wp-83.pdf: the formation of a european electorate

The Formation of a European Electorate

Evidence from Electoral Volatility
Measures, 1970s – 2000s
Daniele Caramani Daniele Caramani The Formation of a European Electorate
Evidence from Electoral Volatility Measures, 1970s – 2000s

Arbeitspapiere – Working Papers
Nr. 83, 2004
Mannheimer Zentrum für Europäische Sozialforschung
Caramani, Daniele:
The formation of a european electorate : evidence from electoral volatility measures, 1970s – 2000s / Daniele Caramani. – Mannheim : MZES, 2004 (Arbeitspapiere - Mannheimer Zentrum für Europäische Sozialforschung ; 83) Not available in book shops. Token fee: € 2,60 Purchase: Mannheimer Zentrum für Europäische Sozialforschung (MZES), D – 68131 Mannheim WWW: http://www.mzes.uni-mannheim.de Daniele Caramani is a research fellow at the Mannheim Centre for European Social Research (University of Mannheim). He received the Ph.D. from the European University Institute (Florence), and has taught previously at the universities of Geneva and Florence. In 2000-02 he was "Vincent Wright Fellow in Comparative Politics" at the Robert Schuman Centre for Advanced Studies (EUI). His research interests include party systems and elections in a comparative and historical perspective, European integration, methodology, and regionalism. He is the author of the book and CD-ROM Elections in Western Europe since 1815 (Palgrave, 2000), and The Nationalization of Politics (Cambridge University Press, 2004), as well as of several international journal articles. I would like to thank Franz Urban Pappi and Hermann Schmitt for their comments on an earlier version of this paper. This paper investigates the thesis according to which the formation of a European electorate which is distinct from national electorates in the European Union member-states translates into an increasing differentiation of the EU party system from national party systems. The main indicators used to test for the existence of an autonomous EU political space are indices of electoral volatility between national and European elections. Data include results of national elections over the last 30 years and European elections from the first election to the European Parliament in 1979 until 2004 for all current member-states. Evidence shows a persistently similar electoral behaviour in the national and European arenas indicating the predominant salience of national issues, parties, and alignments. In interpreting these results in a comparative historical sociology perspective, the paper argues that a "Europeanisation" of electoral politics comparable to its "nationalisation" in the nineteenth and twentieth century is unlikely because of the absence of similar social and political mobilisation Key-words: National and EP elections Volatility Parties European integration Contents
1 Introduction
After the first direct election to the European Parliament (EP) in 1979, Reif and Schmitt (1980) described European elections in terms of "second-order elections." Since then this label implies in the first place that European elections are (perceived as) less important than national or even local elections in the member-states of the European Union (EU). On the other hand, "second-order" also means that European elections are dominated by national factors, with European electorates voting for the EP according to national criteria, with electoral campaigns and media dominated by national issues, with voters being guided in their voting choice by national party affiliations and electoral alignments, and with leaders seeking support on national policy platforms and governmental action or – in the case of opposition parties – seeking support for their arguments against incumbents. In sum, European elections as national politics "with other means." A large part of subsequent research has since then repeatedly supported this view. Most studies on voters' choice in European elections have found that they are primarily national political contests.
Also authors concerned more directly with the organisational dimension of political parties have stressed the structural weakness of European party federations as well as their lack of cohesion in the EP. continues to be predominant in the European electoral contest: transnational parties are lose organisations which do not play any significant role in the nomination of candidates, as well as ideologically and programmatically heterogeneous (Anderweg 1995). To many, therefore, European parties appear as the "empy vessels" described by Katz and Kolodny (1994) concerning American parties, or "baskets of parties" (Mair 200 Such a debate is not purely academic. Without a European electorate and party system – and without cohesive and concurrent European parties – direct elections to the EP cannot provide adequate channels for democratic accountability. It is therefore a debate with strong normative implications concerning the "democratic deficit" ofe democratic confrontation between alternative platforms and policies requires a contest between accountable party fronts competing for votes and facing the sanction by a European electorate. Such a process can develop only together with the creation of an autonomous political space independent from national issues, leaders, and structures. However, the emergence of a "truly European" party system (Anderweg 1995: 67) looks unlikely to 1 For the most complete review and analysis of European elections, see Van der Eijk and Franklin (1996). 2 See Hix (2002) for an overview and review of the literature since the 1970s. Kreppel (2002) argues that European party leaderships have no means to sanction dissent from national fractions concerning candidates' nominations (even though in both the EPP and PES there is a tendency to circumvent the control by national parties by institutionalising electoral processes in the EP). For Pedersen, the absence of "genuine" parties at the level of the EU translates in the lack of stable organisations stucturing the electoral process, "organizations that span and control the electoral linkage" (1996: 17). 3 "The Europarties that emerge in the EP are much more akin to the notion of the basket of parties, being juxtaposed to one another rather than competing with one another in any predictable sense" (Mair 2000: 39). 4 Marquand (1978) was the first to stress that a democratic European polity could only emerge with a structured party system and European-wide electoral alignments replacing politics based on national identities. On most observers, not least because it would face the natural resistance of national parties, but mostly because the development of a European-wide electoral competition and cleavage constellation would require important institutional reforms, with the creation of an elected executive and parliamentary control (Mair 2000: 38). Yet after six direct European elections it is legitimate to ask if there has been some sort of trend over the last 25 years. This paper therefore addresses the question of whether or not European electorates perceive the national and European arenas as increasingly distinct, and whether or not their electoral behaviour increasingly variates between the two national and European orders of elections. Many factors support the plausibility of such a hypothesis. In the course of the 1980s and 1990s, the EP has acquired new (and increased previous) competences. Recent investigations have found that decision- making mechanisms are systematically affected by European parties, and that the structures of party groups in the EP have developed and consolierally, the whole process of European integration has moved at an accelerated pace in the last decade, and the integration versus anti- integration dimension has gained strength in public debates in all countries in the wake of referenda on accession, single currency, and treaties (Hug 2002), initiatives of the Commission (White Paper on Governance), and more generally with the process of "constitutionalisation" of the EU (Weiler 1999). This paper wishes to contribute to this debate in three ways. First, it places the six European elections into a time perspective beginning in the early 1970s and ending with the most recent national elections in 2003–04 and the election to the EP in 2004, as well as in cross-country comparative perspective which includes all 25 EU member-states. Second, the paper focuses on voters and their voting behaviour and introduces an indicator that has rarely been applied to the comparison of national and European elections, namely, electoral volatilityper investigates the theoretical, methodological, and normative implications of empirical results in a comparative historical sociology perspective, by comparing the conditions for the formation of broader spaces of party competition in Europe with earlier developments during the formation of national electorates and party systems in the nineteenth and early twentieth century. This allows to move the first steps towards a tentative "theory of electoral integration" and sets the bases for future analyses. democratic deficit see also Mény (2003). For an overview of the question of whether a "Europe of the parties" has emerged, see Hix and Lord (1997), Kreppel (2002), and Raunio (2002). 5 See Hix and Lord (1997: 18). Kreppel speaks of a "transformation of the EP from a chamber of debate to a legislative body" (2002: 151). Hix, Kreppel, and Noury (2003) come to the conclusion that the European party system is "ready for power." 6 Compared to party organisations and parliamentary groups in the EU, the electoral dimension is often neglected in analyses of the party system at the EU level. For example, Hix, Kreppel, and Noury (2003) distinguish simply an organisational and a competitive dimension in the EP, leaving totally aside the electoral process. Approaches to the Formation of a European Electorate
2.1 First- Versus Second-Order Elections
Analyses of a forming EU electorate and party system conducted in the past decades through the comparison of European and national elections usually consider two main dimensions of electoral behaviour: (1) the support for political parties and (2) electoral participation (turnout). In particular, these two dimensions appear in studies that have compared different orders of elections – between first- and second-order elections – with the goal to explain why voters turn out and vote differently in different types of elections or when elections are held at different times. In this perspective a number of writings have analysed mid-term elections in the U.S. and advanced explanations concerning the (supposed) weaker performance of incumbents compared to Congress elections in Presidential election years. One of these theories – the theory of "surge and declinten stimulated authors to describe the forming EU party system by comparing it to the U.S (even though the theory itself did not receive much support from empirical evidence).
Indeed, in Europe research on different orders of elections has proved useful in analysing elections to the EP.s work has pointed to the lack of salience of second-order (European) elections in the eyes of European electorates insofar as they do not lead to the selection of executives and that the EP has had only limited power (Reif 1985b). According to this theory this led to (1) lower rates of turnout in elections to the EP, (2) a loss of votes for the major parties in European elections, as well as (3) poorer performances of incumbents. Because in European elections there is "less at stake" the propensity to vote sincerely increases, that is, disregarding tactical or strategic considerations which would on the contrary incentivate to vote for large and/or incumbent parties that have a realistic chance to win office.wever, research conducted on the emergence of EU or European-wide issues that would account for a deviating electoral behaviour in elections to the EP with respect to national elections, was unable to find striking supporting evidence (Kuechler 1991, Van der Eijk and Franklin 1991). Concerning turnout, additional explanatory factors to account for the few cases in which turnout in European elections approaches that of national elections, include compulsory voting (Blumler and Fox 1982), positive views on Europe (Van der Eijk and Oppenhuis 1990), and time until the next national election (Marsh and Franklin 1996: 19).
