Resumé

This article summarizes some of the main findings from the 1976 Swedish Election Study. The defeat of the Social Democrat government in 1976 was caused by three issues: bureaucracy, socialization, and nuclear power. Although net changes were small, the actual number of voters switching parties was the largest recorded during the last twenty years. The traditional image of the stable Swedish voter is becoming a myth. Aggregate stability is combined with a large and increasing individual volatility.

After 44 years in power the Social Democrat government was defeated. In this sense the Swedish election of September 19, 1976 was an historical event. Although net changes were small, the combined Social Democrat and Communist loss was sufficient to give the three bourgeois parties an absolute majority in the electorate. After two weeks of intense negotiations a coalition government of the Center Party, the People's Party (Liberals), and the Moderate Party (Conservatives) was formed with Thorbjörn Fälldin as Prime Minister.

It was the third successive election in which the Social Democrat vote declined. The party's popular support dwindled from 50.1 percent in the 1968 election to 42.7 percent in 1976. The setbacks of 1970 and 1973 could largely be explained by a worsening economic situation with increasing unemployment and prices. While evaluations of government performance on bread and butter issues were positive in 1968, they were clearly negative in 1973 (Särlvik 1977:97 ff). Nevertheless these kinds of economic issues were not especially important in 1976. The issue area in which the Social Democrats improved most in 1976 was indeed the one which had contributed most to their setback in 1973. Evaluations of employment, prices, and general standard of living - negative in 1973 - were positive

three years later (Table 1). As is the case for most findings reported in this
article, this conclusion is based on the 1976 Swedish Election Study
(Petersson 1977, 1978).

The visibility of the 1974-1975 economic crisis was low in Sweden. Unemployment was held at bay by large government subsidies to private industries, which enabled managers to keep plants running instead of firing workers. This led to a build-up, even an over-production, of goods. There was no strong opposition to this policy as Prime Minister Palme succeeded in obtaining consensus on the most important issues. Since the 1973 election had resulted in a parliamentary dead-lock, the Social Democrats actively sought support from other parties. As opinion data show, the Social Democrat Party obtained a better rating in 1976 in the area which includes willingness to cooperate with other parties (Table 1). Cooperation included a series of negotiations (the so-called Haga rounds) over taxation policy during which Palme secured the backing of the Liberals, and later also of the Center Party. The unions also participated in the Haga talks, guaranteeing that wage demands would not break the negotiation outcome.

The Social Democrat Defeat in 1976

What then lay behind the Social Democrat defeat in 1976? Three issues
were decisive: bureaucracy, the future ownership of private enterprise,
and nuclear power.

During the inter-election period, the Social Democrat government became more and more identified with a growing state bureaucracy. In 1975 and early 1976 there occurred a sequence of widely publicized events, named 'the affairs'. They included Ingmar Bergman's tax troubles, author Astrid Lindgren's campaign against a tax rate of 102 percent, the bugging of telephones by the secret police, intelligence activities directed against hospital employees, a trade union leader's trip to officially-boycotted Spain, tax deductions claimed by leading Social Democrats, and the arrest of the Social Democrat treasurer by customs officials for smuggling dollars into Finland. The underlying theme of 'the affairs' was that the Social Democrat government was losing its ties with the common people. By its insensitivity to grass-roots critique, the Social Democrat Party was seen as misusing its power. While specific events probably had only a marginal impact on the electoral outcome, the cumulative effect was a growing sentiment of general distrust of the government.

The bourgeois opposition could utilize these feelings of unease by launching an attack against trade union proposals to socialize land and to set up union-controlled profit sharing funds. The opposition accused the Social Democrats of trying to concentrate all economic power in the hands of a few union bosses and state bureaucrats. Since Liberals also directed attacks against big business, 'against both public and private concentration of power', it is accurate to describe the opposition's campaign as petty-bourgeois. There was indeed a populist undercurrent in the electorate. In the interview study respondents were asked whether a number of organizations and social groups had too little or too much power (Petersson 1977:110 ff). The group that most people thought had too little power was 'small businessmen'. The two groups that were considered much too powerful were 'big business' and 'LO', that is the central trade union federation.

