Not only was euroscepticism forced onto the political agenda in Europe: on a more fundamental level, hitherto unasked questions about the ultimate goals and methods of the overall process and justification of European integration surfaced: the Maastricht debacle `forced Europhiles and Eurosceptics alike' to confront the legitimacy of the Union (Bellamy and Castiglione 2001: 1). The immediate appeal of this debate following the first public `no' underlines how the issues of public euroscepticism and legitimacy are often conjoined in almost inseparable union. Thomas Banchoff and Mitchell Smith, for instance, take the troublesome ratification process of the Maastricht Treaty as the point of departure for their study of EU legitimacy (Banchoff and Smith 1999). They perceive the Danish 1992 referendum as a demonstration of a two-dimensional `crisis of legitimacy' within the EU (ibid: 1), reflecting simultaneously the lack of popular identification with the Union and the undemocratic structure of its institutions (ibid). This line of reasoning can be paraphrased as an equation, holding that the lack of recognition by political subjects (cf. Weber 1968) of the raison-ďtre of the EU together with their dissatisfaction with accountability or representation in the Union, translates into euroscepticism, which then in and by itself constitutes a legitimacy crisis. The authors are of course well aware that legitimacy is a contested concept (Philippe Schmitter holds that it is `one of the most frequently used and misused concepts in political science', ranking `up there with `power' in terms of how much it is needed, how difficult it is to define and how impossible it is to measure'; Schmitter 2001: 1), and recognise the need to 69 qualify the link between low levels of public support and a general crisis of legitimacy in the EU (Banchoff and Smith 1999: 8; see also Ehin 2001). The thesis briefly reviews some of the more influential diagnoses of the EU's alleged legitimacy problems, before discussing if, when, and to what extent we may speak of euroscepticism as a pointer that the EU is lacking legitimacy in the eyes of the citizens. Indeed, the ease with which one is able to categorise something as a legitimacy deficit clearly rests on the definition one adopts of legitimacy. 4.2.1 Approaches to EU legitimacy Legitimacy is rarely conceived of as an all-or-nothing concept; instead scholars have sought to clarify and analyse its possible components. With specific reference to the EU, influential distinctions have been drawn between direct and indirect legitimacy; input and output legitimacy (Scharpf 1999); external and internal legitimacy (Bellamy and Castiglione 2001); and legality, normatively justifiability and legitimation (Beetham and Lord 1998). As we shall see below, there are clear links between these approaches. The term legitimacy in itself is difficult as it can be approached from a variety of different perspectives (for instance Weber 1968; Lipset 1984; Beetham 1991). As the object of study of this thesis is the types of scepticism that citizens share vis--vis the EU, it understands EU legitimacy empirically, as the EU's right--in the eyes of citizens--to hold and use power or exert its influence on everyday life. It follows that legitimacy shortcomings or deficits arise if the EU is not perceived by the public to be in accordance with accepted or expected standards, thus a serious shortcoming of a political system with democratic aspirations. The thesis' focus on legitimacy is invariably more concerned with perceptions of legitimacy than with actual correspondences between the functioning of the Union and ideal (theoretical) models of legitimate organisation. Put differently, under scrutiny below is not what scholars may reason is legitimate (normative legitimacy), but rather what can be observed to be perceived as legitimate to the wider public (empirical legitimacy). Direct/Indirect. Studies of EU legitimacy often begin with a conceptual discussion of what kind of polity the EU is, and the resulting conceptualisation (for instance of the EU as an international organisation, a federation, a system of multi-level governance or a super state in the making) has important implications for the kind and degree of legitimacy that the Union needs or can aspire to attain. A crude, almost paradigmatic, demarcation line distinguishes scholars who call for a rethinking of established conceptions of legitimacy to recognise the unique character of the EU (for instance Banchoff and Smith 1999) from those who to a larger extent perceive the Union as the extra arm of the member states, in little need of its own direct legitimacy (perhaps most famously Moravcsik 1998; 70 Majone 1996). The intergovernmentalist approach to European integration, exemplary of the latter perception, maintains indirect legitimacy as the main foundation for the EU. This type of legitimation rests on largely invariable external standards: if the participating member states are recognised as legitimate and their agreements are compatible with international law and norms, then the arrangement they create and join--here the EU--is legitimate (to the extent that politicians with federalist inclinations do not push through with issues such as a Constitution for Europe; see also Bellamy and Castiglione 2001: 8). Scholars conceiving the EU as a `regulatory state' (Majone 1996) also conclude that EU policies are not in need of direct democratic legitimation. In this view, the EU's primary purpose is to promote regulations approximating Pareto efficiency, which do not require, for instance, the direct participation of the governed in policy choices. At least, as long as EU politics are seen as fairly uncontroversial or not salient politically, the need for explicit democratic legitimation is weak. To such lines of thought, legitimacy deficits and public euroscepticism are two rather disentangled concepts--neither democracy, nor a common identity, is required at EU level--and beyond arguably minor anarchistic groupings, there is little dispute as to the legality of the European nation-states. Many scholars have been keen to point out, however, that as the EU with time has increasingly come to resemble a polity with a regime of its own (Bellamy and Castiglione 2001), the sole foundation on indirect legitimacy becomes problematic (also Scharpf 2003). At least since the Single European Act in 1987, the EU has assumed competency over socio-economic issues of arguably high political salience, which has clearly crystallised divisions between the member states. Most citizens, moreover, arguably do perceive the EU as impacting on their daily lives,74 and it is in this contested vacuum between national sovereignty and growing supranational authority that other approaches to Union legitimacy have developed. `One thing has become abundantly clear', Eriksen, Fossum and Menéndez suggest, `we Europeans can no longer keep on pretending that the Union is a mere economic enterprise' (Eriksen et al. 2005: 1). Several scholars have thus called for the EU to live up to both the direct and the indirect criteria of legitimacy (see Hix 1998; Banchoff and Smith 1999; Eriksen and Fossum 2000; Eriksen et al. (eds.) 2005), which necessitates the involvement of citizens in the law-making process. Recognising the need for the EU to achieve its own direct legitimacy allocates a potentially prominent role to various types of euroscepticism. Citizen critique of inadequate institutional channels for involvement, and contestation about the character and scope of the Union, are two examples of public scepticism that could surmount to a perceived deficit of direct legitimacy in the EU. It is clear that these types of scepticism may depart from diverse ideologies and approaches as to what the nature of EU co-operation should 74 As a glance at tabloid papers in the United Kingdom, for instance, may remind us. Whether large-scale impact is in fact real is another issue. 71 be like, which studies of Union legitimacy rarely take onboard. The thesis returns to this discussion below during the examination of the more specific approaches to understanding euroscepticism. Input/Output. Fritz Scharpf introduced input and output as the two core components of legitimacy in democratic Westerns societies (for instance Scharpf 1999). Input legitimacy rests on the trust by the governed that the governing process is responsive to their preferences (`government by the people'); while output legitimacy (`government for the people') rests on the expectation that adopted policies will represent effective solutions to the common problems of the governed (for instance Scharpf 2003: 3). Input and output legitimacy is thus inextricably linked in modern, well-functioning nation-states: trust is offered in return for delivery, and delivery is only justified in as far as it lives up to the invested trust. As regards co-operation in the European Union, Scharpf recognises the absence of a strong EU identity amongst the member states' populations, which is often seen as a prerequisite for trust. However, the Union is not construed as being in need of improving efforts to generate this identity either, which is why Scharpf instead emphasises the importance of output-oriented legitimation. This type of legitimation, it follows from the above, is based on common interests rather than on a common identity. The EU may, in other words, have to come to terms with a chronically low input legitimacy (which could well inspire euroscepticism). As Erik Eriksen and John Fossum point out, it is important to bear in mind that functional results themselves may be in need of further legitimation and that there can hardly be expected to be consensus among the EU member states as to what constitutes the `common interesť (Eriksen and Fossum 2000: 4-7). Contrary to common beliefs, output legitimacy in the EU may, in other words, be as difficult to achieve as input legitimacy, something which may cloud the at least short-term success of recent proposals of a `Europe of results'.75 Internal/External. Another conceptualisation of legitimacy proposed by Richard Bellamy and Dario Castiglione involves an internal and an external dimension. The internal dimension of legitimacy reflects the ways in which people relate to each other and to the institutions governing their lives. It arises from a fit with socially accepted norms, customs and beliefs, and formalised processes of authorisation through direct or indirect forms of consent (Bellamy and Castiglione 2001: 3). The external dimension of legitimacy stems from the justification of the rationale of the political institutions, and their congruence with certain formal and substantive norms, such as legality and human rights (ibid). It follows with regard to public opinion that the EU would be experiencing a problem with its internal legitimacy if citizens did not feel familiar with, or trusted, the EU's 75 A `Europe of Results' has been strongly promoted by politicians such as Barroso, Sarkozy and Fogh Rasmussen. See Chapter 7 for a fuller discussion. 72 institutions, or the populations of other member states, and manifested this lack of trust at EU elections, either by abstaining or by voting for clearly anti-EU groups. Karen De Jonghe and Peter Bursens identify two concrete variants of a lack of internal legitimacy: an institutional variant, pointed at the EU's alleged democratic deficit, and a social variant, manifesting through low popular identification with EU governance (Jonghe and Bursens 2003). External legitimacy problems would arise if citizens questioned the legality of the EU, for instance its compatibility with international law, or perceived of the Union as oppressive or unjust towards outsiders (ibid: 7-9). Legality/Normative Justifiability/Legitimation. The perhaps most comprehensive multi-dimensional conception of political legitimacy within the EU was proposed by Beetham and Lord in 1998. In their view, the EU is a political system in need of citizen support and loyalty, which may be achieved through adherence to certain legitimacy criteria applicable to liberal democracies. The authors identify and discuss three dimensions of legitimate political authority: a political system (a) has to be legal, i.e. its authority must be acquired and exercised in accordance with established rules; (b) has to have normative justifiability, i.e. its rules must be justifiable according to socially accepted beliefs about (i) the rightful source of authority and (ii) the proper ends and standards of government; and finally (c) it must engage in processes of legitimation, i.e. its positions of authority must be confirmed by the express consent or affirmation of the appropriate subordinates as well as from other legitimate authorities (Beetham and Lord 1998: 2-3). Theoretically, euroscepticism could arise at all three levels. Thus, citizens could question the EU's legality, disagree with the EU's aims and standards, or challenge the EU's efficient functioning by refraining from legitimation activities, such as participation in EU elections, or by voicing continuous negative feedback in opinion polls. However, it is in particular the dimension of normative justifiability, which poses questions as to the relationship between euroscepticism and legitimacy by directing our attention to the issues of defining who constitutes the people and what constitutes the rightful scope and ends of authority-- making questions of political identity and inclusion/exclusion crucial for legitimacy (ibid 1998: 5-8). Moreover, while the link between a lack of support for the EU and citizens who do not recognise the legality of the EU, and citizens who do not engage in practices of legitimation, is rather straightforward--it is less so in the dimension of normative justifiability. Indeed, is any critique of the EU's scope and authority, or in other words: any kind of euroscepticism, to be regarded as a legitimacy deficit? Beetham and Lord accommodate this question by identifying a number of possible deficits within the dimension of normative justifiability, divided along the three components of performance, democracy and identity. 73 4.2.2 Legitimacy deficits and euroscepticism In Beetham and Lorďs conception performance deficits involve disagreement both as to the scope of the tasks the EU should undertake and as to the effectiveness with which it is able to perform these tasks. Democracy deficits occur as a result of a lack of accountability or popular authorisation, while identity deficits result from the absence of a sufficient sense of common identity amongst the peoples of the EU, resulting for instance in the lack of trust by a country in a minority position in the belief that the majority coalition will protect their vital interests (which are therefore strongly held on to). 76 Nevertheless, a number of issues with regard to the relationship between euroscepticism and legitimacy deficits are left untouched. It is clear from the above brief accounts that citizen support is central to (empirical) EU legitimacy, and that it is required at various levels. Depending on where a lack of popular support occurs, a different type of legitimacy deficit may enter into play: the critique of the EU's legality is distinct from the critique of its scope, which again differs from inertia at elections because of a lack of identification with the EU. It is, however, central to recognise that these levels are not independent of one another--as Beetham and Lord point out, attempts to rectify a legitimacy deficit at one level may have (negative) implications for another level (Beetham and Lord 1998: 2). It is, nevertheless, as stated above, important to qualify the relationship between euroscepticism and perceptions that the EU lacks legitimacy: A lack of public support does not automatically mean that the EU has a legitimacy problem (for instance Ehin 2001). This statement builds on the thesis' definition of legitimacy, which, to reiterate from page 70, understands legitimacy empirically to mean the EU's right--in the eyes of citizens--to hold and use power. This definition excludes from the classification `legitimacy crisis' the everyday criticism of, say, a particular EU policy: losing out in a political negotiation does not necessarily imply sharing the perception that the winning side has no right to pursue the given policy. The thesis argues that it is in fact only with regard to one manifestation of legitimacy and one manifestation of euroscepticism that it is meaningful to speak of a direct equation between the two concepts. As regards the former, severe disputes about the EU's external legitimacy, or legality, giving way to public perceptions that the EU rests on juridical illegality, human rights oppression or non-compliance with international law call the entire Union into question and pose, 76 In addition, one could argue that Beetham and Lorďs dimension of normative justifiability is distinguished by being the dimension where the political system has the least independent control. One could argue that most political systems have little control over the dimension of legitimation: if voters, for instance, refuse to turn up to elect their representatives, the use of force to rectify this situation will hardly qualify as an act of legitimation. This argument could be applied to the EU on the basis of the generally low turn outs at elections for the European Parliament. However, there are reasons why this argument should not be taken too far, the most important being that if people refuse to turn up to elect a new European Parliament this is very likely to build mainly on the prior absence of normative justifiability and not on deficiencies within the dimension of legitimation. 74 indeed, an indisputable legitimacy crisis. As regards the latter, the thesis suggests the utility of recalling Aleks Szczerbiak and Paul Taggarťs distinction between hard and soft euroscepticism, discussed in Chapter 3. While hard euroscepticism--the principled objection to membership itself--constitutes a direct legitimacy challenge to the EU, the picture is less clear cut when it comes to soft euroscepticism, i.e. the contingent critique of specific issues or developments. Indeed, as will be discussed in the below paragraphs, while it would be misleading to maintain that soft euroscepticism may never constitute a legitimacy deficit, it would certainly also be misleading to consider all soft scepticism towards the European Union a legitimacy deficit. 4.2.3 Vertical and horizontal euroscepticism The hard-soft typology of euroscepticism was useful in narrowing down the focus of the remainder of this section to soft euroscepticism. However, we also saw that it was lacking in terms of the achievement of a fuller understanding of the complex link between euroscepticism and EU legitimacy deficits. In order to sharpen our understanding of this, the thesis finds it useful to introduce a transcending conceptual distinction between vertical and horizontal euroscepticism.77 Vertical euroscepticism, as the figure below illustrates, departs from the nation-state in the sense that the EU is perceived as something exterior--something that is in addition to the `normaľ national political system. The EU is conceived of as inherently foreign, and something abstract it would be possible to do away with, should it be at odds with national preferences. Hard euroscepticism is clearly vertical; however, also citizens sharing a contingent critique of the Union may base their opinions on vertical assessments. To give a brief example of vertical, but yet soft, euroscepticism, consider citizens who accept cooperation in the European Union on the basis that it is beneficial, but who do no support EU involvement in issues they consider central to national integrity. This utilitarian acceptance and appreciation of the EU does not necessarily entail that co-operation is accepted as an integral and given part of political life. If for some reason the EU was perceived as not being able to deliver, alternative forms of co-operation could be envisaged. To such citizens, a proposal by the European Commission to make obligatory the flying of the EU flag on buildings of local authority in the nation-states could meet with resistance, as it could be seen as a step in the direction of weakened national authority. This particular `spin off' from co-operation would thus not be seen as legitimate or, phrased more precisely, it would not be seen as normatively justifiable. Nevertheless, this particular encroachment on national 75 integrity need not be considered `serious enough' to affect the overall support of EU membership. Horizontal euroscepticism, in contrast, departs from the perspective that the EU is an integral and permanent part of political life. This is the `euroscepticism of the day' that may arise as a consequence of a critique of specific policies, actions and/or leaders, and may resemble the patterns of contestation that take place at the national level. Figure 6: Vertical and horizontal euroscepticism EU EU Nation-stateHorizontal euroscepticism Again, a few examples may serve to clarify. Disputing the character and scope of EU politics has been an ongoing activity even prior to the foundation of the European Communities in 1957. Today, treaty negotiations are for instance marked by divisions as to the extent of the social character of the Union. Previously, the single currency was a major issue of contention. It seems fair to say that such disputes are an inherent trait of an evolving polity, which do not necessarily call into question its legitimacy. Neither does a deficit of indirect legitimacy necessarily arise if citizens perceive member state officials in charge of taking decisions at EU level (typically Council ministers) to be non-accountable to domestic parliaments. This scepticism may merely give rise to calls for democratisation of the Union, which in and by themselves are not equal to scepticism of the justification of the undertaking. When it comes to perceptions of a poor correspondence between the values of EU leaders or institutions and the values of citizens, this situation is hardly distinguishable from everyday political quarrels between supporters of the government and supporters of the opposition within nationstates--where, again, the legitimacy of the system is rarely called into question, even by voters of the opposition parties. Public scepticism about a specific outcome, or policy proposal, from the EU--a concrete example could be the services directive on the grounds that it weakens the social character of the EU--is neither in and by itself (however strong) the same as calling into question the legitimacy of 77 Note the distinction between the below definition and Lindberg and Scheingolďs conception of horizontal and vertical interaction in the EU. To their study, horizontal interaction among the EU's publics is indicative of identitive support, while Nation-state Vertical euroscepticism 76 the Union. Rather, it could be seen as `normal political contestation', similar to the disputes that occur within nation-states between, for instance, supporters of different parties.78 David Easton's model of the political system and, more precisely, his conception of political support may help to further clarify the distinction between vertical and horizontal euroscepticism. Indeed, the thesis argues that the existence of `diffuse' (or abstract) support for the political system contributes to distinguish horizontal support from vertical support. Easton famously distinguishes between three objects of support--the political community, the regime, and the authorities--and two types of support: diffuse and specific (Easton 1965, 1975). Diffuse support `forms a reservoir of favorable attitudes or good will that helps members to accept or tolerate outputs to which they are opposed or the effect of which they see as damaging to their wants' (Easton 1965: 273). It is thought to be deep-seated and rather robust attitudes, which enable leaders to weather severe storms when a political system is under stress--it guarantees, in other words, the stability of the system (see also Lindberg and Scheingold 1970; Norris 1999). Specific support is closely related to the actions of political leaders and the day-to-day performance of the system. It is sensitive to the level of perceived output that the system is able to deliver. In a political system, diffuse and specific support may be directed towards three objects: the political community, the regime and/or the political authorities. The political community, which to Easton denotes a group of people bound together by a political division of labour (for instance Schmitt and Thomassen 1999: 13), is perhaps the most basic level. Support here generates a `sense of political community' (see Scheuer 1999: 30), which to Easton represents the highest category of diffuse support for a political system (ibid). Indeed, support at this level is predominantly diffuse, and the stronger it is, the greater are the system's stress reducing capabilities (ibid: 30).79 Citizens may moreover direct their support towards the political regime. Here, Pippa Norris distinguishes between regime principles, processes and institutions (Norris 1999: 74-75), which she, with regard to EU support, operationalises in terms of overall attitudes towards European integration, satisfaction with the way democracy works vertical relations gauges support by these publics of the EU's system (Lindberg and Scheingold 1970: 40). 78 The possibility of fighting day-to-day political battles without calling the entire system into question was praised by the Danish Prime Minister, Anders Fogh Rasmussen (Liberal Party) in his New Year's Address of 1st January 2007, `I know that in our day-to-day lives we can disagree about many things. But if we take a step back and consider the whole picture, it is such a great strength for us Danes that it is possible for us to unite around the larger picture' (own translation; see Prime Minister's Office: www.stm.dk). 79 Scheuer confirms the presence of a substantial sense of community among EU citizens; however, her operationalisation of Easton's sense of political community is one that to a high extent taps a sense of Europeanness as opposed to identification around the (narrower) EU community--indeed, her indicators consist mainly of items polling identification with Europe and other Europeans. Thus, Scheuer's finding arguably does more to confirm Karl Deutsch's broader idea of the existence of a sense of community amongst European citizens than it confirms the existence amongst EU citizens of Easton's more demanding, or `thicker', sense of political community. Scheuer is aware of the distinction between Deutsch and Easton's employment of the terms, and it is a matter of argument that her own use in fact taps the former more than the latter. 77 in the European Union, and trust in the EU institutions (ibid). As these indicators suggest, regime support is likely to meet with both diffuse and specific types of attitudes. Citizens may on the one hand share a diffuse support for the underlying principles and values of a system, which need not be affected by a more specific support for actual institutional workings and processes on the other hand. Finally, citizens may be supportive of the political authorities--that is the incumbent government (or Commission), their preferred political party or individual candidates--as well as their proposed policies. This latter type of support is predominantly specific. Applying this understanding to the thesis' endeavour of establishing when euroscepticism represents a legitimacy deficit to the EU, it is useful to note that vertical euroscepticism is characterised by the absence of diffuse support. Vertical eurosceptics, in other words, have no `reservoir of favourable attitudes' that makes unpopular policies acceptable. Horizontal eurosceptics, on the other hand, share some sense of diffuse support for the EU political system. It should be clear from the above that while vertical euroscepticism constitutes a legitimacy challenge to the EU, horizontal euroscepticism does so no more than national critique of a poor performing government constitutes a national legitimacy crisis. This is not to say that there is no possible connection between specific attitudes and legitimacy--indeed, although efficient policy-making and satisfactory outputs (i.e. specific support), as Michael Marsh puts it, in themselves are hardly enough to render the EU more legitimate in the eyes of its citizens, legitimacy is unlikely to develop in their absence (Marsh 1999: 91). Specific support is thus a `necessary', but not `sufficienť, condition for keeping legitimacy deficits at bay. Importantly, the distinction between vertical and horizontal euroscepticism is conceptual. It was introduced with the intention of clarifying the less than straightforward relationship that exists between euroscepticism and legitimacy deficits. What is important to keep in mind is that to criticise a policy is not the same as calling its legitimacy, or the legitimacy of the entire system of which it is a product, into question. The thesis is careful to underline this lack of a direct relationship between most types of euroscepticism and legitimacy deficits, because while the latter constitutes a serious pointer to any political entity seeking to live up to European democratic standards that it is not well-functioning, euroscepticism may to a large extent have to be seen as an integral part of a democratic political system. The conceptualisation of at least some types of euroscepticism as integral to the EU is developed upon in the below sections, which also, where appropriate, seek to clarify the extent to which the findings from the actual measurement of the various types of euroscepticism represent a legitimacy deficit to the EU. The thesis recognises that a full investigation of whether or not specific types of euroscepticism reflect horizontal or vertical EU attitudes falls outside the thesis' scope and aim. The latter inquiry 78 would, for instance, require an investigation of individual level data and a combination of indicators that is vast enough to constitute an entirely new thesis. The empirical part of the thesis is therefore limited to discussing the likelihood of the various prevalent types of euroscepticism reflecting horizontal or vertical EU attitudes and to evaluating the degree of hard euroscepticism within each type. 4.3 Utilitarianism The broad utility thesis is an obvious point of departure for specific investigations into euroscepticism: chronologically, it represents a common denominator for the onset of endeavours to understanding sceptical public EU attitudes and it continues to be a major approach in the literature. Indeed, the hypothesis that utilitarian motivations are decisive for citizens' opinion of the EU recurs in numerous studies and has done so for decades (for instance Haas 1958; Andersen and Reichert 1996; Gabel and Whitten 1997; Gabel 1998a, b). Central to discussions of utility is the issue of tangible economic benefit. The key rationale is logically deduced: as the EU itself is driven by a largely economic agenda (integration started with coal and steel co-operation, and still today economic growth and development are central motivations for integration, and among the predominant responsibilities of the EU) the public evaluates the EU according to its economic achievements. This economic, utilitarian focus is explicitly shared by many analysts of Europe; for instance, as we shall see below, by advocates of the neo-functionalist and, in a different way, the liberal intergovernmentalist approaches. It arguably also played a prominent role to the EU's so-called founding fathers--Jean Monnet, at least, once commented that he thought the Community would be operating as a `public utility state' (cf. Bellamy and Castiglione 2001: 1). In the words of Eichenberg and Dalton, `if the EC has promised anything, it has promised the enhancement of member states' national economic welfare' (Eichenberg and Dalton 1993: 510; see also Bosch and Newton 1995: 76-77). In general, the importance of the economy for citizen support has been widely recognised by politicians in democratic societies. In the words of Martin Paldam and Peter Nannested, `people know little about the economy ­ but worry about it a loť (Paldam and Nannested 2000). David Easton's classic distinction between diffuse and specific support is often taken as the theoretical point of departure (Easton 1965): as citizens' affective (or diffuse) loyalties remain largely with the nation-state, the EU has to depend on securing utilitarian (or specific) support (for instance Gabel 1998s)--an easily changeable matter. Utilitarian approaches to understanding euroscepticism, importantly, do not all centre directly on 79 `money'. Indeed, important works focus on the importance for public support of EU efficiency; that is, the EU's ability to present itself as an entity capable of carrying out its policies effectively (an example of this logic could be the EU's role in assuring the ability and right to travel and settle freely--and easily--across internal borders). Karl Deutsch's contribution to European integration theory, Transactionalism, foresaw in the late 1950s that the development of horizontal relations across member states (increased transactions and communications, leading for instance to higher levels of trust between populations) would further a `sense of community' that would underscore public support for integration (Deutsch 1957).80 Indeed, integration was defined as the `attainment, within a territory, of a `sense of community'' (ibid: 5), which in turn was defined as `a matter of mutual sympathies and loyalty; of `we-feeling', trust, and mutual consideration' (ibid: 36). Although the thesis' interest lies with `vertical relations' (between a population and the EU) rather than on `horizontal relations' (between populations), it is of relevance with regard to the present section to note that this sense of community, which would further identitive81 recognition of the ties that link European citizens, would be `provokeď by the utilitarian recognition of the advantages from cooperating. Increased benefit and efficiency, in other words, precedes identification; however; both utility and identity were in Deutsch's view necessary for successful co-operation. As the paragraphs demonstrate below, the `information thesis' inherent within Deutsch's perspective contributes to our understanding of utilitarian euroscepticism.82 While acknowledging an increase in the amount of transactions, Deutsch admitted in a study from 1967 that there were little signs of the development of a closer sense of community amongst the populations of the EU. The equation between transactions and identification, in other words, appeared mistaken. Deutsch instead grew pessimistic on the development of the `ensemble of public moods' towards the EC. In 1967, proclaiming the slow-down in integration, he wrote: `[There is] now in European mass opinion a latent clash between the continuing acceptance of the reality of the nation-state and the newly accepted image of some vague sort of European unity. The ensemble of the present 80 Transactionalism has been criticised for assuming that it is in fact possible to measure both the sentimental relations among peoples and the amount of transactions that is thought to further integration. Also, to a large extent Deutsch's focus was on the creation of security communities, rather than on European integration per se (Diez and Wiener 2004: 11). As Diez and Wiener point out, rather few scholars followed Deutsch's lead in focusing attention to the social, rather than the political integration process (ibid). However, re-articulation of integration theories over the past few decades, partly inspired by empirical events directing politicians' attention to public attitudes, has arguably inspired a new interest in particularly the identity aspects of the analyses of Deutsch (Kelstrup 1998: 32). 81 The thesis uses the term `identitive' in the same sense as for instance Richard Sinnott (1994) and Tobias Theiler (2004). It roughly means the same as `identity-baseď. 82 Indeed, it is the underlying assumption of the information thesis inherent in Deutsch's writings on the development of horizontal relations that leads me to `place' him largely within the utilitarian approach to euroscepticism. The thesis does not wish to underestimate Deutsch's insistence on the importance of an identitive sense of community for a successful co- operation. 80 public moods would not be of much help to statesmen who would lead their countries toward a greatly deepened union. Rather, they may facilitate general expressions of good will, combined with policies of temporizing, caution, national consolidation, and only gradual and sectoral advance toward somewhat greater European integration. Bolder steps toward substantially greater European unity would have to be `sold to' mass opinion by the sustained and concerted efforts of leaders and elites--assuming that these elites had the capacity and motivation to do so in preference to other short-run goals' (Deutsch 1967: 250-251). Deutsch is cited in some length as the quotation exposes a number of assumptions about public attitudes towards integration useful for the thesis' attempt to establish a theoretical conceptualisation of euroscepticism. First of all, it is acknowledged that the continuing prevalence of the nation-state as a centre of people's loyalties might come to clash with developing ideas of European unity. Second, by stating that the image of European unity is `newly accepteď, Deutsch indicates that public opinion is a dynamic entity and that positive attitudes may increase with time. Moreover, the quotation stresses the importance for the development of European unity of efforts on the part of elites in selling the EU to citizens, assuming a positive relationship between information and public support. This `information thesis' is very prominent in the literature on EU attitudes and underpins the hypothesis that the `cognitively mobiliseď are the strongest supporters of European integration (see Section 4.4; Inglehart, Rabier, and Reif 1991; Janssen 1991; Anderson and Kaltenthaler 1996). It assumes a positive relationship between (all) information about integration and support for the EU, and thus contends that public support for new developments will follow when people become adequately informed about them (for a critique see for instance Gabel 1998c; Moravcsik 2006). As suggested above, there is arguably a strong utilitarian logic to this equation: if the EU was not perceived as a beneficial, delivering entity, informing about it would hardly translatable into public support. The information thesis has remained prominent, more or less explicitly, in a range of newer approaches to euroscepticism, the difference mainly being that prior to 1992, convincing the public of the EU's merits was not perceived as a complicated or ambiguous endeavour: it was merely a question of the allocation of enough resources to informing about the EU, perhaps combined with a little patience. Following the Danish rejection of the Maastricht Treaty, the ease of the information task was no longer assumed (the Danes counted as one of the more informed publics in the EU), but (inadequate) information as such continue to be perceived as perhaps the main deficit in face of euroscepticism: `Informer, informer, encore informer', as the advice sounded by Guy Carcassonne, a French 81 professor of law, on national TV prior to the referendum on the Constitutional Treaty in May 2005.83 In the wake of the French and Dutch no's, the strong focus on information has been somewhat modified by the complementary stress on `communication'--perhaps in the attempt to inspire sentiments of empowerment in citizens. It has become accepted that generating support is not just a matter of informing citizens, but a two-way street of communicating and debating EU politics. Deliberation, in other words, is seen as the way to counter public euroscepticism. However, ideas of deliberation arguably build on a similar logic, and remain contested (for instance Moravcsik 2006). At least, as mentioned above, the information and communication theses presuppose that citizens share a particular type of euroscepticism. This especially pertains to utilitarian euroscepticism. Indeed, not all types of euroscepticism can be expected to be as malleable. If citizens predominantly build their EU attitudes on democratic grounds, more information about the actual state of the Union might arguably merely reinforce underlying cleavages. Similarly, if it is sovereignty concerns that constitute euroscepticism, information would not be an effective remedy either. The neo-functionalist perspective on European integration also shares the hypothesis that public support for the EU is predominantly tied to perceptions of utility. Indeed, when neo-functionalists talk of the `functionalist goals' of integrating and of the need for `effective performance' in the EU, they do so from a largely utilitarian perspective of co-operation. Neo-functionalism is arguably the most sophisticated and complex of the classic integration theories, and while it has been criticized for being too elitist in focus, it does in fact contain several assumptions about the nature of public opinion towards the EU. The main thrust of the perspective of the early 1970s was that public opinion, largely disinterested or passively compliant, would follow much the same logic as the opinion of political actors: citizens would gradually shift their loyalties from the national to the European level when becoming aware of, and getting used to, the functionalist, utility maximising requirements of policies. Citizen support, in other words, was seen as a function of the efficiency of the EU in producing policy. It was thus assumed that integration would almost automatically foster increased support among the populations of the member states. This support generating process, its proponents admitted, did place some demands on the EU's ability to perform (or produce) efficiently in the eyes of citizens (for instance Marsh 1999: 90). To its perhaps leading proponent Ernst B. Haas, public support might involve increased contact and familiarity with the EC, education and progressively rewarding experiences derived from the activities of the common market (Haas 1971, from Niedermayer and Sinnott 1995: 20). These elements have in common that they build on the assumption that familiarity 83 `Inform, inform, and inform again' (my translation). Television Française 1, 20th April 2005. 82 and information about the EU produces support if tasks of public utility are efficiently performed by the Union. As such, they thus open the door for a type of euroscepticism that is based on the critique of lacking benefits from the EU or the inefficiency of the EU's set up. While the performance-based assumption remained at its core, neo-functionalism abandoned its assumption of an almost automatic spill-over process rather early. Perhaps inspired by Deutsch's forceful insistence on the centrality of sentimental relations among peoples, a (rather short) revisionist phase of neo-functionalism in the late 1960s and early 1970s arguably placed more emphasis on the study of public opinion (see Sinnott 1995: 19-23 for a review of neo-functionalist acknowledgements of the importance of public opinion). In 1971, Haas, for example, addressed the problem of linking `variables that describe the rate of transaction between units with variables that describe the attitudes of masses and members of the elites' (in Sinnott 1995, my emphasis). In 1975, he further explicated his understanding of public opinion towards the Community, maintaining a clear utilitarian logic: `the public is (...) concerned with income, price stability, better working conditions, cleaner air, more recreational facilities (...) [and] does not greatly care whether these are provided by the national government or by Brussels' (Haas quoted in Bellamy and Castiglione 2001: 7). The quote is interesting as it promotes utilitarian considerations as the dominant determinant of public opinion: European co-operation is justified to the extent that it delivers. In Section 4.5, this view is contrasted to that of federalists, who engage a more bottom-up, identitive, perspective on support, and `sovereigntists', who disagree with especially the latter part of the quote by Haas. Another prominent integration theory gaining ground in the early 1990s, Liberal Intergovernmentalism, also opens the door for a utilitarian euroscepticism (though, as we shall see in Section 4.5 on sovereignty-based euroscepticism, the utility type is arguably not its weightiest conception). It offers a somewhat different thrust than neo-functionalist ideas about utility and public opinion: utility is the very raison-ďtre for co-operation in the EU but the utility argument has important limitations, and remains inferior, or insufficient, to other concerns. The deterministic logic of spill-over, central to neofunctionalism, is thus rejected by liberal intergovernmentalists. Nation-states remain sovereign and independent actors in crucial matters even if, from a purely utilitarian calculation, superior economic gains could have been achieved through supranational co-operation. Nevertheless, as stated, utility is still the EU's raison-ďtre: if it were not for tangible benefits, economic and political, co-operation would not be desirable in the first place. Euroscepticism, it follows, emerge as the critique of lacking benefits. Interestingly, Liberal Intergovernmentalism offers an additional perspective on the relationship between public opinion and utility. The public may `deselecť the EU level based on a rational 83 calculation juxtaposing the relative lack of salience of the issues decided at the supranational level with the substantial amount of energy it would take to mobilise and participate. True democratic participation assumes a redefinition of existing political identities, which is both utopian and undesirable. It is simply `extremely expensive' for the individual citizen to deliberate (Moravcsik 2006: 226227), and this explains the current lack of widespread public engagement with the EU--perhaps, in the words of Hermann Schmitt, citizens choose `rational ignorance' (Schmitt 2005: 668). Leaving aside the ambition of providing a high-level theory of integration, contemporary scholars writing on the political community of the EU continue to embrace the equation between utility and public EU attitudes. They introduce an important complementary aspect to the utility thesis, namely the EU's role in providing protection for the rights of citizens. German philosopher Jürgen Habermas' famous concept of `constitutional patriotism' (Habermas 1998) rejects the view that a prior background consensus based on a homogenous culture is necessary for integration in the direction of a postnational society. Instead, shared rights may form the foundation for the emergence and unity of a European people (and even breed patriotic sentiments). To Habermas, peoples emerge with the constitutions of their states (Habermas 2003: 97). Citizens will come to identify with a construction like the EU when they realise that it provides the infrastructure by which all their other attachments (local, national, gender, sexual, occupational...) can be managed and prevented from coming into excessive conflict with one another (see Beetham and Lord 1998: 42, paraphrasing Habermas' argument). In this view, it is an empirical question when and to what extent modern populations understand themselves as a nation based on ethnic membership or as a `nation of citizens' (Habermas 1998: 129-153). Habermas' own position in this regard is deterministic, as it contends that all solidarities will need to shift away from exclusivist, ethno-cultural constructs as contemporary societies become more multicultural (Habermas, quoted by Beetham and Lord 1998: 42). Habermas' perspectives on public EU support are certainly not exclusively rights-based, let alone benefit-based. He recognises that `[e]conomic justifications must at the very least be combined with ideas of a different kind--let us say, an interest in and affective attachment to a particular ethos' (Habermas 1998: 3, my emphasis), and thus that economic advantages are valid as arguments for further construction of the EU only if they can `appeal to a cultural power of attraction' (ibid). However, Habermas' perspectives do build on the assumption that EU citizens to a large extent already share a common identity in the EU's foundational myth of the desire to avoid another world war from arising on European soil, as well as a telos in the common goal of an `ever closer union'. Habermas points to decisive historical experiences that unite European peoples, and holds that democratically structured opinion- and will-formation 84 make possible rational agreement even between strangers (Habermas 1998: 129-153). To Habermas, the stipulation in the EU's Charter of Basic Rights of the rights of EU citizens is thus an `admirable expression' of the awareness by Europeans of what they have in common (Habermas 2001: 10). Karl Deutsch's ideas have also found new expression through Habermas, who argues that dense networks of communication recently have been developed among European nationals, and that `the initial impetus to integration in the direction of a postnational society is not provided by the substrate of a supposed `European people' but by the communicative network of a European-wide political public sphere embedded in a shared political culture' (Habermas 1998: 129-153). Habermas' deterministic thrust has to be seen in context of `globalisation' processes, which in recent decades are thought to have furthered a multicultural setting in most Western societies, at the same time as national sovereignty has encountered limits faced with increased trans-border activity. To Habermas, the complex and overlapping identities that increasingly characterise Western citizens call for something like the EU to play a greater role. The EU, in this sense, obtains its raison-ďtre from being useful.84 To Matthew Gabel, a euroscepticism researcher emphasising the utilitarian dimension, public opinion is contingent on the degree to which citizens believe the EU is able to improve the economic situation of themselves/their country/Europe (Gabel 1998a, b, c). In other words, market liberalisation--which is seen as the EU's primary objective--provides different costs and benefits for different EU citizens, and these differences are decisive for their attitudes (Gabel 1998c). Whether or not citizens experience economic welfare from European integration may depend on a number of socio-demographic indicators, in particular their human and financial capital (education, occupational skill, income) and proximity to other EU markets (border region residence). In this sense, utilitarian euroscepticism associates closely with the polarisation hypothesis discussed among others by Andersen, foreseeing an increasing split between `resourceweak' and `resourcestrong' in public EU attitudes (Andersen 1998). Both national and international economic factors are of relevance to this type of popular attitude (Eichenberg and Dalton 1993). Testing the explanatory power of five different theories of public support for the EU, including 84 To put it more succinctly, similarly to the above discussion about Karl Deutsch' work, Habermas is discussed under the heading of utilitarian euroscepticism (instead of under a more identitive approach) in recognition of his adoption of a largely rights-based (utilitarian) justification of the rationale for a constitutional patriotism. In this vein, what could obstruct constitutional patriotism from developing are strong perceptions about the lack of rights guarantee by the EU. This rationale, for instance, distinguishes Habermas from, for instance Joseph Weiler, who, as we shall see in Section 4.5 below (on identity-based approaches to euroscepticism), has described a conceptualisation of European citizenship in terms of rights as an `end-of-millennium version of bread-and-circus politics' (Weiler 1999: 335). 85 Ronald Ingleharťs theory of the Silent Revolution (see Section 4.4 below), Gabel finds empirical support for the centrality of utilitarian EU attitudes (Gabel 1998c). This influential study lists cognitive mobilisation; political values; class partisanship; and government support as the backdrop to advancing the thesis of utilitarianism as the most potent in accounting for public opinion towards European integration. Indicators are tested using regression analyses of Eurobarometer surveys from the period 1978 to 1992. As dependent variable, Matthew Gabel creates the variable support, by combining two measures he identifies as frequently used, independently, by other (unspecified) researchers as indicators of `support for integration'. They consist of the Eurobarometer poll questions regarding (i) membership of the EC and (ii) unification of Western Europe. However, although Gabel is able to confirm a correlation between this combined measure and support for a number of concrete proposals for European integration (a common defence, the CFSP, the Euro, a European government), a number of questions are left untouched. The first, which Gabel shares with most other causal studies of public EU attitudes, concerns the adequacy of working with one relatively simple dependent variable, or, in other words, the assumption that it is possible to establish a relatively simple measure of public support for, or scepticism towards, integration. The value of Gabeľs empirical test of five independent variables depends, vitally, on the actual suitability of attitudes toward membership and Western unification as a proxy for support for integration. There are, as this thesis strongly argues, reasons to be cautious towards this. First, the two poll questions will remain an unclear measure of support for integration in as far as Gabel does not define what he implies by `integration'. It is not farfetched to state that different understandings of integration exist, and that these may in fact be pointing in rather different directions (see Section 1.3 on terminology). Second, while it is relatively clear that support for membership is largely intertwined with general support for integration, however this may be defined (the first leg in Gabeľs dependent variable), it is less clear that this should be the case with support of Western European unification (the second leg of the dependent variable). Indeed, this poll question may be difficult to categorise (in several prominent studies it is understood as a proxy for affective support for the EU, for instance Hewstone 1986; Inglehart and Reif 1991). It is unclear what `unification' refers to in the poll question. A closer look at how respondents have replied to it throughout the time period surveyed by Gabel reveals discrepancies in the relative positioning of some member states to the two questions making up his dependent variable. The United Kingdom has placed itself in the most sceptical end of the EU-12 when it comes to attitudes towards membership of the Union, but alongside the EU average when it comes to 86 attitudes towards the unification of Western Europe.85 One explanation could be that the British associated the issue of `unification' with the bringing together of East and West Europe following the end of Communism, and therefore supported it to a larger extent than the issue of membership, which is more directly associated with the European Union. In any case, such substantive ambiguity leaves the measure a doubtful proxy for generic support for the EU. The thesis returns with new perspectives on the `dependent variable problem' in Chapter 7. Gabel selects his five independent variables from existing literature on public attitudes, but does not mention whether he sees them as exhaustive accounts. They are likely not to be. The almost intuitively important dimension of sovereignty (see Section 4.5) is, rather surprisingly, not included in the analysis and when he concludes by identifying utilitarian (or economic) motivations as the more convincing explanation of public support for European integration, we are left to consider how the sovereignty dimension would have fared in the analysis. It is moreover important to note that, although published in 1998, Gabeľs data only examines attitudes prior to 1992--a year which the below analysis will reveal appears to be a turning point for public opinion in the EU. However, given that the relative success of utilitarian indicators in Gabeľs study confirm the above testimonies of the importance of economic benefit and effective performance to EU attitudes, it does seem relevant to further investigate the claim that at least part of the publics in the member states of the EU build their opinion about integration on the ability of the EU to be useful--both in terms of financial rewards and in terms of effective policy-making. 4.3.1 Contribution It is clear from the above that a number of studies share the assumption that utilitarian calculations are the determining explanation of support for membership of the European Union. Although the utility thesis is not uncritically accepted by all analysts of European integration, it has since the foundation of the European Communities represented an overwhelmingly weighty common denominator to a variety of different approaches to euroscepticism. The thesis finds, in light of the strengths of these accounts, and the EU's explicitly economic rationale, that it makes sense to also expect that an important part of public opinion in the member states bases its evaluation of the Union on whether it is able to bring benefits. In other words, the thesis expects a utility-based critique to constitute a powerful constitutive 85 In Eurobarometer 27 from 1987, one of the years surveyed by Gabel, 16 percent of the British public were against the unification of Western Europe--thus only somewhat more sceptical than the EU average of 10 percent (it could be mentioned that Danish scepticism was at 47 percent). As regards the membership question, the British distinguished 87 type of euroscepticism, and continues with an investigation of its empirical basis in the case countries in Part Three. In light of the above accounts, however, the thesis finds it appropriate to distinguish between two types of utilitarian euroscepticism: (i) `economic euroscepticism', namely the critique of a lack of economic benefit from co-operation, and (ii) `performance-based euroscepticism', namely the critique of the inefficiency of the EU's setup and functioning. The crude common denominator of the first type is money, while it is the equation of the EU with a bureaucratic and fraudulent polity as regards the second type. 4.4 The Silent Revolution and ideological euroscepticism `Post-materialism' and `cognitive mobilisation' are the two central concepts in Ronald Ingleharťs thesis of the Silent Revolution (introduced in 1971). This was perhaps the thesis the most often used to explain--at the individual level--public feelings towards European integration prior to the Danish referendum on the Maastricht Treaty in 1992. Joseph Janssen, writing in 1991, found that it was frequently cited as `the' explanation for public support for Europe (Janssen 1991: 444). To this perspective on European integration, what constitutes euroscepticism is ideological or value-based contestation about the nature or direction of the EU. Of its two central concepts, `post-materialism' and `cognitive mobilisation', the former underscores the hypothesis that post-materialists have a more supportive attitude towards European integration and the EU, than do materialists. Ingleharťs claim is that the issue of European integration fits in better with the value orientation of post-materialists (or cosmopolitans), and fulfils their intellectual needs and broad horizons. This is in contrast to materialists, who, to Inglehart, are preoccupied with material needs and have less time to devote to abstract issues like integration (Janssen 1991: 445). The latter component of the thesis, cognitive mobilisation, points to the claim that a high level of cognitive skill--political awareness, familiarity with the EU, skills often associated with higher levels of education--is necessary for understanding, and thereby appreciating, European integration. The more political skills one has, or the more one knows about the EU, the more one is able to cash in its benefits and the more one is positive towards integration and the EU. This equation, which is akin to the previously mentioned `information thesis' (see Section 4.3), would account for the frequently cited distinction between `elite' and `broader public' in support for the EU. The elite, i.e. the cognitively mobilised, know how to draw the benefits from integration in the EU, while the broader public is more themselves in a sceptic direction by 15 percentage points, with 26 percent finding membership a bad thing compared to 11 percent in the EU as a whole. 