The Maastricht Treaty, negotiated between EC ministers in December 1991, has quite unintendedly caused a virtual experiment in direct democracy. Not only was the ratification of the Treaty sent to referendums in three of the twelve EC countries (Denmark, France and Ireland), but the Danes even had to vote twice on this issue within a single year. First, the Maastricht Treaty was rejected in Denmark by a narrow margin in the referendum on 2 June 1992, then the Treaty and the Edinburgh Agreement were accepted in the referendum on 18 May 1993. This development raises a number of specific questions about how and why the Danish voters within a year could act so differently. Do the Danes meet the worst Shakespearian expectations about internal doubts and indecisiveness or are they characterized by sensible stability in their attitudes towards European integration? Are the Danes basically anarchistic and unpredictable in their political behaviour or are they reasonably responsible and predictable?

However, the experiment in direct democracy also raises more general questions. Do the apparently contradictory results prove the inability and inadequacy of popular participation in the decision-making process? Do the voters react on irrelevant stimuli or can they really decide an issue rationally? And is it possible for political elites by deliberate means to manipulate the voters?

This article tries to find some preliminary answers to these specific and general questions, but the intention is, in addition, to describe the background for the: Danish May referendum, the campaign leading up to this referendum, and the main characteristics of the result. In the conclusion some of the democratic perspectives that might be drawn from this experiment in direct democracy are discussed.

From Maastricht to Edinburgh

The background for the May 1993 referendum in Denmark comprises two
essential elements: the result of the first referendum in June 1992, and the
events that followed.

The result of the first referendum in June 1992 was a narrow rejection of the Maastricht Treaty (Siune 1993). A solid majority in the Danish Parliament had been in favour of a Danish ratification of the Treaty - 130 voting for and 25 voting against the bill on 12 May 1992. Despite this fact the voters rejected the Treaty at the referendum, 50.7 percent of the votes voted No and 49.3 percent voted Yes. Only 46847 votes separated the two alternatives. However, the two constitutional requirements for a rejection were fulfilled: there were more No votes than Yes votes, and the No votes comprised more than 30 percent of all eligible voters: 41.7 percent, the turnout at the referendum was as high as 83.1 percent.

The immediate reaction of the political elite was one of shock and confusion. Members of the Government could hardly believe that they had been overruled by the people, and during the days following the referendum is was widely discussed how to interpret the result. The opposition to the Maastricht Treaty comprising The Socialist People's Party, The Progress Party and parts of The Christian People's Party together with more or less well-organized popular movements had stressed different arguments against the Treaty. It was quite evident that the Government and the main political parties had lost the referendum, but it was far from evident who had won it - and for what reasons.

As outlined by Nikolaj Petersen, three main alternatives presented themselves to this difficult situation (Petersen 1992, 1993). However, each of these solutions was beset by serious drawbacks as it would be impossible to satisfy the other EC governments, the political establishment in Denmark and the Danish public opinion all at the same time.

One alternative could be termed "Maastricht with Roses", as this solution would involve some kind of solemn declaration interpreting and emphasizing the reservations Denmark had already secured in the Treaty. This would add some new formulations about more democracy, more openness and a more precise application of the principle of subsidiarity. The EC partners as well as the Danish Government would prefer this model, whereas it would be in conflict with the referendum result, as it would - in reality - involve a second referendum on the same Treaty. "Such a procedure would be democratically suspect and involve a tangible risk of another No, which in turn might very well result in Denmark having to leave the Community altogether" (Petersen 1993, 12).

A second alternative could be termed "Maastricht without Thorns", as
this solution would involve a Danish acceptance of the Maastricht Treaty,
but without some of the obligations that caused popular resistance. Denmark

would be part of the union, but on limited conditions. "The main problem with this model was, of course, whether the other EC members would accept a derogation from the Treaty to accommodate Denmark despite the fears of re-negotiation, of a 'Europe å la carte' or a 'Europe at different speeds'" (Petersen 1993, 12).

A third alternative could be termed "The Scent of Maastricht", as this would involve Danish non-membership of the union, but only some kind of association. Whereas Danish participation in most of the practical aspects would be a perfect solution for the majority of Danish voters and for other EC governments that were reluctant to admit Denmark a privileged position, "this solution would be utterly unsatisfactory for the government and most of the political establishment in Denmark because it would relegate the country to a clearly secondary role in European politics" (Petersen 1993, 12).

