Frontiers of Women and Politics Research Seminar

Wednesday 2nd September 1998
John F. Kennedy School of Government, Harvard University
Public Opinion and Political Behaviour
Pippa Norris

The issue of gender differences in political behaviour at mass and elite levels has been a recurrent theme in political science ever since the earliest studies of elections and political representation (Tingsten 1937; Duverger 1955; Lipset 1960). The classic puzzles always were, and remain, how far and why women and men differ in their levels of civic engagement and political preferences. An extensive body of research has evolved during the last fifty years, including the orthodox studies of voting behaviour in the 1960s, the critical literature deconstructing common assumptions in the early women's studies in the 1970s, and the burgeoning feminist scholarship of the 1980s and 1990s (for an excellent review see Carroll and Zerilli 1993). In the United States, and to a far lesser extent elsewhere, the empirical analysis of gender differences in political behaviour has moved from margin to mainstream.

Nevertheless despite the richness and diversity of the literature now available many questions remain unresolved. At the risk of serious over-simplification in this brief note we can highlight certain common methodological conventions and pervasive theoretical assumptions in the study of women and political behaviour which have restricted scholarship. We need to construct more innovative, distinctive and imaginative research designs for the next generation of the discipline. I would suggest that four related principles, in particular, should guide the future agenda, namely:

* to move the emphasis away from the conventional study of individual-level cross-sectional survey data of mass political behaviour towards a new institutionalism linking structures to processes and outcomes;

* to develop integrated multi-method and multi-level designs which allow us to combine cross-sectional survey data with alternative approaches including in-depth qualitative interviews, focus groups, experiments, content and textual analysis, participant observation, multi-wave panel studies, time-series and cross-cultural aggregate analysis;

* to systematically expand the comparative framework so that we can analyze how far theories of gender politics developed within a particular national context are sui generis and how far they can be generalized cross-nationally; and lastly,

*to strengthen the linkages between empirical research and normative issues of democratic citizenship and public policy.

These principles can be applied to any of the major areas of growing scholarship on gender differences in political behaviour- illustrated by the rich diversity of papers at the 1998 APSA annual meeting - such as the study of whether the media creates barriers or opportunities for women in politics; the issue of legislative recruitment and whether more women in office 'make a difference' to the policy agenda; or the role of women in grassroots mobilization, civic engagement and social capital within the community. But in a brief note let me illustrate these principles by focussing on one particular corner of the sub-field: theories of the gender gap in voting behaviour.

The Gender Gap in Voting Behaviour

In the United States the growth of the modern gender gap in voting behaviour and partisan alignments during the early 1980s stimulated a series of studies. Early work sought to document the extent of the gender gap and to explore the major factors leading to differences in women and men's voting behaviour, usually followed the traditional methodology of public opinion research (Smeal 1984; Klein and Ethel 1984; Erie and Rein Martin 1988; Miller and Shanks 1996; Seltzer, Newman, and Leighton 1997). The most common research strategy has been to use individual-level cross-sectional survey data within a particular election or series of elections, usually the NES. The central debate has focussed on the role of structural factors such as labor force participation (Manza and Brooks 1998) and the economic 'autonomy' of women (Carroll 1988), or cultural factors like support for feminist values and the women's movement (Conover 1988; Cook and Wilcox 1991) and attitudes towards the welfare state (Erie and Rein Martin 1988; Shapiro and Mahajan 1986; Seltzer et al. 1997). The most popular approach has commonly sought to 'explain' differences in the voting choice of women and men, which is reduced to analysing significant differences in the size of the gender gap between social groups, hence the 'labour force participation' gap (Manza and Brooks 1998), the 'marital' gap (Mitofsky and Plissner 1980), the 'autonomy' gap (Carroll 1988), the 'feminist' gap (Conover 1988), or even the popular journalistic myth of the 'soccer mom' gap).

