In Denmark the direct election to the EC Parliament on June 7, 1979 had two distinctive features. First the poll of 47 per cent was extremely low compared not only with the normal Danish level, but also with that in the other EC countries, even when the various different electoral systems are taken into consideration. Second, a non-party list with Denmark's withdrawal from the EC as the only plank of the platform achieved 21 per cent of the votes; the largest Danish party, the Social Democratic Party, received only 21.9 per cent of the votes. Together, the parties that are against continuing membership of the EC were supported by 1/3 of the electors.

In this article we shall describe the election result and the participation of the electorate. By comparing the referendum for Danish membership of the EC in 1972 with the last general election in 1977 we shall point out characteristic ecological patterns. Finally, the results will be analysed using survey data collected on the day of election.

For this article we have several types of data: the official national result with an ecological analysis based on the official constituency results, and also a set of individual data from a survey conducted on the day of the election. At four selected polling places we asked the voters to fill in a short questionnaire after they had participated in the official poll. The questionnaire gives us the following voter characteristics: sex, age, occupation, voting at the general election in 1977, voting at the direct election to the EC parliament, and responses to three attitude questions - on continuing Danish membership of the EC, the possible development of the EC into an economic union, and the development of the EC into an actual European Federal State. The four polling districts were very different geographically, socially, and in their political complexion. We cannot say that it is a proper representative study, but on the other hand there is no basis for thinking that the relations and tendencies found in this material are not valid for Denmark as a whole. We collected data on 4876 persons corresponding to an average response rate of 79 per cent.

1. Participation in the Election

The very low participation rate of 47 per cent must be considered surprising. The
Danish electorate normally acknowledges a social norm prescribing participation

in political elections and referenda. Since the introduction of proportional representation in 1918 the lowest participation at a general election has been 74.9 per cent, and during the past 20 years it has not fallen below 85 per cent. Participation at the last general election in 1977 was 88.9 per cent. After a drop in 1974 participation at local elections in 1978 was again above 70 per cent, and even referenda on relatively depoliticized issues have aparticipation rate of about 65 per cent. At the referendum in 1978 on lowering the voting age to 18, 63 per cent of the electors participated.

The extraordinarily low participation in the United Kingdom is partly a result of its special voting system, which may have a very distinct effect when the number of seats being contested is as low as was the case at the direct election to the EC parliament. This was not the case in Denmark where the 15 seats are elected by proportional representation, and one may suppose that only a few of the voters were aware that at the EC election a different method of distribution was used, compared with the one used at the general elections - one which discriminated against the smaller parties.

The low participation in the election in most of the EC-countries may be an expression of a common feeling among electors that the EC, and not least the EC parliament, is something distant, irrelevant, and confusing, and on which the individual has no influence. Opinion polls appear to confirm that this attitude is common in Denmark. A negative or sceptical attitude towards the EC has been noticeable during the entire period of membership since 1973, and was also expressed in the Danish opinion polls before the direct election to the EC parliament. According to most polls, about 1/3 of the electors had no definite idea of how to vote, or did not intend to vote right up to the election. In addition, abstainers are usually underestimated by 10 per cent in opinion polls, so almost half of the electorate had no definite attitude on how they were going to vote. The number of those in doubt was thus larger than before general elections and also larger than before the referendum on Danish membership in 1972. The parties did not succeed in mobilizing their support, and this is the main reason differentiating this EC election from the other elections and referenda mentioned. In 1972 it appeared that on the day of the referendum the undecided voted for the party for which they had voted at the previous general election. Several factors indicate that the convinced opponents of Danish membership were mobilized while the two parties with the most vague profile on EC questions, the Social Democratic Party and the Progress Party, were damaged by the low participation.

2. The Main Results of the Election

Before the referendum in 1972 on Denmark's membership of the EC an umbrella organization of opponents was formed. It comprised mainly the three left-v/ing parties, Communists, Left Socialists, and Socialist People's Party, together with the orthodox liberal Single Tax Party. Furthermore, the People's Movement against the EC, as it was called, had the support of a few members of the Radical Liberal Party which was divided on the EC question, and also a few from distinctly pro-European parties such as the Conservatives and the Liberals. The Social Democrats were also split on this issue and some formed their own 'anti' organization which at no time worked with the popular movement.