While energy questions had not entered the political arena during the 1973 election, three years later the scene had changed. The parties had engaged in far-reaching study campaigns and propaganda efforts so that in 1975 the Riksdag was able to take a decision upon an energy program for the coming decades. It is significant to note that cleavages among the parties did not follow traditional left-right lines. Social Democrats and Conservatives were both favorable to nuclear power. The Liberals also

took a positive stand, though they supported the construction of a more limited number of plants. The Center Party opposed nuclear power, with its youth organization vigorously engaged in a fight against the construction of as many as thirteen nuclear plants. Combining resistance against atomic energy with a vision of environmental protection, energy conservation and small, self-sufficient local communities, the Center Party represented the 'green wave' in Swedish politics. Communists were internally split on the issue of nuclear power: the Moscow-oriented minority group opposed the 'ecological' line of the majority and advocated a policy which resembled that of the Social Democrats.²

Despite rather clear differences of opinion among the political parties on the subject of nuclear power, it was not at all obvious at the beginning that this would be one of the main controversies of the 1976 election. Indeed, one month before the election it looked as if the nuclear issue had been forgotten. In particular, the Social Democrats did not want to raise the question as their followers were split. The polls showed that a significant minority of Social Democrat voters took positions different to the official party line. Thus there were at first very few references to the energy question in the public debate, but on August 25 the Center leader, Fälldin, broke the silence. In a press conference which was given wide coverage in the newspapers, radio, and television, Fälldin launched a tough program on energy conservation. Claiming that all nuclear power plants could be dismantled by no later than 1985, he also gave the nuclear issue a special moral aspect by stating that he would never enter a government that initiated the operation of a single additional nuclear power plant. From this point on, nuclear power was the issue that dominated the rest of the campaign. According to the previous polls, the Center Party had lagged behind, even coming close to disappearing from the focus of the campaign. But this trend was reversed when the Center Party entered the nuclear debate: its initiative forced the Social Democrats into a defensive position. The interview study proves that the Center strategy was successful.³ By the end of the campaign the Social Democrats had lost some support and the Center Party had captured voters on the nuclear issue not only from the Social Democrats but also from the Liberals and Conservatives. A major electoral defeat was thus avoided and the Center Party suffered a loss of only one percentage point in comparison to its 1973 performance.

To summarize, six months before the election the situation was very dismal for the Social Democrats. Polls in the Spring of 1976 showed that popular support was clearly below the already low level of the 1973 election. If nothing occurred, the bourgeois parties were bound to sweep

the election. But the trend did change. In the late Spring support for the Social Democrat Party began to increase. Unemployment was not, as in 1973, a drawback. But the bourgeois anti-socialization campaign and the nuclear debate took the wind out of the Social Democrat sails, and shortly before the election popular support sank again. By election day the bourgeois parties had secured a majority of ihe electorate. Palme resigned, and Sweden had a new government.

Aggregate Stability and Individual Volatility

Stability is often said to be the hallmark of Swedish party politics. Superficially, the 1976 election was no exception as net changes were quite modest. Voter turnout was high with 91.8 percent of the electorate participating in the election. Today the party spectrum in Sweden looks much the same as it did fifty years ago. No new parties have attracted any significant voter support. In this respect the Swedish situation stands in contrast to that of many other Western political systems in which protest parties have lately been a typical feature.

But the traditional image of the stable Swedish voter is now misleading.
Aggregate stability is combined with a large and increasing individual
volatility.

In the mid-1950s the proportion of voters switching parties between election was 7 percent. This figure has risen continously. Between 1973 and 1976, 19 percent of the voters changed their party choice (Table 2). Behind the small net changes there were large flows of voters; but as the trends had opposite directions, to a large extent the effects cancelled each other out. In addition, today's voters make their decisions closer to

election day. In 1964, 18 percent of the voters made their voting decision
during the election campaign, whereas in 1976, 28 percent of the voters
determined their preferences during the campaign.

A growing number of Swedes distrust parties and politicians. In 1968,48 percent agreed with the statement 'Those Members of the Riksdag whom we elect quickly lose contact with common people'. Five years later the figure in agreement was 58 percent, and in 1976 it was 66 percent. A similar pattern holds for all measures of alienation and distrust (Table 3). There has been, in short, a growing sense of dissatisfaction, a widening gap between voters and representatives.

The basis for this alienation is political rather than social. High distrust is found among voters who think that there are important economic and social problems in society to be solved, but who also have a pessimistic view of the capability of their own party to do something about it. Compared to this trend of decreasing trust, the drop in party identification has not been as sharp. Alienation has grown also among party identifiers. But although many voters recognize the failures of politicians, they still cling to the old parties. Distrust is still contained within the established party system.

One obvious explanation for the absence of grievance or protest parties in Sweden is the predominance of the Social Democrat Party. The bourgeois opposition has remained an untried viable alternative. Now the parliamentary situation has been reversed. The bourgeois parties run the country, yet taxes are still high, and inflation is still in double-digits. The immediate future of Swedish party politics will to a large extent be determined

by how the bourgeois voters react. Will they still support the
bourgeois bloc, in spite of the faet that changes in society are smaller than
many of them had hoped?

Four Voter Types

The present reshaping of the Swedish electorate can be illuminated by a
typology consisting of four voter types. They are defined as follows:

Party Activists are characterized by very high voter turnout. They make their decisions long before the election and they switch parties infrequently. They have high trust in politicians and parties. In this group both social class and political attitudes are strongly correlated with party choice. Party Activists are most frequently found among older voters.

