Irita Kise

The Decline of Class-Based Voting

The importance of voting behaviour studies is widely recognized. It has resulted in many books, researches and theories. Big number of scientists have tried to explain the factors which determine the party choice. A number of theories has been built on analyses of survey data, very often the interpretations have been contradictory.

This century is said to be the century of changes. Indeed, the difference between life in the beginning of this century and these days is enormous. Ever changing circumstances make propounding of a theory which explains everything practically impossible. That is the case in studies of voting behaviour. Changing social structure and party systems have challenged even the most well-founded theories.

In my essay, I try to describe the changes which have occurred in the Western societies since 1950s and their impact on voting behaviour. As a starting point I take the cleavage model of Lipset and Rokkan. That model for long time was the most authoritative explanation of existing party systems. As new developments changed the stable party alignments also new explanations were put forward. Although they are rather impressive and are based on reliable data they do not explain entirely why people vote for one party or another. I come to conclusion that old explanations are still important while new ones gradually gain the ground. My answer to the question then would be - class-based voting has been supplemented by value-based voting and in some segments of society replaced by it.

Social structure and party system

It has been for a long time an unquestionable truth that social structure of society accounts for the strength of parties in given country. Membership in a population group leads to a specific attitude profile, either through socialisation or through common stimuli of interest aggregation, and these political attitudes have a decisive influence on voting. Rokkan and Lipset have argued in their popular study that the cleavage structure which underlay party alignments was the same in 1960s as it was in 1920s. Such party stability was achieved because of the capacity of parties, once established, to retain their hold on sections of electorate. Old social cleavages created a party system which thereafter structured electoral choice itself. New voters had to choose between already existing parties and they did it so more or less relying on their class bias. Data proves that in 1950s more than half of electorate voted for their natural parties, in some countries even more, for example, Great Britain and Scandinavian countries. Although class has been regarded as the most significant determinant of party choice, Rokkan and Lipset identify more dimensions of cleavages: conflicts between landed and industrial interests, conflicts between different religious groups or between pro-church and anti-church elements (Ware, 1987:19). All these cleavages came about in two revolutions - national and industrial. National revolutions resulted in centralising process in which elites imposed their authority on groups accustomed to greater autonomy. That stimulated the formation of parties designed to defend the interest of the periphery or the church against the centralising elite, these are called “pre-industrial” cleavages. Industrial revolution created new conflicts between employers and workers. Besides social class “industrial” cleavages comprise also education and income. The growth of the state since 1945 led to the expansion of the public sector. Divisions were growing between those who worked for the state and those who worked for private sector firms, between those who were directly dependent on the state for financial support and those who were not. These later cleavages are regarded as secondary because the older ones such as class and religion have firmly established themselves. The visibility of these cleavages has been proved by relatively stable party systems in Western Europe, at least until 1970s.

In different countries different cleavages are dominant. For example, religion is the most important cleavage in Belgium, the Netherlands and Germany. The urban-rural cleavage is the most important cleavage in Ireland. Social class is the most correlated to party preference in Britain (Knutsen, 1988:337).

Social class

Class is traditionally regarded as the most important. Not only have class differences historically determined the very structure of modern party systems but, until recently, class differences were also viewed as the pre-eminent source of political conflict within western societies. Lipset concludes:

“Even though many parties renounce the principle of class conflict or loyalty, an analysis of their appeals and their support suggests that they do represent the interests of different classes. On a world scale, the principle generalization that can be made is that parties are primarily based on either the lower classes or the middle and upper classes.” (Quoted by Hayes, 1995:70).

One problem which has resulted in different survey data is the division between classes. Goldthorpe’s basic three category classification differentiates between service, intermediate and working class. The service class is composed of professionals, administrators and managers. The intermediate class is confined to routine non-manual workers, personal service workers, the “petty bourgeoisie”, and supervisors of manual workers. Agricultural labourers and skilled, semi-skilled, or unskilled manual wage workers compose the working class. More often there is a need to divide between working class and middle class. For example, Swedish classification procedures have resulted in a rather broad operationalization of the working class. Not only manual workers, but also a large segment of lower level nonmanuals have been assigned to the working class. The nonmanuals consist of shot assistants and low level public employees in transport and certain services. Thus the working class does not rely solely on the distinction between manual and non-manual workers. Instead it quite close corresponds to the occupational groups organized by the Swedish Confederation of Trade Unions. Special problem is pensioners. The Norwegian and Swedish election surveys classify pensioners according to their former occupations. By contrast, pensioners have frequently constituted a separate category in Danish election studies (Sainsbury, 1987). This discussion is in regard of objective class position which is in most cases used in surveys. Subjective class identification is also important, if not even more important than objective class position. Perception of ones position may often be different from the actual position. Hayes presents interesting data on United Kingdom and Germany about the differences in class position and class self-identification.