7 The theory of "surge and decline" was proposed by Campbell (1960 and 1966). It seeks to explain the performance of parties in presidential and mid-term elections. See also Niemi and Wiesberg (1993: 207–21). 8 This comparison is, for example, mentioned in Schmitter (2000: 70). For a more systematic juxtaposition (concerning the control of the nomination of candidates) see Raunio (2002: 273). 9 On second-order elections see – besides Reif and Schmitt (1980) – Reif (1985a, 1985b, 1997), Marsh (1998), and Van der Eijk and Franklin (1996). 10 Schmitt (1990) has described the tentency of avoiding the "waste" of votes as "voting with the heart." In a similar vein a number of authors interpret voting cycles between first-order (national) and second-order elections in terms of a punishment effect against incumbent parties in national governments (Erikson 1988, Reif 1984). 11 Two more elements are often present in the literature on second-order elections which also emerge in Campbell studies on surge and decline. First, the loss of votes for incumbent parties is caused by the lower 2.2 The Nation-State's "Model"
More recently a number of writings have attempted to adapt theories of formation and structuring of party systems at the national level to the EU party system. These attempts represent an important step forward insofar as they recuperate the rich conceptual and theoretical apparatus of what often goes under the label of "comparative historical sociology" in the wake of the work by Karl Deutsch, Stein Rokkan, and others. Theories of state formation and nation-building, cleavage structures, and centralisation, constitute in this view promising "models" for the interpretation of European unification (Klausen and Tilly 1997). Several of these writings look in particular at Stein Rokkan's macro-theory of state formation, nation- building, and mass politics in Europe as a powerful model for interpreting processes of system- building. As for nation-states, European unification can be understood as a process of external boundary-building and of dismantling internal boundaries – above all judicial and economic – combined with an increasing centralisation of decision-making structures and procedures. The interaction between external boundary-building and internal boundary-dismantling was developed from Hirschman's twin-concepts exit and voice and Rokkan's concept of political structuring (Caramani 2004: 15–particularly in the legal sphere, the reduction of "selective exit" through the jurisprudence of the European Court of Justice mainly, has led to the development of channels for the expression of voice (institutions for the representation of various social and territorial groups) as well as to the development of political oppositions, differentiations, and cleavages – namely, in the form of contrasts between nation-states.
The national dimension, however, is only one of the possible dimensions within a forming European- wide cleavage constellation. As Bartolini has stressed (2002: 130–55), territorial resistances to centre formation – economic, legal, cultural – will closely interact with functional differentiations within the new "higher level" forming system. In particular, the territorial dimension over more versus less EU political control or more versus less integration cuts across the traditional class or left–right dimension – the principal functional differentiation in all European party systems. The forming of a European turnout of the groups supporting incumbents in mid-term or European elections which on the contrary turn out in concomitance of the election of incumbents offices. Second, second-order elections are often interpreted as referenda on the performance of the incumbent government which, however, do not affect its survival (see Reif 1985b: 7–15). 12 Until now the process of dismantling internal boundaries has not affected the political and cultural dimensions (the other two major social sub-systems in Rokkan's theory based on Parsons' categories) to the same degree. See Flora (1999: 88–91) for a first outline of the explanatory potential of these categories if applied to the process of European unification. 13 See Weiler (1999) for this interpretation of Hirschman's scheme and more generally for the first application of the "exit" and "voice" concepts to European integration. See Bartolini (2002) for a development of the implications for European unification of the interaction of the two mechanism in a historical sociology perspective. electorate and party system will therefore depend strongly on whether or not the left–right dimension will impose itself over the "sovereignty dimensi Although many of these points may be disputed, the above contributions seem to agree on the importance of the territorial dimension in the forming EU political space, and indicate that it will play an important role in structuring political oppositions in Europe.level, processes of "electoral integration" consist of the transformation from territorial into functional politics oppositions progressively transform into – or are replaced by – "higher-level" alignments and cleavages. Among the theories of electoral integration seeking to explain the formation of wider spaces of party competition, the theory of the nationalisation of politics has provided the most complete set of concepts, hypotheses, and empirical material on the formation of national spaces of electoral competition. As its name reveals, this theory is mainly concerned with the integration of electorates and party systems at the national level, that is, the formation of national cleavage constellations and nation-wide attitudes, issues, and organisations. However, in this case too, it is possible to grasp ideas which were developed in the frame of the nation-state and propose a transposition to European unification. On the basis of work produced recently (Caramani 2004), it may therefore be fruitful to explore the possibility to adapt some of the categories of the theory of the "nationalisation" of electoral politics to interpret the process of "Europeanisation." The Persistence of National Politics in European Elections
Accordingly, the paper wishes to contribute to the debate by exploring the possibility of thinking of Europeanisation in the same terms as of nationalisation. The development of a European political space and the introduction of direct elections for a supranational parliament, stimulate the temptation to adapt the same categories and instruments of the nationalisation of politics to the emergence of a 14 Also according to Marks and Steenberger (2002) the interaction between cleavage "residues" from the nineteenth and twentieth century – especially the left–right dimension – and the pro/anti-European dimension will determine the nature and shape of the European-wide party system. Besides the main class dimension, as a second functional or sectoral dimension of differentiation, Schmitter (2000: 68) points to agriculture – a cleavage that has disappeared or has been incorporated in other alignments in national constellations – but that at the EU level is re-emerging as a consequence of the important resources for the Common Agricultural Policy controlled by the EU and through alliances of "integration losers" (economically weak groups and peripheral regions). 15 Schmitter points to the salience of the territorial dimension at the EU level in the form of regionalist movements within and across national borders (2000: 69). On the contrary, the religious dimension seems to have moved "outside" the European political space or, more precisely, as a common element characterising European identity in opposition to other border cultures. On the renaissance of the territorial dimension in Europe see also Kohler-Koch (1998). 16 Term "electoral integration" is used to describe processes of formation of wider spaces of party competition (Caramani 2003). The term therefore encompasses both the integration of national party systems and electorates (which took place in the nineteenth and early twentienth century), and the incipent integration of a European party system and electorate. 17 The theory of the nationalisation of electoral politics originated in the United States. Among the most significant writings see Schattschneider's classical book The Semisovereign People (1960), in particular Chapter 5 entitled "The Nationalization of Politics. A Case Study in the Changing Dimensions of Politics," pp. 78-96. Of particular relevance are Stokes' articles on the variance components model (1965 and 1967). For a European electorate and party system. However, the radical differences between the conditions of the nationalisation of elections and their possible Europeanisation make a direct application of the same tools problematic. In particular, the absence of European-wide parties and electoral lists and the confinement of parties, candidates, and campaigns into striclty national borders hinders the use of measures of "territorial homogenisation." For this reason a different indicator is used in this paper. 3.1 Indicators, Data, and Cases
This paper looks at the relationship between first- and second-order elections in Europe with an eye to its long-term evolution over the last 25–30 years, by using the "differential" between electoral behaviour in national and European elections as main indicator. Two dimensions of electoral behaviour are considered: (1) party support: this is measured through the levels of electoral volatility; (2) turnout levels, that is, the differences in the levels of electoral participation between national and European elections.