88 prone to experiencing its uncertainty and drawbacks. Ingleharťs theory thus assumes a positive relationship between information about integration and support for the EU (see also Inglehart, Rabier, and Reif 1991). Though the two components within Ingleharťs thesis are designated at the level of the individual, they have implications for the meso- and macro-levels (Janssen 1991: 446). The name `the Silent Revolution' indicates the occurrence of change, a change which according to Inglehart is happening on two fronts. First, there is a generational change (the meso-level) from materialism to post-materialism as pre-war generations, which due to exposure to war and hunger have a materialist value orientation, decrease in proportion compared to post-war generations, which are expected to be occupied with more postmaterialist values--and thus more supportive of the EU. Second, recent general increases in levels of skill amongst the total population in the member states (the macro-level), due to, for instance, generally higher levels of education, should also over time generate higher levels of support for European integration and the EU (Janssen 1991: 446). In the early 1990s, however, precisely these ideas of Inglehart were thoroughly criticised for being empirically unfounded (Janssen 1991), and the entire thesis of the Silent Revolution was arguably dealt a serious blow by the Danish rejection of the Maastricht Treaty in 1992. If post-materialist values are characteristic of modern European nation-states, they are arguably especially so in the affluent Scandinavian countries. At least, in the early 1990s, post-material issues such as environmental protection, social security provisions and gay rights were to the forefront in Denmark to perhaps a higher extent than anywhere else in the EU-12. Christopher J. Anderson even found empirical proof for the claim that post-materialism was negatively correlated to support for EU membership (Anderson 1998b: 586, cf. Anderson and Reichert 1996); and Matthew Gabel was able to refute the centrality of both the cognitive mobilisation and the post-materialism components of the thesis (Gabel 1998c, postmaterialism was operationalised as political values). A lack of consistent empirical support in the data and the absence of controls for other, potentially confounding factors such as education were Gabeľs grounds for refuting Ingleharťs hypotheses (ibid, see also de Vreese 2004: 3, 10). Bernard Wessels moreover asserts the relevance of intra-societal diffusion as a counter argument to the assumption of generational, or demographic, change (Wessels 1995). Wessels' argument holds that differences in attitudes between socio-economic and other status groups towards the EC tend to decline as a consequence of growing familiarity within European co-operation, not as a function of age (ibid). Moreover, Danes were, as mentioned, consistently among the EC populations that declared themselves the most knowledgeable about the EU, and turn-out in both national and European 89 elections have generally been higher in Denmark than in most other member states.86 According to Ingleharťs thesis, Danes should be keen integrationists. The rather obvious counterproof of this provided by the Maastricht referendum may have contributed to spur the development of rival theories of euroscepticism (see following sections). Today, Ingleharťs thesis can hardly be said to be among `the' explanations of euroscepticism. In his writings on euroscepticism, Anthony Forster has sought to broaden the rather narrow focus on post-materialism to a more general focus on the role of ideology in constituting attitudes. Forster shares a two-dimensional approach to euroscepticism, arguing that in addition to an `ideology axis', it is possible to identify a `sovereignty axis' of the phenomenon (Forster 2000: 100-101, 2002). The sovereignty axis represents those citizens for whom the centrality of the nation-state is a main reason for scepticism (see Section 4.5 below), and the ideology axis represents citizens sceptic of the type of Europe on offer. What distinguishes ideological euroscepticism is, in other words, the misfit between the values of citizens and the values represented by the EU-of-the-day. There is, however, little agreement as to what constitutes the values or ideologies that are contested in the Union, and more broadly what should be defined as an ideology or a value. Previously, academics attempting to evaluate the broader role of `ideology' on support for the EU, have retorted to somewhat dubious indicators. Miles Hewstone, a social psychologist, for instance, included support for the European Parliament as a main indicator in an influential study of public attitudes towards the EU (Hewstone 1986: 23). However, it is unclear what this indicator actually tells us. At least, with regard to Forster's ideology-sovereignty dichotomy, it could reflect both axes. While opinions of the European Parliament may reflect an ideological positioning similar to that characterising opinions towards the composition of national parliaments, it can certainly be imagined that such opinions may also be reflecting concerns about national sovereignty, as the European Parliament represents a supranational element of co-operation (as Juliet Lodge writes, the strengthening of the EP's powers has `in particular been interpreted as symptomatic of a qualitative leap from intergovernmental co-operation (...) to supranational integration presaging a federal future'; Lodge 1996: 12). Then again, adherence to national sovereignty may also be portrayed as an ideology or a value--as may, in fact, utilitarianism. This uncertainty risks conflating `ideological euroscepticism' (or `value-based euroscepticism') with a universal. 86 Turn-out at Denmark's six referenda on the EC/EU has not been lower than 75 percent. See EU Information Centre of the Danish Parliament: www.euo.dk. 90 4.4.1 Contribution Up until the early 1990s the thesis of the Silent Revolution was indeed a prominent approach to understanding euroscepticism. However, not only does this thesis concur with scholars like Janssen, Gabel and Anderson in finding that empirical events have since seriously questioned central aspects of the theory (Janssen 1991; Anderson 1998b; Gabel 1998c), it more importantly finds that it is not appropriate to pursue with a direct operationalisation of this thesis as a single constitutive type of euroscepticism. Indeed, even if post-materialist values and cognitive skills can be confirmed as being relevant explanations of contemporary public EU attitudes, they are not well conceptualised as being euroscepticism. This is because they predominantly attempt to map socio-demographic characteristics of the individual eurosceptic person, as opposed to mapping a common denominator of what euroscepticism is. Indeed, one of the theory's two legs--the idea that cognitive mobilisation leads to more supportive EU attitudes--has to do with citizens' level of education and/or their interest in politics, which represent a level of analysis that it is beyond the scope of this study to engage. Nevertheless, these comments are voiced with regard to the overall thesis of the Silent Revolution. Certainly, its inherent assumption that `ideology', as a vaguely defined, broad heading, plays a central role with regard to EU support and scepticism (for instance Forster 2000, 2002) can certainly not be readily ignored--given that care is taken to narrow down what is meant by ideology. Below this thesis argues that several of the central variables of the theory of the Silent Revolution in fact continue to be employed in contemporary studies of euroscepticism, albeit in different guises. Moreover, newer approaches to understanding euroscepticism offer precisely this more specific touch on ideology-- understood as value-based, contestable political orientations. Indeed, while the democratic deficit thesis (Section 4.7) rests on the assumption that citizens treasure modern European values about the form of government, understandings of euroscepticism as contestation around the left-right axis of politics (Section 4.8) also departs from a largely ideological frame of reference. Thus, although the thesis does not pursue a direct operationalisation of the theory of the Silent Revolution, its legacy will be clear in several of the below sections. 4.5 National integrity A different approach to understanding euroscepticism maintains the centrality of `national integrity' (Moravcsik 1998, 2002; Weiler 1999; Smith 2001, 2005). Integrity is here understood as `the state or quality of being entire or complete', and national integrity to refer to public perceptions of the `unimpairedness' of the nation-state. In terms of the EU's impact on national integrity, this is of course 91 most visible through the pooling of national sovereignty. However, the sections below demonstrate that debates on sovereignty-based euroscepticism are often coalesced with the issue of national identity. As Liesbet Hooghe and Gary Marks suggest, the strongest territorial identities today remain national, wherefore it is reasonable to expect that such identities constrain preferences concerning European integration (Hooghe and Marks 2005: 423). Sovereignty issues reflect one of the main political science questions evoked by European integration (Rosamond 2000: 1), namely the future role and centrality of the nation-state. The EU involves the rendering of national sovereignty in certain areas to supranational institutions as well as some harmonisation of national standards, and as such it is not unreasonable to expect that the integration process may activate public concerns about national integrity. Increasingly, scholars are pointing to the likely prominence of identity concerns with regard to support for the EU87 (Forster 2000; Haesly 2001; McLaren 2002, 2005; Medrano 2003; Hooghe and Marks 2004, 2005; Kritzinger 2003). Rather than being the function of economic utility, or political ideology, this approach suggests that euroscepticism represents reluctance to increasing (certain) competencies of the EU and thereby potentially weakening national integrity. Importantly, not all forms of co-operation in the EU meet with this concern, and, as the thesis returns to discuss, what constitutes a `sensitive' national issue varies from member state to member state.88 More broadly, the national integrity thesis portrays a battle between support for two different types of European Union: the intergovernmental version and the supranational version (cf. Siune et al. 1992: 78; Worre 1995: 249-251; Medrano 2003: 3; the thesis returns to this discussion in the conclusion). Nationalism researcher Anthony D. Smith, who has recently become more involved with euroscepticism, relies on the dictionary to define a `sceptic' as someone `who inclines to disbelieve', and adds with specific regard to euroscepticism that in general parlance it signifies `an emotional detachment from particular claims, doctrines and ideals' (Smith 2005: 1). The more complex task then becomes to define the nature of the ideals and doctrines that are disbelieved. Smith first turns to support for the EU, where he identifies two sets of overlapping doctrines as forming the centre of support (ibid). The first sees the EU as an irreversible and potentially beneficial economic and political project; the second holds that `Europe' constitutes an underlying cultural identity which is being realised, and accepts that loyalty to 87 Lauren McLaren notes that the importance of identity concerns on euroscepticism have long been neglected (McLaren 2005: 5). 88 This argument, moreover, is double-edged as some scholars perceive the EU as in fact protecting national identities in a globalizing world--the thesis returns to this point below). 92 and identification with Europe will eventually come to subsume national loyalties and identities, even if it will never eradicate them (ibid: 1-2). While the former set represents a more utilitarian or value-based support, and the latter a more identity-based support, Smith defines opposition to both doctrines as a largely emotive phenomenon. It is thus not dissatisfaction with few economic benefits from integration that defines opposition to the former doctrine, but an emotive stance. Public euroscepticism, to Smith, thus seconds economic concerns and illustrates instead a lack of `fiť between the Union and a person's identitive attachments. Smith, however, does not operationalise the sets; neither does he engage in a discussion of the possibility of varying degrees of scepticism. Liesbet Hooghe and Gary Marks go one step further in this regard. They conclude their paper `Does Identity or Economic Rationality Drive Public Opinion on European Integration' with the finding that citizens `do indeed take into account the economic consequences of European integration, but conceptions of group membership appear to be more powerfuľ (Hooghe and Marks 2004: 415). They thereby suggest that while both theories are strong, concerns about national identity override. This would imply that the EU cannot leave it up to good performance or delivery of results to secure public support. While Jürgen Habermas, as we saw in Section 4.3, construed rights as crucial for the emergence and unity of a European people, attempts to secure EU support by reference to its utility represents, to the national integrity thesis, in the words of Joseph Weiler: `a view which is concerned with the degradation of the political process, of image trumping substance, of deliberative governance being replaced by a commodification of the political process, of consumer replacing the citizen, of a Saatchi and Saatchi European citizenship. To conceptualize European citizenship around needs (even needs as important as employment) and rights is an end-of-millennium version of bread-and-circus politics' (Weiler 1999: 335). The balancing between utility and national integrity is also implicit in the Liberal Intergovernmentalist approach to integration theory, albeit with a more deterministic thrust: while utility is the very raisonďtre for co-operation in the EU, nation-states will continue as sovereign and independent actors, as they constitute the most practical and natural locus for politics. The study of public attitudes towards the EU has generally been a low priority to Liberal Intergovernmentalism, even following the events of 1992, where the Danish `no' to the Maastricht Treaty threatened to block the continuation of the integration process. However, an implicit, fundamental assumption of the approach does involve the lack of preference--elite, but also public-- for `federalisť developments. When its leading proponent Andrew Moravcsik, with reference to the ratification process of the Maastricht Treaty, labels Denmark `antifederalisť (Moravcsik 1998: 381), and sharing `the British governmenťs extreme opposition to nearly all the proposals considered in the Maastricht negotiations, 93 not least EMU' (ibid: 419), he is arguably underlining the importance of the manifestation of public euroscepticism at the Danish referendum. A vast majority in the then Danish government and parliament had prior to the referendum endorsed the Maastricht Treaty, and it was supported almost unanimously by the Danish business sector, newspapers and interest organisations (for instance Svensson 2002: 741). The governmenťs aspiration with the subsequent opt-outs was thus to alleviate popular, not elite, euroscepticism. In itself, the fact that non-state actors (such as `the people') are able to influence the European integration process (and bother to, cf. Moravcsik 2006) is not readily contained by Liberal Intergovernmentalism. 89 However, the demonstration of public scepticism towards federalist developments--a frequent interpretation of the source of the problems encountered by the Maastricht Treaty--fits in well with its assumptions about the centrality of the domestic arena. The argument here is that patterns of international co-operation will adapt to stubborn national patterns of identity, rather than the other way round. A rather different strand within European integration theory, Federalism, also contributes--albeit more reluctantly--to an understanding of sovereignty-based EU attitudes. In fact, the early federalist movement of the late 1940s can hardly be said to have been indifferent to public opinion (Spinelli 1967; Ray 2007). Rather, it began with a bottom up conception of mass opinion as a driving force behind integration (ibid). At least, the federalist assumption that the finalité politique of the EU is federal (Burgess 2004: 25) implies that there will be no major, unsurpassable public opposition to that development. Citizens, it is assumed, are either already inclined to prioritising the general welfare of the EU to that of their own locality, or they will become inclined to doing so almost automatically during the course of integration. Federalists, in short, assume the existence, or smooth development, of enough loyalty among citizens of previously independent nation-states to allow them to constitute a political whole (ibid). However, while early federalists could interpret modest advances towards true federalist unification as a product of mass attitudes being too incoherent and vague (cf. Ray 2007: 2), recent developments and understandings of EU attitudes constitute more serious pointers as to the presence of strong and persistent opposition to federalist goals of a United States of Europe. Federalist imprints on the Maastricht Treaty, for instance, are likely to have constituted a major reason behind its failed public ratification in Denmark in 1992. Moreover, Altiero Spinelli wrote (already in 1967) that `federalists 89 It would, however, be mistaken to hold that liberal intergovernmentalists rule out that politicians are motivated by considerations about voter positioning, and thus that they do not take public opinion into consideration. 94 had deceived themselves with regard to the speed with which the resistance of national traditions could be overcome' (Spinelli 1967: 328). Some scholars took heed of the anti-federalist public sentiments that were expressed in the aftermath of the Danish referendum. Conor O'Brien, for instance, emotively argued that: `if the federalists were able to press on towards their USE [United States of Europe] through a serious attempt at dismantling the national sovereignties, they would find that they had done the opposite of what they intended. They would have awakened the sleeping giant of nationalism (...). A serious attempt to impose a USE would have been a bonanza to the Le Pens of every nation ­ including the most dangerous ones: those of Germany' (O'Brien 1997: 84). In this view, strong public scepticism towards the (at least formal) weakening of national sovereignty is perceived to constitute a powerful strand of euroscepticism. To Joseph Weiler, `the Danish referendum was a watershed in the battle for democracy in the Union, a well aimed `shot over the bows' which changed the process of IGCs forever ­ and for the better' (Weiler 1998). Weiler underlines his appreciation of the active role of public opinion by introducing Maastricht and its aftermath as the most important constitutional moment in the history of the European Union--not in terms of its content, but because of `the public reaction, frequently and deliciously hostile, and the public debate which followeď (Weiler 1999: 3-4). Of importance to this section, this debate about democratising the EU, which the thesis returns to discuss in Section 4.7 below, sparked contention as to the development of, and the possible nature of, a European identity, or demos. The influential `no demos thesis' takes as its point of departure the interplay between democracy and identity, holding, in line with O'Brien, that the absence of a strong EU identity puts serious limits on federalist ideas and ideals. It was perhaps most famously discussed by Joseph Weiler, who dubbed it the embarrassing and sad position of the German Constitutional Court with regard to its `Maastricht Decision' (Weiler 1995c: 2). The `no demos thesis' holds that the people of a polity have a subjective, socio-psychological component, which is rooted in objective, organic conditions (ibid: 4)--a `thick' conception that precludes the existence of a European demos (or Volk). Critically, to this way of thinking, the Volk is the basis for the modern democratic state: majority rule is only legitimate within a demos, which makes demos a condition of democracy (ibid: 5). A somewhat `thinner' version of the argument is put forward by Christopher Lord, who holds that `Democracy requires at least enough of an identity for people to accept that they should deliberate and vote as a group, yet, there is no guarantee that an adequate sense of identity will be available at the exact moment that a political system comes under pressure to democratise its decision-making' (Lord 2000: 5; see also Christiansen 1997; Portillo 95 1998: 15). Roger Eatwell adds that redistributive policies in particular require some form of affective identity (Eatwell 1997: 261). Lord, nevertheless, opens the door for a possible future European identity being willed into existence (also Beetham and Lord 1998: 39), which would then make true democratisation possible. This perspective divides scholars. Anthony D. Smith (1991), for instance, does not preclude that nation-state and national identity are constructed concepts. However, he sees them as `frozen' political identities that are non-transferable to the European level. The nation-state is the carrier of a `special loyalty' that has been able to contain and arbitrate more diffuse identities in a manner which has made democracy possible (ibid). Today's heads of states and governments, at least, do not possess the mechanisms and means with which to construct a common identity that were available to their colleagues two centuries ago. The essence of Smith's stance is not that co-operation in the EU is a bad idea--it is not a `hard scepticism' towards integration. However, it stresses the existence of a critical tension between the nation-state and supranational integration (see also Hansen 2001), the central question being whether or not this tension is perceived as manageable or insurmountable, and on which side priorities lie. It follows that the `no demos thesis' in both its strong, organic and thinner versions has important implications for the prospects of democratising the European Union. Making the EU fully democratic is simply not possible--and more importantly, desirable--as (long as) there is no real European demos. Theoretically, at least, the democratic deficit thesis (see Section 4.7) and this perspective on the national integrity thesis are likely to be mutually exclusive (see Moravcsik 2006 for a different view on this argument). This seeming contradiction inspires much of Weiler's writings of the 1990s. His own position, departing from the recognition that the EU has already gone far enough to require democratic guarantees, favours at the same time democratisation and the safeguarding of the individual demoi of the member states (Weiler 1995b, c). Weiler's conclusion is an encouragement to rethink the concept of demos, allowing for the simultaneous belonging of individuals to multiple demoi, based on different subjective factors of identification (Weiler 1995c, 1999). This conception of multiple demoi is attractive to Weiler, as it: `preserves the boundaries, preserves the Self and preserves the Other (...), [attempting] to educate the I to reach to that Other (...). To this demos, one cardinal value is precisely that there will not be a drive towards, or an acceptance of, an overarching organic-cultural national identity displacing those of the Member States. Nationals of the Member States are European citizens, not the other way arounď (Weiler 1999: 344-348). 96 National identity, to Weiler, may well in part be based on ethno-cultural sources, but this does not impede nationals from sharing a European identity based on civic values (ibid). In other words, a `thick' national identity and a `thin' European identity serve different purposes and are not in competition with one another. The euroscepticism that was voiced through the Maastricht ratification debacles could thus in Weiler's view represent the perception that the new Treaty broke with the `cardinal value' (cf. above quote) of peaceful coexistence with national identities--perhaps through the Maastricht Treaty's launch of a European citizenship. Taking a somewhat more pragmatic approach to multiple demoi, William Wallace argues that multiple loyalties and identities follow logically from the multi-level government of the EU, and compares this to citizens who may `define themselves as Bavarian in some contexts, German in others, and European in perhaps the broadest political contexť (Wallace 1997: 44-45). Importantly, while European elites--or the cognitively mobilized--move easily between such levels, European publics are less convinced of the benefits brought about by integration. On the contrary, they are more susceptible to its psychological costs: `national identities shaken, the link between citizens and accountable governments weakened by the displacement of policymaking into the transgovernmental maze of the Community process' (ibid). Wallace identifies an underlying crisis of national identity in most European member states, expressed through different forms of popular disillusionment with established institutions and elites (ibid). Importantly, he does not see any indication of the transfer of loyalties to European institutions, for which the early idealists of European integration had hoped. In this account of multiple loyalties, the average citizen feels a widening gap to policy-makers and a psychological loss from what is perceived to be a shaken foundation for the nation-state. This is thought to give rise to a largely identity contingent euroscepticism, or what Wallace calls expressions of popular disillusionment (Wallace 1997; see also McLaren 2002). As will be explained below, this type of disillusionment is conducive to an exclusivist, xenophobic variant of sovereignty-based euroscepticism, as citizens `retrench' in their nation-states. In fact, the perception of threats posed by other cultures has even been argued to constitute a `great parť of the reason why `people are hostile toward the European projecť (McLaren 2002: 551, but see also McLaren 2005). 4.5.1 Exclusionary populism Literature on a possible link between certain variants of euroscepticism and hostility towards immigrants is scarce, despite the individual prominence of each of the two issues on the public and political agendas in many European countries (de Vreese 2004 is an important exception). Researchers 97 have disagreed with regard to whether or not a connection exists. Jrgen Goul Andersen answers his own rhetorical question, `Is there a relationship between the election result [the Danish general election of November 2001, which was interpreted as a victory for anti-immigration sentiments] and more EU-sceptic attitudes ­ do nationalism, xenophobia, and hostility towards Europe form a kind of syndrome (...)?' with a definitive and unambiguous `no' (Andersen 2001: 8). There is, Andersen argues, `little connection' between antiimmigration attitudes and anti-European attitudes (Andersen 2001: 8). Nevertheless, the situation is unlikely to be unambiguous. Already with regard to the Maastricht Treaty, a Danish study of the referendum suggested that questions about refugees and immigrants could be brought into a future referendum as a powerful and secure `no' argument (Siune et al. 1992: 95). At least, the EU and contemporary patterns of non-Western immigration have in common that they may both contribute to a blurring of the historical constellation of the unity of state, society, economy and culture. Thus, nonWestern immigration and European integration may be highly related issues in the public perception (Hjlund 2000: 59; see also Togeby 1998: 190), as both point to the growing multicultural, mutiethnical and multireligious composition of modern European societies (ibid). Importantly, this perception need not be based on rational arguments; as it may well be in the psychological area that we find citizens' sense of a loss of national integrity (Lyck 1992: 14). Karen Siune et al. found that about 20 percent of Danish voters following the `yes' vote on the Edinburgh Agreement in 1993 believed it meant that Denmark would experience a large-scale increase in the number of refugees and immigrants (Siune et al. 1994a: 145). Claes de Vreese, moreover, confirms a strong statistical relationship between his measures of anti-immigrant sentiments and low support for the EU (de Vreese 2004). Globalisation pressures, by virtue of their border transgressing nature, can be thought to challenge the factual sovereignty of nation-states. It has been argued that one value-related consequence of today's decreasing autonomy of the nation-state is that increased loyalty is formed around the national, creating a form of boundary around what is local (for instance Gundelach 2002: 48). Zygmunt Bauman has argued that although physical distances in space are becoming blurred, the near-faraway dichotomy is as prevalent as ever. What is close and nearby is one's domestic or local setting; a place that is certain and where relatively few efforts are required to fit in. What is `faraway' is associated with risks, an illdefinable location requiring effort, uncertainties and craft. It does not seem farfetched to apply it to the `country-EU' level (Bauman 1998). Along similar lines, Manuel Castells argues that the `age of globalisation' is also the `age of nationalist resurgence' (Castells 1997: 27). In Europe, the perceived threats of globalisation are enhanced by the growing multiethnicity and multiculturalism of European societies. These materialise in the expanding 98 powers of the EU, as people affirm their identity both against a supranational state and against cultural diversification (Castells 2000: 357-8). As Hooghe and Marks put is, `European integration reinforces multiculturalism' (Hooghe and Marks 2005: 423). Nationalism, Castells argues, is thus the concomitant development of European integration, as citizens `retrench in the countries' (Castells 2000: 359). In this sense, some segments of euroscepticism may, contrary to Andersen's claim, be somehow related to hostility towards immigration. This finding also supports the analyses of Lauren McLaren, who argues that antipathy toward other cultures stemming from nationalistic attachments constitutes an obviously important, but often overlooked, factor determining levels of support for, or hostility toward, the process of European integration (McLaren 2002: 551; McLaren, however, moderates her conclusions somewhat in her article from 2005, where she asserts the overarching prominence of economic concerns). Also accepting this premise, Gerard Delanty argues that the almost parallel increase in prominence over the past few decades of the EU and non-Western immigration into Europe has provoked a `nationalism of resistance', and inspired a situation where nationality is increasingly being defined in opposition to immigrants and foreign influences (Delanty 1996: 1.1; see also Hjlund 2000: 57-9). Ole Borre identifies the EU and immigration as constituting the two major components of political alienation in Denmark in the 1990s (Borre 2000: 285). Both are `critical issues' in that they count among the most distrust generating issues in modern Danish politics (ibid: 290). Borre even perceives the EU and immigration as the two most powerful issues at elections, often revealing large gaps between the position of the main political parties and mass opinion (ibid: 290-2). The issues represent a `new type of concern', which transgresses the traditional left-right dimension as well as economic divides. As Borre writes, although the EU and immigration also have economic implications, concerns over their impact on the national budget are probably of minor importance compared with nationalistic feelings (Borre ibid: 286-294). Threats to the continued coherence of the nation-state may be perceived to stem from a multitude of sources, and the one of interest to the thesis remains of course the EU. It is a debate among scholars whether or not the pooling of sovereignty in the EU institutions in fact constitutes an encroachment or a strengthening of the factual sovereignty of a member state (cf. for instance Beck and Giddens;90 Lidegaard;91 also Hedetoft 1994: 14-20). However, in line with Wallace's distinction between the elite and the average citizen in Section 4.5, it is hardly surprising if, to the latter, the 90 Beck, Ulrich and Anthony Giddens (2005): `Nationalism has now become the enemy of Europe's nations', The Guardian, 4th October 2005. 99 possibility of being voted down in the Council of Ministers and subsequently having to implement a directive that is counter to national decisions, or of having to change from the national currency to the Euro, may breed some concern about national integrity. It should be mentioned that, whether or not infused with xenophobic undertones, sovereignty-based euroscepticism is sometimes construed as a national `exceptionalism', which is hypothesised to make participation in the EU `unnaturaľ (see for instance Denman 1996; Lawler 1997; stergaard 2000, Forster 2002; Lang-Jensen 2003; Spiering 2005: 127-149). Using this rhetoric, Kenneth Minogue, for instance, argues with reference to the United Kingdom that `We have become so accustomed to it [the European Union] as to forget how remarkable, indeed, how unnatural, it is to be able to induce established states to hand over their power to an institution largely composed of foreigners' (Minogue 1996: 261). A somewhat different thrust to exclusivist euroscepticism stems from the past centuries' history of European wars, which has inspired a deep mistrust in some parts of public opinion of a strong Germany (see Lang-Jensen 2003 for an account of Danish anti-German sentiments at the time of Denmark's accession to the EU). Danish autonomy, for instance, has sometimes been seen as synonymous with the ability to avoid German influence (Mouritzen 1996: 67, see also stergaard 2000). Forster points out that anti-German sentiment also inspired early British euroscepticism (Forster 2002: 13-14). 4.5.2 Contribution On the basis of the above discussions, the thesis expects national integrity concerns to constitute a prominent type of euroscepticism: European integration indeed touches on issues that can be seen as sensitive to national integrity, wherefore sentiments of disapproval towards furthering integration in certain areas seem to logically constitute an independent type of scepticism. Moreover, we may ponder the extent to which studies confirming the dominance of other types of concern would have had to alter their findings, had they only included sovereignty indicators as independent variables. Matthew Gabeľs confirmation of the utility thesis (Section 4.3), and Robert Rohrschneider's affirmation that the democratic deficit is central in accounting for low EU support (Section 4.7) importantly overlook or disregard this possible factor. Already at this stage, however, a cautionary remark about the likelihood of fully operationalising this type of euroscepticism should be voiced. The possibility of xenophobic euroscepticism is yet 91 Bo Lidegaard: Speech at the `Europe Conference ­ 23rd March 2007: 50 Years of the Treaty of Rome' at Arken, Denmark. 100 relatively unperceived of, and clearly politically incorrect.92 Even if the data-set relied upon by the thesis did contain indicators adequate for measuring such attitudes (a hypothetical example could be: `Are you planning to vote no to the Constitution because you think this will mean fewer immigrants on the streets'), it is because of the sensitivity of the issue possible that interviewees deliberately express what they believe is the politically correct answer, even if this is not their actual opinion. However, neither the Standard Eurobarometer polls, nor any other data material that the thesis is aware of, allows for an adequate direct investigation of xenophobic euroscepticism. This may, in other words, be a case of the theory expecting xenophobic attitudes to contribute to constituting a variant of euroscepticism and of the data not being able to examine the hypothesis fully. The thesis operationalises national integrity concerns as sovereignty-based euroscepticism, based on the assumption that identity concerns are reflected herein. To citizens concerned about losing national identity, supranational co-operation is, in other words, likely to be seen as largely synonymous with a pressure on identity. In addition, and more pragmatically, this focus for operationalisation steers the thesis free of the host of problems that are involved with setting common denominators for when something in the EU invokes a concern about national identity across the case countries. Indeed, in some countries, fears may surround the prospect of the national language being used less and less, while in other countries it may be the idea to fly the EU flag on public buildings that especially evokes identity concerns. There is likely to be considerably less ambiguity surrounding what constitutes an encroachment on sovereignty than what constitutes an encroachment on identity--and, to reiterate, as the former is assumed to be capturing the latter, 93 the thesis labels this type of euroscepticism sovereignty-based. 4.6 The national context The experience of direct elections to the European Parliament, and the growing number of referenda on EU issues, inspired a new approach to explaining euroscepticism, which is mostly concerned with the analysis of voting behaviour. It gained ground in the early 1990s, and takes its point of departure in the national political context of the EU's member states. While it does seek to account for `no' votes at referenda and eurosceptic votes at EP elections, and thus euroscepticism, the approach has little to do 92 The thesis could add that in the United Kingdom, commentators have acknowledged that euroscepticism may have a xenophobic dimension, but that this should not be seen as a racist, mindless attitude, but merely as the reflection of Britain's geographical status as an island, separate from the continent (Sharpe 1996: 306-7). 101 with the EU itself. The nation-state, in other words, is regarded as the central player in influencing public EU attitudes. The assumption that the EU is evaluated based on an assessment of national factors (for instance Franklin et al. 1994) implies that citizens do not form their attitudes about the EU independently of the national level (for instance Kritzinger 2003). Euroscepticism therefore reflects negative attitudes to national developments, including the incumbent government, rather than sceptical attitudes toward the EU. The negative portrayal of the EU in the national media, or by politicians engaging in a `blame game', where European integration is evoked as the excuse for unpopular domestic reforms, combined with inadequate knowledge of the EU, is thought to contribute to preventing most citizens from forming independent attitudes towards the Union. Information about the EU usually has to be obtained through the national media, and this `national lens' rules out the possibility of independent attitudes. The national context thesis usually finds its empirical backing in the outcomes of elections to the European Parliament or referenda, such as the recent French referendum on the Constitutional Treaty, where it was claimed that citizens were protesting against the unpopular Chirac-Raffarin leadership and were being guided by concerns about France's domestic situation (see for instance Ivaldi 2006). At elections to the European Parliament, scholars have underlined the length of stay in power of the incumbent government and the general national political and socio-economic climate as the decisive motivations for voters (Schmitt 2005). This `subdimension' of the national context thesis, dealing exclusively with voting behaviour, is often referred to as the protest--or government punishment-- thesis. The EU is a tool for citizens to demonstrate their dissatisfaction with the incumbent national government. Member state governments have traditionally been pro-EU and advocated a `yes' in EU elections: To date, there has never been one EU referendum, where the national government has been opposed to ratification, and at EP elections, candidates from the governing parties have campaigned for the continuation of a generally pro-EU line in the European Parliament. The protest thesis holds that citizens who are dissatisfied with the government will use such elections to punish and weaken the leaders in office (for instance by voting `no' or for a eurosceptic party). Importantly, such votes are assumed to have little or nothing to do with the EU; the EU is so to say a casualty of the domestic setting. Schneider and Weitsman moderate this position somewhat, acknowledging the existence of a great dilemma for the electorate between the possible utility of, say, a new treaty on the one side, and the value of influencing domestic politics on the other (Schneider and Weitsman 1996: 582). As the title 93 Indeed, while sovereignty concerns may exist independently of identity concerns, it is unlikely that identity concerns exist independently of sovereignty concerns. 102 of their article suggests `The Punishment Trap. Integration Referendums as Popularity Contests', domestic issues do play a role at EU referenda, however, Schneider and Weitsman are careful to point out that-- rational--voters always care, to some extent, about the international dimension (ibid: 605). Given the frequent success of minor anti-EU parties at EP elections, lines of thought on the prevalence of the national context appear relevant to a broader study of voter motivations. They need, however, to be cautioned by the finding that European voters often seem to position themselves in a position less positive to the EU than the position of their party. In Denmark, Ole Borre and Jrgen Goul Andersen found in 1997 that two-thirds of voters declared themselves more eurosceptic than their party (Borre and Andersen 1997). This suggests that when EU elections run counter to the recommendation of parties and governments, we may `merely' be witnessing a difference in stance, and not a protest. Prominent advocates of the protest thesis include Mark Franklin et al. (for instance 1994), who set out with three possible explanations to popular scepticism, as this was demonstrated at the three referenda on the Maastricht Treaty in 1992 (Denmark, Ireland, France). The first possible explanation was that the popular sentiment on Europe had never been as positive as had been generally expected by EU leaders and academics (the permissive consensus thesis). In other words, the narrow Maastricht referenda in Denmark and in France should not have surprised us, as they merely portrayed a longterm, general pattern. The second possible explanation was that voters became eurosceptic during a campaign because they, after listening to the arguments, did not like what pro-EU supporters were telling. This type of scepticism would reflect high public awareness and suggest that integration had reached a limit in the eyes of most citizens. The authors are (perhaps too) quick to point out that neither of these possible explanations is likely to hold empirically. Danes, for instance, were contrary to what has sometimes been claimed, not well informed about the EU and the proposed new Treaty (Franklin et al. 1994). Franklin et al. thus lean toward their third explanation, which holds that the results of the Maastricht referenda are best understood in terms of domestic party competition (ibid: 456-8). Voters made up their minds based on short-term national concern, rather than on long-term European considerations (ibid: 470). The thesis suggests that Franklin et al.'s rejection of especially their first potential explanation may be unfounded as euroscepticism seems to be a much more rooted and long-term phenomenon than is often assumed. At least, it is unlikely to be an explanation that holds for both the Danish `no' and the strong showing of euroscepticism in France. The thesis returns to this hypothesis following the empirical analysis; however, casual findings from examining two widely used Eurobarometer 103 indicators--the question of support for membership and the question of perception of benefit-- tentatively support the thesis' hypothesis as Danish scepticism on both indicators was at its lowest in precisely 1992 (an overview of these Eurobarometer figures is presented in Chapter 7). Palle Svensson voiced a thorough critique of the Franklin thesis in 2002, arguing with reference to a study of five Danish referenda on the EU that citizens in fact develop consistent values on salient issues. He holds that independent EU attitudes of Danish voters, more so that the popularity of the government, determined referendum behaviour (Svensson 2002; see Garry, Marsh and Sinnott 2005 for a similar argument with regard to the Irish referenda on the Nice Treaty in 2001 and 2002). The relative reluctance of supporters of the Danish Socialist People's Party to follow their party line and shift to a `yes' between the referenda on the Maastricht Treaty (1992) and the Edinburgh Agreement (1993) is for instance used as an argument in favour of citizens holding firm and independent anti-EU beliefs (Svensson 2002: 744). The national context thesis is not exclusively concerned with voting behaviour. The centrality of the nation-state in accounting for everyday eurosceptic attitudes is also assumed by for example Andrew Moravcsik as an integral part of his understanding of the European integration process. Indeed, when Moravcsik argues that it is too costly for rational citizens to substantively engage in deliberation about the EU (see Section 4.3 on utilitarian euroscepticism), he does so based on the assumption of the overriding centrality of the national context (Moravcsik 2006). What matters to citizens takes place at the national level, which allocates Europe a second order status. An important assumption underpinning the national context thesis is thus the perception of EU issues as `second order issues'. Evidence in favour of the second order thesis is oriented towards EU elections--especially elections to the European Parliament have been characterised as being second order (see Reif and Schmitt 1980; Schmitt 2005). This basically means that they `are perceived to be less important, because there is less at stake' (Schmitt 2005: 651). Second order elections are known for their relatively low rates of participation, low degree of politicisation and success of non-government, and generally smaller, parties (ibid: 651-652). Hermann Schmitt, an early proponent of the second order hypothesis, finds evidence that also the latest EP election of June 2005 was second order (ibid). As a consequence of the perceived second order status of many issues dealt with in the EU, as well as in consideration of the lack of knowledge and information about the EU that is thought to be characteristic of many citizens, Christopher Anderson has suggested that citizens use proxies when evaluating the European Union. These proxies relate to the performance of the national political system (Anderson 1998b), as citizens rely on something they know well (or better; i.e. proxies) to make 104 judgements regarding something they know less well (see also McLaren 2002: 553). This study argues that the national context thesis carries with it at least three problematic assumptions, which invite for some caution in allocating it a too prominent role, even at EU elections. First, it assumes that citizens' EU attitudes are not formed in an abstract way, but instead reflect specific characteristics of the surroundings in which people live (for instance Kritzinger 2003). Though advocates of the approach convincingly point to the predominance of domestic issues at elections to the European Parliament (for instance Schmitt 2005), an accompanying finding--their low salience compared with general elections or referenda--may foremost suggest that the predominance of domestic issues at these elections has more to do with a lack of interest in the institution of the Parliament than with attitudes towards the EU in general. EU referenda, with their generally higher turnout and prominence, as well as the critical nature of the EU issue in many countries (Borre 2000), at least indicate that EP elections should not uncritically be interpreted as indicators of general attitudes towards the EU--just as low attendance at a regional election need not be suggestive of low turnout at a general election in a given country. Second, as mentioned above, the national context thesis assumes that inadequate public knowledge or awareness of the EU refrains or hinders citizens from forming independent attitudes towards the Union. The European integration process is thought to be too complicated and too distant. Information and knowledge, according to this line of thought, are necessary conditions for being able to form independent views. While it would rightly be nave to assume that the `average citizen' has the time or interest to study details about the functioning and policies of the EU, it seems, however, unclear why this circumstance should refrain citizens from passing a judgement about the Union that is independent of the performance of their country, or the stance of their party or government representatives. Moreover, while citizens may have clearer ideas about domestic policies than EU policies, we may ponder the extent to which policies are actually central in constituting EU attitudes. In other words, when considering a supranational entity like the EU, not all citizens may base their attitudes on a rather complex subject matter like policies. Indeed, there are other issues that could inform attitudes, for instance the rejection of supranational co-operation or the generic impression that co-operation is going too fast. Whether or not a subjective impression like the latter will prevail after the citizen has studied in more depth the `true' nature of the EU, does not remove the fact that the initial impression was based upon some evaluation of the merits of European integration, and not necessarily modulated by the nation-state or a mere plebiscite on the performance of national governments (Garry, Marsh and Sinnott 2005). With regard to referenda, it moreover seems 105 problematic to completely disentangle the EU dimension from protest votes: if voters decide to protest against an incumbent government in a(ny) referendum, this most likely follows a previous evaluation that it is possible to live without the realisation of the particular referendum topic. If, for instance, a referendum was held on the re-introduction of the death penalty, few would probably find it appropriate to let their vote unilaterally depend on a need to protest against a poor mid-term performance of their government. Finally, the claim that co-operation in the EU is about issues of too low salience for citizens to take an interest is questionable as a general thesis. Again, low turnout at elections to the European Parliament is no good yardstick--we may reiterate the above argument that it is not possible to infer about turnout at a general election from turnout at a regional or local election. Importantly, referenda about EU issues attract a high turnout in many member states, and generally invoke such strong emotional responses in the electorate that it seems unfounded to relegate the EU a second order standing. Indeed, although careful scrutiny of the EU's actual competencies may result in the claim that the EU only has real influence on minor issues of low salience (Moravcsik 2006), citizens (and this part of the argument is in fact in line with Moravcsik) are likely not to have engaged in such careful scrutiny, and instead associate the presence of the EU flag, or media stories about EU regulation on for instance consumer goods, with a strong `Brussels influence'. British citizens, at least, are known for combining little knowledge about the EU (and little discussion about the EU at general elections) with a strong perception that the Union is (far too) powerful. Commission President José Barroso's nomination in 2006 as `Britain's most powerful man' by a BBC poll94 is hardly indicating that the EU is perceived to be occupied with only low salience issues. In Denmark, moreover, even though the EU takes up almost no space in general elections, it is, as discussed above, still a `critical issue' (Borre 2000), characterised by a strong difference between elite-mass opinions and a strong emotional touch.95 Thus, while it remains a puzzle why the EU as an issue is often disentangled from most general elections,96 this finding--given the prominence of EU referenda--is not adequately explained by assuming a second order status for the EU. The national context approach also encompasses as a possible explanation for euroscepticism general 94 BBC 4 Radio poll, quoted on BBC's website: http://news.bbc.co.uk/2/hi/uk_news/4574968.stm. The BBC story is quick to point out the fact that Barroso is not democratically elected. 95 It could in passing be mentioned that the thesis has found its walkthrough of readers' letters in Danish newspapers to confirm the emotional status of the EU issue (cf. the archive of EU related reader's letters developed by the Danish European Movement). 96 Gilles Ivaldi speaks of the European issue in France being `constantly present in voters' minds and a structuring dimension of the attitudinal and ideological space in French politics', while nevertheless not really emerging `as a salient issue per se in first-order elections' (Ivaldi 2006: 65). 106 sentiments of public disenchantment, or lack of trust, with regard to politics--a connection which has been forcefully argued by Siune et al., who have stated with reference to the Danish rejection of the Maastricht Treaty in 1992 that `the higher the degree of lack of trust, the higher the no-percentage' (Siune et al. 1992: 108, own translation from Danish) and, with specific regard to the hypothesis of democraticallybased euroscepticism, that `[l]ack of trust played a much larger role [in explaining the outcome of the referendum] than democratic concerns' (ibid). To disenchanted citizens, the rejection of an EU treaty in a referendum is not just a protest against a particular government, it reflects an underlying, deeper expression of societal frustration. Euroscepticism, as Szczerbiak and Taggart point out and as referred to in the introduction to the thesis, in this sense becomes a bell-weather for understanding the tenor of politics in a climate of sceptical or distrusting public sentiments (Szczerbiak and Taggart 2003: 22). 4.6.1 Contribution The national context thesis understands euroscepticism from the perspective of the nation-state. It excludes that the EU plays a substantive role in `no' votes at EU referenda and in determining the choice of candidates at elections to the European Parliament. Instead, the EU has second order status in the minds of most citizens, who base their votes on an evaluation of the national context or a general disenchantment with politics. Insisting on a stringent definition of euroscepticism (see Chapter 3), the thesis does not pursue with an empirical analysis of this approach--it is in other words not included as a constitutive type of euroscepticism. This is not to posit that EU elections are never characterised by protest votes, votes purely determined by the national context, or broader societal disenchantment. Voting analysis could be elaborated to more adequately examine these points. Instead, the decision not to include the approach in the thesis' conceptualisation is informed by the three-fold recognition that (i) protest votes are hardly indicative of everyday public attitudes towards the EU; that (ii) their main concern is not the EU as such: it would in other words not be euroscepticism that was being examined, but some other form of political dissatisfaction; and that (iii) this dissatisfaction would suggest a general malaise of European societies rather than the thesis' definition of euroscepticism. More specifically, the national context thesis, as its name so eloquently suggests, is about the national context. It is only about the EU in as far as it involves a down-playing of the importance of EU co-operation or the equation by disenchanted citizens of the EU with politics more generally. (Deliberate) ignorance about the EU, perceptions about the low salience of EU issues, and universal disenchantment fall outside the thesis' definition of euroscepticism. Moreover, a related point, adequate empirical illustration of the national context thesis is not possible without resort to voting analysis. As 107 elections invite considerations about, for instance, strategic voting and reasons for abstention, pursuing empirical analysis of this approach would confuse the aims and arguments of the thesis, which is concerned with patterns within general everyday public opinion. Protest voting is not assumed to be a major characteristic of responses to Eurobarometer polls. 4.7 The democratic deficit The idea of `dissatisfied democrats' (for instance Milner 2000: 11), or the `democratic deficiť thesis, became widely popular as a means to understanding public euroscepticism following the problems with ratifying the Maastricht Treaty in 1992 (see for instance Christiansen 1997, note 2, for an early list of contributions). It takes its point of departure in the assumption that the transfer to `Brussels' of many, increasingly important, government functions requires the EU to live up to certain democratic standards (cf. Weiler 1995c: 7). What standards are contested in the literature (for instance Weiler 1995c; Moravcsik 2002; Mény 2003b; Fllesdal and Hix 2005a; Taggart 2006: 9), perhaps in part because the discussion is difficult to distinguish from talking about what the EU really is, and is not (Dahrendorf 2003: 101).97 As Frank Decker is right to point out, and the thesis elaborates upon in the below, there is an `optimistic and integrationisť variant of the democratic deficit debate, perceiving the deficit to be a relatively minor and transient ailment; as well as a `euro-pessimistic' variant, where the democratic deficit is merely used as an argument reinforcing fundamental objections to the integration process (Decker 2002: 256-257). At the public level, manifestations of the democratic deficit are thought to be revealed through public opinion polls, electoral absenteeism at EU elections and the emergence of new forms of mobilisation and political parties (Mény 2003a: 3). It is important to note that the democratic deficit thesis did not arise in connection with the Maastricht Treaty. Juliet Lodge, for instance, traces its relevance back to political discussions prior to the introduction of direct elections to the European Parliament in 1979 and sees it as one of the motivations behind the reforms introduced by the Treaty on European Union in the in the early 1990s (Lodge 1996: Chapter 1; also Taggart 2006: 10). Yves Mény points out that the expression (the democratic deficit) was originated by British political scientist David Marquand in 1979, in criticism of the European Assembly's poor 108 representation of the member states' populations (Mény 2003a, b: 58).98 Mény also suggests that democratic scepticism only gained in prominence following the introduction of direct elections. We can thus note that attempts to rectify a perceived democratic deficit at least twice has been followed by even more powerful accusations of democratic deficiencies (aside the introduction of direct elections, the Convention method to drafting the Constitutional Treaty was also followed by new accusations of democratic shortcomings after the French and the Dutch rejections). 4.7.1 Versions of the deficit Capsule versions of the democratic deficit are proposed by a number of scholars. To one of its initial proponents, Joseph Weiler (Weiler 1995c: 7-10), it consists of several interrelated issues. One is the diminution of the relative political gravity of each individual when a policy area, such as environmental protection, comes exclusively or predominantly within Community responsibility. The inclusion of a larger population reduces the individual weight of each member. While the extent to which this is perceived as a democratic problem arguably depends on the nature of the policy area, this aspect of European co-operation is likely to be readily perceivable to EU citizens (perhaps especially in small member states). The thesis returns to the importance of this observation below. Another prominent aspect of the democratic deficit thesis is according to Weiler that the volume, complexity and timing of the Community decision-making process make national parliamentary control `more an illusion than reality' (Weiler 1995c: 8). The European Parliament is not seen as offering an effective or adequate substitution to these shortcomings, both in light of its still emerging powers and in view of the low turn-outs at elections to it. Lodge even portrays national parliaments and the European Parliament as engaged in a zero-sum game over EC legislation (Lodge 1996: 13). The result is that Union governance involves a net empowerment of the executive branch of the State (ibid: 8). Although the European Parliament has witnessed an increase in its competencies since Weiler voiced his critique in 1995, a similar conclusion can probably still be maintained today. However, it is arguably highly unlikely that the `average citizen' is knowledgeable about the processes of parliamentary control with EU policies, or the relative weight of the European Parliament in the EU system. In addition to the above two issues, Weiler lists as characteristic of the EU's democratic deficit the absence of an opportunity for the electorate to `throw the scoundrels [here unpopular EU politicians] ouť; the proportionate over-representation of smaller member states in the European Parliament as well as 97 This section seeks to isolate debates about democracy from debates on democratic legitimacy, which were discussed in Section 4.2. This does of course not preclude that pronounced democratic deficits at EU level may constitute a lack of Union legitimacy. 