Mavericks participate in elections fairly regularly, make their decisions late during the campaign, change parties frequently, and display rather low trust in politicians. Attitudes are more important than social class in explaining party choice. The typical social feature of Mavericks is high education.

Ritualists also participate quite regularly, but the great majority have already made their voting decisions before the opening of the campaign. Voting stability is high and political trust widespread. Class is a better predictor of party choice than attitudes. Ritualists have lower education than the average voter.

Passivists participate more irregularly. If persons in this category vote,
they make their decisions shortly before election day. Inter-election stability
is low and alienation is high. Neither class nor attitudes are very good

explanations of voting decisions. Passivists are most frequently
among younger electors.

found

Party Activists comprised about 40 percent of the electorate in 1976. The size of each of the three other voter types was about 20 percent. During the last decade the number of Mavericks has increased, while Ritualists have become fewer. This is one reason why the Swedish electorate has become more volatile.

Trends in Class Voting

One of the most important findings of the election study is that class voting is decreasing. Voters today cross traditional class boundaries more often than some twenty years ago. But while this can be recognized as an ongoing process, it should also be pointed out that each party still has a characteristic social base. The class base of the Swedish parties is by no means completely eroded away.

The important cleavages of Swedish party politics still reflect the contradictions upon which the party system was founded in the nineteenth and early twentieth centuries. Although the important political conflicts in Sweden could be summarized by a single left-right scale, a two-dimensional map gives a fuller description of the political front-lines. Figure 1 is based on voters' perceptions of the social support of the five parties (see Table 4 and Legend for further information). The horizontal, west-east

axis captures contradictions between labor and capital, between lower and upper class. The vertical, north-south dimension represents conflicts between industrial center and rural periphery. In the 1890s political struggles were mainly fought between urban and rural interests. At the turn of the century, however, the labor movement entered the political arena. Contradictions between labor and capital came to dominate political life.

The modal support of Social Democrats is found in the north-west area of the social map, among industrial workers. This result holds when the analysis is based on voters' perceptions as well as in actual voter support. The Center Party's strongest support is in the 'south', in the rural part of the population. The stronghold of the Conservatives is located in the 'east', in the upper class. The social support of the Communists and Liberals is more diffuse. But in voters' perceptions the Communists are located in the 'west' and the Liberal Party is placed somewhere between the Center Party and the Conservatives.

As was pointed out earlier, the trend during the last two decades has been toward a weaker association between class and voting (Table 5). One of the most significant examples of this trend is the expansion of the Center Party. The old Farmers' Party, renaming itself in the late 19505, has risen from 9 to 25 percent in less than twenty years. The Center support is especially strong among first-generation workers with farm background. But the decreasing class voting is explained not only by this change in working class behavior. There is an equally important process going on in old bourgeois strongholds.

In the mid-1950s the upper class almost completely supported the non-socialist parties. In 1956 no less than 96 percent in 'Social Group I' voted for the Conservative Party or the Liberals.⁴ But twenty years later it could be seen that the political sympathies of this group had changed drastically. In 1976 only 57 percent voted for these two bourgeois parties. During the same period the socialist share of Social Group I increased from 4 to 23 percent.

The increased tendency to cross traditional class lines in voting behavior is to a large extent an expression of generational change. Class voting is strong among the older voters, weak among younger (Table 6). The greatest political differences between generations are found in the upper class. Among salaried employees in higher positions, 52 percent in the oldest age group (51-80 years) voted for the Conservatives; the corresponding

figure for the young voters (18-30 years) in this social group was
only 17 percent (Petersson 1977:19). The Conservative Party's traditional
core group is crumbling.

The changes in social support have had a particularly profound effect for the Communist Party. In the early 1960s the typical Communist voter was still an old worker. Today the Communists have retained a base in the working class, winning increasing support from younger workers. But it is also true that today workers constitute a proportionally smaller part of the Communist electoral support. During the 1970s the Communist Party has gained a completely new stronghold: young employees with academic background, especially in social work, education, mass media, and culture.

The decreasing class voting can to a certain extent be explained by changes in the social structure. Social mobility during the last decade has resulted in a situation where today many voters have a different class position from that of their parents. Despite this pattern the election study shows that parents' class positions still make a strong impact upon voting behavior. This inter-generational 'political memory', combined with a high degree of social mobility, has had a depressing effect on the measure of class voting. However, structural change remains only a necessary, not a sufficient, explanation of the decrease in class voting. The new tendency of left voting within traditionally bourgeois groups cannot be explained by social mobility alone.

