As Lowi (Reference Lowi1992) observed more than two decades ago, political science often has shifted its focus in tandem with changes in the political and social environment. Just as gender equality has slowly but tangibly increased in recent decades, so has the study of gender politics in American political science (McClerking and Philpot Reference McClerking and Philpot2008; Tolleson-Rinehart and Carroll Reference Tolleson-Rinehart and Carroll2006). At the same time, however, as Parenti (Reference Parenti2006) observed, the discipline often has proved rigid against change and diversity of thought. The stunted incorporation of scholars belonging to marginalized groups, including women, constitutes one potential symptom of such rigidity—and, if widespread, such bias may pose a threat to the study of politics. If political scientists are to study politics genuinely and comprehensively, their ranks should represent—descriptively and substantively—the diverse groups and voices that comprise the polity at large. This article investigates gender representation within a key segment of the American Politics literature: the readings assigned in introductory graduate seminars.
Imbalances in representation within the discipline have been documented previously. Although the number and proportion of female political science faculty have increased in recent decades (APSA 2011; Tolleson-Rinehart and Carroll Reference Tolleson-Rinehart and Carroll2006), various forms of bias in scholarship have persisted with regard to gender (Breuning and Sanders Reference Breuning and Sanders2007; Mathews and Anderson Reference Mathews and Anderson2001; Young Reference Young1995). Most recently, Teele and Thelen (Reference Teele and Thelen2017) observed that women’s representation in 10 prominent political science journals lags behind the proportion of women in the profession. Given these circumstances—and a host of potential issues enumerated by Beaulieu et al. (Reference Beaulieu, Boydstun, Brown, Dionne, Gillespie and Klar2017)—gender bias in political science graduate instruction seems likely.
Presentation of the politics of marginalized groups in course content may prove just as essential to the health of the discipline. Wahlke (Reference Wahlke1991) emphasized this fact in his report on the political science profession, recommending that the politics of diverse groups be mainstreamed rather than confined to specialized courses. Echoing these concerns 20 years later, the Report of the Task Force on Political Science in the 21st Century called for an increased focus on “the study and teaching of issues related to diversity and inclusion” (APSA 2011, 53). The mainstreaming of these topics is likely to play a large role in attracting members of marginalized groups to the discipline (Cassese, Bos, and Duncan Reference Cassese, Bos and Duncan2012); however, bringing the politics of women into the core curriculum continues to be a work in progress.
We further elucidate these issues, specifically at the graduate level in the American Politics subfield, through a unique approach. Using an original dataset, constructed from 56 graduate syllabi provided by highly-ranked political science departments (Diament, Howat, and Lacombe Reference Diament, Howat and Lacombe2017), we examined the status of gender representation in the American Politics “canon” as presented to current graduate students. We found substantial gaps (and variation) in the representation of women in terms of both authorship (i.e., the proportion of assigned works with female authors) and the substantive content of the readings.
Both forms of representation may have far-reaching consequences. As we argued previously (Diament, Howat, and Lacombe Reference Diament, Howat and Lacombe2017), graduate coursework may prove especially influential in the “reproduction of power” in the field because doctoral students are explicitly pursuing careers in political science scholarship. These students will likely retain the lessons gleaned from their early courses for years, if not decades, to come. Among these lessons, the content of assigned readings communicates beliefs about what politics are worth studying—including (or not) the politics of marginalized groups. Simultaneously, authors’ own characteristics and experiences influence how they study politics, potentially leading to biases in their research approaches and conclusions. Finally, the characteristics of assigned authors may send an implicit message, likely unintended, about who is considered capable of studying politics. In these ways, introductory course readings may leave a lasting impression on new graduate students.
Among these lessons, the content of assigned readings communicates beliefs about what politics are worth studying—including (or not) the politics of marginalized groups.
Before presenting our analysis, we want to make clear what the purpose of this article is not. Although we believe our findings reveal consequential gender bias within the American Politics literature (at least as presented in introductory graduate syllabi), we make no attributions of blame for this state of affairs. Most important, we emphasize that we do not consider any representational biases that we uncovered to be the fault of the instructors who assembled the syllabi comprising our data. If an essential work on an important topic happens to have been written entirely by white men, for example, we cannot reasonably expect instructors to exclude it from their syllabi on that basis alone. Thus, rather than to place blame, our project wants to achieve (1) greater awareness of the gender bias that exists in the literature, so that it can be considered in discussions of these important works; and (2) greater motivation among scholars at all levels to work—as best they can—toward a more diverse and representative discipline.Footnote 1
DATA AND METHOD
In the fall of 2013, we requested syllabi from instructors of core American Politics courses—or core-course sequences—at each of US News and World Report’s top 75 political science departments.Footnote 2 We sent a total of 88 requests and ultimately received 63 syllabi—an impressive response rate of almost 72%. Of the 75 schools to which at least one request was sent, 57 (or 76%) provided at least one syllabus. In the case of schools with a multi-course sequence, only those for which we received a complete sequence were included in the analysis to avoid over-representing either behavior or institutions. We analyzed 44 core (single-course) syllabi and six two-course sequences, for a total of 56 syllabi.