109 in the voting system; the compromised status of constitutional courts in countries that have such control; and the lack of transparency of the EU system (Weiler 1995c: 9-10). In itself, the claim that the European Parliamenťs powers are still `extremely weak' (Plattner 2003: 54) is perceived to constitute an important aspect of the EU's democratic deficit (ibid). Andreas Fllesdal and Simon Hix concur with most of Weiler's points in their five-point version of the `standard democratic deficiť. Indeed, this includes (i) the decrease in national parliamentary control; (ii) a too weak European Parliament (especially in comparison to governments in the Council); (iii) the absence of European elections that are truly about the EU; (iv) the fact that the EU institutionally and psychologically remains `too distanť from citizens; and, largely in consequence of these four factors, (v) a `policy drifť from citizens' ideal policy preferences (Fllesdal and Hix 2005: 4-6). Fllesdal and Hix, however, distance themselves from the argument that the EU has to live up fully to all these criteria in order not to be undemocratic. They thus go some way in agreeing with, for instance, the defence of the democratic deficit advanced by Andrew Moravcsik (Moravcsik 2002; see below). The fourth point, however, is maintained as crucial for `even the `thinnesť theories of democracy' (Fllesdal and Hix 2005: 4): For the Union to be democratically acceptable there should be electoral contest about political leadership at the European level or the basic direction of the EU policy agenda (ibid: 18). To the extent that citizens are sceptic of the democratic deficit, this is likely to be a ready-toperceive deficiency Yves Mény offers a more succinct account of the democratic deficit, portraying public critique of EU democracy as the result of an imbalance between what he sees as its two pillars: popular democracy and constitutional democracy (2003; see also Wind 2006). While the EU in Mény's perception has sought to develop the most sophisticated and elaborate forms of constitutionalism, it has not come close to being able to combine this with strong popular input (Mény 2003b; a focus which is tied to Section 4.2's discussion on legitimacy). Robert Rohrschneider is perhaps the most direct in his portrayal of the democratic deficit and its consequences for mass support for the EU (and more specifically an EU-wide government; Rohrschneider 2002). He claims to find empirical proof that `when citizens perceive that they are unrepresented, their support for the EU is reduced independent of their economic perceptions' (ibid: 463). This claim, the thesis 98 The Wikipedia entry on the 'democratic deficiť traces the term to a pamphlet published in the 1980s by Bill Newton Dunn, a British member of the European Parliament (correct as of January 2007, on www.wikipedia.org). 110 finds, has to be qualified with some reservations--most importantly the absence in the study of adequate indicators for sovereignty-based concerns (we cannot infer much about euroscepticism from the indicator of `national pride'). As Section 4.5 demonstrated, concerns about national integrity may not cease even if citizens feel perfectly represented in the Union. Dieter Fuchs also points to the criterion of utility barely being considered as a predictor in Rohrschneider's study (Fuchs 2003). However, the thesis is not engaged with testing which argument would win in a `yes-no' referendum situation. As discussed below, if citizens do indeed, as Rohrschneider claims, perceive themselves to be seriously un(der)represented by the EU's institutions, this is an expression of scepticism of the EU-ofthe-day that is relevant to comprehend if the aim is to be `close to citizens'. On the overall level, the thesis therefore concurs with Rohrschneider that fundamental dissatisfaction with EU democracy is an important component of euroscepticism. 4.7.2 Defence of the deficit Not all scholars agree that democratic deficiencies are at all critical for the EU, and thus important to understand with regard to public euroscepticism. Andrew Moravcsik, for instance, has explicitly argued against rectifying the version of the democratic deficit that Fllesdal and Hix label as standard (see above; Moravcsik 2002). This view is advanced in extension of the theory of Liberal Intergovernmentalism, which accepts as its basic premise that national governments run the EU (Moravcsik 1998). However, Moravcsik's defence of the democratic deficit thesis not only underlines his perception that the EU is not a state. It also reflects his view that democratisation of the Union would not lead to less euroscepticism. Rather the exact opposite situation is likely to hold: `democratising the EU would be expected, if it has any effect at all, to render it less popular and legitimate in the eyes of publics' (Moravcsik 2006: 233, original emphasis). This claim builds on the argument that: `increased opportunities to participate do not, as a rule, generate more intensive and informed public deliberation or greater public trust, identity and legitimacy'. `There is no reason (...) to assume that increases in opportunities to participate necessarily generate participation, deliberation, legitimacy, or popularity' (Moravcsik 2006: 219-222). Giandomenico Majone adds to the defence of the democratic deficit that since the EU can be characterised as a regulatory agency, its policy-making is not in need of being democratic in the usual meaning of the term (for instance Majone 1996). In fact, democratisation of the European Union could contribute to politicise questions that are not suitable to partisan logic (Bartolini 2001: 16). As Stefano Bartolini paraphrases Majone's perspective, taking decisions according to a partisan and majoritarian logic in the EU would constitute a negative development, as integration should be protected from its consensus seeking political mechanisms, which hinder the reaching of the most efficient and optimal 111 solutions (ibid). 4.7.3 Scepticism about the deficit Yet other authors point to critical flaws in the very premises of the democratic deficit debate (for instance Christiansen 1997). In Thomas Christiansen's view, flaws occur on three accounts. First, the deficit debates include a tendency to recommend models of parliamentary democracy to the EU, which are intimately linked to the emergence of the nation-state and which cannot as such be transferred without modification to the entirely different environment of European integration (see also Mény 2003a; Dahrendorf 2003: 107). Second, recommendations to `overcome' the democratic deficit often neglect that the EU is not necessarily less democratic than the average European nation-state (see Moravcsik 2002 on this point). Finally, the debates are flawed since they tend to ignore that democratising the Union would have serious consequences for its efficiency, which may contribute (even more) to underline its popularity (Christiansen 1997: 2). In a somewhat blunter vein, Eric Hobsbawn has noted that: `The development of the EU was only possible in the absence of democracy. The idea that there is a `democratic deficiť is absurd as it assumes that democracy was part of the original equation. The EU was never intended to be democratic' (cited in Leonard 1998: 13). Evaluating the overall contribution of the detailed accounts of the democratic deficit to an understanding of euroscepticism, it could be added that they are likely to be more apt in accounting for organised or elite scepticism than they are capable of capturing public scepticism--simply because greater interest in, and more knowledge about, the EU political system (an inherent assumption of most literature on the alleged deficit) can be more readily assumed at the level of for instance political parties. Concerns about the democratic standards of the EU institutions have for instance been invoked as an explanation for the equivocal stance towards the EU that several Green parties in the EU member states have traditionally assumed. As Hooghe, Marks and Wilson quote Elisabeth Bomberg for arguing, `Greens in Europe (...) face a strategic paradox: the incentives to work through the EU are great, yet how can they work through institutions that inherently violate green principles?' (in Hooghe, Marks and Wilson 2002: 19-20). Opinion polls surveying people's correct answers to a number of basic questions confirm that detailed knowledge about the EU's institutional set-up is an unlikely characteristic of the greater public. 4.7.4 Contribution As the above section demonstrates, studies of the EU's `democratic deficiť have informed much of the literature on euroscepticism following the Danish rejection of the Maastricht Treaty in 1992, the 112 underlying assumption being that the EU's failure to live up to citizens' ideals about democratic political systems cause them to become eurosceptic (i.e. typically to find membership itself a `bad thing'). Based on the above readings, the thesis finds it reasonable to expect that citizens are eurosceptic precisely because of these democratic shortcomings--or in other words that democratic concerns in themselves can be construed as an independent type of euroscepticism. Although there is little clarity as to what exactly constitutes the democratic deficit of the EU, it is important to note that this interpretation of euroscepticism actually constitutes a rather dramatic ontological shift in the general understanding of public opinion towards the European integration process. To put it crudely, from conceiving public opinion as largely passive, uninterested and compliant--and therefore arguably as rather unimportant--it almost overnight became widely assumed that citizens were knowledgeable about their country's number of seats in the European Parliament, the decision-making rules in the Council, the number of policy areas decided by co-decision, subsidiarity, and other aspects of co-operation amenable to democratisation. As opinion poll after opinion poll, however, reveal that large numbers of EU citizens do not give the correct answer to presumably simple knowledge indicators (such as the question of how many countries are members of the Union), this expectation about the broad public opinion may be aiming too high. This is not to say that large numbers of citizens are not acutely aware of democratic shortcomings in the EU, let alone that vast number of citizens should not be eurosceptic on democratic grounds--indeed, the thesis finds it relevant to conceptualise this as an independent type of euroscepticism. However, it does suggest the utility of adopting a rather broad understanding of this variant of scepticism in order to reflect that the critique of the democratic functioning of the Union may in fact be rather abstract. Similar to the discussion in Section 4.2 on the relationship between euroscepticism and EU legitimacy, it is crucial to note that the thesis here is concerned with perceptions of a deficient EU democracy, and not with whether or not there in fact are such deficits. 4.8 Political contestation The classic axis of political contestation in West European countries has for more than a century been the left-right division. Scholars have debated whether or not this axis also characterises public attitudes towards the EU, and increasingly, as the thesis returns to below, public attitudes within the EU. Some disagree. Classic International Relations approaches to European integration (including Liberal Intergovernmentalism), for instance, downplayed the importance of ideological positions, including those reflecting the left/right axis of politics, in shaping contestation about integration. 113 Instead, geopolitical positioning has been seen as paramount (see Marks and Steenbergen 2004). As Gary Marks and Marco Steenbergen paraphrase these approaches: pursuit of the national interest, rather than a domestic ideological positioning, determines whether national leaders support or oppose further European integration (Marks and Steenbergen 2004: 5). Marks and Steenbergen underline that Liberal Intergovernmentalists do maintain that domestic conflict is important in explaining why some governments support and others oppose integration. But that conflict is about the gains and losses from trade, and is independent from the left/right dimension that structures much domestic contestation (Marks and Steenbergen 2004: 5). The scarcity of studies on political contestation at the EU level has led scholars to lament that while Europe has become more politicised, `the current wave of scholarship on Europe persists in more or less ignoring politics in the treatment of European integration' (Sudbury and Laffan 2006: 2, quoting Mair 2006: 344). There is a need, as Radaelli succinctly puts it, to `bring politics back in' (Radaelli 2006). Models of political contestation in the EU, however, do exist. Simon Hix and Christopher Lorďs model from 1997 (Hix and Lord 1997), for instance, perceive of an orthogonal relationship between the left/right dimension and a more/less integration (or sovereignty-based) dimension.99 They thus see the two dimensions as independent of one another. In the words of Hooghe, Marks and Wilson 2002, this involves a conception of the issue of European integration as engaging national sovereignty questions and mobilising territorial groups, which compete on where authority should be located, on the one hand, and on the other hand left-right contestation involving the allocation of values along functional interests (Hooghe, Marks and Wilson 2002: 6). In a fusion of the two dimensions, sometimes referred to as the `regulation modeľ (Marks and Steenbergen 2004), the political left is seen as being supportive of the EU's current development towards more regulation, while the right increasingly opposes it. Contestation in the EU is here recognised to be about more than a `yes-no' positioning or the exclusive reference to the national interest. Another, more restrictive, variant of this model of political contestation in the EU, developed by Hooghe, Marks and Wilson, conceives left-right location as related only to a sub-set of European issues. In the view of these scholars, left-right contestation shapes positioning only on European policies that are concerned with redistribution and regulating capitalism: `Hence the center-left supports European integration in cohesion policy, social policy, unemployment policy, environmental regulation, and upgrading the European Parliament, while the right supports market integration but opposes European re-regulation' (Hooghe, Marks and Wilson 2002: 6). 99 Note that this model largely concurs with Anthony Forster's typology of euroscepticism as consisting of an ideological and a sovereignty axis (Forster 2000, discussed in Section 4.4). 114 In a later study, Hix, together with Andreas Fllesdal, suggests that, because of a number of democratic deficiencies, there may be a policy drift in the integration process, which means that the EU adopts policies that are not supported by a majority of citizens in many or even most member states (this argument was hinted at in Section 4.7 above; Fllesdal and Hix 2005: 6). Of importance with regard to this approach to euroscepticism, they interpret the drift as skewing EU politics to the right of domestic policy status quos, which breeds social democratic euroscepticism (ibid). In a different version of this argument, Hooghe, Marks and Wilson paraphrase Fritz Scharpf in noting that there is an in-built asymmetry in the European Treaties favouring market-deepening to market-correcting policies: `Social democrats expect that this asymmetry can be countered to achieve moderate reform. For the radical left, the European Union is beyond repair' (Hooghe, Marks and Wilson 2002: 8). Perhaps spurred by the persistence of this imbalance, the authors perceive of increasing political contestation at the EU level, in particular in the European Parliament, where MEPs to a greater extent than previously vote with their European party colleagues and against their national party leaderships when these two sets of interests are in conflict (ibid: 19). Also the Council of Ministers is believed to be witnessing increased policy contestation, with the left-right axis determining government positions alongside the traditional pro-/anti-Europe axis (ibid). Fllesdal and Hix conclude that there is real potential for left-right battles over the EU policy agenda (ibid: 20). This suggests that opinions about the EU may increasingly be determined by political orientation. It does indeed seem fair to suggest that the left/right dimension, for its part, has come to play an increasingly central role in shaping political contestation about EU integration over the past decades (Hix and Lord 1997; Hix 1999; see Marks and Steenbergen 2004: 6), and empirical studies provide some evidence that the political right, highly supportive of the integration process of the early decades, seems to be turning more eurosceptical, while the political left experiences the reverse development (Marks and Hooghe 1999; Pelinka 2005: 207-225). It may even be possible to hypothesise that the importance of left-right contestation for our understanding of euroscepticism becomes more acute as the EU takes onboard an increasing number of policy areas, in this sense enhancing its influence on the everyday lives of Europeans. To Hooghe and Marks, European integration has in most countries even become a left-leaning project because it holds out the prospect of continental-wide regulation (Hooghe and Marks 2005: 425; also Hooghe, Marks and Wilson 2002: 9). Generally, in these lines of thought, public positioning towards the EU is connected to perceptions of which issues are important, which are not, and what political colour, they are given. 115 Such perceptions are not exclusively determined by the left-right axis. Indeed, contestation may reflect a host of cleavages, including religion and ethnicity, of which the left-right axis is but the most prominent. The introduction of new members to the Union may also bring about new lines of contestation. Southern Enlargement to Greece, Spain and Portugal in the 1980s, for instance, may have enhanced the Protestant-Catholic schism, while Eastern Enlargement in 2004 may be intensifying `moraľ contestation--opposing, for instance, adherents of traditional family values and adherents of more post-modern values. However, it cannot be assumed that cleavages at the EU level merely reproduce existing national cleavages. Indeed, the EU may inspire entirely new lines of contestation. Simon Hix writes that the result of the `collapse' of the permissive consensus is `a more complex pattern of social and political interaction. (...) New transnational socioeconomic and value-based divisions are shaping people's attitudes towards the EU' (Hix 2005a: 173). Stefano Bartolini asks whether the process of European integration, perhaps in conjunction with the fall of Communism and the accelerating process of transformation from industrial economies to service economies, may even represent a new `critical juncture' in Europe's history, with profound influence on existing political cleavages (Bartolini 2001: 23). He is himself rather sceptical that this is the case (see especially p. 43); however, the question remains central to political scientists interested in the nature of contestation in the EU. Indeed, it should not surprise us that lines of contestation change over time. It is generally recognised in the literature on the formation and structure of political cleavages (for instance Bartolini 2001) that in the formative years of a political system, cleavages essentially emerge around what should be the competencies of the new centre. This process entails a somewhat consolidation and legitimation of its boundaries. Following the formation of the political system, cleavages tend to emerge on the basis of the resistance felt by, for instance, minority groups, towards the values and identity of the emergent system. Only in more mature political systems does a third type of cleavage emerge, which is based on a process of interest differentiation along functional lines (typically conflicts about market and redistributive mechanisms, thus bearing on the left-right division; Bartolini 2001: 23-24). It is this sketch of the possible chronology of political cleavage formation that may lead us to ponder whether contestation in the European Union may be developing towards an increasing abandonment of the more-less integration dimension, highlighted by for instance Simon Hix and Christopher Lord (1997), in favour of more `normaľ political contestation. It is, to elaborate, not particularly innovative to suggest that citizens' left-right positioning may have some bearing on their 116 attitudes towards certain aspects of the European integration process--and that, with regard to certain other aspects, the EU issue may cut across existing cleavages, including the left-right divide. What is intriguing, however, is whether (and to what extent) the EU's long-term existence may be bringing about a weakening of the more-less integration divide--leaving euro-contestation a matter of political value preferences. Such a development would entail that the EU's oft-acclaimed `sui generis' image cede to partisan, value-based discussions of the merits of specific policies and their implementation. This type of euroscepticism is clearly horizontal (see Section 4.2), as it reflects a type of contestation that accepts the EU as a normal political battleground. Research published in the 2000s speaks affirmatively of an emerging `political turn' in the EU (Sudbury and Laffan 2006: 9), and efforts to identify the nature of contemporary transnational cleavages have intensified (ibid; Hooghe, Marks and Wilson 2002; Marks and Steenbergen 2004). Imogen Sudbury and Brigid Laffan (2006) suggest an issue of contestation peculiar to the EU to be that of redistribution, which sets apart net-receivers and net-contributors. One may add to the list: geo-political divisions, institutional divisions (degree of fit between `Brussels' and one's national institutional set-up), and even mobile/non-mobile divisions (those who are able to benefit (or have an interest in benefiting) from negative integration and those who are not). Immigration, energy policy and protecting the environment may constitute other issues, which both nationally and Europe-wide inspires new cleavages. Hooghe et al. (2002) have explicitly conducted empirical tests to examine the strength of a potentially emerging `new politics dimension' at the EU level. This dimension is largely non-economic and occurs over issues such as sustainable development and individual liberty (Hooghe, Marks and Wilson 2002: 11; see also Sudbury and Laffan 2006: 5). While we may recall from Section 4.4 that Ronald Inglehart already in the 1970s divided supporters and opponents of European integration along largely similar lines (the materialist/post-materialist distinction), the contemporary dimension identified by Hooghe et al. conforms more broadly to a divide between those ascribing to green/alternative/libertarian values (the `GAL pole') and those who maintain traditional/authoritarian/nationalist values (the `TAN pole'; Hooghe, Marks and Wilson 2002: 11). Based on an expert survey of political party positioning, the authors are able to confirm the strength of the GAL/TAN dimension of contestation, on which basis they rephrase Clausewitz' famous statement to suggest that `European politics is domestic politics by other means' (Hooghe, Marks and Wilson 2002: 23). It is clearly beyond the scope of the thesis to examine the nature of all possible new lines of contestation in great detail. Nevertheless, it is important to retain two things from this discussion, 117 namely the growing recognition that (i) political orientations shape euroscepticism; and (ii) that the classic left-right cleavage is coexisting within a complex cleavage structure (Sudbury and Laffan 2006: 5). 4.8.1 Contribution In addition to perceptions of economic shortcomings, fraud and inefficiency, the fear of losing national sovereignty and concerns about a democratic deficit, the thesis finds it reasonable to expect that a variant of euroscepticism consists of a political, value-based positioning with regard to the issues debated within the realm of co-operation in the European Union. The above section sought to identify some of the cleavages that this type of euroscepticism could reflect. At issue in particular is the traditional left-right axis, but the literature on contestation also suggests that new and EU-peculiar cleavages may be emerging. It is, to repeat, beyond this thesis to examine all these dynamics in greater detail. From the above accounts, however, the thesis finds there is reason to expect that many of these contemporary lines of political contestation in the EU are in fact captured by debates about a `Social Europe'--a growing focus of recent years.100 Indeed, the thesis proposes that identifying indicators measuring citizens' degree of dissatisfaction with the EU's current social credentials is a helpful operationalisation of Hooghe, Marks and Wilson's green/alternative/libertarian dimension of contestation: indeed, while the social-liberal divide in the EU clearly touches these cleavages, it arguably goes even further in capturing critical public EU attitudes, as it also incorporates the more noneconomic issues of social justice and solidarity--in short the wish for a rather closely-knit homogenous Union that is opposed to more market-oriented integration. While it was arguably rather uncontroversial between the 1950s and 1980s to maintain that the EU was predominantly about market integration (as its then name also testified, focus was on the economic community), there is today growing disagreement as to what weight should be placed on market integration. Although there presumably continues to be little disagreement as to the centrality and importance of a close internal market for something like the EU, desire for a more social Europe has grown in intensity and is arguably still largely unfulfilled (the other side of the social-liberal coin, 100 In an inductive manner, the thesis examined the open-ended (spontaneous) Eurobarometer question--`What are all the reasons why you are opposed to the European Constitution?'--for potential powerful variants of value-based, political euroscepticism that might have been overlooked in the existent literature on euro-political contestation. The only other reply category that could constitute a variant of this type of scepticism read 'No reference to the Christian roots of Europe'-- alluding to a religious or moral variant of value-based euroscepticism. However, this concern was not at all voiced in the first survey the question was posed (Eurobarometer 62), and by only six percent in the EU25 as a whole in the subsequent survey (EB63, 2005). As for the three case countries, the figure from Denmark in 2005 was two percent, while in France it was three percent and in the United Kingdom four percent--figures that according to the schema for classifying absolute 118 namely perceptions that the EU has come to embrace too much social redistribution, is expected to be captured by economic euroscepticism; see Section 4.3 above). To the thesis, social euroscepticism takes as its point of departure the broad perception that the EU is paying too much attention to achieving the benefits of market integration at the expense of social welfare. It is important to note that this is likely to be a contemporary type of euroscepticism, emerging in response to the changing nature of EU co-operation. It is unlikely to have been adequately perceived of in the early decades of predominately economic integration. 4.9 Constitutive types: a summary In the above, the thesis grouped existing studies on public euroscepticism according to their dominant views about its nature. The studies could be divided along six independent headings, which largely reflected a chronological evolution of scholarly focus. In the early decades of co-operation utility-based explanations dominated alongside the hypothesis of post-materialist support inherent within the thesis of the Silent Revolution. In the aftermath of the Maastricht ratification debacle, theories about protest voting, the democratic deficit and the concern about national sovereignty became dominant, while more recently, ideas that euroscepticism resembles domestic political contestation are gaining ground. On this basis, the thesis posits that euroscepticism comes in five main ideal types. In existing literature, in other words, we find grounds for expecting that past and contemporary euroscepticism may be economic, performance-based, sovereignty-based, democratic and/or social, as simplistically illustrated by the table below. The fact that the types are theoretically-derived ideal types implies that it is not possible to rule out some degree of empirical overlap between them (an issue that the next chapter discusses in more detail). Table 6: Five theoretical expectations as to the nature of public euroscepticism Theoretical expectation Main objection Utilitarian I: Economic Lacking economic benefits Utilitarian II: Performance Inefficient set-up Sovereignty-based Supranational integration Democratic Undemocratic structures Political (here: Social) Non-accordance with political orientation (here: Too liberal) levels of euroscepticism adopted by the thesis (see Chapter 5) are considered to represent the absence of euroscepticism (i.e. below five percent). 119 As the above sections demonstrate, the five types are not evenly represented by existent theoretical approaches. Based purely on the number of contributions, utilitarian (especially the economic variant) and sovereignty-based euroscepticism appear the weightiest accounts, whilst markedly less has been written about the social-liberal divide. However, this may merely underline the existing lack of an indepth combined theoretical and empirical study of euroscepticism, as well as the dynamic nature of the object of study, which is constantly being re-examined in light of political events. In the absence of an existing focused study of these `euroscepticisms', the thesis wishes to make no prior assumption about the relative empirical weight or prevalence of the five types. Importantly, the types need not be mutually exclusive. A person, if asked, may criticise the EU for the lack of economic benefit she perceives to derive from it while also not supporting what is perceived to be encroachments by the EU on national sovereignty. Analyses of voting behaviour at EU referenda; of the campaigns preceding a vote; or focus groups, may, of course, reveal certain types of arguments to be more decisive than others in influencing a `yes/no' vote. It should be clear that voting behaviour is not the object of study of the present thesis. The deduction of five types of euroscepticism concludes the theoretical endeavour of the thesis. More specifically, it concludes the drafting of the second level in Gary Goertz concept design, which was introduced and discussed in Chapter 2. We may recall that the relationship among the five types of euroscepticism is one of family resemblance and that their relationship to the basic level is one of ontology. Chapter 5 now completes the three-level concept structure by developing operational indicators that stand in for each of the secondary level types. 120 Part Three: Analysing euroscepticism Part Two deduced five constitutive types of euroscepticism from the literature. The aim of Part Three is to assess and analyse the empirical relevance of this conceptualisation of euroscepticism and to discuss the resulting new understanding of the phenomenon. The part first elaborates on its operationalisation. In Chapter 5 it is specified how indicators standing in for the various types of euroscepticism are selected, how they are treated, and what information the thesis is looking for in the analysis. The issue of falsifiability is also discussed, and tests are run to ensure the validity and strength of the indicators. The analysis then proceeds in Chapter 6 by applying the corroborated conceptualisation on three EU member states, Denmark, France and the United Kingdom, in order to demonstrate different and perhaps even contradictory patterns of euroscepticism. The chapter concludes by accounting for comparative findings, before Chapter 7 assesses what can be learned from the identified trends. 5 Operationalisation: the indicator level This chapter completes the three-level concept structure introduced in Chapter 2. At issue is the third, or indicator, level in Gary Goertz's design, which involves the operationalisation of the types of euroscepticism into statements that can be subjected to empirical analysis. In the thesis, this is done through the identification of substitutable indicators from public opinion polls and the running of statistical tests to check their validity. Five issues are discussed below. Section 5.1 discusses the scope, constraints and shortcomings of the chosen data-set (Eurobarometer polls). Section 5.2 turns to the selection, handling and validity of indicators (opinion poll questions) of euroscepticism. Section 5.3 considers the issue of falsifying, or disproving, the conceptualisation; while Section 5.4 summarises the discussion on operationalisation, and draws up a checklist of the broad observations to be made by the analysis. A table provides a schematic overview of the operationalisation of euroscepticism. Finally, Section 5.5 tests the validity of the chosen indicators. 121 5.1 The data-set: advantages and shortcomings In accordance with the method of structured, focused comparison, the same data-set and the same questions are used across the cases. The data-set used in the analysis is the European Commission's Standard Eurobarometer polls. Eurobarometer polls have existed since the early 1970s.101 They are conducted on behalf of the Directorate-General for Press and Communication, and most are publicly available on the Commission's website.102 The polls have included Greece since autumn 1980; Portugal and Spain since autumn 1985; the former German Democratic Republic was integrated in the polls since autumn 1990; Austria, Finland and Sweden since spring 1995; and the Czech Republic, Cyprus, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia since autumn 2004. The thesis' time perspective does not include polls following the entry of Romania and Bulgaria in 2007. An identical set of questions is asked of representative samples of the population aged fifteen years and over in each member state. Eurobarometer surveys come in four main types.103 The Standard Eurobarometer poll, by far the most influential of the types and the one employed in the thesis, have been published biannually since spring 1974, with a number of questions recurring in most polls. Typically, a standard survey covers over 40 questions with a commentary. Special Eurobarometer polls have since the early 1970s surveyed public opinion towards specific issues, such as Economic and Monetary Union, enlargements, and xenophobia. Flash Eurobarometer are ad hoc polls aimed at providing relatively quick, focused surveys of topical issues, such as post-referendum surveys or attitudes toward the European Convention, perhaps from specific target groups. Finally, Eurobarometer has begun to produce qualitative studies, for instance through focus groups, with the aim of investigating the motivations of specific social groups towards a given subject in a more in-depth way. To facilitate analyses, Eurobarometer provides a Eurobarometer Trends volume, offering a ten year overview of regularly posed questions, as well as an accessible interactive search tool on its website, allowing the visitor to search for longitudinal and comparative trends among 43 more or less recurrent questions. On a more technical note, the regular sample size (in the sense of completed interviews) of a Standard Eurobarometer poll is 1000 respondents per country, except for the United Kingdom (1300 respondents, of which 300 are from Northern Ireland) and Luxembourg (5-600 respondents; cf. 101 1974 is the year of the first poll available on the European Commission's website (Eurobarometer 1). There were, however, roughly comparable predecessors to this poll (see for instance Schmitt 2003). 102 http://www.europa.eu.int/comm/public_opinion/index_en.htm. Since 2004, they have been conducted by TNS Opinion and Social; prior to that they were conducted by national institutes associated with INRA EUROPE (see respective polls). 103 The introduction below builds on information available at the Commission's Eurobarometer website, op. cit. 122 Eurobarometer data service guide).104 The sampling design is a multi-stage, random (probability) one. The sampling is thus based on a random selection of sampling points after stratification by the distribution of the national, resident population in terms of metropolitan, urban and rural areas (ibid). Primary sampling units are selected from the administrative regions in every country, after which a cluster of addresses is selected from each sampled unit. Addresses are chosen systematically using standard random route procedures, beginning with an initial address selected at random (and then taking, for instance, every tenth street). In each household, a respondent is selected by a random procedure, such as the first birthday method. Up to two recalls are made to obtain an interview with the selected respondent. The reason for using random sampling is that it reduces the risk of selecting a sample `that is seriously biased in some way, thus leading to inaccurate inferences about the population'. Indeed, in random samples, `everyone has the same chance of inclusion' (Agresti and Finlay 1997: 20). Interviews are conducted face-to-face in people's homes and in the appropriate national language. For each country a comparison between the sample and the universe is carried out (data is derived from Eurostat or national statistics offices). A national weighting procedure is carried out one the basis of this description of the universe.105 Over time, the size of a Standard Eurobarometer poll has grown; the number of questions posed has simply increased. As will also be (all too) clear from Chapter 6, some Eurobarometer questions are asked in every poll, while others are asked randomly or only once. The nature of the questions asked by Eurobarometer has also, in certain respects, evolved. In an attempt to reflect the evolution of European integration, certain questions have been dropped and others added to reflect if not specific events, then at least the official concerns in Brussels at a given time.106 Eurobarometer polls offer a unique and extensive data-set for comparative and longitudinal studies of EU public opinion, and have been widely used by scholars. Hermann Schmitt writes that with the possible exception of American National Election Studies (ANES), `there is not a single survey data source that has been used or cited in a greater number of scholarly publications than the Eurobarometers' (Schmitt 2003: 245). He adds that `without any doubť the Eurobarometers have become the prime research tool in the comparative study of public opinion at the domestic level (ibid). By means of the Eurobarometers, data on public support for European integration is more comprehensive than for any other international body (Hooghe and Marks 2004), and in the endeavour to establish a spatial and temporal 104 http://www.gesis.org/EN/data_service/eurobarometer/guide/index.htm. 105 This information is based on the Eurobarometer data service guide as well as technical notes included in the Standard Eurobarometer polls. 106 NIRA Research Output 2002 Vol. 15 (1), see: http://nira.go.jp/publ/routp/v15n01.html. 123 conceptualisation of euroscepticism, there is no competitor to the Standard Eurobarometer polls.107 Usage of the cross-national polls, however, is not unproblematic. Researchers for instance have to pay attention to potential problems with regard to data quality, varying national contexts and (unevenly) missing data. It seems fair to say that the data quality of the Standard Eurobarometer could be improved on substantive as well as methodological grounds. As pointed out on more than one occasion by this thesis, as regards the far majority of Eurobarometer question formulations, they change often and are rarely one-dimensional (i.e. they tap a multitude of concerns that are difficult to decipher. Schmitt moreover notes that the kind of questions asked is often not suited to the fact that the public at large are not experts on the policy process. `[R]eading through Eurobarometer questionnaires', he writes, `one sometimes wishes that they would concentrate somewhat more on public preferences towards policy ends rather than means' Schmitt 2003: 248). The replacement of one word in a survey question may make it unusable or at least unreliable in longitudinal analysis. Instead, research-based question formulations would be a great improvement for analysts, allowing them to draw more elaborate and precise conclusions about European public opinion. There is also scope for methodological improvement of the Eurobarometer polls. Pure probability sampling as employed by for instance the European Social Survey (ESS)--as opposed to Eurobarometer's multi-stage random sampling design outlined above--as well as larger samples and more efforts to secure higher interview completion rates (including a higher number of recalls to get through on the telephone) would result in better data quality and higher representativeness--although effects on actual survey findings remain unclear (Schmitt 2003: 248). Moreover, the risk of bias resulting from the fact that the Eurobarometer surveys are sponsored by the European Commission deserves mentioning. Potentially, the Commission could have a political and strategic interest in bestowing the Union with an aura of having the backing of citizens. It was for instance pointed out in November 2006 by EUobserver, the largest online daily news site focused on the European Union (www.euobserver.com), that the Commission delayed the publishing of a Flash Eurobarometer survey on energy policy because its findings were thought to be too pessimistic in the midst of attempts to move the EU forwards during the Period of Reflection (the survey found that a majority of EU citizens were in favour of keeping decisions on energy policy at the national-level, despite this policy area having been declared the new hub for co-operation). However, the thesis' 107 Other cross-national surveys of European attitudes do exist, for instance the European Social Survey or the European Values Survey, however, given their lack of particular focus on attitudes towards the EU they are to the thesis' purpose 124 reliance on multiple indicators, and its interest in relative scepticism levels (scepticism in relation to the EU-average, see below), are expected to reduce the negative implications of a potential bias. Even if a particular survey question was framed in manner expected to favour `friendly' responses, it is still to be expected that the Danes, the French and the British would position themselves in a consistent manner vis--vis each other, and that the examination of the question in connection with other indicators would nuance the picture. When dealing with cross-national surveys it is moreover important to be sensitive to the possibility that the same wording of questions may have different meanings in different countries. In this connection, it is important for the thesis' purpose to distinguish between actual ambiguity inherent within a question formulation, which could potentially threaten the validity of measurement (see Adcock and Collier 2001: 534; and the discussion in Section 2.3.2 above), and mere differences between countries that are in fact interesting to bring to light in order to understand the particular variant of euroscepticism in a member state. An example of the former could be the Eurobarometer question: `In general, are you for or against efforts being made to unify Western Europe?', which was a classic in the first decades of the surveys (according to the thesis' estimates, it was with two or three exceptions posed continuously up until Eurobarometer 28 (autumn 1987), where after it has up to today only been posed four times). This question may to West Germans have brought connotations about the rifts caused by the Second World War and thus the possibility of a historical (and geographical) unification with East Germany. To Danes, however (who as we shall see below are wary of ceding sovereignty to the EU), it may have given ideas about a full-fledged federal union. It is unclear from the question wording itself what is being implied, and the measure thus appears problematic to use as an indicator of euroscepticism. Notwithstanding this kind of ambiguity, it is certainly clear that questions surveying attitudes towards, for example, social issues will meet with differing nationality contingent environments. In countries with existing high levels of social protection and equality, proposals for more redistribution at the European level may be met with more resistance than in countries with a low level of social guarantees. Though it is important that the researcher is aware of differing national contexts when employing sensitive Eurobarometer questions, uncovering national variation with regard to (in this case) the desire for social redistribution at the EU level is only furthering an aim of the thesis. Missing data is a problem encountered by most survey designs, and not just by cross-national surveys. Whether or not it is caused by the inability to get through to a person (who may not be at home), the (despite of potential methodological superiority) not a competitor to Eurobarometer. 125 person's refusal to participate, or incomplete questionnaires, missing data is problematic as it may endanger the representativity of a poll (for instance Risbjerg Thomsen 1998). Indeed, the tendency to be away from home a lot, or to refuse to participate in polls, may be peculiar to specific groups within society, who may hold rather different opinions than the groups who are easier to get hold of, or more willing to be polled. Polling institutes usually endeavour to counter potential problems involved with missing data by applying various weighting procedures. Sren Risbjerg Thomsen, for instance, mentions demographic weighting, which allocates additional weight to underrepresented demographic groups (ibid). However, cross-national surveys face the additional problem that missing data may be of a systematically different nature from country to country. While in one country older citizens may be resentful towards stating their political opinions to unknown people there may in another country be a large population of homeless people who are very difficult to include in the sampling design. It is virtually impossible for researchers relying on secondary data-sets to rule out the possibility of flaws with regard to how representative polls are. Although it does not solve this problem, the finding that opinion polls are often reliable predictors of election outcomes (ibid)--perhaps because of similar configurations within groups easy to get hold of and groups difficult to get hold of--is, however, a soothing indicator of the general trustworthiness of many contemporary polls. The same can be said of the relatively stable positioning of EU member states with regard to individual indicators. As we shall see below, even over a 30-year period and a considerable evolution of scepticism levels, the Danes, the French and the British position themselves in a stable manner vis--vis each other. In a similar vein, a person's failure to respond to some of the questions posed by a quantitative survey may result in much wasted information, as many software programs subsequently used to analyse data-sets ignore cases for which observations are missing for at least one of the variables used (Agresti and Finlay 1997: 24). This poses particular problems for surveys engaging a relatively small-n. However, it would also affect how representative a large-n poll is, if it could be found that it was members of a certain socio-demographic group who consistently failed to respond to a certain poll question. The thesis returns to discuss how it handles missing data below. Moreover, with specific regard to the present thesis, and its ambition to propose a thorough conceptualisation of euroscepticism, Eurobarometer, it is clear, poses a finite number of questions. In theory, it is of course possible that the surveys completely fail to account for a significant part of public opinion towards the EU. While inductive studies of euroscepticism would find difficulty establishing the appropriateness of the scope of the standard Eurobarometer series, the thesis does not approach the data-set with the intention of discovering different types of euroscepticism from scratch. Instead, it 126 departs from theoretically derived expectations of what euroscepticism is, and it is these expectations which guide indicator selection. As could be expected from the vast number of different questions that have been engaged by Eurobarometer questions over the years, and as will be demonstrated below, Eurobarometer contains enough survey questions for analysing each identified type of euroscepticism. Importantly, if the analyses should reveal that only few, or none, of the types are prevalent and significant in the case countries, this would point to a flawed theoretical understanding of euroscepticism--a flaw which invariably spills over to the indicator level (the issue of falsification and possible lack of corroboration is taken up by Section 5.4). 5.2 Indicator selection and handling Multiple one-dimensional and direct indicators are required in order to achieve an adequate measure of a multifaceted concept. Indicators should in other words be directly substitutable with the relevant type of euroscepticism; logically appear to reflect one and only one type, and be additive in the sense that the more indicators standing in for one type that indicate scepticism, the more prominent is that type. Multiple indicators give the thesis more leverage and reliability in establishing the empirical prevalence of the various types of euroscepticism: if multiple questions point in the same direction it increases the likelihood of a sound interpretation of the individual indicators. The criterion of unidimensionality excludes survey questions that are otherwise frequently used by journalists, politicians and academics as indicators of euroscepticism. For instance, the questions `Do you support the European Constitution?' and `Are you in favour of the Single Currency?' are not valid indicators of any type of euroscepticism, since they do not permit the inference of particular reasons for being sceptic towards or opposing the Constitution or the Euro. These questions could in theory be referring to all or none of the types. Mixed indicators, which could fall under two or more types, are also rejected by the thesis. Appropriate indicators therefore necessarily draw a sharp analytic line among the types--one could say that the types become `ideal types' of euroscepticism. It should be noted that while necessary for clarity, this analytic distinction is likely not to be as clear cut in `reality', where some overlap between, for instance, sovereignty-based scepticism and economic scepticism, or democratic scepticism and social scepticism, may occur. The issue of the relationship among the types is resumed below. Although the ambition is not to create a composite measure for each secondary level type of euroscepticism, the use of multiple indicators nevertheless poses questions regarding intensity and weight. The thesis recognises that Eurobarometer questions are of varying intensity, that is, some questions are 127 worded in a way that attracts a higher particular response than others. It is, for instance, more likely that the poll question `the EU should give more attention to social justice' will attract confirmatory responses than the poll question `the Constitutional Treaty should be rejected because it is not social enough'. This is why the thesis analyses euroscepticism in relative terms (see below). The thesis, however, assumes that all one-dimensional, direct indicators have the same weight. Recalling from the discussion in Chapter 2 that the relationship among indicators standing in for a type is one of family resemblance, this means that, in themselves, each indicator found to be relatively significant is taken to reflect euroscepticism within a country. It should be recalled that as the aim is not to generate a measure for when a population is sceptical enough to vote `no' in an EU referendum, weights refer to the indicators' ability to reflect the presence of euroscepticism in a country. It goes without saying that a type assumes more weight if all indicators standing in for it prove to be relatively significant. A case may be characterised by more than one type, which is relatively significant. The establishment of which type(s) is dominant within a country--or secondary or tertiary--is simply done by comparing the number of sceptical indicators (in terms of percentages).108 Indicators have been identified through a study of the list of contents and list of tables of all available Standard Eurobarometer polls, i.e. from EB1 in 1974 to EB65 in 2006, offering more than a hundred different questions and sub-questions. The thesis finds that Eurobarometer typically provides between four and six direct indicators for each of the thesis' types of euroscepticism as well as for the measurement of the prevalence of hard euroscepticism. Though highly useful for the balanced evaluation of the types, it is in fact coincidental that Eurobarometer provides a rather even distribution of indicators. It was also not a given that a substantial number of Eurobarometer questions could in fact be used as indicators for the conceptualisation. Indeed, question wordings are rarely onedimensional. The exhaustive examination of polls, however, resulted in a number of indicators that the thesis finds sufficient for the analysis. To treat each type evenly, the thesis has chosen four indicators for each type (where a higher number was possible, it chose the indicators with the widest time span). The thesis numbers the indicators of the four types of euroscepticism 1 to 16. Annexed to the thesis is a list of all one-dimensional indicators and the years they were surveyed. 108 Apart from serving the thesis' aim of classification, distinguishing between `mixeď types and `strong' types will be relevant for future research into euroscepticism: through qualitative surveys, the relationship, or possible trade-off, between two types can be illuminated in order to determine, for instance, which type is likely to `win' during a referendum battle. Moreover, the relative strength of the various types has implications when it comes to targeting the right audience in communication strategies. 128 As developed in Chapter 3, the thesis distinguishes between hard and soft euroscepticism. While the latter is directly observable from the 16 Eurobarometer indicators standing in for the various types of euroscepticism--these indicators clearly show the presence of some scepticism--the latter (its intensity) is not. Indeed, the vast bulk of opinion poll questions do not allow inferences about the strength with which an opinion is held. There is, for instance, little telling from an indicator of moneybased euroscepticism whether it equals the desire to withdraw from the EU. To give an example, perceiving the EU's impact on the exports sector as `very baď could surmount to a wish to withdraw from the Union if one were depending on exports in one's professional life. However, there is no telling of this from the indicator. Hard euroscepticism is independently perceived of in the Eurobarometer polls, however, only in a manner which allows the researcher to acknowledge its prevalence, and thus not in a manner which contributes to increasing our substantive knowledge about the various grounds constituting hard euroscepticism. The thesis tackles this issue by pursuing an empirical analysis of euroscepticism as a whole. In itself, this examination of data will not allow many inferences about the intensity of scepticism (the thesis will, of course, note the absolute scepticism level under the presentation and discussion of each indicator). Subsequent to this analysis, however, the thesis turns to an examination of four indicators, numbered A-D, identified to measure the extent of hard euroscepticism. By cross-tabulating an indicator of hard euroscepticism with an indicator from each of the prevalent types (latest year polled), it can obtain an idea of what types are (today) the most and the least conducive to the wish to withdraw completely from the Union. Indeed, although those who reject the idea of the Union are likely to be sceptic along most indicators of euroscepticism (this to some extent goes with being rejecting the Union), some types of scepticism are likely to have a larger proportion of opponents than others. This examination thus qualifies the thesis to review the analysis of the prevalent types and arrive at an idea about the probable level of hard euroscepticism in the case countries as well as of where it is most and least likely to be rooted. 5.2.1 Relativity It is reasonable to assume that no member state of the EU is characterised by zero scepticism towards its membership or developments on the EU agenda. In other words, some degree of euroscepticism is expected always to be present within a country. The neutral point permitting the classification of a country as more or less eurosceptic is found, for each indicator, by means of the mathematical average for all EU-15 countries. If a country's score departs from the average (neutral point) in a negative 129 direction, it may be classified as being relatively eurosceptic toward the issue under scrutiny.109 A divergence from the average of plus/minus five percentage points is not considered to be significant,110 while a divergence of more than ten percentage points is considered to constitute a marked scepticism. This means of classification permits the thesis to identify the relative significance of the types and to distinguish between varying degrees of euroscepticism across countries. Establishing in this manner whether a country is eurosceptic within a given type has one main drawback: it does not in itself allow room for determining whether a given indicator is met with a generally high or low level of scepticism. An example may highlight this problem. The EU average rejecting joint decision-making on educational matters was 61 percent in 2001. The French population was not relatively sceptical in this regard, as 63 percent were against this idea. Danish figures were at 72 percent, thus 11 percentage points above EU average, which would classify Denmark as relatively sceptical. The issue, of course, is whether it would not also make sense to classify the French as being sceptical in this regard, as more than 60 percent of the French population is opposed to the particular idea. The example could of course also work the other way around, with, say, a 20 percent score being classified as relatively eurosceptical if the EU average is 11 percent. It is no simple task to bridge the interest in determining relative scepticism with the need to establish criteria for when a general scepticism level is high enough to be classified as eurosceptic, or too low to fall under the classification. Again, the thesis does not endeavour to establish when scepticism is `severe enough' to result in a `no' at an EU referendum, but instead to examine various types of scepticism and cross-country differences. To this end, it is significant to learn whether the Danes are relatively more sceptical than the French towards pooling sovereignty in various policy areas. Nevertheless, to ensure a reasonable employment of the term euroscepticism, the thesis will for each indicator consider the general level of scepticism in the EU, alongside the nature of the question wording, and moderate its concluding discussion about the various implications of the findings accordingly. The thesis therefore adopts the following crude measure of absolute levels of euroscepticism: 0 ­ 4 percent: no euroscepticism. 5 ­ 24 percent: low euroscepticism. 25 ­ 49 percent: medium euroscepticism. 50 ­ 74 percent: high euroscepticism; and 75 ­ 100 percent: very high 109 Likewise, if it departs from the neutral point in a positive direction it may reversely be classified as characterised by a permissive consensus. However, the study of EU positive attitudes falls outside the scope of this thesis. 110 Plus-minus five percent allows for a maximum sampling error. With typical Eurobarometer sample sizes ranging between 700 and 1200 respondents per country, and a 50/50 percentage distribution, the sample error is between 3.7 and 4.9 percent at a 99% confidence level (figures from Argyrous 2005: 298; Standard Eurobarometer polls include reference to a confidence limit of up to +/- 3.1 points). Unless otherwise indicated, the EU average consists of the number of member states at the time of polling. It should be noted that newly acceded countries may initially slightly skew the EU average in a sceptical direction as past Eurobarometer polls show newcomers tend to be particularly sceptical in the first years following 130 euroscepticism. As mentioned, these measures are applied with caution as the various indicators are worded with such varying intensity that degree comparisons are, if not entirely obsolete, then at least difficult. 5.2.2 Durability Finally, with regard to indicator selection, the long-term perspective of the thesis must be considered. In accordance with the definition of euroscepticism developed in Chapter 3, a trend should be durable (allowing for a plus/minus five percentage point variation) to be classified as eurosceptic. Tracking the development of indicators over time ensures that momentary upheavals (which could be due to the immediate shock effects of an event, or the odd high statistical deviation) are not classified as euroscepticism. The longitudinal perspective also allows the thesis to examine the dynamics of euroscepticism. Unfortunately, the data does not allow for all indicators to be traced over a long time period. Some are only posed once, and some of the types of euroscepticism were seemingly only discovered by Eurobarometer during the past decade (democratic and social euroscepticism). It is rare that one Eurobarometer poll serves more than five indicators in total, and, consequently, rare that one type is represented by more than two questions in one poll. This accounts for the thesis' main reliance on cross-tables as opposed to more sophisticated, but more restrictive methods of analysis, requiring, in particular, the same sample. Factor analysis, for instance--an otherwise appropriate means to uncovering latent structures among indicators--is not plausible for this analysis as there are simply not enough direct indicators available within one polling sample. However, each type of euroscepticism has at least two indicators, which have been posed by Eurobarometer on more than one occasion. 5.2.3 Data validity The scores of the indicators are examined and given meaning in relation to the constitutive type of euroscepticism that they are a substitute for (cf. Adcock and Collier 2001: 531). They can be seen as evidence material, which has the ability to falsify the theoretically-derived expectations as to the natures of euroscepticism (ibid: 532). The thesis concurs with Robert Adcock and David Collier (2001) in that the validity of a study has to be understood as part of `a larger process of validation that integrates "multiple sources of evidence" and requires the combination of "logical argument and empirical evidence"' (Shepard quoted by Adcock and Collier 2001: 537). Its assessment of measurement validity thus includes what the authors refer to as `content validation', i.e. the examination of the relationship between an indicator and the their accession. However, newcomers are typically so few in relative numbers that their impact on the EU average is rather small. 