Class position	United Kingdom	Germany
service	25.6%	21.7%
intermediate	39.7%	42.1%
working	34.8%	36.2%
Class self-identification
middle	27.9%	59.8%
working	72.1%	40.2%

(Hayes,1995)

The working class is the second largest in both countries, comprising about one third of the populations. The tendency for self-placement in the “working class” is highest in Britain. While members of the service class are more likely to have a middle-class self-image, working class identifiers tend to hold a working class objective class position. Class self-identification appears more salient then class position when it comes to party choice. It was rather significant until 1970s when many Western European party systems began to change. Earlier quite stable party systems now had change as their most characteristic feature. In Belgium, Switzerland and Britain new electoral support was given to parties based upon linguistic and ethnic cleavages long thought to have been depoliticized. In other countries new political parties championed different causes that cut across existing party lines. Many countries saw a decline in the electoral support given to left parties (britain and ireland); but in other countries left parties have done unexpectedly well in recent years. In all countries a set of environmental issues were placed on the political agenda by “green” parties (Franklin, Mackie and Valen, 1992:7). Two controversial explanations have been discussed in literature: electoral changes are due to social changes and the opposite view - social structure is irrelevant. Proponents of the social structure relevance focus on proportions of social classes in society and changes in cleavage patterns. Social cleavages are seen as defining the supporters of competing parties and structural changes automatically affect such cleavages. On this view, economic growth that leads to a larger middle class automatically increases the support for bourgeois parties; and deindustrialization reduces the sizes of groups normally thought of as supporting socialist parties. In countries with agrarian parties, a decline in rural population would weaken their support. Other changes in electoral politics might similarly be predicated upon changes in the size of particular social groups (Franklin et al., 1992:8). The evidence is the declining electoral results for left wing parties. Many countries have become predominantly white-collar. According to this view socialist parties loose their basis of support as working class declines. This is a result of technological development which can be compared with industrial revolution in the extent of changes it brings in structure of society.

The decline of class voting

Nevertheless, class-based voting is reported as sharply declining. In countries where left parties retain their aggregated strengths, they are nevertheless failing to gain the support of their natural constituency to the extent they once did. Alford introduced one way how to measure class-based voting. Alford index is obtained by taking the percentage of the working-class voting for the left wing party and subtracting from this the percentage of the middle-class who vote for the party of the left. In his original study of voting in the mid 1950s, Alford gave scores of 41 for britain. In continental Europe, class-voting is much less prevalent than in britain. However, in Scandinavia class voting exceeds even the British level: the average Alford index for Denmark, Norway and Sweden was around 55 in the mid 1950s though this had fallen to 35 by 1980 (Harrop and Miller, 1987:186). The decline of traditional working class communities, increasing geographical and social mobility, the growth of home ownership and rising education standards have lead to a declining proportion of workers fitting the traditional class stereotype. The growth of welfare state has also helped to soften the sharpness of cleavage distinctions. The blurred distinction between classes has led to decreased class loyalty. Cleavage politics depend for their existence on loyalties to social groups. Such loyalties are generally inculcated during childhood and reinforced by repeated affirmation later in life. Class voting is reported to be lower among the young than the old. Changed social situation influences more young people who are open to such changes because they are not immunised against them by repeated affirmations of loyalties. If this reflects generational rather than life-cycle effects, class voting is likely to continue its decline as older generations die off (Harrop and Miller, 1987:187). The assumption that class-based voting is declining leads to conclusion that social structure does not matter. If individuals are no longer voting on the basis of group loyalties, then we would not expect changes on the strength of political parties to reflect changes in the size of social groups. It is true that only with help of social structure little can be explained in these days when electoral volatility grows, new parties are created, traditional ones are loosing support. Now parties must mobilize electorate before every election, no more they can rely on constant segment of society and only try to maintain its support.