Since the 1972 referendum the People's Movement against the EC has continued

to work for a Danish withdrawal from the EC, and nominated its own list at the direct election to the EC parliament. Of the 20 candidates on this list 7 were not party members, while of the remainder, three were social democrats, two radicals, and from the distinct pro-parties one liberal and two conservatives. The Communists did not have their own list and were represented on the People's Movement's list with two candidates.

The heterogeneous list of the People's Movement, a single issue list, was the second largest party with 21.0 per cent of the votes, surpassed only by the Social Democrats with 21.9 per cent. The anti-EC parties received a total of 32.6 per cent of the votes. A comparison of voting at the 1977 general election and the 1979 EC election shows a remarkable agreement with a few large deviations. The Social Democrats account for the largest difference with a drop of 15 per cent in its support; the Progress Party lost more than half of its support with a decrease of nearly 9 per cent. Both parties are split on the EC issue. However, the Progress Party nominated supporters as well as a few opponents, while the Social Democratic Party nominated only supporters of Danish membership. On the other hand, the People's Movement's list nominated some prominent social democrats who, furthermore, were excluded from the party as a result of this nomination. Apart from the considerable number of votes for the People's Movement, one can see a meaningful success for the Conservative Party which increased its vote by almost 6 per cent as compared with the general election in 1977.

Table 1 gives the distribution of both the national vote and seats among the parties. The relationship between votes and seats was not as straight forward as in general elections. At general elections Denmark uses the largest remainder method to determine the number of additional seats necessary to ensure proportional representation for those parties which have passed one of the three thresholds in the electoral law. At the direct election to the European Parliament it was decided to use the d'Hondt formula for the 15 seats to be distributed in Denmark outside Greenland. Compared to the largest remainder method this discriminates against smaller parties.

The whole country formed one electoral district. Electoral alliances were allowed, and two were formed. One consisted of the four bourgeois proponents of Danish membership: the Conservative People's Party, the Liberals, the Centre Democrats, and the Christian People's Party. The other comprised the four parties or lists against Danish membership: the People's Movement against the EC, the Socialist People's Party, the Left Socialist, and the Single Tax Party. The two alliances won 6 and 5 seats respectively while the Social Democratic Party won 3 and the Progress Party 1. As seats were also distributed within electoral alliances according to the same formula, the bigger parties gained more than their proportionate share, with 4 seats to the People's Movement and 3 seats to the Liberals in the other alliance. In both cases this was one seat more than they would have received if the largest remainder method had been used. Without electoral alliances these seats would have been won by two parties on the anti-EC side, the Single Tax Party and the Left Socialists. The formula used at general elections would have ensured the election of 6 members of the European Parliament opposed to the EC, compared to the 5 now elected.

3. Ecological Patterns

At the time of writing only constituency level data are available for studying aggregate voting behaviour. Denmark has 103 constituencies, consisting of one or more municipalities or parts of the larger municipalities. The constituencies are normally used only as nomination districts at general elections, but at the elections to the European Parliament they were used as counting units and therefore voting results could fairly easily be made available at this level.

The main points to be discussed in this section are: (1) turnout, which was extremely low when compared to normal Danish standards, (2) the relative success of the People's Movement against the EC, and (3) the remarkable losses of the Social Democratic Party compared to other Danish elections.

On the first point, it can reasonably be argued that those political-ecological structures, which, for example, at one election produce a low turnout, will also tend to do so at other elections at the same level. We shall begin by considering the direct elections to the European Parliament as more comparable to general elections than to other elections. The first hypothesis tested was: constituencies with a low turnout in the 1977 general elections are those with a low turnout at the direct elections. However, when turnout at the two occasions are plotted against each other, we find a rather weak relationship between the two variables. * This plot was made on the basis of all 103 constituencies, but one should also note the possible existence of differences between the three main regions into which Denmark is divided for (other) electoral purposes, i.e. the Capital, the Islands, and Jutland

with 19, 39, and 45 constituencies, respectively. At the regional level, the connection
between the two sets of turnout is more obvious, at least outside the Capital.

