Danish politics during the 1960s was a close race between socialist parties (Social Democrats and Socialist People's party) and non-socialist parties (Conservatives, Agrarian Liberals, and Radical Liberals). A numerical balance between these two sides was established in 1960, when the Socialist People's party entered the Folketing as the first major party to emerge to the left of the Social Democrats since the decline of the Communist party from 1947. But it was not until 1966 that the Social Democrats responded to this challenge by abandoning their traditional alliance with the Radical Liberal center party to rely on support from the new leftist movement in what was dubbed the 'red cabinet' by the bourgeois opposition, now joined by the Radical Liberals. This bourgeois three-party coalition, led by the Radical Liberal H. Baunsgaard, came into power in January 1968, when a splinter group breaking away from the left wing of the Socialist People's party helped to bring down the Social Democratic government. Except for in this election, which gave a 54-40 majority to the bourgeois-liberal side, the two sides were never more than a few percentage points removed from one another in the vote division or the parliamentary division of seats. The close race became a stalemate at the election in September 1971, from which emerged a Social Democratic government pinning its majority on support from candidates from Greenland and the Faroes, besides the Socialist People's party. Amazingly parallel developments took place in Norway and Sweden, where elections in 1973 resulted in ties in their Parliaments.

In Denmark the election period from September 1971 to December 1973 was dominated by two major issues. One was the heated discussion about Denmark's entry into the European Community, leading to the referendum in October 1972.¹ The other was the discussion about increasing taxes and governmental expenditures. What was most salient to the non-farm population during the early months of 1973 was a squeeze on their real disposable incomes due both to rising prices and to an expanding public sector. Simultaneously, the Progress party appeared, led by tax lawyer M. Glistrup, whose incomparable remarks attacking the income tax, the bureaucracy, and the incapacities of the old parties were well publicized by the mass media. Within a few months, the Progress party rose to the 25 percent level in the polls. During the summer and autumn months there were signs of decline, but during the election campaign the movement regained its momentum.² The Social Democratic prime minister, Anker Jørgensen, to whom J. O. Krag had handed over the reins of the government on the day after the successful EEC referendum, intrepidly declared to a Social Democratic congress that his party was placed 'to the left of the center'. But throughout October and early November 1973 it was evident that the shaky socialist majority might disappear at any moment. The election was triggered when Erhard Jakobsen, a prominent Social Democrat known to be opposed to his party's dependency on the Socialist People's party and to be particularly concerned with the increasing tax on one-family houses, failed to appear at a legislative roll call. The next day, the Center Democrats were formed, with Jakobsen as the leading figure. Polls taken both before and after this occasion forecasted success for the new party at the cost particularly of the Social Democrats and the Progress party. In addition., large numbers of voters appeared to be undecided; but as election day approached this indecision declined, and three minor parties emerged with good chances of surpassing the 2 percent threshold for representation. Of these, the Christian People's party campaigned for the first time in 1971 against

secular trends on such issues as abortion, pornography, religious education, and sex education. The Single-Tax party and the Communist party had not been in the Folketing since 1960, although they put up candidates at every general election. Although differing widely in other respects, their 1973 campaigns were dominated by attacks on Denmark's membership in the EEC.

The spectacular result of the 1973 election is presented in Table I and compared with some earlier elections. Both the traditional party blocs went down from the 46 percent level to the 32 percent level of voting strength. The remaining parties increased their combined voting strength, as shown in the bottom line, from 7 to almost 36 percent. The two parties that were formed within the last year, the Progress party and the Center Democrats, together received almost one quarter of the votes. Of the old parties in the bloc system, the Social Democrats suffered the most in terms of absolute number of votes, and the Conservatives suffered the most relative to its previous strength.

These are partisan changes of an unusual magnitude, and we shall search for common denominators in the explanation for them. The election result stirred political observers internationally. Most interpretations ran in terms of voter protest and resentment of the Danish welfare state, while some spoke of a right swing as a reaction against the socialist build-up of governmental power and interference. We shall be looking mainly for psychological indicators that may express such notions more accurately, and relate them to indicators of participation and partisan preferences. The data come from a postelection survey of 535 respondents who had been interviewed twice previously, just before

and after the election in September 1971.3 Although, naturally, the limited size of this sample prohibits a detailed account of switching among the several parties, population breakdowns, etc., it is hoped that it will be sufficient for a discussion of broad changes in the Danish electorate.