To measure the levels of volatility, the index of "total volatility" devised and used in Bartolini and Mair (1990) has been applied. This index is constructed by summing up the absolute differences (disregarding plus and minus signs) between the percentage of votes for each party in election "t" and the percentage in "t+1." The sum is then divided by two to avoid double counting (gains for one party meaning losses for another or several other parties), and to make the index vary between 0 and 10 Volatility indices have been computed always on the previous election. For example, the level of volatility in the British 1997 national election has been computed with respect to the 1992 national election. Similarly, the levels of volatility in the European election in Britain of 1999 has been computed with respect to the European election in Britain of 1994. And so on. Overall – by combining national and European elections – three types of volatility can be computed: • Volatility between national elections only (in the different graphs this type of volatility is represented through solid lines). • Volatility between European elections only (dashed lines). • Volatility between national and European elections: the levels of volatility in national elections have been computed with respect to the previous European election and, conversely, the levels of volatility in European elections have been computed with respect to the previous national election. This type of volatility is called here "mixed volatility" (and is represented through dotted lines). theoretical and methodological critique of Stokes' work see Katz (1973). Claggett, Flanigan, and Zingale (1984) propose the more articulated reflection on the dimensions composing the concept of nationalisation. 18 Because this dimension has already been discussed in a great deal of work the paper focusses mainly on the dimension of party support. 19 The construction of the index is described in Bartolini and Mair (1990: 20–21). The formula is the following: Total Volatility = ∑ Pi…n t – Pi…n t+1 / 2 where P is the percentage of votes for parties "i" to "n" in elections "t" and "t+1." Values of volatility are computed for each election. The number of countries considered varies over time according to accession (see Table 1). For the first 1979 election to the EP, there were 10 countries. Greece joined the EU in 1981 and voted to elect its representatives to the EP after accession. This has been considered together with the 1979 vote in the other nine countries. Similarly, Portugal and Spain voted in 1987 after accession and this vote has been considered together with the 1984 European election in the other 10 countries. The same applies to the subsequent round of accession in 1995 when Austria, Finland, and Sweden joined the EU reaching a total number of 15 member states. In Sweden the vote was held in 1995 whereas in the other two countries in 1996. The three elections have been considered as part of the 1994 European- wide election to the EP.2004, finally, all 25 member states voted almost simultaneously over four All six European elections have been considered since 1979, the first direct election of the EP. In addition, to compute the levels of volatility in national elections, as well as mixed levels of volatility between national and European elections, a number of national elections have been selected in the interim of European elections. Two national elections before the 1979 election to the EP have also been consider1 includes the years of national election in order to provide the information concerning the temporal contiguity between national and European elections. It has been considered that not all national elections were needed for the analysis. Between two European elections only one national election has been considered even when more than one took placal results concern the election of lower houses. Concerning the selection of parties to be included in computations, the criterion has been to consider parties receiving at least five per cent nation-wide as well as those receiving at least five per cent within at least one constituency. This allows to include regionalist parties and more generally small parties which – in European elections – score better than 20 A number of more specific problems concern the changing alliances and the splits of parties between two elections. For example, the Italian Democrazia Cristiana after 1992 divided into several parties. The Partito Popolare Italiano has been considered its "successor" for computation purposes. 21 For the 2004 European election data are provisional and minor adjustments might prove necessary in subsequent analyses. 22 Two national elections have been considered before the first European election because two (subsequent) elections are required to compute one national volatility index. 23 In some cases national and European elections took place during the same year. If a previous national election was available, this has been preferred. However, this has not always been possible. In Luxembourg this is particularly problematic as the two orders of elections take place on the same day (implications are discussed below). Furthermore, the 2003 Dutch national election has been preferred as the 2002 vote was heavily influenced by the murder of a party leader. 24 For data sources see notes to Table 2. Concerning national elections, in Germany the Zweitstimmen have been used (on the basis of which the allocation of seats is carried out by PR), and in Italy since 1994 the list votes have been used according to which seats are allocated proportionally (instead of the single-member constituency votes). For France and Hungary results for first ballots have been used. In Belgium Wallon and Flemish parties have been considered together in the computations. Table 1: Elections and countries
Year of national elections considered in relation to EP elections 1974 1979 Netherlands 1977 1979 Ireland 1973 1977 UK-Britain Netherlands 1981 Netherlands 1986 1989 12 France 1986 Netherlands 1989 Netherlands 1998 1999 15 Denmark 1998 1996 2001 Ireland 1999 2003 Lithuania 1996 2000 Sweden a) Before the 1979 European election two elections were considered to be able to compute national volatility rates. The same applies for the 10 Central and East European new member-states before the 2004 European election. In 2004 voting was carried out over four days according to countries (10–13 June). b) Greece voted in 1981 after accession. c) Portugal, and Spain voted in 1987 after accession. d) In Luxembourg European and national elections are held on the same day. e) British October election. Northern Ireland excluded from all British figures. f) Austria, Finland joined the EU in 1994 but voted for the EP in 1996. In Sweden the election was held in 1995. 3.2 Hypotheses and Overall Figures
First, if the hypothesis of the progressive formation of a European electorate increasingly independent of national politics is true, we expect to find that the divergence between party vote in national and European elections grows over 25 years. This would mean that voters perceive two distinct arenas or "spaces" in which different lists – or, at least, the same parties but campaigning and positioning themselves on European rather than national issues – compete on dimensions that cut across national cleavages. If, for example, in a given country we find a low level of volatility in national elections but a high volatility when these are compared to European elections, we ought to conclude that – in elections to the EP – voters refer to different issues, platforms, and perhaps leaders that cut across the usual national alignments. In other words, if in an evolutionary perspective over six elections to the EP since 1979 the European competition sphere has developed progressively, we expect that the mixed volatility increases over time Second, if the hypothesis of the formation of a European electorate is true, we expect the rates of electoral participation to increase over time with voters perceiving these elections as increasingly relevant. More precisely, given the long-term trend of declining turnout in national elections, we would expect to find a convergence of electoral participation between national and European elections. This point is dealt with in a subsequent section of the paper. Concerning party support, over the last 25 years electoral behaviour in European elections has not been dramatically different from electoral behaviour in national elections. This basic result of the analysis appears in the first row in Table 2 where overall figures are given. For the entire period from the beginning of the 1970s until the last European election in 2004, the level of volatility in all types of elections, national and European, is 21.55.appears in the comparison between different types of volatility is that there is only a slight difference in the levels of volatility between national, European, and mixed volatility. The lowest level of volatility can be observed concerning European elections (18.20) applicable only to countries in which at least two elections to the EP took place, that is, Western Europe: From one European election to the next there is greater stability than from one national election to the next. The important information concerning the hypothesis of the "autonomisation" of the European sphere of competition, however, is given by the comparison between national and mixed volatility. The mixed volatility between national and European elections 25 It could be considered that the similarity of electoral behaviour between national and European elections does not necessarily indicate the absence of a European electorate because, in a multi-level system, the same cleavages and alignments exist both at the national and European level. A high volatility indicates the existence of a different dimension at the European level – for example, a pro/anti-European dimension – but a low volatility could also mean that a European electorate and party system exists, but based on the same dimension – namely, the left–right dimension – which is predominant at the national level (see Hooghe and Marks, 2001: Appendix tables). 26 The "N" in Table 2 indicates the number of pairs of elections ("t" and "t+1") on which volatility indices have been computed. The overall "N" of 317 includes volatility between pairs of national elections, pairs of European elections, and pairs of national and European elections in all countries. Table 2: Levels of electoral volatility by country
Difference N Overall N (mixed – volatility volatility volatility national) 21.43 (100) 18.20 (64) 23.04 (153) 21.55 (317) countries) Standard 20.12 (90) 18.20 (64) (297) 2.29 28.77 (22) 15.26 27.97 (22) 11.47 25.25 (22) -0.52 17.87 (22) -0.67 Netherlands 19.71 (6) 17.84 (22) -2.27 Luxembourg 18.66 (6) 16.64 (22) -2.33 17.38 (19) -5.73 13.95 (13) -6.02 21.28 (19) -6.05 New member- 33.21 (10) 31.91 (10) 32.56 (20) -1.30 34.10 (1) – – 6.20 (1) – – 0.50 (1) – – 27.40 (1) – – 21.55 (1) – – 45.50 (1) – – 50.10 (1) – – 46.40 (1) – – 49.00 (1) – – 51.30 (1) – – Notes: Countries are odered by difference between mixed and national volatility (last column). In Luxembourg elections are held simultaneously with EP election (the national election 2004 has been included as well). This was also the case in Belgium (1999), Greece (1989), and Ireland (1989). National elections: Election results to compute volatility indices have been taken from Caramani (2000) with updates from official statistics. European elections: Fragmented – and often unprecise – sources had to be used for compiling the file on European elections results. National Statistical Yearbooks (several years), National Ministries of Interior (electoral services), National Statistical Offices, European Parliament (poor data), Grunberg, Perrineau, Ysmal (2000), Lodge (1986, 1990, 2001), Mackie (1990), Mackie and Craig (1985), Perrineau and Ysmal (1995), as well as newspapers reports for the last 2004 elections. I am grateful to Reinhart Schneider for providing me with data from the Infas-Report Wahlen European election study (1989). (23.04) is 1.50 higher than the overall level of volatility, wher There are therefore only small differences between electoral behaviour in national and European elections. On the contrary, these figures give a first indication that – since the beginning of direct 27 The relatively low levels of the standard deviation further indicate that there is no major variation among cases around the levels of national, European, and mixed volatility. elections to the EP – electoral behaviour in the elections to the EP and elections to national parliaments is fundamentally similar, and speak against the hypothesis of a distinct European arena. This initial result is confirmed by trends over time. Figure 1 displays the temporal evolution of the three levels of volatility – national, European, mixed – from the 1970s until the present. As far as national elections only are concerned (solid curve), there is an increase of volatility between the mid-1970s and the second