If we examine the relationship between turnout in the referendum about Denmark's entry into the EC in October 1972 and turnout in the present election, a closer connexion is apparent. At the national level, the slope of the regression line is 1.04 and the simple linear model explains 14 per cent of the variation in 1979 turnout. This might lead us to the tentative conclusion that in this respect the direct elections to the European Parliament have more in common with the referendum seven years ago than with more recent national elections. This conclusion is substantiated when Jutland is considered alone. For the Islands, the results for the two comparisons are fairly identical, but for the Capital they are more inconclusive, a fact which might be connected with the small number of constituencies in that region.

Turning then to the second main question, the relative strength of the opponents of Danish membership of the EC, we would expect that it - the strength of the opponents - would be connected both to the combined strength of the four political parties opposing the EC (as measured by their combined 1977 vote) and to the relative strength of No-votes in the 1972 referendum. Again we have plotted the percentages of the electorate voting with these groups at the various elections and the referendum against each other. From the inspection of the scattergrams it becomes evident that the relationship between percentage of No-votes in 1972 and percentage voting for the EC opposition in the direct elections, and between percentage voting for either the leftist parties or the Single Tax Party in the General Election of 1977 and percentage voting for the EC opposition in June 1979, is indeed very strong.

This is not only the case when we consider all 103 constituencies. The, evidence is equally impressive within each of the three regions. For the whole country we find a slope of .00, when 1977 and 1979 data are taken together, and r2r² is as high as .79. We can therefore talk of a very strong and important relationship, even though it is not very surprising. The same overall picture is found within each of the three regions, but of course the results vary a little across regions.

When we then turn to the relationship between the 1972 referendum and voting results in 1979, we find the same picture both on the national level and within each of the three regions: the slopes of the regression lines are smaller but the percent of variation explained is higher than when we examined the two elections of 1977 and 1979. The plot of the 1972 referendum with the 1979 election figures for the 103 constituencies is shown in Figure 1. The slope here is .66 and r2r² is .86, indicating again that the simple linear model can explain not less than 86 of the variation.

These findings - about both turnout and strength of the opposition against the EC - point to a partial conclusion that the direct elections to the European Parliament have as much in common - or probably even more - with the referendum than with a national election. The political-ecological analysis thus agrees with a point sometimes made, that the European dimension of Danish politics is easily separated from other- and more salient - political dimensions by the voters.

A third point worth considering is the poor performance of the Social Democrats, far behind their normal results at elections and their recent opinion poll figures. Here we have run the 1977-1979 Social Democratic losses (in per cent of the total electorate) against drop in turnout at the same two elections, the following

of the People's Movement, and the 1972 No-vote. The results with regard to a possible connection between Social Democratic losses on the one hånd and drop in turnout and 1979 turnout on the other are more inconclusive than was the case above. This is not surprising, since anumber of factors related to these phenomena tend to point in opposite directions. These results are not only found when all 103 constituencies are taken together, but also in the largest of the three regions, Jutland. The two remaining regions do, however, reveal a relationship.

When we then compare Social Democratic losses at the constituency level with the percentage voting for the People's Movement against the EC, as well as with the 1972 No-percentage of the total electorate, much more uniform patterns are found. Both nationally and regionally the evidence is very clear. The slopes and r²'s are very similar; when we plot No-vote in 1972 against Social Democratic losses the slope is .46 and r2r² is as high as .70 (p < .001). The main conclusion here is as obvious as it was when turnout and strength of the EC-opposition were discussed: the existing social, political, and cultural structures tend to produce patterns of participation and voting which at the constituency level link this first direct election to the European Parliament with the referendum in 1972. This is in accord with the view that these elections have at least as much in common with the 1972 situation as with recent general elections.