1. Alienation from the System of Representation

The political debate in Denmark throughout 1973 centered on the notion of a general 'confidence gap' between political leaders and the public. The EEC debate of 1972 already carried overtones of a cleavage beyond the accustomed partisan cleavages - one separating establishment and non-establishment people, who were identified as supporters and opponents of the membership, respectively. This cleavage broadened during 1973 into a crisis of political leadership. The Ruling Class, published in the autumn by welfare economist Jørgen S. Dich, pointed out how the educated sectors had expanded the welfare system far beyond what was expedient for society.⁴ Thus, professionals and bureaucrats as well as politicians came under fire for having exploited and alienated the general public.

Alienation has several meanings, but certainly two of its important dimensions that have been investigated in a political context are political distrust or cynicism, on the one hand, and perceived political powerlessness or lack of personal efficacy, on the other. Time series data for the United States show a dramatic increase in measures for both of these attitudes and indicate their relation to such issues as the Vietnam war, racial confrontations, and urban problems.⁵ Our questionnaire included, both in 1971 and 1973, some of these alienation items that are relevant to these general orientations. The data appear in Table 11.

It is evident that items most directly focusing on the output side of the political system show increasing dissatisfaction from 1971 to 1973, despite a high level already in 1971. Admittedly, questions of this type tend to be leading; the frequency of agreement is therefore presumably far less reliable than the change. Still, to find that, by 1973, 91 percent are against the governmental budget leaves the impression that there is indeed 'something rotten in the state of Denmark'.

The two items reflecting personal political efficacy suggest that much less change took place in attitudes connected with the input side of the political system. Some may have thought the number of competing parties bewildering, but others probably found that the 1973 campaign simplified politics greatly and that reassured the voter of his power. The complaint that the politicians care too little about voter opinions may be taken to relate both to the input side and the output side of politics. The moderate increase in agreement with this statement may be due to the experience by many voters in the campaign that the politicians indeed seemed to care quite a lot.

The question on party identification was included in this table because of its presumed ability to measure long-range support for the existing parties. As is well known, research of the U.S. electorate has demonstrated that both the level and partisan division of such psychological party ties can remain undisturbed by short-run defections in the actual voting pattern.⁶ In Denmark, losses in the socialist bloc in 1968 may well have been of a similar transient character, since the polls showed a recovery of the defeated parties beginning already a few months after the election. An exodus from the established parties on the scale that was seen in 1973, however, can hardly be expected to occur unless party identifications also suffer badly. The actual decrease from 57 to 48 percent of the sample who consented to be termed 'adherents' of one party or another seems surprisingly small.⁷ In the United States, however, a similar change has not been observed even with short-term forces being very strong. Moreover, the 48 percent is probably an overestimate because only eligible voters from 1971 were among our respondents. The three youngest age cohorts of 1973 voters were excluded, two because of the time elapsed and the third because a lowering of the voting age from 21 to 20 years was passed simultaneously with the EEC referendum.

Finally, the two last items in Table 11, repeated from the pre-election survey in 1971, are intended to suggest some authoritarian responses to the 'confidence gap'. Both of these items point to an inclination among the voters to restrict the channels of political influence. Both items, and especially the 'strong man' item, show fairly positive correlations with the political cynicism items. The increase in agreement with these statements naturally has implications for the wider democratic norms.

2. Participation and Vote Decision

The voting turnout at the 1973 election was 89 percent, slightly lower than at the EEC referendum one year earlier (90 percent) but slightly higher than at the 1971 election (87 percent). Our sample overrates these figures, as is often the case in voter surveys. But several indicators of involvement and campaign activity, presented in Table 111, show marked increases from 1971. And, as anticipated, the high level of activity was coupled with widespread indecision about the voting choice, which is also seen in this table.

A slight increase in political involvement might be expected, other things being equal, because of dropout of indifferent respondents on the panel. But the amount of increase suggests that other things were not equal; and nobody complained that the brief 1973 election campaign had been dull, unlike the previous campaign. The figures in Table 111 fit partly with those in Table 11. The combination of widespread dissatisfaction with political leadership and sustained feelings of political efficacy indicates a mood of active

protest in the electorate, and the decline of psychological attachment to a party leads
one to expect that large numbers of voters were undecided and searching for new alternatives.