Despite the considerable insights which this body of literature has generated, the methodological assumptions have proved capable of provided only limited answers to certain key puzzles. Think about the following. Why did more women than men in the United States support Republican presidential candidates in the 1950s, and the gap then subsequently first converge then reverse in the 1980s? Why did the modern switch in established partisan alignments in the US first become apparent in the early 1980s, and not earlier, or later? Once the gender gap in voting behaviour became apparent, why did it continue, and how did the parties respond as organizations to reinforce and consolidate their electoral base? Given the stability of the voting gap since 1980, why have has the gender gap usually been framed as a switch in the partisan allegiances of women, while other elections have been depicted as shifting votes among men (angry, white, or otherwise). And, most importantly, why did the gender gap in voting become an established feature on the American political landscape during the 1980s, and yet, at least until recently, this pattern appeared exceptional compared with other advanced industrialized societies? (Mayer and Smith 1985; deVaus and McAllister 1989; Jelen et al. 1994; Inglehart and Norris 1998)

These are serious puzzles which, if we think hard about them, throw a spanner into the works of many conventional and popular accounts. We cannot tackle these puzzles by another conventional and unreflective study of 'the gender gap in the 1998 mid-term U.S. elections' which tidily rearranges the individual-level variables in the equation, like the chairs on the Titanic. Instead we need more imaginative theoretical and methodological approaches which allow us to understand the process of gender realignment as one involving long-term secular changes in the structural and cultural aspects of women and men's lives and the way that the political system, particular parties and candidates, respond systemically to consolidate these changes. In American gender politics, 1980 was a critical election, which then stabilized and consolidated into a new long-term pattern with significant consequences for party competition, for the predominant policy agenda, for women's representation in office, and for the power of women and men at the ballot box. Without being able to develop the full argument here, (for details see Inglehart and Norris 1998) it can be suggested that we need to understand gender differences in the electorate as part of a long-term process which occurs over decades, driven by a process of value change which has gradually altered the culture of advanced industrialized societies. The gender gap in American politics is not just a change at the level of individual voters, instead developments in the electorate occur within the context of the system of party competition, which connects to the structure of American politics and the outcome of the policy process.

Given the available resources, how can we start to move beyond individual-level analysis of the gender gap in America? To expand the research agenda we need to adopt multi-method and multi-level approaches which locate voters within a broader context. Let me give some concrete examples. One fruitful approach would be the reanalysis of aggregate trends from public opinion polls which provide a rich source of time-series data to consider the dynamics of changes in gender politics since the early 1940s and the critical realignment in the early 1980s. Panel surveys allowing us to understand the process of electoral choice over successive waves, both outside and then within the campaign, are another important resource. To break the classic problems of causality which have bedevilled cross-sectional survey analysis, experimental designs are becoming increasingly popular again. These allow us to analyze how women and men respond to alternative messages - candidate ads, network news broadcasts, web pages, or presidential debates - within a controlled setting. We can monitor the response of women and men to different types of issues, candidates and formats. Moreover we need to monitor changes in party competition, using resources like the Comparative Manifesto Project to see how far parties have changed strategic programmatic appeals on issues prioritized by women and men. Another approach would be the use of in-depth qualitative interviews and focus groups which allow us to explore whether women and men structure their thoughts and beliefs about political issues, candidates and government in a similar, or different, fashion. Women and men may come to the same voting choices but for different reasons. Or that women and men take account of different considerations when evaluating candidates or prioritizing issues. Given what we know about sex differences in speech patterns, it seems wholly plausible that women and men talk about politics in distinctive ways. Qualitative approaches provide some of the richest ways to explore these avenues of research.

Methodological innovations need to be driven by theoretical concerns. Three are particularly important. First, the gender gap in the United States varies over time, in size and direction. A comprehensive theory needs to account for these variations. Many of the early studies were limited to static, cross-sectional survey data, focussing on one, or perhaps two successive, elections. Explanations of gender realignment need comparisons over time to explain why this voting shift occurred in the early 1980s, rather than before or after. Moreover any analysis of cross-sectional data needs to be sensitive to the context of particular campaigns - each with a particular mix of issues, candidates and events - rather than assuming that the gender gap is constant and invariable.

