Frontiers of Women and Politics Research Seminar

Wednesday 2nd September 1998
John F. Kennedy School of Government, Harvard University
Challenges for Empirical Researchers in the Subfield of Political Behavior
Susan J. Carroll, Rutgers University

Like women’s studies more generally, women and politics scholarship, in my view, should and must have as its goal the transformation of knowledge. Scholars of my generation have laid the groundwork by establishing women and politics as an active and (usually) legitimate area of research. We have made great strides in establishing the importance of gender as a variable in the empirical study of gender and political behavior (both elite and mass). Particularly in the area of electoral politics, we have greatly enhanced the discipline’s understanding of women’s relationship to the political arena. We have learned a great deal about similarities and differences in women’s and men’s political attitudes and behavior, the impediments to women’s political participation, and the contributions of women to political life.

Now it is up to those of you who are attending this research seminar, as the next generation of women and politics scholars, to move us further along toward the goal of transforming how political scientists think about and study politics. Unlike the scholars of my generation, many of you have been exposed to feminist scholarship as part of your formal training. You have a rich, imaginative body of interdisciplinary feminist work to draw upon in conceptualizing your work. My hopes are that in 20 years the empirical study of women and politics, as a result of your research, will have moved in new directions and raised and addressed questions that scholars of my generation cannot possibly foresee.

One major challenge to your generation of researchers is to do a far better job than my generation has done in bringing the insights of interdisciplinary feminist theory to bear on empirical research. We need to find ways to bridge the ever-increasing gap between feminist theory and the empirical study of political behavior. Perhaps because it is explicitly normative and consequently not constrained by (the illusion of) objectivity in the way that our empirical work often is, the work in feminist theory has been especially imaginative and provocative. As empirical researchers we can benefit from the more advanced state of feminist theorizing by drawing upon feminist theory in forging new frameworks for our own work which combine disciplinary and feminist interdisciplinary insights.

In general, work on gender and political behavior is undertheorized from a feminist perspective. As feminist scholars, we need to be especially clear in our research about our underlying theoretical models and assumptions regarding gender. One of the common practices that irritates me the most (very prevalent, for example, in analyses of the gender gap) is when researchers control for variables such as income, education, party identification, and marital status and then (and only then) attribute whatever variance is left to gender. This practice is irritating not only because it makes gender appear less important, but also because it represents the researcher’s failure to conceptualize the underlying model of the relationships among gender and other independent variables. By controlling for variables such as income or marital status in examining gender differences, researchers are implicitly assuming that income and marital status are causally prior to gender. But, of course, this makes no theoretical sense. If we, as feminist scholars, want nonfeminist scholars to consider the theoretical assumptions underlying their use of gender as a variable, then we have to set an example by being especially clear and explicit in our own work about how we think about gender and its relationship to other variables and/or identity categories (such as race, class, ethnicity).

One of the most serious challenges that contemporary feminist theory poses to empirical researchers is the need to rethink and complicate our use of the category "women." The category has come under attack in much of contemporary feminist theory (and feminist scholarship more generally) from both anti-foundationalists and women of color (not, by the way, mutually exclusive groupings). Their work has revealed the fictional and artificial nature of the category; yet, the category "women" continues to be employed uncritically in most contemporary empirical research. Many feminist scholars have argued very convincingly that race, class, and gender are mutually constitutive and that the effects of gender cannot be separated from the effects of race and class. From this perspective, the prevalent practice in social science research of statistically controlling for race, income, and education in examining gender differences in order to account for variation based on race and class misses the point entirely; this practice is based on erroneous theoretical assumptions about how these categories relate to each other. Feminist theory and feminist interdisciplinary scholarship more generally suggest that we need to unpack the category "women" in our empirical work, to rely less on analysis that deals with "women" as a group, and to focus more on researching the political behavior of specific subgroups of women (African-American, Christian Right, working-class white) in particular contexts. Much of contemporary feminist theory would suggest, as Shane Phelan does in her work, that we need to get more specific.

Just as we need to open up and critically examine the category "women," we also need to open up and expand the category "politics." Politics happens in locations other than voting booths and government institutions, but it would be difficult to tell that by reading the political science literature. Almost a decade ago Barbara Nelson pointed to the problem of the discipline’s "overemphasis on electoral democracy" and suggested that "The discipline needs a two-fold strategy, emphasizing that all political subjects are gendered while also giving special attention to those areas and concerns where women have traditionally put their political energy" ("Women and Knowledge in Political Science: Texts, Histories, and Epistemologies," Women & Politics, 9(2), 1989, p. 21). Scholars of women’s history through their research have helped to expand their discipline’s conception of politics by showing the ways in which women’s activities have had political consequences, and through our research we too need to demonstrate the importance of examining political behavior in non-electoral and non-governmental settings.

Another challenge for us as researchers is to move beyond the study of individual-level gender differences and devote more attention to the gendered context in which individual political behavior takes place. Gender underlies and is embedded in the very fabric of political institutions, symbols, images, and discourses. We cannot have a comprehensive understanding of the relationship between gender and political behavior until we have more research and analysis examining the ways gender helps to constitute, and also is constituted by, the contexts in which political behavior takes place.

Finally, all of these challenges suggest the need to expand not just the substantive scope of our research, but also our methods. Our understanding of gender and political behavior has been hampered by an over-reliance on survey research and quantitative analysis. As a scholar who has made ample use of both, I would hardly suggest abandoning either surveys or quantitative methods. Nevertheless, if we are to expand the substantive scope of our work to deal with the challenges I have described above, we must employ a greater diversity of methods, and I believe we all need to be trained in, and become comfortable with, a variety of methodological techniques. As researchers, we all should strive for methodological flexibility, choosing the methods best suited to our research questions.