Resumé

In direct elections to the European Parliament, the behaviour of the Danish electorate, unlike that of other member countries of the EC, deviates strongly from the usual party pattern of national, general elections, thus producing a very differently structured Euro-party system. The existence of this distinct Euro-party system has been confirmed in the two consecutive Euro-elections of 1979 and 1984. This article analyses the special electoral behaviour of the Danish voters in Euro-elections. The deviation is explained primarily by the presence of a European cleavage in the Danish electorate which partly counteracts normal party identification. While supporters of Danish membership of EC vote just as they do in national, general elections, opponents of EC behave in a very different way. The voting in Euro-elections has become a mixture of election and referendum behaviour.

Introduction

The introduction, in 1979, of direct, universal elections to the European Parliament was an experiment in popular participation in international relations. The outcome in Denmark was very different from that of other European countries. In most countries the result of the Euro-election was similar to those of normal national elections. But in Denmark it deviated greatly from the usual pattern of general elections, and a differently structured Euro-party system emerged. The nature of this system was confirmed by the following Euro-election in 1984, proving the stability of its distinctive electoral behaviour.

The present study analyses this Danish Euro-party system on the basis of
survey data and examines the reasons for its strongly deviant voting
behavior.1

The European Elections

The European Parliament, the representative assembly of the European Community, is a mainly advisory and opinion-forming body with little power. Originally, its members (MEPs) were appointed by the national parliaments of the member states, with the Treaty of Rome stipulating a subsequent transition to universal, direct elections by all citizens of the member countries. The introduction of direct elections took more than 20 years, however, because it was a rather controversial issue.

---

*This article is a revised version of a paper delivered at a workshop on the European Election at the congress of ECPR at Barcelona, 25-30 March, 1985.

The purpose of direct elections was a democratization of the institutions of the Community, and two further consequences were expected: the voters would obtain a kind of direct European citizenship, which would produce an identification with the Community, and the popular mandate would endow the parliament with a legitimacy of its own, thus strengthening its demand for more influence and, eventually, real legislative power over European affairs. Thus, direct elections were considered a step in the direction of further European integration. They were, therefore, favoured by supporters of integration and opposed by those who were against political integration.

However, the direct elections in 1979 proved a great disappointment to the integrationists. They had expected an inter-European event, where the transnational party federations would campaign across borders on common platforms dealing with European issues,² while 170 million European voters would cast their votes on the basis of the European policies of the parties. These hopes were not fulfilled. The European voters manifested little understanding of or interest in the Euro-elections. Both in the first direct election in 1979 and in the following one in 1984 the turnout was extremely low. The voters tended to stick to their usual national party loyalities. And as political parties in any election seek to get as many votes as possible, the parties accommodated their electorates by concentrating their campaigns on national issues (Siune, McQuail & Blumler 1984). Euro-elections became a kind of opinion poll on party popularity.³ Like local and regional elections, they were only secondary ones, where many of the voters, lacking any clear idea of Euro-politics, tended to project party sympathies from the primary political arena, the national one. Some scholars have even made generalizations about the dependence of party strength in Euro-elections on party position in national politics, e.g., that the support for a party in government varies with the electoral cycle; or that small, new and radical parties get more votes in an election where there is little risk of political consequences (Reif 1984, 1985b).

Thus, the Euro-elections produced results which differed little from those of national elections and even less from concurrent opinion polls on party popularity. Only in Denmark were the results radically different from the normal voting pattern of the national party system. This was mainly due to the presence of a European cleavage within the Danish electorate.

The European Cleavage

Danish membership of the European Community is a highly controversial issue. But a time series of opinion polls on attitudes to EC (Table 1) shows that when Denmark first applied for membership in 1961, there was nearly a consensus on the matter: only eight per cent were opposed to EC while as

many as 40 per cent had no opinion. At the time, membership was considered a question of economic expediency or even necessity,⁴ while few were concerned with integration and national sovereignty. The application was supported by all the major parties in parliament in accordance with a tradition of consensus in foreign policy.⁵

But as the negotiations on membership were in their final phase in 1970-72, a surge of popular opposition arose. It was mainly concerned with the preservation of Danish sovereignty, but it was also marked by doubts aboui: the economic benefits, fear of involvement in great power politics and art assertion of national and cultural identity. As nearly all major parties and interest organizations were in favour of EC, the opposition was a kind of anti-establishment manifestation. Originally, the opposition had been limited to a number of small left-wing parties. But now anti-EC factions emerged even within the pro-European parties, and they were particularly successful in the Social Democratic and the Radical parties. The opposition forces were coordinated by an umbrella organization, the People's Movement Against EC.