Still, one might reason that the decline of party identification showed a failure of the parties to mobilize interest and participation in the election, and hence one might expect signs of apathy among the voters. This would be in line with the typical findings in voting research put forward in the early community studies.⁸ However, in a multiparty system with good opportunities for introducing new parties and with an attentive TV and press, there may be short-term forces with much stronger mobilizing effect than traditional party allegiance. The TV audience, for example, had occasion to view the tied roll call, the calling of the election by the prime minister and, minutes later, the arrival of Erhard Jakobsen, explaining that his automobile had run out of gas. On the whole, one might say that the surge of the Progress party and the Center Democrats in the course of a few months would not have been possible without active reporting in the news media.

On the whole, therefore, the pattern of the 1973 election up to a point follows the mechanism of 'surge' election, as described by Angus Campbell.9 The increase in general stimulation chiefly affects voters with weak or no party identification and causes large partisan changes. In this case, however, the short-term forces were directed against the established parties as a whole, and the response was to favor new parties and hitherto neglected parties.

3. Party Switching and Preferences

Where did the new or formerly unrepresented parties recruit their voters? Our sample of 535 respondents is naturally inconclusive concerning a detailed investigation of this question. Furthermore, the exclusion of young voters and the under-representation of non-voters dictate that conclusions from these data be tentative and limited.

The party switching of our respondents is presented in Table IV. The figures given for switching between individual parties are absolute numbers of respondents. Switching between blocs or groups of parties is given as the percent of those respondents who reported their partisan vote for both elections.

These percentages suggest two conclusions: that the previously unrepresented parties
drew support roughly equally from the established party blocs and that very few voters
switched between the socialist and the bourgeois blocs.

From the absolute figures it appears fairly certain that the Progress party recruited substantial numbers of voters from each of the three bourgeois parties as well as from the Social Democrats. The Center Democrats drew support especially from the Social Democrats, and it is likely that the Communist party recruited mostly from the socialist bloc, whereas the Christian People's party recruited mostly from the bourgeois bloc.

Corresponding tendencies were found in a survey comprising some 800 respondents in the town of Aarhus a few days before the election in December 1973.¹⁰ One might observe that they are not consistent with the standard version of the 'right swing'. That is, the gaining parties do not fit particularly well into an ideological left-right pattern in such a way that the electoral shift can be described as a neat 'chain reaction' toward the right-hand side of the spectrum. With respect to ideological attitude measures carried over from the 1971 study, some indeed exhibit an overall increase in ideological conservatism during the 2-year span, while others have moved in the opposite direction. To cite examples, agreement with the statement 'High incomes should be taxed more strongly than they are today' rose from 51 to 68 percent; agreement with the statement 'The government has too little control over private investments' rose from 51 to 60 percent; and agreement with the statement 'In politics one ought to strive to give everybody the same economic conditions, no matter what their education or occupation is' remained at 53 percent in both August 1971 and December 1973.

Rather than following the conventional ideological dimension, the 1973 election introduced a protest or distrust dimension. The established parties were blamed especially for economic mismanagement during past election periods, during which all of the five parties had, at one time or another, been involved in governing the country. The movement led by Mr. Glistrup was the first to seize this dimension and the most successful in channeling the feelings of dissatisfaction into mass voting behavior. On the four distrust items and the first of the two authoritarian-like items appearing in Table II above, Progress party voters scored substantially higher than other voters.¹¹ When extending the analysis from voting for the Progress party to relative preference of the Progress party by means of rank-order preference data, a high ranking of the Progress party among our total set of respondents is clearly positively correlated with indicators of distrust but only vaguely related to indicators of conservative ideology or middle-class identification.12

By means of more refined analyses, we may eventually hope to clarify which issues caused negative psychological orientations toward politics and established parties to develop. However, once the Progress party demonstrated the saliency of this distrust or protest dimension for voting behavior, the road was open for partisan instability on a broader scale, and four more parties were pulled into the Folketing. When the electoral laws of a political system facilitate a representation of new parties, an outcome of this sort seems the plausible response from a dissatisfied electorate receiving strong stimulus from the mass media.

University of Aarhus

Ole Borre

---

NOTES

1. See Nikolaj Petersen and Jørgen Elklit, 'Denmark Enters the European Communities', pp. 198-213 in Scandinavian Political Studies, Vol. 8/73. Oslo, Universitetsforlaget, 1973.