We centralized syllabus content in a single spreadsheet with a line for each reading: 6,266 in total, with 6,259 in which the author’s gender could be ascertained. A given reading received a line for each syllabus on which it was assigned, effectively giving proportional weight to more widely assigned works. We then labeled each reading according to the topic it covered; all three of us had to agree on a topic label before it was applied to a reading. Each work was coded under only the one topic deemed most appropriate. Because some subject areas within American Politics tend to be more unified than others, we divided some topics into subtopics—for example, Identity Politics contains Gender, Intersectionality, LGBTQ, Race and Ethnicity, and Religion.
To assess gender representation in the core graduate canon, we first coded whether each work had at least one female author. This measure followed other assessments of female authorship in political science (Evans and Moulder Reference Evans and Moulder2011; Young Reference Young1995), and we expected it to produce a conservative estimate of gender bias because the presence of only one female author on a given work counts as much as a work whose authors are majority or entirely female. As a more stringent test, we additionally coded whether a work’s first author was female, also in keeping with other work (Breuning and Sanders Reference Breuning and Sanders2007; Evans and Moulder Reference Evans and Moulder2011; Mathews and Anderson Reference Mathews and Anderson2001; Young Reference Young1995). Taken together, these two operationalizations provided a clear picture of female-author representation in the American Politics canon.
GENDER REPRESENTATION IN THE CANON
We begin the analysis with simple percentages of works with female authors—overall and within each topic (table 1). These percentages do not paint an optimistic picture of gender representation in the subfield: only 18% of the works in our dataset reach even the weak standard of having at least one female author and only 11% have a female first author. To be sure, female representation appears strong within a small number of topics—for example, assigned works on Participation and Identity Politics boast 32% and 43%, respectively, with at least one female author. However, these percentages decrease to 15% and 31% with the stronger standard of female-first authorship. The proportion of works by at least one female author on the politics of gender (as well as intersectional works concerning multiple identities) exceeds 80%. However, these topics comprise a small proportion of the field as presented in introductory syllabi—Participation is 5%, Gender is a mere 1%, and Identity Politics overall is less than 8%. (For perspective on topic size, table 1 also notes the number of readings in each one from the total of 6,259.)
These percentages do not paint an optimistic picture of gender representation in the subfield: only 18% of the works in our dataset reach even the weak standard of having at least one female author and only 11% have a female first author.
Table 1 Works with Female Authors by Topic
Female authorship proved far less prevalent in the largest topics. Of all topics that comprised 5% or more of the assigned literature, only Courts, Identity Politics, Interest Groups, and Participation exceeded the overall percentage of works with a female author (18%). For most of these major topics, 11% to 14% of the readings had at least one female author, with the classic topic of Voting a notable outlier at 4%. Regarding female-first authorship, these major topics ranged from 3% to 14%. In summary, within the most voluminous American Politics topics—which also tended to be those most consistently covered in introductory courses (Diament, Howat, and Lacombe Reference Diament, Howat and Lacombe2017)—assigned readings skewed heavily toward works with only male authors or male first authors.Footnote 3
We checked the robustness of these findings in two ways. First, we calculated the proportion of female authors for each reading. Across all works, as well as coauthored works only, the average proportion of female authors was 12%—which, unsurprisingly, was lower than the 18% total calculated using the weaker standard of at least one female author. Second, we examined the proportion of female authorship among unique readings in our dataset. Of the 2,668 readings, 606 (23%) had at least one female author. Thus, when we did not give extra weight to frequently assigned readings, female representation increased but did not approach gender parity.
A similar pattern held when we confined our analysis to the most frequently assigned authors and works (Diament, Howat, and Lacombe Reference Diament, Howat and Lacombe2017). Of the top 10 author lists for each of the 15 major topics, seven contained at least one female author. However, within any given topic, we found no more than two women among the most-assigned authors. Among the top 10 most frequently assigned works on these topics, we found similar evidence: 10 topics included at least one work by at least one female author. However, except for Courts (which included three top works with a female author), none showed more than two.
Together, these metrics indicate a canon literature written largely by men. Of course, we must exercise caution in our judgment. The presence of women in academia has grown over time, and a number of classics in the field were published in a time when female political scientists were far less common (APSA 2011). Thus, although our data comprise only a one-time snapshot of the canon literature, we attempted to consider changes over time using the publication dates of the readings in our dataset.
Table 2 compares the proportion of works with female authors published before 1990 with those published in 1990 or later. We chose the cutoff date of 1990 because it demarcates the publication of the majority of assigned works in our dataset on the topic of Gender. This suggests that considerable growth in the study of gender politics occurred at that time—perhaps as a result of greater female involvement in the field in the 1970s and 1980s (Tolleson-Rinehart and Carroll Reference Tolleson-Rinehart and Carroll2006). It also coincides with increases in female authorship in top journals (Young Reference Young1995), women’s participation at the APSA Annual Meeting (Gruberg Reference Gruberg2006), and the beginning of an increase in the proportion of female political science faculty (APSA 2011). We considered this pre- and post-1990 division to be another conservative test of gender bias. After several decades as a discipline, political science certainly had ample opportunity to become more inclusive of female perspectives in the works considered most important for graduate instruction.