The final decision on Danish membership was made by referendum on 2 October, 1972, after a long and heated campaign, in which opinion on EC became extremely polarized. Opinion polls showed a close run, but at the real polls a comfortable majority of 63 per cent voted yes to entry into the EC.

Soon after entry, general disappointment with the benefits, in the context of the world-wide economic crisis, caused a reversal of the trends. In the late seventies there was an even balance between supporters and opponents of EC, while nearly all opinion polls in the eighties have shown a considerable majority against the EC.⁶

The Danish political parties may be grouped into three blocs in terms of their European policies. The four bourgeois parties, which have formed the present government since 1982, Conservatives, Liberals, Centre Democrats and the Christian People's Party, are committed and unreserved supporters

of the EC and are not unfavourable towards further integration. Three other parties, the Social Democrats, the Radicals and the Progressives, are, in principle, supporters of EC, but they are reserved and hesitant about it, wishing to preserve the institutional status quo and rejecting any further transfer of power, political union and limitations of the veto. Finally, four small left-wing parties, the Socialist People's Party, the Left Socialists, the Communists and the Justice Party, oppose Danish membership of EC on principle, wishing to leave the Community and advocating a new referendum on withdrawal.

Table 2 shows that, although there are supporters and opponents of EC among the voters of all parties, the balance of opinion differs from party to party and, on the whole, -corresponds to party policy. There is only limited opposition to EC within the bourgeois parties, 8 to 26 per cent in the government parties and 33 per cent among the Radicals. Among both the Social Democrats and the Progressives, the opponents have clear majorities with margins of 11 to 13 per cent. Finally, there is an overhelming majority of opponents in all the anti-EC parties.

Table 2 shows that between the two Euro-elections the opposition to EC
has grown among the voters of nearly all the parties.

One consequence of the opinion balances in Table 2 is that the Danish parliament, the Folketing, is not very representative of the voters, as far as EC is concerned. While half of the voters are against Danish membership of EC, only one sixth of them vote for a party opposed to it. This is a consequence of the very low salience of European issues in Danish national elections: even if voters are against the EC, they vote without scruple for a pro-European party. But in European elections this is not necessarily the case.

The

The introduction of direct elections was opposed by the anti-EC parties, which claimed this would 'strengthen EC integration and give it a deceptive gloss of democracy'.⁷ The status quo parties were reluctant. The Social Democrats succeeded, as a government party in the seventies, in delaying direct elections for years and tried to link them to national elections in two ways: the same election day and 'double membership', i.e. membership of the Danish Folketing would be a condition for membership of the European Parliament. The bourgeois parties favoured the direct elections, wishing to increase the prestige and influence of the European Parliament.

For the political parties the direct elections were a challenge, because they had to face the voters in very unfamiliar circumstances. The situation was especially problematic for the EC opponents, who faced a number of controversial, strategic choices. The first one was whether to run at all. Some felt that participation would imply a tacit acceptance of the direct elections and thus would contribute to legitimizing them. So they advocated abstention and even started a boycott movement to campaign for it: a low turnout would be the best manifestation of opposition to the EC.

But the anti-European parties did not want to leave Danish representation exclusively in the hands of the EC supporters, and they were tempted by the opportunity to mobilize the widespread opposition to EC. 'The Justice party does not seek elections to the European Parliament because it expects to influence anything down there. Our purpose is to offer the Danish people an opportunity to send a representation to the European Parliament which reflects the attitude of the Danes to the EC, says one party platform. But participation implied another problem. The anti-EC parties belong to the socialist left wing, which is unable to get more than 15 per cent of the vote in general elections. It seemed improbable that even this occasion would enable them to attract any sizeable increase of the vote from voters unsympathetic to their radicalism. Therefore, it was decided - alongside the party tickets - to enter an additional cross-party ticket of anti-European candidates. For this purpose was chosen the old referendum-campaign organization, the People's Movement Against EC (PM).

