What Is CPG?

Any effort to delineate the boundaries of an area of inquiry necessarily includes some people and excludes others. The delineation of CPG offered here is meant to be as inclusive as possible while staying true to the analytical consequences of selecting gender, rather than women, as the defining feature of this field of study. As with most concepts in comparative politics, gender can be both a dependent and an independent variable. Thinking of gender as an independent and dependent variable helps to distinguish CPG scholarship from other work in the subfield and highlights important variations in how CPG scholars integrate gender into their work. Figure 1 presents a two-by-three box, with independent variables on one side and dependent variables on the other. Independent variables are gendered or ungendered, while dependent variables are divided into three categories: women, gendered, and ungendered. CPG encompasses all work in which gender is an independent or a dependent variable (denoted by light gray in the figure). Two boxes are gender-free. One is the box on the bottom left, comparing women (CW), and the other is the blank box on the bottom right, in which most comparative politics research falls. This mapping captures the possibility that gender may not be relevant to many political outcomes, but it also suggests that gender is pertinent to subjects in comparative politics that seemingly have nothing to do with women or gender.

Figure 1

Mapping a comparative politics of gender

Before addressing the ways in which gender figures into CPG work as a dependent and independent variable, a few words on what distinguishes CPG work from CW research are in order. A central goal of feminist research has been to make women visible, and most CPG scholars share a strong commitment to this goal. The exclusion of CW from CPG research rests on conceptual distinctions between women and gender (see Karen Beckwith's contribution for a definition of gender). Research on women does not necessarily incorporate gender as its subject of study or as an explanatory variable. CW research focuses on explaining (variations in) outcomes pertaining to women but uses ungendered explanatory variables to elucidate them. For example, CW scholars might explain women's political mobilization, (under)representation, or voting behavior; the success or failure of women's movements; or the adoption of policies that raise (or lower) women's status, as products of age, income, education, or party orientation (left/right).

CPG work also retains a commitment to many of the questions posed by CW scholars. The difference is that CPG research about women incorporates gendered independent variables. The vast majority of CPG work falls into this category. Explaining outcomes about women usually involves highlighting how the gendering of ideas, institutions, and structures contribute to these outcomes. For example, Elisabeth Friedman shows how “gendered political opportunity structures” explain the varying influence of women's movements in Venezuela, Lousie Chappell highlights how the gendering of institutions affects the success of women's groups in attaining their goals, and Lisa Baldez illuminates the role of gender frames in explaining women's political mobilization.Footnote ¹

Some CPG scholars incorporate gender not only as an independent variable but also as a dependent variable. The shift from women to gender in the dependent variable necessarily involves reframing some of the ways in which scholars pose questions. What makes a dependent variable “gendered” is not immediately obvious. Unlike the categories “man” and “woman,” gender does not refer to particular human subjects. In other words, women are subjects—living beings—who can act. Women are affected by public policy. Women can win political office. Women have rights. Gender has none of these qualities because it is an analytic and descriptive category that illuminates how the social organization of sexual difference, both material and discursive, structures the lived experiences of human beings, often in ways that produce unequal power relations, inequalities, and injustice. Luckily for comparativists, the contours of these gendered structures vary in a multitude of ways.

Moreover, the contours of gendered structures are extremely complex due in part to the intersectional nature of identity. Dara Strolovitch illustrates this point nicely with the example of women's concentration in low-wage jobs.Footnote ² Not only gender shapes the contours of low-wage work, but also class, race, and sexuality. Intersectionality necessarily implies variation among women within countries. In addition to the usual stable of cross-national and temporal variations that comparativists draw on, we can therefore also exploit intersectionality to inform our case selection. Comparativists thrive on explaining variation, so many exciting opportunities for selecting gendered dependent variables as the subject of our research exist. But how can gendered outcomes be operationalized as dependent variables?

When the dependent variable involves inequalities that can be expressed as proportions—for example, gender pay gaps or gender segregation in labor markets—gender is built into the dependent variable in an easily quantifiable way. Comparative work then seeks to explain variations in that dependent variable across different contexts (time, countries, sectors of the labor market, etc.). Of course, much of what interests us cannot be quantified so easily, so scholars have also focused on important qualitative differences pertaining to gender. Dramatic variations exist in the gendering of citizenship rights, rights in the family, and social policy.Footnote ³ One strategy is to explain how the roots of different “gender regimes” or sets of gendered rights and policies come about.Footnote ⁴ In some cases, CPG work produces typologies that capture significant gendered variations, and these ideal types then become—or could become—dependent variables. The shift in focus from explaining the impact on women or the achievement (or loss) of a right for women, to explaining how particular sets of gendered policies and rights came about or produce, reproduce, and alter gender relations, sets CPG work with gender as a dependent variable apart from that with women as the dependent variable.