Table 2 Works with Female Authors by Topic: Before and After 1990
As table 2 illustrates, the degree to which the discipline seized that opportunity appears to vary substantially. The table shows the percentage of works with at least one female author—overall and within topics—published during each period, as well as the percentage-point increase and percentage growth between them.Footnote 4 For works published in 1990 or later, female authorship reached 23.6% overall—an increase of about 18 percentage points (or 356% growth) from 5.2% in the pre-1990 period. Some topics boasted a significant difference in female authorship between the two eras. Most prominently, only 3.1% of works in Participation published before 1990 had at least one female author, whereas among works published in 1990 or later, that proportion reached 39.6%. Interest Groups shows a difference almost as significant, increasing from 7.7% to 39.3%. Other sizable topics, including Bureaucracy and Campaigns and Elections, increased from no female representation to more than 20% with a female author. The only category that experienced a reduction was Gender, from 100% to 91.8% after 1990, which indicates that male political scientists also developed interests in studying gender in politics. However, some of the largest topics again lagged behind: Parties and Voting showed respective increases of 12 and 7 percentage points (64% and 32% growth, respectively); and Congress, although displaying 15-point and 813% increases, reached only 17% female authorship for works published in the latter period.
This evidence suggests a substantial gender bias among widely taught American Politics works, diminished but far from eliminated among works published in recent decades. Furthermore, in accordance with previously observed patterns in course content (Cassese and Bos Reference Cassese and Bos2013; Cassese, Bos, and Schneider Reference Cassese, Bos and Schneider2014; Olivo Reference Olivo2012; Tolleson-Rinehart and Carroll Reference Tolleson-Rinehart and Carroll2006), we found that explicit readings on the politics of gender—far from being mainstream—occupied only 1% of assigned works. Based on our data, gender politics seems far from becoming a central topic of political science inquiry.
Gender, of course, is only one of several identity categories. Course weeks about identity politics focus not only on gender but also on multiple racial groups, as well as other topics including religion, immigration, and LGBTQ. Of the six syllabi that included an Identity Politics week focusing on more than racial politics, only one included a week addressing gender politics (whereas 19 included a week on race). Given the time constraints inherent in a single introductory course (or even a two-course sequence), this attention to some forms of identity politics is somewhat encouraging. However, it appears that Wahlke’s (Reference Wahlke1991) recommendation that ethnic and gender politics be “mainstreamed” in political science curricula, rather than relegated to specialized courses, has yet to be broadly followed. As our discipline progresses, however, we do expect more introductory American Politics seminars to have weeks devoted to these topics. What was once a small albeit dedicated study of the African American experience in American politics has paved the way for more inclusive studies that address the politics of the nation’s diverse population.
[T]he underrepresentation of female authors suggests that the American Politics subfield must go a considerable distance before it can credibly claim to reflect the polity it studies.
CONCLUSION
Our analyses of introductory American Politics graduate syllabi revealed substantial shortfalls in the representation of women, with disproportionately low female authorship of “canon” works, even by lax standards. Additionally—and perhaps partially a result of biases in authorship—the study of gender politics lags in volume behind other topics. Although our examination of works published in more recent decades suggests that these biases have attenuated over time, it appears that male voices still dominate the study of American Politics.
As noted at the outset, the subfield’s core graduate curriculum may prove especially influential in reducing or perpetuating imbalances in representation. These courses provide many students’ first in-depth look at extant American Politics scholarship; what they read at this formative point in their career may shape their view of the field—as well as their own research agendas—for decades to come. This potential further underlines the importance of representing diverse groups within the American Politics canon in terms of both authorship and subject matter, and it makes our findings all the more troubling. As a simple first step toward rectifying these disparities, we echo Beaulieu et al.’s (Reference Beaulieu, Boydstun, Brown, Dionne, Gillespie and Klar2017) recommendation that instructors (of introductory as well as other courses) check their syllabi for inadvertent gender bias and, wherever possible, include works by female scholars on the topics of interest.
Our results provide some cause for optimism. Restricting our analysis to introductory graduate seminars to some extent may overstate the underrepresentation of topics related to identity in general and gender specifically, given that they often receive far greater attention in more specialized courses. Still, the underrepresentation of female authors suggests that the American Politics subfield must go a considerable distance before it can credibly claim to reflect the polity it studies. To that point, our examination of gender representation within graduate syllabi also provides a useful template for similar work. Investigating the inclusion of other underrepresented groups, such as racial minorities and LGBTQ individuals, presents its own unique challenges, but these efforts are necessary if we are to foster a more balanced and diverse discipline.
ACKNOWLEDGMENTS
We thank Jamie Druckman, Molly Patterson, and Nina Kasniunas, as well as participants at the 2016 Midwest Political Science Association Annual Meeting, for their valuable guidance on this project. We also thank instructors who graciously provided their syllabi and our anonymous reviewers for helpful suggestions.