The PM had no platform, and its candidates were pledged to abstain from voting on any question not concerning the powers of the EC. This was a very unusual pledge from a party participating in an election, but according to its election platform, 'The People's Movement is not a political party, and the European election is not a party affair, but an opportunity to demonstrate the people's attitude to EC. It is important to underline that the Folketing today does not represent the people's attitude to EC.'⁸ The purpose was, thus, to transform the Euro-election into a new referendum for or against EC-membership.

The participation of the PM entailed a third controversy. Most of the anti

parties wanted the PM to nominate a list of independent personalities of distinguished reputation as well as dissidents belonging to the pro-European parties. But the PM decided to also include members of the anti-EC parties on its ticket, who would thus be competing with their own parties. The most controversial step, however, was the decision of the Communists not to enter their own ticket, but to recommend their members to vote for Communist candidates on the PM list. This provoked allegations that they wanted to transform the PM into a vehicle of Communist influence. But at the second election another minor party, the Justice Party, made a similar decision. The strategy proved successful: while neither of the two parties was able to win any of the 179 seats in the 1984 election to the Danish Parliament, they each won one of the 16 Danish seats in the European Parliament as PM candidates.

The entry of the PM constituted a threat to those EC parties which included many opponents within their ranks, especially the Social Democrats. The Social Democrats, therefore, adopted a very different approach, which emphasized that the Euro-election had nothing to do with Danish membership or the institutions of the EC. These matters were the exclusive prerogative of the Folketing. The aim of the Euro-election was to influence the general politics of Europe in the same direction as national policy, and therefore people ought to vote for the same party in European as in national elections. On the other hand, the Social Democrats were worried by the prospect of a mass defection of EC opponents among its voters. Therefore, in the first Euro-election it allowed candidates who were against EC membership to be nominated alongside other candidates who were in favour of it. Candidates were permitted, furthermore, to campaign for their own point of view and even to set up separate campaign organizations.

The disappointing results of the first election caused the Social Democrats to alter their position during the second: the candidates were no longer allowed to express personal views; the party should speak with one voice; EC membership was considered a closed matter not to be discussed during the campaign, but reservations regarding political union were emphasized. Most of the organized EC opposition within the Social Democratic Party joined the People's Movement.⁹

The Radicals and the Progressives were in a similar dilemma, and they faced it in the same way: they nominated both pro- and anti-European candidates and stressed their opposition to integration. The bourgeois parties presented a strongly pro-European attitude, and the Centre-Democrats, in particular, launched themselves as unconditionally the most super-European party.

The 16 Danish members of the European Parliament are elected by a
system of proportional representation, in which the candidates on each ticket

are listed in rank order, and in which the entire country constitutes a single
constituency.

Danish parties were not accustomed to nominating nationwide tickets, but most of them established a two-stage procedure, where candidates were proposed at county level, while the final ranking of the candidates was decided at a national convention. Most county candidates were local party workers unknown to the public, but the parties tried to include at least one well known and popular name as a figurehead. The Social Democrats chose their vice-chairman, and several parties selected retired party leaders, some minor parties even proposed their current leaders.¹⁰ At the second Euroelection most parties renominated their MEPs. But as there had been little publicity surrounding their work in Strasbourg, they had not built much of a reputation.

With the whole country serving as a single constituency, all the voters have the possibility of voting directly for any of the candidates. As only a few of the candidates had any nation-wide reputation, the direct, personal votes were concentrated around one or two persons on each party ticket. In the first election, one fourth of the entire vote was cast as personal votes for the three most popular candidates. That exceeded the total vote for six minor parties. Since almost all of the personal votes went to the figurehead candidates, who headed the ballots, the second and third places were decided by very few personal votes.

