Sociology Of Voting Research Paper

1. Paradigms Explaining Voting Choice

Electoral geography was the first, and for a long time the only, clue for understanding political behavior, linking voters’ choices to the social structures and the religious and political traditions of the place where they lived. Since the earliest studies of Andre Siegfried, Charles Gosnell, or Herbert Tingsten, this ‘contextual’ or ‘ecological’ approach, based on the analysis of aggregate electoral data, has gained in statistical sophistication, as illustrated by the school of Johnston in Britain or King in the United States. Its limitation lies in the fact that the unit of analysis considered is the group—people living in the same constituency, the same district, the same region, etc.—and that the correlations observed at the aggregate level do not necessarily hold true at the level of the individuals, as Robinson’s (1950) famous article on the ‘ecological fallacy’ reminds us. Opinion polls provide a better tool for understanding individuals’ behavior, asking them directly how and why they vote. Since World War II, survey-based research has given a new impetus to electoral sociology, especially in the United States, where the pioneer studies and the classic models explaining vote choice were developed.

The ‘sociological’ model originates from the local surveys conducted by Paul Lazarsfeld and his colleagues at the Bureau for Applied Social Research of the University of Columbia. Repeated interviews with a randomly selected panel of voters, first in the district of Erie County, Ohio during the 1940 Roosevelt– Wilkie presidential campaign, then in the town of Elmira, New York, at the time of the Truman–Dewey contest reveal, to their surprise, that the campaign has but a limited effect on voters’ choice. Mostly it reinforces long-standing and diffuse political preferences which reflect the norms of the groups they belong to. Constructing an index of political predisposition based on social class, religion, and rural/urban residence, the authors of The People’s Choice can predict accurately how people vote. Those combining a rural residence, Protestant affiliation, and high social status tend to support the Republicans, while those who are working class, nonprotestant, and residing in an urban area mostly vote Democrat. The main conclusion of their book is that ‘A person thinks, politically, as he is, socially. Social characteristics determine political preference’ (Lazarsfeld et al. 1944). And this social anchoring accounts for the relative stability of voting choice, except among the cross-pressured or poorly integrated voters. However, in his later empirical and theoretical work such as Voting (Berelson et al. 1954), Lazarsfeld was to develop a more complex view of action in general and more particularly of voting behavior (his research evolution is related in Lautman and Lecuyer, 1998).

Reacting against what they see as a social determinism, Angus Campbell and his colleagues from the Survey Research Center (SRC) at the University of Michigan propose a psychological model focusing on individual decision-making. Reconstructing the chain of events that converge in what they call the ‘funnel of causality,’ they give special attention to the attitudes formed towards candidates, issues, and parties. On the basis of nationwide panel surveys conducted in the presidential elections from 1948 to 1956, they state in The American Voter (Campbell et al. 1960) that party identification is the key factor orienting vote choice. By that they mean an affective attachment to a specific party, formed early in life, generally passed over from parents to children, and reflecting the individual’s social and religious background. It functions as a filter coloring their political perceptions and guiding their votes, especially for the least sophisticated: ‘Like the automobile buyer who knows nothing of cars except he prefers a given make, the voter who knows simply that he is a Republican or Democrat responds directly to his stable allegiance without the mediating influence of perceptions he has formed on the objects he must choose between’ (Campbell et al. 1960). Lazarsfeld portrayed voters as social conformists; the Michigan team sees them more as political believers. Yet their conclusions are very similar: the average voters are far from the democratic ideal of a rational and active citizen. Most of them are not interested in politics, display low levels of ideological awareness, and are unfamiliar with the candidates and the issues at stake. They vote according to long-standing political preferences shaped by their early socialization.