half of the 1980s. During this decade electoral instability has been growing. Overall, for the 10–12 countries considered during this period, the volatility in the second half of the 1980s is double with respect to that of the mid-1970s mainly due to important changes occurred in almost all European party systems.ater, during the 1990s until the last elections in 2003–04, volatility levels decrease again and remain stable. The high levels of the mid-1980s affect also European elections (dashed curve). The first election for which this type of volatility can be computed is 1984 with respect to the first direct election to the EP in 1979. As it appears in the figure, the level of volatility (as for national elections) decreases in 1989 and stabilises during the 1990s. But the most important information in Figure 1 is conveyed by the evolution of mixed volatility (dotted curve). The trend over time of electoral volatility between national and European elections follows very closely the other two types of volatility – between national elections only and between European elections only. At no moment, during the 25 years of European elections, electoral behaviour in elections to the EP has drastically differred from electoral behaviour in elections to national parliaments. The curves follow the same trajectory over time. Furthermore, the trends of the curves do not suggest any type of divergence which would indicate that European electoral behaviour progressively "departs" or "detaches" itself from national patterns. According to this finding, therefore, it is possible to conclude that even though there have been six elections to the EP over the last quarter of the twentieth century, a European electoral and partisan arena – or a competitive sphere autonomous from the national parties, issues, and leaders – has not as yet emerged. Data suggest on the contrary that, overall in Europe, the same electoral lists run for both national and European elections and that these lists receive very similar levels of support in national and European elections. This indicates that voters do not modify their behaviour on the basis of the issues that are raised before elections to the EP and that they do not react to the European platforms of parties as presented in campaigns before European elections. As a first conclusion, therefore, it is possible to say that national politics still predominates in European elections. 28 This result relativises to some extent the conclusions reached by the famous study by Bartolini and Mair (1990) on the absence of a recrudecence of electoral instability in the recent decades. Whereas their analysis ends with 1985, the data presented here show that there is a peek of instability shortly thereafter which, however, does not persist in the 1990s. As other literature has suggested, these patterns of volatility might suggest a re-alignment taking place in the second half of the 1980s and then stabilising in the 1990s (Dalton et al. 1984, Ersson and Lane 1982, Shamir 1984). The first date appearing in Figure 1 is 1976. Yet the level of volatility is computed with respect to the previous 1972–74 elections (depending on the country) so that almost the whole of the 1970s are covered. 29 Unlike most countries, in the Italian case electoral and party system instability is higher in the 1990s than in the 1980s. The same applies to the Netherlands. For a more detailed comparison see below Figure 2. Figure 1: Levels of volatility in 25 countries (1970s – 2000s)
tylitilaovral to 3.3 Country
In spite of these overall results attesting to a fundamental similarity between national and European electoral behaviour, the volatility between national and European elections varies according to countries. The country comparison shows that in a number of cases the mixed volatility is larger than in other cases. Figures concerning the country comparisons are presented in the lower half of Table 2. What follows in this section looks at the three types of volatility and charts the differences and similarities between countries. First, national volatility is described in order to have a reference point (see the first column in Table 2). The most stable countries with the lowest levels of volatility between the mid-1970s and the present day are Austria, Germany, Finland, and Britain. On the other hand, Portugal and Italy clearly stand out with high levels of electoral instability. Whereas in Portugal this is mainly due to the high values of volatility for the 1985 electiolatility scores have been more erratic but persistent. In most countries the evolution over time follows the general pattern displayed above in Figure 1. There is a high of volatility in the second half of the 1980s and then levels decrease and stabilise in the 1990s. Besides the mentioned cases of Italy and Portugal, exceptions to this pattern are mainly Ireland and Luxembourg where volatility peeks in the 1987 and 1989 national elections respectively, and the 30 In the 1985 national election in Portugal, it is mainly the rise of the Partido Renovador Democrático (with a score of 18.53 percent of the votes) that causes both high levels of national volatility (with respect to the 1980 election) and mixed volatility (with respect to the 1984 EP election). In the 1987 national election, this party received 5.05 percent of the votes nation-wide. Netherlands where electoral instability is a more recent phenomenon that becomes more consistent towards the end of the 1990s (see solid lines in Figure 2 on the different national patterns). National volatility is particularly high in the new accession countries of Central and Eastern Europe where values range between 21.55 in the Czech Republic to 51.30 in Hungary. Volatility rates are high in spite of the fact that – because of the fluid nature of these forming party systems in the early 1990s – only elections since the second half of the 1990s have been considered in the computations. The two new Mediterranean states Cyprus and Malta, on the contrary, are characterised by a great stability of national electoral patterns. Second, the number of cases for the volatility between European elections is smaller than for national volatility levels. In the case of the 10 new accesion countries no value could be computed because only one election took place so far. Generally, the levels of volatility between European elections are lower than those between national elections (18.20 against 21.43). However, in this case too there are country differences. Volatility between pairs of European elections are highest in Italy and France, as well as in Britain and Denmark. On the contrary, this type of volatility is particularly low in Finland and Portugal.second type of volatility is concerned, Figure 2 indicates for most countries a stable and declining trend in the 1980s and 1990s (dashed lines). The main exceptions are France, for which there is an increase in the volatility between the European elections of 1989, 1994, and 1999, Italy, where the 1990s are characterised by a high level of volatility in all orders of elections due to the drastic changes in the party system, Ireland and the Netherlands with an increase in the 2004 election, and Spain between 1989 and 1994. Also in Britain the levels of volatility in European elections remain rather high. Third, and more interestingly, we must consider the volatility between national and European elections (mixed volatility). This indicator tells us in which countries the differential of party support between national and European elections is larger, that is, where there is a more pronounced tendency of voters to separate the two arenas in their voting behaviour. Of the three types of volatility, the mixed volatility is the largest. However, as already notes, it is not dramatically different from national volatility, indicating a basic similarity of behaviour in the two arenas (23.04 against 21.43 and 18.20 of the national and European volatility respectively). Among West European countries for which indices can be computed on larger time periods and which have consolidated party systems, the highest levels of mixed volatility can be found in France and Denmark (36.38 and 32.80 respectively). Also in Britain and Italy (as well as in Ireland, Greece, and Portugal) is the mixed volatility high compared to most other countries. On the other extreme, in Finland this type of volatility is limited to 11.48. Whereas in France and Denmark European elections seem to constitute a different arena with respect to the national ones, in Finland, Spain, or 31 In Portugal national volatility was particularly high in the second half of the 1970s after the democratic transition and stabilises in the beginning of the 1980s (see Figure 2). European elections in this country – since 1984 – are therefore not affected by the initial national volatility. Luxembourg, as well as in most other countries, there is a clear "overlap" between national and European arenas with very few votes changing between the two types of elections. Furthermore, as it is possible to see in Figure 2 (dotted lines), the pattern over time shows again that only in few countries the mixed volatility is higher than the national one. The two cases are again Denmark and Franere there has always been a higher level of volatility and, to some extent, Britain, Greece and Germany where a slight increase in the mixed type of volatility can be observed in the last European election. However, even though this information provides a first indication about the tendency to modify voting behaviour according to the order of election in each country, it might be that the levels of mixed volatility simply reproduce the "endemic" or "congenital" levels of electoral instability in each country. For example, in Italy the level of mixed volatility is comparatively higher than in other countries (24.33), but if juxtapposed to the levels of national volatility in Italy, it appears as "normal" and therefore does not indicate a specific propensity to shifting votes when the arena is the European rather than the national parliament. For this reason it is more appropriate to consider the difference between mixed volatility and national volatility. These data are given in the last column of Table 2 for all countries. Contrary to what one would have expected in the case of a European electoral arena independent of national factors, the difference between mixed volatility and national volatility is overall very small (1.61 if all countries are considered). The difference is even more reduced for Central and East European countries (-1.30) indicating that volatility is a peculiarity of electoral behaviour in these systems in all orders of elections, and not the sign of independent arenas of competition. Over time this appears clearly also in Figure 1 above where the dashed curve of mixed volatility follows over the entire period very closely that of national volatility. Both curves increase from the 1970s until the end of the 1980s and then stabilise at lower levels of electoral change until the present day. The absence of a divarication between the two curves, rejects therefore the thesis of a progressive autonomisation of the European electoral sphere. 32 For France, however, this is true especially in the 1980s and less today. Figure 2: Levels of electoral volatility by country: EU-12 (1970s – 2000s)
National volatility European volatility Mixed volatility Only in two cases in particular the levels of volatility between European and national elections (the mixed volatility) are much higher than the "endemic" national volatility one can observe between national legislative elections. The first of these two cases is Denmark, where the mixed volatility is 36.38, that is 14.26 higher than the volatility in the elections to the Folketing (21.12, a comparatively "average" national volatility). Of all countries, this is the clearest case of a separation in voting behaviour between national and European arenas. What are the causes of this high differential? The main factor of volatility is the creation of a specific anti-EC movement which participates as an electoral alliance in European elections in Denmark since 1979 (Folkebevægelsen mod EF). This alliance does not contest national elections and received in European elections in 1979–89 around 20 percent of the votes. In the subsequent elections of 1994, 1999, and 2004, however, its strength declined to 10.3 and 7.3 and 5.2 percent respectively. Anti-EC groups first mobilised in occasion of the referendum for accession in 1972. The movement (a conglomerate of local committes and political parties) was defeated in the referendum but was revived before the 1979 European election.
most of the groups participating in this alliance are left-wing, the second factor accounting for a high mixed volatility in Denmark is the weakness of Social Democrats in European elections. In national elections they score between 30 and 38 percent, whereas in European elections between 15 and 23 percent (whereas significant variations do not occur for Conservatives and Liberals).