The decline of class theory is not without its critics. They say that the decline of class-party nexus is relative and there is still substantial evidence to suggest that political parties remain divided on class lines. Recent study of British electorate by Marshall (1988) reassures the primary importance of class as a predictor of voting patterns (Hayes, 1995:71). Philip Norton argues that although the central variable of class - occupation - has declined in importance in structuring party voting, class-supportive variables have become more important, notably housing tenure and trade union membership (Norton, 1984:36). These are also quite convincing arguments but I would rather agree with Clark’s conclusion:

“Social classes may not have died, but their political significance has declined substantially. Out point is that social class has declined in its ability to explain social and especially political processes. But it still lives.” (Quoted by Hayes,1995:71)

If social class does not account for electoral changes in 1970s and later on, new explanations are needed. Indeed, in literature there is no lack of theories. Declining cleavage politics can be regarded as a consequence of the successful resolution by political systems of deep-seated conflicts of social interests. Many countries have succeeded in politicizing their major social conflicts in the form of cleavage-based parties. Politicization in this specific form not only served to express group-based conflicts of interest, but simultaneously promoted their resolution by integrating them into a single national political system (Eijk et al., 1992:423). This statement makes sense but many authors have challenged it by proposing that instead of old cleavages new ones have emerged: differences based not on traditional economic issues but arising from non-economic factors such as freedom of speech, the environment, women’s rights. The very creation and success of the German Green party in 1980s, points both to the emerging importance of these non-economic issues in the minds of the young, well-educated and post-materialist West German voter as well as the increasing irrelevance and rejection of the traditionally class-based parties in meeting their electoral demands. The enormous increase in issue diversification, emergence of new issues and new organizations representing such interests correspond to new groupings of individuals based on differences in values and social priorities which cannot in any simple fashion be reduced to traditional differences in objective social positions. There has been a specialization of life. There is no more one unifying source of information which could provide basis for sense of solidarity. As Alvin Toffler calls it, “de-massification” has occurred in Western societies.

Post-industrialism is regarded as the fourth wave of change (after national and industrial revolutions and growth of the state) which creates new lines of cleavages. As affluence spreads, as education standards rise, there is some evidence of a shift in Western democracies from class politics to value politics - to a politics which emphasises the quality of life rather than just material well-being.

Parties themselves have also changed in response to declining social group loyalties. More and more parties adopt image of “catch-all” parties. They must rely on charismatic leadership and well-organized electoral campaigns. There is no more very big difference between consequences of electing one or another party. The consensus between dominant parties has been reached on major issues of state. For example, in Sweden no important decision is taken before consultations with opposition party. The inherent determination of political process limits the party’s chance to make quick changes in policy direction. If some decades ago welfare state was one issue on which different parties had different policies, now there is consensus that providing social security is important function of the state.

There are some doubts about how persistent these new cleavages are. Empirical evidence shows that in times of recession post-materialist values tend to loose the importance in party choice but they do not fade away entirely. They still play great role. Materialist - post-materialist cleavage is the most important cleavage in Denmark and Italy and the second most important variable in France, the Netherlands, Germany and Greece (Knutsen, 1988:337). Great Britain seems to be an exception here. The New Politics cleavages had little impact on British politics. Not only have political movements based around these “quality-of-life” issues been extremely slow to develop but, attitudes towards these issues were just as likely to reflect the same traditional class cleavages as the original economic issues of the old agenda (Hayes, 1995:72). Several reasons might be put forward to account for this apparent lack of impact: stress on aesthetic and intellectual needs is not new in Britain; relative rigidity of a long-established two-party system provides situation where meaningful change could only come from individuals switching their votes from one major party to another (Franklin, 1985:164).

There is no doubt that class as determinant of voting choice is declining. Process of pluralization has taken place. Perhaps, some countries are going through these changes earlier than others and that explains differences between countries. Nevertheless, traditional class-based voting has been supplemented by other motives of choice. Values are not in the last place to determine voters choice of party.

Literature:

Franklin, Mark (1985) “The Decline of Class Voting in Britain”, Oxford:Clarendon Press

Franklin, Mark et al. (1992) “Electoral Change. Responses to Evolving Social and Attitudinal Structures in Western Countries”, Cambridge University Press

Harrop, Martin and Miller, William (1987) “Elections and Voters. A Comparative Introduction”, Macmillan

Hayes, Bernadette (1995) “The Impact of Class on Political Attitudes” from European Journal of Political Research, vol.27, pp.69-91

Knutsen, Oddbjørn (1988) “The Impact of Structural and Ideological Party Cleavages in West European Democracies. A Comparative Empirical Analysis” from British Journal of Political Science, vol.18, pp.323-352

Norton, Philip (1984) “Britain: Still a Two-Party System?” from “Party Politics in Contemporary Western Europe” ed. By S.Bartolini and P.Mair

Sainsbury, Diane (1987) “Class-Voting and Left Voting in Scandinavia” from European Journal of Political Research, vol.15, pp. 507-526

Ware, Alan (1987) “Political Parties. Electoral Change and Structural Response”, Basil Blackwell