131 constitutive type of euroscepticism that it is a substitute for, and `convergent/discriminant validation', i.e. the statistical examination of the intra- as well as inter-type fit of the indicators. This section discusses the criteria that the thesis employs to assess the validity of the chosen indicators as measures of euroscepticism. Convergent validation is achieved if the scores produced by the multiple indicators of a constitutive type of euroscepticism are empirically associated (and thus convergent; Adcock and Collier 2001: 540542). Correlation-based approaches to validity assessment are useful to assess this convergence in quantitative studies. Accordingly, a statistical indication of the validity of the thesis' data is obtained by finding the poll where most indicators for a type of euroscepticism coincide, obtaining the relevant data-sets from the Danish Data Archive, and running, through SPSS for Windows, cross-tables, gamma tests and, when more than two indicators are present, reliability analysis for the relevant intra-type indicators (measured through Cronbach's Alpha). Cross-tabulations (or contingency tables) compare the distribution of sceptical and non-sceptical answers for two indicators. They thus represent the first step in the investigation of association (see Agresti and Finlay 1997: 251). Gamma111 is a symmetric measure of the association between two variables at ordinal level or higher. A high gamma indicates the likelihood of, by knowing the ranking along one variable, knowing the ranking on the other variable (Argyrous 2005: 100). Therefore, if gamma, for instance, is .458, we may say that knowing one variable reduces our errors in predicting the rank of the other variable by 45.8 percent (Garson112 ). In other words, gamma can be said to give a mathematical value to the observations that can be obtained from the visual inspection of a cross-table. Gamma is an appropriate measure of association when there is a high number of units and few values for each variable--that is, tables with many units in one cell (Hellevik 1997: 227).113 111 Gamma is calculated by the formula: ^ DC DC + = where C is the total number of concordant pairs of observations, and D is the total number of discordant pairs of observation (a pair of observations is concordant if the subject who is higher on one variable also is higher on the other variable; Agresti and Finlay 1997: 274-276). The value of gamma falls between -1 (negative association) and +1 (positive association). 112 Multivariate Analysis for Applied Social Science. The URL for this work in progress is http://www2.chass.ncsu.edu/garson/pa765/statnote.htm. Obtained from personal webpage of David Garson: http://www2.chass.ncsu.edu/garson/PA765/vita.htm. 113 Gamma is `the most popular ordinal measure of association', yet it is `one of many ordinal measures of association for contingency tables with ordered categories', others being Kendalľs tau-b and Spearman's correlation (Agresti and Finlay 1997: 277). While the thesis has found that Spearman's correlation is often used when the association of causal variables is being investigated, gamma is preferable to Kendalľs tau-b when a study contains many occurrences of ties in a bivariate frequency distribution (Hellevik 1997: 227). However, a potential problem involved with using gamma-values is that gamma tends to be larger when a variable is measured with few categories (ibid: 277). When setting the thesis' requirement for minimum gamma value, the thesis takes this characteristic into account. It should be noted with regard to Chapter 6's data analysis that the thesis has 132 Finally, Cronbach's Alpha, also known as the `reliability coefficienť, indicates how well a set of items measures a single latent construct. It measures the level of mean inter-correlation weighted by variances, taking into account the number of items (Garson op.cit.). There is no consensus in the literature on the use of statistics in political science as to the correct size of gamma and alpha values. The thesis accepts indicators where the cross-tabulations are sound; gamma is .600 or above114 with a significance level of p<0.01;115 and Cronbach's Alpha, when more than two indicators coincide, is above .600. 116 Tests are run both for the three case countries individually and across the entire Eurobarometer data-set (i.e. the EU average). The former runs allow the thesis to scrutinise possible differences in the types of euroscepticism characterising individual member states--indeed, it is the hypothesis of the thesis that large variation in scepticism occurs among countries--while the latter gives us a broader idea of the robustness of the association among the indicators. With regard to the fit of the indicators, this is evaluated on the basis of the overall results. The original coding of the 20 survey questions used as indicators ranges from dichotomous items (yes--no; good--bad; more power--less power) to five-point Likert scales (very much, to some extent, neutral, not very much, not at all) to gradations (`on a scale of one to seven, where seven is the most in favour, where do you position yourself?'). Moreover, the indicators are of differing levels of measurement. Some of the survey questions are nominal, allowing, for instance, respondents to state the reason why they rejected the European Constitution (not democratic enough; did not like the further loss of national sovereignty; etc.), while others build on ratio-scales to poll the extent to which a proposal is favoured. Bearing in mind that the indicators are intended to specifically capture the content of the constitutive types of euroscepticism, the thesis chooses to recode them all to form dichotomous variables, with 0 always being the `not-sceptical reply' and 1 always being the `sceptical reply'. In the same line of thought, as the interest is to single out scepticism from a `permissive consensus' around the integration process, `don't know' (and `neither-nor' in the rare cases where this is permitted as a reply category) is treated as system missing (this increase in missing values is not seen as checked values for Kendalľs tau-b, and it is my evaluation that the study would have reached virtually similar conclusions about indicator validity had the thesis used this measure (with a minimum threshold of .300, given the different calculation premises of Kendalľs tau-b). For information, Kendalľs tau-b equalled without exception Spearman's Correlation with my data. 114 This may seem a relatively high gamma value; however, given the low number of categories in the thesis' data and its ambitions, it is arguably better to set the threshold for concluding the existence of an association higher, rather than lower, than `standarď. Values are rounded off to the nearest ten. 115 Given the scattered nature of the Eurobarometer data, the thesis chooses the strict confidence interval of 99 percent in order to be as precise as possible when the data allows for tests. 116 Alpha generally increases with the number of items (Garson--op. cit.). As the thesis' number of testable items necessarily is low; and as high correlation is not an absolute requirement (see elsewhere), .600 is deemed ample. 133 a problem as the overall Eurobarometer sample is very large, typically more than 15,000 individuals for the entire poll, and more than 1000 for each individual country--and, as Chapter 3 discussed, indifference towards the EU is not defined as euroscepticism. See, however, Section 5.1 for a discussion of potential problems involved with missing values). On a more general level, this recoding of indicators naturally reflects a trade-off between parsimony and completeness (Adcock and Collier 2001: 539). The thesis is aware that information is, so to say, thrown away by the dichotomisation of, for instance, a five-point scale; however, the thesis deems that the effort to keep the full information contents of an indicator is not justified in respect of the thesis' specific research interest in singling out euroscepticism and ensuring indicator equivalence. Arguably, as the relationship among the indicators is hypothesised to be one of family resemblance (see Chapter 2), statistical correlation is not an absolute requirement as long as there is theoretical justification for each indicator. Substantive validation, in other words, is what is crucial.117 Neither can high correlation between indicators within a type always be expected--it is, for instance, unlikely that there be a uniform critique of the EU's democratic deficit. Indicators measuring this type of euroscepticism may focus on a host of different aspects of the alleged deficit, which citizens do not value evenly (some may find the absence of a directly elected Commission to be a serious democratic shortcoming, others may find this characteristic irrelevant and instead build their critique on a feeling of estrangement from Members of the European Parliament, and so forth). However, to avoid possible bias, it is deemed that the above requirements of convergent validation should be valid to the extent that they can be carried out. One further statistical requirement is applicable, and relates to the possibility of co-variance between two or more types of euroscepticism. Although the indicators are one-dimensional, it cannot be ruled out that there is no correlation across the types. They do, after all, all measure the same thing (euroscepticism). Indeed, as discussed above, hard eurosceptics are likely to some extent to be found across the types, and this makes some overlap across the types unavoidable. Moreover, the types of scepticism are theoretically derived ideal types, and it would be unrealistic to expect empirical `reality' to be as clear cut. Especially sceptical replies to broadly formulated indicators are likely to have some overlap across types. While this again underlines the importance of the substantive validity of the indicators on the one hand, and the scrutiny of the consistent behaviour of the cases across the four indicators measuring a type of scepticism on 117 Some kind of causal relationship between the ideal types of euroscepticism is of course not unlikely. For example, the relationship between democratic euroscepticism and sovereignty-based euroscepticism could be mutually exclusive: in a country where euroscepticism embodies the critique of an undemocratic EU, there might--resultantly--exist a willingness to accept more power to the European Parliament (lack of sovereignty-based euroscepticism). While it is not a main focus of the thesis to provide an account of possible interaction chains, the examination of the dynamics of euroscepticism may nevertheless serve to highlight likely relationships, which could form the basis for future research. 134 the other hand, an indication of the level of co-variance can be established by cross-tabulating an indicator from one type with an indicator from another type. This is what Adcock and Collier refer to as `discriminant validation' (2001: 540), which examines whether the association between inter-type indicators is weaker than the association between intra-type indicators. To underscore that the thesis' four types are independent of one another, or put differently: that they are acceptable groupings, intratype gamma should be higher than inter-type gamma.118 This point is linked to the validity of the theoretically-derived types of euroscepticism, which is moreover scrutinised by analysing a case country's scores on the various indicators of a type. If they reveal contradictory patterns in comparison to the EU average (if, for instance, Danes are much more sceptical than average on two indicators of a type, but much less sceptical than average on the other two indicators of that type), or there proved to be no difference among the cases, the rationale of the conceptualisation would have to be questioned. 5.3 Establishing, and dealing with, a lack of corroboration The task of measuring euroscepticism includes the possibility of finding a lack of corroboration between the theoretically derived types and the empirical analyses. Four findings of varying gravity would `falsify' the conceptualisation and/or the thesis' approach: * Statistical non-corroboration. As set out in this section, the empirical data should conform to a number of statistical requirements. Failure to obtain the necessary values renders it impossible for the thesis to fully confirm the validity of its conceptualisation of euroscepticism. If this proves to be the case, the thesis will, depending on the extent of the non-conformity, where it occurs and what it suggests, discuss where in the thesis' assumptions and approach the root of the problem could be found, as well as possible remedies. * One type is not reflected in any case country. The three main cases were in combination expected to capture the five types of euroscepticism. If one type is not relatively significant in any country, two steps can be taken before the conceptualisation as such is disregarded. First, the thesis must examine the actual (i.e. not relative) scores for the direct indicators of the absent type: do they suggest a powerful general level, which could be evenly strong across all countries? This would 118 The thesis tests for co-variance purely by means of gamma as alpha is highly sensitive to the number of items. Due to this fact (and based on the expected presence of some hard euroscepticism within all types), the introduction of items from other types could be expected to produce confounding, or at least unclear, alpha values. 135 suggest its relevance regardless of the fact that none of the member states demonstrate a relatively high level of scepticism. Second, unlike in tests of causal relationships, it would not undermine the conceptualisation if more cases were included to demonstrate the prevalence of a type of euroscepticism. If a generally strong type proved not to be relatively significant in any of the three case countries, but instead relatively significant in another of the fifteen old member states, it would merely be the case selection process, rather than the conceptualisation of euroscepticism, which was weak. However, if neither of these two steps is able to demonstrate the relevance of the type, it would have to be taken out of the conceptualisation, and the theoretical lessons from this would be discussed. * The indicators of a type of euroscepticism point in scattered directions. This would be the case if a case country is markedly eurosceptical on two indicators of a type of scepticism and much less sceptical than average on the two remaining indicators. This would point to a flaw or weakness in the conceptualisation or operationalisation of that type, regardless of possible high statistical associations. In particular, it would challenge the assumption that the relationship between the secondary level types of scepticism and the indicator level was one of substitutability (see Chapter 2). Depending on the actual results, a new look at the literature or data-set is required. The theoretical approaches from which the type was derived could be flawed, as could the scope of the data-set. * None of the types of euroscepticism are prevalent in a case country. Some euroscepticism is assumed to characterise Denmark, France and the United Kingdom. If this assumption is challenged by the data, it points to a flawed understanding of euroscepticism. An entirely new conceptualisation of the phenomenon would be required, or perhaps more appropriately, an inductive study examining survey questions not included in the thesis' analyses as well as results from inquisitive qualitative studies, and so forth. Questions to be asked would include with what alternative data euroscepticism could have been measured; what the achieved findings may be telling; and whether other types than those identified in the initial conceptualisation could account for euroscepticism. It goes for all these possible avenues of non-corroboration that what initially may appear as negative evidence can spur refinements that ultimately enhance validity (Adcock and Collier 2001: 540). In the words of Adcock and Collier: [R]esearchers may conclude that divergence suggests the indicators measure different (...) concepts and may reevaluate the conceptualisation that led them to expect convergence' (Adcock and Collier 2001: 540). 136 The thesis returns to evaluate the fit of the conceptualisation following the presentation and examination of the data material in Chapter 6. 5.4 Summary of operationalisation and observables In summary, four one-dimensional, direct indicators have been selected from the Standard Eurobarometer polls to measure each type of euroscepticism. The indicators were selected with the criteria of face validity--below, statistical tests will provide evidence as to whether or not the types are in fact coherent and independent. Euroscepticism is prevalent in a case if an indicator departs from the EU average in a long-lasting, sceptical way; its weight being determined by how many indicators are relatively eurosceptical. The basic observations from the analysis of euroscepticism in each case country should thus include: 1) Whether the country is relatively sceptic on one or more types 2) How strong each type is 3) How each type has fared over time In the comparative conclusion, the following must be established: 4) To what extent, and how, does euroscepticism differ and compare across cases? 5) What pattern(s) can be discerned over time? 6) Are there any void or unclear types of euroscepticism? With these observables in mind, a schematised overview of the operational statements and indicators can be presented. Table 7: Operationalisation Expectation 1 Utility I: economic euroscepticism Operational statement Indicator 1 This variant of utilitarian euroscepticism represents scepticism towards the perceived inability of the EU to bring tangible benefits (at macro and/or micro level). Feeling of having benefited from membership: `No benefiť (1983 ­ 2006) 137 Indicator 2 Indicator 3 Indicator 4 Tests Internal: Cross-type: Intensity: Meaning of the EU: `A waste of money' (2003 ­ 2005) Effect of the EU in specific economic areas: `Bad effecť and `Very bad effecť (1983, 2005) Role of the EU in difference areas: `The economic situation ­ negative role' (2003) Indicators 1, 2 and 4 (EB60, 61, and 62) ­ Gamma and alpha values Type 2: Indicator 1 with Indicator 8 (EB60) ­ Gamma values Type 3: Indicator 1 with Indicator 11 (EB38) ­ Gamma values Type 4: Indicator 1 with Indicators 13 (EB42) and 16 (EB60) ­ Gamma values Type 5: Indicator 1 with Indicator 19 (EB44) ­ Gamma values Indicator 1 with Indicator A (EB44) ­ Gamma values Expectation 2 Utility II: performance-based euroscepticism Operational statement Indicator 5 Indicator 6 Indicator 7 Indicator 8 Tests Internal: Cross-type: Intensity: This variant of utilitarian euroscepticism represents scepticism towards the perceived inability of the EU to act efficiently (at macro and/or micro level). Meaning of the EU: `A lot of bureaucracy' (2003 ­ 2005) Fraud concerning the EU and its budget happens very rarely: `Disagree' (2003) Fears connected to integration: `Decisions taken more slowly because of bureaucracy' (1995, 2001) Effectiveness of 15 EU policies:119 `Not effective' (2003) Indicators 5, 6 and 8 (EB60) ­ Gamma and alpha values Type 3: Indicator 6 with Indicator 9 (EB60) ­ Gamma values Type 4: Indicator 6 with Indicator 13 (EB60)­ Gamma values Type 5: Indicator 5 with Indicator 20 (EB61) ­ Gamma values Indicator 6 with Indicator A (EB60) ­ Gamma values 119 The 15 policy areas are: Maintaining peace and security; the Euro; Guaranteeing individual rights and respect for democratic principles; Guaranteeing food quality; Consumer protection; Fighting terrorism; Welcoming new member countries; Asserting the EU's global importance; Protecting the environment; Fighting organised crime and drug trafficking; Getting closer to the citizens; Fighting illegal immigration; Fighting poverty and social exclusion; Reforming the EU's institutions; Fighting unemployment. 138 Expectation 3 Sovereignty-based euroscepticism Operational statement Indicator 9 Indicator 10 Indicator 11 Indicator 12 Tests Internal: Cross-type: Intensity: Euroscepticism is the perception of threats posed by the EU to the continued relevance, integrity and identity of the nation-state. It is the scepticism towards developments that are thought to have a negative impact on national sovereignty. A European Government? `No supporť (plus `no neeď in 1996) (1987, 1989, 1990, 1992, 1993, 1995, 1996) Reasons for opposing the Constitutional Treaty: `Loss of national sovereignty' (2004, 2005) EU integration is a threat to national identity: `Agree' (1992) Joint or national decision-making (10 policy areas): `national level only' (1999, 2001) Indicators 9 and 11 (EB38) ­ Gamma values. Type 4: Indicator 9 with Indicator 13 (EB42) ­ Gamma values Type 5: Indicator 9 with Indicator 19 (EB44) ­ Gamma values Indicator 9 with Indicator A (EB60) ­ Gamma values Expectation 4 Democratic euroscepticism Operational statement Indicator 13 Indicator 14* Indicator 15 Indicator 16 Tests: Internal: Democratic euroscepticism is scepticism towards (perceived) democratic shortcomings in the EU. Satisfaction with democracy in the EU: `Not very satisfieď and `Not at all satisfieď (1993 ­ 2005) Reasons for opposing the Constitutional Treaty: `Not democratic enough' (2004, 2005) Is the EU democratic?: `No' (1989) The European Parliamenťs ability to protect citizens: `Not welľ and `Not at all welľ (1994, 1997, 1998, 2001) Indicators 13 and 16 (EB42, EB60, EB61) ­ Gamma values 139 Cross-type: Intensity: Type 5: Indicator 13 with Indicator 19 (EB61) ­ Gamma values Indicator 16 with Indicator A (EB60)­ Gamma values Expectation 5 Social euroscepticism Operational statement Indicator 17 Indicator 18* Indicator 19 Indicator 20 Tests: Internal: Cross-type: Intensity: Social euroscepticism is citizens' scepticism towards the EU's social engagement Priorities of the EU: `The EU should give more help to poor and socially excluded people within the EU'/`EU should pay less attention to the economy and more to social justice' (1996) Reason for opposing the Constitution: `Not enough social Europe' (2004, 2005) Fears about the EU: `The loss of social benefits' (1997, 1999, 2000, 2004, 2005) EU propositions: `There should be closer co-operation between member states in social matters ­ Agree' (2003) It is not possible to carry out statistical test on the internal coherence of these four indicators, as the thesis did not find them to coincide in one survey.120 (See under Expectations 1-4 above) Indicator 19 with Indicator A (EB61) ­ Gamma values Intensity Prevalence of hard euroscepticism Operational statement Indicator A Indicator B Indicator C To what extent is there in the case countries a principled, or hard, euroscepticism equal to the desire to do away with the EU. Opinion about membership of the EU: `Bad thing' (1974 ­ 2005) Reasons for opposing the Constitutional Treaty: `Against Europe/European construction/European integration' (2004, 2005) Personal feelings about the EU: `Rejecting iť (2002 ­ 2005) 120 Eurobarometer does contain other indicators related to attitudes towards social issues (for instance, whether or not it is an EU priority to fight unemployment or poverty), which overlap with at least one of the above four indicators. However, the substantive validity of these indicators is too weak for them to function as direct indicators: In order to be indicative of a critique of the EU's current social engagement, the question wording must contain a `more' or a `less' (cf. Indicator 1). 140 Indicator D Tests: Internal: EU referendum: `Leave the EU' (1995) Indicators A and C (EB63) ­ Gamma values Indicators A and D (EB44) ­ Gamma values *) This indicator was not included as a question in the Standard Eurobarometer data-sets, wherefore the thesis cannot employ it in the statistical tests. 5.5 Statistical tests and results This section is technical, and consists of the running of the intra- and inter-type tests listed in above table. Its purpose is to examine whether the five theoretically-deduced expectations as to what euroscepticism is are in fact coherent, independent types of the phenomenon. The intra-type validity of each of the five expectations is examined, before the thesis turns to the assessment of discriminant validity--i.e. the possibility of co-variance of one prevalent type with other prevalent types of euroscepticism. This section also examines the coherence of the indicators measuring the level of hard euroscepticism. The completeness of the tests listed in the above table is, as elsewhere mentioned, compromised by the unavailability of exhaustive Eurobarometer polls. It is not possible for the thesis to check indicator coherence and independence over as many years as would have been desirable. While the thesis maintains that the tests are of a large enough number to allow it to assume that similar results could have been obtained from, say, a poll five years later--and that they provide the authority with which it can reject or confirm the relevance of the theoretical expectations--it wishes to emphasise that the tests should be seen as attempts to falsify, and not `eternally verify', the conceptualisation. 5.5.1 Intra-type associations (coherence) * Expectation 1: Economic benefit Indicators 1, 2 and 4 have been posed by the same Eurobarometer poll on three consecutive occasions. It is suitable to check gamma values to see if an association between them can be confirmed, and alpha values to see if they form a reasonably coherent scale. Results are calculated by SPSS for the individual 141 case countries as well as for the EU as a whole. In 2003 (EB60), the latest year the three questions coincided, the gamma and alpha values were as follows (p<.000 throughout):121 GAMMA Denmark France UK EU No benefit ­ Waste of money: .750** .600** .720** .730** No benefit ­ Negative economic role .860** .780** .810** .770** Waste of money ­ Negative economic role .690** .680** .670** .600** ** p<0.00 CRONBACH'S ALPHA The three items together .680 .640 .680 .650 On the whole, we note that the gamma and alpha values live up to the requirements stipulated in Chapter 5. The thesis therefore deems that it is justified to speak of a coherent type of euroscepticism.122 We may moreover note from the gamma and alpha values that the three testable indicators especially form a good scale in Denmark and the United Kingdom. * Expectation 2: Performance Indicators 5, 6 and 8 coincided in the same Eurobarometer poll in 2003 (EB60). The thesis obtained the following values from the crosstab statistics and reliability analysis: GAMMA Denmark France UK EU Bureaucracy ­ Fraud .240* .310* .370** .150** Bureaucracy ­ Ineffectiveness .310** .300** .190* .230** Fraud ­ Ineffectiveness .280* .360** .230* .220** CRONBACH'S ALPHA The three items together .300 .280 .270 .220 ** p<0.00 * p<0.01 121 Gamma and alpha values for the two subsequent surveys (EB61 and EB62) are very similar to the above results. 142 Importantly, the indicators of this expected type of euroscepticism fail to live up to the thesis' requirements of fit, i.e. the gamma and alpha coefficients are consistently too low in the year where Eurobarometer allowed for a test between three indicators (2003). This suggests either that the type is flawed, and that there is no coherent, independent type of scepticism of the EU's ability to produce efficiently (or, as this seems unlikely, that the theoretical foundation was too crude and the type should in fact have been split in two to account for, on the one hand, bureaucracy and, on the other hand, ineffectiveness), and/or that the existing conceptualisation is relevant but that its operationalisation is weak; or that there simply are not enough adequate indicators in the Eurobarometer polls to produce a highly correlated pair or scale. Examining the indicators, the thesis deems both possibilities realistic: Each indicator seems to have something to do with an evaluation of the EU's overall effectiveness, but there is not necessarily high correlation between perceptions of bureaucracy and evaluations of the EU's effectiveness in specific policy areas. It does, however, come as a surprise to the thesis that the indicators of perception of bureaucracy, fraud and ineffectiveness are not correlated more strongly that the above figures show. Because of the strong theoretically-based expectation of performance-based euroscepticism, the thesis, instead of immediately rejecting the expectation of a coherent, independent type of performance-based euroscepticism, proceeds to evaluate the relevance of the identified indicators in terms of their correlation with indicators of hard euroscepticism, as well as their possible correlation with indicators of the remaining four expectations of euroscepticism. High correlation in the first instance would suggest that indicators of performance-based euroscepticism, although not forming a coherent type, are important for understanding the wish to withdraw from the EU (hard euroscepticism), while high correlation in the second instance would suggest that the critique of the EU's performance should be seen as integral to another type of euroscepticism, instead of as an independent type. Correlation with hard euroscepticism The thesis first examined the strength of the association between proposed indicators of performancebased euroscepticism (indicators 5 (meaning of the EU: bureaucracy); 6 (fraud); and 8 (ineffectiveness)) and an indicator polling the desire to withdraw from the EU (indicator A: opinion about membership). Gamma values are as follows: 122 To further check the validity of the items, the thesis ran a factor analysis on the three indicators (EB60), which came up with one component (explaining 61 percent of the total variance in Denmark and the United Kingdom, 57 percent of the variance in France, and 59 percent in the EU as a whole). 143 GAMMA (EB60) Denmark France UK EU Bureaucracy ­ Membership bad thing .520** (.260) .390** .500** Fraud ­ Membership bad thing .390** (.170) (.160) .160** Ineffectiveness ­ Membership bad thing .660** .470** .420** .370** ** p<0.00 ( ): not significant at p<0.01 Although there in most cases is an association between the thesis' proposed indicators of performancebased euroscepticism and the issue of membership in the Danish case, the general figures show the association to be too low to be considered significant by the thesis. Correlation with other types of euroscepticism As another possible remedy, the thesis examined whether the indicators of performance-based euroscepticism correlated highly with indicators of the other types of euroscepticism--as mentioned this would suggest the appropriateness of incorporating concerns about inefficiency into other concerns, instead of conceptualising these as an independent type. The thesis selected the indicator surveying attitudes towards EU effectiveness, as this was the indicator that correlated most strongly, and most consistently, with the membership indicator in the above analyses. However, as there was no one-dimensional indicator of social euroscepticism in this particular Eurobarometer poll (EB60), Eurobarometer 61 is used to test the association between performance-based euroscepticism and social euroscepticism. Gamma values are shown below: GAMMA Denmark France UK EU Ineffectiveness ­ No benefit .490** .240* .300** .280** Ineffectiveness ­ No EU government .410** .330** .270** .310** Ineffectiveness ­ EU democracy poor .540** .350** .490** .310** Bureaucracy ­ Fear loss of social benefits (.130) .450** (-.130) .120** ** p<0.00 * p<0.01 ( ): not significant at p<0.01 Although we note large differences between countries as regards to whether or not an association exists between proposed indicators of performance-based euroscepticism and indicators of the remaining expected types of euroscepticism, we note that there is not one instance where the gamma value is high 144 enough to be considered significant by the thesis. In relative terms, the association is strongest with regard to democratic euroscepticism. The thesis therefore has to accept that its expectation about the prevalence of a coherent, independent type of performance-based euroscepticism could not be corroborated by statistical evidence, wherefore it does not undertake a comparative analysis of its indicators in the three case countries. The thesis suggests that future studies into public opinion would benefit from engaging qualitative analyses, such as focus groups, in order to clarify how the critique of poor performance links to euroscepticism, and more precisely in what context the chosen performance-based indicators should be seen. Indeed, designing surveys that allow for the investigation of nuances within the criticism of EU effectiveness would assist our understanding of its potential contribution to a conceptualisation of coherent types of euroscepticism. * Sovereignty-based euroscepticism In 1992, Indicator 9 and 11 coincided. The gamma values below demonstrate a clear association between the two questions. GAMMA Denmark France UK EU Identity threatened by EU ­ No EU government .760** .750** .710** .710** ** p<0.00 In Chapter 4, the thesis found a growing expectation in the literature as to an association between sovereignty-based euroscepticism and resentment towards immigration, or xenophobia. As mentioned in that connection, this was not an expectation that the thesis expected to be able to adequately test. In the examination of Eurobarometer questions, however, the thesis was able to find, in 2003, the concomitant appearance of the survey question: `For each of the following statements, please tell me if you totally agree, tend to agree, tend to disagree or totally disagree: Immigrants are a threat to our way of life' and Indicator 9 of sovereignty-based euroscepticism (no EU government). The below gamma values show no significant association between the two questions: GAMMA Denmark France UK EU No EU government ­ Immigrants are a threat (.140) .270** .170* .170** ** p<0.00 * p<0.01 ( ): Not significant at p<0.01 145 The thesis acknowledges that these values are not sufficient to rule out in an authoritative way any possible association between xenophobia and some parts of euroscepticism--only it is able to demonstrate the lack of a strong association between the thesis' measure of sovereignty-based euroscepticism and a particular poll question gauging hostile opinions towards immigrants. More data would be required to further test the existence of a relationship. * Democratic euroscepticism Following the same procedures as with the above expectations of euroscepticism, the thesis tested the strength of the association between indicators of democratic euroscepticism. Indicators 13 and 16 coincide in 1994 (EB42), 2003 (EB60) and 2004 (EB61). As the below figures show, overall gamma values are acceptable to the thesis.123 EU democracy poor ­ The European Parliament is not able to protect interests: GAMMA Denmark France UK EU In Eurobarometer 42 .750** .500** .640** .640** In Eurobarometer 60 .590** .650** .750** .630** In Eurobarometer 61 .680** .680** .820** .710** ** p<0.00 * Social euroscepticism As mentioned in Table 7 above, none of the four indicators of this type coincide in one Standard Eurobarometer poll. The thesis thus has to rely on the face validity of the indicators of this type of euroscepticism, which all have a clear reference to the social aspect of EU co-operation. * Measure of hard euroscepticism It finally remains to test whether the four indicators of the level of hard euroscepticism are coherent. The thesis tests this by checking the strength of the association between the `testable' indicators of hard euroscepticism, namely Indicators A (membership bad thing) and C (rejection of EU), which coincided 123 It should be noted that out of the 12 runs, gamma values are twice somewhat below the stipulated .600 criterion: in France in EB42 and in Denmark in EB60 (although the latter discrepancy is very small). The thesis' decision to pursue with this type is based on the overall picture, which reveals a strong association between the two indicators, also in the two countries where values once fall below the threshold. 146 in 2003 (EB63), and Indicator A with Indicator D (`no' in referendum), which coincided in 1995 (EB44). GAMMA Denmark France UK EU Membership bad thing ­ Rejection of EU 1.000** .960** .930** .950** Membership bad thing ­ Would vote `no' in referendum on membership .990** .990** .990** .990** ** p<0.00 The gamma values show a uniformly strong association between the indicators of the level of hard euroscepticism, even attaining perfect association in Denmark.124 5.5.2 Status Having examined whether the five theoretical expectations about the nature of euroscepticism could be corroborated by statistical tests--testing as to whether they form coherent types--the thesis is able to pursue an investigation of four types of euroscepticism: Economic euroscepticism (`Type 1'), Sovereignty-based euroscepticism (`Type 2'), Democratic euroscepticism (`Type 3') and Social euroscepticism (`Type 4'). It was also possible to corroborate the thesis' measure of the level of hard euroscepticism. On the contrary, the tests led the thesis to abandon further analysis of performancebased euroscepticism--this theoretical expectation was, so to say, falsified. It is now relevant to test whether the four types of euroscepticism are also independent of one another, by examining the strength of the association between indicators of different types. 5.5.3 Inter-type associations (independence) This section turns to examine if co-variance between the four prevalent types of euroscepticism can be found. As discussed above, some degree of co-variance is expected, and may even be significant, as the types are `ideal types' and can hardly be expected to be clear-cut empirically. To reiterate, the criteria central to the thesis is that inter-type associations should be consistently weaker than intra-type associations. 124 The gamma value 1.000 reflects that all respondents who associate the EU with the feeling of rejection find that membership is a `bad thing'. This perfect association should of course not be mistaken for the two opinions being evenly widespread: as will be clear from the examination of the indicators below (Section 6.5), markedly fewer respondents associate the EU with the feeling of rejection, than membership being a `bad thing'. 147 ˇ Type 1 and Type 2 (Economic and Sovereignty-based euroscepticism) In 1992, in Eurobarometer 38, the question of perceived benefits from co-operation was polled alongside an indicator of sovereignty-based euroscepticism, namely the perception that European integration is a threat to national identity. Cross-tabulating these two indicators, we obtain the following gamma values: GAMMA Denmark France UK EU No benefit ­ Identity threatened by EU .750** .540** .500** .520** ** p<0.00 The gamma values suggest that an association between the two indicators does exist, perhaps due to both types having strong representation of hard euroscepticism (the thesis returns to measure this in Section 6.6)--although in France, the United Kingdom and overall in the EU it is considerably lower than the intra-type gamma values for economic (and, as we shall see below, sovereignty-based-) euroscepticism. However, in Denmark there seems to be a particularly strong association. A look at the contingency table for Denmark (see below) reveals that it is particularly within the segment of the surveyed population that claims not to benefit from EU membership that we find a strong association with the perception that a true Union represents a threat to national identity. Hardly anyone who perceives of `no benefiť from the EU rejects that it does not pose a threat to identity. However, of interest to discriminant validation, we may also note that there is still a considerable amount of respondents (276 people out of a total of 687 people) who believe that they have benefited from EU membership whilst still being afraid that the Union might come to compromise national integrity (cell C1R2). This is a smaller amount of respondents than that predicted from a random association. Yet, the number is high enough to suggest the existence of a considerable group of respondents who are not critical of the amount of benefit from the EU, but who are nevertheless concerned about the Union's impact on national identity. 148 Table 8: Contingency table ­ Identity threatened by EU * No benefit The EU has benefited DenmarkDENMARK Gamma = .750 `Agree' `Disagree' Total `Agree' Count 277 17 294 % within Benefit 50% 13% 43% `Disagree' Count 276 117 393 The EU does not pose a threat to national identity % within Benefit 50% 87% 57% 553 134 687 Total Count % within Benefit 100% 100% 100% This reasoning does not remove the fact that there is a high association between the two indicators in the Danish case. However, it does leave some interest in further investigating the motivations of the respondents figuring in cell C1R2. Given the acceptable gamma values in France, the United Kingdom and overall in the EU, the substantive argument of there being a difference between sovereignty concerns and economic concerns, as well as the consistent behaviour of the indicators in the actual data analysis (see Section 6.1 below), the thesis deems it is acceptable to maintain the distinction between the two types of euroscepticism. * Type 1 and Type 3 (Economic and democratic euroscepticism) In EB42, an indicator of economic euroscepticism--the feeling of benefit from membership-- coincided with an indicator for democratic euroscepticism--the satisfaction with EU democracy. The two indicators produce the below figures. In EB60, it was moreover possible to test the association between two different indicators of the two types, namely the perception that the EU is a `waste of money' and the feeling that the `European Parliament does not protect one's interests.' GAMMA (EB42 and EB60 respectively) Denmark France UK EU No benefit ­ EU democracy poor .610** .450** .630** .600** Waste of money ­ EP does not protect interests .460** .600** .630** .540** ** p<0.00 149 Once again, it is demonstrated that there is some overlap between the thesis' ideal types of euroscepticism as well as considerable variation across the cases. As the figures show, perceptions of `no benefiť and `dissatisfaction with EU democracy' produce a gamma value of .600, while perceptions of the Union as a `waste of money' and the feeling that the `European Parliament does not protect interests' have a gamma value of .540. In the three case countries, there seems to be some overlap between the types in especially the United Kingdom. However, as these gamma-values remain lower than the intra-type gamma-values of economic and democratic euroscepticism in the two countries, and given the arguable strong substantive validity of maintaining the independence of the two types (see sections 4.3 and 4.7), the planned analysis is continued, and a return will be made to the discussion of the possibility of overlap following the investigation of the actual behaviour of the indicators in the case countries (i.e. in Section 6.3). Moreover, the construction of even more specific indicators for the two types would considerably improve our understanding of possible overlap. As mentioned, the thesis at this stage accepts the finding of some overlap, as it is lower than the intra-type associations. * Type 1 and Type 4 (Economic and Social euroscepticism) As Section 6.4 below shows, Eurobarometer only started inquiring about citizens' concerns about losing social benefits in the mid-1990s, and the thesis has found no Standard Eurobarometer where two direct indicators of the type coincide. It is nonetheless possible to cross-tabulate an indicator of social euroscepticism with an indicator of economic euroscepticism. The below gamma values reveal the strength of the association (figures from Eurobarometer 44) between the feeling of not having benefited from membership and the fear of a loss of social benefits as a result of the process of European integration. GAMMA Denmark France UK EU No benefit ­ Fear of losing social benefits .520** .350** .290** .370** ** p<0.00 With gamma values clearly below the intra-type gamma values, the thesis is able to demonstrate the likely absence of a strong association between economic and social euroscepticism. 150 ˇ Type 2 and Type 3 (Sovereignty-based and democratic euroscepticism) GAMMA Denmark France UK EU No EU Government ­ EU democracy poor .460** .480** .590** .480** ** p<0.00 The above gamma values (derived from the Eurobarometer 42 dataset) demonstrate that the inter-type association between sovereignty-based euroscepticism and democratic euroscepticism is weaker than the intra-type association among the indicators of sovereignty-based euroscepticism. Not only is gamma lower than within the two respective types: at .480 it is moreover considered relatively weak by the thesis. * Type 2 and Type 4 (Sovereignty-based and social euroscepticism) GAMMA Denmark France UK EU No EU Government ­ Fear of losing social benefits (.110) .170** (.010) .130** ** p<0.00 ( ): Not significant at p<0.01 The gamma values are obtained from the Eurobarometer 44 data-set, where the indicator on support for a European government (sovereignty-based euroscepticism) coincided with the indicator on the fear of losing social benefits (social euroscepticism). With a value of .130 in the EU as a whole, and similarly low gamma values in the individual case countries, we may argue that there is only a weak association between the two types of scepticism. * Type 3 and Type 4 (Democratic and Social euroscepticism) The indicator of satisfaction with EU democracy and the indicator for Social euroscepticism--the fear of losing social benefits--coincided in 2004 in Eurobarometer 61. They produce the following gamma values in the case countries and in the EU as a whole: 151 GAMMA Denmark France UK EU Satisfaction EU democracy ­ Fear losing social benefits .510** .490** .290** .420** ** p<0.00 The gamma values show that the association between these two types of euroscepticism that is weaker than their individual intra-type associations. 5.5.4 Statistical conclusions It should be clear from the above that the results of the statistical tests of the coherence and independence of the various expectations of euroscepticism permit the thesis to proceed with a comparative analysis of four types, namely economic, sovereignty-based, democratic- and social euroscepticism. To reiterate, contrary to theoretical expectation, the thesis failed to find statistical evidence of a coherent type of performance-based euroscepticism. Importantly, statistical tests are not, on their own, sufficient for confirming the adequacy of the thesis' conceptualisation of euroscepticism--indeed, they say little about actual figures and can only supplement, and not replace, substantive arguments about coherence and independence. Final confirmation of the conceptualisation, therefore, awaits the analyses of empirical data, most notably whether the intra-type indicators behave consistently. Chapter 6 now turns to this examination. 152 6 Measuring euroscepticism in Denmark, France and the United Kingdom 6.1 Utilitarian euroscepticism: economic benefit125 Below, all four indicators of economic euroscepticism are introduced, illustrated, and discussed. The section concludes with a summary. Indicator 1 GENERAL BENEFIT FROM MEMBERSHIP. `Taking everything into consideration, would you say that (OUR COUNTRY) has on balance benefited or not from being a member of the European Union?' Not benefited Introduction Graph 1. Years: 1983 ­ 2006: Every third year, and final year, illustrated The figures This indicator, a classic in studies of public EU attitudes, reflects in a broad, general way whether or not citizens perceive an aggregate benefit as a result of their country's membership of the Union. It does not specify a particular type of benefit or direct attention to a specific sector or group of citizens; wherefore we may say that it is a rather abstract indicator. The indicator has been posed regularly since 1983, and is rather crude, with only a positive, a negative and a `don't know' reply category. Feeling of no benefit 31 22 30 21 26 19 24 23 15 57 50 42 37 44 39 38 47 44 21 24 21 32 38 31 28 34 40 25 27 28 29 38 32 30 35 33 0 10 20 30 40 50 60 1983 EB19 1986 EB25 1989 EB31 1992 EB37 1995 EB43 1998 EB49 2001 EB55 2004 EB61 2006 EB65 Percent DK UK FR EU The indicator shows pronounced dynamism, with opinion moving in different 125 It is interesting to note that economic opposition to the European Constitutional Treaty was not spontaneously voiced by respondents to the open question of why the document was being opposed. The thesis' remaining three types of euroscepticism were all spontaneously evoked. This could either suggest that citizens simply forgot economic concerns or that these were disentangled from the issue of the Constitution (See Eurobarometer 63). 153 Denmark The United Kingdom directions in the three case countries: In Denmark and the United Kingdom, opinion has become less sceptical over the 23-year period, while in France it has become more sceptical. Despite this dynamism, we note the relative stability of the three countries' positioning vis--vis each other: Since 1992, the most sceptical country has been the United Kingdom, followed by France (which may however, be on its way to becoming the most sceptical of the three cases, should the above-mentioned current trend continue). In 1989, for instance, Danes were nine percentage points more sceptical than the French. In 2006, the French were 25 percentage points more sceptical than the Danes. We may also note that, except for a simultaneous increase in scepticism across the cases in 1995, and a subsequent simultaneous fall in scepticism between 1995 and 1998, peaks of low and high scepticism have not followed the same pattern in the three countries. Whilst the number of British citizens who do not perceive EU benefit, for instance, consistently decreased throughout the measured years in the 1980s, it fluctuated considerably in Denmark while remaining stable in France. On the whole, absolute scepticism is generally within the medium level, currently bordering the low level in Denmark, and the high level in the United Kingdom (and currently also a high level in France). Danes are generally the most stable in their opinions, with scepticism ranging between 31 percent (1983) and 15 percent (2006). Danish scepticism has in general fallen: for the past 14 years, it has not been higher than 26 percent. This distinguishes Denmark from the EU average, showing a less sceptical viewpoint, with the difference often superseding ten percentage points. In the United Kingdom, scepticism has also decreased over the surveyed period--from a high of 57 percent who did not perceive of EU benefit in 1983 to 44 percent who shared that opinion in 2006. A low of 37 percent were sceptic in 1992, the year where the Danes rejected the Maastricht Treaty. The UK is consistently distinguished from the EU average by showing a more sceptical viewpoint. However, the gap has decreased from 32 percent in 1983 to 11 percent in 2006. 154 France The EU average French scepticism has gone in the opposite direction to Danish and British scepticism. Between 1983 and 2006, it rose rather consistently from 21 percent (the lowest in the period surveyed) to 40 percent (the highest in the period surveyed). The rise appears to commence in 1992, which we may recall was the year where British scepticism was at its lowest. Despite the significant increase in French sceptical opinion, scepticism has only been noticeably different from the EU average on two occasions: in 1989, the French were less sceptical than average, while in 2006, the French are for the first time distinguished in a more sceptical direction. This reflects a general increase in sceptical opinion in the EU as a whole. Scepticism on this indicator has risen from 25 percent in 1983 (the lowest of the time period) to 33 percent in 2006. It has largely followed the same ups and downs as French opinion, with a high of 38 percent in 1995. 155 Indicator 2 MEANING OF THE EU. `What does the European Union mean to you personally?' (Multiple answers possible): A waste of money Introduction Graph 2. Years: 2003 ­ 2005 The figures This poll question also provides a general, abstract indicator of perceptions of the EU as an economically unsound undertaking, this time from the perspective of the individual as opposed to the country as a whole. The reply categories were pre-given and interviewees could select more than one reply. The poll question was posed with slightly different reply categories in 1988 (EB29), 2001 (EB55) and 2002 (EB57), before, as illustrated below, being posed with the above formulation by four consecutive polls from 2003 (EB60). Meaning of EU: Waste of money 22 27 22 19 27 31 26 26 28 27 25 2524 26 23 23 0 5 10 15 20 25 30 35 2003EB60 2004EB61 2004EB62 2005EB63 Percent DK UK FR EU There is no marked development in opinion over the four surveyed polls, which cover a one and a half year time period. The initial figures change neither in a sceptical nor positive direction by more than five percentage points. Neither is any of the case countries significantly distinguished from the EU average. However, this finding should not hide the fact that Denmark is consistently the least sceptical case country, while France and the UK share the position as the most sceptical cases. The absolute scepticism level is low in Denmark and just reaching the medium-level in France and the United Kingdom. 156 Indicator 3 SPECIFIC BENEFITS FROM MEMBERSHIP. `People disagree about the advantages and disadvantages of (OUR COUNTRY) belonging to the European Union. I am going to read out some points and, for each one, I would like you to tell me if (OUR COUNTRY) being in the European Union has a very good, fairly good, fairly bad or very bad effecť: `Our exports; Our Industry; Our Agriculture; Our standard of living'. Bad effect and Very bad effect Introduction Graph 3. Years: 1983 and 2005 The figures This is a specific poll question indicating whether or not the EU's influence on a number of named economic areas is perceived to be beneficial. The possibility of two sceptical reply categories of different intensity should in combination give a nuanced indication of scepticism in this regard. The poll question has been posed twice with a large time gap in between: in 1983 and in 2005. In 2005, attitudes towards the EU's effect on the service sector and on employment were also questioned. Bad effect in specific areas, 1983 and 2005 16 36 27 23 7 29 30 23 20 53 29 29 14 46 49 37 21 49 22 23 16 30 50 36 25 56 40 39 24 54 53 48 0 10 20 30 40 50 60 DK UK FR EU Percent EXPORTS INDUSTRY STANDARD OF LIVING AGRICULTURE The graph above illustrates significant and contrasting developments in opinion on this indicator: briefly, scepticism has risen significantly in France and on average in the EU, while it has fallen significantly in Denmark and in the United Kingdom. Across the three cases, `standard of living' has experienced the greatest overall change in scepticism level, albeit, interestingly, with a different prefix: it is the area with the steepest increase in scepticism in France (and in the EU as a whole) and the area with 157 Denmark The United Kingdom France the steepest decrease in sceptical replies in the United Kingdom. As the paragraphs on the individual countries will specify in more detail, the absolute scepticism level varies from area to area, from case to case and over time. Danish scepticism was, and continues to be, low in all four areas, both in relative and in absolute terms. Between 1983 and 2005, sceptical replies have fallen by more than five percentage points in two of the four areas (exports, where it at seven percent has become almost negligible, and industry). Danish scepticism is highest towards `agriculture' (25 percent in 2005), which could in part be due to the close links between this area and the EU's controversial Common Agricultural Policy. However, on all four areas scepticism is more than 15 percentage points below the EU average. Moreover, although EU support is not the topic of this thesis, it is worth mentioning that Danish positive evaluations of three of the four areas in 2005 were the highest in EU-25.126 British scepticism, like Danish scepticism, was considerably lower in 2005 than in 1983. In three of the four areas, sceptical replies have fallen by more than five percentage points. `Agriculture' has, like in Denmark, witnessed the lowest fall in absolute scepticism. Despite the general decrease in scepticism, British opinion in 2005 was still more sceptical than the EU average in three of the areas (exports, industry and agriculture). A 19 percentage point drop in sceptical opinion towards the EU's effect on standards of living leaves the UK six percentage points less sceptical than average. Scepticism is highest in the area of agriculture (54 percent in 2005), and lowest with regards to exports (29 percent in 2005). Although not illustrated in the above graph, the British exhibit the most `non-scepticaľ view compared to the EU average in one of the areas that was only included in the survey in 2005, namely the employment situation (by seven percentage points). Curiously, this is precisely the area where the French show the widest deviation from the average in a sceptical direction (by 17 percentage points). In France, opinion has moved in the opposite direction to the trend in Denmark and 126 88, 80 and 70 percent of Danes thought that the EU had a `very gooď or `gooď effect on, respectively, exports, industry and agriculture. The EU average was at 64, 51 and 40 percent respectively. 158 The EU average the United Kingdom. Feelings of negative EU impact have increased in all four areas, topped by a 28 percentage point increase in precisely that area where scepticism dropped by 19 percentage points in the UK, namely the standard of living. While the EU average has also generally increased, the growth in French scepticism has been steeper, which means that from being in line with the EU average in 1983 in all four areas, France is today significantly more sceptical than average in three areas (again, agriculture is the exception), with scepticism superseding the average by more than ten percentage points when it comes to industry and standard of living. This evidently means that the Danes and the French diverge considerably on this indicator. Generally, Danish scepticism is `low', while French scepticism is `medium' or `high'. In two of the areas, the difference between the two cases even supersedes 30 percentage points (standard of living and industry). In 1983, French scepticism ranged between 22 and 40 percent; in 2005 it ranged between 30 and 53 percent. Scepticism in France is most pronounced on agriculture, closely followed by standard of living and industry. Scepticism towards the EU's effect on exports is significantly lower--as is the case in Denmark and the UK. Overall in the EU, scepticism has increased on three of the four areas (industry, standard of living and agriculture), while remaining stable within the remaining area. The increase has been of about 10 percentage points. Both in 1983 and in 2005, scepticism was highest towards the EU's impact on agriculture (39 and 48 percent respectively) and lowest towards the EU's impact on exports (23 percent in both years). In absolute terms, scepticism is mainly within the medium-level. 