The other country for which the mixed volatility is much higher than the national is France. Whereas the level of volatility in elections to the Assemblée Nationale is on average 21.33, if national and European elections are combined, volatility peeks to 32.80 (11.47 more). As in the Danish case, the causes are the strong fluctuations of some parties between the two orders of elections, as well as the creation of "ad hoc" alliances and new lists on European themes. This concerns particularly the 1994 and 1999 European elections in which the Gaullist Rassemblement pour la République built and alliance with the liberal Union pour la Démocratie Française, and in which a number of new lists were of these is Energie Radicale (led by a former buisinessman) which received 12 percent of the votes, another is Autre Europe, a right wing splinter of the Gaullist party and, finally, the Rassemblement pour la France et l'Indépendance de l'Europe. These parties contested also the 1999 election (with the exception of Energie Radicale), but the alliance between RPR and UDF was In the French case, however, more than in the Danish case, the mixed volatility is also caused by two national elections – the 1981 and 1986 (see graph in Figure 2). With respect to the 1979 election in 33 I have chosen to consider the anti-EC movement as an independent alliance as it regroups candidates and groups from different parties (although most of them are left of centre parties: Socialistisk Folkeparti, Socialdemokrater, Venstresocialisterne) as well as the liberal Retsforbundet (or Justice Party). This alliance represents a good early case of a European issue and cleavage cutting across national alignments and, therefore, an example of the formation of an autonomous European electoral space with respect to national political spaces. 34 The Radikale Venstre too scores less in European elections during the 1980s in particular (they recover strength in the 1990s) and the Centrum Demokraterne enters the electoral scene for the first time in the 1979 European election. which the UDF increased its votes to 26.6 percent, in the 1981 election it returned to around 20 percent and, in the 1984 European election, dropped to 10 ove all, however, the 1981 national election has been particularly volatile because of the strong increase of support for the Socialist Party (from 22–24 percent in the previous national and European elections to 37.3 percent). This percentage diminished to 20.8 percent in the subsequent European election of 1984, but then increased again to 31.5 percent in the 1986 election (the only national election during this period with PR instead of a two-ballot formula). The 1984 European election is also the first election in which the extreme right-wing and anti-European Front National increases its votes from close to nothing to more than 10 percent – a level that it has maintained until the present cating, as in Denmark, a significant dimension in the party system between pro- and anti-Europeans. The other countries in which mixed volatility is higher than national volatility are Britain (7.78 more) and Ireland (4.27 more). In Britain, there are two moments of high mixed volatility: the first half of the 1980s, and since the 1999 European election (see graph in Figure 2). In both cases, it is the combination of variation of support for the large parties and the increase of votes for small parties in European elections to cause high levels of mixed volatility. The European election of 1989 is characterised by the strong decline of the Conservative Party (from 42.30 to 33.00 percent) as well as of the Liberal Party (almost half the votes). On the contrary, the Labour Party increases its support by eight percent. In the 1989 election, furthermore, the Greens score a high of 14.50 percent, a level that they will never reach again. In the 1989 election, furthermore, the Greens score a high of 14.50 percent, a level that they will never reach again. In the subsequent 1992 national election, the main change consists of the creation of the Liberal Democratic Party which receives 17.85 percent of the votes. Concerning the late 1990s, the main factor of volatility is the Labour Party which passes from 43.21 percent in the 1997 national election to 28.00 percent in the 1999 European election and then again to 40.70 percent in the 2001 national election. In 2004, finally, the UK Independence Party (UKIP) did particularly well with 16 percent of the vote. In Ireland only the first two European elections show a discrepancy with national electoral behaviour mainly through the fluctuations of Fianna Fáil's support. This party, in the late 1970s and early 1980s, is strongly penalised in European election Since then, however, the levels of mixed volatility decline and correspond to those of low national 35 The RPR and the UDF were two separate parties until they unified last year as Union pour un Mouvement Populaire. The UDF still exists formed by the group of Liberals led by Alain Madelin that did not join the UMP. 36 In part this was a consequence of the rise of the Centre pour l'Europe (Centre des Démocrates Sociaux) which received 13 percent in the 1984 European election. 37 A peek of support for the Front National takes place in the 1997 national election (15.1 percent). 38 A number of changes take place in the Irish party system – although the do not affect significantly the levels of volatility – principally with the rise of the Workers' Party, the Democratic Left, and the Progressive Democratic Party. 39 In Germany the slightly higher levels of mixed volatility in the 1980s (3.72 more than national volatility) are caused mainly by the growth of the Greens in European elections and, in the late 1990s, by the bad performance of the Social Democrats in European elections. This would support the thesis of a "punishment" of incumbent parties in second-order elections. The Sozialdemokratische Partei Deutschlands wins the 1998 and 2002 national elections but decreases to 30.70 percent in the 1999 European election and to 21.50 in the 2004 European election. The Christlich-Demokratische Union – on the contrary – receives less than 30 For a number of cases Figure 2 shows that the mixed volatility has increased between the last 2004 European election and the previous national election. Among the 12 countries for which longer time series are available this appears to be the case in Belgium, Germany, Ireland, Portugal, Spain, and, most particularly the mentioned case of Britain where the UKIP scored very well. Besides the UKIP, European-specific lists emerged in several countries: examples are Europa de los pueblos in Spain (a coalition of Catalan, Basque, and other regionalists), the anti-European June List in Sweden, and Europe of Transparancy in the Netherlands. In Poland, two anti-European League of Catholic Families and Selfdefence were particularly successful. Concerning the new member-states, the mixed volatility is hihg in Estonia, Latvia, Lithuania, Slovenia, and Poland. However, in these new forming systems a great proportion of volatility seems to be caused by their national instability. This appears in the last column of Table 2. If one leaves out Cyprus and Malty which have extremely stable national volatility patterns among the eight Central and East European countries, the two electorates that stand out for having a different behaviour in the 2004 Eurpean election compared to previous national elections are Estonia (18.90 more volatility tha in national elections) and Slovenia (8.80). In Estonia this is due mostly to the increase of votes between the 2003 national election and the 2004 European election of the Fatherland's Union and of the People's Party (from 7,30 and 7,00 percent respectively to 20.05 and 36.08 percent), as well as of the lower scores of the Estonian Centre Party and Estonian Reform Party (from 25.50 and 17.7 respectively to 17.05 and 12.20). The Estonian election has also been characterised by a particularly low turnout (26.9 percent as it appears in Table 3 below). Overall, Figure 3 indicates that there is a basic correspondence between national levels of volatility and the changes which occur in European elections. Only few cases deviate which are mainly those which have been documented. The close correlation (r=.54) that is displayed between the electoral instability from one national election to the next and the instability that occurs in combination with European elections, indicates that the latter – in most countries – do not constitute a radically different arena in which voters change their preferences on the basis of a different offer in the form of lists or platforms. the contrary, such figures rather indicate that European elections are still based on national political themes and allegiances in most countries. percent of the votes in the 1998 and 2002 national elections and increases to 36–39 percent in the 1999 European elections. The slight bifurcation we see in Germany between the last two national and European elections appears also in Greece and Britain (see Figure 2). 40 In three cases in Western Europe, even, the level of mixed volatility is lower than the average national volatility. In Spain, Portugal, and Finland, voters seem to change preference more from one national election to the next than between national and European elections (see negative scores in the last column of Table 2). This is also the case in Lituania, Poland, Slovakia, and Hungary as far as the new accession countries are concerned. Figure 3: Relationship between national and mixed levels of volatility
Lithuania Czech Rep. National volatility Note: Pearson's r = .54 (significant at the .01 level). Legend: Italics used for 10 (mostly Central and East European) new member-states. Orthogonal lines: mid-ranges. What Figure 3 also shows is a clear division between Western Europe – a group to which also Cyprus and Malta seem to belong – and Central and East European countries characterised by high electoral instability (the Czech Republic is the exception among these eight countries). All countries in which both types of volatility are above the mid-range are Central and East European countrie more countries (Slovakia and Hungary) it is the national volatility which is predominantly high, and Portugal approaches these two countries. As already shown by Table 2, in addition, this figure indicates the countries that deviate from the correspondence between national and mixed volatility. In particular, for Denmark, France, Estonia, and Cyprus the mixed volatility between national and European elections is higher than volatility in national elections. To some extent this is the case also for Britain, Czech Republic, Slovenia, and Malta. 41 The orthogonal dashed lines have been placed in the middle of the range between the lowest and highest value of each of the two volatility distributions (Finland and Estonia concerning mixed volatility, and Malta and Lithania concerning national volatility). 3.4 Turnout
Further insights concerning the formation of a European electorate can be gained from the observation of turnout figures – on which a great dela of literature is available and of which also the wider public is aware. As mentioned, we expect the evolution towards an incrasingly aware and informed electorate that forms over six European electoral campaigns to translate in the convergence between the levels of electoral participation in European elections and turnout rates in national What appears in Table 3, however, disconfirms once again the thesis of a forming European electorate. The left-hand side of the table shows that turnout in European elections is historica In Western Europe over six elections to the EP, the average rate of electoral participation is 60.8 percent. Turnout is particularly low in Britain, Portugal, and Finland (below 30 percent of turnout in the 1999 election to the EP). The crucial information, however, concerns the large difference with turnout in national elections. Overall, in European elections turnout is around 18–25 percent less than in national elections, although there are significant country variations (see the last two columns of the table). Over the entire period, the West European countries in which the difference between national and European electoral participation is larger are Sweden, the Netherlands, Denmark, and Austria. In these countries, the number of voters is 35–45 percent less in elections to the EP. In three more countries – Finland, Portugal, and Germany – voters in European elections diminish of more than a third with respect to elections to national parliaments, and in most countries voters are a fourth less (around 25 percent). The only countries in which the levels of electoral participation in European elections do not differ dramatically from national turnout rates are those in which voting is compulsory (Belgium, Greece, and Luxembourg) and – in particular – in Luxembourg where national and European elections are held on the same day.