The actual solution found to the dilemma caused by the June referendum was the so-called "National Compromise" including seven of the eight parties in the Folketing, the Progress Party was the only outsider. The decision-making process of this compromise was an extreme manifestation of the peculiar arrangement of Danish minority parliamentarism (cf. Damgaard & Svensson 1989).

In October 1992 the Socialist People's Party invited the Social Democrats and the Radical Liberals to discuss the possibility of a "National Compromise" on the Maastricht Treaty (see further Petersen 1993, 13-19). The Government had not only lost the initiative, but actually expressed the view that it was most appropriate for finding an overall solution that the Government "was put outside the door". Within a few days the three opposition parties agreed on a document which the Government more or less had to accept. The document, entitled Denmark in Europe was subsequently presented by the Danish Foreign Minister, Uffe Ellemann to the individual EC governments, while the Social Democratic leader, Poul Nyrup Rasmussen, went on his own private diplomatic mission to the Socialist and Social Democratic Parties in the Community.

The document contained three substantial parts, one on issues of common interest, one on the particular interest of an agreement for Denmark and one on the Danish position in discussions of enlargement of the Community. The main content of the "National Compromise" concerned the second part and involved a demand for four exemptions from full Danish participation in the European Union: (1) Denmark does not participate in the so-called defence policy dimension involving membership of the Western European Union and a common defence policy or a common defence; (2) Denmark does not participate in the single currency and the economic policy obligations linked to the third stage of the EMU; (3) Denmark is not committed

in relation to union citizenship; and (4) Denmark cannot accept the transfer
of sovereignty in the area of justice and police affairs.

As a consequence of these four exceptions it was emphasized that "the objectives of the Union set up in the common provisions of the Maastricht Treaty will not apply to Denmark on the above points". The last major demand was that the Danish agreement should be legally binding to all twelve EC Member states for an unlimited period.

The exact implications of these formulations were not exactly clear. For instance, according to the Maastricht Treaty Denmark was not obliged to participate in the third phase of the EMU. However, whereas such a participation according to the Treaty was conditioned by a future referendum, it was now stated that Denmark would definitely not participate. Furthermore, concerning defence policy as well as justice and police policies, the Treaty only contained broad formulations involving no specific obligations. However, instead of postponing decisions about Danish participation in these matters, it was immediately refused by the National Compromise. Concerning union citizenship the formulation mainly involved a symbolic rejection of the idea, as Denmark either had given or indicated a firm willingness to give EC residents the rights involved.

The European Council at its meeting in Edinburgh on 11-12 December reached an agreement which to a large extent sanctioned the demands made in the document on Denmark in Europe. The issues of common interest put forward by the Danish Government concerning subsidiarity, openness in the decision-making process and the enlargement of the Community were covered in the Presidency's main conclusions, and the solution to the specific Danish demands was dealt with in a "decision" by the Heads of State and Government and in declarations of the European Council and the Danish Government.

The results of the Edinburgh summit were presented by the media in Denmark as a clear victory for the Danish points of view. The outcome was accepted not only by the Social Democrats and the Radical Liberals but also by the leaders of the Socialist People's Party. Only the Progress Party maintained its opposition. The immediate reaction of public opinion was favourable, 54 percent stated that they would vote Yes at a coming referendum, while 24 percent would vote No (Gallup poll, Berlingske Tidende, 17 December 1992).

The Campaign

After the Danish Folketing had adopted the Edinburgh Agreement, a
rather short and dull campaign followed up to the 18 May referendum. All
the main arguments for and against the Treaty had been presented previously

and the new elements primarily concerned three issues: (1) Did the Edinburgh Agreement actually change anything or were the Danes to vote on the same issues once again? (2) Would a new Danish No involve the definite defeat of the Maastricht Treaty or would it lead to Denmark's withdrawal from the Community? (3) To what extent did the "National Compromise" and the Edinburgh Agreement meet the reasons for opposition and the doubts of many Danes, in particular of a large number of Social Democrats and most of the Socialist People's Party's voters?