4. Individual Party Choice

Our account of individual voting behaviour at the 1979 election is based on a survey of the voters at four polling stations: one was located in Enghave, an old and densely populated district of Copenhagen; another in Svendborg, a town of about 25,000 inhabitants on the island of Funen; and the remaining two in the village of Kjellerup and a nearby rural district, Vinderslev, in central Jutland. At these stations the voters, after voting, passed through a room in which they were asked by a staff of interviewers from the Danish National Institute of Social Research to fill in a short questionnaire. A total of 4876 responses were collected, giving a response rate of 76 per cent in Enghave, 74 per cent in Svendborg, and 90 and 91 per cent in Kjellerup and Vinderslev. This amounts to an overall response rate of 79 per cent of the turnout after excluding the postal vote. An analysis of these responses was presented in the TV broadcast of the election on the night of June 10, the intention being to give the public more information on electoral behaviour than could be gathered from the official results.

The four polling stations were naturally selected with a view to giving a combined breakdown of the vote for the parties roughly approximating the national election result. But as we had to select polling stations of a manageable size with a view also to the arrangement of the buildings etc., and as we also wanted to cover the three main regions in the country, it can be no surprise that the objective of a party distribution as close as possible to the nation vote was reached only to some degree. The figures in Table 2 permit a comparison of the partisan distribution of the survey with the four polling stations combined, and finally with the national result. The correspondence between the survey and the combined result of the four stations is as good as it probably can be, and it is seen that the deviation between the survey and the national results is due only to the difference between the four polling stations and the national result. The main differences are the överrepresentation of the anti-EC electoral alliance by 6 per cent and a corresponding

underrepresentation of the bourgeois alliance. Of the various parties and lists it is
particularly the People's Movement against the EC and the Social Democrats
which had better support in these districts than in the whole country.

The questionnaire and the whole arrangement at the polling stations were so designed as to permit the voters to reply in writing in a few minutes. The questionnaire gave information on the voter's sex, age group, occupational group, partisan choice at the EP election, partisan choice at the last general election, in February 1977, and attitude toward Danish membership in the EEC and toward European integration as measured by two-way responses to the following three questions:

'Are you for or against the European Common Market developing into an
economic union with common economic policy for all EEC countries?'

'Are you for or against the European Common Market in due time developing into
a federal state, that is, a kind of the United States of Europe?'

'Are you for or against Denmark's continued membership of the European Common
Market?'

The pattern of responses to these three questions suggests that they tend to tap the same dimension. Of those in favour of a federal state, 93 per cent were also in favour of an economic union, and 96 per cent were in favour of continued membership. Of those in favour of an economic union, 92 per cent were in favour of continued membership. Hence, once continued membership is accepted, the idea of an economic union can be looked on as a further step toward European integration, and the idea of a federal state can be regarded as a still further step. On the question of continued membership 48 per cent were in favour and 47 per cent against, with 5 per cent don't knows. With regard to an economic union, 40 per cent were in favour and 50 per cent against, with 10 per cent don't knows. Finally, with regard to a federal state 28 per cent were in favour and 62 per cent against, with 10 per cent don't knows.

The character of the EP elect iono_n as a choice for or against Danish membership of the EC, even though this was not formally an issue, is disclosed by Table 3. Among those in favour of continued membership, only 2 per cent voted for one of the four lists which made up the anti-EC alliance. Among those opposed to continued membership, 79 per cent did so, 51 per cent choosing the People's Movement and 28 per cent one of the three minor parties.

This high level of consistency between the voter's attitude on the European issue and his or her choice of party goes a long way toward explaining the success of the People's Movement and the relative failure of the two largest parties in the system, the Social Democratic Party and the Progress Party. Both included a large proportion of EC opponents among their 1977 voters and would therefore be vulnerable in a situation in which the electorate became polarized on the issue of Danish membership of the EC. The flow of the vote between the 1977 election and the EP election is shown in Table 4.

When the 1977 and 1979 vote are cross-tabulated with each other for those respondents who reported a partisan choice on both occasions, the parties being grouped according to their alliances at the latter election, we find that almost all who gave their 1977 vote for one of the anti-EC parties went on to vote in 1979 for either the People's Movement or the parties in the anti-EC election alliance. By

contrast, the Social Democrats lost almost one-fourth of their former voters to the People's Movement and another 7 per cent to the anti-EEC parties, while the Progressives and Radical Liberals lost almost one sixth of their former voters to the People's Movement. Only the four parties that formed the Bourgeois election alliance managed to hold on to a large majority of their former voters, and in addition these parties attracted over one fifth of the former Progressive and Radical Liberal voters. It should be observed that because of the bias in the survey in favour of the anti-EC alliance and against the Bourgeois alliance (cf. Table 2), the estimates in the first two columns are likely to be too high and those in the fifth column too low. Yet there can be no doubt that clear stands for or against the EC were rewarded handsomely by the voters at this election.