CPG work with a gendered dependent variable can lead to different findings than research that focuses on women. My own study of the feminization of the manufacturing workforce in developing countries shows that shifting the question from “why do employers hire women?” to “why do employers make gendered hiring choices?” has important consequences.Footnote ⁵ Previous studies had focused almost exclusively on export-oriented, labor-intensive, and female-dominated industries, so it is unsurprising that scholars concluded that these market features, which impelled employers to lower labor costs, drove feminization. I selected industries that varied in their gendered employment patterns and market orientation and made gendered hiring choices the focus of the research. By doing so, the study cast doubt on the prevailing cheap labor argument dominant in the literature and showed the independent effects of gendered institutions and gendered discourses of work on men's and women's employment in manufacturing.

Some CPG research incorporates gender in the dependent variable but relies on ungendered independent variables to explain the outcome. Mounira Charrad, for example, analyzes why civil and family law in Algeria, Tunisia, and Morocco regulates gender relations in distinct ways.Footnote ⁶ The title indicates a focus on “women's rights” but in fact she arrays the cases based on whether individuals have rights independently of their gender or whether rights depend on their gendered location in the family. To explain this gendered outcome, Charrad utilizes an ungendered independent variable—the degree of state autonomy from tribal groupings. Similarly, Kimberly Morgan traces variant welfare state polices affecting the gender division of labor in the household to an ungendered variable, the secularization of politics.Footnote ⁷

Finally, some CPG work only incorporates gender as an independent variable. In other words, the outcome is ungendered—at least on the surface—but one or more of the independent variables is gendered. This research has primarily critiqued existing work for omitting gender as part of the explanation for the outcome in question. For example, in her review of the comparative historical work on democratization, Pamela Paxton demonstrates that while scholars usually include universal suffrage as part of their definition of democracy, their operationalization of democracy rests on universal male suffrage.Footnote ⁸ She shows that setting the bar at universal suffrage not only shifts dates for the transition to democracy but also has implications for the validity of theories of democratization. Scholars may have found “that women are ‘far less important’ in the known history of democratization,” she argues, because scholars have made them so by focusing primarily on male suffrage and class.Footnote ⁹ Building on Paxton's work, I argue further that even if scholars narrowed their claims to explaining the extension of suffrage to all men, their theories would still be problematic since the inclusion of (white) men is often deeply intertwined with continued exclusions based on gender and race.Footnote ¹⁰ In other words, a convincing comparative historical theory of democratization must consider the gendered dynamics of the extension of suffrage. Some of the feminist comparative welfare state literature does the same thing by suggesting that the historical evolution of welfare states cannot be understood without a consideration of the battles surrounding the state's regulation of families and the gendered ideas of policymakers and the courts.Footnote ¹¹

CPG work could push the envelope further by integrating masculinity and femininity into our analytical arsenal more frequently and by denaturalizing the hetero-normative assumptions that shape family policies. Beckwith's work on the Pittston coal miners' strike taps into the burgeoning literature from other disciplines on masculinity and shows, convincingly, that the transformation of the miners' collective action repertoire depended on deploying gender frames that reconciled the masculine identities of the miners with non-violent forms of action.Footnote ¹² The invocation of masculinity is rare in comparative politics; it is provocative because it demonstrates that the political action of men is gendered and thus illustrates that gender is not just about women and that we ignore it at our peril. A focus on masculinity could open up many exciting avenues of research—for example, to what extent is masculinity useful in understanding ethnic conflict, civil war, and men's political mobilization and voting behavior? Similarly, the gendered analysis of the family could sharpen its analytical tools by denaturalizing the hetero-normative assumptions that shape family policies. States regulate family structures not only by promoting and reinforcing norms about maleness and femaleness but also through hetero-normative ideas about what constitutes a “family.” Most studies of the family replicate the hetero-normative assumption that families are comprised of a man and a woman in a legally recognized union. States encourage, reward, and punish distinct family forms; incorporating the intersectional character of how states define legitimate family structures would greatly enrich our research agenda and our understanding of how and why states intervene in family formation.

CPG scholars continue to place enormous emphasis on women as a subject of research. By highlighting gender, CPG promises to come up with more complete and more nuanced understandings of women and politics. But CPG includes a much wider array of subjects than CW research. CPG research promises to shed light not only on women but also on men and to offer more persuasive accounts of gender inequality and variations in gendered outcomes. CPG scholars have shown that gender matters for explaining political phenomena that seem to have nothing directly to do with women or gender, which suggests that gender could be an important overlooked variable in many areas of research.