42 Exceptions are the countries in which voting is – according to different provisions – compulsory. Voting is compulsory in Belgium, Greece, and Luxembourg. In the Netherlands it has been compulsory until 1967 and in Italy – to some extent – until 1992 when the legislation was changed. In Austria, compulsory voting is a matter of Land legislations. For details see Caramani (2000: 63–65). 43 This effect of simultaneous elections is confirmed by the Irish figure for 1989, when concurrent national and European elections were held. Table 3: Turnout levels by country and differential between national and European elections (percentages)
Turnout in European elections Difference with previous national election Netherlands 57.8 a) Compulsory voting. In the Netherlands until 1967 and in Italy until 1992. In Austria matter of Land legislation. Countries ordered by overall difference between national levels of turnout and levels of turnout in European elections (last column). To compute differences with national elections, the last national election before each European election has been considered in each country. Figure 4: Levels of turnout in national and European elections
This situation – as extensively reported by the press – is the same, and even more accentuated, in Central and East European countries. For the 2004 European election turnout has been 53.3 percent less with respect to the previous national election in Slovakia, 41.6 percent less in Slovenia, and a third less in countries like Hungary, Estonia, Latvia, and the Czech Republic. From an evolutionary perspective, instead of increasing turnout rates have been decreasing from 1979 to 2004, from 67.5 percent (in the 10 countries then member of the European Economic Community) to 47.8 percent (25 member-states of the EU). The progressive and accelerating process of European integration has therefore not produced effects on the awareness of the relevance of the EP and European issues and policies in the perception of the voters. Quite the opposite occurred with electorates being increasingly indifferent to the election of members of the EP. The left-hand side of Table 3 shows that, in this case too, differences between countries are important. Until 1999, as general pattern in all countries turnout in European elections has diminished but particularly in the Netherlands, where electoral participation decreased from 57.8 percent in 1979 to 29.9 in 1999, in Austria (from 67.7 to 49.0 percent in the two elections of 1994 and 1999), and in Finland where between the two elections of 1994 and 1999 voters literally halved (from around 60 to 30 percent). As far as the 2004 election is concerned, however, low turnout rates in Central and Eastern Europe accounts for most of the overall drop of electoral participation between 1999 and 2004. On the contrary, rates have recuperated to some extent in Western Europe (where a comparison with previous European elections is possible) where turnout has increased in the Netherlands, Finland, Britain, Ireland, and Luxembourg, or remained stable (Belgium, Portugal, Germany, and Sweden). Again, the most important information is whether this corresponds to a more general trend of political apathy that can be observed also in national electorates, or whether the phenomenon is more specific to European elections. Figures on the right-hand side of Table 3 – as well as Figure 4 – show a clear trend towards an increasing divergence between national and European levels of participation. This means that if on the one hand there is a general trend towards absentionism in all election orders, this trend is more accentuated in the case of European elections. This disconfirms the thesis on the formation of a European electorate. The last row of the table, however, also shows that this tendency is particularly present in the last elections of 1999 and 2004 (an overall turnout difference with respect to national elections of 25.3 and 25.1 percent respectively). Even though between 1979 and 1994 there has been an increase of the divergence between national and European turnout (from 18.6 to 21.3 percent less in European elections), this tendency has been more fluctuating than decreasing 4 The EU as a Forming Democratic System
For the time being, the empirical evidence presented above does not need to be further dwelled to confirm that an autonomous EU party competition with European-wide alignments and issues has not as yet emerged. It is therefore to some extent paradoxical that much of this literature has been concerned with the explanation of the differences between party choice in European and national elections rather than trying to account for the persistent national character of European elections. Not surprisingly, research on second-order elections was able to find only little evidence supporting the decline of major and incumbent parties in second-order elections anre, a more "sincere" vote that would cause a "deviation" of electoral behaviour in European elections from national elections. On the other hand, also indicators of protest vote, decline of support for government parties, and losses of votes for the major parties, do not show significant differences with national elections (Marsh and Franklin 1996: 13–15). Cyclical models too – in which a "punishment effect" for large parties is supposed to be at work – have found only partial confirmation in empirical analyses (Erikson 1988).ermore, evidence above has shown that only in few cases party systems in second- 44 Even though incumbents in 2004 did badly. Anti-incumbent vote took place in Great Britain, Germany, Sweden, as well as Italy and France. In Central and Eastern Europe the vote too (with the exception of Slovakia) has been strongly against incumbents in particular in the Czech Republic and in Poland. Yet volatility does not seem much higher. 45 Schmitt (1990) shows that governing and large parties in European elections were not less effective than other parties at mobilising their voters and not less performant in the elections' outcome. order elections include new or "ad hoc" parties forming against European themconfirms the results of research conducted on issue voting and ideologies which found no or little evidence on the formation of a separate arena. European-wide issues do not constitute salient factors in voting choice in European electionere large differences in behaviour exist between national and European elections – namely when it comes to turnout – analysts have been more successful in accounting for country differences by introducing factors such as compulsory voting, positive attitudes towards the EU, or temporal proximity with national elections. The lack of empirical support for these hypotheses suggest that the wrong question has been As intriguing as the "puzzle" might be, therefore, this type of research seems to miss the crucial point. The focus on the explanation of the sources of differences in behaviour between national and European elections, leads to the omission of the more important question of "why isn't there a European electorate?," in other words the sources of persistence of nationalised electorates and their resistance to European-wide integration. As mentioned, previous work done on the "nationalisation" of politics (Caramani 2004) might shed, comparatively, some light on (the lack of) "Europeanisation" processes from a mass-electoral perspective. Such a comparison might also be useful in future and more extensive research beyond the mere exploration of this paper. Nationalisation processes started off as soon as external state boundaries were consolidates and internal voice channels structured through the introduction of elected parliaments and the organisation of electoral committees, candidacies, and parties. The fact that nationalisation takes place from the first phases after the transition to competitive election, suggests that the transition from territorial to functional nation-wide cleavages politics is typical of forming democratic systems. The EU is today in a similar situation with consolidating external boundaries, integrated internal structures (economic and juridical above all), and electoral channels for the expression of voice. How then can we explain the absence of a European electorate? In the light of the theory of the nationalisation of politics, the absence of a similar process today at the European level is not surprising. On the contrary, a comparable European-wide shift of issues, political affiliations, party organisations, leadership, etc. from the national to the European level is highly unlikely. What this paper argues is that a European electoral and party system will not exist for long. The basic point is that most of the factors that led to the to the integration of fragmented and localised electorates into national ones during a period of great social and political mobilisation are absent today in Europe. The nationalisation of politics took place during a period of great social and geographical mobilisation in the wake of industrialisation and urbanisation processes which uprooted pre-industrial 46 EP estimates for 2004 attribute to these parties a maximum of 10 percent of seats in the EP: 12 seats for the UKIP, two Dutsch and one for Danish anti-European, together with seven and four Nationalists in Poland and Latvia respectively. 47 See, on this point, Blumler and Fox (1982) who also argue that voters have only little knowledge on the affiliation of their preferred party in European party federations. See also Kuechler (1991), and Van der Eijk and Franklin (1991). 48 As is often the case in political science where explanations for phenomena are sought for before having social structures. At the same time, the nationalisation period corresponded to unprecedented levels of political mobilisation through processes of nation-building and the development of competitive mass The fundamental changes through which the process of national electoral integration can be explained, are absent today in the process of European integration. Although we witness a clear tendency towards the centralisation of policy-making structures and processes in the EU, as well as toward increasing powers to the directly elected EP, the conditions for the inclusion and mobilisation of the masses are not comparable to those of the nineteenth and beginning of twentieth century. The absence of strong factors of mass mobilisation makes the disruption of consolidated (national) electoral structures, alignments, and identifications much more difficult than the nineteenth century "virgin" masses highly mobilised by industrialisation and urbanisation, encapsulated in heavy party and union organisations, and socialised in rigid and powerful ideological schemes such as Socialism and Nationalism. Political-electoral allegiances were "frozen" and encapsulated into national structures and identities in the first phases of mass mobilisation. They proves therefore much more resistant than past local and regional identities which had never been mobilised politically.rst mobilisation was a national mobilisation, a total-mass mobilisation, as well as the first mobilisation constraining future generations. The simultaneous social and political mobilisation of the nineteenth and twentieth century is absent. This simultaneous processes involve the incorporation in the political system (through the lowering of entry barriers such as the franchise and the electoral formula) of new working masses mobilised by the Industrial Revolution and the subsequent urbanisation, as well as by the national mobilisation and the creation of national citizenships against peripheral, cultural, regional resistances. The stability in the post-World War II period of the nationalised electorates that emerged from the crucial moments of social and political mobilisation confirms the strength of these processes. This period witnesses a fundamental stability of the territorial configurations of the vote in Europe, with strong similarities with other long-term electoral analyses – namely those of the "freezing hypothesis" (1970: 72–144): territorial structures have crystallised after World War I and remained stable in the following decades. Not only was mass mobilisation a factor of nationalisation of politics, but also of stabilisation of functional electoral alignments (Bartolini and Mair 1990). Cleavage constellations as well as territorial structures proved stable in the period since World War II. No factor intervening after World War II was able to modify the existing nationalised territorial structures:e further development of communication technologies (through electronic media in particular), nor the transformation of social structures from agrarian societies into industrial and service societies (with the shift towards a service-based economy) or the process of secularisation, nor finally the transformation 49 This runs against the argument (for example, Schmitter 2000: 66–71) suggesting that the European party system in the last two decades resembles roughly to the national politics in the mid-nineteenth century. 