The third approach stems from Downs’s seminal essay, An Economic Theory of Democracy (1957). It postulates, in the line of the neo-classic economic paradigm, that voters are rational and that their choice is based on a cost-benefit calculation. Voters cast their votes for the party or the candidate from which they expect the maximum ‘utility’ or benefits. The electoral decision between party A in office and opposition party B boils down to a simple equation, the difference between the benefits expected from the victory of party A (Ua) and the benefits expected from party B (Ub). Voters will favor party A if they see the difference between Ua and Ub as positive, party B if the difference is negative, and not bother to vote if they see no difference. Since Downs’s pioneering work, elaborate econometric models have developed. Some correlate aggregate electoral results with global economic indicators such as purchasing power, inflation rates, and unemployment rates. Others use survey data to analyze individual pocket-book voting, based on the evaluation of candidates past (retrospective voting) or future performances (prospective voting). What they have in common is to shift attention from the people’s demand to the supply side of politics, the specific goods offered on the electoral marketplace by competing parties and candidates.

2. Alignment, Dealignment, Or Realignment?

These three models have been guiding electoral research agendas. In the wake of the sociological model attention was first given to the effects of social class and religion, even more so in Europe where party systems were historically built on the basis of social and religious conflicts and cleavages (see Lipset and Rokkan 1967). Rose’s (1974) classic study of voting patterns in 15 democracies shows that class in Scandinavian countries and religion in the others are by far the most predictive factors explaining postwar electoral behavior, a working class position predisposing to a left-wing vote and regular church attendance to a right-wing vote. Then, in the 1960s, the Michigan paradigm became the leading model, inspiring nationwide research surveys in Europe, especially in Britain where Butler and Stokes (1969) explained long-term alignments of British voters by their party identification, itself rooted in their social background. What became known as the ‘two class, two party’ model does not fit as well in countries such as France with multiparty systems. There, Left Right orientations are the substitute for party identification, equally rooted in religion and social class (Michelat and Simon 1977).

Neither model though can account for the increasing electoral change that appeared in the United States by the end of the 1960s, and in Europe at the end of the 1970’s. On the basis of the SRC presidential surveys between 1956 and 1974, the authors of The Changing American Voter (Nie et al. 1976) revealed a decline of party ties. The proportion of ‘independent’ voters, who define themselves as neither Republican nor Democrat, has risen from 23% in 1964 to 40% in 1974. Among those who still identify themselves with one side or the other the proportion who declare a ‘strong’ attachment or have a positive image of a party is declining, and the very link between party preference and voting is weakening. Voters seem more interested in politics, more aware of the issues at stake and more inclined to choose their candidates according to their political stands, whatever party they belong to.

Parallel studies in Europe detect similar trends of electoral volatility or instability of voting choice and partisan dealignment or gradual moving away from all parties (Crewe and Denver 1985), seen as a structural feature of the emerging postindustrial society. Social and geographical mobility is blurring the traditional class cleavages and loosening community ties. Progress in education and exposure to the media have increased the average levels of political sophistication. The rise of the permissive and individualistic set of values coined as postmaterialist by sociologist Inglehart (The Silent Revolution, 1977) encourage a new style of politics, more demanding and protestprone, and promotes new post materialist issues— environmental, feminist—that cut through traditional party lines. All these trends converge to make citizens less dependent upon existing parties to make their choice. Voters are seen as ‘autonomous,’ ‘strategic,’ ‘rational,’ or ‘reasoning’ and the developing cognitive sciences are increasing studies on political reasoning mechanisms, the way voters process and organize information, the cues they rely on to make a decision, etc. (Sniderman et al. 1991). Short-term issue or candidate centered voting could well be taking the place of the former stable ‘cleavage voting’ with a partisan, religious, or social base (Franklin et al. 1992).

