Class Voting Research Paper

Historically, interest in class voting arose in response in part to the failed agenda of Marxism. For classic texts in political sociology electoral politics was an expression of ‘the democratic class struggle’ (Anderson and Davidson 1943) that replaced, at least for the short-term, the predicted class-based revolution. This in turn led to a focus on class voting as a dispute between just two classes, the working class and the middle class, and their representatives, parties of the left and right. Early research also relied on data obtained at the level of electoral constituencies. Given the well-known problems with inferring individual vote choices from aggregate data this work had to make strong assumptions about how voters in different classes actually voted. To help overcome this limitation, modern social scientific studies of class voting have focused on its measurement through surveys of voters. From its beginnings in a few national election studies this data base has expanded through time and space to create a large body of evidence on class voting that now informs academic debate.

1. The Two Class Two Party Era

The baseline for many current studies was set by Robert Alford’s (1967) analysis of trends in class voting in four Anglo-American democracies (Australia, the UK, Canada, and the USA) between 1936 and 1962 using a measure of social and economic position in which occupations were aggregated to form a dichotomous manual nonmanual class division. He also introduced the most commonly used, cited, and criticized measure of the level of class voting: the ‘Alford index.’ The Alford index is simply the difference between the percentage of manual workers that voted for left-wing political parties on the one hand and the percentage of nonmanual workers that voted for these parties on the other. This became standard practice in the cross-national and over-time analyses that followed, and which usually concluded that class voting is in decline. Lipset (1981), for example, presented evidence of a downward trend in the UK, Germany, and the USA between 1945 and 1980; while Inglehart (1997) points to evidence of a continuing decline from the early 1980s onwards. Indeed, whether or not class voting in modern industrial societies has declined is, for many commentators, no longer an issue. The question is why has it declined? A plethora of explanations have been proposed to account for this outcome (see Goldthorpe 1996, Manza et al. 1995): some assert that it has lost its importance as a determinant of life-chances and in consequence its role as a source of divergent political interests because the working class has become richer, white-collar workers have been ‘proletarianized,’ or because of extensive social mobility between classes. Others argue that ‘postindustrial’ cleavages, such as gender, race, ethnicity, public versus private sector, and various ‘identity groups,’ are emerging and replacing class-based conflict. Postindustrialism also impacts on class voting through the rise of postmaterial values as a basis of party preference, which has led to the ‘new left’ drawing its support from the middle classes thus weakening the class basis of leftright divisions; while rising levels of education and ‘cognitive mobilization’ have supposedly produced voters that are calculative and ‘issue oriented’ rather than driven by collective identities such as class.

Whatever their strengths and weaknesses—and there are many questions and doubts that remain unresolved in each case—all of these explanations assume that there is indeed a widespread, secular decline in class-based political action which then has to be explained. However, answers to questions about change have been strongly conditioned by choices of method and measurement. The literature on the decline of class voting that followed Alford’s work has been characterized by shortcomings in both these respects.

For example, one problem with the manual nonmanual distinction is that it obscures variations in the composition of the manual and nonmanual classes. Any changes in their composition may lead to spurious change in estimates of class voting. Thus, if skilled manual workers are more conservative than unskilled workers and the number of skilled workers increases, the difference between manual and nonmanual workers will decline even if the relative political positions of skilled, unskilled, and nonmanual workers remain the same. With this and similar indices the measurement of the class–vote association is open to confounding by changes in the shape of the class structure and in the general popularity of the parties. In other words, this type of index confuses differences in the marginal distributions of the variables with differences in the association it is supposed to measure. The same problem applies to the apparently more sophisticated ordinary least squares regression techniques used in, among others, Franklin, Mackie and Valen’s (1992) ambitious 16 nation study that represents, arguably, the culmination of this tradition.