159 Indicator 4 ROLE OF THE EU. `And for each of the following issues in (OUR COUNTRY), do you think that the European Union plays a positive role, a negative role or neither positive nor negative role?' `The economic situation': negative role. Introduction Graph 4. Year: 2003 The figures This indicator specifically surveys whether citizens perceive the EU to have a negative impact on their country's economic situation. The wording of the question, however, is vague (economic situation can refer to a multitude of things), and therefore it is arguably rather `easy' to give either a negative or a positive reply. However; the availability of a neutral reply category in addition to the `don't know' reply arguably refines the survey results. This wording of the poll question has, it appears, only been posed once. The economic situation - negative role 19 35 40 33 0 5 10 15 20 25 30 35 40 45 2003 EB60 Percent DK UK FR EU This recent indicator shows France to be the most sceptical of the case countries, and with seven percentage point more sceptics than the EU average, this represents a significant relative euroscepticism to the thesis. The United Kingdom is in line with the EU average, while Danes are significantly less sceptical: 19 percent perceive the EU's role on the domestic economic situation to be negative, which is 14 percentage points lower than the EU average and 21 percentage points lower than in France. In absolute terms, scepticism is thus within the lower level in Denmark, while in the medium level in France, the United Kingdom, and on average in the EU. 160 6.1.1 Summary Economic euroscepticism registers marked differences across the three cases and over time. Today, the United Kingdom and France count as relative sceptics, while scepticism in Denmark on the contrary is much lower than on average in the EU. The difference between France and the UK on the one hand, and Denmark on the other hand, is more than 20 percentage points in three of the four indicators. France is the most sceptical case on two of the four indicators, while the UK is on the other two. However, the increase over time of French sceptical opinion combined with a somewhat decrease in British scepticism, suggests that France could in the years to come emerge as the most sceptical of the three countries as regards the economic benefits of membership. Interestingly, the French continue not to exhibit considerable difference from the EU average in the abstract indicator of economic euroscepticism. However, French opinions seem to conform to a general increase in the number of EU citizens who feel their country has not benefited from membership. Thus, there has been between 1983 and 2004 a gradual increase in sceptical opinions of 13 and 10 percentage points respectively. Also on the indicator of EU effect on specific economic areas (Indicator 3), the EU average has registered an increase of sceptical opinions in three of the four areas. The discussion of the development of abstract and specific indicators will be returned to in Chapter 7. Although France and the UK emerge as the most eurosceptical cases in 2005, it is important to note that they still differ considerably in the individual indicators. Indicator 3, for instance, showed that precisely that area where the British showed the widest divergence from the EU average in terms of a non-sceptical attitude (the employment situation), was the area where the French showed the widest divergence from the average in a sceptical direction. This underlines the importance of having multiple indicators to illustrate a type of euroscepticism. * France and the United Kingdom are eurosceptic, but neither is distinguished by scepticism in all four indicators. The two countries diverge considerably within the individual indicators * Denmark is distinguished by less scepticism than the EU average in all four indicators. Indeed, it is not (and has hardly ever been) appropriate to speak of relative Danish euroscepticism in this regard * Marked changes can be seen over time, with a general increase in euroscepticism on average in the European Union. Generally, Denmark and France have `switched places', with France today consistently emerging as the more sceptical of the two countries * Peaks of low and high scepticism have not followed the same patterns in the case countries. Both Denmark and the United Kingdom experienced low points in 1992, the year the Maastricht Treaty 161 was being ratified. The rise of French euroscepticism on the long-term indicators of this type appears to begin in that year 162 6.2 Sovereignty-based euroscepticism In line with the above section, this section introduces and discusses the four indicators of sovereigntybased euroscepticism, concluding with a summary (the thesis keeps the numbering of indicators shown in Table 7, although as explicated in Section 5.5.3, it was not able to pursue with an analysis of performance-based euroscepticism). Indicator 9 EUROPEAN GOVERNMENT. `Are you for or against the formation of a European Government responsible to the European Parliament?' Against (plus no need in 1996). Question wordings varied slightly in some years. In 1992, for instance, it read: `Are you for or against the formation of a European Union with a European government responsible to the European Parliament?' Introduction Graph 5. Years: from 1987 to 1996, with a gap of three surveys (in EB37, the question was not posed. The thesis includes EB38 instead) Governments are central features of sovereign states, and the question of a government at EU level is therefore likely to underline the supranational aspects of co-operation. The dichotomous question surveying attitudes to the formation of an EU government was posed by almost every Eurobarometer between 1987 and 1996. In the latter year, an additional poll question was added, namely whether citizens at all thought an EU government was necessary for the functioning of the Union. Interestingly, the question has not appeared in a Eurobarometer poll since 1996. Against an EU-government 64 53 61 67 69 62 64 45 33 37 56 49 42 43 19 13 19 31 28 18 2424 17 19 28 27 23 24 0 20 40 60 80 1987 EB28 1989 EB31 1990 EB34 1992 EB38 1993 EB40 1995 EB43 1996 EB46 Percent DK UK FR EU 163 The figures Graph 6. Year: 1996 The figures Throughout the seven surveyed years, Danes typically exhibit a higher degree of scepticism than the EU average by an astounding 40 percentage points. This easily makes Denmark the most sceptical of the three cases. It is followed by the United Kingdom, which is also consistently and significantly more sceptical than the EU average. France, however, is not different from the average in any of the surveyed years. The level of scepticism towards an EU government is rather stable in all cases, and stability also characterises the EU average. A sudden leap in scepticism, however, occurred in 1992, at the time of the Maastricht ratification process. From 1990 to 1992, British scepticism surged by 19 percentage points, from 37 to 56 percent; French scepticism increased by 12 percentage points, from 19 to 31 percent, while Danish scepticism went up by six percentage points, from 61 to 67 percent. The increase at the EU level was at nine percentage points. By 1995, however, the level of scepticism had `normaliseď again. In Denmark, the absolute level of scepticism is consistently high; it is medium in the United Kingdom (except for in 1992, where it reached the high level); and lower-medium, or low, in France as well as on average in the EU. The EU does not need a government 82 39 29 31 0 10 20 30 40 50 60 70 80 90 1996 EB45 Percent DK UK FR EU It is clear from this additional question that Danes' vision of EU co-operation does not involve an EU government: 82 percent were of that opinion in 1996. Although the British are also higher than the EU average of scepticism on this indicator, an astounding 43 percentage points still separate the two case countries. 51 and 53 164 percentage points more Danes than the EU average (31 percent) and the French (29 percent) believed that there was no need for a government at EU level. 165 Indicator 10 EU CONSTITUTION. `What are all the reasons why you are opposed to the European Constitution?' Loss of national sovereignty Introduction This survey question, which Eurobarometer posed twice in connection with the ratification process of the Constitutional Treaty prior to its rejection in France and the Netherlands, opened the possibility for a range of reply categories. The question was posed only to the segment of the population declaring itself opposed to the Constitution. Replies were not pre-given but spontaneous, and multiple replies could be given. `Loss of national sovereignty' is the reply category to this poll question relevant for the measuring of sovereignty-based euroscepticism. In fact, as the chapter demonstrates, this open-ended indicator can be used to survey three out of the thesis' four types of euroscepticism. As this is the first introduction to the indicator, it is worth mentioning that, all in all, in the EU-25, replies fell into 14 reply categories (see EB63 annex). The reason the most frequently given in the EU-25 for opposing the Constitutional Treaty was `lack of information' (32 percent in 2005). With regard to the thesis' three case countries, Danes were mostly opposed because of the `loss of national sovereignty' (42 percent); the French because of `not enough social Europe' (27 percent); and the British, like the Danes, were mostly opposed because of the `loss of national sovereignty' (42 percent) that was perceived to result from ratification of the Constitutional Treaty.127 127 The least-evoked reasons were as follows: The EU-25 as a whole: `The Constitution does not go far enough' (three percent). Denmark: `The Constitution does not go far enough', `Opposition to the national government/certain political parties', and `No reference to the Christian roots of Europe' (all at two percent). France: `The Constitution does not go far enough' (two percent). The United Kingdom: `The Constitution does not go far enough', and `Not enough social Europe' (both at two percent). 166 Graph 7. Years: 2004 and 2005 The figures The EU Constitution: Against because of loss of national sovereignty 32 42 50 42 23 18 37 32 0 10 20 30 40 50 60 2004 2005 Percent DK UK FR EU25 Responses to this reply category are substantial in number, with some 35 percent on average in the EU's member states claiming to oppose the Constitutional Treaty because of the fear of losing national sovereignty. In 2005, the United Kingdom and Denmark were distinguished from the EU average in a sceptical direction (by 10 percentage points). France is consistently less sceptical (by 14 percentage points; scepticism thus falling in the low level). Though there is some movement in replies between the two surveyed years, there is general stability with regard to the most and the least sceptical countries. 167 Indicator 11 THE EU AND NATIONAL IDENTITY. `There is a lot of talk about national identity and European identity in the countries of the European Community. Some say (A) If a real European Union ever came about, it would mean the end of our national cultural identities and their diversity (UNION ENDS IDENTITIES). Others say (B) The only way to protect our national cultural identities and their diversity, is through the countries of Europe caning to a real European Union (UNION PROTECTS IDENTITIES). Do you feel closer to the first (A) or to the second (B) of these two statements?': Closer to the first Introduction Graph 8. Year: 1992 The figures This indicator is included to add the identity perspective to the sovereignty type of euroscepticism. Arguably, citizens who perceive EU integration as a threat to national identity are likely to feel that integration has a negative impact on their country's sovereignty. Based on this study's examination of Eurobarometer polls, this direct and specific framing of the question has been posed only once, namely in 1992, the year of the ratification process of the Maastricht Treaty. The poll question allowed a seven-point reply scale. The thesis has taken the three most sceptical replies to constitute euroscepticism. Integration is a threat to national identity 45 42 34 30 0 10 20 30 40 50 1992 EB38 Percent DK UK FR EU Denmark is the most sceptical of the three cases, closely followed by the United Kingdom. With respectively 15 percentage points and 12 percentage points to the EU average, it is a pronounced relative scepticism. France is not distinguished from the average, which is at 30 percent. 168 Indicator 12 POLICIES: NATIONAL OR EU LEVEL DECISION-MAKING? `For each of the following areas, do you think that decisions should be made by the (NATIONALITY) government, or made jointly within the European Union?' (10 policy areas) National level Introduction Graph 9. Year: 1999 The question of whether specific policy areas should be decided upon at national level or in combination at the EU level is a Eurobarometer classic. However, question formulation and the identity of the policy areas under scrutiny have varied considerably. The below illustrations are of ten policy areas subjected to public evaluation in 1999 and 2001. Decision making level - national 1999 0 10 20 30 40 50 60 70 80 90 100 DK UK FR EU DK 71 81 78 86 65 54 34 34 54 68 UK 63 51 77 67 66 47 41 44 42 61 FR 35 52 63 68 40 38 23 25 35 41 EU 42 52 62 62 46 37 29 28 34 44 Accept- ing refugees Cultural policy Education Health & social w elfare Immi- gration policy Protecting the environm Scientific & techn. research Fight against drugs Fight against poverty Fight against unemploy 169 Graph 10. Year: 2001 Figures in 2001 Difference from 1999 Decision making level - national 2001 0 10 20 30 40 50 60 70 80 90 100 DK UK FR EU DK 68 80 72 82 65 44 32 27 38 62 UK 59 52 73 63 65 44 35 37 36 58 FR 36 51 63 69 44 32 24 22 32 45 EU 43 49 61 59 48 33 27 26 30 43 Accept- ing refugees Cultural policy Education Health & social w elfare Immi- gration policy Protecting the environm Scientific & techn. research Fight against drugs Fight against poverty Fight against unemploy There is some variation across the policy areas as regards which of the case countries is the most intent on leaving decision-making exclusively at the national level. In 2001, Danes are the most national-minded in half of the areas; in three areas, it is the British, while in the remaining two areas (immigration and protecting the environment), the two countries are even. The French are thus never the most national-minded in any of the policy areas. On the contrary, the French are less supportive than the EU average of retaining decision-making at the national level on half the policy areas, and there is only one area, health and social welfare, where the French to a higher degree than the EU average support the retaining of decision-making at the national level. Both Denmark and the United Kingdom are more inclined than average to support the retaining of decision-making power at the national level in all policy areas. However, Denmark is the only case country where the difference to the average sometimes supersedes 20 percentage points. Compared to the 1999 level, Danish sceptical attitudes towards EU involvement had in 2001 dropped with regard to five of the policy areas: education (by six percentage points), protecting the environment (by ten percentage points), the fight against drugs (by seven 170 percentage points), the fight against poverty (by 16 percentage points), and the fight against unemployment (by six percentage points). In the United Kingdom, scepticism witnessed a drop in three of the areas: scientific research (by six percentage points), the fight against drugs (by seven percentage points) and the fight against poverty (by six percentage points). With the exception of education and the fight against unemployment in the Danish case, we may note that the policy areas witnessing a decrease in the wish to maintain decision-making nationally in Denmark and the United Kingdom are largely areas in which there is already considerable EU co-operation and which are arguably not the most sensitive to national sovereignty. This brings us back to Chapter 4's discussion about the likelihood of a balancing between the possible benefits achievable by European co-operation with the caseto-case evaluation of the impact of co-operation on national integrity. In France, the wish to maintain national decision-making with regard to protecting the environment had increased by six percentage points--apart from this, opinions were stable. Stability across the ten policy areas characterise the EU average. 6.2.1 Summary Above, four indicators measure EU citizens' scepticism towards the idea of sharing national sovereignty in the Union. All indicators present unequivocal differences among the three cases, or rather between France and the EU average on the one hand, and Denmark and the United Kingdom on the other. While France generally shares a `low to medium' level of euroscepticism in this regard, Denmark--the most sceptic case--has never experienced a low level of scepticism on any of the indicators in any of the measured years. Figures serve to recall the differences among the cases within this type of euroscepticism. In 1992, 48 percent of Danes perceived a `real European Union' to mean the end of national identity. The figure in France was 34 percent, while 45 percent of the British shared that opinion. In 1996, 63 percent of the Danes were against the creation of an EU government responsible to the European Parliament--compared to 18 percent in France and 32 percent in the UK. That year Eurobarometer moreover asked citizens whether they thought such a government was necessary or not. 29 percent in France, and 39 percent in Britain, did not think so. The figure for Denmark was 82 percent. With regard to the level of decision-making, France is only relatively sceptical towards EU decision-making in one of the 17 policy areas (health and social welfare). Denmark and the UK are more strongly opposed than average to EU decision-making in all of the policy areas. The indicator on the 171 Constitution from 2005 shows that 14 percentage points fewer Frenchmen than the EU average of 32 percent opposed the document because of its implications for national sovereignty. In Denmark and the United Kingdom, 24 percentage points more respondents expressed this opinion, namely 42 percent. Movement in opinions within this type could most clearly be measured within the indicator on an EU government from 1987 to 1996. Here, Denmark, France and the EU average report stable opinions. Denmark is consistently within the category of `high' euroscepticism, while France and the EU average are in the `low' category. The UK is a bit more variable, but always within the `mediumleveľ category. Although this is not a study of elite euroscepticism, it seems fair to note that a frequent interpretation has emerged of French politicians at EU summits being very concerned about French sovereignty and national identity in the integration process. It is interesting to see the thesis' findings on the absence of sovereignty-based euroscepticism within the French public in this light. * Denmark and the United Kingdom are eurosceptic, in that order of intensity. France is not eurosceptic in this regard * The level of scepticism is stable on the one indicator allowing a long time perspective; however, a sudden momentary leap in scepticism occurred in all case countries, as well as on average in the EU, around 1992 and the Maastricht ratification process 172 6.3 Democratic euroscepticism Indicator 13 EU DEMOCRACY. `On the whole, are you very satisfied, fairly satisfied, not very satisfied or not at all satisfied with the way democracy works in the European Union?' Not very satisfied and Not at all satisfied. Introduction Graph 11. Years: 1993 ­ 2004. (Due to smaller gaps in the years where Eurobarometer has posed this indicator, the only regular interval was a 45 year one) The figures This poll question was asked for the first time in 1993 (EB39), perhaps as a direct reaction to the Danish rejection of the Maastricht Treaty in June 1992, which was generally interpreted as a critique of the EU's democratic standing (see discussion above). It is always posed in connection with a question surveying attitudes towards national democracy--a question Eurobarometer had been posing for a number of years prior to 1993. The EU variant of the question almost immediately achieved permanent status in the Eurobarometer polls. As such, it has been posed in 18 of the 26 Eurobarometers published since 1993. The poll question surveys satisfaction with the general level of democracy in the EU as a whole, wherefore it is a rather broad indicator. Dissatisfaction with EU-democracy 0 10 20 30 40 50 60 70 Percent DK 48 46 58 57 35 34 UK 47 47 36 33 37 34 FR 50 51 46 35 40 41 EU 47 48 44 37 36 37 1993 EB39 1995 EB43 1997 EB48 2000 EB53 2002 EB58 2005 EB63 Over this 12-year period, Denmark is the only case country that has been more sceptical than the EU average (in 1997 and in 2000, by 14 and 20 percentage points respectively), while the United Kingdom is the only case to have shown a less 173 sceptical tendency (in 1997, by eight percentage points). Over the past four years, all three cases are in line with the EU average. The overall movement has been towards less dissatisfaction with EU democracy: Since 1993, scepticism in all three cases has dropped: by 14 percentage points in Denmark, 13 percentage points in the UK and nine percentage points in France. The EU average is in line with this development, having witnessed a rather consistent ten percentage point drop in scepticism over the 12 surveyed years. In terms of absolute scepticism, the indicator is generally within the medium level. 174 Indicator 14 EU CONSTITUTION. `What are all the reasons why you are opposed to the European Constitution?' Not democratic enough Introduction Graph 12. Years: 2004 and 2005 The figures As mentioned in Section 6.2 this open poll question was posed twice in connection with the ratification process of the Constitutional Treaty (prior to the French and Dutch rejections). Its multiple reply categories specifically included dissatisfaction with the democratic credentials of the Constitutional Treaty. The EU Constitution: Not democratic enough 0 4 8 12 16 20 2004 EB62 2005 EB63 Percent DK UK FR EU25 Again, Denmark is the only case to be significantly distinguished from the EU average in a sceptical direction. In 2005, 17 percent of Danes opposed the Constitutional Treaty for its lack of democratic credentials, as opposed to 11 percent in the EU as a whole. Both France and the United Kingdom are in line with the EU average; however, in absolute numbers, their percentage of sceptical replies is so low that it is almost not relevant to speak of scepticism in this regard, especially considering that the survey question allowed interviewees to give multiple reasons for their opposition to the Constitutional Treaty. It is possible that if reply categories had been pre-given, as opposed to spontaneous, sceptical figures would have been higher, but this does not hide the conclusion that democratic concerns were hardly the foremost concern of most citizens with regard to the Constitutional document. 175 Indicator 15 EC DEMOCRACY. `Is the Community democratic?' Very little and Not at all Introduction Graph 13. Year: 1989 'The figures It is worth noting that this one-off question was posed already in 1989, and thus prior to the Danish rejection of the Maastricht Treaty, which otherwise provoked the surge in the attention given to the EU's alleged democratic shortcomings. It is the only question surveying such opinions prior to 1993, where the general question on attitudes to EU democracy (cf. Indicator 1) was introduced. While the question is very blunt, it does allow for two sceptical reply categories (`very little' and `not at alľ). The Community is not very democratic 28 20 29 20 0 5 10 15 20 25 30 35 1989 EB32 Percent DK UK FR EU 28 percent in Denmark and 29 percent in France had a clear view of the then European Community as not very democratic. This makes both countries relatively eurosceptic, while the United Kingdom is not distinguished from the EU average of 20 percent. These levels moreover place the UK in the low level of scepticism (alongside the EU average), and Denmark and France in the lower part of the medium level. 176 Indicator 16 INTEREST PROTECTION. `As a European citizen, do you think that the European Parliament protects your interests...?' Not very well and Not at all well Introduction Graph 14. Years 1994, 1997, 1998, 2001, 2004 (Note: uneven gaps among years) Denmark This indicator has been posed at irregular intervals by Eurobarometer since 1994 (see Annex for years). In the graph below, the thesis has nevertheless chosen to illustrate it with a continuous line, in order to gain some sense of its level and development. However, because of the uneven gaps between the points of the graph, it should of course only be used cautiously for inferences about its development.128 The EP: Bad and very bad interest protection 46 55 48 46 40 31 33 35 40 49 37 34 36 38 33 32 0 10 20 30 40 50 60 1994 EB42 1997 EB47 1998 EB50 2001 EB54 Percent DK UK FR EU Again, Denmark is the case country that shows the greatest variance from the EU average in a sceptical direction. Consistently throughout the surveyed period of seven years, Danes are relatively eurosceptic, and with figures around 50 percent, the level is rather high. In all surveyed years, the difference from average is ten percentage points or higher. 128 It could be noted that similar questions have been posed by Eurobarometer in additional years. In 2004, for instance, EB61 had a range of agree/disagree questions, including: `for each of the following statements, please tell me whether you totally agree, tend to agree, tend to disagree or totally disagree: The members of the European Parliament are good at protecting your interests'. The thesis has chosen to examine only polls that have the same question wording/reply categories. 177 France The United Kingdom The EU average The French exhibit a greater level of scepticism than the average in one of the surveyed years (1997), albeit to a lesser extent than the Danes. Nevertheless, throughout the four polls, the French are some four percentage points more sceptical than the EU average. Scepticism has decreased somewhat, from 40 percent who did not think the European Parliament was good at protecting their interests in 1994, to 34 percent who shared that opinion in 2001. The least sceptical case country is generally the United Kingdom. It has not been relatively eurosceptic in any of the surveyed years, and was even less sceptical than average in 1997 (note that this was the year where French scepticism topped). Contrary to France, scepticism has risen by eight percentage points from 1994 to 2001. The EU average registers no great change in scepticism. 6.3.1 Summary With regard to the three case countries, a number of common denominators emerge. Scepticism is generally within the medium-level (this is also the case for the EU average) and rather stable over time. There are, however, interesting variations with regard to their levels of relative scepticism, with differences between the most and least sceptical case often superseding ten percentage points. Put crudely, within this type of euroscepticism, Denmark is the most sceptical compared to the EU average, while the United Kingdom is the least sceptical. In the years surveyed, the Danes can be seen to exhibit a higher level of scepticism than the EU average in all four indicators; the French have on two, while the British have on none. Danes, for instance, differ from the British and the French in sharing a relative scepticism towards the European Constitution's perceived lack of democratic credentials. Moreover, Denmark is the only case where the difference from the average has often superseded ten percentage points. It should be mentioned, however, that on the indicator where the Danes have shown the greatest difference from the EU average (dissatisfaction with EU democracy), there can be observed a marked decrease in the absolute level of scepticism over time, which has fallen from 54 percent in 1994 to 34 percent in 2005. The UK is only found to be different from the EU average in one of the polled years: in 1997, the British were less dissatisfied than average with democracy in the EU. That year saw a peak in the 178 number of French citizens who were sceptical of the European Parliamenťs ability to protect citizens. We may thus again note a contrast between the scepticism of France and that of the UK. The French, moreover, shared a more pessimistic view than average on EC democracy in 1989. * Denmark is the most eurosceptic case country, but was for the first time in 2005 no longer distinguished on all four indicators * France sends mixed signals, and is generally at odds with the UK with regard to its evaluation of the EU's democratic credentials * In recent years there are, however, no marked differences across the cases Following this discussion of democratic euroscepticism, it is appropriate to return to the finding of moderate co-variance with the economic type of euroscepticism in the statistical analysis in Section 5.5.3. Indeed, does the actual behaviour of the indicators representing the two types strengthen the suspicion that, although distinct theoretically, the two types are in fact more or less expressions of the same euroscepticism? The thesis does not take this to be the case. First of all, within each of the two types, the indicators behave logically. That is, across the four indicators, the three cases are consistently placed vis--vis each other, and vis--vis the EU average. A merger of the indicators of the two types would not have produced a clear picture. Indeed, the finding that French euroscepticism on the specific indicators of economic euroscepticism has sharply increased, while it has decreased on the indicators of democratic euroscepticism, is to the thesis one example in support of the two types being independent of one another--just as it was envisaged by Sections 4.3 and 4.7. Moreover, and perhaps more importantly, it is possible that there exists a causal relationship between the two types of euroscepticism, which contributes to the relatively high gamma values revealed by the statistical tests in Section 5.5.3 above. The idea is not novel. With specific regard to public attitudes towards the EU, Pippa Norris has furthered the argument that there might in fact exist a causal relationship between opinions towards the Eurobarometer indicators of `EU benefiť and `satisfaction with EU democracy' (and thus, implicitly, the two types of scepticism discerned by the thesis). She finds that `overall policy performance lies at the heart of (dis-) satisfaction with EU democracy. If people feel that there are direct benefits from membership then they are more likely to feel positive about the broad principles of the regime' (Norris 1999: 88). Norris, in other words, does not question the distinct and independent nature of benefit evaluations on the one hand, and democracy evaluations on the other--rather she construes the former as the independent variable and the latter as the dependent variable. It is beyond the scope of the thesis to pursue Norris' argument. However, it acknowledges the possibility of causal 179 relationships between the types of euroscepticism, and can only agree that this constitutes an interesting avenue for further research (see Chapter 7 in this regard). In this connection we may also note the finding by Jens Henrik Haahr that an ethno-nationalist discourse underlies Danish conceptions of democracy (Haahr 2003), which could explain why the degree of overlap between sovereignty-based euroscepticism and democratic euroscepticism was relatively low in Denmark. Haahr notes that the perceived equivalence in Denmark between democracy and one homogeneous people, and the understanding of democracy as parliament, result in `the total absence of arguments for a strengthening of parliamentarism at the European level in response to a `democratic deficiť' (Haahr 2003: 40). In this sense, if the wish to protect the powers of the national parliament breeds sovereignty-based euroscepticism, this wish may simultaneously make criticism of the Union's democratic credentials less pronounced. 180 6.4 Social euroscepticism This section accounts for the fourth and final of the thesis' types of euroscepticism, namely the critique of a Union that is too weak on the social dimension. Indicator 17 EU PRIORITIES. `Some people expect the European Union to become (even) more active than now in certain areas. For each of the following, please tell me if you consider it a key priority or noť. `Giving more help to the poor and the socially excluded in the European Union'; `Paying less attention to the economy and more to social justice' (multiple replies possible). A priority Introduction These two sub-questions within the general poll question on EU priorities allow citizens to voice their wish for more EU action in the social welfare domain. The first sub-question calls for more redistribution to ensure equality amongst EU citizens, while the second explicitly targets the classic relationship between liberalism and socialism. Eurobarometer surveyed citizen priorities towards a whole range of pre-given areas, and multiple priorities could be voiced. The `priority question' can thus be expected to facilitate many replies in favour of more action: it is not difficult to state that an area is important and that one would like more to be done. In particular to the first sub-question, `positive' replies merely suggest in an unspecified and uncommitted way that one thinks the general level of well-being of disadvantaged people in the EU should be increased. The poll question on EU priorities has been posed by several Eurobarometer polls; however, the two sub-questions of interest to the measurement of social euroscepticism were to my knowledge only posed in the mid 1990s. It is the specific formulations of the chosen reply categories to the `priority' indicator that allow their usage in a thesis on euroscepticism: the replies call for more things to be done, thus arguably hinting at a current dissatisfaction with the EU's level of engagement. 181 Graph 15. Year: 1996 The figures EU priorities: More social Europe 0 20 40 60 80 100 Percent DK UK FR EU DK 70 58 UK 65 41 FR 84 64 EU 73 55 More aid to socially excluded in EU, 1996 (EB46) More attention to social justice, 1996 (EB46) There is some variation in replies to the two sub-questions. Nevertheless, the distribution of the three case countries in relation to the EU average is consistent. Not surprisingly, the amount of people who think the EU should give more help to socially excluded people is generally high. Denmark agrees with the EU average on both sub-questions. France has the most citizens who desire a stronger EU focus on social issues, while the United Kingdom has the least. The difference between the two countries is 19 and 23 percentage points respectively. French calls for a more social Europe are higher than the EU average in both sub-questions, while the exact opposite situation characterises the United Kingdom. 182 Indicator 18 EU CONSTITUTION. `What are all the reasons why you are opposed to the European Constitution?' (Spontaneous, multiple replies possible) Not enough social Europe and Economically speaking, the Constitution is too liberal Introduction Graph 16. Years: 2004 and 2005 The figures This indicator has already been introduced above (see Section 6.2 on sovereigntybased euroscepticism). It also includes two specific reply categories directly illustrative of social euroscepticism, namely the critique that the EU is too weak on the social dimension and the critique that the EU is too liberal (arguably, Eurobarometer could have merged the two reply categories into one). In the graph below, the thesis has added the scores of these two questions. The EU Constitution: 'Not social enough' and 'Too liberaľ 4 19 4 5 36 48 17 24 0 10 20 30 40 50 Not enough social Europe and Too liberal 2004 Not enough social Europe and Too liberal 2005 Percent DK UK FR EU The indicator reveals a sharp difference between France and Denmark, and, especially, France and the United Kingdom. In both 2004 and 2005, France is the only case country with a relative scepticism in this regard, while scepticism levels in the UK in both years is negligible. With 48 percent of the respondents found to be sceptics in 2005 in France, and five percent to be sceptics in the United Kingdom, there is a 43 percentage point difference between the two countries. The EU average in 2005 is 24 percent, which is somewhat in line with the situation in Denmark, where 19 percent found the Constitutional Treaty `not social enough' and/or `too liberal.' Taking both surveyed years into consideration, 183 scepticism is in absolute terms low in the EU as a whole and in Denmark, negligible in the UK and in the high-medium area in France. 184 Indicator 19 SOCIAL BENEFITS. `Some people may have fears about the building of Europe, the European Union. Here is a list of things which some people say they are afraid of. For each one, please tell me if you, personally, are currently afraid of it or noť: The loss of social benefits ­ `afraiď Introduction Graph 17. Years: 1997, 1999, 2000, 2004, 2005 (Note: uneven gaps between years) The figures This indicator illustrates whether co-operation in the EU brings about fears of a loss of, or deterioration of, social benefits, perhaps because integration is perceived to proceed according to a strongly market-based rationale. Questions surveying fears involved with co-operation have been frequently posed by Eurobarometer, but with varying question formulations. Surveys with a formulation that is useful for the measurement of social euroscepticism have been rather scattered and only prevalent from the mid-1990s onwards. Fear: loss of social benefits 0 10 20 30 40 50 60 70 80 Percent DK UK FR EU DK 46 54 52 49 44 UK 43 53 49 44 46 FR 61 67 65 64 66 EU 50 53 51 53 50 1997 EB47 1999 EB51 2000 EB54 2004 EB61 2005 EB63 There is little dynamism on this indicator across the five surveyed years. Taken as a whole, scepticism is markedly stable in all three case countries as well as on average in the EU. With scepticism consistently at the high level, the French are the most cautious of the three cases, and throughout all the surveyed years, more than ten percentage points distinguish them in this regard from the EU average. Denmark and the United Kingdom are either in line with, or less fearful, than the EU average. 185 In 1997 and in 2004, fewer British citizens than on average in the EU feared the continued existence of their social benefits, while the Danes were for the first time less afraid than average in 2005. Danish and British scepticism is within the highermedium level. 186 Indicator 20 PROPOSITIONS ON THE EU. `Do you tend to agree or tend to disagree with each of the following statements?' (multiple replies possible) `There should be closer co-operation between member states in social matters'. Agree Introduction Graph 18. Year: 2003 The figures This poll question measures whether, in the eyes of the citizens of the member states, there is scope for more EU co-operation on social matters. The question formulation does not specify what shape this co-operation should take, and it is thus a rather broad and `easy' indicator. It was posed as an either-or question alongside a number of other propositions on the EU in 2003. More cooperation on social matters 66 63 75 76 0 10 20 30 40 50 60 70 80 2003 EB59 Percent DK UK FR EU Many citizens, it seems, are supportive of more co-operation on social matters in the EU. There is, however, a significant difference in relative support on the issue across the three cases: France is in line with the EU average, while respectively ten and 13 percentage points less people in Denmark and the United Kingdom favour the proposition. 6.4.1 Summary Together, the four indicators measuring whether or not citizens find the EU too weak on the social dimension spanned a period from 1996 to 2005. Discounting the indicator on reasons for opposing the Constitution, average levels of scepticism in the EU are high. In the three case countries, however, they 187 range from `medium' to `very high', with the highest relative scepticism in France and the lowest level in the United Kingdom. On the indicator on the Constitution, scepticism because of too little social Europe was in 2005 negligible in the UK (two percent) but 12 percentage points above the EU average in France (27 percent). In fact, while it is not possible to speak of relative euroscepticism in either Denmark or the UK on any of the indicators, France is distinguished by relative scepticism on three of them, and often by more than ten percentage points. There is moreover an interesting contrast in the dynamics of opinion on the indicator allowing a nine-year time span--measuring fears of a loss of social benefits as a consequence of co-operation between 1996 and 2005--with a gradual decrease in scepticism of 12 percentage points in Denmark but virtually no movement in France. The result is a 22 percentage point difference between responses in the two countries. Once again, there is generally a large discrepancy between French and British opinions. For instance, 30 percentage points separated the French and the British when, in 1997, Eurobarometer asked if the Union should pay less attention to the economy and more to social justice. * France is eurosceptic * The United Kingdom exhibits less scepticism than the EU average, and its scepticism-level is negligible in one indicator (i.e. below five percent) * There is thus marked difference in attitudes between France and the UK. Indeed, the difference is never below ten percentage points, and even supersedes forty percentage points on one indicator * The level of scepticism is stable on the one indicator allowing a long time perspective 188 6.5 The level of hard euroscepticism Finally, this section accounts for the level of hard euroscepticism. As with the above sections, it begins by illustrating and discussing each indicator and concludes with a summary. Indicator A SUPPORT FOR EUROPEAN UNION MEMBERSHIP. `Generally speaking, do you think that (OUR COUNTRY'S) membership of the European Union is ... ?' A bad thing Introduction Graph 19. 1974 ­ 2005 (Every sixth indicator illustrated. Note: first year with uneven gap, as not all early Eurobarometers are accessible) Denmark and the United Kingdom This is the classic Eurobarometer question. It is often used in isolation as the dependent variable in studies of the causal power of some factor on euroscepticism. It is moreover the only Eurobarometer question that has been posed by every single poll since 1974. It surveys general attitudes to the very idea of membership of the European Union. EU membership: A bad thing 31 25 31 31 29 18 26 25 24 16 16 39 34 48 30 29 13 22 26 24 25 28 5 8 11 6 7 7 13 13 14 12 1514 12 17 12 11 7 13 15 14 11 17 0 10 20 30 40 50 60 1974 EB1 1979 EB11 1981 EB15 1985 EB23 1988 EB29 1991 EB35 1994 EB41 1997 EB47 2000 EB53 2003 EB59 2005 EB64 DK UK FR EU As Denmark and the United Kingdom both joined the EC in 1973, Danish and British figures between 1974 and 1985 should be observed with caution, as the thesis holds that it takes a while for public opinion in newly acceded member 189 France countries to stabilise. It is likely that it takes time to adjust to membership. It can be seen from the graph that both Denmark and the UK in the first ten years of the surveyed period were sharply critical of their new acquaintance. In general, both cases have been relatively hard sceptics on this indicator, although a drop in the negative opinions of both countries can be observed from 1981 to today. Interestingly, both countries reached a low in hard euroscepticism in 1991, a year characterised by the Maastricht negotiations. Danish opposition continued to be distinguished from the EU average in a sceptical direction until 2003. Generally, Danish hard euroscepticism has been at its lowest ever in the 2000s. In fact, it was for the first time ever surpassed by the EU average in 2005, although this was merely by one percentage point. British hard euroscepticism continues to be distinguished from the EU average. From the mid-1970s to the late 1980s, French hard euroscepticism has been less pronounced than the EU average. From 1991 onwards, it has been in line with the average. 190 Indicator B EU CONSTIUTION. `What are all the reasons why you are opposed to the European Constitution?' Against Europe/European construction/European integration Introduction Graph 20. Years: 2004 and 2005 The figures All three formulations within this reply category indicate hard euroscepticism towards the idea of European co-operation. The EU-Constitution: Against because of rejection of the EU 31 19 34 25 19 9 22 12 0 5 10 15 20 25 30 35 40 2004 2005 Percent DK UK FR EU Both the United Kingdom and Denmark are consistently differentiated by their intensity of hard euroscepticism on this indicator: the UK by an average of 13 percentage points and Denmark by an average of eight percentage points. France is not distinguished from the average in either year. Although there is some fluctuation across the two polls, scepticism in France and on average in the EU is consistently low. In Denmark, it varies between the low and medium level of euroscepticism, while it is consistently within the latter level in the United Kingdom. 191 Indicator C REJECTION. `Does the European Union give you personally the feeling of...?' (multiple replies possible) Rejecting it Introduction Graph 21. Years: 2002 ­ 2005. (Note: uneven gaps between years) The figures General attitudes towards the EU are sought by this indicator, which has been asked regularly by Eurobarometer in the 2000s, albeit not by every survey. Feeling EU inspires: Rejecting it 4 3 3 4 9 12 10 12 4 4 4 4 5 6 5 6 0 2 4 6 8 10 12 14 2002 EB57 2003 EB60 2004 EB62 2005 EB63 Percent DK UK FR EU Four polls, all in all covering a three-year period, are traced in the above graph. It shows British hard scepticism being significantly more sceptical than the EU average in all of the surveyed years, while Danish and French hard scepticism is not. 192 Indicator D A REFERENDUM ON MEMBERSHIP. `If there were a referendum tomorrow, asking whether (OUR COUNTRY) should stay in the European Union or leave the European Union, how would you vote?' Leave the European Union Introduction Graph 22. Year: 1996 The figures This indicator is a straightforward measure of hard euroscepticism. It aims to simulate a referendum situation allowing citizens to decide upon their country's continued adherence to the European Union. It was posed in 1996. As this is only shortly after Austria, Finland and Sweden became members, the thesis has taken the average for the EU-12. EU-Referendum: 'Leave the EU' 30 30 15 14 0 5 10 15 20 25 30 35 1996 EB45 Percent DK UK FR EU According to this indicator, 30 percent of the population in Denmark and Britain would have wished to leave the EU if there were a referendum on membership in 1996--that is 16 percent more than on average in the EU. France is not distinguished from the average. In absolute terms, there is medium-level hard euroscepticism in Denmark and the UK, and low-level hard euroscepticism in France and on average in the EU. 6.5.1 Summary As mentioned in Chapter 3, the hard manifestation of euroscepticism is distinguished by its intensity, and overall levels were expected to be low. The figures from the four indicators confirmed this expectation. Indeed, all the indicators suggested a low average level of hard euroscepticism in the EU. 193 On the indicator allowing for a long-term perspective on the development of hard euroscepticism (indicator A, attitudes towards membership), an initially high level of relative hard euroscepticism in Denmark has over the years decreased. In absolute terms it went from 31 percent to 16 percent, while a low level of hard euroscepticism in France has generally been climbing, in line with the EU average, from five percent who found membership a `bad thing' in 1974 to 15 percent who shared that opinion in 2005. It seems fair to say, however, that hard euroscepticism is still weaker in France than in Denmark. Until today, France has never differentiated itself by a relatively strong hard euroscepticism, while Denmark continues to be differentiated on the indicator on the Constitutional Treaty, which measures the amount of people opposing the document because they do not `support the principle of the EU.' Only the United Kingdom is distinguished by a relatively strong level of hard euroscepticism on all four indicators. 12 percentage points more Britons than average find membership of the EU a `bad thing.' 13 percentage points more than average oppose the Constitution because they do `not support the principle of the EU.' 10 percentage points more than average would have voted in favour of leaving the EU if there was a referendum on membership in 1996 (this number would have been 21 percent if Northern Ireland was discounted), while `rejecting iť came to the mind of six percentage points more Britons than average in the EU when they thought about the Union. * France is not characterised by hard scepticism on any of the indicators, however, its level of scepticism is increasing * The United Kingdom reveals hard scepticism on all indicators * Both Denmark, the UK and the EU average reached a low-point in levels of hard euroscepticism during the Maastricht negotiations; French hard euroscepticism, however, appears to begin its general increase in this year * There is considerable change in opinion over time 194 6.6 Degree of hard euroscepticism within the four types Having established the statistical relevance of the four types of euroscepticism, as well as of the indicators measuring the prevalence of hard euroscepticism, what remains to be examined is the degree to which each of the four types of euroscepticism is `harď--that is, if hard euroscepticism is equally strong within all four types, or if one type is more representative of hard euroscepticism than the rest. As Chapter 5 stipulated, this examination is carried out by cross-tabulating the indicator of support for EU membership with an indicator from each of the types. * Economic euroscepticism GAMMA (EB60) Denmark France UK EU No benefit ­ Membership bad thing .980** .940** .970** .970** ** p<0.00 There is an exceptionally strong association between the two indicators of EU benefit and membership in all three case countries as well as on average in the EU. This means that citizens who do not perceive to benefit from their country's membership of the EU are also likely not to find membership itself to be a good thing. This finding, however, should not cloud the fact that the two poll questions produce very different absolute figures. While an EU average of 33 percent in 2006 shared the opinion that they had not benefited from their membership, considerably fewer, namely 17 percent, shared the opinion that membership in itself was a bad thing. It should moreover be noted that both questions are formulated in an abstract way. There is, for example, no specification of what kind of benefit one does not perceive of. It is therefore relevant to check whether the gamma values for more specific indicators of economic euroscepticism and hard euroscepticism confirm the strength of the association. In EB60, two other indicators of economic euroscepticism were polled (indicator 2: waste of money; and indicator 4: economic role), as was another indicator of hard euroscepticism, namely the feeling of rejection towards the EU. As expected, results generally confirm a strong association, however, one that is considerably weaker than the one identified with the more abstract indicators. Gamma values are as follows: 195 GAMMA Denmark France UK EU EU rejection ­ Waste of money .750** .590** .780** .790** EU rejection ­ Negative economic role .780** .690** .600** .740** ** p<0.00 The figures underline the importance of keeping survey question formulation in mind when making inferences about the prevalence of hard euroscepticism. In the case of the economic type of euroscepticism, the prevalence of hard euroscepticism is clear, but mostly so with regard to abstract indicators. * Sovereignty-based euroscepticism GAMMA (EB60) Denmark France UK EU No EU Government ­ Membership bad thing .940** .910** .910** .920** ** p<0.00 As with regard to economic euroscepticism, there is a very strong association in all case countries between the opposition to an EU government and the perception that EU membership is a `bad thing'. Below, the contingency table for the EU as a whole is included as an illustration. Table 9: Contingency table ­ No EU Government * Membership bad thing (EU-15, 2003) Membership? (Gamma = .920) `Not a bad thing' `Bad thing' Total Count 7264 596 7860`Yes' % within Membership 88% 25% 74% Count 944 1788 2732 An EU Government? `No' % within Membership 12% 75% 26% Count 8208 2384 10592Total % within Membership 100% 100,0% 100,0% 196 It is clear from Column 2, row 2 that many more of the respondents (than what would have been expected from a random association), who hold the opinion that membership is a bad thing, are also against the creation of an EU government. * Democratic euroscepticism GAMMA (EB60) Denmark France UK EU The EP is not able to protect interests ­ Membership bad thing .670** .880** .760** .740** ** p<0.00 There is a strong association between democratic euroscepticism and hard euroscepticism, albeit the association is less strong than with regard to the two above types. We may moreover note large differences across the three surveyed countries: the association is considerably stronger in France (gamma = .880) than in Denmark (gamma = .670). This means that dissatisfaction with EU democracy to a larger extent is independent of the feeling that membership is a bad thing amongst Danes than amongst the French. * Social euroscepticism GAMMA (EB61) Denmark France UK EU Fear of losing social benefits ­ Membership bad thing .790** .560** .750** .590** ** p<0.00 As the above gamma values show, there is also an association between social euroscepticism and hard euroscepticism, but we may note that it is generally the weakest of the four types. Again, we are able to note large-scale country variation. However, contrary to democratic euroscepticism, we find the strongest association between social euroscepticism and hard euroscepticism in Denmark (gamma = .790), and the weakest association in France (gamma = .560, which is below the .600 limit introduced by the thesis as threshold for a significant association). Fears about the future of social benefits in the EU thus exist somewhat independently of the criticism of membership of the Union itself in France (and on average in the EU). 197 6.6.1 Discussion To summarise, these figures, all from the 2000s, show that it is especially within the economic and sovereignty-based types of euroscepticism that we find citizens wishing to do away with the EU. There is a less strong, but still consistently significant association (gamma above .600) between democratic scepticism and hard scepticism. The prevalence of hard euroscepticism is weakest within the type of social euroscepticism--however, we may note that the wish to do away with co-operation in the EU exists in conjunction with all types of euroscepticism. With regard to the three case countries, we ca observe that Denmark and the United Kingdom have a significant level of hard euroscepticism within all the four types, while the level in France, and on average in the EU, falls below the .600 gamma value with respect to social euroscepticism. Why is hard euroscepticism more prevalent within the types of economic and sovereignty-based euroscepticism than within the democratic and social types of euroscepticism? The thesis interprets this finding as underlining the relevance of the conceptual distinction introduced in Chapter 4 between horizontal and vertical euroscepticism. We may recall that vertical euroscepticism denotes citizens who essentially evaluate the EU as something external, something which needs not be an integral part of political life, while horizontal euroscepticism takes as its point of departure that the Union is a permanent and natural extension of the political system. The thesis argues that sovereignty-based scepticism, and to some extent economic scepticism,129 represent more vertical EU attitudes than democratic and social euroscepticism. Indeed, if, to citizens sharing a predominantly economic logic with regard to the integration process, co-operation once and for all is not seen as a beneficial enterprise--perhaps if one is a fisherman negatively affected by EU fishing quotas--this is likely to remove the basic desire for co-operation. Likewise, citizens who are very concerned about preserving national integrity are likely to be fundamentally opposed to developments perceived to enhance supranational co-operation in the Union. To citizens critical of the EU's level of democratic credentials, however, there is likely to remain a stronger perception that the situation may be realistically rectifiable (by more democratisation), following which scepticism would diminish. Moreover, this type of scepticism is likely to only develop following the evaluation that the EU is a justified political entity, which is in fact worthy of sharing democratic standards otherwise only expected of modern European nation-states. It is not surprising that it is the social type of euroscepticism that seems to represent the weakest level of hard euroscepticism. Indeed, this is horizontal euroscepticism per se, as citizens' 129 Chapter 7 resumes the discussion of whether indicators of economic scepticism are vertical or horizontal. 198 political value orientations form the basis of social euroscepticism. The discussion of these findings is pursued in Chapter 7. It should be noted with regard to inferences about the degree to which hard euroscepticism is prevalent within the four types of euroscepticism that more purposeful survey questions and/or targeted focus groups would be able illuminate to a greater extent the nature of the relationship. The above statistics are able to show us that indicators of hard euroscepticism do correlate with sceptical replies to indicators of the four specific types of euroscepticism, and that the correlation is stronger within some types (and member states) than within others. However, individual survey questions could be designed to poll more accurately when a particular scepticism is `strong enough' to constitute a de facto wish to withdraw from the Union. Indeed, although very few among the employed indicators come close to capturing this on their own, what is required is in fact only a substantial elaboration of the indicators of hard euroscepticism: whether membership, for instance, is considered a bad thing because of the perception of undemocratic structures. 6.7 Comparative findings This section does three things. First, on the basis of the above findings--the tests of coherence, independence and the logical consistency of the intra-type indicators--it presents the final conceptualisation of euroscepticism, a schematised overview of relative scepticism in Denmark, France and the United Kingdom, as well as an overview of the development of euroscepticism over time. Second, for reasons of further clarification, it offers a brief `walkthrough' of euroscepticism country by country, type by type. Finally, the thesis builds on the above comparative and longitudinal analysis to respond to some specific issues about public euroscepticism that have characterised and confused contemporary debates. The thesis' data, for example, allow clarification as to whether or not public euroscepticism is a recent phenomenon: if we can adequately speak today of a need to reconnect citizens to the EU, as well as to whether or not, given the prolonged existence of the EU, there is evidence that euroscepticism has come to resemble `normal political contestation'--if concerns about national sovereignty, and membership itself, have ceded to other more political concerns. Public euroscepticism, this chapter has demonstrated, comes in four broad types: citizens may base their scepticism toward the EU-of-the-day on economic, sovereignty-based, democratic and/or social grounds. Each of these types of scepticism can assume a hard character, meaning that there may be a 199 200 significant correlation between the particular substantive critique and the desire to withdraw from the European Union. The chapter also analysed the level of `hard euroscepticism' in absolute terms. Below, these findings are referred to as the `level of hard euroscepticism' and should not be confused with a substantive type of scepticism. Importantly, the Eurobarometer indicators of the level of hard euroscepticism do not in themselves allow the reader to make inferences about its nature. 6.7.1 Euroscepticism, an overview The complete three-level structure of euroscepticism outlined in Chapter 2 is presented below: OR OR 201 Figure 7. The concept of euroscepticism Indicator level Secondary level Basic level OR OR No benefit from membership EU is a waste of money Negative role on economic situation Against an EU government Integration is threat to national identity Satisfaction with EU democracy The EU is not democratic Utilitarian Sovereignty-based Democratic Euroscepticism Social Bad EU effect in specific areas Against CT because not democratic Against CT because not social enough Fear of loss of social benefits in EU Closer social co-operation Against Constitutional Treaty (CT) because of loss of sovereignty Only national-level decision-making The EP does not protect citizens More social emphasis in the EU Legend: : Ontology OR is the logical term for family resemblance Builds on Goertz 2005. : Substitutability Table 10 shows a crude overview of relative euroscepticism in Denmark, France, and the United Kingdom. It is interesting to note that the United Kingdom, despite its reputation, is not the most eurosceptic within all types. The UK is, however, the only of the cases where we find a relatively high level of hard euroscepticism. Table 10: Overview of relative euroscepticism Relative euroscepticism? Economic Sovereignty Democracy Social Intensity level (hard) Denmark No Yes Mixed No Mixed France Mixed No Mixed Yes Low (Not significant with regard to social euroscepticism) United Kingdom Mixed Yes No No High Yes/No means that all four indicators show euroscepticism/not euroscepticism; Mixed means that while one or two indicators show scepticism/not scepticism, the other are neutral (latest year polled). This simplified table shows that euroscepticism differs markedly across the three member states. All four types of scepticism are empirically relevant, but not in the same country. We may note that no type is relevant in all the case countries. This behaviour of the data measuring the prevalent types of euroscepticism allows the thesis to refute the four grounds for falsification set out in Chapter 5. First, the above investigation confirmed the statistical relevance of economic, sovereignty-based, democratic and social euroscepticism: Where measurable, intra-type gamma values were high, and lower than inter-type gamma values. Ideally, there would have been an even lower co-variance across the types--at least, as discussed in the above sections, the presence of overlap suggests that it would be fruitful in future studies to investigate whether part of the co-variance is in fact due to the existence of a causal relationship amongst the various motivations making up the four types of scepticism. Second, the intra-type indicators behave in a logical and uniform manner--that is, in the case of a mixed type in a country, where two of the four indicators show euroscepticism, the not-sceptical indicators are neutral as opposed to distinguished by a relative lack of scepticism. Third, each of the types of euroscepticism is relevant for at least one case country. It arguably further strengthens the relevance of the conceptualisation (and more precisely the case selection process) that no type is simultaneously present in all three case countries. Finally, as the thesis departed from an expectation that euroscepticism was manifest in each of the case countries, it is 202 relevant to note that both Denmark, France and the United Kingdom are characterised by the presence of at least one strong type of euroscepticism. Half of the total of 20 indicators examined by the thesis in this chapter allowed the tracing of euroscepticism over a longer time period. These indicators are able to give us an indication of trends in the development of scepticism.130 As the below table summarises, there is no stable picture. The EU average, in itself, is the most stable across the different types as well as with regard to the level of hard euroscepticism, witnessing no increase or decrease larger than ten percentage points.131 Danish and British scepticism have interestingly decreased on six out of the ten indicators, and in the remaining four, opinions have not fluctuated by more than five percentage points. French scepticism has somewhat decreased as regards the democratic deficit, but increased as regards economic euroscepticism. The level of hard euroscepticism in France is increasing; in Denmark and in the United Kingdom it is decreasing. 