Campaigning in Euro-elections was an unfamiliar task for the party organizations, and the Danish electorate was rather confused with regard to the issues in Euro-politics, even at the second election. Euro-elections receive the same TV coverage as general elections, but they are followed with much less interest and attention. As it was not easy to find Europolitical issues of general interest for the voters, the campaign focused on the

advantages or disadvantages of membership as such. A content analysis of the television campaigns in 1979 and 1984 shows a marked shift in the focus from European to domestic issues. While the candidates in 1979 devoted 63 per cent of their TV time to purely European themes, these took up only seven per cent in the 1984 campaign.₁₁

The campaigns were not successful in arousing the interest of the voters. Table 3 shows that attention to the election campaign in television was much lower than in general elections, and that it declined considerably from the first to the second Euro-election. Discussion of the election with friends and family was much less frequent. Nevertheless, half of the voters had discussed the election, and two thirds had seen at least one TV programme. However, the most strikning expression of the low involvement in Euro-elections was the turnout: while there are usually 10 to 14 per cent non-voters in general elections, only half of the eligible voters cast their votes in the Euroelections.

The Euro-Party System

While the European elections in all the other countries produced results very close to the pattern of their national general elections, the Danish results exhibit considerable differences (Table 4).¹¹ The most conspicuous was the success of the new party existing only at the European level, the People's Movement Against EC, which obtained 21 per cent of the vote. Three bourgeois parties received more votes than in national elections; the Liberals and the Conservatives in 1979 increased their vote by 16 and 12 per cent, respectively, and strongly pro-European Centre-Democrats nearly doubled their vote. The other parties lost many votes in the Euro-election: the Socialist People's Party retained only 80 per cent of its usual vote; the Social Democrats, the Radicals and the Progressives just a little more than half. The results were most disastrous for the largest Danish party, the Social Democrats, which in the first Euro-election polled only 21.9 per cent, the lowest figure in this century, and five years later lost an additional 2.3 per cent, taking only third place, with less votes than the Conservatives and PM.

The pattern of deviation in the first Euro-election was repeated in the second one, in 1984. Of course, there were some differences, but they corresponded to the fluctuations in party popularity reflected in the national election results of that year: an increase for the Conservatives and People's Socialists, a decline for the Social Democrats and the Progressives. The fluctuations from one election to the other were, however, much smaller in the case of the Euro-elections: only 51 per cent of the national variation for the Conservatives; and for the other three, respectively, the percentage was 75, 36 and 31.

Fluctations in party popularity on the national level are apparently only partly reflected in the Euro-elections, somewhat similar to the way in which they are reflected with only half strength in local elections.¹² There has been a certain convergence between 1979 and 1984, because the changes in the general system approached the structure of the Euro-party system.

There is a simple pattern behind the differences between the national and the European levels of the party system, which is shown in the bottom section of Table 4. Here the parties are grouped according to their European policies, which seem to provide the clue to their performance in the Euroelections: the committed pro-European, bourgeois parties do better in the Euro-elections than in the concurrent national elections; also while the anti parties do lose some votes, the entire anti-EC bloc, including the PM, doubles its vote, compared to what the anti-EC parties can muster in national elections.

Finally, the status quo parties suffer a crushing defeat, with only 57 per cent of their usual national vote. So it seems that the two committed wings in the European cleavage have good Euro-election results, while the reserved Europeans in the centre are unable to transfer their usual national following to the Euro-level.

The differences between the two party systems should, however, be considered in connection with the fact that the turnout in the Euro-elections is only half as large as in national elections. This means that no party is able to mobilize more than a fraction of its normal vote: the bourgeois parties two

thirds, the anti-EC parties less than half and the status quo parties only one
third.

In order to understand the deviations of the Euro-party system, it is,
therefore, essential to consider turnout.

The Turnout

A lower turnout in Euro-elections is to be expected because much less is at stake than in general elections. Euro-elections involve an advisory body with few Danish members and very little impact on the voters, who are unfamiliar with and uninterested in the issues of Euro-politics, and who pay very little attention to the election campaign (cf. Table 2). The motivation to vote is much weaker than in national elections, even local ones.