2. Recent Innovations

In recent years, therefore, the rather atheoretical manual/nonmanual representation of class voting has been to a large degree superseded as researchers have taken into account more nuanced ideas about class structure. Most influential has been the class schema developed by John Goldthorpe and his colleagues (Erikson and Goldthorpe 1992). First used to study patterns of class voting over time in Britain by Heath et al. (1985) this instrument, unlike many other measures of class, has been empirically validated. The main classes identified by the Goldthorpe schema are the petty bourgeoisie (small employers and selfemployed), the service class, or salariat (professional and managerial groups), the routine nonmanual class (typically lower grade clerical ‘white-collar workers’), and the working class (foremen and technicians, skilled, semi-, and unskilled manual workers). The principal distinction underlying the distinction among the employee classes in the schema is between a service contract in which employees receive not only salaried rewards but also prospective elements—salary increments, job security and pension rights, and, most importantly, well-defined career opportunities—and a labor contract, in which employees supply discrete amounts of labor, under supervision, in return for wages which are calculated on a ‘piece’ or time basis. As the employment relationship of the service class in relatively advantageous in terms of employment and payment conditions, occupational security, and promotion prospects, its members have a stake in preserving the status quo. In contrast, the disadvantages of the labor contract can explain why the working class provides a basis of support for the redistributive programs of the left.

This shift to greater complexity in the measurement of class has been accompanied by a similar move away from the measurement of political choice as a dichotomy of left versus right (or left versus non-left) to a fuller representation of the voters’ spectrum of choice at the ballot box. Apart from its simplicity, the main reason for the use of a dichotomy to represent voter choice seems to have been a desire to make systematic cross-national and over-time comparisons. Unfortunately, the selective nature of what is being compared undermines any true comparability. The problem is analogous to that faced in the analysis of class position. Changes in the composition of composite categories such as ‘left’ or ‘non-left’ may lead to spurious changes in estimates of class voting. The use of dichotomies to represent vote choices and social classes also precludes from observation any processes of class–party realignment. The concept of class realignment in voting implies a change in the pattern of association between class and vote without any change in the overall strength of this association—i.e., without class dealignment (or, of course, increased alignment). But this cannot be discerned if the distinction between realignment and dealignment is prevented by restricting the numbers of parties and classes to two.

The final innovation of the 1980s and 1990s is in the statistical measurement of the class–vote association, with a move from Alford type indices to ones that are based on odds ratios, which measure the strength of the relationship between class and vote independently of the general popularity of political parties or changes in the sizes of classes. The use of logistic modeling techniques likewise capitalizes on this basic feature of the odds ratio and allows the relationship between more complex class and party systems to be summarized, thus complementing and facilitating greater sophistication in the representation of class and vote.

Research using these advances comes to rather different conclusions than those in the two-class, two-party tradition. Studies in the UK (i.e., Heath et al. 1995) have found little evidence of declining class voting and have concluded that ‘trendless fluctuation,’ or at most a ‘one-off’ change, best captures the pattern of association in class voting over time. The USA, by comparison has seen the growth of a new class–vote cleavage—between those who vote and those who do not—as a deficit of working class interest representation that has been associated with increased class differences in participation (Hout et al. 1995). Various studies reported in Evans (1999) show that although there is evidence for a linear decline in left versus nonleft class voting (Nieuwbeerta and de Graaf 1999), it is not typical. For example, when examined over the long-term, levels of class voting in the UK were found to have increased during the 1940s and 1950s before falling back in the 1960s; while the 1990s have seen increased levels of class voting in new post-communist democracies as these societies have undergone the rigors of marketization. Only in certain Scandinavian countries is there robust evidence of a decline from an unusually high degree of class voting to levels similar to those in other Western democracies.

3. New Directions

The debate on class voting has been strong on evidence and weak on theory. Most scholars have assumed a sociological, relatively deterministic account of the transition to industrial and postindustrial politics, but there are those who have rejected these in favor of more voluntaristic models. Kitschelt (1994), for example, argues that the electoral fortunes of European social democratic parties are largely determined by their strategic appeals, rather than by secular trends in the class structure—a line of reasoning that echoes Sartori’s (1969) influential emphasis on the importance of organization, and especially parties, in the creation of class constituencies. From this perspective there is reason to believe that even in advanced industrial societies class voting might increase as well as decrease. It also implies that the adoption of class-relevant policy programs should be associated with an increase in the class basis of partisanship. Some evidence for this has been presented by Evans et al. (1999) who show a relationship over time between left-right polarization in parties’ manifestos and the extent of class voting.