130 It should be recalled that because of the inconsistency of the Eurobarometer data material, the indication of a fall or rise in euroscepticism across the types has no fixed point of beginning or end. To give an example, when speaking of a decrease in democratic euroscepticism with regard to the indicator of `satisfaction with EU democracy', the frame of reference is the mid-1990s to today. When speaking of an overall increase in economic euroscepticism (with reference to the indicator of `benefit from the EU'), the frame of reference is the mid-1980s to today. 131 This seems to corroborate ideas by Andersen 2002; Page and Shapiro 1992; and Fishkin 1995 that random fluctuations in individual views (here the populations of the various member states) tend to `cancel each other ouť; see Section 1.3. 203 Table 11: Overview of the development of euroscepticism over time Type of scepticism Indicator EU DK FR UK Time span 1 (No benefit) 23 years 2 (Waste of money) - - - - 3 years 3 (Bad effect, specific economic areas) 1983 & 2005 Utility 4 (Economic situation: negative role) n/a n/a n/a n/a n/a 5 (EU government: no support) - - - - 10 years 6 (Fear of losing national sovereignty) n/a n/a n/a n/a n/a 7 (Perceived identity threat) n/a n/a n/a n/a n/a Sovereignty 8 (Against joint decision-making) - - 2 years *) 9 (Dissatisfaction with EU democracy) 11 years 10 (Constitution is not democratic) n/a n/a n/a n/a n/a 11 (EU is not democratic) n/a n/a n/a n/a n/a Democracy 12 (The EP does not protect interests) - - - 7 years 13 (Should be more social justice) n/a n/a n/a n/a n/a 14 (Constitution not social enough) n/a n/a n/a n/a n/a 15 (Fear of losing social benefits) - 8 years Social 16 (More social policy co-operation) n/a n/a n/a n/a n/a A (Membership is a bad thing) - 31 years B (Against European integration) n/a n/a n/a n/a n/a C (Feeling of rejection toward EU) - - - - 3 years Intensity level D (Want to leave the EU) n/a n/a n/a n/a n/a : Decrease : Decrease by more than ten percentage points : Increase : Increase by more than ten percentage points - : Overall stability fluctuation does not exceed four percentage points) n/a : The data does not allow a longitudinal perspective *) In Denmark, this decrease occurred on five out of the ten surveyed policy areas. In the United Kingdom, it occurred with regard to three of the policy areas 204 Section 6.7.2 now reviews the findings from the data analyses, country by country and type by type, before the thesis turns to discuss a number of perspectives about the possible underpinnings of the trends revealed in the two above tables. 6.7.2 Euroscepticism walkthrough * The EU as a whole Economic euroscepticism. Economic euroscepticism has generally increased since Eurobarometer started polling perceptions of EU benefit in 1983. Indeed, the lowest figure of scepticism on this indicator in the years surveyed in Chapter 6 is precisely from 1983 (25 percent). In absolute terms, however, scepticism has remained close to (and never higher than) the medium level on all the indicators, thus between 25 and 49 percent. Sovereignty-based euroscepticism. Sovereignty-based euroscepticism is also stable on the long-term indicators. In absolute terms, it is generally within the lower medium level--only on the question surveying attitudes to joint decision-making on a wide number of policy areas does scepticism within two areas, education and health and social welfare, fall within the high level (50 to 74 percent). Democratic euroscepticism. The intensity of the critique of the democratic deficit varies from indicator to indicator. On the two `yes-no' questions--whether the European Constitutional Treaty was opposed because it was not democratic enough, and whether the entire co-operation in itself is perceived as undemocratic--there is a low scepticism level (five to 24 percent). Meanwhile, on the two multiple reply questions--satisfaction with EU democracy and the European Parliamenťs ability to protect interests--scepticism is within the medium level. It appears to be slightly decreasing over a ten-year period. Social euroscepticism. There seems to be a high potential for the type of euroscepticism, which criticises the EU's level of social engagement. However, one has to recognise from the discussion in Part Two that this is also the type of scepticism which is most similar to the types of contestation characterising the nation-states (where consensus around political decisions are rarely expected across the political spectrum). Generally, a high and rather stable number of EU citizens believe the EU should do more to guarantee social justice and equality. We may nevertheless note that only 13 percent would have rejected the Constitution on the grounds that this specific document did too little to promote a social Europe. 205 The EU as a whole: The level of hard euroscepticism. Hard euroscepticism is low on all indicators throughout the surveyed years. It is stable on both indicators allowing a longer time perspective. In general, there is strongest correlation in the three surveyed countries between hard euroscepticism and economic and sovereignty-based euroscepticism. There is generally least hard euroscepticism with respect to social euroscepticism. The below graph illustrates these dynamics of euroscepticism, using the longest polled indicator for each type or euroscepticism, as well as for the level of hard euroscepticism.132 These are: * Economic euroscepticism: Feeling of benefit from membership (1983 ­ 2005) * Sovereignty-based euroscepticism: Support for an EU government (1987 ­ 1996) * Democratic euroscepticism: Satisfaction with EU democracy (1994 ­ 2005) * Social euroscepticism: Fear of losing social benefits from integration (1999 ­ 2005) * Level of hard euroscepticism: Attitude towards membership (1974 ­ 2005) It should thus be borne in mind that the graph only reflects the development of one indicator for each type. Since the indicators have divergent intensity level, it is not possible to infer from the graph about the positioning of the types vis--vis each other. 132 The questions were polled with uneven gaps, which is why the thesis has grouped some years in the graph--and why the line for some of the types is broken between certain years. 206 Graph 23: Dynamics of Euroscepticism (EU) European Union 0 10 20 30 40 50 60 70 80 1974 EB11975 EB4 1979 EB11 1981 EB15 1983 EB19 1984/1985 1986 1987/1988 1989 1990/1991 1992/1993 1993/1994 1995/1996 1996/1997 1998/1999 1999/2000 2000/2001 2001 EB56 2002 EB58 2003 EB59 2004 2005 Percent Economic Sovereignty-based Democratic Social Level of hard euroscepticism * Denmark Denmark is strongly sceptical--all four indicators showing relative scepticism--in the sovereignty type of euroscepticism. This scepticism is generally stable over time, albeit there has been a slight decrease over the surveyed years in the percentage of Danes wishing decision-making to remain at the national level on a number of policy areas (see Section 6.2). Denmark is to some extent characterised by democratic euroscepticism; however, this type of scepticism has slightly decreased over time. There is no relative euroscepticism in Denmark concerning the lack of a social EU, or concerning the lack of economic utility of co-operation. Where measurable, the level of hard euroscepticism appears to be decreasing in Denmark. However, on two indicators it still remains higher than the EU average. It is particularly within the economic and sovereignty-based types of euroscepticism that strong Danish desires to withdraw from EU co-operation can be found.133 This desire is weakest (albeit still significant) within the type of 133 This is no contradiction to the finding that there is a low level of economic euroscepticism in Denmark: What it suggests is that within the (relatively low) percentage of Danes who criticise the EU's benefit, there are a high number of citizens who do not find membership itself a good thing. 207 democratic euroscepticism. This means that criticism of the Union's democratic credentials is least likely to constitute hard euroscepticism in Denmark. As with regard to Graph 23 above, the below graph illustrates the development of euroscepticism using the longest polled indicator for each type. Graph 24: Dynamics of Euroscepticism (DK) Denmark 0 10 20 30 40 50 60 70 80 1974 EB11975 EB4 1979 EB11 1981 EB15 1983 EB19 1984/1985 1986 1987/1988 1989 1990/1991 1992/1993 1993/1994 1995/1996 1996/1997 1998/1999 1999/2000 2000/2001 2001 EB56 2002 EB58 2003 EB59 2004 2005 Percent Economic Sovereignty-based Democratic Social Level of hard euroscepticism * France France shows strong absolute and relative scepticism within one euroscepticism type, namely the scepticism towards the lack of a social EU. The French are moreover relatively sceptical on two out of the four indicators measuring democratic euroscepticism and scepticism about the EU's economic utility. Within the former type, scepticism has, however, somewhat decreased over time, while within the latter type, it has markedly increased. Sovereignty-based euroscepticism is stable and continues to be low in France. Although France is not more sceptical than the EU average as regards the level of hard euroscepticism, there has nonetheless been an increase in this level. It is especially relevant to speak of hard euroscepticism within the economic type of euroscepticism, whereas the otherwise more prominent type of social euroscepticism correlates considerably less well with desires to withdraw from the Union (see Section 6.6). 208 Graph 25: Dynamics of Euroscepticism (France) France 0 10 20 30 40 50 60 70 80 1974 EB11975 EB4 1979 EB11 1981 EB15 1983 EB19 1984/1985 1986 1987/1988 1989 1990/1991 1992/1993 1993/1994 1995/1996 1996/1997 1998/1999 1999/2000 2000/2001 2001 EB56 2002 EB58 2003 EB59 2004 2005 Percent Economic Sovereignty-based Democratic Social Level of hard euroscepticism * United Kingdom The United Kingdom is strongly sceptical--all four indicators showing relative scepticism--on sovereignty grounds, and well as with regard to the level of hard euroscepticism. There is moreover some relative scepticism in the UK on the indicators of the economic type. There is less scepticism than on average in the EU towards the alleged democratic deficit and the lack of social credentials in the integration process. Moreover, Danish and French levels of scepticism within, respectively, the sovereignty-based and the economic type of euroscepticism, are often higher than British levels of scepticism. As mentioned above, these findings are noteworthy to keep in mind before describing the UK as the undoubtedly most eurosceptic member state. It does in fact seem plausible that this observation (and thus the UK's eurosceptic reputation) is in part due to a conflation of `don't know' replies and sceptical replies. True, studies of EU support reveal low levels in the UK, but this is often due to a high number of people who have not made up their minds, and not only due to high levels of scepticism. Recalling the definition of the concept and non-concept of euroscepticism in Chapter 3, `not knowing' does not equal euroscepticism. Certainly, however, the fact that the scepticism that does 209 prevail within the four substantive types appears to breed a considerable degree of hard euroscepticism is likely to constitute another important component of the UK's eurosceptical reputation. Hard scepticism in the United Kingdom is strongly present with regard to all four substantial types of euroscepticism; however, overall levels have decreased over a 30-year period. It is interesting to note from the graph below that the United Kingdom is the only case where the four types of euroscepticism seem to coalesce around the approximate same absolute level--at least compared to the situation in Denmark, France and on average in the EU. As Chapter 5's tests allow us to rule out the suspicion that it to a high degree are the same eurosceptical respondents who give the sceptical replies to the four types, this occurrence of consistency in scepticism-levels is probably coincidental. Graph 26: Dynamics of Euroscepticism (UK) United Kingdom 0 10 20 30 40 50 60 70 80 1974 EB11975 EB4 1979 EB11 1981 EB15 1983 EB19 1984/1985 1986 1987/1988 1989 1990/1991 1992/1993 1993/1994 1995/1996 1996/1997 1998/1999 1999/2000 2000/2001 2001 EB56 2002 EB58 2003 EB59 2004 2005 Percent Economic Sovereignty-based Democratic Social Level of hard euroscepticism Again, more adequate data material could have brought more detailed conclusions with regard to the strength and coherence of the individual types of euroscepticism, as well as with regard to the degree of co-variance amongst the types. If more direct indicators, for instance, had coincided within one survey, it would have been possible to run a factor analysis to see if the indicators of the respective types 210 loaded on factors in accordance with the thesis' theoretically-derived expectations. This would have further strengthened the argument of statistical independence of the types. 6.7.3 Analysing the development of euroscepticism We saw from Table 10 above that whilst the EU average as a whole has witnessed rather small variation in the level of scepticism over time (a less than ten percentage points increase in economic scepticism, and a less than ten percentage points decrease in democratic and social scepticism), the individual member states have witnessed larger-scale variation. These developments have occurred over a long enough time period to make it reasonable to argue that the resulting patterns are not merely ascribable to particular economic reforms or down periods, nor to the popularity of specific governments (we may recall a formerly mentioned-- contested--debate in the literature as to whether or not periods of domestic economic affluence, or incumbent support, affect eurosceptic sentiments). The present section builds on Chapter 4's discussion of existing theories of euroscepticism, and the present chapter's analyses, in order to explain this, and other, trends discernable from the overall dynamics of soft and hard euroscepticism. The thesis' analyses demonstrate that over the past thirty years there has been a rather sharp decrease in Danish economic euroscepticism combined with a decrease in the level of hard euroscepticism, while in France there has been the reverse development: a consistent increase in economic euroscepticism combined with an increase in the level of hard euroscepticism. As regards sovereignty-based, democratic and social euroscepticism, these types have remained largely stable, if not witnessing a small decrease in especially the Danish case. Revisiting Chapter 4's discussion of EU support and legitimacy, it seems plausible to suggest that this development supports the `Eastonian'-inspired argument on the relationship between vertical and horizontal attitudes. Vertical euroscepticism, to reiterate from Section 4.2, departs from a `them and us' perception of the EU, perceiving the Union as something exterior to the national political system, while horizontal euroscepticism approaches the EU as an integral and permanent part of political life, which is evaluated on the basis of day-to-day politics. In short, system support or rejection (here polled through the prevalence of hard euroscepticism) generally builds on vertical attitudes, which are emotive and rather resistant to change, while support for day-to-day policies (here represented by the economic type of scepticism) to a larger extent builds on horizontal, and more specific, attitudes, which are susceptible to short-term variations, depending, for instance, on the economic situation. However, if specific attitudes are strongly sceptical over a long 211 time period, this may eventually decrease vertical support, and reverse. Citizens, the argument holds, gradually grow disillusioned with the overall political system if they over a longer time period have been consistently dissatisfied with specific policy elements. We may speculate whether this development explains the French and Danish cases.134 A founding member of the Union, France's membership has for decades almost been taken for granted.135 This seems to correspond to the perspective of the general public. The Eurobarometer data scrutinised in Section 6.6 above reveals a low and stable level of hard scepticism throughout the 1970s and 1980s, which indicates that co-operation in the EU as such was not questioned by the bulk of French public opinion. Moreover, a closer look at the individual indicators making up the types of scepticism showed that the more abstract, or vertical, the question wording, the lower the level of French sceptical replies. As regards the economic type of euroscepticism, which is of interest to the present section, French scepticism is lowest on the two more abstract indicators, `feeling of having benefited from membership' and `meaning of the EU: a waste of money', and highest on the two more specific indicators, `the EU's effect in specific economic areas' and `the EU's effect on the economic situation'. Over recent years, however, there has been a slow, but consistent, increase in more vertical scepticism in France. This is most directly witnessed by the level of hard euroscepticism. Here, French sceptical attitudes towards the membership question were fluctuating between four and 11 percent prior to 1992. After 1992, they have fluctuated between 12 and 15 percent--a new, higher level which was confirmed by the indicator of hard euroscepticism from 1996, where 14 percent of the French would have voted in favour of leaving the EU were there to be held a referendum on membership, and in the run up to the referendum on the Constitution in 2005, where 14 percent would have voted `no' on the grounds that they did not support the European Union. Moreover, even within the output oriented type of economic euroscepticism, there has on the indicators capturing vertical attitudes been an increase in scepticism over the same time period: prior to 1992, feelings of not having benefited from EU membership fluctuated between 21 and 28 percent in France. Following 1992, between 28 and 39 percent of the population have shared this feeling. It should be mentioned that while the feeling that the EU is a `waste of money' over the three years this question was polled increased by three 134 In the United Kingdom, it makes little sense to apply Easton's thesis, as the EU has consistently been highly contested at both `system' level and on most other specific indicators. 135 It was striking to note in connection with the planning of the referenda on the EU's Constitutional Treaty that in making predictions for what action the Union could take in the face of a `no' result, withdrawal was mentioned as a possible solution or consequence in the case of a British `no', while this option was generally not perceived off with regard to France. 212 percentage points in France, this represents a too small increase to be considered significant by the thesis. This increase in French scepticism on the vertical indicators of economic euroscepticism has importantly been occurring alongside a sharp increase in scepticism on the specific indicator on the EU's effect on specific economic areas, which is likely to capture more horizontal attitudes (the remaining specific indicator of economic scepticism does not allow a time perspective). Again recalling the data examination in Section 6.1, French sceptical perceptions in this regard rose by up to 28 percentage points between 1983 and 2005; leaving scepticism within the high level (the area marked by the steepest increase was `standard of living'. That it was not a unilateral perception that the EU was performing worse in this area in 2005 as compared to in 1983 is revealed by Danish sceptical opinions decreasing over the same time period by five percentage points). Thus, overall, France seems to have witnessed a rather steep increase in horizontal scepticism alongside a less steep, but consistent, increase in vertical scepticism. Danes' sceptical opinions towards the EU as such (hard euroscepticism) have decreased over the past years: since 2001, perceptions of EU membership as a `bad thing' have not been higher than 16 percent. Prior to 2000, they often surpassed 25 percent (curiously, an all-time `low' during this period-- 14 percent--came in 1992, the year of the Danish rejection of the Maastricht Treaty). It is important to note that contrary to levels of hard euroscepticism, economic scepticism in Denmark has in fact rarely been higher than the EU average throughout the more than 20-year time period surveyed by Eurobarometer. As pointed out above, it has in other words not been appropriate to speak of relative Danish euroscepticism in this regard. Notwithstanding this relatively low point of departure, scepticism on all three long-term indicators measuring economically-oriented EU attitudes has over the past years witnessed a slow but consistent decline. Since 1996, fewer than 25 percent of Danes feel that they have not benefited from membership (thus leaving scepticism in the low level). Impressions that the EU is a `waste of money' decreased by three percentage points between 2003 and 2005 (a decrease that is, however, too small to be considered significant by the thesis); while perceptions of a negative effect of the EU on specific economic areas dropped by up to nine percentage points between 1983 and 2005. The result is that contemporary Danish economic euroscepticism is low both in absolute and in relative terms. In summary, there is thus a long-term concomitant decrease of economic and hard euroscepticism in Denmark. 213 On the basis of these observations, it does seem plausible to argue that there is a relationship between the development of vertical and horizontal scepticism. More expressly, it does not seem far-fetched to suggest that this relationship is causal: periods with high horizontal scepticism may lead to growing vertical scepticism. In other words, the level of diffuse support in the `reservoir of EU attitudes' identified by Lindberg and Scheingold (1970; see also page 36 of this thesis), sinks if exposed to a long period with draught. A more detailed study involving a larger number of vertical and horizontal indicators, or qualitative analysis, would be required to adequately confirm this derived hypothesis. The observation of an opposite development as regards the levels of economic and hard euroscepticism in France and Denmark moreover adds a new perspective to the critique of Ronald Ingleharťs generational logic, referred to in Chapter 3 and 4 above. According to this line of thought, hard euroscepticism should eventually completely phase out. Instead, the thesis' analysis of the development of euroscepticism supports the argument that citizens do form independent, issue specific perceptions about whether or not the EU-of-the-day merits support or scepticism, which, given the long-term perspective used, do seem independent of generational change. While several sociodemographic studies and opinion polls suggest the continuous presence of some kind of euroscepticism amongst also younger age groups,136 the above analyses demonstrate on a macro-level that not only is hard euroscepticism not on the decrease in the EU--importantly, its development varies considerably from country to country, with a slight overall increase in the EU average between 1974 and 2006. Judging from the above analysis, it appears that the EU-of-today is more in line with what Danes want, than with what the French like. We may further speculate that this development is influenced by the marked difference that became manifest between Danish and French opinions towards the 2004 Eastern Enlargement (see Section 7.3 below). On a more fundamental level, contradictory patterns as regards the development of hard euroscepticism have consequences for the assumption that the EU is increasingly coming to resemble a political system with `normal political contestation'. Were citizens ever connected? Before turning to this discussion, we may note that a consequence of the thesis' demonstration of a long-term presence of euroscepticism is that it makes little sense today to speak of a general need to reconnect the EU to its citizens. Indeed, public euroscepticism in one form or the other has been 136 In the Spanish referendum on the Constitutional Treaty, for instance, the `no' percentage decreased with higher age; Flash Eurobarometer 168. See also a recent Danish poll (published March 2007) revealing considerable opposition amongst young people towards abolishing the Danish opt-outs: www.duf.dk. 214 powerful at least throughout the time span of the Eurobarometer polls. It is, however, relevant to note the idiosyncrasy of euroscepticism situations across different time periods. In other words, while it is inappropriate to speak of a general need to reconnect citizens, it is nonetheless a correct observation that the euroscepticism situation of the 2000s differs from the euroscepticism situation of, say, the 1980s. This change has occurred across all types of euroscepticism, as well as with regard to the level of hard euroscepticism. Is it at all possible to speak of a major rupture in euroscepticism throughout the history of the Eurobarometer? The thesis, we may recall from Chapter 3, operates with the requirement that survey dynamics have to be durable in order to be classifiable in terms of euroscepticism. This rationale was an important reason behind its longitudinal focus: a casual look at the development of scores to most Eurobarometer poll questions reveal frequent smaller fluctuations and rather few, but clear, long-term trends--a finding which underlines the risk of one-survey studies falling prey to spurious `odd year' deviations. The two questions surveying attitudes to EU membership and the feeling of benefit from the EU are relevant examples as they offer the longest possible overview of opinion (see graph illustrating 11 surveys with regard to the membership question page 189--and graph page 153 illustrating nine surveys with regard to the benefit question). It goes for both indicators that one year stands out as marking a genuine shift in euroscepticism, namely, as hinted at in the analysis of the data above, the year 1992. Including a Eurobarometer survey per year for every year the questions have been polled, the `lows' and `highs' perceiving (i) membership as a `bad thing' and (ii) `no benefit from the EU' before and after 1992 can be illustrated as follows in the three case countries: Table 12 (i and ii): Movement in perceptions of (i) EU membership as a `bad thing'; and (ii) `no benefit from the EU' before and after 1992 Between 1974 and 1991 Between 1992 and 2002137 (i) Membership Low High Average Low High Average Deve- lopment Denmark 18 35 27,3 14 31 21,5 France 4 11 6,7 9 19 13,4 UK 13 54 31,6 13 28 22,3 215 Between 1983 and 1991 Between 1992 and 2004(ii) Benefit Low High Average Low High Average Deve- lopment Denmark 21 34 29,0 16 32 22,8 France 18 29 23,7 25 40 32,3 UK 34 57 45,4 35 52 42,9 The tables show that in all three case countries, the development on the two indicators is consistent, albeit with a varying strength and prefix for the individual countries. In Denmark, the average perceiving membership as a bad thing has decreased by six percentage points, and generally reached lower `lows and highs'. In the United Kingdom, the average has decreased by nine percentage points, and while the low of 13 percent finding membership a bad thing is consistent across the two time periods, there has since 1992 not been more than 28 percent who share this opinion--the high prior to 1992 was 54 percent. On the contrary, in France, the average has increased by seven percentage points, and reached higher `lows and highs'. This trend is repeated for the benefit indicator, although here, France is witnessing the most marked development (the average across the two time periods perceiving `no benefiť has increased by nine percentage points; in the UK, the development is considerably less marked, namely about three percentage points decrease). Relating these figures to empirical events, 1992 is of course a year characterised by the ratification process of the Maastricht Treaty. It was the first year in the history of the European Communities that a member state (Denmark) voted `no' to an integrationist development in a public referendum, and the year where the French almost voted `no'. It appears that this strong unprecedented showing of euroscepticism `legaliseď expressions of dissatisfaction with the EU, as if more citizens came to reflect upon the possibility that EU co-operation might be lacking in certain regards. The Maastricht Treaty, moreover, was an obvious trigger: in the immediate aftermath of key events in European history it radically and visible reformed the underpinnings of co-operation. It does not, however, appear that the member state populations' reactions to the trigger are readily comparable: Judging from this chapter's analysis of the indicators of the four types of euroscepticism, scepticism in France has for instance assumed a social and economic character, while in Denmark sovereignty concerns have continued to be prominent. To return to the issue of whether it is possible to speak of a historical connection between the EU and its citizens, we may thus specify that the year 1992 might well mark a rupture for the French, 137 The most recent year is available in the Eurobarometer Interactive Search System. 216 but that on the contrary, it makes little sense to perceive of the Danes and the British as `less connecteď today than twenty years ago. Towards normal political contestation? In Part Two we saw that the rejection by French and Dutch voters of the Constitutional Treaty has spurred increasing scholarly focus on patterns of political contestation as a means to understanding and explaining public euroscepticism (Section 4.8). The assumption behind this approach is that--to the European publics--the EU is increasingly becoming a locus of normal political contestation, implying that `yes-no' debates cede to value-based partisan arguments. This development predicts a decreasing prevalence of hard euroscepticism, as well as of sovereignty-based concerns, as a result of the so-called `yes/no' debate losing its prominence. This decrease in prominence would possibly be followed by the increasing prevalence of democratic and social euroscepticism: if the distinction between the EU and the national level loses its significance, then, presumably, democratic shortcomings of the Union appear less tolerable, whilst the Union also would be a more prominent battleground when it comes to politicised issues of social regulation. To once again recall the conceptual distinction between vertical and horizontal euroscepticism, normal political contestation clearly assumes the dominance of the latter manifestation. We noted from Section 6.6.1 (see also Section 7.2 below) that horizontal attitudes characterise all indicators of democratic and social euroscepticism as well as, to some extent, indicators of the economic type of euroscepticism. That is, scepticism within these three types may depart from an evaluation of cooperation in the EU as an integral, and normal, part of everyday political life. This assumes that EU contestation takes place along other, or at least additional, lines than the oft-assumed more/less integration schism. The nature of these additional lines is debated in the literature. Simon Hix and Christopher Lord have suggested the salience of the classic left/right axis; Liesbet Hooghe et al. have perceived of a split between supporters of `regulated capitalism' and supporters of `neoliberalism'; adding in a later study that at least with regard to elite euroscepticism, entirely new lines of contestation--the so-called `new politics dimension'--may, in fact be `the most general and powerful predictor of party positioning on the issues that arise from European integration' (Hooghe, Marks and Wilson 2004: 121, 140). Of importance to the thesis' research questions is the evaluation of whether it is possible to conclude from the data analysis that public euroscepticism is increasingly resembling `normal political contestation'. Interestingly, there is little sign that this is the development characterising the bulk of sceptical public opinion today. This is most evidently demonstrated by `counterfactual evidence', namely the 217 findings that in the EU as a whole there is no marked movement in the saliency of hard euroscepticism, nor with regard to the level of sovereignty-based euroscepticism (we may recall that both the level of hard euroscepticism and the sovereignty type of euroscepticism reflect vertical attitudes). This indicates that the more/less integration axis continuous to be prominent amongst EU citizens, or at least that it is stable. Moreover, neither can we note an increase in prominence in the EU--on the few indicators allowing the polling of this--of democratic and social euroscepticism.138 This EU picture generally fits the euroscepticism situations of the case countries, although there is important cross-country variation. The situation in France virtually resembles that of the EU average. In Denmark and the United Kingdom, however, the development of euroscepticism is in this regard less clear. On the one hand, it suggests the increasing relevance of horizontal (or `normaľ) contestation about the EU at the `expense' of vertical contestation. At least, we can note a general decrease in Denmark and, to a lesser extent in the UK, of the level of vertical euroscepticism, which has been occurring alongside year-long intensive debates about the merits of EU membership, as well as in juxtaposition with crucial developments in the Union itself.139 As day-to-day euroscepticism has to do with more or less informed evaluations of the nature of the European Union (see definition in Chapter 3), it is indeed possible that this development has to do with a growing acceptance of the way in which the EU is currently moving.140 However, as sovereignty-based euroscepticism remains strong and stable, the more/less integration axis can not be ignored. 138 It is important to stress that this finding is limited to public opinion--it is certainly possible that the situation is different with regard to `organiseď contestation. 139 Lise Togeby noted a long-term trend of increasing support for EU membership in Denmark in 2004 (Togeby 2004: 83). 140 Given the continued strong prevalence of several types of euroscepticism in the United Kingdom, as well as the continued high level of hard euroscepticism--despite a decrease over time--it makes less sense to speak of horizontal EU attitudes. However, given the past years' significant decrease in absolute and relative scepticism levels, it will certainly be interesting to follow the British situation over the coming years. 218 Part Four: Consequences and Conclusions 7 Euroscepticism revisited The thesis has demonstrated the extent to which, and how, public euroscepticism is a multifaceted phenomenon, and the ways in which it varies from member state to member state. What people want (and do not want) from the EU is simply very different across populations. These characteristics of euroscepticism have consequences for the ambition to bridge the frequently declared EU-citizen gap. However, not only is euroscepticism multifaceted, Chapter 6 also pointed at potentially contradictory patterns of scepticism across the three case countries. These mean that not only do populations differ in terms of prevalent types of scepticism, but what one population is against, may be precisely what another population wants. In this concluding chapter, the thesis applies its typology of euroscepticism in order to close the `dependent variable problem', explain the reasons behind shortcomings in existing literature, and evaluate its implications. It does four main things: Section 7.1 first offers a review and a reconsideration of three largely empirical studies of euroscepticism/public EU support that were briefly introduced in Chapter 4. The aim is to demonstrate through the new conceptual perspectives gained by the thesis that the contributions of these studies are in fact considerably narrower and less comparable than what is being claimed. In a similar vein, Section 7.2 revisits three theoretical approaches to the EU's challenge of securing public support, namely David Beetham and Christopher Lorďs model of EU legitimacy, Jürgen Habermas' call for a `constitutional patriotism', and Joseph Weiler's assumptions as to what constitutes the `deep sense of malaise and public disaffection' that he refers to on several occasions (for instance Weiler 1999: 329). The section relies on the thesis' empirical examination of euroscepticism to explain why an application of these three approaches might run counter to the authors' expectations. The examples are intended to underline the importance for researchers of bridging the frequent divide, pointed to in the introduction to the thesis, between highly theoretical studies of the integration process on the one hand and highly empirical studies of public attitudes on the other. With these implications for the research field in mind, Section 7.3 continues on a more practical note by applying the developed typology of euroscepticism to three recent large-scale EU debates and developments, namely the reflections on the existence of broad `EU visions'; Eastern Enlargement of 2004; and the `Period of Reflection' to bringing the EU forward following the rejection of the 219 Constitutional Treaty by the French and the Dutch voters in 2005. Here, the aim is to show how the thesis' conceptualisation of euroscepticism contributes to a better understanding of the position of the EU's publics on these issues. Finally, Section 7.4 concludes the inquiry into euroscepticism by evaluating the thesis' overall approach, its strengths and weaknesses, and discussing a number of possible avenues for further research into euroscepticism. 7.1 Revisiting the dependent variable problem Having explicated the thesis' answer to the two first basic research questions set out in Chapter 1-- namely to what extent, and how, euroscepticism is a multifaceted phenomenon and how patterns of euroscepticism differ between member states and over time--this section now turns to the two remaining basic research questions, namely `Why has existent research failed to capture the nature and dynamics of euroscepticism?' and `What consequences does a refined understanding of euroscepticism have for the research field, and for the European integration process?'. In line with the rationale behind concept analysis (see Chapter 2), the thesis argues that a central theoretical contribution is particularly found in its enhancement of the cumulativeness of studies--something which the below paragraphs demonstrate can also be applied retrospectively. A motivation for this study was the puzzlingly consistent confusion, despite decades of literature, as to what public euroscepticism is and why it is--a confusion that on several occasion has produced contradictory findings as to these very `whať and `why' questions (see Parts One and Two for discussions). In Chapter 2, the thesis labelled this problem the dependent variable problem. Indeed, the modest contribution of otherwise sound studies with regard to explaining why euroscepticism emerges and develops (and who is eurosceptic) essentially resides in a neglected attention to the multifarious nature of the object being examined. The problem is serious, as it obstructs the cumulability of individual studies, and thus ultimately an actual theory of euroscepticism from developing and evolving. The thesis suggested a detailed theoretical and empirical conceptual analysis of euroscepticism as a remedy to the problem, and presented in Section 6.7 a schematised summary, and a more detailed `walkthrough', of the results of the thesis' investigation. In this section, the thesis seeks to demonstrate how these findings solve the aforementioned dependent variable problem, and show that this has consequences for both existing studies, as well as for future work. Indeed, while the thesis would advise future studies into public euroscepticism to be sensitive to the four independent types of scepticism 220 identified here, using perhaps only one or two of the types as dependent variables in a given study, it is possible to revisit the choice of dependent variables of previous studies using the thesis' typology as framework. This endeavour serves to narrow down and specify the most likely reach and nature of their contributions. While it is clearly beyond the scope of this section to revisit all the studies of euroscepticism introduced in Chapter 4 above, the thesis illustrates the argument with reference to three studies: one engaging solely the membership question as the dependent variable (Andersen and Reichert 1996), one relying on attitudes towards common decision-making as a proxy for euroscepticism (McLaren forthcoming 2007), and one measuring EU support with reference to satisfaction with EU democracy and support for an EU government (Rohrschneider 2002).141 On a general note it is clear that studies engaging a single dependent variable to account for a complex phenomenon are likely to capture only a particular manifestation of euroscepticism, while studies combining several indicators into one dependent variable run the risk of mixing or merging what are in fact distinct types of scepticism. I. Christopher J. Anderson and M. Shawn Reichert set out in their study from 1996 to explain `differences in public support for membership in the European Union across the twelve EU member states' (1996: 231). This, they argue, is important as `Europe is difficult to unify if there are significant differences in support for integration across the member states of the European Union' (ibid: 232; note the lack of discussion of the difference between study object (support for membership) and justification (support for integration)). We saw in Chapter 2 that Andersen and Reichert seemingly go against two common conclusions in the literature on euroscepticism by stating (i) that age and gender are no consistent determinants of EU support (ibid: 241); and (ii) that there `may be a negative relationship between postmaterialism and EU supporť (ibid: 235)--findings which have evoked considerable debate in the scholarly community. However, more careful attention to the authors' dependent variable reveals that, in fact, their sole focus is on the issue of membership as a measure for pro- or anti-EU attitudes.142 Importantly, this narrows down their study's applicability to instances of hard euroscepticism. The target group of their study is thus citizens who reject the EU in itself. We saw in Chapter 6 that the overall size of this group is relatively small. As regards perceptions of EU membership as a `bad thing', the range of people 141 Note that Matthew Gabeľs choice of dependent variable in this 1998c article was discussed at some length in Section 4.3. 142 Generally, vigilance is recommended when interpreting the correlation coefficients achieved by studies employing only the membership question as the dependent variable. Support for membership is indicative of the level of hard euroscepticism in a member state, and as the thesis holds that the types of concern shared by hard eurosceptics are the same as those shared by soft eurosceptics--namely the four types identified above--it is indeed plausible that indicators of the level of hard euroscepticism correlate strongly with indicators for each of the substantive types. Citizens who give sceptical replies to indicators of hard euroscepticism are likely to also give sceptical replies to most other indicators of EU attitudes. 221 sharing this opinion in the case countries ranged from 28 percent in the United Kingdom to 15 percent in France (the EU average was at 17 percent). Thus, unlike what is often assumed, there is not necessarily a contradiction between Anderson and Reicherťs study and studies finding age, gender and post-materialist values to be determinants of `EU supporť. Indeed, examples of studies concluding the latter have, for example, been investigating explanations for public rejection of the Maastricht Treaty (and thus not necessarily hard euroscepticism; see for instance Siune 1993). Andersen and Reicherťs study `merely' suggests that these variables have little importance for whether or not membership itself is supported, and not that they are without significance for a study of euroscepticism generally. A re-run of the studies with standardised dependent variables is what would be required before an authoritative conclusion can be drawn as to whether or not there in fact is a contradiction. The comparability of euroscepticism studies, in other words, is dependent on the employment of comparable dependent variables. II. Lauren McLarens' submission for a forthcoming issue of Acta Politica investigates the `degree to which Euroskepticism is driven by (a) feelings about national institutions, (b) distrust of supra-national institutions, (c) fears about the loss of national identity, and (d) interest-based utilitarianism.' She concludes that `Attitudes to European integration appear to be driven by feelings about EU institutions rather than attitudes to national institutions' (McLaren forthcoming: abstract; note the subtle switch between the expression of an interest in investigating `Euroskepticism' and the concluding statement on `attitudes to European integration'). McLaren herself notes the contested nature of most existing theories and findings on euroscepticism, and on this basis she sets out to `re-analyze' the major contending approaches (ibid: 1), which are perceived to be `cognitive mobilization and knowledge of the EU', `general malaise over the functioning of government and the poverty of EU Institutions', `identity' and `egocentric utilitarianism.' (ibid: 2-7). Her line of reasoning for choice of a measure for euroscepticism is that: `Euroskepticism [can be] conceptualized in terms of opposition to or reserved support for the process of European integration. The process of European integration, in turn, implies policy integration--that an increasing number of policy areas are decided at the European level rather than at the national level. Thus, the measure chosen for the concept, Euroskepticism, is a composite of responses to the standard policymaking question: `For each of the following areas, do you think that decisions should be made by the (NATIONALITY) government, or made jointly within the European Union?'' (ibid: 7-8). In terms of the framework developed by this thesis, the quote serves to disclose that McLaren's actual object of study may in fact not be the broad phenomenon of euroscepticism but only a variant of it, namely sovereignty-based euroscepticism. Indeed, while it appears questionable that support for 222 European integration as a whole can be subsumed under the question of support for joint decisionmaking, what the Eurobarometer question engaged by McLaren does highlight is attitudes towards supranational co-operation. The thesis therefore proposes that her conclusions about the primary role played by `feelings about EU institutions' in the shaping of attitudes towards European integration should be restricted to the type of sovereignty-based scepticism. In a similar vein, it would arguably be more useful to approach the finding that `many EU citizens view the European integration process as a threat to their key terminal identities' (ibid: 15) as an instance of her two indicators `exclusive national identity' (independent variable) and `the wish to retain policy making at the national leveľ (dependent variable) being highly correlated internally--and thus two aspects of the same concern--rather than to view them in terms of a causal relationship. Continuing the review of McLaren's study, her finding that perceptions of benefit from the EU lead to more support for EU policy making (ibid: 12) would then be indicative of a possible causal relationship between two of the thesis' types of euroscepticism, which it would be fruitful to investigate in future studies (see Section 6.3.1 for a similar argument of causality between types of scepticism). III. In his article from 2002, Robert Rohrschneider sets out to investigate two hypotheses, namely if `Perceptions of under-representation reduce EU support independent of economic factors' and if `The quality of national institutions mediates the effect of the representation deficit on EU supporť. `Two indicators measure EU-supporť, Rohrschneider writes, and take these to be satisfaction with EU democracy and support for an EU government. Following the typology of euroscepticism advanced by this thesis, it should be clear that these two questions are understood to address two independent types of euroscepticism, namely democratic and sovereignty-based euroscepticism. If Rohrschneider had been attentive to their different natures, the study would have been a solid contribution to our understanding of how feelings of under-representation impact on these two types respectively. However, neither does he discuss the possibility of his dependent variable not being a general measure of mass EU support (we can therefore wonder where, for example, utilitarian concerns enter the picture), nor does he seem to associate the issue of a European government with sovereigntybased concerns. Instead the question is taken to gauge support `for a central institutional variant presently considered for the EU' (Rohrschneider 2002: 466). Such conceptual uncertainties, in short, cloud the specific contribution of the article to our understanding of euroscepticism or public EU support. The thesis suggests that Rohrschneider's conclusion: `when citizens perceive that they are unrepresented, their support for the EU is reduced independent of economic perceptions' (ibid: 463) should be modified to the more 223 specific (though still interesting) claim that `feelings of being unrepresented evenly breed democraticas well as sovereignty-based euroscepticism, independently of economic perceptions'. Importantly, the above account of the works of three prominent euroscepticism researchers is critical in so far as the studies are presented as instances of public euroscepticism as a whole. They are not, and as a result careless comparisons will lead only to contradictions or spurious conclusions. Rather, the first study contributed to our knowledge of what impacts on the level of hard euroscepticism, the second study to our knowledge of sovereignty-based euroscepticism, and the latter study to our knowledge of similarities between democratic and sovereignty-based euroscepticism. This specification of the likely reach of the various studies of euroscepticism would greatly increase the cumulability of knowledge and thereby contribute to the development of a coherent theory of euroscepticism. 7.2 Bearings on research into public EU attitudes In Chapter 4, works of David Beetham and Christopher Lord, Jürgen Habermas and Joseph Weiler served as part of the thesis' platform for deriving and developing its typology of euroscepticism. In light of this typology begin subsequently confirmed, this section returns to the assumptions about public EU attitudes inherent in these works to exemplify the implications of a combined theoretical and empirical understanding of euroscepticism to attempts to theorise about the Union's legitimacy and public justification.143 I. Beetham and Lorďs conception of EU legitimacy (Beetham and Lord 1998) consists of three dimensions--legality, normative justifiability, and legitimation--which can be presented as in this simplified model (Figure 8). The dimension of normative justifiability is divided into three components, namely those of performance, democracy and identity (ibid). It has been discussed already that Beetham and Lord identify a number of possible Legality Performance Democracy Identity Legitimation N o r m a t i v e J u s t i f i a b i l i t y 143 This is thus a review purely focused on the respective authors' perspectives on public euroscepticism and no attempt to review their overall theoretical perspectives. 224 legitimacy deficits pertaining to these three components. It should be recalled, however, that the thesis narrowed down the definition of a legitimacy deficit to apply to the perception, in the eyes of citizens, of the EU's right to hold and use power or exert influence on everyday life. This restriction narrows down the applicability of the label `legitimacy crisis' to instances of hard euroscepticism (within all four types), as well as to the vertical indicators characteristic of sovereignty-based euroscepticism and to some extent of economic euroscepticism (these are discussed in more detail below; see also Section 6.6.1). Indeed, to the thesis, democratic and social euroscepticism depart in their essence from the acceptance of some supranational co-operation, unless the intensity of the scepticism within these types surmounts to hard euroscepticism. We saw in Chapter 6 that especially with regard to social euroscepticism, there is only a low association with hard euroscepticism. In other words, some euroscepticism (to the extent it does not succumb to outright rejection of EU membership) may characterise a member state without posing a legitimacy crisis to the Union. It should be noted that Beetham and Lorďs book does not focus in any great depth on public EU attitudes. However, they do include a list of `factors that may explain identification and support for the European Union' (ibid: 48-49). The list counts (i) `learning and habituation', i.e. the `passage of time'; (ii) domestic political leadership, i.e. the national context (cf. Section 4.6); and (iii) cognitive mobilisation, i.e. Ronald Ingleharťs idea that support depends on `the capacity of publics to understand (...) somewhat remote political systems' (ibid). It should be clear from Chapters 4 and 6 that the thesis understanding of EU attitudes does not concur with this list. Instead, with the above considerations in mind, it shares the following perspectives on the extent of agreement between its conceptualisation of euroscepticism and Beetham and Lorďs conceptualisation of EU legitimacy. * The dimension of legality reflects the rejection of the very rationale of co-operation and thus hard euroscepticism. * The performance component of the dimension of normative justifiability refers both to the nature and scope of the tasks the EU should undertake, and to the effectiveness with which it is able to actually perform these tasks (ibid). Criticism with regard to each of these referents reflects two independent types of euroscepticism. Social euroscepticism includes the criticism that the EU does not respond adequately to social concerns, while economic euroscepticism is the criticism of the perceived inability of a political system to bring benefits to its citizens (we may note that in Beetham and Lorďs conception, the thesis' uncorroborated expectation of 225 performance-based euroscepticism as a distinct variant of utilitarian euroscepticism appears as relevant as its corroborated variant, namely economic euroscepticism). * The democracy component concerns issues such as accountability or popular authorisation. Criticism here reflects democratic euroscepticism, namely discontentment with improper standards of co-operation. * Deficits within the identity component may result from the absence of a sufficient sense of common identity amongst the peoples of the EU. Such a deficit may breed sovereignty-based euroscepticism to the extent that citizens find supranational co-operation in the EU to be compromising the position of nation-states as the ultimate source of authority. * The dimension of legitimation, requiring the active support of citizens in the political process, is arguably largely restricted to elections. The national context approach to public EU attitudes, discussed in Chapter 4, is relevant for understanding potential deficits within this dimension of EU legitimacy. There are thus important agreements between euroscepticism and the possibility of legitimacy deficits in the European Union, although the thesis wishes to make no assumption about a direct link between the two. The agreements open for a number of perspectives on how this study's empirical analyses can respond to some of the issues that remain unanswered by Beetham and Lorďs predominantly theoretical contribution. In particular, the thesis can assess in what dimensions the EU does, and does not, appear legitimate in the eyes of the publics, and how these public perceptions differ across the Union. This cursory and focused review of Beetham and Lorďs book should thus be seen in light of the thesis' attempt to bridge the divide that often exists between, on the one hand, purely theoretical suppositions, and, on the other, non-theorised empirical survey-studies. The above analyses have shown that all types of euroscepticism are prevalent, albeit in different combinations in different member states. A casual observation, thus, is that perceptions of EU legitimacy deficits are likely to be contextual, and at least that they vary considerably across the member states, depending on their type(s) of euroscepticism. Keeping in mind the thesis' finding that the types of euroscepticism may pull in contradictory directions, this poses a challenge to the prospect of simultaneously rendering the EU legitimate in all of Beetham and Lorďs dimensions. This brings the thesis in line with Beetham and Lord, who are well aware that improvements within one dimension of legitimacy may reduce legitimacy in another dimension (for instance ibid: 23; see also discussion in Section 4.2). They do not expect, in other words, that a perfectly legitimate Union 226 is realistic (just as this thesis is aware that a Union without any euroscepticism is unrealistic). More specifically, the nature and dynamics of euroscepticism revealed by the thesis suggest that within the dimension of normative justifiability, rectifying a democratic deficit could for instance increase an identity deficit, and, likewise, that accommodating a performance deficit, by adapting the nature and scope of the EU's tasks, could simply change the profile of the `typical eurosceptic citizen'. If, for instance, calls for a more social Europe were heeded, this profile could `merely' shift from being citizens with a left-wing orientation to being citizens with a right-wing orientation, as the development could increase concerns that the EU was moving away from a predominantly economic rationale. These findings imply that improving legitimacy within one of the dimensions identified by Beetham and Lord is no guarantee for a more publicly supported Union--also even if this improvement does not bring about reverse effects within other dimensions. In fact, the thesis' empirical analyses suggest that some kind(s) of legitimacy deficits are in fact seen as desirable by eurosceptical citizens. What kinds of deficits that are considered desirable differ from one member state to another, depending on its combination of euroscepticism types. In Denmark, it appears that a majority within the population considers the EU legitimate with regard to Beetham and Lorďs performance component. Indeed, there is neither relative economic scepticism, nor relative social scepticism (i.e. the two types of euroscepticism pertaining to this component). Instead, the Union appears questioned with regard to the identity component. In France, however, it is precisely the performance component that meets scepticism, and thus poses a potential legitimacy deficit, while there appears to be no identity-based deficit. In the United Kingdom, the analyses first of all noted a relatively high level of hard scepticism, and thus the likely presence of a critique of the very legality or raison-ďtre of the Union--a serious overarching legitimacy deficit. Indeed, hard euroscepticism is a direct undermining of the EU's public legitimacy. With regard to the individual types of euroscepticism, possible legitimacy deficits are found with regard to one leg of the performance component, namely the effectiveness with which the EU is able to bring benefits, and with regard to the identity component (sovereignty-based euroscepticism). Judging from the thesis' analysis, it thus appears that both the Danes and the British would be happy for maintaining some form of identity deficit in the EU. The main task facing proponents of rendering the EU more legitimate may therefore be to find the trade-off amongst the three components of normative justifiability that appears the most suitable to the largest number of people. In this regard, the awareness of the nature of the euroscepticism situation of the EU's member states is a helpful tool. 227 II. As Chapter 4 discussed, Jürgen Habermas encourages the development of a `constitutional patriotism' as the backbone of public support for the European Union (for instance Habermas 1998). Habermas argues that `[a]s a political collectivity, Europe cannot take hold in the consciousness of its citizens simply in the shape of a common currency'. Instead, what citizens need is `symbolic crystallization', which only a Constitution can give (Habermas 2001: 2). An EU Constitution, importantly, would guarantee citizens of those rights that unite them as Europeans--rights, in other words, are to Habermas a way of expressing what Europeans have in common (ibid: 10). To Habermas (writing in 2001), the achievement of a true European Constitution is realistic, based on his rejection of the `familiar objection of the Eurosceptics' that there is yet no European people (Habermas 2001: 7; cf. the no-demos thesis discussed in Section 4.5). To Habermas, there exists a core European identity, which has developed through Europe's painful historical learning process, and the results it has brought about. What bring European nation-states together today are the challenges that they all equally face (Habermas 2001: 10). The Constitution that Habermas calls for would foster the necessary shifts that give the European Commission the functions of a government, the Council that of a second chamber and the European Parliament a more visible exercise of competencies (Habermas 2001: 8). However, importantly, such shifts in the focus of politics `from national capitals to European centres' would also be in response to the `pressure from the streeť, namely `protests no longer merely by farmers or truck-drivers, but arising from the initiatives of citizens at large'. Indeed, to Habermas, `[r]elevant interests formed along lines of political ideology, economic sector, occupational position, social class, religion, ethnicity and gender (...) fuse across national boundaries' (ibid). In light of the thesis' analyses of public euroscepticism, ideas that citizens need symbolic crystallization of the EU's powers, that recognition of common challenges provide scope for a fullfledged European government, and that `citizens at large' call for more power to European centres, appear largely uncorroborated. Indeed, in two of the surveyed member states, Denmark and the United Kingdom, it seems fair to say that the EU is to a high degree seen as a threat to national identity, and that there is a consistently strong resistance toward the ideas of a European government and more EU decision-making power. It appears that at least in these member states, citizens at large do not ascribe to Habermas' conception of a sufficiently strong European demos. Moreover, the thesis has shown that there are other `eurosceptic objections' than the perception of no demos. In 2005, the French did not subscribe to Habermas' arguments in favour of a Constitution, despite the absence in France of the Danish and British widespread reluctance 228 towards more supranational integration. Instead, the analyses suggested that French voters to a large extent rejected the European Constitution because of a lack of faith in the EU's ability to provide the social rights that they demanded. This perception may not, as in the Danish and British case, be a product of a fundamental disagreement with Habermasian calls for an `ever closer union'--rather it may reflect pragmatic concerns about the nature and development of today's Union. Indeed, an argument of the thesis has been that euroscepticism is an expression of discontent with the `EU of the day', and as the Section 7.3 discusses in more detail, the French rejection of the Constitution has to be seen in light of the `EU of 2005', which importantly had just enlarged to ten Central and East European countries. Curiously, Habermas' above-cited deliberations on the values that unite the Europeans, including his 2001 paper `Why Europe needs a Constitution', rarely considered the imminent Eastern Enlargement or the likely distinction between European and EU identities or values. Thus, as with the review of Beetham and Lorďs approach to EU legitimacy, this focused assessment of Habermas' assumptions about the nature of public EU-attitudes underlines the importance of empirical input to theories about the European integration process. III. Joseph Weiler, we saw in Chapter 4, rejects the conceptualisation of a European citizenship or identity around needs and rights. This conception, to recall, is seen as the `commodification of the political process' (for instance Weiler 1999: 335). In this sense, Weiler's views are opposing those of Jürgen Habermas, who as we have seen precisely construes rights as a unifying base of support. It is therefore relevant to also return to Weiler's assumptions about public EU attitudes in order to exemplify some of the implications of the thesis' conceptualisation. Weiler's pessimistic view on the ability of rights and good performance to foster legitimacy at the EU level contributes to substantiate his otherwise regrettably lacking discussion of what constitutions the `deep sense of malaise and public disaffection' that he refers to on several occasions (for instance Weiler 1999: 329). To Weiler, utilitarian benefits are simply not the main concern of most EU citizens; on the contrary, it is matters related to national identity that `speak' to the average citizen. Weiler thus ascribes to a sovereignty-based conception of public attitudes--rhetorically he asks: `Is the European debate in Great Britain really moved by economic differences on the desirability of EMU or, instead, by political differences concerning identity and control of national destiny?' (ibid: 327). This hint at the domination of `national identity' motivations in Britain was corroborated by the thesis' empirical analyses of euroscepticism. However, this hint as to the nature of British euroscepticism does not go far in explaining--theoretically or empirically--what dissatisfies the average EU citizen. We gain no 229 clear understanding of whether `public disaffection' is a one-dimensional phenomenon, or whether it varies in nature from member state to member state--and only a superficial account of the dynamic of public disaffection. Reflecting about what happened during the ratification crisis surrounding the Maastricht Treaty, Weiler asks: `What accounts for this attitude, for this change in fortune towards the idea of European integration?' (ibid: 329). However, before attempting to answer such a question, we must ask ourselves on what grounds we can speak of a change in fortune. Weiler seems to suggest that public euroscepticism is a product of, or at least increased in connection with, the Treaty on European Union. However, as suggested above, euroscepticism in the very country that rejected the Treaty had existed long before 1992, was at a `low' on several indicators in precisely 1992, and has considerably decreased since then with regard to several types of euroscepticism. In Denmark, and to some extent in the United Kingdom, the EU's `change in fortune' was, from the perspective of EU supporters, to the better. Moreover, we may note with regard to French euroscepticism that attempts to secure EU support by reference to its utility might in fact, contrary to Weiler's ideal, go some way in countering an otherwise sharply increasing economic euroscepticism. These findings serve as a reminder of the importance of knowing the nature of the various types of euroscepticism, their intensity, as well as in what combinations they occur in the member states. Attempting to explain public dissatisfaction with reference to a uniform phenomenon of euroscepticism is likely not to get very far in capturing its dynamics. In general, we may note from these accounts that while the thesis' understanding of public euroscepticism has varied consequences for the individual existing approaches, an overall conclusion is the demonstration of the necessity of integrating empirical findings into theoretical deliberations. Indeed, disregard of the findings that euroscepticism is a multifaceted phenomenon, in what ways it is multifaceted, how it develops, and in which combinations it is prevalent within the various member states, poses problems for the validity of approaches attempting to theorise about the state of the EU's integration process and the inclinations of the wider public. The thesis now turns to evaluate the consequences of this understanding of euroscepticism on a number of key EU developments and events. 