In general elections participation depends on the voter's level of involvement and political interest. This is also the case in Euro-elections. The range of the turnout, however, is much greater in the Euro-elections: it varies from 70 per cent among those who are most interested to only 18 per cent among the least interested. In national elections, the respective percentages are 96 and 62. This difference in variation, even so, has hardly any effect on the political results.

Table 6 shows that political factors influence the turnout, too. Supporters of EC are more motivated to participate than opponents: their turnout was eight per cent higher in the first and 12 per cent higher in the second Euroelection. Voters with no opinion on EC have by far the lowest turnout.

There are considerable differences between the parties' voters, too. The committed parties on both wings have a much higher turnout than the reserved status quo parties. In 1979 the participation was 60 per cent for both wings, compared to only 43 per cent for the centre; and in 1984 the difference increased, as turnout rose six to seven per cent on the wings, but only two per cent in the centre.

According to Table 6, the turnout is higher among the voters for the anti
parties than the pro-EC parties. This seems to be incompatible with the

finding that participation is higher among supporters of EC than among opponents. The explanation is found in a combination of the two variables, party preference and attitude to Europe. Voters who agree with their party on European policy have a greater incentive to vote. Voters who disagree with their party on the EC are subject to conflicting influences, to cross-pressures, and thus inclined to abstain.

Table 7 shows the combined effect of party preference and EC attitude on turnout in the Euro-elections. It proves that within both the bourgeois parties and the centre parties the turnout is considerably higher among supporters of EC who agree with their party, than among opponents who disagree. In 1979 the differences were 15 per cent and five per cent, respectively, and in 1984 they had increased to 20 per cent and 15 per cent in the two party groups. Among the anti-EC parties the relation is reversed: here the turnout is highest among the opponents who are in agreement with their party. Therefore, participation peaks among the supporters of party policy within both committed blocs, namely the pro-EC bourgeois voters and the EC

opponents in the anti-EC parties. The lowest participation occurs among EC
opponents from the centre parties: 20 per cent below the peak groups in the
first election and 30 per cent in the second.

The connection between attitude to Europe and turnout is not a unique Danish phenomenon; it occurs in all EC countries. But in most of these the opposition to EC is limited to a few per cent, so it does not affect the result. Only in Britain is the opposition of the same size as in Denmark. Here the turnout in the first Euro-election was 48 per cent among supporters and 24 per cent among opponents, and this difference contributed considerably to the Conservative victory (Inglehart & Rabier 1984, 497).

Differential turnout, seen as the combined effect of attitude to Europe and the compatibility of this attitude with normal party preference, seems to acount for much of the difference between national and the Euro-party systems. These very same factors also effect party switching.

Party Switching

Only 63 per cent of the Danish voters vote for the same party in national
and European elections.

Table 8 shows a very similar pattern of vote shifting between party blocs in the two Euro-elections. The bourgeois parties retain nearly all their usual voters (88 per cent), but lose a few to the PM (6-8 per cent). The centre parties retain far fewer of their voters (65 per cent), they lose some voters to the bourgeois bloc (7 per cent) and a few to the anti-EC parties, but many more to the PM (22-25 per cent). Finally, the anti-EC parties have the lowest stability of all: they lose half of their voters to the PM.¹³ It seems from Table 8 that it is mainly the varied attraction of the PM that explains the differences in voting stability between the party blocs.

This pattern of defection is evidently related to differences in attitude to Europe. Table 9 shows that at national elections EC supporters give two thirds of their vote to the bourgeois parties and one third to the centre parties. In Euro-elections both bourgeois and centre parties retain their share; of the EC supporters. The EC opponents, on the other hand, give one sixth of their vote in national elections to the bourgeois parties, one half to the centre parties and only one third to anti-EC parties. It is the two thirds of the EC opponents, who in national elections vote for pro-EC parties, that constitute the decisive group in the Euro-elections, the marginal voters. Table 9 shows that support for the bourgeois parties among EC opponent voters is reduced to nearly half its size in Euro-elections, the centre parties lose two thirds, and the anti-EC parties one third, while the PM succeeds in mobilizing half of the EC opponents' vote. Thus the strategy of running a crossparty anti-EC ticket proved successful: the pro-EC parties which were supported by two thirds of the anti-EC voters in the national elections kept only one fourth of them in Euro-elections. But this remnant, combined with the effect of the higher turnout among EC supporters, was sufficient to obtain two thirds of the vote and MEP seats for the pro-EC parties.