But a second research trend is taking an opposite stand, questioning the extent of this dealignment process by casting doubts on the way it is measured. The most heated controversies have developed around the supposed decline of class voting. Until now the simplest and most widely used indicator has been Alford’s index, which measures the difference between the proportion of manual and nonmanual voters voting for the Left. For instance, if all the British working-class voters supported the Labour party and none among the middle classes in a given election, the value of the Alford index would be 100%, indicating perfect class voting. Measured by this index, the downward trend is undeniable. Class voting across post-World War II elections has decreased, for in-stance, by almost half in Britain, and more than two-thirds in Germany. But critics underline several weaknesses of such an index. Its dichotomous nature does not take into account the complexity of post- industrial occupational structure nor the complexity of party systems. It is statistically biased because it does not consider the changes in the marginal distributions of the two classes and of the two votes from one election to another. And it only measures dealignment, while various processes of electoral realignment along new cleavages such as private sector vs. public sector, self-employed vs. wage-earners, or educated vs. noneducated voters, seem to be taking place in postindustrial democracies, as shown by multiclass approaches using detailed occupational classifications and sophisticated measures such as odds-ratios or loglinear models. In Britain for instance, in the long run, Heath et al. (1991), contrary to previous findings, find no decline of class voting, just ‘fluctuations without trend.’ In France since the 1980s, the salaried middle classes have tended to support the Left, compensating for the drop of working-class support, attracted by the extreme right embodied by the National Front (Boy and Mayer 1997). In the United States, Hout, Brooks, and Manza dispute Clark, Lipset, and Rempel’s claims about the declining political significance of social class, by showing the historical realignment in presidential post-War elections since 1968, as professionals and nonmanagerial whitecollar workers moved from supporting the Republicans to voting for the Democrats (special issue of International Sociology, 8(3), 1993, on class voting). Lastly, the new institutionalism approach attempts to reintegrate voters’ choices in their political context, analyzing their response to changes in the electoral rules of the game, in the party structure, and in the strategies of political actors, especially concerning their uses of the media.

3. The Puzzle Of Electoral Participation

An analysis of voting choice would not be complete without taking into account the alternative behavior: not voting. Whatever the election, a sizable proportion of eligible citizens are not registered or choose not to go to the polls. The proportion ranges between less than 5% in countries where voting is compulsory to 40% or more in Switzerland or in the United States. Whatever the country, the profile of these nonvoters, revealed by aggregate as well as survey data, is quite similar. Educated voters and members of the upper and middle classes participate more than the uneducated and the working classes, and middle-aged voters go more willingly to the polls than the very young or the very old. What has become known as the ‘socioeconomic status (SES) model’ (Verba and Nie 1972) postulates that well-integrated and upper-status individuals are more likely to participate because they have more resources to invest in politics—time, money, information, networks—and because they are more likely to develop the ‘civic’ attitudes (see Almond and Verba, The Civic Culture: Political Attitudes and Democracy in Fi e Nations, Princeton University Press, Princeton, 1963) that predispose to do so, such as interest in politics, a sense of duty, and a feeling of political efficacy. Political participation is socially skewed. There are compensating factors, however, such as group consciousness or ethnic awareness. With the development of feminism, the difference between female and male turnout rates has disappeared, and ‘black-political empowerment’ in the United States since the Voting Right Acts has reduced the difference between whites and African– Americans, the latter participating even more than whites when differences in SES are taken into account (Verba and Nie 1972). On the whole, the increasing educational level of the electorate and the growth of the middle classes should gradually raise electoral participation in postindustrial societies. But paradoxically the opposite is happening. There has been a clear drop in turnout rates among American voters, from 63% of the eligible voters in the 1960 Kennedy–Nixon election to 50% in the 1988 Bush–Dukakis contest. A similar decline is developing in European democracies but the trend is recent, milder, and educational differences are far less marked (Topf 1995).

This ‘participation puzzle’ has revived interest in rational choice models. Because one vote has but an infinitesimal influence on the electoral outcome, the rational behavior, at first sight, would be not to bother going to the polls, and since Downs, a large body of literature has been devoted to what appeared as the ‘voting paradox.’ It can only be understood if one extends the notion of rationality and if one admits that voting, like most human actions, is not only guided by interest (‘instrumental’ rationality) but by principles, and values (‘axiological’ rationality) (Boudon 1997, Friedman 1996). At the beginning of the twenty-first century, along the same line, several studies try to understand the paradox of nonvoting. Some focus on the costs and benefits of voting turnout and particularly institutional factors such as registration formalities, compulsory voting, electoral rules, party competition, and all the political factors that facilitate or make meaningless the act of voting (Powell 1980). Others see nonvoting as a deliberate strategic behavior, expressing political discontent with the electoral choices available (Subileau and Toinet 1993). Studies of electoral turnout detecting nonvoters on official voting records, such as the series conducted in France by the National Institute for Statistics and Economics Studies (INSEE), are more reliable than survey data because voters are reluctant to confess what is considered as an uncivic behavior. Yet they confirm the tight imbrication of the social and political factors leading to nonvoting.