230 7.3 Multifaceted and contradictory: consequences of contemporary euroscepticism To explain how the thesis' conceptualisation of euroscepticism contributes to understanding public positioning towards the EU as well as towards specific EU issues, this section applies the above findings about the euroscepticism situation of Denmark, France and the United Kingdom in order to: (i) examine how the four types of euroscepticism apply to contemporary accounts of broader EU visions, reflecting normative ideas as to what kind of Union is desired, which may form the underlying basis for euroscepticism if the EU is perceived to be moving in a direction further away from ones ideals; (ii) retrospectively explain differences in support amongst the Danes, the French and the British towards the 2004 Eastern Enlargement of the EU--perhaps the most significant event of European integration in the past decade, not least in the eyes of the citizen; and finally, to (iii) put forward some considerations on the likely reach of strategies suggested during the recent `Period of Reflection' to winning the hearts of the Europeans. This period was the EU's official response to the French and the Dutch rejections of the Constitutional Treaty, having so far included plans to democratise the Union (the `Plan Ď) and strengthen its output dimension (the idea of a `Europe of Results'). 7.3.1 Euroscepticism and EU visions The `average citizen' rarely elaborates in great detail on what in her opinion constitutes the rightful scope of the EU, or, to put it differently, on which kind of Union she favours. This, of course, does not preclude the existence amongst citizens of broad types of ideas or ideals as to what form co-operation in Europe should take, similar to the EU visions one can expect from EU politicians. The interest of the present discussion is to propose a preliminary investigation of whether the four delineated types of scepticism `fiť developed accounts of `public EU philosophies.' This may give us cues about the different `kinds of Europe' that underlie the euroscepticism situation of the various member states. A word of caution, however, is required. Detailed inferences about coherent EU visions cannot be expected from an account of the various motivations that make up euroscepticism. What this focus can tell us is whether, for instance, the euroscepticism situation of the British public supports the established mantra that the United Kingdom firmly ascribes to an intergovernmental vision of cooperation and so forth. This tells us something about the broader foundation of the developed types of euroscepticism. For an account of `elite' EU visions, this study relies on a recent PhD thesis engaged with the deciphering of public EU philosophies as these were expressed in the European Convention drafting the Constitutional Treaty (Olsen 2006). Tore Vincents Olsen defines a public philosophy as a theory attempting to justify to the citizens why the decisions of a political order are authoritative and therefore 231 binding (ibid: 2), and perceives of five such philosophies in the Convention: Interparliamentarians, Intergovernmentalists, Christian Civilisation, Liberalists and New Socialist Federalism. This particular study is relevant to the thesis both as it gives us a contemporary perspective on various EU philosophies, and as it surveys the views of a rather mixed group of politicians (typically members of the national parliaments of the member states) precisely occupied with setting out an overall future path for European co-operation. A potential challenge involved with the comparison is the crossnational nature of negotiation in the Convention, which could conflict with the thesis' reliance on member state populations as its unit of analysis. In this connection, however, it is important to distinguish between the thesis' theoretical conceptualisation and its choice of illustrative units of analysis. Indeed, it should be clear that the four types of euroscepticism in themselves are crossnational and applicable to studies engaging all kinds of unit of analysis, and thus not just country populations (see Chapter 2 for the discussion of country populations as unit of analysis). As the below paragraphs demonstrate, there do indeed appear to be links between the four types of euroscepticism identified by the thesis and the various EU philosophies that found expression through the drafting process of the Constitutional Treaty. Economic euroscepticism, first, seems especially to capture dissatisfied adherents of what Olsen labels the Liberal approach as well as to some extent adherents of the so-called Intergovernmental approach. The latter values the continued sovereignty of the member states, whilst strongly emphasising the utilitarian aspects of EU co-operation. It even allows economic and efficiency arguments to override sovereignty arguments in certain sensitive areas (Olsen 2006: 159). The former approach has as its core the Union's ability to guarantee the rights of individuals, in particular her private freedom and autonomy (ibid: 207). It is thus a strongly utility-oriented approach, even at the expense of both national sovereignty (ibid: 211, 217) and substantial social protection (ibid: 208). It is interesting to note that both approaches strongly emphasise efficiency arguments alongside more economic and financial arguments (we may recall that the thesis was unable to find statistical corroboration for the existence of a coherent type of efficiency-based euroscepticism). It is rather straightforward to infer that adherents of a market-based or market-run Europe will, if dissatisfied, share an economic euroscepticism. Economic euroscepticism, to turn the argument around, thus seems to be (reflective) of a marketdriven EU vision. Sovereignty-based euroscepticism is captured by Olsen's category of Interparliamentarians. To this group of politicians in the European Convention, the continued sovereignty of the member states is not only valued, it also overrides utility arguments. It is arguably also to some extent represented by the Christian Civilisation approach, which is occupied with the need for an extensive definition, and 232 protection, of the member states' national identity, while at the same time holding that the EU is the carrier of a `project of civilisation' from which it would be almost inconceivable to withdraw (ibid: 193). Common to both these approaches is that if national integrity appears threatened by continued integration, sovereignty-based euroscepticism results. One could imagine that to the first approach this sovereignty-based euroscepticism will be strongly correlated with hard euroscepticism, i.e. it would be so intense that it equals the wish to withdraw from the EU; while to the latter approach the intensity of scepticism will be unlikely to surmount a `no' vote in a referendum on continued membership. Democratic euroscepticism is interestingly not accounted for in Olsen's categorisation. This could suggest that Olsen did not identity a distinguishable, independent group in the European Convention that was united in construing the Union in a predominantly democratic rhetoric. A focused review of Olsen's data might be able to reveal the extent to which rectifying the EU's alleged democratic deficit was a dominant concern of the respective public philosophies. Indeed, its absence as an independent approach might merely reflect its integrity to the remaining approaches--however, its lack of centrality in his study does seem to suggest that such concerns were not amongst the most structural determinants of alignment in the Convention (a finding which seemingly confirms this study's observation that democratic euroscepticism is not a dominant type of public euroscepticism). Finally, social euroscepticism in the main coincides with the ideas about EU justification characterising the New Socialist Federalism approach. This approach sees the guarantee of citizens' social rights as being as essential to co-operation as economic and political guarantees (ibid: 170). It is clear that to adherents of a vision of a social Europe, the perception of the current Union as being too liberal equals social-based euroscepticism. With Olsen's total of five public philosophies thus all being found in the thesis' conceptualisation of euroscepticism, it is interesting to return to the thesis' units of analysis, namely the country populations of Denmark, France and the United Kingdom. Here we have seen that the Danes are particularly eurosceptic with regard to issues of national sovereignty, followed by democratic issues; that the French are particularly eurosceptic with regard to social issues, followed by economic and democratic issues; and that the British are particularly eurosceptic with regard to issues of national sovereignty, followed by economic issues (see table 10, Section 6.7.1). Judging from the foremost concern of each of these countries, the Danes and the British appear in line with the Interparliamentarian and/or Intergovernmental approaches, 144 and the French with the New Socialist Federalism approach. 144 Indeed, given the `mixeď and `high' level of hard euroscepticism in Denmark and the United Kingdom respectively, the link to the Christian Civilisation approach appears weaker. 233 However, incorporating also the euroscepticism types that to some extent characterise Denmark, France and the United Kingdom, the picture becomes less clear. The Danes' relatively high concern about the level of EU democracy does not fit the Interparliamentarian approach, whilst neither being easily reflected within the Intergovernmental approach. Neither is growing French economic euroscepticism easily accounted for by New Socialist Federalism. At least, it is important to note that the French critique of too little social Europe does not seem to come with an openness towards paying more to the EU budget. Only British public euroscepticism with its mix of sovereignty and economic concerns seems somewhat consistent with especially the Intergovernmental approach. This brief discussion underlines that although there is no straightforward equation between public EU philosophies and the various motivations constituting euroscepticism in Denmark, France and the United Kingdom, similar arguments about the raison-ďtre and rationale of the Union do seem to be at play. In-depth qualitative interviews with survey respondents would permit a more thorough investigation of the possibility that the various types of euroscepticism are in fact representative of coherent underpinning EU visions. Such a study would contribute to furthering our understanding of the general environments that EU integration meets with in the various member states. Indeed, while the thesis would like to repeat that care has to be taken before comparing accounts of scepticism with accounts of support or EU justification, knowledge of how the types of euroscepticism tie in with broader visions of what kind of Union is desired is useful for more fully understanding how recent initiatives to increasing public support for the EU are met. The thesis now turns to examine a number of these initiatives. 7.3.2 Euroscepticism and Eastern Enlargement In the years leading up to the EU's `big bang enlargemenť with ten central and east European countries in May 2004, Eurobarometer consistently polled reactions to this development in the existing 15 member states. While the question of support for enlargement in itself does not allow many inferences about the nature of public EU attitudes, it is possible that the more nuanced knowledge about the latter, gained in Chapter 6, can help elucidate why `enlargement scepticism' differed from member state to member state. This section, in other words, examines whether the euroscepticism situations of Denmark, France and the United Kingdom could have predicted what today is common knowledge, namely that the French were rather sceptical of Eastern Enlargement, the Danes largely supportive, and the British, if not strong supporters, then at least less sceptically inclined towards enlargement than towards many other issues on the EU agenda. 234 Eurobarometer figures serve to underline the differences. As the below graphs confirm, the French were among the most strongly opposed populations to enlargement, while the Danes were among the most positively inclined: In the spring of 2002, 17 percent of the French population saw EU enlargement as a priority. The figures were 65 percent in Denmark, and 26 percent in the United Kingdom, the latter not being significantly different from the EU average, which was 27 percent. Graph 27 (i and ii): Attitudes prior to Eastern Enlargement Enlargement: 'A priority' 65 17 26 27 0 10 20 30 40 50 60 70 Year: 2002 Percent Denmark France UK EU15 Enlargement: 'Againsť 23 47 30 35 0 10 20 30 40 50 60 70 Year: 2002 Percent The question directly surveying whether or not citizens were `for or againsť enlargement showed, also in 2002, that 47 percent of the French, 23 percent of the Danes and 35 percent of the British were `againsť. The French figure, the most sceptical of the EU-15, was 11 percentage points more `againsť than the second-most sceptical country (a position that was shared by Germany and Austria). As noted in the text accompanying the Eurobarometer poll, France was moreover distinguished by being the only member state with more people who opposed enlargement than people who were in favour of it (Standard Eurobarometer 57: 84). Further indicators largely confirm this picture. In eight polls between 1997 and 2001, Eurobarometer asked citizens in the old member states what criteria they saw fit for a country's accession to the Union (Eurobarometer Search System). One possible reply category was that `its joining should not be costly for existing member countries' (economic euroscepticism). An average for these eight polls shows the French and the British to be in line with the EU average, with around 80 percent finding the criteria important. In Denmark the figure was 58 percent. In 2002 (EB57), Eurobarometer asked citizens whether they were of the opinion that `the more member states in the EU, the more unemployment there would be' in their own country. 41 percent on average in the EU-15 thought this would be the case. In Denmark, the figure was 23 percent, thus markedly below the 235 average, while in France it was 48 percent, thus somewhat above the EU average--and 25 percentage points more likely to agree with the statement than in Denmark (at 40 percent, the British figure was in line with the average). Of the four types of euroscepticism identified by this thesis, there is reason to believe that Eastern Enlargement especially has a bearing on economic and social concerns, but also, inversely, on sovereignty-based concerns. Economic concerns arise quite simply because enlargement is costly from a short- and medium-term perspective. Partly as a consequence of enlargement, net-receivers amongst the existent member states faced becoming net-contributors (this has for instance been the case with the Danish EU contribution over the past decade), or at least a reduction of their share of the EU's structural funds. Existent net-contributors faced an increase in their expenses. It is likely to have been common knowledge that all ten new members were poorer than the EU average.145 A diverse and sizable enlargement can also be assumed to impact on evaluations of the social character of the Union. Not only were the ten newcomers relatively more poor than the old member states, their systems of social security were also less developed--a fact which citizens seemed strongly aware off. The question of opening up labour markets to workers from the new member states, for instance, spurred a number of persistent and powerful fears in the populations of the old member states, most of which had their roots in the fear of a negative impact on social equality (Vestergaard and Srensen 2004). Normative ideas that the EU should be about social homogeneity and regulation were under pressure from the `big-bang' enlargement, which `changed the nature of the EU in a way that still eludes definition' (Whitman 2005: 679). Moreover, Eastern Enlargement could hardly in itself be perceived as a development furthering a full-fledged federal EU. Indeed, as wider integration is often seen as the opposite to deeper integration, it is plausible that enlargement spurred feelings that federal developments would be at least paused. This could be hypothesised to represent a welcome development in countries where there is a marked reluctance towards supranational co-operation because of its impact on national sovereignty.146 145 Eastern enlargement can certainly also be expected to make some citizens more content with the EU on economic grounds--indeed, a central argument behind enlargement was the benefits obtainable from a big single market. The focus of this section, however, is exclusively on the possible characteristics of Eastern Enlargement that can be thought to make some citizens sceptic. 146 To some citizens and politicians, an important rationale for co-operation in the European Union may surely be the perception that co-operation in fact strengthens national sovereignty. Indeed, particularly with regard to France, it has been commonplace to state that European integration has been perceived as a means of strengthening French influence in Europe. To such ambitions, Eastern Enlargement may certainly breed fears about losing relative influence. In this sense, the 2004 enlargement could merely have been seen as the latest step in a general process lessening French influence in the EU, which may, in fact, have already started in connection with the EC's first enlargement in 1973. The thesis acknowledges that such concerns are inadequately captured by the present conceptualisation of sovereignty-based euroscepticism, which, as Chapter 4 explained, is concerned with fears that supranational co-operation in the EU in itself is compromising national 236 The remaining type of euroscepticism, the critique of the democratic deficit, appeared largely distinct from the issue of enlargement. On the basis of these expectations, the below table illustrates the euroscepticism combination likely to give rise to the most pro-Enlargement and least pro-Enlargement background. Table 13: Euroscepticism and Eastern Enlargement Euroscepticism type: Utility Sovereignty Democracy Social Pro-enlargement No Yes n/a No Against enlargement Yes No n/a Yes Juxtaposing this table with the findings from Chapter 6 (especially Table 10), we see that French euroscepticism concurs fully with the `againsť combination. Given its mix of economic and social euroscepticism and absence of sovereignty-based euroscepticism, France could have been expected to be strongly sceptical of Eastern Enlargement. Danish euroscepticism fits the pro-enlargement combination, wherefore the absence of large-scale scepticism was to be expected. Denmark is characterised by neither an economic nor a social euroscepticism, but on the contrary by a sovereigntybased scepticism. The situation in the United Kingdom is somewhat less clear. The presence of sovereignty-based euroscepticism and the absence of social euroscepticism suggested largely positive attitudes towards enlargement; however, the presence of some economic euroscepticism moderated the picture somewhat. As these findings support the actual Eurobarometer figures on attitudes towards Eastern Enlargement, knowledge of the types of euroscepticism characterising a member state appears at least in this case indicative of support for current EU developments. The thesis, however, accepts that further tests would be necessary to authoritatively rule out the possibility of other, more powerful explanations, which could render the above findings spurious or at least part of a bigger picture. sovereignty (the above-mentioned French concerns about sovereignty are likely not to lie with supranational co-operation in itself--indeed, at least to the extent that the EU is seen as reflecting French values, supranationalism does not appear to be a contested issue in France. Therefore, a more refined causal analysis of reactions to Eastern Enlargement would do well in attempting to capture varying bases for support for supranational co-operation in the EU. Staying with the relation between sovereignty concerns and EU enlargement, it is noteworthy to consider claims that wider integration may in fact contribute to strengthening the European Commission (and thus supranational elements in the EU), as it is looked to for overview in the midst of a bigger, more dispersed Union (for instance EUobserver, 27th February 2007). 237 7.3.3 Euroscepticism and initiatives during the Reflection Period Whether or not Eastern Enlargement has in fact contributed to increasing public euroscepticism in France (and elsewhere), and spurred no-votes at the referenda on the Constitutional Treaty, the EU's official response to the ratification crisis was the initiation of a `period of reflection' and, more specifically, a Commission-led Plan Democracy, Debate and Dialogue (Plan D147 )--both initiatives which to a greater or lesser extent are still on-going at the time of writing. The idea of the reflection period was for EU leaders to take the time to thoroughly listen to what the citizens wanted with the EU, and from the EU--in line with the above discussion on EU visions, it seems fair to say that the underlying aim of the declared ambition of the reflection period was to identity the EU visions of the citizens. During the past 18 months, EU leaders have used the reflection period as the basis for voicing several suggestions as to which way forward the Union in their view should take. It is possible to identify at least three prominent approaches aired during this period: * One call has been for the EU to focus on generating results. This suggestion has, inter alia, been strongly supported by politicians like the Danish Prime Minister, Anders Fogh Rasmussen, Commission President José Barroso, French Presidential candidate Nicolas Sarkozy and leader of the Conservative Party in Britain, David Cameron. It largely supposes a liberal or intergovernmental vision of co-operation. * Another approach has been the call for a new `grand projeť, or vision, for the EU (or at least for a core group within the EU), as for instance expressed in the recent book by Belgian Prime Minister Guy Verhofstadt, entitled the `United States of Europe', published November 2005. Interestingly, it appears to especially be political leaders from the founding member states of the European Community who call for overarching new and grand visions for co-operation. Being somewhat less specific than Verhofstadt, German Chancellor Angela Merkel has for instance called for a rethink of the EU and a `new rationale' or mission to underpin co-operation (EurActiv 15th May 2006); while Jacques Barrot, French vice-president of the European Commission, has identified a `need for big EU projects (...) that will make EU citizens become aware of the European added value' (EurActiv 16th 2006). On similar lines, French Minister for European Affairs, Catherine Colonna, in a rather pessimistic address on the state of the EU, called for the Union to do less small things and more grand things.148 Returning to Olsen's account of EU visions, it seems fair to say that the call for a 147 European Commission (2005). 148 `Discours ľoccasion de la Conférence des Ambassadeurs', Paris, 29th August 2006. 238 `grand projeť emanates from the assumptions of either New Social Federalism, Liberalism or Christian Civilisation. * Yet another approach has been to focus on improving the existing structures of the EU. Indeed, the core ambition of the aforementioned Plan D is to engage in democratisation and communication activities in order to engage the public and `restore public confidence in the European Union' (European Commission 2005: 3). Its motivation seems to be the interpretation of the French and the Dutch referenda results as largely protest-based and uninformed: `What influenced the choice of "no" voters in both countries the most were the concerns about the country's economic and social situation' and, with specific regard to the Netherlands, `it appears that many "no" votes were also motivated by inadequate understanding of the real impact and meaning of the Constitution'.149 Plan Ďs explicit rationale is that `[a]ny vision of the future of Europe needs to build on a clear view of citizen's needs and expectations' (European Commission 2005: 2). It is clear that these three suggestions are potentially strongly linked. Indeed, it is likely to be inconceivable to unite Europeans around a `grand projeť without being able to demonstrate its beneficial qualities to the public--something which calls for some kind of `Europe of results'. Likewise, the Commission has argued that one result of Plan Ďs efforts to listen to the public has precisely been the discovery that `citizens want a Europe of results'.150 However, on a more fundamental level, these three approaches to tackling the French and the Dutch rejection of the Constitutional Treaty carry with them strong assumptions as to the nature of public euroscepticism. Importantly, they each seem to depart from the assumption of a rather uniform euroscepticism situation across the various member states. Citizen support is seen to be achievable by focusing either on the delivery of concrete results, by the generation and promotion of a unifying vision, or by more deliberation and democracy. Given the diversity of the euroscepticism situations in Denmark, France and the United Kingdom revealed by Chapter 6, however, it is only reasonable to expect that, like Eastern Enlargement, suggestions about a `Europe of Results', a `grand projeť or a `Plan Ď, will meet with differing responses, and thus differing success-rates, in the member states. Central to an initiative like Plan D, for instance, is the rarely questioned (though as this thesis has suggested: not thoroughly analysed) belief that democratic deficiencies are a main foundation for public euroscepticism. Chapter 6 demonstrated the relevance of democratic euroscepticism; however, it also 149 European Commission (2006): `Communication from the Commission to the European Counciľ, COM(2006)212, 10th May. 239 demonstrated the likelihood that this particular type hardly constitutes the most prominent EU concern to the Danes, the French and the British. Indeed, there is little sign that democratic euroscepticism has gained in prominence over the past years (if anything, scepticism seems to decrease slightly), and neither was one of the case countries in relative terms strongly eurosceptic in this regard. Other types of euroscepticism appeared markedly more prominent. Thus, to the extent that citizens are aware of democratisation initiatives of the EU (in this connection the Plan D, which lists six initiatives to stimulating public debate, four initiatives for promoting EU democracy, and three tools to generating dialogue, cf. European Commission 2005: 7-10) and perceive of them as successful, they are indeed likely to reduce the salience of the scepticism that is based on the critique of the EU's democratic deficit. If this type of scepticism is not prominent in a country, the initiatives may, however, not play a particularly efficient role in diminishing overall levels of euroscepticism, and perhaps even contribute to the contrary. Taking another look at the indicators for `democratic euroscepticism', Denmark is the only case country with a relatively strong initial scepticism in this regard. It is, for instance, the only case with a relatively significant opposition to the European Constitution because of its perceived lack of democratic credentials. However, we note on the long-term indicator on satisfaction with EU democracy that sceptical Danish perceptions have dropped rather considerably over recent years: While sceptical evaluations in 1999 were shared by 60 percent of the population (22 percentage points above the EU average), Danish figures of dissatisfaction have now dropped to 40 percent, which is in line with the EU average. In France and the United Kingdom, negative perceptions of EU democracy have since the mid-1990s consistently been in line with the EU average, and there has not been a marked change in this particular manifestation of scepticism. Thus, only in the case country where scepticism towards the EU's level of democracy was relatively strong in the mid-1990s, has there been a significant decrease in this type of scepticism. This could indicate that democratisation efforts (see Chapter 1) have an impact, albeit a rather targeted and thus limited impact. There is, for instance, no marked decrease in Danish sovereignty-based euroscepticism, which continues to be prominent. Plan D, certainly, appears to be as much about informing and communicating as about democratising. As the `information thesis' carries with it strong assumptions about the utility of European co-operation (see Chapter 4), it seems fair to relate the underlying logic of Plan D not only to the type of democratic euroscepticism, but also to that of economic euroscepticism. Recalling the scores for the indicators of this type of scepticism, we were able to note a sharp recent increase in French sceptical replies, which has shown no sign of abiding in 150 European Commission (2006): Memo/06/192, May. 240 the newest polls (see Section 6.1). We were, however, able to simultaneously note a recent decrease in Danish scores on the exact same indicators, which could suggest that the very same communication efforts, or events, impact differently on the member states. The prevalence and persistency of multiple types of euroscepticism in the EU's member states is thus a pessimistic reminder for EU leaders hoping to rely on Plan D to bring the Union closer to the French, the Danes, the British, and 24 other populations. Most likely, it is a win-lose dilemma--a point that affects not only the potential success of Plan D, but also most other approaches which, by one strategy, seek to stimulate greater support across the member states. Thus, if in response to the French `no' to the Constitutional Treaty, politicians try to accommodate the critique of a lack of social engagement by the Union, they may succeed in making the French more content. However, they are likely to simultaneously increase euroscepticism in the United Kingdom, as the British are relatively unconcerned about a more social Europe and would rather be concerned about the initiatives' consequence for the EU's budget (economic euroscepticism) or national social policy (sovereigntybased euroscepticism). What the EU populations want from co-operation is very different. This, indeed, also poses difficulty for a `Europe of Results', which, to repeat, is perhaps today the most popular discourse among many EU leaders seeking a way forward for the Union. It builds on a clearly utilitarian understanding of public euroscepticism--indeed, the EU, according to this line of thought, has to focus on producing concrete results in areas where citizens want the EU to act. European Commission President José Barroso's own strategy, dubbed the `Elvis Presley strategy', calls for `A little less conversation, a little more action' (Financial Times: May 11, 2006). In Barroso's words, `The way to strengthen public confidence in Europe is through results'.151 Not less forcefully, Danish Prime Minister Anders Fogh Rasmussen argues that `We need to show our citizens that the EU is first and foremost about creating results (...). Our citizens see the sense of common institutions and common policies if they deliver results'.152 However, not surprisingly in light of the above accounts, Eurobarometer surveys indicate that member state populations may want EU action in very diverse fields. In Germany in autumn 2005, for instance, 74 percent believed unemployment to be one of the two major issues of today. A mere eight percent shared that opinion in Ireland. 32 percent in Denmark mentioned terrorism. This figure was but one percent in Lithuania. A similar situation, at least in the absence of a major external threat, is likely to meet the attempt to generate an encompassing `grand projeť to simultaneously strengthen 151 European Commission 2006: IP/06/595, 10th May. 152 Prime Minister's Office, Address by Prime Minister Anders Fogh Rasmussen at a Conference of Speakers of EU Parliaments in Copenhagen, 20th June 2006. 241 member state populations' ties to the EU. These points support the discussion in Section 4.2.1, where it was suggested on the basis of Fritz Scharpf's distinction between input and output legitimacy that, contrary to common beliefs, it may in fact be as difficult for the EU to achieve output legitimacy as it is to achieve as input legitimacy. 7.4 Evaluation and suggestions for further research This thesis theorised, analysed and discussed public euroscepticism. It clarified the extent to which, and how, euroscepticism is multifaceted; demonstrated how patterns of euroscepticism differ between member states and over time; explained through its findings why existing research has fallen short in accounting for today's dynamics of euroscepticism; and evaluated the consequences of its new understanding on the field of euroscepticism research as well as on today's European Union. This section summarises, evaluates and suggests possible avenues for future research. Euroscepticism, the data confirmed, comes in four broad types: it may be economic, sovereignty-based, democratic and/or social. Economic euroscepticism is the perception of too few benefits from cooperation; sovereignty-based euroscepticism is the objection to pooling national sovereignty in the EU; democratic euroscepticism is the critique of the existent structures of the Union; while social euroscepticism is the perception that the EU does too little with regard to the social protection of its citizens. These four types of euroscepticism vary considerably between the EU's member states. What citizens disapprove of in the Union is simply very different from country to country. The thesis demonstrated this point through the longitudinal examination of multiple indicators of each type of euroscepticism in Denmark, France and the United Kingdom. These three countries have all been labelled eurosceptic, a reputation which is old with regard to Denmark and the UK, and more recent with regard to France. Importantly, the thesis' analyses revealed that the label `eurosceptic' may be the only thing that unites the three countries when it comes to sceptical public EU attitudes. Indeed, there are few overlaps with regard to which types are prominent in the case countries, as well as with regard to the intensity with which citizens are sceptic. The Danes react with scepticism when the transfer of national sovereignty is at stake, and keep a critical eye on the EU's democratic credentials; the French are worried that the Union is not lucrative enough, as well as with regard to what they see as too little emphasis on a Social Europe; while the British share the Danes' concern about sovereignty, but in 242 conjunction with an economic euroscepticism. With relatively little democratic and social euroscepticism in the UK, low economic euroscepticism in Denmark, and low sovereignty-based euroscepticism in France, it seems fair to say that the picture could hardly be more diverse. A further finding which contributes to distinguishing the three case countries from one another in terms of euroscepticism is the level of hard scepticism: this level is strong in the United Kingdom, low in France, with Denmark assuming the middle position between these two countries. The longitudinal indicators are coherent in their message. They show that euroscepticism is increasing in France, as well as, albeit to a less sharp degree, in the EU as a whole. Denmark and the United Kingdom go against this trend, and in fact experience decreasing euroscepticism on the indicators that poll a longer time period. Euroscepticism, in other words, is not static: as the EU develops, so do people's attitudes--and thus, logically, euroscepticism. These diverse portraits and patterns of euroscepticism present the EU with a `win-lose' dilemma, which contributes to explain how the initiatives that the EU has undertaken in recent years (for instance enlargement, democratisation, the Constitutional Treaty) have equally contributed to increasing scepticism in some member states and reducing it in others. Indeed, precisely that area where one population wishes the EU to focus, risks being that area where another population fears its influence. It will depend on the type(s) of euroscepticism characterising a country. Knowledge about the euroscepticism situation of a member state assists predictions about public reactions to events on the EU agenda. As the level of hard euroscepticism is generally low, the future of euroscepticism depends on the direction that the EU is taking. Bringing the public in and being close to the citizens are prominent EU priorities at the time of writing, even to the extent where public euroscepticism has become a bargaining chip in treaty and policy negotiations, and referenda are advertised on contentious issues--virtually allocating euroscepticism a role as veto player in the integration process. But if the EU aims to continue as a unity in all regards, whilst being proactive and productive, its leaders may, at least in the foreseeable future, have to accept that they cannot all leave the negotiation table as a winner in the eyes of their citizens. What people want from the EU is simply very different. To their consolation, this is not some unique pathology of the Union: horizontal contestation, at least, is in fact a constitutive characteristic of any democratic political system. 7.4.1 Evaluation When I started research on this thesis three years ago, it was in fact with the intention to demonstrate the causal impact of sovereignty concerns on public euroscepticism. However, a closer investigation of this research design soon revealed its tautological nature. Indeed, if the concern about losing national 243 sovereignty is not the expression of some sort of scepticism towards the integration process, then what is? Moreover, it soon became clear from the French and the Dutch `no' votes in May and June 2005, occurring a year after the present study had begun, that voting `no' to a key European development was not contingent on dissatisfaction with membership itself, which is otherwise the sole dependent variable of the majority of existent research on euroscepticism (it was the dependent variable, or part of it, in nine out of the 12 studies listed in Table 2). Since the question on attitude to EU membership is so entangled with existing research, it is unclear in what way these studies actually contribute to our knowledge of why so many French and Dutch citizens, founding members of the Union, came to reject the Constitutional Treaty. This growing dissatisfaction with the actual explanatory power of otherwise thorough studies was what led the thesis to adopt a different approach, namely a constitutive one. The rationale is that authoritative conclusions on the causes of euroscepticism cannot be drawn as long as our understanding of the nature of euroscepticism is scattered and incomplete. Embarking on a conceptualisation of euroscepticism entailed establishing an overview of a vast field of literature and investigating the adequacy of existing arguments, independent variables and dependent variables, in order to discover if, and in what ways, they could contribute to the thesis' own conceptualisation. A number of unforeseen challenges were encountered, as accounted for in the above chapters. One involved singling out adequate expectations in existing literature as to what euroscepticism is. Indeed, the thesis came to the conclusion that some specific approaches to euroscepticism were not useful for further analysis. Although (variants of) the thesis of the Silent Revolution, for instance, featured prominently alongside other accounts of the causes of euroscepticism in existent literature, the thesis found that it was inappropriate to conceptualise as an independent type of euroscepticism, because of the risk of mixing levels of analysis (see Section 4.4). Another challenge was the lack of statistical corroboration for the performance-based type of euroscepticism. Indeed, here was a theoretical expectation of a coherent type of scepticism, which fell short of meeting the thesis' statistical threshold for a strong association. As described in Section 5.5, and as I develop below with regard to avenues for further research, the thesis was without the required means to run additional tests, which could have shed more light on the actual suitability of the indicators of this uncorroborated type. While the thesis, therefore, took the consequence of not pursuing the analysis of performance-based euroscepticism, it does not preclude that future analysis finds scope for improving our theoretical understanding, or empirical operationalisation, of this expectation--and thereby succeeds in establishing its coherence and independence as a type of euroscepticism. 244 Indeed, a related challenge that should be emphasised here is the extremely scattered nature of the Eurobarometer data, which reduced the nature and number of statistical tests that the thesis was able to run. Indeed, with regard to the type of social euroscepticism, the thesis was in its examination of the lists of content and annexes of all available Standard Eurobarometer surveys unable to find a dataset with more than one of the four indicators present. Without the same sample, the statistical tests foreseen by the thesis were not possible. Moreover, the measurement of both social and democratic euroscepticism suffered from these types only being asked about by Eurobarometer since the 1990s. Usage of cross-national polls also posed the issue of context specificity (see Section 2.3.2). To this potential problem, the thesis' approach was pragmatic. It did not preclude that words have different meanings in different contexts, and thus that answers to the same Eurobarometer poll question might reflect somewhat different concerns. Rather it sought to delimit as clearly as possible its intention with each referent of the concept of euroscepticism, whilst maintaining the important contribution of cross-national polls in revealing variations in attitudes to the same poll question over time. As Section 7.4.2 elaborated, an interesting avenue for further research would certainly be the investigation of why there is a particular variation between countries, as well as why a type of euroscepticism at a specific point in time experiences a sharp increase or decrease in a country. A final challenge that should be mentioned here posed itself in terms of the existence of a significant association between some inter-type indicators. This was the case between indicators of sovereignty-based euroscepticism and economic euroscepticism, as well as between democratic euroscepticism and economic euroscepticism. As the inter-type associations were still lower than the intra-type associations, the thesis' criterion for pursuing with the types was, however, fulfilled--and, moreover, the subsequent examination of data provided evidence in favour of keeping the types separate. Nevertheless, as suggested in Section 6.3.1, the overlap does suggest the utility of further investigating whether, for instance, there is a causal relationship between these particular types of euroscepticism. 7.4.2 Suggestions for further research I propose that future studies into the reasons behind today's manifestations of euroscepticism--that is, studies that so to say take a step further to the left of the thesis' concept structure (see Figure 7) and investigate why it is, for example, that Danes share a strong sovereignty-based euroscepticism, while the French do not, can use the thesis' conceptualisation of euroscepticism as their platform. Such studies might, for instance, investigate the impact of specific events with a bearing on sovereignty, or examine the explanatory power of national political cultures. 245 Inspired by existing research, a number of possible future causal investigations suggest themselves for examination. It would be interesting, for example, to further investigate Claes de Vreese's hypothesis of anti-immigration sentiments emerging as a key independent variable for explaining euroscepticism (de Vreese 2004). Indeed, it is plausible that anti-immigration sentiments correlate differently with the four different types of euroscepticism, and that the degree of the correlation varies across member states. Also, Leonard Ray's study of the relationship between incumbent support and EU attitudes would be relevant to nuance in terms of the thesis' conceptualisation (Ray 2003). Other studies could target the context specific differences that are likely to exist among member states. It would in this regard be relevant to examine, for example, if sovereignty-based euroscepticism is dependent on whether national identity in a country is predominantly ethnic or civic in character (cf. Beetham and Lord 1998), or if economic euroscepticism is contingent on the size of a member state. In this latter regard, we may cursorily recall the relatively significant economic scepticism in the two large member states surveyed by the thesis (France and the United Kingdom), and the virtual absence of this type of scepticism in the small member state (Denmark). In this case, a study could investigate the degree to which differences in, for example, the economical autonomy of states or the composition of the EU budget, favour small or big states, as well as whether such potential bias is reflected in economic euroscepticism. Marlene Winďs study of how the Nordic conception of apolitical courts leads to a suspicion in Denmark towards constitutional democracy and the centrality of the European Court of Justice; and Lene Hansen's argument that Danish EU attitudes must be seen in light of the way in which the key concepts of nation, state and people have been constructed, offer further inspiration as to the independent variables that could be used to quantitatively test influences on the four types of euroscepticism (Wind 2006: pp. 44-56; Hansen 2001: 113). Indeed, both qualitative and quantitative studies could be designed to examine and explain influences on public euroscepticism. Qualitative strategies could use knowledge of the four types of euroscepticism as an analytical framework for conducting semi-structured interviews and focus groups, or content analysis of relevant writings (readers' letters, political speeches, and so forth). They would have the advantage of not being dependent on a given question formulation of a survey--a disadvantage that quantitative studies using existing poll material as indicators invariably encounter. A more inductive, or discursive, approach to euroscepticism could in this sense contribute to suggesting emerging variants of euroscepticism and inspire new question formulation in quantitative polls. 246 Quantitative analyses, on the other hand, could operationalise various hypotheses regarding the causes of euroscepticism, and test these against the four types by means of regression analyses. Such analysis could, for example, serve to investigate the reasons behind the thesis' disclosed dynamics of euroscepticism--whether, for instance, events such as national elections, or periods with economic recession, impact on specific types of euroscepticism, and the extent to which member states react similarly to such events. Moreover, the investigation of the extent to while party-based, `elite' euroscepticism conforms to the thesis' four types of public euroscepticism would be valuable. Indeed, the thesis' finding of low sovereignty-based euroscepticism in France appears to stand in contrast to frequent accounts from EU meetings of French politicians being strongly concerned about the need to secure French identity in a diverse Union.153 Although specialised, extensive Eurobarometer polls of elite euroscepticism do not exist, 154 studies of elite euroscepticism could contribute to clarifying possible overlaps and discrepancies. The same would be the case with regard to the extent to which there is a discrepancy between the thesis' portrait of day-to-day euroscepticism and `no' votes at EU referenda. As Section 4.6 discussed, a strategic dimension can be assumed to enter the picture at elections--however, the extent to which this results in `protest voting' is unclear, as is the specific EU attitudes of such protest voters (if protest voters, for instance, are generally eurosceptic or not; if protest voting is mostly characteristic of adherents of specific types of euroscepticism; if protest voters generally consider the EU unimportant, and so forth). Clearly, the study of the euroscepticism situation in the remaining member states would also be interesting. This study would allow filling in the `eurosceptical map of the EU' and check for regional patterns. It would moreover be able to reveal if there are large discrepancies between the 15 `olď member states of the EU and the 12 new member states in terms of their prevalent types of euroscepticism. In a similar vein, our knowledge of euroscepticism would be strengthened by studies of the socio-demographic profiles of the different types of euroscepticism, which could clarify if particular societal groups are more represented in one of the four types than in the rest. To return specifically to the existing literature that gave the thesis input for the development of the four types of euroscepticism, the analysis revealed particular strong dynamics over time with regard 153 And with French politicians' expectations to French public euroscepticism--Francois Bayrou, presidential candidate at the 2007 elections, for instance, has been quoted for claiming that the Constitutional Treaty inspired French fears that the Constitution was a threat to their national identity, wherefore they voted `no'. The thesis' analyses have not corroborated this claim. 154 The European Commission did in 1996 undertake a survey of elite attitudes towards European integration, called the Top Decision-Makers Survey. Some results were included in Eurobarometer 46. 247 to economic euroscepticism. Reviewing this particular body of literature, I find little evidence that explains why this should be the case--why, for instance, France today experiences a large-scale increase in this type of scepticism, at the same time as scepticism is falling in Denmark and in the United Kingdom. Studies of the impact of antecedent events or occurrences (for instance particular domestic socio-economic situations, or, as hinted at in Section 7.3, public reactions to Eastern Enlargement) could contribute to explaining why contemporary euroscepticism is particularly fluctuating within this type. * * * The paths are many and diverse, but hopefully increased cumulability of studies into public euroscepticism will facilitate further clarification of this dynamic and multifaceted concept. It is common and sound advice never to issue guarantees; however, the thesis shall nevertheless venture to conclude with one, which has only grown stronger by political events over the past decade. Such studies promise not only to remain pertinent over the next many years: they may in fact constitute the foundation behind to the Union's idealistic aspiration that `[f]or any of its policies, including enlargement, the EU has to win the support of its citizens' (European Commission 2006: 23). 248 Bibliography Abramson, Paul R. and Ronald Inglehart (1987): `Generational Replacement and the Future of PostMaterialist Values', The Journal of Politics, Vol. 49 (1), pp. 231-241. Adcock, Robert and David Collier (2001): `Measurement validity: A shared standard for qualitative and quantitative research', American Political Science Review, Vol. 95 (3), pp. 529-546. Adler-Nissen, Rebecca and Rasmus Vanggaard Knudsen (2005): Forenet i Mangfoldighed? EU's Forfatningstraktat i demokratiteoretisk perspektiv, Copenhagen: Djfs Forlag. Agresti, Alan and Barbara Finlay (1997): Statistical Methods for the Social Sciences, New Jersey: PrenticeHall, Inc. Andersen, Jrgen Goul (1998): `Analyse af EU-afstemningen den 28. maj 1998', Ugebrevet Mandag Morgen, October. Andersen, Jrgen Goul (2000): `Euro-survey september 2000', Ugebrevet Mandag Morgen. Andersen, Jrgen Goul (2002): `Denmark and Europe: Did the November elections change anything?', Working Paper, Paris: Notre Europe. Andersen, Jrgen Goul (2003): `Evige skeptikere, men bedre europere. EU-modstanden i Danmark, 1972-2002', In Danish Foreign Ministry (2003): `Danmark 30 r i EU ­ et festskrifť, Gyldendal. Andersen, Svein S. and Kjell A. Eliassen (1996): The European Union: How Democratic is it?, London, Thousand Oaks, New Delhi: Sage Publications. Anderson Christopher J. (1998b): `When in Doubt, Use Proxies. Attitudes Toward Domestic Politics and Support for European Integration', Comparative Political Studies, Vol.31 (5), pp. 569-601. Anderson, Christopher J. (1998a): `Parties, Party Systems, and Satisfaction with Democratic Performance in the New Europe', Political Studies, Vol. 46, pp. 572-588. Anderson, Christopher J. (2002): `Consent and consensus: The contours of public opinion toward the Euro', Paper presented at the conference `The Year of the Euro', Nanovic Institute for European Studies, University of Notre Dame, December 6-8, pp. 1-24. Anderson, Christopher J. and Karl C. Kaltenthaler (1996): `The Dynamics of Public Opinion toward European Integration 1973-93', European Journal of International Relations, Vol. 2 (2), pp. 175-199. Anderson, Christopher J. and M. Shawn Reichert (1996): `Economic Benefits and Support for Membership in the European Union: A Cross-National Analysis', Journal of Public Policy, Vol. 15 (3), pp. 231-249. Argyrous, George (2005): Statistics for research, Sage Publications, Second edition. 