It is apparently the combination of party preference and attitude to EC
that provides the key to an understanding of voting behaviour in European
elections. In Table 10 both the bourgeois and the centre party voters are

subdivided according to their attitude toward EC. A very similar pattern
appears in both blocs.

The EC supporters vote nearly unanimously for their usual party, and the minor defections that occur go to the other pro-EC bloc. And while defection from the firmly pro-EC bourgeois parties to the centre is only five per cent, the losses suffered by the more ambivalent centre parties to the bourgeois bloc amount to 12 per cent. On the other hand, both blocs lost more than half of their EC opponents to the PM. In 1979 they lost another 10 per cent to the anti-EC parties, the centre retaining only one third and the bourgeios parties only on fourth of their usual vote. In 1984 both blocs were a bit more successful in keeping their anti-EC voters.

The preceding analysis demonstrates that the differences between the national and the Euro-party systems are the result of both differential turnout and party switching, and that both are influenced by the same set of variables: attitude to EC, normal party preference and the compatibility of the two. Table 11 shows the combined effects of party preference and EC attitude to both Euro-party choice and turnout, and thus combines all the factors examined in this analysis. It can be considered as a complete model of Euro-electoral behavior and is able to account for nearly all the differences between the national and the European level of the party system. What is particularly notable is the combined effect of a very high degree of abstention and a very high rate of defection to the PM among the opponents of EC from the bourgeois and, in particular, from the centre parties. The latter lost 80 per

cent of the EC opponents among their usual voters. At the other extreme, a low degree of abstention and a high degree of stability among the bourgeois EC supporters allow these parties to retain 62 to 71 per cent of their usual voters.

The good Euro-election results of the bourgeois parties are the effect of overwhelming support for EC among its voters, as well as a large turnout and no defections among its EC supporters. The poor results of the centre parties are due to a majority of EC opponents among their voters, an extremely high degree of abstention and a very high rate of defection among the opponents, but also considerable abstention and defection among its EC supporters. The: voter loss of the anti-EC parties is due to very heavy defection to the PM, which is only partly compensated by the very large turnout. Finally, the People's Movement is able to attract half of the votes cast by opponents in all three blocs.

Conclusion

Two European elections proved that the Danes, unlike their European
neighbours, behave very differently in national and European elections, and

that the party structure that emerges from the two kinds of election also
differs.

This study has examined the causes of these differences in voting pattern. It was established that they were mainly due to the influence of the ECcleavage: many voters consider Euro-elections as a kind of referendum, an opportunity to say yes or no to Danish membership of EC. This makes the EC opponents within the pro-EC parties the decisive marginal group in the Euro-elections, a group subject to cross-pressures of conflicting inclinations. Constituting one third of the electorate, they provide the key to the result: their extremely low turnout and heavy defection to the PM account for most of the differences between the two levels of the party system.

There are two remarkable features of the deviant Danish Euro-election behaviour. One is that it has no impact whatsoever on party popularity in national politics, not even on the opinion polls taken at the same time as Euro-elections. The defection in the Euro-elections does not alienate the voters from their normal party preference. The other feature is the stability of the deviation. In 1979 one might have expected it to be a short-term, protestlike form of behavior, to be followed by a return to the normal pattern, when people became accustomed to the new institution. But all the tables in this study show that behaviour in the second election, five years later, followed exactly the same pattern as in the first. It was a confirmation of the existence of a distinct Danish Euro-party system based on deviant voting behaviour.

---

NOTES

1. This study is based on a number of election surveys conducted by a research group from the Universities of Copenhagen and Århus. In 1979 there were both pre-and post election surveys. The field work was carried out by the Danish Gallup Institute. In 1984 a postelection survey was conducted through Gallup. Data from a telephone survey carried out by AIM for the Danish Radio have also been consulted. Each survey included about 1000 respondents.