The main issue was whether the Edinburgh Agreement really contained
anything new. Was it - to apply Nikolaj Petersens terminology - a "Maastricht
with Roses' or rather a "Maastricht without Thorns"?

The supporters argued that it was more of the latter than of the former. The supporters - including the whole political establishment, all political parties represented in the Folketing with the exception of one, almost all major interest organizations, and almost all newspapers - claimed that Denmark had obtained important exceptions, not included in the Maastricht agreement, that Denmark had not only stated its reservations, but that the other EC governments had accepted these reservations, and that Denmark would not be legally obliged to proceed into any further integration. The Government and the Folketing had, in fact, been very responsive to the public on the EC issue, and all future changes in Danish EC policy would be submitted to referendums. In short, the Danish public had no cause for fear about the further EC development.

The opponents - including the Progress Party and a few spokesmen of the Socialist People's Party, but mainly comprising the so-called "June Movement", the leading anti-Union grassroots organization - argued that the Edinburgh Agreement did not really represent anything new, that it was not legally binding (if Denmark was brought before the European Court of Justice), and that the Danish exceptions would only be of limited duration, as new initiatives for further integration would soon come up and place Denmark in a new situation of "take-it-or-leave-it". In short, the political establishment tried to manipulate the voters to support a European Union which was only demanded by a few and disliked by the great majority.

The supporters further argued that this was Denmark's last chance. The other EC governments had been very generous to Denmark, and if the Danish voters declined this opportunity to join the European Union at minimum terms, Denmark would put its EC membership at stake, as ten or eleven EC countries would proceed on further integration without Denmark.

Against this line of argument the opponents claimed that a new Danish
No would not lead to any forced withdrawal from the EC, but rather to
the definitive collapse of the Maastricht Treaty. This argument was based

on the statements from the British Government that another Danish rejection
of the Treaty would imply that the Treaty would neither be introduced
in nor passed by the House of Commons.

As this debate and the divergent interpretations of the Edinburgh Agreement clearly reflected more basic attitudes towards further EC integration, the development in public opinion was scrutinized very closely. As indicated above, and as shown in Figure 1, it seems beyond doubt that the "National Compromise" and the Edinburgh Agreement shifted the public opinion in Denmark significantly. Although a growing opposition and a slight decline in the support could be detected - at least until the middle of April - the lead of the supporters was never in doubt.

The Result of the Referendum

The outcome of the referendum on 18 May confirmed the expectations raised by the opinion polls. Of the votes cast 56.7 percent were in favour of the Maastricht Treaty and the Edinburgh Agreement, and 43.3 percent were against this manner of Danish membership of the European Union. In fact, the net change from June 1992 to May 1993 was no more than 7.4 percent and perhaps even smaller when the turnout and the new voters are taken into account.

As shown in Table 1, the pattern of the May 1993 referendum was quite similar to the pattern of previous EC referendums. The strongest support for further EC integration was found in Jutland and the strongest opposition in the metropolitan area (Copenhagen and Frederiksberg).

The turnout in the May 1993 referendum was higher than that in the June 1992 referendum. Whereas 83.1 percent of the voters participated in June 1992, the turnout increased to 86.5 percent in May 1993. Although the turnout did not reach the 90.4 percent record of 1972 it nevertheless gave the result a high degree of democratic legitimacy. As shown in Table 1, the turnout pattern in 1993 was also similar to that of previous EC referendums. The turnout was higher on the islands and lower in the metropolitan area.

A closer ecological analysis of the referendum result on the basis of the votes cast in the 103 constituencies confirms a strong stability in the regional pattern of voting on the EC. If the No percentages in 1972 and 1993 are compared as in Figure 2, it is evident that constituencies voting No to Danish EC membership in 1972 are also most negative towards the Maastricht Treaty as supplemented by the Edinburgh Agreement in 1993, the regression coefficient being .66 and the explained variance 80 percent. This is quite similar to the relationship between the voting pattern in 1972 and 1992, albeit a little weaker than the relations between 1972 and 1986 (cf. Siune et al. 1992, 25-26).

In sum, the pattern of turnout and the structure of voting in the Danish

EC referendums show strong elements of stability which do not indicate the prevalence of shifting attitudes towards the European Community in the Danish public. In order to investigate this point further it is necessary to turn to data at the individual level.¹

Trends in Voting Behaviour

The voting behaviour of the main social and political groups in May 1993 as compared with that in June 1992 is presented in Table 2.² The overall impression is one of a general move towards lower opposition against the European Union in all social and political groups. Some interesting differences can be observed, however.