249 Arnaud, Jean-Louis (2000): `France and Europe ­ The European debate in France at the start of the French presidency', Notre Europe, Research and Policy Papers, No. 10, July. Arnold, Christine and Madeleine O. Hosli (2006): `Public Opinion Heterogeneity and Euroskepticism', Paper presented at the ECPR Joint Sessions, Nicosia, April 25-30. Babbie, Earl (1998): The practice of social research, Belmont, CA: Wadsworth Publishing Company, Eighth edition. Babbie, Earl and Fred Halley (1995): Adventures in Social Research. Data Analysis using SPSS for Windows, Pine Forge Press. Baker, David and David Seawright (eds.) (1998): Britain for and Against Europe? British Politics and the Question of European Integration, Oxford University Press, Clarendon. Baker, David and Nick Randall (2002): `From Zero Sum Game to Negotiable Asset ­ and Back: National Sovereignty and British Political Culture', Paper presented to the Conference `Britain and the European Union: At the heart of Europe or on its edge', University of Oklahoma, October 24-27. Baldwin, David A. (1985): Economic Statecraft, Princeton University Press. Banchoff, Thomas and Mitchell P. Smith (eds.) (1999): Legitimacy and the European Union. A Contested Polity, London: Routledge. Barbiero, Daniel (1996?): `Further Remarks on Explanatory Adequacy and the Problem of Reproduction', Contribution to the Huen-Barbiero exchange: http://www.personal.kent.edu/~pbohanbr/Webpage/New/Huen-Barbiero/barbiero2.html. Bartolini, Stefano (2001): `La structure des clivages nationaux et la question de ľintégration dans ľUnion Européenne', Politique européenne, No. 4, pp. 15-45. Bauman, Zygmunt (1998): Globalization: The human consequences, Cambridge: Polity Press. Baylis, John and Steve Smith (2005): The Globalization of World Politics, Oxford University Press, Third edition. Beetham, David (1991): The Legitimation of Power, Macmillan Education Limited. Beetham, David and Christopher Lord (1998): Legitimacy and the European Union, Longman. Bellamy, Richard and Dario Castiglione (2001): `Normative Theory and the European Union: Legitimising the Euro-polity and its Regime', Queen's Papers on Europeanisation, No. 13. Bennett, Andrew and Alexander L. George (1997): `Research Design Tasks in Case Study Methods', Paper presented at the MacArthur Foundation Workshop on Case Study Methods, Belfer Center for Science and International Affairs (BCSIA), Harvard University, October 17-19. Benoit, Bertrand (1997): Social Nationalism: An anatomy of French euroscepticism, Aldershot: Aldgate. Berger, Arthur Asa (ed.) (1989): Political Culture and Public Opinion, New Brunswick, Oxford: Transaction Publishers. 250 Betz, Hans-Georg (2001): `Exclusionary populism in Western Europe in the 1990s', Paper presented at the UNRISD Conference, September 3-5, Durban. Bhaskar, Roy (1997): A Realist Theory of Science, New York and London: Verso, Second edition. Blondiaux, Loc, Dominique Reynié et al. (2001): ĽOpinion Publique. Perspectives anglo-saxonnes, Herms No. 31 (edited journal volume), Paris: CNRS Editions. Bogart, Leo (1989): `Mass Communications and World Opinion'. In Berger, Arthur Asa (ed.): Political Culture and Public Opinion, New Brunswick, Oxford: Transaction Publishers, pp. 51-78. Bogdanor, Vernon (2005): `Footfalls echoing in the memory. Britain and Europe: the historical perspective', Institutional Affairs, Vol. 81 (4), pp. 689-701. Bomberg, Elizabeth and Alexander Stubb (eds.) (2003): The European Union: How Does it Work?, Oxford University Press. Borre, Ole (2000): `Critical Issues and Political Alienation in Denmark', Scandinavian Political Studies, Vol. 23 (4), pp. 285-309. Borre, Ole and Jrgen Goul Andersen (1997): Voting and Political Attitudes in Denmark, Aarhus University Press. Bosch, Agusti and Kenneth Newton (1995): `Economic Calculus or Familiarity Breeds Content?'. In: Niedermayer, Oskar and Richard Sinnott (eds.): Public opinion and internationalized Governance, Oxford University Press, pp.73-104. Bourdieu, Pierre (1973): `Ľopinion publique n'existe pas', Les Temps Modernes, No. 318, pp. 1292-1309. Boy, Daniel and Nonna Mayer (1993): The French Voter Decides, Ann Arbor, The University of Michigan Press. Brady, Henry E. and David Collier (eds.) (2004): Rethinking social inquiry: Diverse Tools, Shared Standards, Lanham: Rowman and Littlefield. Branner, Hans and Morten Kelstrup (eds.) (2000): Denmark's Policy towards Europe after 1945: History, Theory and Options, Odense Universitetsforlag. Buch, Roger and Kasper M. Hansen (2002): `The Danes and Europe: From EC 1972 to Euro 2000 ­ Elections, Referendums and Attitudes', Scandinavian Political Studies, Vol. 25 (1), pp. 1-24. Burgess, Michael (2004): `Federalism'. In Wiener, A.ntje and Thomas Diez (eds.): European Integration Theory, Oxford University Press, pp. 25-44. Castells, Manuel (1997): The Information Age: Economy, Society, and Culture. Vol. 2: `The Power of Identity', Oxford and Malden, MA: Blackwell Publishers. Castells, Manuel (2000): The Information Age: Economy, Society, and Culture, Oxford and Malden, MA: Blackwell Publishers, Second edition. 251 Cautrs, Bruno (2006): `Mapping attitudes towards European integration in the French political space: how many worlds of `euroscepticism' is there?', Paper presented at the ECPR Joint Session of Workshops, Nicosia, April 25-30. Cautrs, Bruno and Dominique Reynié (eds.) (2000) [2001, 2002]: Ľopinion européenne 2000 [2001, 2002], Presses de Sciences-Po, Foundation Robert Schuman. Cautrs, Bruno and Nonna Mayer (eds.) (2004): Le nouveau désordre électoral. Les leçons du 21 Avril 2002, Presses de Sciences-Po. Chan, Kenneth (2004): `Central and Eastern Europe in the 2004 European Parliament Elections. A Not So European Evenť, Sussex European Institute Working Papers, No. 81. Chierici, Catia (2005): `Is there a European Public Opinion? Public support for the European Union, theoretical concepts and empirical measurements', Paper presented at POLIS 2005, Paris. Christenson, Reo and Robert McWilliams (1962): Voice of the People ­ Readings in Public Opinion and Propaganda, McGraw-Hill Book Company. Christiansen, Thomas (then forthcoming 1997): `The limits of legitimating European Integration', Irish Studies In International Affairs, Vol. 8. Christin, Thomas (2005): `Economic and Political Basis of Attitudes towards the EU in Central and East European Countries in the 1990s', European Union Politics, Vol. 6 (1), pp. 29-57. Çíftçí, Sabrí (2005): `Treaties, Collective Responses and the Determinants of Aggregate Support for European Integration', European Union Politics, Vol. 6 (4), pp. 469-492. Collignon, Stefan (2006): `Democracy and Europe's Crisis of Legitimacy', Intervention. Journal of Economics, vol. 3 (1), pp. 1-15. Coser, Lewis A. (Winter 1984-1985): `Reviewed work: Social Science Concepts: A Systematic Analysis, by Giovanni Sartori', Political Science Quarterly, Vol. 99 (4), pp. 735-736. Crum, Ben and Harmen Binnema (2006): `Euroscepticism as a Carrier of Elite-Mass Incongruence: The Case of The Netherlands', Paper presented at the Colloquium `Resisting Europe. Euroscepticism and national civic cultures', Brussels, March 16-17. Daddow, Oliver J. (2006): `Euroscepticism and the culture of the discipline of history', Review of International Studies, Vol. 32, pp. 309-28. Dahrendorf, Ralf (2003): `The Challenge for Democracy', Journal of Democracy, Vol. 14 (4), pp. 101-114. Dalton, Russell J. (1998): `Political Support in Advanced Industrial Democracies', Paper posted at the eScholarship Repository, University of California, http://repositories.cdlib.org/csd/98-01. Danish Foreign Ministry (ed.) (2003): Danmark 30 r i EU ­ et festskrift, Gyldendal. Danish Institute of International Affairs (DUPI) (2000): Udviklingen i EU siden 1992 p de omrder, der er omfattet af de danske forbehold, Copenhagen: DUPI. 252 Decker, Frank (2002): `Governance beyond the nation-state. Reflections on the democratic deficit of the European Union', Journal of European Public Policy, Vol. 9 (2), pp. 256-72. Delanty, Gerard (1996): `Beyond the Nation-State: National Identity and Citizenship in a Multicultural Society - A Response to Rex', Sociological Research Online, Vol. 1 (3), www.socresonline.org.uk/socresonline/1/3/1.html. Denman, Roy (1996): Missed chances ­ Britain and Europe in the 20th Century, Indigo. Deutsch, Karl W. et al. (1957): Political Community and the North Atlantic Area: International Organization in the Light of Historical Experience, Princeton University Press. Deutsch, Karl W. et al. (1967): France, Germany and the Western Alliance: A Study of Elite Attitudes on European Integration and World Politics, Charles Scribner's Sons, New York. Diez, Thomas and Antje Wiener (2004): `Introducing the Mosaic of Integration Theory'. In: Wiener, Antje and Thomas Diez (eds.): European integration theory, Oxford University Press, pp. 1-24. Dinan, Desmond (1994): Ever Closer Union? An Introduction to the European Community, Lynne Rienner Publishers Inc., Macmillan Press Ltd. Dinan, Desmond (2002): `Governance and Institutions: Debating the EU's future', Journal of Common Market Studies, Vol. 40, Annual review, pp. 29-43. Dobson, Lynn and Albert Weale (2003): `Governance and Legitimacy'. In Bomberg, Elizabeth and Alexander Stubb (eds.): The European Union: How Does it Work?, Oxford University Press, pp. 156-173. Dosenrode, Sren (2002): `The Danes, the European Union, and the Up-coming Presidency', Report presented to Groupement ďétudes et de recherches, Paris: Notre Europe. Dosenrode, Sren (red.) (2000): Danmark og MU'en. Politiske Aspekter, Systime/Rdet for europisk politik: rhus. Drake, Helen (ed.) (2005): French Relations with the European Union, London and New York: Routledge. Easton, David (1965): A Systems Analysis of Political Life, New York: John Wiley and Sons. Eatwell, Roger (ed.) (1997): European political cultures. Conflict or Convergence?, London and New York: Routledge. Ehin, Piret (2001): `Determinants of public support for EU membership: Data from the Baltic countries', European Journal of Political Research, Vol. 40, pp. 31-56. Eichenberg, Richard and Russell J. Dalton (forthcoming 2007): `Post-Maastricht Blues: The Transformation of Citizen Support for European Integration, 1973-2004, Submission for a special issue for Acta Politica. Eichenberg, Richard C. (1999): `Measurement Matters: Cumulation In The Study Of Citizen Support for European Integration', Research note, Tufts University. 253 Eichenberg, Richard C. and Russell J. Dalton (1993): `The Dynamics of Public Support for European Integration', International Organization, Vol.47 (4), pp. 507-534. Eijk, Cees van der and Mark Franklin (2001): `The Sleeping Giant: Potential for Contestation on European matters at National Elections in Europe', Paper presented at the annual meeting of the American Political Science Association, San Francisco, September. Eijk, Cees van der, Mark Franklin and Michael Marsh (1996): `What Voters Teach Us about EuropeWide Elections: What Europe-Wide Elections Teach Us about Voters', Electoral Studies, Vol. 15 (2), pp. 149-166. Elkins, Zachary (2000): `Gradations of Democracy? Empirical Tests of Alternative Conceptualizations', American Journal of Political Science, Vol. 44 (2), pp. 287-294. Eriksen, Erik Oddvar and John Erik Fossum (2000): `The EU and Post-National Legitimacy', Arena Working Papers, No. 2. Eriksen, Erik Oddvar and John Erik Fossum (2002): `Democracy through Strong Publics in the European Union?', Journal of Common Market Studies, vol. 40 (3), pp. 401-424. Eriksen, Erik Oddvar et al. (eds.) (2005): The European Constitution: the Rubicon crossed?, Arena, Oslo. Eriksen, Erik, John Fossum and Agustín Menéndez (2005): `Introduction. Crossing the Political Rubicon'. In: Eriksen, Erik Oddvar et al. (eds.): The European Constitution: the Rubicon crossed?, Arena, Oslo. Eurobarometer (multiple years): The European Commission's Eurobarometer Surveys can be obtained here: http://ec.europa.eu/public_opinion/index_en.htm. (The data-sets can be acquired through the Danish Data Archive). European Commission (2005): `The Commission's contribution to the period of reflection and beyond: Plan-D for Democracy, Dialogue and Debate', Communication from the Commission to the Council, the European Parliament, the European Economic and Social Committee and the Committee of the Regions, Brussels, October 13, COM(2005) 494 final. European Commission (2006): `Enlargement Strategy and Main Challenges 2006 ­ 2007', Communication from the Commission to the European Parliament and the Council, COMM(2006) 649, November. Fallend, Franz (2002): `Opposing Europe: Euroscepticism of Political Parties in Austria', Paper prepared for the ECPR Joint Sessions of Workshops, Turin, March 22-27. Feld, Werner J. and John K. Wildgen (1976): Domestic Political Realities and European Unification. A Study of Mass Publics and Elites in the European Community Countries, Boulder, Colorado: Westview Press. Feldbk, Ole (ed.) (1992): Dansk Identitetshistorie 4: Danmark og Europa 1940 ­ 1990, C. A. Reitzels Forlag. Fishkin, James S. (1995): The Voice of the People ­ Public opinion and democracy, Yale University Press. 254 Flockhart, Trine (2005): `Critical Junctures and Social Identity Theory: Explaining the Gap between Danish Mass and Elite Attitudes to Europeanization', Journal of Common Market Studies, Vol. 43 (2), pp. 251-71. Flood, Christopher (2002): `Euroscepticism: A Problematic Concepť, Paper presented at the UACES 32nd Annual Conference and 7th Research Conference, Queen's University Belfast. Flood, Christopher (2002): `The Challenge of Euroscepticism'. In: Gower, Jackie (ed.): The European Union Handbook, Fitzroy Dearborn Publishers, Second edition, pp. 73-84. Flood, Christopher (2005): `French Euroscepticism and the politics of indifference'. In: Drake, Helen (ed.): French Relations with the European Union, London and New York: Routledge. Flood, Christopher and Simon Usherwood (2005): `Positions, Dispositions, Transitions: A model of Group Alignment on EU Integration', Paper presented at the 55th Annual Conference of the Political Studies Association, University of Leeds, April 5-7. Fllesdal, Andreas (2000): `The Future Soul of Europe: Nationalism or Just Patriotism? A Critique of David Miller's defence of Nationality', Arena Working Paper, No. 7. Fllesdal, Andreas (2004): `Legitimacy Theories of the European Union', Arena Working Paper, No. 04/15. Fllesdal, Andreas and Simon Hix (2005): `Why There is a Democratic Deficit in the EU: A Response to Majone and Moravcsik', European Governance Papers (Eurogov), No. C-05-02. Forster, Anthony (2000): `Euroscepticism: What, When, Why and Where?', Paper presented at the colloquium `Le Royaume-Uni et la construction européenne', Université ďOrléans, November 24-25. Forster, Anthony (2002): Euroscepticism in Contemporary British Politics, London and New York: Routledge. Fortune, Anne E. and William J. Reid (1999): Research in social work, New York: Columbia University Press, Third edition. Fossum, John Erik (2006): `Conceptualizing the European Union Through Four Strategies of Comparison', Comparative European Politics, Vol. 4, pp. 94-123. Franklin, Mark (2002): `Learning from the Danish case: A comment on Palle Svensson's critique of the Franklin thesis', European Journal of Political Research, Vol. 41, pp. 751-757. Franklin, Mark, Michael Marsh and Lauren McLaren (1994): `Uncorking the bottle: Popular Opposition to European Unification in the Wake of Maastrichť, Journal of Common Market Studies, Vol. 32 (4), pp. 455-72. Friis, Lykke (1999): `EU and Legitimacy - The Challenge of Compatibility. A Danish Case Study', Cooperation and Conflict, Vol. 34 (3), pp. 243-271. 255 Fuchs, Dieter (2003): `Das Demokratiedefizit der Europäischen Union und die politische Integration Europas: Eine Analyse der Einstellungen der Bürger in Westeuropa'. In: Brettschneider, Frank, Jan van Deth and Edeltraud Roller (eds.): Europäische Integration in der öffentlichen Meinung', Leske und Budrich, Opladen. Gabel, Matthew (1998a): Interests and Integration: Market Liberalization, Public Opinion, and the European Union, University of Michigan Press. Gabel, Matthew (1998b): `Economic Integration and Mass Politics: Market Liberalization and Public Attitudes in the European Union', American Journal of Political Science, Vol. 42 (2), pp. 936­53. Gabel, Matthew (1998c): `Public Support for European Integration: An Empirical Test of Five Theories', Journal of Politics, Vol. 60 (2), pp. 333­354. Gabel, Matthew and Christopher J. Anderson (2001): `Exploring the European Demos (or lack thereof): the Structure of Citizen Attitudes and the European Political Space', Jean Monnet Centre for European Studies (CEuS), University of Bremen, Working Paper No. 4. Gabel, Matthew and Guy D. Whitten (1997): `Economic Conditions, Economic Perceptions, and Public Support for European Integration', Political Behavior, Vol. 19 (1), pp.81-96. Garry, John, Michael Marsh and Richard Sinnott (2005): ``Second-order' versus `Issue-voting' Effects in EU Referendums. Evidence from the Irish Nice Treaty Referendums', European Union Politics, Vol. 6 (2), pp. 201-221. George, Alexander L. (1979): `Case Studies and Theory Development: The Method of Structured, Focused Comparison', In: Paul Gordon Lauren (ed.): Diplomacy. New Approaches in History, Theory, and Policy, The Free Press, Macmillan Publishing Co., Inc. George, Alexander L. and Andrew Bennett (2005): Case Studies and Theory Development in the Social Science, Cambridge: MIT Press. Gerring, John (1997): `Ideology: A Definitional Analysis', Political Research Quarterly, Vol. 50, (4), pp. 957- 994. Gerring, John (2001): Social Science Methodology, A Criterial Framework, Cambridge University Press. Gerring, John (2004): `What Is a Case Study and What Is It Good for?', American Political Science Review, Vol. 98 (2), pp. 341-354. Gilje, Nils and Harald Grimen (1993): Samfunnsvitenskapenes Forutsetninger. Innfring i samfunnsvitenskapenes vitenskapsfilosofi, Universitetsforlaget, Oslo. Gill, Mark, Simon Atkinson and Roger Mortimore (2004): The Referendum Battle, London: The Foreign Policy Centre/MORI. Gillingham, John (2003): European integration, 1950-2003. Superstate or New Market Economy, Cambridge University Press. Goertz, Gary (2005): Social Science Concepts: A User's Guide, Princeton University Press. 256 Gowan, Peter and Perry Anderson (eds.) (1997): The Question of Europe, London: Verso. Gowland, David and Arthur Turner (2000): Reluctant Europeans. Britain and European integration 19451998, Pearson Education Limited, Longman. Gowland, David, Basil O'Neill and Richard Dunphy (2000): The European Mosaic, Pearson Education Limited, Second edition. Grabbe, Heather (2001): `The Irish shock to enlargemenť, Centre for European Reform, No. 08/06. Gramacho, Wladimir and Iván Llamazares (?): `Euroskeptics among Euroenthusiasts. An Empirical Exploration of Lukewarm Europeanism in Greece, Portugal, and Spain', Draft paper, University of Salamanca. Groom, A. J. R. and Paul Taylor (1975): Functionalism, Theory and Practice in International Relations, University of London Press Ltd. Groom, A. J. R. and Paul Taylor (1990): Frameworks for International Co-operation, Pinter Publishers: London. Gstöhl, Sieglinde (2002): Reluctant Europeans. Norway, Sweden, and Switzerland in the process of integration, Lynne Rienner Publishers, Inc. Gundelach, Peter (ed.) (2002): Danskernes vrdier 1981 ­ 1999, Copenhagen: Hans Reitzels Forlag. Haahr, Jens Henrik (2003): `Our Danish Democracy': Community, People and Democracy in the Danish Debate on the Common Currency', Cooperation and Conflict, Vol. 38 (1), pp. 27-47. Haas, Ernst Bernard (1958): The Uniting of Europe, Stanford University Press. Habermas, Jürgen (1993): `Statsborgerskab og national identiteť, Grus, No. 39. pp. 80-98. Habermas, Jürgen (1998): The Inclusion of the Other, Cambridge: MIT Press. Habermas, Jürgen (2001): `Why Europe Needs a Constitution', New Left Review, Vol.11, Online edition. Habermas, Jürgen (2003): `Toward a Cosmopolitan Europe', Journal of Democracy, Vol. 14 (4), pp. 86- 100. Haesly, Richard (2001): `Euroskeptics, Europhiles and Instrumental Europeans', European Union Politics, Vol. 2 (1), pp. 81-102. Hague, Rod, Martin Harrop and Shaun Breslin (1992): Comparative government and politics, an introduction, Macmillan, Third edition. Hansen, Bernhard (2002): `Eurofili vs. Euroskepsis i dansk partipolitik ­ hvilken rolle spiller partimedlemmernes holdniger?', Politologiske studier, Vol. 5 (4), pp. 50-62. Hansen, Kasper Mller and Jacob Andersen (2006): `Fra Folkehringen til Folketingeť, Rapport over deltagernes svar p sprgeskemaerne i forbindelse med Folketingets og TV2s Europahring 2006, April 29-30, Dansk Kommunikation. 257 Hansen, Lene (2001): `Historie, identitet og det danske europadilemma', Den Jyske Historiker, No. 93, pp. 113-131. Hansen, Lene and Michael C. Williams (1999): `The Myths of Europe: Legitimacy, Community and the `Crisis' of the EU', Journal of Common Market Studies, Vol. 37 (2), pp. 233-249. Hansen, Lene and Ole Wver (eds.) (2002): European Integration and National Identity, The challenge of the Nordic states, Routledge. Harmsen, Robert (2002): `Europe and the Dutch Parliamentary election of May 2002', Opposing Europe Research Network, Election Briefing, No. 3. Harmsen, Robert (2005): `A Dual Exceptionalism?: British and French Patterns of Euroscepticism in Wider Contemporary Perspective, Paper presented at the Workshop `National Identity and Euroscepticism: A Comparison Between France and the United Kingdom', Oxford, May 13. Harmsen, Robert and Menno Spiering (2004) (eds.): Euroscepticism ­ Party politics, national identity and European integration, Amsterdam, New York: Rodopi. Hay, Colin (2002): Political Analysis, Houndmills, Basingstoke, Hampshire and New York: Palgrave. Hayward, Katy (2002): `Not a Nice Surprise: An analysis of the Debate on the Nice Treaty in the Republic of Irelanď, Irish Studies in International Affairs, Vol. 13, pp. 167-186. Hayward, Katy (2003): `"If at First You Don't Succeed...": The Second Referendum on the Treaty of Nice, 2002', Irish Political Studies, Vol. 18 (1), pp. 120-132. Heath, Anthony, Robert Andersen and Richard Sinnott (2003): `Do less informed voters make political mistakes? Political knowledge and electoral choice', Revue de la Maison française ďOxford, Vol. 1 (1), pp. 69-86. Hedetoft, Ulf (1994): `The State of Sovereignty in Europe: Political Concept or Cultural Self-Image'. In Zetterholm, Staffan (ed.): National Cultures and European Integration: Exploratory Essays on Cultural Diversity and Common Policies, Oxford: Berg. Hellevik, Ottar (1997): Forskningsmetode I Sosiologi og Statsvitenskap, Oslo: Universitetsforlaget. Hewstone, Miles (1986): Understanding attitudes to the European Community. A social-psychological study in four member states, Cambridge University Press. Hewstone, Miles (1991): `Public Opinion and Public Information Campaigns: the Value of the Eurobarometer'. In: Reif, Karlheinz and Ronald Inglehart (eds.): Eurobarometer: The Dynamics of European Public Opinion, Macmillan Academic and Professional Ltd, pp. 81-100. Hix, Simon (1998): `The Study of the EU II: the `new governance' agenda and its rivaľ, Journal of European Public Policy, Vol. 5 (1), pp. 38-65. Hix, Simon (2005a): The Political System of the European Union, Palgrave Macmillan, Second edition. 258 Hix, Simon (2005b): `Understanding European Parliament Elections: Punishment or Protest?', Paper presented at the Ninth Biennial International Conference of the European Union Studies Association, Austin, Texas, March 31-April 2. Hix, Simon and Christopher Lord (1997): Political Parties in the European Union, Basingstoke, Macmillan Press Ltd. Hobolt, Sara Binzer (2003): `Europe and the Ballot: Voting Behaviour in Referendums on European Integration', Paper presented at the Second ECPR conference, Marburg, September 18-21. Hjlund, Niels (2000): Det evindelige forbehold: Danmark i den store verden, Gyldendal. Holm, Hans Henrik (1996): `Denmark's active internationalism', DUPI Working Paper, No. 3. Holm, Ulla (2006): Det gamle Frankrig i det nye Europa, Danish Institute for International Studies. Holmes, Martin (ed.) (1996): The Eurosceptical Reader, Palgrave Macmillan. Holmes, Martin (ed.) (2002): The Eurosceptical Reader 2, Palgrave Macmillan. Hooghe, Liesbet (2001?): `The Renaissance of European Integration Studies', article based on the prologue of Hooghe's then forthcoming book (2002): The Commission and the Integration of Europe. Explaining Preferences, Cambridge University Press. Hooghe, Liesbet and Gary Marks (2004): `Does Identity or Economic Rationality Drive Public Opinion on European Integration?', PSOnline, www.apsanet.org. Hooghe, Liesbet and Gary Marks (2005): `Calculation, Community and Cues. Public Opinion on European Integration', European Union Politics, Vol. 6 (4), pp. 419-443. Hooghe, Liesbet and Gary Marks (forthcoming 2007): `Europe's Blues: Introduction', Submission for a special issue for Acta Politica. Hooghe, Liesbet, Gary Marks and Carole J. Wilson (2002): `Does Left/Right Structure Party Positions on European Integration?', Then forthcoming in Comparative Political Studies, Vol. 35 (8). Hooghe, Liesbet, Gary Marks and Carole Wilson (2004): `Does Left/Right Structure Party Positions on European Integration?' in Marks, Gary and Marco Steenbergen (eds.): European Integration and Political Conflict, Cambridge University Press, pp. 120-40. Hug, Simon and Pascal Sciarini (2000): `Referendums on European Integration. Do Institutions Matter in the Voter's Decision?', Comparative Political Studies, Vol. 33 (1), pp. 3-36. Hvidt, Kristian (2001): Dem om os, Copenhagen: Udenrigsministeriets Informationssekretariat. Ingebritsen, Christine and Larson, Susan (1997): `Interest and Identity. Finland, Norway and the European Union', Cooperation and Conflict, Vol. 32 (2), pp. 207-222. Inglehart, Richard (1970): `Public Opinion and Regional Integration', International Organization, Vol. 24 (4), pp. 764-795. 259 Inglehart, Ronald (1970): `Cognitive Mobilization and European Identity', Comparative Politics, Vol. 3 (1), pp. 45-70. Inglehart, Ronald (1971): `The Silent Revolution in Europe: Intergenerational Change in Post-Industrial Societies', American Political Science Review, Vol. 65 (4), pp. 991-1017. Inglehart, Ronald and Karlheinz Reif (1991): `Analyzing Trends in West European Opinion: The Role of the Eurobarometer Surveys'. In: Reif, Karlheinz and Ronald Inglehart (eds.): Eurobarometer: The Dynamics of European Public Opinion, London: Macmillan, pp. 1-26. Inglehart, Ronald, Jacques-René Rabier and Karlheinz Reif (1991): `The Evolution of Public Attitudes Toward European Integration'. In Reif, Karlheinz and Ronald Inglehart (eds.): Eurobarometer: The Dynamics of European Public Opinion, London: Macmillan. Ivaldi, Gilles (2006): `Beyond France's 2005 Referendum on the European Constitutional Treaty: Second-Order Model, Anti-Establishment Attitudes and the End of the Alternative European Utopia', West European Politics, Vol. 29 (1), pp. 47-69. Jachtenfuchs, Markus (2001): `The Governance approach to European integration', Journal of Common Market Studies, Vol. 39 (2), pp. 245-264. Jachtenfuchs, Markus and Beate Kohler-Koch (eds.) (1996): Europasche Integration, Leske und Budrich. Jachtenfuchs, Markus et al. (1998): `Which Europe? Conflicting models of a legitimate European political order', European Journal of International Relations, Vol. 4 (4), pp. 409-445. Jachtenfuchs, Markus, (1995): `Theoretical reflections on the efficiency and democracy of European governance structures?', www.ecsanet.org/conferences/ecsaworld2/Jachtenfuchs.htm. Jakobsen, Mogens, Regin Reinert and Sren Risbjerg Thomsen (2001): `Afstemningen om den flles mnt - social baggrund og holdninger', Politica, Vol. 33 (1), pp. 66-88. Janssen, Joseph H. (1991): `Post-Materialism, Cognitive Mobilization, and Support for European Integration', British Journal of Political Science, Vol. 21, pp. 443-68. Jonghe, Karen De and Peter Bursens (2003): `The Quest for more Legitimacy in the European Union as a Multi-Level Political System: a Conceptual Framework', Paper presented at the ECPR congress in Edinburgh, March 28 ­ April 2. Karpantschof, Rene (2002): `Populism and Right Wing Extremism in Denmark 1980 ­ 2001', Sociologisk rapportserie, No. 4, Copenhagen University: Sociologisk Institut. Keersbergen, Kees Van and Catherine Netjes (forthcoming 2007): `Interests, Identity and Political Allegiance in the European Union', Submission for a special issue for Acta Politica. Kelstrup, Morten (1996): `What kind of European Union - reflections on political structures and processes in the new Europe', CORE Working Paper, No. 4, Copenhagen: Institut for Statskundskab. 260 Kelstrup, Morten (1998): `Integration Theories: History, Competing Approaches and New Perspectives'. In Wivel, Anders (ed.): Explaining European Integration, Copenhagen Political Studies Press, pp. 15-55. Kelstrup, Morten (2000): `Legitimacy, Democracy and the European Union: Perspectives in the Normative Discussion of EU's Future Political Structure', CORE Working Paper, No. 1. Kersbergen, Kees van (1997): `Double Allegiance in European Integration. Publics, Nation-States, and Social Policy', European University Institute, Working Paper, RSC No. 15. King, Gary, Robert O. Keohane and Sidney Verba (1994): Designing Social Inquiry. Scientific Inference in Qualitative Research, Princeton University Press. Kopecky, Petr and Cas Mudde (2002): `The Two Sides of Euroscepticism: Party Positions on European Integration in East Central Europe', European Union Politics, Vol. 3 (3), pp. 297-326. Kritzinger, Sylvia (2003): `The influence of the Nation-State on individual support for the European Union', European Union Politics, Vol. 4, pp. 219-241. Kritzinger, Sylvia (2004): `European Identity Building from the Perspective of Efficiency', Paper presented at the Second pan-European ECPR Conference, Bologna, June 24-26. Lang-Jensen, Sebastian (2003): `Den danske venstrefljs modstand mod EF 1957-1972', Jean Monnet Skriftserie No. 6, University of Aarhus. Laursen, Finn (2004): `Denmark, the European Convention and the IGC 2003-04: The handling of a difficult two-level game', draft article for Carlsen, Per and Hans Mouritzen (eds.) (2005), Danish foreign policy yearbook 2004, Danish Institute for International Studies. Lawler, Peter (1997): `Scandinavian Exceptionalism and the European Union', Journal of Common Market Studies, Vol. 35 (4), pp. 565-594. Leonard, Mark (1998): Rediscovering Europe, London: Demos. Lijphart, Arend (1971): `Comparative Politics and Comparative Methoď, The American Political Science Review, Vol. 65 (3), pp. 682-693. Lindberg, Leon (1967): `The European Community as a Political System. Notes toward the Construction of a Modeľ, Journal of Common Market Studies, Vol. 5 (4), pp. 344-387. Lindberg, Leon and Stuart Scheingold (1970): Europe's would-be polity. Patterns of change in the European Community, New Jersey: Prentice-Hall, Inc. Lippmann, Walter (1922): Public Opinion, New York: The Free Press, London: Collier-Macmillan Ltd. Lipset, Seymour M. (1984): `Social Conflict, Legitimacy and Democracy' in William Connolly (ed.) (1984): Legitimacy and the State, New York University Press, pp. 88-103. Lise Lyck (ed.) (1992): Denmark and EC Membership Evaluated, New York: St. Martin's Press. 261 Lodge, Juliet (1993): The European Community and the Challenge of the Future, London: Pinter Publishers, Second edition. Lodge, Juliet (ed.) (1996): The 1994 elections to the European Parliament, Pinter. Lord, Christopher (2000): `Legitimacy, Democracy and the EU: when abstract questions become practical policy problems', Policy Paper No. 03, http://www.one-europe.ac.uk/pdf/P3Lord.PDF. Lord, Christopher and David Beetham (2001): `Legitimizing the EU: Is there a `Post-parliamentary Basis' for its Legitimation?', Journal of Common Market Studies, Vol. 39 (3), pp. 443-62. Lubbers, Marcel and Peer Scheepers (2005): `Political versus Instrumental Euro-scepticism', Mapping Scepticism in European Countries and Regions, European Union Politics, Vol. 6 (2), pp. 223-242. Luedtke, Adam (2005): `European Integration, Public Opinion and Immigration Policy. Testing the Impact of National Identity', European Union Politics, Vol. 6 (1), pp. 83-112. Machiavelli, Nicolo (1961 [1514]): The Prince, Penguin Classics. Magnette, Paul (2003): `European Governance and Civic Participation: Beyond Elitist Citizenship?', Political Studies, Vol. 51, pp. 144-160. Mair, Peter (2005): `Popular Democracy and the European Union Polity', European Governance Papers (EUROGOV), No. C-05-03. Majone, Giandomenico (1996): Regulating Europe, London and New York: Routledge. Manners, Ian (2002): `Normative Power Europe: A Contradiction in Terms?', Journal of Common Market Studies, Vol. 40 (2), pp. 235-258. Marcussen, Martin and Mette Zlner (2000): `The Danish EMU-referendum 2000: Business as Usuaľ, Kommission Statskundskab, Working Paper, No. 17. Marcussen, Martin et al. (1999): `Constructing Europe? The evolution of French, British and German nation state identities', Journal of European Public Policy, Vol. 6 (4), Special issue, pp. 614-633. Marks, Gary and Liesbet Hooghe (1999): `National Parties and the Contestation of Europe'. In Banchoff, Thomas and Mitchell P. Smith (eds.): Legitimacy and the European Union, London: Routledge. Marks, Gary and Marco Steenbergen (eds.) (2004): European Integration and Political Conflict, Cambridge University Press. Marsh, Michael (1999): `Policy Performance'. In: Schmitt, Hermann and Jacques Thomassen: Political Representation and Legitimacy in the European Union, Oxford University Press, pp. 90-110. Marthaler, Sally (2005): `The French referendum on ratification of the Constitutional Treaty 29 May 2005', Referendum Briefing Paper, No. 12, EPERN. Marthaler, Sally (2005): `The French referendum on ratification of the Constitutional Treaty', Referendum briefing paper, No. 12, European Parties Elections and Referendums Network (EPERN). 262 Mayer, Nonna and Pascal Perrineau (1992): Les comportements politiques, Paris: Armand Colin. McLaren, Lauren (2002): `Public Support for the European Union: Cost/Benefit Analysis or Perceived Cultural Threat?', The Journal of Politics, Vol. 64 (2), pp. 551-566. McLaren, Lauren (2005): `Explaining Mass-Level Euroskepticism: Identity, Interests, and Institutional Distrusť, Paper presented at the conference: Euroskepticism--Causes and Consequences, Amsterdam, July 1-2. McLaren, Lauren (forthcoming 2007): `Explaining Mass-Level Euroscepticism: Identity, Interests, and Institutional Distrusť, Submission for a special issue for Acta Politica. Medrano, Juan Diez (2003): Framing Europe: Attitudes to European integration in Germany, Spain, and the United Kingdom, Princeton University Press. Mény, Yves (2003a): `De La Démocratie en Europe: Old Concepts and New Challenges', Journal of Common Market Studies, Vol. 41 (1), pp. 1-13. Mény, Yves (2003b): `The Achievements of the Convention', Journal of Democracy, Vol. 14 (4), pp. 57-70. Miller, Vaughne (2001): `The Irish referendum on the Treaty of Nice', House of Commons Research Paper 01/57, June. Milner, Susan (2000): `Introduction: A Healthy Scepticism?', Journal of European Integration, Vol. 22 (1), pp. 1-13. Milward, Alan (1997): `The Springs of Integration'. In Gowan, Peter and Perry Anderson (eds.): The Question of Europe, London: Verso, pp. 5-20. Minogue, Kenneth (1996): `National Self-Hatred and the EC'. In Holmes, Martin (ed.): The Eurosceptical Reader, Palgrave Macmillan, pp. 261-268. Mller, Per Stig (1992): `Jeg tvivler, alts er jeg europer', In Srensen, Christen (ed.): Europa. Nation ­ Union, Fremad. Moravcsik, Andrew (1998): The Choice for Europe: Social Purpose and State Power from Messina to Maastricht, Cornell University Press. Moravcsik, Andrew (2005): `The European Constitutional Compromise and the neofunctionalist legacy', Journal of European Public Policy, Vol. 12 (2), pp. 349-386. Moravscik, Andrew (2002): `In defence of the `democratic deficiť: Reassessing legitimacy in the European Union', Journal of Common Market Studies, Vol. 40 (4), pp. 603-624. Moravscik, Andrew (2006): `What Can We Learn from the Collapse of the European Constitutional Project?', Politische Vierteljahresschrift, Vol. 47, pp. 219-241. Morgan, Glyn (2005): The Idea of a European Superstate. Public Justification and European Integration, Princeton University Press. 263 Mouritzen, Hans (1996): `Denmark'. In Mouritzen, Hans, Ole Wver and Hkan Wiberg (eds.): European Integration and National Adaptations. A Theoretical Inquiry, Nova Science Publishers, Inc. Mouritzen, Hans (ed.) (2003): Er vi s forbeholdne? Danmark overfor globaliseringen, EU og det nre, rhus Universitetsforlag. Mouritzen, Hans, Ole Wver and Hkan Wiberg (eds.) (1996): European Integration and National Adaptations. A Theoretical Inquiry, Nova Science Publishers Inc. Nedergaard, Peter (2001): Organiseringen af den Europiske Union, Handelshjskolens forlag, Third edition. Nedergaard, Peter (2004): EU Leksikon, Jurist- og konomforbundets Forlag. Nehring, Niels Jrgen (2000): `Europapolitik og -debat efter 28. september', Udenrigs, Vol. 4, pp. 1-11. Netjes, Catherine E. and Kees van Kersbergen (2005): `Interests, Identity and Political Allegiance in the European Union', Paper presented at the Euroscepticism conference in Amsterdam, July 1-2. Nicoll, William Sir and Trever C. Salmon (2001): Understanding the European Union, Pearson Education Limited. Niedermayer, Oskar (1991): `Public Opinion about the European Parliamenť. In Reif, Karlheinz and Ronald Inglehart (eds.): Eurobarometer: The Dynamics of European Public Opinion, London: Macmillan, pp. 27-44. Niedermayer, Oskar and Bettina Westle (1995): `A Typology of Orientations'. In Niedermayer, Oskar and Richard Sinnott (eds.): Public opinion and internationalized Governance, Oxford University Press, pp.33- 50. Niedermayer, Oskar and Richard Sinnott (eds.) (1995): Public opinion and internationalized Governance, Oxford University Press. Nielsen, Rasmus Leander and Laura Lund Olsen (2006): `Why Did the French Reject the European Constitutional Treaty?', CFES Working Paper, No. 26. Norris, Pippa (1999): `The Political Regime'. In: Schmitt, Hermann and Jacques Thomassen (eds.): Political Representation and Legitimacy in the European Union, Oxford University Press, pp. 74-89. Norris, Pippa (1999): Critical Citizens, Oxford University Press. Nugent, Neil (2003): The government and politics of the European Union, The European Union Series, Palgrave, Fifth edition. O'Brien, Conor Cruise (1997): `Pursuing a Chimera'. In Gowan, Peter and Perry Anderson (eds.): The Question of Europe, London: Verso, pp. 77-84. Obradovic, Daniela (1996): `Policy Legitimacy and the European Union', Journal of Common Market Studies, Vol. 34 (2). Olsen, Tore Vincents (2006): Union Philosophies. A study of different approaches to the European Union's public philosophy identified in the Convention on the Future of Europe, PhD Thesis, Copenhagen University. 264 Page, Benjamin and Robert Shapiro (1992): The Rational Public: Fifty Years of Trends in Americans' Policy Preferences, The University of Chicago Press. Paldam, Martin and Peter Nannestad (2000): `What do voters know about the economy? A study of Danish Data', Electoral Studies, Vol. 19 (2-3), pp. 363-391. Panebianco, Stefania (1996): `European Citizenship and European Identity: from the Treaty of Maastricht to public opinion attitudes', Jean Monnet Working Papers in Comparative and International Politics, No. 3, University of Catania. Pedersen, Thomas (ed.) (2002): Europa for folket? EU og det danske demokrati, Magtudredningen, Aarhus Universitetsforlag. Pelinka, Anton (2005): `Austrian Euroscepticism: The Shift From the Left to the Righť. In: Harmsen, Robert and Menno Spiering (2004): Euroscepticism ­ Party politics, national identity and European integration, Rodopi: Amsterdam, New York, pp. 207-224. Pierce, Roy (1983): `Referendum Voting Behavior: The Norwegian and British Referenda on Membership in the European Community', American Journal of Political Science, Vol. 27 (1), pp. 43-63. Plattner, Marc F. (2003): `Competing Goals, Conflicting Perspectives', Journal of democracy, Vol. 14 (4), pp. 42-56. Pollack, Mark A. (2000): `A Blairite Treaty: Neo-Liberalism and Regulated Capitalism in the Treaty of Amsterdam'. In Neunreither, Karlheinz and Antje Wiener (eds.): European Integration After Amsterdam. Institutional Dynamics and Prospects for Democracy, Oxford University Press, Chapter 14. Portillo, Michael (1998): Democratic Values and the Currency, London: IEA. Puchala, Donald (1972): `Of Blind Men, Elephants and International Integration', Journal of Common Market Studies, Vol. 10 (3). pp. 267 ­ 284. Quaglia, Lucia (2003): `Euroscepticism in Italy and centre-right and right wing political parties', Sussex European Institute Working Paper, No. 60. Radaelli, Claudio (2006): `Europeanization: Solution or Problem?, In Michelle Cini and Angela K. Bourne (eds.): Palgrave Advances in European Studies, Basingstoke: Palgrave Macmillan, pp. 56-76. Ragin, Charles C. (1994): Constructing Social Research, Thousand Oaks: Pine Forge Press. Ragin, Charles, C. and Howard S. Becker (eds.) (1992): What is a case? Exploring the Foundations of Social Inquiry, Cambridge University Press. Rasmussen, Jens Rahbek and Nils Arne Srensen (1997): Briterne og Europa, Danish Institute of International Affairs. Rasmussen, Morten (2005): `The Hesitant European. History of Denmark's Accession to the European Communities 1970-73', Journal of European Integration History, Vol. 11 (2), pp. 47-74. 265 Rasmussen, Sren Hein (1997): Sre Alliancer. Politiske bevgelser i efterkrigstidens Danmark, Odense Universitetsforlag. Rattinger, Hans (1994): `Public attitudes to European integration in Germany after Maastricht. Inventory and Typology', Journal of Common Market Studies, Vol. 32 (4), pp. 525-540. Ray, Leonard (2003): `Reconsidering the Link between Incumbent Support and Pro-EU Opinion', European Union Politics', Vol.4 (3), pp. 259-279. Ray, Leonard (2007): `Public Opinion, Socialization, and Political Communication'. In Jrgensen, Knud Erik, Mark Pollack and Ben Rosamond (eds.): Handbook of European Union Politics, Sage Publications. (Note: As I was not able to obtain a copy of this recent book prior to deadline, page-numbers refer to Ray's draft chapter. Reif, Karlheinz and Hermann Schmitt (1980): `Nine Second Order National Elections: A Conceptual Framework for the Analysis of European Election Results', European Journal of Political Research, Vol. 8 (1), pp. 3-44. Reif, Karlheinz and Ronald Inglehart (eds.) (1991): Eurobarometer: The Dynamics of European Public Opinion, London: Macmillan. Ricard-Nihoul, Gaëtane (2005): `Le "non" français du 29 mai 2005: comprendre, agir', Etudes et recherches, No. 44, Paris: Notre Europe. Riishj, Sren (2004): `Europeanisation and Euro-scepticism. Experiences from Poland and the Czech Republic', Central European Political Studies Review, Vol. 4 (6), pp. 1-22. Risse, Thomas (2004): `Neofunctionalism, European Identity, and the Puzzles of European Integration', Journal of European Public Policy, Vol. 12 (2), pp. 291-309. Rohrschneider, Robert (2002): `The Democratic Deficit and Mass Support for an EU-wide Governmenť, American Journal of Political Science, Vol. 46 (2), pp. 463-475. Rosamond, Ben (2000): Theories of European Integration, Macmillan. Rosamond, Ben (2005): `The uniting of Europe and the foundation of EU studies: revisiting the neofunctionalism of Ernst B. Haas, Journal of European Public Policy, Vol. 12 (2), pp. 237-254. Santos, J. Reynaldo (1999): `Cronbach's Alpha: A Tool for Assessing the Reliability of Scales', Journal of Extension, Vol. 37 (2), Online version. Sartori, Giovanni (1970): `Concept misformation in comparative politics', American Political Science Review, Vol. 64 (4), pp. 1033-53. Sartori, Giovanni (1974): `Philosophy, Theory and Science of Politics', Political Theory, Vol. 2 (2), pp. 133-162. Sartori, Giovanni (1991): `Comparing and Miscomparing', Journal of Theoretical Politics, Vol. 3 (3), pp. 243-257. 266 Sartori, Giovanni (1997): `Understanding Pluralism', Journal of Democracy, Vol.8 (4), pp. 58-69. Sartori, Giovanni (ed.) (1984): Social Science Concepts. A Systematic Analysis, Beverly Hills, London, New Delhi: Sage Publications. Scharpf, Fritz (1999): Governing in Europe: Effective and Democratic?, Oxford University Press. Scharpf, Fritz W. (2003): `Problem-Solving Effectiveness and Democratic Accountability in the EU', Max Planck Institute für Gesellschaftsforschung, Working Paper, No. 03/1. Scheuer, Angelika (1999): `A Political Community'. In: Schmitt, Hermann and Jacques Thomassen (eds.): Political Representation and Legitimacy in the European Union, Oxford University Press, pp. 25-46. Scheuer, Angelika (2003): `Is knowing it loving it? How people's views on the EU are shapeď, Revue de la Maison française ďOxford, Vol. 1 (1), pp. 169-188. Schild, Joachim (2001): `National v. European Identities? French and Germans in the European MultiLevel System', Journal of Common Market Studies, Vol. 39 (2), pp. 331-351. Schmitt, Hermann (2003): `The Eurobarometers', European Union Politics, Vol. 4 (2), pp. 243-251. Schmitt, Hermann (2005): `The European Parliament Elections of June 2004: Still Second-Order?', West European Politics, Vol. 28 (3), pp. 650-679. Schmitt, Hermann and Jacques Thomassen (eds.) (1999): Political Representation and Legitimacy in the European Union, Oxford University Press. Schmitter, Philippe C. (2001): `What is there to legitimize in the EU... And how might this be accomplished?', European University Institute, January. Schmitter, Philippe C. (2005): `Ernst B. Haas and the legacy of neofunctionalism', Journal of European Public Policy, Vol. 12 (2), pp. 255-272. Schneider, Gerald and Patricia Weitsman (1996): `The Punishment Trap: Integration Referendums as Popularity Contests', Comparative Political Studies, Vol. 28 (4), pp. 582-607. Schwellnus, Guido (2001): `"Much ado about nothing?" Minority Protection and the EU Charter of Fundamental Rights', Constitutionalism Web-Papers, ConWeb, No. 5. Seawright, Jason and David Collier (2004): `Glossary'. In: Brady, Henry E. and David Collier (eds.): Rethinking social inquiry: Diverse Tools, Shared Standards, Lanham: Rowman and Littlefield, pp. 273-313. Sharpe, L. J. (1996): `British Scepticism and the European Union: A Guide for Foreigners', In Holmes, Martin (ed.): The Eurosceptical Reader, Palgrave Macmillan, pp. 261-268. Sinnott, Richard (1994): `Integration theory, subsidiarity and the internationalization of issues: the implication for legitimacy', Paper presented at the Second ECSA World Conference, Brussels, May 5-6. Sinnott, Richard (1995): `Bringing Public Opinion Back In'. In Niedermayer, Oskar and Richard Sinnott (eds.): Public opinion and internationalized Governance, Oxford University Press, pp. 11-32. 267 Sitter, Nick (2002): `Opposing Europe: Euro-scepticism, opposition and party competition, Sussex European Institute, Working Paper, No. 56. Siune, Karen (1993): `The Danes said NO to the Maastricht Treaty. The Danish EC Referendum of June 1992', Scandinavian Political Studies, Vol. 16 (1), pp. 93-103. Siune, Karen, Palle Svensson and Ole Tonsgaard (1992): Det blev et nej, Politica, rhus. Siune, Karen, Palle Svensson and Ole Tonsgaard (1994a): Fra et nej til et ja, Politica, rhus. Siune, Karen, Palle Svensson, and Ole Tonsgaard (1994): `The European Union: The Danes Said `No' in 1992 but `Yes' in 1993: How and Why?', Electoral Studies, Vol. 13 (2), pp. 107-116. Smith, Anthony D. (1991): National Identity, London: Penguin. Smith, Anthony D. (1992): `National Identity and the Idea of European Unity', International Affairs, Vol. 68 (1), pp. 55-76. Smith, Anthony D. (2005): `Set in the Silver Sea': English National Identity and European Integration', Paper presented at Workshop: National Identity and Euroscepticism: A Comparison Between France and the United Kingdom, Centre for the Study of Democratic Government, Oxford University, May 13. Srensen, Henning and Ole Wver (1992): `State, society and democracy and the effect of the EC'. In Lise Lyck (ed.): Denmark and EC Membership Evaluated, New York: St. Martin's Press, pp. 3-25. Srensen, Catharina (2004): `Danish and British Popular Euroscepticism Compared: A Sceptical Assessment of the Concepť, Danish Institute for International Studies, Working Paper No. 25. Srensen, Catharina (2006): `Elsker, elsker ikke? Euroskepsis i forandring'. In Anne Mette Vestergaard and Malene Fenger-Grndahl (eds.): EU's Dilemmaer ­ Hvad sker der?, Informations Forlag. Spiering, Menno (2004): `British Euroscepticism'. In: Harmsen, Robert and Menno Spiering (2004): Euroscepticism ­ Party politics, national identity and European integration, Amsterdam, New York: Rodopi, pp. 127-150. Spinelli, Altiero (1967): `European Union in the Resistance', Government and Opposition, Vol. 2, pp. 321- 329. Stone Sweet, Alec and Wayne Sandholtz (1997): `European integration and supranational governance', Journal of European Public Policy, Vol. 4 (3), pp. 297-317. Sudbery, Imogen and Brigid Laffan (2006): `Politicisation ­ an emerging dynamic in EU integration research', Annual State of the art paper of EU-Consent WP II/III, EU-Consent. Surel, Yves (2000): `Ľintégration européenne vue par ľapproche cognitive et normative des politiques publiques', Cahiers européen de Sciences Po, No. 2. Svensson, Palle (2002): `Five Danish referendums on the European Community and European Union: A critical assessment of the Franklin thesis', European Journal of Political Research, Vol. 41, pp. 733-750. 268 Szczerbiak, Aleks (2001): `Polish Public Opinion: Explaining Declining Support for EU Membership', Journal of Common Market Studies, Vol. 39 (1), pp.105-122. Szczerbiak, Aleks (2002): `After the election, nearing the endgame: The Polish euro-debate in the run up to the 2003 EU accession referendum', Sussex European Institute, Working Paper, No. 53. Szczerbiak, Aleks and Paul Taggart (2003): `Theorising Party-Based Euroscepticism: Problems of Definition, Measurement and Causality', Sussex European Institute, Working Paper, No. 69. Taggart, Paul (1998): `A Touchstone of Dissent: Euroscepticism in Contemporary Western European Party Systems', European Journal of Political Research, Vol. 33, pp. 363-388. Taggart, Paul (2006): `Keynote Article: Questions of Europe ­ The Domestic Politics of the 2005 French and Dutch Referendums and their Challenge for the Study of European Integration', Journal of Common Market Studies, Vol. 44, Annual Review, pp. 7-25. Taggart, Paul and Aleks Szczerbiak (2002): `The party politics of euroscepticism in EU member and candidate states', Sussex European Institute, Working Paper, No. 51. Taggart, Paul and Aleks Szczerbiak (2005): `Three Patterns of Party Competition Over Europe', Paper presented at the conference `Euroskepticism--Causes and Consequences', Amsterdam, July 1-2. Takle, Marianne (2002): `German Policy on Immigration and European Integration: Two challenges to the nation-state and the role of the political community in the democratic process', Arena-seminar, March 19. Tamvaki, Dionysia (2003): `The Copenhagen Criteria and the evolution of popular consent to EU norms: from legality to normative justifiability in Poland and the Czech Republic', EUI workshop, November. Teschner, Julia (2000): `No longer Europe's Europhiles? Euroscepticism in Germany in the 1990s', European Integration, Vol. 22, pp. 59-86. Theiler, Tobias (2004): `The origins of Euroscepticism in Germany-speaking Switzerlanď, European Journal of Political Research, Vol. 43 (4), pp. 635-656. Thompsen, Birgit N. (ed.) (1993): The odd man out: Danmark og den Europiske Integration 1948-1992, Odense Universitetsforlag. Thomsen, Sren Risbjerg (1998): `Meningsmlingernes Trovrdigheď, Politica, Vol. 30 (2), pp. 204-227. Togeby, Lise (1998): Man har et standpunkt... Om stabilitet og forandring i befolkningernes holdninger, Magtudredningen, Aarhus Universitetsforlag. Uffe stergaard (2000): Europa. Identitet og identitetspolitik, Copenhagen: Rosinante. Usherwood, Simon (2001): `Opposition to the European Union in the UK and France', Paper first presented at the ECPR First General Conference, University of Kent at Canterbury, September 6-8. 269 Usherwood, Simon (2002): `Opposition to the European Union in the UK', Government and Opposition, Vol. 37 (2), pp. 211-230. Usherwood, Simon (2004): `Left-wing opposition to the European Union in the UK: Explaining its under-developmenť, Paper presented at the Political Studies Association 54th Annual conference, University of Lincoln, April 6-8. Vestergaard, Anne Mette and Catharina Srensen (2004): studvidelsen og arbejdskraften: Myter og realiteter, Danish Institute for International Studies. Vreese, Claas de and Hajo Boomgaarden (2006): `Media Effects on Public Opinion about the Enlargement of the European Union', Journal of Common Market Studies, Vol. 44 (2), pp. 419-36. Vreese, Claes de (2004): `Why European Citizens Will Reject the EU Constitution', Center for European Studies, Working Paper, No. 116. Vreese, Claes H. de and Holli A. Semetko (2004): `Influences on the vote in the Danish 2000 euro referendum campaign', European Journal of Political Research, Vol.43, pp. 699-722. Wver, Ole (1996): `European security identities', Journal of Common Market Studies, Vol. 34 (1), pp. 103- 131. Wagner, Markus (2005): `Euroscepticism in the French Socialist Party: Opposition with a Soft Voice?', Paper for UACES Student Forum, 6th Annual Conference, Oxford, April 7-8. Wallace, William (1997): `The Nation-State ­ Rescue or Retreat?' In Gowan, Peter and Perry Anderson (eds.): The Question of Europe, London: Verso, pp. 21-50. Weber, Max (1968 [1921]): Economy and Society: An Outline of Interpretative Sociology, Berkeley, CA: University of California Press. Weiler, Joseph (1995a): `The State "über alles". Demos, Telos and the German Maastricht Decision', NYU School of Law, Jean Monnet Papers. Weiler, Joseph (1995b): `Europe after Maastrichť, Jean Monnet Program Paper, NYU School of Law. Weiler, Joseph et al. (1995c): `European democracy and its critique', West European Politics, Vol. 18 (3), pp. 4-39. Weiler, Joseph H. H. (1998): `Europe. The Case Against the Case for Statehooď, European Law Journal, Vol. 4 (1), pp. 43-62. Weiler, Joseph H. H. (1999): The Constitution of Europe. "Do the new clothes have an emperor" and other essays on European integration, Cambridge University Press. Weiler, Joseph, Ian Begg and John Peterson (2003): Integration in an expanding EU. Reassessing the fundamentals, Blackwell. Wendt, Alexander (1998): `On constitution and causation in international relations', Review of International Studies, Vol. 24, Special issue, pp. 101-117. 270 Wendt, Alexander (1999): Social Theory of International Politics, Cambridge University Press. Wessels, Bernhard (2005): `Euroskepticism ­ Instrumental, Anti-European, or just Nescience', Paper presented at the conference: Euroskepticism--Causes and Consequences, Amsterdam, July 1-2. Wessels, Bernhard (1995): `Development of support: Diffusion or demographic replacement?'. In Niedermayer, Oskar and Richard Sinnott (eds.): Public opinion and internationalized Governance, Oxford University Press, pp.105-136. Wessels, Bernhard (2000): `Interests Groups in the EU: the Emergence of Contestation Potentiaľ, Paper presented at the conference: Dimensions of Contestation in the European Union, Chapel Hill, April 30 ­ May 2. Weyland, Kurt (2001): `Clarifying a Contested Concept: Populism in the Study of Latin American Politics', Comparative Politics, Vol. 34 (1), pp. 1-22. Whitman, Richard (2005): `No and after: options for Europe', International Affairs, Vol. 81 (4), pp. 673- 687. Wiberg, Hkan (1994): `Europe. The Western Project and the Hesitant North', Research Institute for European Studies, Research Paper, No. 7. Wiener, Antje and Thomas Diez (eds.) (2004): European integration theory, Oxford University Press. Wind, Marlene (2001): `The Commission White Paper. Bridging the gap between the governed and the governing?', A paper forming part of the contributions to the Jean Monnet Working Paper No.6/01, edited by C. Joerges, Y. Mény and J. H. H. Weiler. Wind, Marlene (2003): Den stille revolution ­ mod en europisk forfatning, Copenhagen: Forlaget Politiske Studier. Wind, Marlene (2006): Den Europiske Forfatningskamp ­ fr, under og efter 2 x nej, Forlaget Thomson. Winter, Lieven De and Marc Swyngedouw (1999): `The Scope of EU Governmenť. In Schmitt, Hermann and Jacques Thomassen (eds.): Political Representation and Legitimacy in the European Union, Oxford University Press, pp. 47-73. Wivel, Anders (ed.) (1998): Explaining European Integration, Copenhagen Political Studies Press. Worcester, Robert (1999): `Europe: The State of Public Opinion', Paper presented at the European Journalists' Seminar, European Parliament Office, London, February 26. Worre, Torben (1995): `First No then Yes: The Danish Referendums on the Maastricht Treaty 1992 and 1993', Journal of Common Market Studies, Vol. 33 (2), pp. 235-257. Zetterholm, Staffan (ed.) (1994): National Cultures and European Integration: Exploratory Essays on Cultural Diversity and Common Policies, Oxford: Berg. Zuckerman, Alan (1975): `Political Cleavage: A Conceptual and Theoretical Analysis', British Journal of Political Science, Vol. 5 (2), pp. 231-248. 271 Appendix: One-dimensional Eurobarometer indicators The below table lists one-dimensional indicators for all the thesis' expectations of euroscepticism, as well as for the level of hard euroscepticism. It is based on my examination of the list of contents and annexes of the Standard Eurobarometer polls--from Eurobarometer 1 (1974) to Eurobarometer 64 (2005). Eurobarometers 2, 5, 6, 7, 8, 9 are not available. 272 Eurobarometer number: 1 3 4 10 11 12 13 14 15 16 17 Indicators with reply categories relevant for more than one type of euroscepticism Meaning of the EU Fears related to integration X EU-Priorities Maastricht Treaty: specific areas For/against specific EU issues Perceived role of the EU 1: statements 2: Constitution Propositions about the EU Economic euroscepticism Benefit from membership X Effect of the EU in specific areas Performance-based euroscepticism Effectiveness of EU-policies Democratic euroscepticism Satisfaction with EU-democracy EP's ability to protect citizens 1:Is EU democratic? 2:Dem deficit Feeling that voice counts in EU Social euroscepticism Support for a social charter Sovereignty-based euroscepticism European government Support for Political Union X X Decision-making level X X X X Right of veto 1:Integration threats id. 2: culture 2 EP preference/role 2: Full-fledged EU now Support for 1:EU Tax 2: U.S.E. Unification of Western Europe X X X X X X X X X Level of hard euroscepticism Membership: good/bad X X X X X X X X X X X X: EU feeling 2:Sad if EUdissolved Abstention at EP-election (reason) EU-referendum (Maastricht ref) X? Regret if EU was scrapped X X X 273 Continued... 18 19 20 21 22 23 24 25 26 27 28 29 30 Indicators with reply categories relevant for more than one type of euroscepticism Meaning of the EU X Fears related to integration EU-Priorities Maastricht Treaty: specific areas For/against specific EU issues Perceived role of the EU 1: statements 2: Constitution Propositions about the EU Economic euroscepticism Benefit from membership X X X X X X X X X X X X X Effect of EU in specific areas X Performance-based euroscepticism Effectiveness of EU-policies Democratic euroscepticism Satisfaction with EU-democracy EP's ability to protect citizens 1:Is EU democratic? 2:D. deficit Feeling that voice counts in EU Social euroscepticism Support for a social charter Sovereignty-based euroscepticism European government X X X? Support for Political Union X X Decision-making level X X Right of veto 1:Integration threats id. 2: cult. 1? 1? EP preference/role X X X X X X X X 2: Full-fledged EU now 2 Support for 1:EU Tax 2: U.S.E. 2 2 Unification of Western Europe X X X X X X X X X X Level of hard euroscepticism Membership: good/bad X X X X X X X X X X X X X X: EU feeling 2:Sad if dissolved 2 Abstention at EP-election X EU-referendum (Maastricht ref) X Regret if EU was scrapped X X 274 Continued... 31 32 33 34 35 36 37 38 39 40 41 42 43 Indicators with reply categories relevant for more than one type of euroscepticism Meaning of the EU Fears related to integration X EU-Priorities Maastricht Treaty: specific areas X X X X For/against specific EU issues X X X Perceived role of the EU 1: statements 2: Constitution Propositions about the EU Economic euroscepticism Benefit from membership X X X X X X X X X X X X X Effect of EU in specific areas Performance-based euroscepticism Effectiveness of EU-policies Democratic euroscepticism Satisfaction with EU-democracy X X X X X EP's ability to protect citizens X X 1:Is EU democratic? 2:D. deficit 1 2 Feeling that voice counts in EU Social euroscepticism Support for a social charter X X X X X X Sovereignty-based euroscepticism European government X X X X X? X X X X X X Support for Political Union X? Decision-making level X X X X X X X X X X Right of veto 1:Integration threats id. 2: cult. 1 1 EP preference/role X X X X X X X X X X X 2: Full-fledged EU now Support for 1:EU Tax 2: U.S.E. 1 Unification of Western Europe X X Level of hard euroscepticism Membership: good/bad X X X X X X X X X X X X X X: EU feeling 2:Sad if dissolved Abstention at EP-election X X EU-referendum (Maastricht ref) 2 Regret if EU was scrapped X X X X 275 Continued... 44 45 46 47 48 49 50 51 52 53 54 55 56 Indicators with reply categories relevant for more than one type of euroscepticism Meaning of the EU X Fears related to integration X X X X EU-Priorities X X X X X X X X Maastricht Treaty: specific areas For/against specific EU issues X Perceived role of the EU 1: statements 2: Constitution 1 Propositions about the EU Economic euroscepticism Benefit from membership X X X X X X X X X X X X X Effect of EU in specific areas Performance-based euroscepticism Effectiveness of EU-policies Democratic euroscepticism Satisfaction with EU-democracy X X X X X X X EP's ability to protect citizens X X X X X X 1:Is EU democratic? 2:D. deficit Feeling that voice counts in EU Social euroscepticism Support for a social charter Sovereignty-based euroscepticism European government X X X Support for Political Union Decision-making level X X X X X X X X Right of veto 1:Integration threats id. 2: cult. EP preference/role X X X X X 2: Full-fledged EU now Support for 1:EU Tax 2: U.S.E. Unification of Western Europe X X Level of hard euroscepticism Membership: good/bad X X X X X X X X X X X X X X: EU feeling 2:Sad if dissolved Abstention at EP-election X X EU-referendum (Maastricht ref) X Regret if EU was scrapped X X 276 Continued... 57 58 59 60 61 62 63 64 Indicators with reply categories relevant for more than one type of euroscepticism Meaning of the EU X? X X X X X Fears related to integration X X X X X EU-Priorities X X X X X X X Maastricht Treaty: specific areas For/against specific EU issues Perceived role of the EU X 1: statements 2: Constitution 2 Propositions about the EU X Economic euroscepticism Benefit from membership X X X X X X X X Effect of EU in specific areas Performance-based euroscepticism Effectiveness of EU-policies X X Democratic euroscepticism Satisfaction with EU-democr. X X X X X X EP's ability to protect citizens /X /X /X 1:Is EU democratic? 2:D. deficit Feeling that voice counts in EU X X Social euroscepticism Support for a social charter Sovereignty-based euroscepticism European government Support for Political Union X X X X Decision-making level X X X X X X Right of veto X X X X X 1:Integration threats id. 2: cult. EP preference/role 2: Full-fledged EU now Support for 1:EU Tax 2: U.S.E. 1 1 Unification of Western Europe Level of hard euroscepticism Membership: good/bad X X X X X X X X X: EU feeling 2:Sad if dissolved X X X X X Abstention at EP-election EU-referendum (Maastricht ref) Regret if EU was scrapped X X X 277