2. According to the monthly Gallup polls, published in Berlingske Tidende, the following percent of the voters would vote for the Progress party: February 1973, 5.8 percent; March, 21.2 percent; April, 26.0 percent; May, 17.4 percent; June, 16.6 percent; August, 13.1 percent; September, 11.1 percent; October, 13.3 percent; and November, 9.4 percent. According to the Observa panel polls, published in Jyllands-Posten, the figures were as follows: January 31, 4 percent; March 2, 7 percent; April 2, 25 percent; May 2, 22 percent; June 14, 18 percent; August 2, 17 percent; August 30, 16 percent. These time series are summarized in Information, October 8, 1973, up to that date.

---

2. According to the monthly Gallup polls, published in Berlingske Tidende, the following percent of the voters would vote for the Progress party: February 1973, 5.8 percent; March, 21.2 percent; April, 26.0 percent; May, 17.4 percent; June, 16.6 percent; August, 13.1 percent; September, 11.1 percent; October, 13.3 percent; and November, 9.4 percent. According to the Observa panel polls, published in Jyllands-Posten, the figures were as follows: January 31, 4 percent; March 2, 7 percent; April 2, 25 percent; May 2, 22 percent; June 14, 18 percent; August 2, 17 percent; August 30, 16 percent. These time series are summarized in Information, October 8, 1973, up to that date.

3. Papers reporting on this survey include those by Ole Borre and Daniel Katz, 'Party Identification and Its Motivational Base in a Multiparty System: A Study of the Danish General Election of 1971', pp. 69-112 in Scandinavian Political Studies, Vol. 8/73. Oslo, Universitetsforlaget, 1973; and Ole Borre, 'Party and Ideology in Denmark' and Structural Effects in Danish Voting Behavior', Institute of Political Science, University of Aarhus, 1973 (mimeograph).

4. Jørgen S. Dich. Den herskende klasse (The Ruling Class). Copenhagen, Borgens forlag, 1973.

5. Arthur H. Miller. 'Political Issues and Trust in Government: 1964—1970', paper delivered at the 1972 Annual Meeting of the American Political Science Association.

6. See, e.g., Philip E. Converse, 'The Concept of a Normal Vote', pp. 9-39 in Angus Campbell, Philip E. Converse, Warren E. Miller, and Donald E. Stokes, Election and the Political Order, New York, John Wiley and Sons, 1966.

7. Judging from surveys conducted in connection with the EEC referendum, party identification appears to have declined already during the first half of the election period 1971— 73. A survey in May—June 1972 showed 50 percent identifying with a party (N = 905). A survey in October 1972, just after the referendum, included respondents who had been interviewed in May—June 1972 and respondents who had been interviewed in October 1971. Put together, the October 1972 survey showed 49 percent identifying with a party (N = 1,275). These data were presented to me by Jørgen Elklit.

8. See, e.g., Paul F. Lazarsfeld, Bernard Berelson, and Hazel Gaudet, The People's Choice, 3rd ed. New York, Columbia University Press, 1968, p. 68; Bernard R. Berelson, Paul F. Lazarsfeld, and William N. McPhee, Voting. Chicago, The University of Chicago Press, 1954, p. 26.

9. Angus Campbell. 'Surge and Decline: A Study of Electoral Change', Public Opinion Quarterly, 24, Fall 1960 (chap. 3 of Campbell et al., op. cit.).

10. Ole Borre, Jørgen Elklit, and Ole Tonsgaard, 'Danmarks Radios Vælgerundersøgelse 1973' (The 1973 voter survey by the Danish Radio). Danmarks Radio og Institut for Statskundskab, Aarhus Universitet, December 4, 1973 (mimeograph).

11. Among Progress party voters, the proportion giving responses that indicate distrust on the four items in Table II was as follows: 97 percent agreed that 'the politicians are too generous with the taxpayers' money'; 88 percent agreed that 'the politicians generally care too little about what the voters think'; 72 percent agreed that 'people who want to get to the top in politics have to give up most of their principles'; and 54 percent disagreed that 'in general one may trust our political leaders to make the right decisions for the country'; furthermore, 82 percent agreed that 'it would be sensible to have a strong man seize power in a situation of economic crisis'; Ns = 65 respondents.

12. A study of the dimensions of the Danish party system in the period 1971-73 is under preparation by Jerrold G. Rusk and Ole Borre for European Journal of Political Science.