Women seem to lean somewhat more strongly towards Yes than men, which means that the small gender difference in 1992 tends to become even smaller. Somewhat surprising perhaps is that the young voters are the most stable in their referendum voting. One might think that the young would be the most unstable, tending to follow the general movement among the voters more strongly than the elderly. This was, however, by no means the case. The strongest movement from No to Yes appeared among those above 50 years of age, perhaps because they were most worried about the possible consequences of a second Danish No.

It is also of some interest to observe that the more educated changed their mind to the same degree as the less educated. If the elite had been able to manipulate the voters to change their mind, the expectation would have been that the less educated would have been easier to manipulate than the better educated. Thus, this pattern does not support the case of elite manipulation. The privately employed with a higher education - one of the groups most in favour of the Maastricht Treaty in 1992 - in particular turned strongly from No to Yes, indicating an increased enthusiasm rather than manipulation.

The resistance towards the European Union is still strongest among the workers and the publicly employed, but even among these groups the opposition weakened from 1992 to 1993. In particular, the unskilled workers changed, surprisingly, from No to Yes. As the unskilled workers comprise the opposition group with least resources, this marked change can be taken as an effect of elite manipulation.

A most interesting observation can be made with regard to the groups that are more or less interested in politics. Whereas in June 1992 the opposition towards the Maastricht Treaty increased with declining interest in politics, this relationship disappeared in 1993. The voters with the smallest interest in politics presumably changed their minds most. The absolute number in this group is not large (N = 22), but it seems reasonable

to assume that this group has been the one most receptive to the argument of the political elite that the Danes had been given a new chance and that it would have horrendous consequences to say No a second time. If a case could be made for elite manipulation of the voters this seems to be the most obvious one.

On the other hand, the opposite story could be told as far as the voters for individual parties are concerned. Although the leaders of the Socialist People's Party changed over from No to Yes they were followed by only a few of their voters. Four out of five voting for the Socialist People's Party in the 1990 Folketing election did not follow the advice of their political leaders, but voted No once again. Somewhat more successful were the Social Democratic leaders. However, the changes with the "National Compromise" and the Edinburgh Agreement only convinced Social Democratic voters to the extent that whereas two out of three voted No in June 1992, half of them did so in May 1993.

Among the bourgeois voters two groups increased their No vote, but presumably for quite different reasons. The Conservative voters increased their (small) opposition to the European Union, undoubtedly because the Danish membership had been weakened and because they did not approve of the negative self-commitments imposed on the Government by the centre-left opposition. The voters for the Progress Party, on the other hand, increasingly followed their party leaders in the continued resistance against the Maastricht Treaty.

In sum, the Danish voters behaved quite independently of their party in the EC referendums. This point is most clearly demonstrated by the Socialist People's Party, but also by the Social Democrats. Half of the Social Democrat voters did not follow the "party line" despite the fact that their leaders claimed that the basis for the referendum had been changed in order to comply with the views of - in particular - the Social Democrat voters.

If many voters did not follow the line chosen by their party leaders it might be assumed that they acted on other grounds. At this point the individual attitudes of the voters towards EC integration naturally come into mind. Perhaps the Danes are not - after all - characterized by internal doubts and indecisive ness, but rather by fairly stable attitudes towards European integration?

In Table 3 the main ideological attitudes towards EC integration have been identified by the voters' views on the areas to be included in ECcooperation. The table reveals two important points. First, there is an unambiguous pattern in the orientation towards EC integration among Danish voters. Whereas a majority favours economic integration, the support clearly declines the more political the integration becomes. Second, there is a remarkable stability over time in these orientations. The rank order of areas preferred to be included in EC cooperation is basically the

same from May 1992 to May 1993, that is, from before the first referendum to after the second referendum. The main differences are that the single market has moved to the top position and that support for the social dimension has increased. Undoubtedly, these changes were mainly a result of the campaign up to the first referendum.