50 Data on nationalisation show for example that "new regionalisms" are weak and sporadic phenomena if inserted in a long-term perspective. of political parties from mass parties with heavy organisations and strict ideologies into broader catch- all parties deprived of solid socio-electoral bases, dense organisational networks, and constraining ideologies.n the contrary, the current period is characterised by depolitisation processes, lower rates of participation, ideological dilution, blurring of cleavage lines, and weakening of organisational party structures. The concomitant timing between political democratisation with the extension of voting rights to new masses mobilised by the Industrial and National Revolutions, and the process of state formation and nation-building proved crucial in forging national electorates. The process of democratisation and political mobilisation was a crucial element of the formation of nationhood. The dominant ideology of the nineteenth century – Liberalism – combined elements of national unification (Italy, Germany, Switzerland, etc.) or independence from foreign occupation (Belgium, Italy again, Finland, Norway, etc.) with state formation (centralisation) and democratisation (in all European countries against privileges of church and aristocracy). Up to the present, therefore, there is a strong association between the nation and democracy, between political citizenship and the nation. This same association applied recently (15 years ago) to Central and East European countries. The Europeanisation of electorates faces therefore a double challenge. First, the formidable difficulty to uproot nationalised electorates, and to fight against such powerful mobilisation, politicisation, and socialisation factors that forged electorates into the national mould. It appears extraordinarily difficult to uproot and denationalise such strongly consolidated structures and identities created through the macro-transformations of the industrial society. Second, current trends are not only not comparable in scope to those processes, but in addition, the current phase of demobilisation – rather than mobilisation – with low turnout rates, declining membership, and dilution and blurring of cleavages, is unable to create a new European allegiance. 5 Conclusion
There is one last point that finally deserves to be discussed: the general trend of volatility that can be observed in most European member-states from the 1970s until the end of the 1980s, and the subsequent stabilisation of electoral behaviour during the whole of the 1990s and the beginning of the 2000s. Most of the country graphs in Figure 2 reproduce the general pattern of Figure 1, although often the causes of an increasing instability at the end of the 1970s and mid-1980s differ from country to coide the theoretical implications described above this has also methodological and normative implications. 51 See, on these organisational transformations, Kirchheimer (1966) and Katz and Mair (1995). 52 See both the solid and dotted lines for national and mixed volatility respectively. Exceptions to this general trend are mainly Italy and the Netherlands where high levels of volatility persist in the 1990s. Methodologically, comparative politics – as all "quasi-experimental" methods in the social sciences – bases its explanatory control on the co-variation (logical, as in Mill's tables, or statistical) between phenomena which, naturally, leads to focus on differences between the basic analytical cases which in comparative politics are countries. This emphasis on variations and differences leads to overlook the general transformations that take place within European electorates and societies, and too rarely there is sufficient awareness of the trade-off between increasing the number of cases in order to have more variation in the dependent and independent variables, and the observation of fundamental changes that are common to all political systems.concentration on a multi-variate logic of causality goes partly to the disadvantage of the comprehension of broad common evolutions of societies and electoratekkan's model itself – the only that really thought about the genesis of Western party systems – bases its explanatory method on subsequent "deviations" or "variations" between West European states in the development of party families. His model too is much less concerned with the explanation of common aspects, namely, the rise and hegemony of the left–right cleavage and workers' parties. This is related to the assumption of the independence between cases.s several electoral evolutions have demonstrated that – rather than being independent – between European countries there are several parallel evolutions (of which similar volatility trends are only one aspect), suggesting that mechanisms of "diffusion" or "emulation" are at work. not only include long-term processes such as the rise of class politics in the wake of social mobilisation triggered by the Industrial Revolution, or the development after World War II of broad confessional people's parties, but also the recent institutionalisation of "new politics" movements – in the form of new parties developing out of value and generational changes (Inglehart 1990) –, the emergence of new right-wing populist protest parties (the "losers of modernity") as a response to the challenges posed by modernisation and globalisation processes (Betz 1994), or new forms of regionalism developing in the spaces opened up by decentralisation processes – in Belgium, Britain, Italy, Poland, and Spain in particular – and the weakening of the normative and identity role of the nation-state (Keating 1998).57 53 On the problem of increasing the number of observations, see King, Keohane, and Verba 1994: 35–38) and – for an application to European integration – Moravcsik (1998: 79–80). 54 This is obviously different from – but not totally unrelated to – the so-called "n=1" problem (see Caporaso et al. 1997), that is, the explanation of unique phenomena such as "regional integration" in Europe. In EU studies too the tendency in the last years has been to give up "grand theories" and to disaggregate the case regional integration into several cases of different inter-governmental bargaining, adoptions of social regulation, voting of European MPs, policy-implementation, etc. 55 Statistically, but also historically. Charles Tilly's critique to the Rokkanian model of party system structuring – as well as of the formation of nation-states – points to Rokkan's failure in genuinely analysing the interactions between countries, even though Tilly acknowledges that Rokkan made decisive moves forward from comparisons in which cases stood as logically independent (Tilly 1984: 129; see also Flora 1999: 10). 56 Within the vast methodological literature on "diffusion" processes and on "Galton's problem", see Klingman (1980), Naroll (1965), and Wellhofer (1989). 57 The temporal dimension plays a crucial but problematic role, for – when we speak of broad socio-political transformations – we mean that they take place simultaneously in the different analytical units which the research design regards as adequate (here, national political system), that is, when there is no cross-sectional Normatively, therefore, it is somewhat of a paradox that this paper concludes with the result that "there is no European electorate" while it points to major commonalities between national electorates, synchronically (same cleavages and party families in all countries) and diachronically with similar volatility trends over time. These common trends would rather point to the existence of a European- wide electorate rather than to its absence, which is sensible to the same value and structural changes, rather than focussed on nation-specific themes. One explanation for this paradox, however, lies in the fact that whereas it is possible to observe common trends within the European society, these will not translate in a European electorate for there is no institutional link between the European electorate and the political system through an electoral process with accountability and responsibility mechanisms. European societies move together (sociologically and "behaviourally"), they are confronted with the same value changes (secularisation, post-materialism), and suffer from the same structural changes (aging, unemployment, economic globalisation), but react in the consolidated arenas – namely, the national ones. Although the stimuli are the same all over Europe, therefore, the response is given through national channels. Whereas there clearly is a behavioural society, it is the electorate in the institutional meaning of the term which is still missing.
Earlier processes at the national level, once again, provide a useful comparison. The nationalisation of electoral alignments and parties, has meant the transition from a fragmented type of politics with strong autonomous local political figures, to "system-wide" mechanisms of accountability in which candidates are submitted to controls and sanctions from system-wide electorates. National party organisations, with strict vertical controls over local branches, gave the masses the possibility to directly influence national decision-making processes. Candidates no longer merely represented their constituencies but rather nationwide functional interests and values. The formation of nation-wide electoral alignments and party organisations in control of the behaviour of single personalities therefore increased the responsiveness in a political process that was inserted in stable structures in which programmes and policies were debated through a much larger mass participation at all levels. Electoral integration implied therefore the formation of a political and democratic citizenship, or a political and democratic nationhood. Also at the European level today, the construction of a political and democratic citizenship – with a similar development to the creation of an economic and legal citizenship – is at the centre of the debates on the "democratic deficit" of the European Union mentioned at the beginning, in which the unbalance between "input" and "output" is still large. The integration of "system-wide" electorates and party systems represents therefore a crucial step towards the structuring of party politics necessary for an accountable system to emerge. It is only with the transformation of a highly territorialised politics time variation. Explaining phenomena that do not vary cross-sectionally but only over time in all countries at the same time leads to the problem of historical multi-collinearity – a correlation between two variables that co-vary over time – which is particularly acute with general and ubiquitous phenomena "wherever no cross-unit variance is available" (Bartolini 1993: 157), and when the convergence and similarity between countries weakens the chances of success of a research strategy of the type "the earlier … the higher" or "the later … the faster." 58 See, on this point, Mair (2000). through the predominance of functional-ideological politics that a "truly European" electorate will acquire the capacity to sanction directly EU policy-makers, that is, from a Europe des patries (in which politics is structured around national identities and interests) to a Europe des partis (the famous phrase by Marquand 1978). However, as this paper has tried to show, the strong and persistent political structures, behaviours, and cultures that mobilised, politicised, and socialised European electorates at the national level combined with the lack of a significant social and political mobilisation in Europe today, hindered so far a trajectory towards the "Europeanisation" of electorates and party systems and will probably make it unlikely even after major institutional reforms. References