Long-Term and Short-Term Change

Let us now return to our initial question. Why did the Social Democrats
lose the 1976 election? In the post-election debate two different explanations
have been suggested.,5

According to one opinion the defeat must be seen in a longer perspective.
The traditional electoral base of the Social Democrats is vanishing.

Structural change, largely initiated by Social Democrat economic policy itself, has meant that the strongest electoral bases have begun to wither away. The shift in population away from northern Sweden coupled with rapid urbanization has dissolved the cohesion of the original labor movement and has resulted in gains for the bourgeois parties.

Another opinion argues that the Palme government was defeated by specific political issues. In this respect, the electoral setbacks of the 1970s can be explained by the fact that the bourgeois opposition was increasingly able to exploit public discomfort over certain aspects of Social Democrat policy, with the energy controversy of 1976 standing as but one example. The content of certain policies, not structural change, explains the electoral outcome.

Which is correct? How should the 1976 election be explained? Was it a result of long-term social forces or short-term political factors? The answer is that both opinions are correct in a certain sense. The two explanations are compatible.

The short-term analysis of 1976 is correct. The Palme government lost because many voters were worried about bureaucracy and boss rule. In addition, the bourgeois offensive on energy and socialization was successful.

But the long-term explanation is also correct. There have been changes in the social structure and these have altered the preconditions of voting behavior. However, they have not been entirely negative for the Social Democrat Party. The latter's decreasing support in the working class has been largely balanced by an increasing support among salaried employees.

The significance of the long-term processes of change is that old stabilizing forces are being dissolved. The electorate has become more volatile. The image of the stable Swedish voter is today becoming a myth. More and more voters abandon their old parties. Class voting decreases. A growing number of voters make their decisions shortly before the election. The gap between electors and elected widens.

This ongoing development will have important consequences not only for voters but also for the parties. With a more volatile voting support, party strategies will probably be increasingly flexible. There are also signs that the importance of party organization is declining.⁶ Instead we may see a growing significance of mass media and campaign planning. The political agenda of the 1980s will include partly new issues. The 1976 election showed that 'green' issues as the nuclear debate can have a decisive impact on an electoral outcome. Furthermore, if economic growth continues to be low, conflicts over distribution will increase. Resources for

health, child care, pensions, schools etc. will be scarce and politicians are going to face severe priority problems. Demands for increased influence from workers, employees, tenants, and other groups will be even stronger. Though some of these issues might be contained within the traditional left-right framework, many may result in new political cleavages. Thus, there are indications that nev/ coalition patterns might emerge in Swedish party politics.

---

NOTES

1. This article is basedon the 1976 Swedish Elec :ion Study. The study wasjointly carried out by the Department of Government at the University of Uppsala and the National Central Bureau of Statistics with the author as project director. A nation-wide sample of about 2,700 persons was interviewed, half of the sample before and the other half after the election in September 1976. The results are reported in Petersson (1977). A technical description is given in Petersson (1978). The previous election studies have been directed by Professor Bo Sårlvik. This article is a slightly revised version of c. paper presented at the ECPR workshop on 'Social Structure andPoliticalChange', Grenoble, April 6-12, 1978.1 am grateful to Susan Opper for her comments and suggestions.

2. In 1977 the minority Moscow group broke away and formed a new party, APK (The Labor Party Communists).

3. Similar results are reported in Holmberg et al. (1977).

4. Social Group I consists of salaried employees in leading positions, executives, professionals, estate owners, etc. Today it comprises about 8 percent of the electorate.

5. See, for example, the discussion between Jan Lindhagen and Walter Korpi in Tiden, the theoretical journal of the Social Democrat Party: Lindhagen (1976, 1977), Korpi (1977a, 1977b).

6. This trend is not unique for Sweden. Cf. e.g. Maisel et al. (1975).

REFERENCES

Holmberg, S., Westerståhl, J. and Branzén, K. 1977. Väljarna och kärnkraften. Stockholm:
Liber.

Korpi, W. 1977a. 'Vad hotar arbetarrörelsen?' Tiden 3, 133-151.

Maisel, L. and Sacks, P. M., eds. 'The Future of Political Parties'. Sage Electoral Studies
Yearbook, Volume 1. Beverly Hills, Calif.: Sage Publications.

Petersson, O. 1977. Väljarna och valet 1976. Valundersökningar, Rapport 2. Stockholm:
Statistiska centralbyrån/Liber.

Petersson, O. 1978. Valundersökning 1976. Teknisk rapport. Valundersökningar, Rapport 3.
Stockholm: Statistiska centralbyrån/Liber.

Särlvik, B. 1977. 'Recent Electoral Trends in Sweden', pp. 73-129 in Cerny, K., ed.,
Scandinavia at the Polls. Recent Political Trends in Denmark, Norway and Sweden.
Washington, D.C.: American Enterprise Institute for Public Policy Research.