CPG and Comparative Politics

The comparative subfield can be categorized in many ways—by the dependent variable studied (e.g., a policy outcome, democratization), the theoretical approach or method (e.g., historical institutionalism, rational choice), or geographic area. CPG scholars study a variety of dependent variables, adopt divergent theoretical approaches, and study different parts of the world. The only unifying element among us is that we integrate gender in our work as independent or dependent variables. CPG research on women's movements has little in common with CPG work on gendered labor markets and will consequently be more theoretically engaged with the social movement literature than with CPG work that studies different subjects. The wide dispersion in the subjects that CPG scholars study has important implications for how we are integrated in the subfield.

Figure 2 presents a Venn diagram that illustrates CPG's relationship to comparative politics. It identifies a non-exhaustive set of subjects studied in comparative politics, picking those where most CPG research is concentrated. The overlap between the CPG and social movement circles in the diagram indicates that some but not all CPG scholars work on social movements, and vice-versa. The area with no overlap includes CPG scholars working on subjects that have more difficulty engaging with existing work in the discipline because political scientists have neglected these subjects.

Figure 2

CPG and comparative politics

Although CPG work has flourished over the last decade, CPG scholars express almost universal dissatisfaction about our isolation in comparative politics. While often successful in our careers, we lament that our work is rarely assigned or cited by scholars who do not work on gender, even when they are engaged in similar theoretical debates. We have also failed to convince most comparativists that gender is a critical category of political analysis. The best indication of this is that CPG work is rarely defined as “core.” A survey of syllabi from top political science programs revealed only three that assigned CPG work.Footnote ¹³ The situation is only slightly better in specialized seminars. In spite of the rich vein of CPG work on democratization, seminars on comparative democratization often assign no CPG readings and those that do usually assign just one.

A variety of mutually reinforcing explanations exist for the segregation of CPG work from the rest of comparative politics. The first is institutional. As scholars in a marginalized area of study, we pursued the path of creating independent institutions in the discipline to support our work. At APSA, the Women and Politics subsection provided a safe space for CPG scholars, and journals such as Women & Politics, Politics & Gender, and The International Journal of Feminist Politics provided venues where CPG scholars were more likely to have their work reviewed by sympathetic scholars. A consequence of this institutionalization, however, is that CPG is now seen as a freestanding area of study rather than one that overlaps with areas of scholarship dominated by non-gendered work. This sense of CPG being a separate field of study has made it possible for scholars who do not work on gender to ignore it without repercussion. As Lisa Baldez points out in her contribution, scholars of democratization writing review essays seem to be unaware of the existence of the vast body of literature on women, gender, and democratization. Of course, gender is not always pertinent to a subject of study, but CPG scholars have established the importance of gender in many subject areas, and this should be recognized as an integral part of these areas of study. In addition to participating in the Women and Politics section at APSA, CPG scholars should also be more active in comparative politics sections, working to increase our visibility by participating in section governance and nominating CPG work for awards.

The second set of reasons relates to the nature of much CPG work. For CPG scholars working in theoretical areas that political scientists have largely ignored, it is difficult to make the case that other scholars must engage with it. These scholars must do what feminists have always done—push the boundaries of what the field considers to be political. Such endeavors inevitably involve banging one's head up against these disciplinary walls. I have been told by reviewers, editors, and prominent but well-meaning people that my work was not really political science, and that perhaps I should consider a career in sociology. Perseverance does pay off and good work will usually find a home somewhere, but perhaps not in places where the disciplinary gatekeepers are most vigilant. Of course, this has consequences for the visibility of CPG scholars, which will be discussed in more depth later. For the huge slice of CPG work that is firmly situated in well-established theoretical debates in comparative politics, CPG scholars can do more to make our case to the subfield. While some CPG scholars have demonstrated how our research contributes to existing theoretical literatures in comparative politics, many do not. For example, much of the work on the participation of women's movements in democratic transitions has focused on explaining why women mobilized—an extremely important subject—but it rarely took the extra step of showing how this contributed to larger debates in the democratization literature. Scholars not working on gender will confidently define such work as purely CPG with no implications for their field of study unless we meticulously demonstrate how gender unsettles existing explanations and understandings of democratization.

Perhaps the only area where CPG scholars have succeeded in becoming an integral part of an existing theoretical body of work is the welfare state literature. Leading (male) scholars of the welfare state have acknowledged the importance of the gendered intervention and some have begun to incorporate gender into their analyses.Footnote ¹⁴ It took over a decade of intense research that produced dozens of books and hundreds of articles written by scholars in at least three disciplines—history, political science, and sociology. From the beginning, CPG scholars in this area were extremely self-conscious about engaging the existing work on welfare states head on and illustrating how the politics of welfare states could not be understood without a consideration of gender. But I am inclined to agree with Georgina Waylen that the example of the welfare state literature is the exception that proves the rule. Since the analytical link was through the family, the welfare state was an easier nut to crack than other areas of study focusing on the public sphere. And in spite of this success, none of the general seminars in comparative politics that assigned readings on the welfare state included a single required reading on gender and the welfare state.