Anderweg, R., 1995. The Reshaping of National Party Systems. West European Politics 18: 58–78. Bartolini, S., 1993. On Time and Comparative Research. Journal of Theoretical Politics 5: 131–67. ——, 2002. Institutional Democratization and Political Structuring in the EU. Lessons from the Nation- State Development, in Cavanna, H., ed. Governance, Globalization, and the European Union. Which Europe for Tomorrow? Dublin: Four Courts Press. Bartolini, S. and P. Mair, 1990. Identity, Competition, and Electoral Availability. The Stabilisation of European Electorates: 1885–1985. Cambridge: Cambridge University Press. Bendix, R., 1977. Nation-Building and Citizenship. Studies of Our Changing Social Order (new enlarged edition). Berkeley: University of California Press. Betz, H.-G., 1994. Radical Right Wing Populism in Western Europe. Basingstoke: Macmillan. Blumler, J. and A. Fox, 1982. The European Voter. Popular Responses to the First European Community Elections. London: Policy Studies Institute. Campbell, A., 1960. Surge and Decline. A Study of Electoral Change. Public Opinion Quarterly 24: ——, 1966. Surge and Decline. A Study of Electoral Change, pp. 40–62 in Campbell, A., Converse, P., Miller, W. and D. Stokes, eds. Elections and the Political Order. New York: Wiley. Caporaso, J, Marks, G., Moravcsik, A. and M. Pollack, 1997. Does the European Union Represent an n of 1? ECSA Review 10: 1–5. Caramani, D., 2000. Elections in Western Europe since 1815. Electoral Results by Constituencies [supplemented with CD-ROM]. London-New York: Macmillan. ——, 2003. ‘The End of Silent Elections. The Birth of Electoral Competition, 1832–1915'. Party Politics ——, 2004. The Nationalization of Politics. The Formation of National Electorates and Party Systems in Western Europe. Cambridge: Cambridge University Press. Claggett, W., W. Flanigan, and N. Zingale, 1984. Nationalization of American Electorate. American Political Science Review 78: 77–91. Dalton, R., Flanagan, S. and P. Beck, eds., 1984. Electoral Change in Advanced Industrial Democracies. Realignment or Dealignment? Princeton: Princeton University Press. Deutsch, K., 1953. Nationalism and Social Communication. An Inquiry into the Foundations of Nationality. Cambridge: The MIT Press. Erikson, R. 1988. The Puzzle of Midterm Loss. Journal of Politics 50: 1011–29. Ersson, S. and J.-E. Lane, 1982. Democratic Party Systems in Europe. Dimensions, Change, and Stability. Scandinavian Political Studies 5: 67–94. Flora, P. 1999, Introduction and Interpretation, pp. 1–91 in Rokkan, S. State Formation, Nation- Building, and Mass Politics in Europe. Oxford: Oxford University Press. Grunberg, G., P. Perrineau, and C. Ysmal, 2000. Le Vote de Quinze: Les Elections Européennes du 13 Juin 1999. Paris: Presses de Science Po. Hix, S., 2002. Parties at the European Level, pp. 280–309 in Webb, P., D. Farrell, and I. Holliday, eds. Political Parties in Advanced Industrial Democracies. Oxford: Oxford University Press. Hix, S., Kreppel, A. and A. Noury, 2003. The Party System of the European Parliament. Collusive or Competitive? Journal of Common Market Studies 41: 309–31. Hix, S. and C. Lord, 1997. Political Parties in the European Union. London: Macmillan. Hooghe, L. and G. Marks, 2001. Multi-Level Governance and European Integration. Lanham, MD: Rowman and Littlefield. Hug, S., 2002. Voices of Europe: Citizens, Referendums, and European Integration. Lanham, MD: Rowman and Littlefield. Infas Report-Wahlen, 1989. Europawahl 1989. Wahl der Abgeordneten aus der Bundesrepublik Deutschland am 18. Juni 1989. Analysen und Dokumente. Bonn: Infas. Inglehart, R., 1990. Culture Shift in Advanced Industrial Societies. Princeton: Princeton University Katz, R., 1973. The Attribution of Variance in Electoral Returns. An Alternative Measurement Technique. American Political Science Review 67: 817–28. Katz, R. and R. Kolodny, 1994. Party Organization as an Empty Vessel. Parties in American Politics, pp. 23–50 in Katz, R. and P. Mair, eds. How Parties Organize. London: Sage. Katz, R. and P. Mair, 1995. Changing Models of Party Organization and Party Democracy. The Emergence of the Cartel Party. Party Politics 1: 5–28. Keating, M., 1998. The New Regionalism in Western Europe. Territorial Restructuring and Political Change. Cheltenham: Edward Elgar. King, G., Keohane, R. and S. Verba, 1994. Designing Social Inquiry. Scientific Inference in Qualitative Research. Princeton: Princeton University Press. Kirchheimer, O., 1966. The Transformation of the European Party Systems, pp. 177–200 in La Palombara, J. and M. Weiner, eds. Political Parties and Political Development. Princeton: Princeton University Press. Klausen, J. and L. Tilly, 1997. European Integration in Social and Historical Perspective. 1850 to the Present. Lanham, MD: Rowman and Littlefield. Kohler-Koch, B., 1998. La Renaissance de la Dimension Territoriale en Europe. EUI Working Papers Klingman, D., 1980. Temporal and Spatial Diffusion in Comparative Analysis of Social Change. American Political Science Review 74: 123–37. Kreppel, A., 2002. The European Parliament and Supranational Party System. A Study in Institutional Development. Cambridge: Cambridge University Press. Kuechler, M. 1991. Issues and Voting in the European Elections 1989. European Journal of Political Research 19: 81–104. Lodge, J., 1986. Direct Elections to the European Parliament 1984. London: Macmillan. ——, ed., 1990. The 1989 Election of the European Parliament. New York: St. Martin's Press. ——, ed., 2001. The 1999 Elections to the European Parliament. Basingstoke: Palgrave. Mackie, T., ed., 1990. Europe Votes 3. European Parliamentary Election Results 1989. Aldershot: Mackie, T. and F. Craig, eds., 1985. Europe Votes 2. European Parliamentary Election Results 1979– 84. Chichester: Parliamentary Research Services. Mair, P., 2000. The Limited Impact of Europe on National Party Systems. West European Politics 23: Marks, G. and M. Steenberger, 2002. Understanding Political Contestation in the European Union. Comparative Political Studies 35: 879–92. Marquand, D., 1978. Towards a Europe of the Parties. Political Quarterly 49: 425–45. Marsh, M., 1998. Testing the Second Order Election Theory After Four European Elections. British Journal of Political Science 28: 591–607. Marsh, M. and M. Franklin, 1996. The Foundations: Unanwered Questions from the Study of European Elections, 1979–94, pp. 11–32 in Van der Eijk, C. and M. Franklin, et al. Choosing Europe? The European Electorate and National Politics in the Face of Union. Ann Arbor: The University of Michigan Press. Mény, Y., 2003. De la Démocratie en Europe. Old Concepts and New Challenges. Journal of Common Market Studies 41: 1–13. Moravcsik, A., 1998. The Choice for Europe. Social Purpose and State Power from Messina to Maastricht. Ithaca: Cornell University Press. Naroll, R., 1965. Galton's Problem: The Logic of Cross-Cultural Analysis. Social Research 32: 428–51. Niemi, R. and H. Wiesberg, eds., 1993. Controversies in Voting Behavior. Washington, D.C.: Congressional Quarterly Press (third edition). Pedersen, M., 1996. Europarties and European Parties. New Arenas, New Challenges and New Strategies, in Anderssen, S. and K. Eliassen, eds. The European Union. How Democratic Is It? London: Sage. Perrineau, P. and C. Ysmal, eds., 1995. Le Vote des Douze. Les Élections Européennes de Juin 1994. Paris: Presses de la Fondation Nationale des Sciences Politiques. Raunio, T., 2002. Political Interests. The EP's Party Groups, in Peterson, J. and M. Shackleton, eds. The Institutions of the European Union. Oxford: Oxford University Press. Reif, K.-H. and H. Schmitt, 1980. Nine Second-Order National Elections. A Conceptual Framework for the Analysis of European Election Results. European Journal of Political Research 8: 3–44. Reif, K-.H., 1984. National Electoral Cycles and European Elections 1979 and 1984. Electoral Studies ——, 1985a. National Electoral Cycles and European elections 1979 and 1984. Electoral Studies 3: ——, 1985b. Ten Second-Order National Elections, pp. 1–36 in Reif, K.-H., ed. Ten European Elections. Aldershot: Gower. ——, 1997. European Elections As Member State Second-Order Elections Revisited. European Journal of Political Research 31: 115–24. Rokkan, S., 1970. Citizens, Elections, Parties. Oslo-New York: Universitetsforlaget-McKay. ——, 1999. State Formation, Nation-Building, and Mass Politics in Europe. The Theory of Stein Rokkan. Oxford: Oxford University Press (edited by Flora, P., with S. Kuhnle, and D. Urwin). Schattschneider, E., 1960. The Semisovereign People. A Realist's View of Democracy in America. New York: Holt, Rinehart, and Winston. Schmitt, H., 1990. Party Attachment and Party Choice in the European Elections of June 1989. A Cross-National Comparative Analysis of the Post Electoral Survey of the European Voters Study 1989. International Journal of Public Opinion Research 2: 169–84. Schmitter, P., 2000. How to Democratize the European Union… and Why Bother? Lanham: Rowman and Littlefield. Shamir, M., 1984. Are Western European Party Systems "Frozen"? Comparative Political Studies 17: Stokes, D., 1965. A Variance Components Model of Political Effects, pp. 61–85 in Claunch, J., ed. Mathematics Applications in Political Science. Dallas: The Arnold Foundation. ——, 1967. Parties and the Nationalization of Electoral Forces, pp. 182–202 in Chambers, W. and W. Burnham, eds. The American Party Systems. Stages of Political Development. New York: Oxford University Press. Tilly, C., 1984. Big Structures, Large Processes, Huge Comparisons. New York: Russell Sage Van der Eijk, C. and M. Franklin, 1991. European Community Politics and Electoral Representation. Evidence from the 1989 European Elections Study. European Journal of Political Research 19: 105–28. Van der Eijk, C. and M. Franklin, et al., 1996. Choosing Europe? The European Electorate and National Politics in the Face of Union. Ann Arbor: The University of Michigan Press. Van der Eijk, C. and E. Oppenhuis, 1990. Turnout and Second-Order Effects in the European Elections of June 1989. Evidence from The Netherlands. Acta Politica 25: 67–94. Weiler, J. 1999. ‘The Transformation of Europe', reprinted on pp. 10–101 in Weiler, J. The Constitution of Europe. Cambridge: Cambridge University Press. Wellhofer, S., 1989. The Comparative Method and the Study of Development, Diffusion, and Social Change. Comparative Political Studies 22: 315–42.

Source: http://www.mzes.uni-mannheim.de/publications/wp/wp-83.pdf

Microsoft word - tesi-capitoli-14a.doc

POLITECNICO DI TORINO III Facoltà di Ingegneria Corso di Laurea Specialistica in Ingegneria Informatica Tesi di Laurea Specialistica Individuazione di situazioni di rischio in dati clinici rilevati da sensori Relatrice Prof.ssa Elena Baralis Candidati Daniele Apiletti Giulia Bruno Desideriamo ringraziare Tania, per l'aiuto e la disponibilità


HORHAM & ATHELINGTON PARISH COUNCIL: Tel: 01379 384625REDLINGFIELD PARISH MEETING (www.redlingfield.suffolk.gov.uk): Athelington, Horham Tel: 01379 678835 Email: [email protected] MID SUFFOLK DISTRICT COUNCIL (www.midsuffolk.gov.uk): Tel: 01449 724500 Email: [email protected] (0845 606 6067 & [email protected])