The role of the People's Movement appears to be that it provided a home - perhaps of a temporary kind - for EC opponents who would not normally vote for either a left-wing party or the Single Tax Party. Further analysis shows thcit the strength of the People's Movement in many subgroups is very nearly half the proportion of EC opponents in the corresponding subgroup. For example, the People's Movement received 39 per cent of the vote, of those aged 18-29, 22 per cent of those aged 30-54, and 17 per cent of those aged 55 or more. This parallels the proportion of these same age groups who were against Denmark's continued

membership in EC, these proportions being 70 per cent of those aged 18-29,45 per cent of those aged 30-54, and 38 per cent of those aged 55 or more. With respect to occupational group the opposition against the EC was strongest among blue-collar workers and students and weakest among self-employed, in particular farmers. This pattern is mirrored in the support for the People's Movement which varies from 37 per cent among blue-collar workers to zero among the 134 farmers in our sample.

Opposition against the EC within the various parties also goes far toward explaining the differential success of the People's Movement in gaining votes from these parties, as witnessed by Table 4. Of the 1977 voters for the Communist party or the other three parties that went into an election alliance with the People's Movement, no less than 96 per cent were against Denmark's continued membership of the EC. Of the 1977 Social Democrats 50 per cent were against EC membership, of the Progressive voters 33 per cent, and of the Radical Liberal voters 31 per cent. From each of these parties the People's Movement received a share that is just about half the number of EC opponents. Only the parties forming the Bourgeois election alliance succeeded in preventing a larger proportion of their EC opponents from defecting to the People's Movement. These four parties in combination had 11 per cent who were opposed to continued membership, yet had only a loss of 3-4 per cent to the People's Movement. Such minor irregularities detract very little from the general pattern in which opposition to the EC can be seen as an intervening variable between social background and the 1977 vote on one hand, and support for the People's Movement on the other.

5. Election or Referendum?

As has been shown this first direct election to the European Parliament in Denmark differed in several ways from the direct elections in the other member countries. Turnout was extraordinarily low in Denmark, and in addition there was a list - which received more than 20 per cent of the vote - with Danish withdrawal from the EC as its sole issue.

In general, the analysis of aggregate voting statistics shows a closer relationship between the result of this European Election and the 1972 referendum on Danish membership in the EC than between the former and the last Danish general election in 1977. We found a neat correspondence between the attitude of the individual voters on Danish membership and their vote in this direct election. From this we may conclude that the Danish electors saw this as a traditional election to only a limited degree. In a parliamentary election party candidates are selected on the basis of a set of political view-points, covering stands on a number of issues. It looks much more as if the voters perceived the EC-election as some kind of a new referendum on Denmark's membership of the Community.

This situation will have a number of important consequences for Danish politics. One is that specific policies of the EC may not continue to be part of the political debate in Denmark, and therefore the poor performance of the Social Democratic Party will probably not be reflected either in the public opinion polls or in the next general election. This conclusion depends of course on how the party leadership solves the problems connected with the evident discrepancy between their own positive stand on the EC and the rejection of continued membership found among at least half the rank and file of the party. Finally, it can be argued that the direct

elections and the success of the People's Movement have giv. n strength and momentum to the Danish opponents of the EC. Thus, Danish polics towards the EC will still have the background of a certain discrepancy between popular and parliamentary support of continuous membership.

---

NOTE

1. The slope of the linear regression line is .35 and r2r² is only .03. The regression coefficient is significantly different from 0 (p = .05), as are all other regression coefficients referred to in this article. In most cases p is considerably lower, i.e. p < .001.

Scandinavian Political Studies, Vol. 2 - New series - No.

3 1979

ISSN 0080-6757

© Nordic Political Science Association