This last point brings us to the third reason, which is the politics of defining “the canon.” My comments here will focus on the general field seminar, but they also apply to more specialized areas of study. Most professors feel an obligation to prepare our graduate students to become well-trained professionals in their field of study, and this means exposing them to foundational works. In a subfield as unwieldy as comparative politics, professors cannot include every important area of study. We draw from past experience—usually what we read in our field seminars as graduate students—and our own predilections to define the most important areas of study and to select canonical works. Moreover, we place new research into these already defined subject areas. So the first obstacle that CPG work faces is historical—it was never part of the canon.

Still, field seminar syllabi have changed somewhat over the years, so the question of how some work breaks into the canon must be addressed. For example, it is striking how well work on ethnicity has been incorporated into field seminars. The most likely explanation is that over the last decade a profusion of comparative work on ethnic conflict and identity has been published by top university presses and the most prestigious journals in comparative politics, that these scholars often have jobs at top-twenty political science departments, and that citation counts escalate as this area of scholarship expands. Becoming canonical is therefore in part an issue of visibility, and CPG scholars have not met with much success on this count.Footnote ¹⁵ Both the diversity of CPG research and the neglect of CPG work by non-CPG scholars produce low citation counts, even for impressive pieces of research. CPG scholars also have a small presence at top faculties in the discipline. Of the top twenty-five political science graduate programs, as ranked by US News and World Report, only six have active CPG scholars—either newly minted Ph.D.s who wrote CPG dissertations, or more senior faculty with multiple CPG publications (Cornell, Indiana, Minnesota, Washington University, Wisconsin, and Yale). Of the eight faculty members at these six institutions, five have tenure. CPG scholars also rarely publish their work in top venues in the subfield. Cambridge University Press is the top university press in comparative politics, but it publishes hardly any CPG scholarship. In 2005–2008, the Cambridge Studies in Comparative Politics series published nearly thirty new books and none had a CPG focus.Footnote ¹⁶ One included some analysis of gender inequalities in labor markets. Cambridge published about one hundred new comparative politics books outside of this series in the same period; almost four per cent were CPG. Comparative Political Studies published four CPG research articles in 2005–2008, World Politics two, and Comparative Politics three. This represents about two per cent of published research articles in CPS and World Politics, and five percent in Comparative Politics. As Tripp observes, CPG scholars are inclined to submit to journals perceived to be more sympathetic to gendered work and are unrepresented on the editorial boards of top journals.Footnote ¹⁷ More of us need to take the plunge and submit to these journals, and the Women & Politics section should make placing its members on the editorial boards of top journals a priority.

Overcoming this segregation requires action not only by CPG scholars but also by other scholars in comparative politics. In developing syllabi and reflecting upon our work, comparativists can take several steps to increase the visibility of gender. The first is to investigate whether scholarship exists in your field of study that shows how apparently ungendered variables have gendered consequences. For example, there is a mountain of research showing how the design of electoral systems has consequences for women's representation, yet this literature is rarely assigned in courses on political institutions. Similarly, CPG scholars have highlighted the mixed consequences of democratization for women's rights, which has obvious connections to debates about democratic quality. The second is to examine existing research in your subject and to assess whether gender might be an overlooked variable in explaining outcomes that are on the surface ungendered. The gendered critique of the comparative historical work on democratization, for example, has made a strong case for how gender is an important overlooked variable in explaining past episodes of democratization. CPG scholars of welfare states have also argued that gender is a critical part of the explanation for some features of welfare states. The third is to assess whether gendered assumptions that underlie a particular body of research lead to fundamental misconceptions or to overlooking a vital structuring element of variation. The gender and the welfare state literature shows persuasively that political scientists missed critical axes of differentiation among welfare states by ignoring gender. Finally, the normative importance of a subject should also be considered. Explaining different forms of gender inequality has not been a primary area of research in comparative politics, but such questions are profoundly political and of critical normative value. To reach a CPG that is fully integrated into comparative politics requires more of us to value research on gender and to set aside concerns about the “canonical” status of what we assign to our students and reference in our work.

The call for a deeper integration of gender into comparative politics, however, is not a plea to “add gender and stir.” Feminists and gender-conscious scholars have shown the disruptive potential of gendered analyses for an array of subjects, both inside and outside of political science. Should gendered work gain more visibility in the discipline, some tension will inevitably arise between scholars adopting a more critical feminist perspective and those that see gender as just another field of study or as an independent variable that can be added into existing theoretical frameworks. The debates arising out of such tensions, however, promise to make both us and the rest of the discipline better scholars, and would be an improvement over the current situation—one in which we have fought for and obtained space for our work but in which the rest of the subfield ignores us.