2. The parties of the EC countries cooperate within transnational party federations (e.g., Socialists, Liberals, Christian Democrats), which have common platforms and constitute special groups in the European Parliament.

3. Electoral behaviour in European elections has been analysed country by country in Reif (1985) and in a special issue of Electoral Studies (No. 3, 1984). Comparative analyses are found in Reif & Smidt (1980), Reif (1984, 1985), Inglehart & Rabier (1979), Blumier & Fox (1982).

4. Denmark had to follow Britain, its main market.

5. The intemal political arena in Denmark is dominated by a cleavage between two political camps, alternating in government, the Social Democrats and the Bourgeois parties (Liberals and Conservatives). But since entry into NATO in 1949, there has been a tradition of cooperation between both camps on foreign policy.

---

6. In spite of this, there has been a majority of supporters in both Euro-election surveys; and in the second European referendum on the EC package on 26 February, 1986, there was another yes majority, this time 56 per cent.

7. Election manifesto of the Socialist People's Party, 1978.

8. Quoted from the election manifesto of the People's Movement adopted by its national convention on 29 October, 1978.

9. The organized Social Democratic EC opponents suggested that the People's Movement should participate in the Danish general elections and even considered starting an anti-European Social Democratic splinter party.

10. In 1984 the Progress Party ticket was headed by the party leader, who had been convicted and imprisoned for tax fraud and expelled from parliament, and who now conducted his Euroelection campaign from his prison cell. The Centre Democratic leader has been a member of both European Parliaments.

11. General elections were held in the same years as both the European elections, and they afford a very convenient basis for comparison between the national and the Euro-party systems. On the other hand, four or five months elapsed between the two elections, so some of the differences in the results might be due to fluctuations in party popularity in the meantime. But in the Euro-election survey the respondents were also asked how they would vote in a hypothetical general election, and the results of this poll differ only slightly from the actual general election.

12. Analysis of the fluctuations in Danish general and local elections shows that the changes in party popularity on the primary, national level, on average, are followed by changes of half the size at the secondary, local level (Risbjerg Thomsen 1984). One election period is, of course, insufficient for making any generalization about the European level.

13. It should be remembered that some of the smallest of these parties do not participate in Euroelections, so the most natural choice of their voters is the PM.

REFERENCES

Blumler, J.G. & Fox, A.D. 1982. The European Voter. Popular Responses to the First Community
Election. London: Policy Studies Institute.

Borre, O. 1984. 'Denmark', Electoral Studies 3, 268-73.

Borre, 0., Elklit, J. & Tonsgård, O. 1979. 'The Danish Election to the European Parliament in
June 1979: A New Referendum?", Scandinavian Political Studies 2, 299-310.

Elklit, J. & Tonsgård, O. eds., 1984. Valg og vælgeradfærd. Århus: Politica.

Inglehart, R. & Rabier, J.R. 1979. 'Europe elects a Parliament: Cognitive Mobilization, Political
Mobilization and Pro-European Attitudes as Influences on Voter Turnout', Government and Opposition
14, 479-507.

Reif, K. 1984. 'National Electoral Cycles and European Elections 1979 and 1984', Electoral
Studies 3, 244-255.

Reif, K. ed., 1985a. Ten European Elections. Aldershot: Gower.

Reif, K. 1985b. 'Ten Second-order Elections', in Reif (1985a), 1-36.

Reif, K. & Smitt, H. 1980. 'Nine Second-Order National Elections: A Conceptual Framework for
the Analyses of European Election Results', European Journal of Political Research 8, 3-44.

Siune, K. 1984. 'EF-valg: Vælgerpåvirkninger og vælgerreaktioner', in Elklit & Tonsgård (1984),
353-68.

Siune, K. McQuail, D. & Blumler, J.G. 1984. 'Broadcasting European Elections', Electoral Studies
3, 256-63.

Sørensen, CL. 1984. 'Danmarks EF-valg i 1979 og 1984', in Elklit & Tonsgård (1984), 338-52.

Thomsen Risbjerg, S. 1984. 'Kommunale valg' in Elklit & Tonsgård (1984), 288-297.