Second, studies need to be sensitive to social context and therefore significant diversity among groups of women, - by race, generation, and class, as well as between self-identified feminists and others, - rather than treating women as a single, homogeneous group.

Lastly, convincing theories of trends in American public opinion need to be evaluated within a cross-national context. This is particularly true of accounts which suggest 'essentialist' explanations based on the changing role of women as child-rearers and homemakers. The tendency for women to favour right-wing parties during the 1950s was regarded as a cross-cultural pattern (Duverger 1955). In contrast, during the 1990s the gender realignment in American politics has seen parallel developments in many advanced industrialized societies, although in developing societies the traditional (rightwing) gender gap remains (Inglehart and Norris 1998). Cross-cultural differences emphasize the role of systemic factors influencing the relationship between social cleavages and party loyalties, including differences in party systems and party competition, electoral systems, political culture, and the agenda and strength of the organised women's movement. A comparative perspective also highlights the need to be aware of 'top-down' explanations of gender realignment, based on changes in party/candidate strategies, policies and images, as well as 'bottom-up' explanations based on changes in men and women's attitudes, values and policy priorities.

Therefore gender realignment has now become an established part of American elections, taken for granted by commentators, journalists and politicians. The scholarly literature on women and politics has come a long way during the last forty years. Nevertheless, there is a great deal of work which remains to be done. We need to look to the next generation to bring fresh ways of understanding the complex relationship between gender, voting behavior and public opinion.

Carroll, Susan and Linda M.G.Zerilli. 1993. 'Feminist Challenges to Political Science'. In Political Science: The State of the Discipline II, ed. Ada Finifter. Washington, DC: APSA.

Conover, Pamela. 1988. "Feminists and the Gender Gap." Journal of Politics 50:985-1010.

DeVaus, David and Ian McAllister. 1989. "The Changing Politics of Women: Gender and Political Alignments in 11 Nations." European Journal of Political Research. 17: 241-262.

Duverger, Maurice. 1955. The Political Role of Women. Paris: UNESCO.

Erie, Stephen P. and Rein Martin. 1988. "Women and the Welfare State." The Politics of the Gender Gap, ed Carol M. Mueller. Newbury Pk, CA: Sage.

Inglehart, Ronald and Pippa Norris. 1998. Gender Gaps in Voting Behavior in Global Perspective. Paper for the Annual Meeting of the American Political Science Association, Boston 3-6 September 1998. See http:/www.ksg.harvard.edu/people/pnorris/paper.htm

Jelen, Ted G., Sue Thomas and Clyde Wilcox. 1994. "The Gender Gap in Comparative Perspective." European Journal of Political Research. 25: 171-186.

Klein, Ethel. 1984. Gender Politics. Cambridge, MA: Harvard University Press.

Lipset, Seymour M. 1960. Political Man; the Social Bases of Politics. Garden City, N.Y: Doubleday.

Manza, Jeff and Clem Brooks. 1998. 'The Gender Gap in U.S. Presidential Elections: When? Why? Implications?' American Journal of Sociology 103 (5): 1235-66.

Mayer, Lawrence and Roland E. Smith. 1985. 'Feminism and Religiosity: Female Electoral Behaviour in Western Europe'. In Sylia Bashekin Women and Politics in Western Europe. London: Frank Cass.

Miller, Warren E. and J. M. Shanks. 1996. The New American Voter. Cambridge, MA: Harvard University Press.

Mitofsky, Warren J. and Martin Plissner. 1980. Campaign '78 : a Comprehensive Political Handbook. New York: Arno Press.

Mueller, Carol M. The Politics of the Gender Gap, Newbury Pk, CA: Sage.

Seltzer, Richard A., Jody Newman, and Melissa V. Leighton. 1997. Sex As a Political Variable. Boulder, Co.: Lynne Reinner.

Shapiro, Robert and Harpreet Mahajan. 1986. "Gender Differences in Policy Preferences." Public Opinion Quarterly 50:42-61.

Smeal, Elleanor. 1984. Why and How Women Will Elect the Next President. New York: Harper & Row.

Tingsten, Herbert. 1937. Political Behaviour. London.