This is not to say that sociological changes have no impact, however. Changes in the relative sizes of classes have been thought to have implications for party strategy: most importantly, in a change to a ‘catch-all’ strategy by parties on the left in response to the shrinking class basis of support for those parties (Przeworski and Sprague 1986). In some political systems such moves leave open the space for left parties to attract support from marginalized working class groups; in others, such as first-past-the-post systems, we might expect that the start-up costs for electorally viable left parties would be too great. Unfortunately, many of these arguments still await rigorous tests using survey analysis. Research that tries to unravel why voters in different classes vote differently is in short supply, though Weakliem and Health (1994) provide evidence of the role of rational choice and inherited preferences, while Evans (1993) models the effects for future expected rewards and the likelihood of promotion conditioned by position in the life-cycle. A more explicit link between models of class voting and their dependence on advances in theories of voting behavior more generally is an area where further development might usefully occur.

Bibliography:

1. Alford R 1967 Class voting in Anglo-American political systems. In: Lipset S M, Rokkan S (eds.) Party Systems and Voter Alignments: Cross-National Perspectives. Free Press, New York, pp. 67–94
2. Anderson D, Davidson P E 1943 Ballots and the Democratic Class Struggle. Stanford University Press, Stanford, CA
3. Erikson R, Goldthorpe J H 1992 The Constant Flux: A Study of Class Mobility in Industrial Societies. Clarendon, Oxford, UK
4. Evans G 1993 Class, prospects and the life-cycle: explaining the association between class position and political preferences. Acta Sociologica 36: 263–76
5. Evans G (ed.) 1999 The End of Class Politics? Class Voting in Comparative Context. Oxford University Press, New York
6. Evans G, Heath A, Payne C 1999 Class: Labour as a catch all party? In: Evans G, Norris P (eds.) Critical Elections: British Parties and Voters in Long-term Perspective. Sage, London, pp. 87–101
7. Franklin M N, Mackie T T, Valen H 1992 Electoral Change: Responses to Evolving Social and Attitudinal Structures in Western Countries. Cambridge University Press, Cambridge, UK
8. Goldthorpe J H 1996 Class and politics in advanced industrial societies. In: Lee D J, Turner B S (eds.) Conflicts about Class: Debating Inequality in Late Industrialism. Longman, London, pp. 196–202
9. Heath A F, Evans G, Payne C 1995 Modelling the class party relationship in Britain, 1964–92. Journal of the Royal Statistical Society, Series A 158: 563–74
10. Heath A, Jowell R, Curtice J 1985 How Britain Votes. Pergamon, Oxford, UK
11. Hout M, Brooks C, Manza J 1995 The democratic class struggle in the United States. American Sociological Review 60: 805–28
12. Inglehart R 1997 Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton University Press, Princeton, NJ
13. Kitschelt H 1994 The Transformation of European Social Democracy. Cambridge University Press, Cambridge, UK
14. Lipset S M 1981 Political Man: The Social Bases of Politics (expanded and updated edition). Heinemann, London
15. Manza J, Hout M, Brooks C 1995 Class voting in capitalist democracies since World War II: dealignment, realignment or trendless fluctuation? Annual Review of Sociology 21: 137–62
16. Nieuwbeerta P, de Graaf N D 1999 Traditional class voting in 20 postwar societies. In: Evans G (ed.) The End of Class Politics? Oxford University Press, New York, pp. 23–57
17. Przeworski A, Sprague J 1986 Paper Stones: A History of Electoral Socialism. University of Chicago Press, Chicago
18. Sartori G 1969 From the sociology of politics to political sociology. In: Lipset S M (ed.) Politics and the Social Sciences. Oxford University Press, Oxford, UK, pp. 65–100
19. Weakliem D L, Heath A F 1994 Rational choice and class voting. Rationality and Society 6: 243–70