The rising number of emerging or transitional democracies and the multiplication of cross-national data bases such as the Eurobarometers, the Central and Eastern Eurobarometers, the Latino Barometer and the Russia Barometer, provide fresh opportunities to test in different political cultures the validity of the theories and concepts elaborated in Western democracies. They could also be an occasion to explore what most studies leave aside, the noninstrumental meanings the act of voting can take. Although the political outcome of elections is essential, they also have ritual, expressive, and social functions, detailed by Edelman in his classic essay about The Symbolic Uses of Politics (University of Illinois Press, Urbana, 1964).

Bibliography:

1. Berelson B R, Lazarsfeld P F, Mac Phee W N 1954 Voting: A Study of Opinion Formation in a Presidential Campaign. Chicago University Press, Chicago
2. Boudon R 1997 Le ‘paradoxe du vote’ et la theorie de la rationalite. Revue Francaise de Sociologie 38(2): 217–27
3. Boy D, Mayer N 1997 L’electeur A Ses Raisons. Presses de Sciences Po, Paris
4. Butler D, Stokes D 1969 Political Change in Britain: Forces Shaping Electoral Choice. St. Martin’s, New York
5. Campbell A, Converse P, Miller W, Stokes D 1960 The American Voter. Wiley, New York
6. Crewe I, Denver D T (eds.) 1985 Electoral Change in Western Democracies. St. Martin’s, New York
7. Franklin M N, Mackie T T, Valen H et al. 1992 Electoral Change: Responses to Evolving Social and Attitudinal Structures in Western Countries. Cambridge University Press, Cambridge, MA
8. Friedman J (ed.) 1996 The Rational Choice Controversy. Economic Models of Politics Reconsidered. Yale University Press, New Haven, CT
9. Heath A, Curtice J, Jowell A et al. 1991 Understanding Political Change: The British Voter 1964–1987. Pergamon Press, Oxford, UK
10. Lautman J, Lecuyer B P (eds.) 1998 Paul Lazarsfeld (1901–1976). La Sociologie de Vienne a New York. L’Harmattan, Paris
11. Lazarsfeld P F, Berelson B R, Gaudet H 1944 The People’s Choice. Duell, Sloan, and Pearce, New York
12. Lipset S M, Rokkan S (eds.) 1967 Party Systems and Voter Alignments. Free Press, New York
13. Michelat G, Simon M 1977 Classe, Religion et Comportement Politique. Presses de la FNSP Editions Sociales, Paris
14. Norris P (ed.) 1998 Elections and Voting Behavior: New Challenges, New Perspectives. Aldershot, Brookfield, VT
15. Nie N H, Verba S, Petrocik J R 1976 The Changing American Voter. Harvard University Press, Cambridge, MA
16. Norpoth H, Lewis-Beck M S, Lafay J D (eds.) 1991 Economics and Politics: The Calculus of Support. University of Michigan Press, Ann Arbor, MI
17. Powell G, Bingham Jr 1980 Voting turnout in thirty democracies. In: Rose R. (ed.) Electoral Participation. Sage, Beverly Hills, CA
18. Robinson W S 1950 Ecological correlations and the behavior of individuals. American Sociological Review 15: 351–7
19. Rose R 1974 Electoral Behavior: A Comparative Handbook. Free Press, New York
20. Sniderman P M, Brody R A, Tetlock P E 1991 Reasoning and Choice. Explorations in Political Psychology. Cambridge University Press, Cambridge, UK
21. Subileau F, Toinet M F 1993 Les Chemins de L’abstention: Une Comparaison Franco-Americaine. La Decouverte, Paris
22. Topf R 1995 Electoral participation. In: Fuchs D, Klingemann H D (eds.) Citizens and the State (Beliefs in Government). Oxford University Press, Oxford, UK, Vol. 1
23. Verba S, Nie N H 1972 Participation in America: Political Democracy and Social Equality. Harper & Row, New York