As shown in Figure 3 there is a close relationship between the voters' attitudes towards EC integration and their behaviour at the referendums. The larger the number of areas preferred to be included in the EC, the lower the No percentages. The curve for 1993 is a little below the curve for 1992, indicating that the Edinburgh Agreement appeased a number of sceptics voting No at the first referendum.

Conclusions

It has been shown in this article that the Danish voters within a year did not act very differently on the issue of the European Union, although the outcomes were different. In fact, the net change was only 7 percent and the pattern of turnout and the structure of voting in the Danish EC referendums had strong elements of stability. The Danish voters behaved quite independently of their party, they had stable orientations towards EC integration, and they behaved rationally on the basis of these orientations

taking the changes involved in the Edinburgh Agreement into account. It would be wrong, consequently, to claim that the Danes are characterized by internal doubts and indecisiveness in their attitudes towards European integration. The Danes seem to be quite responsible and predictable in their attitude and voting behaviour concerning the EC rather than anarchistic and unpredictable.

It is still a matter of discussion and interpretation based on different values and orientations whether and to what extent it was possible for political elites to manipulate the voters. The answer to this question hinges on the answer given to the question whether the Edinburgh agreement really represented anything new? If the political aspects are stressed in relation to the formal and judicial aspects, a strong argument can be put forward to the effect that the Danish voters did not have to vote on exactly the same thing once again. If this argument is accepted, it is difficult to claim, however, that the voters were manipulated by the elite. In addition, the analysis of the groups changing most strongly from No to Yes did not unambiguously support the manipulation hypothesis.

It is also a matter of discussion and interpretation based on different values whether and to what extent the apparently contradictory results of the Danish referendums in 1992 and 1993 proved the inability and inadequacy of popular participation in the decision-making process. However, if a vital element in the democratic process is - in Robert A. Dahl's terminology - "effective participation" that throughout the process of making binding decisions, citizens ought to have an adequate and equal opportunity for expressing their preferences about the final outcome (Dahl 1989, 109), and if another key characteristic of a democracy - or a polyarchy - is the continuing responsiveness of the government to the preferences of its citizens (ef. Dahl 1971, 1), then it could certainly be argued that the Danish ELC referendums demonstrate a democracy at work.

In any case, the two Danish EC referendums seem to have stimulated a new debate on the feasibility of a more direct democracy (cf. Højlund 1993; Schmidt 1993), indicating that a dialogue and even a conflict between the rulers and the ruled may contribute towards keeping the democratic process dynamic and viable.

---

NOTES

1. The data applied for the following analysis were collected by the commercial opinion research institute AIM on the basis of a questionnaire worked out by Karen Siune, Palle Svensson and Ole Tonsgaard and financed by The Research Fund at the University of Aarhus and the Social Science Research Council. The post-referendum questions were answered by 953 respondents. The data have been weighted with relation to sex in order to obtain a representative sample.

---

2. Respondents who did not report how they voted in the referendums have been excluded from this table. This means that Yes and No percentages sum up to 100.

REFERENCES

Dahl, R. A. 1971. Polyarchy. Participation and Opposition. New Haven and London: Yale
University Press.

Dahl, R. A. 1989. Democracy and Its Critics, New Haven: Yale University Press.

Damgaard, E. and Svensson, P. 1989. 'Who Governs? Parties and Policies in Denmark',
European Journal of Political Research 17, 731-745.

Højlund, N. 1993. Folkeafstemninger - en trussel mod demokratiet! Copenhagen: Gyldendal.

Petersen, N. 1992. 'Den rette forhandlingsstrategi for Danmark': Danmark og Den Europæiske
Union efter 2. juni. Copenhagen: Den Danske Europabevægelse, September 1992.

Petersen,N. 1993. 'Game, Set, and Match': Denmark and the European Union from Maastricht
to Edinburgh, rev. ed., September 1993. Institute of Political Science, University of Aarhus.

Schmidt, M. 1993. Direkte demokrati i Danmark. Copenhagen: Nyt Nordisk Forlag Arnold
Busck.

Siune, K. 1993. 'The Danes Said No to the Maastricht Treaty. The Danish EC Referendum
of June 1992', Scandinavian Political Studies 16, 93-103.

Siune, K., Svensson, P. and Tonsgaard, O. 1992. Det blev et nej. Århus: Politica. Statistiske
Efterretninger (1992, 1993).