DATA
To assess the distribution of book authorship by gender, we started with what is plausibly the universe of book publishers in political science: all presses that hosted a booth at the 2016 APSA Annual Meeting, plus any other publisher listed in Garand and Giles (Reference Garand and Giles 2011). We then compiled lists of books that these publishers classified as "political science" between 2004 and 2015. To do so, we contacted each publisher and requested a list of said books. If we did not receive a response, we compiled the books by scraping publishers' websites. This generated a list of 25,898 books from 34 university and 22 commercial presses.
University presses published about 32% (8,250) of all books during this period and commercial presses published the remainder (17,648). We assigned a subfield to each title—American politics, international relations (IR), comparative politics, political theory, or methodology—by making educated guesses using titles, abstracts, and/or information that publishers provided about the content of each book. Given the volume of books, we coded subfields only for university-press books.
We then classified authors' gender using the "Gender Balance Assessment Tool" created by Jane Lawrence Sumner at the University of Minnesota (Sumner Reference Sumner 2018). This program assigns a gender to names probabilistically. Using online searches of authors' websites, we manually checked all names that the program gave lower than 90% probability and assigned the likeliest gender identity using photographs and gendered pronouns in descriptions of authors' work.
The following discussion reports figures for both commercial and academic presses. However, most of this article focuses on university-press books, which are more important for scholarly advancement. Doing so also provides a reasonable "apples-to-apples" comparison to results that Teele and Thelen (Reference Teele and Thelen 2017) presented for journal articles: similar to articles but unlike most commercial-press books, university-press books are peer reviewed.
RESULTS: GENDER PUBLICATION GAPS
Combining university and commercial presses, books with only male authors comprise 69.1% of the total. Books with only women among authors comprise only 19.6%, and those with mixed-gender author teams comprise 11.2%. Table 1 divides the data into university and commercial presses, according to one of five categories: man solo author, woman solo author, all-men coauthorship, all-women coauthorship, and mixed-gender coauthorship. Combining the first two categories, we found little difference between commercial and university presses—one or more men write 68.5% versus 70.6% of all books. Women are more likely to publish a book on their own with a university press than a commercial press (20.7% versus 15.8%) but are less likely to publish a book written with another woman with a university press (1.5% versus 2.7%). The most significant difference appears in the last category: women are only about half as likely to write a book with at least one man at a university press than at a commercial press. These figures suggest that women are even less well integrated into collaborative writing networks for university-press books than they are for journal articles, regardless of whether they are working with other women or with men (Teele and Thelen Reference Teele and Thelen 2017).
About 30% of university-press books published from 2004 to 2015 had at least one woman's name on the byline. Does that make women underrepresented? Determining a benchmark is not straightforward. On the one hand, the proportion of women among book authors is higher than the share of articles with women as first authors in 10 top journals between 2000 and 2015 (26.7%). Yet, on the other hand, the proportion is still significantly lower than the two benchmarks that Teele and Thelen (Reference Teele and Thelen 2017) suggested: women's share of political science PhDs granted (38% in 2016) and women's share of APSA members (38% in 2015).
It is worth noting that the gender gap in book publishing has narrowed somewhat over time: overall, the total share of women among authors rose by one third between 2004 and 2015, from 21% to almost 28%. However, this is still less than the proportion of women in the discipline. Figure 1 tracks the time trends of the five authorship categories for university presses. In the early 2000s, books by one man accounted for 63% of the total; however, by 2015, that proportion had declined to 54%. Most of the increase in the share of women among authors came from growth in the proportion of women's solo publications (19% to 23%) and participation in mixed-gender teams (5% to 8%).
This minor increase in the proportion of mixed-gender collaborations is the only evidence of an increase in women's coauthorship of books. Yet, in stark contrast to patterns for articles described by Teele and Thelen (Reference Teele and Thelen 2017), the proportion of books coauthored by either men-only or women-only teams has barely increased in recent years. Although the proportion of mixed-gender collaborations has increased to 8% of the total, it is still only one third the level observed for mixed-gender coauthorship in top journals (24.5%). In short, as solo book authors, women have made significant advances on their own (20% of the book total versus only 15% of the journal-article total), but not as collaborators—either with women or with men.
…as solo book authors, women have made significant advances on their own, but not as collaborators—either with women or with men.
The shrinking gender publication gap over time raises the question of whether women are more or less underrepresented in different ranks. From the 5,631 university-press books published between 2004 and 2012 (i.e., six years before our research because we also collected information on citations gaps), we drew a random sample of 1,170 titles: 1,000 from the full dataset, balanced on authorship pattern and subfield, and an additional 170 randomly oversampled from coed and all-women teams, which have relatively fewer books. This provided a sample of 21% of the total. We kept all books authored by academic political scientists (about 6% were not) but eliminated reprints, including those written by long-deceased scholars (137) and deleting paperback editions of previously published books (65 more), resulting in 613 books.
Using biographical information on each book's back cover (cross-checked with institutional or personal web pages), we collected information on the rank of first authors at the time of the book's publication. We found that the gender publication gap, when compared with the proportions of women at each rank, is wider in more senior ranks: at the assistant-professor level, women are 38% of first authors and 43% of APSA members; at the associate-professor level, 35% of first authors and 40% of APSA members; and at the full-professor level, only 16% of first authors and 28.6% of APSA members.
Subfield Patterns
Table 2 addresses the question of whether gender publication gaps vary across subfields. Baseline information on women's share in subfields is from field-specific APSA memberships listed in Hardt et al. (Reference Hardt, Kim, Smith and Meister 2019). IR has the smallest share of women among book authors (21%), whereas comparative politics has the largest (28%). Yet, in both groups, there is an approximate 14-point gap in women's book publications vis-à-vis their presence in the subfield. There are relatively few methodology books; within that subfield, women authors are 21% of the total. This means that in methodology, women comprise a larger proportion of book authors than APSA members in that subfield; however, we should not overstate this finding given the small number of methods books. In all other subfields, women are underrepresented among authors relative to their numbers in those fields.
Figure 2 explores the proportion of women among authors by subfield over time. We excluded methodology because low numbers mean that the publication of a single book by a woman in a given year throws off the proportions. The figure shows that in political theory, the proportion of women among authors has remained constant. In contrast, the proportion of women among authors has increased in American politics (20% to 25%), IR (18% to 25%), and—most notably—in comparative politics (22% to 32%). However, despite these gains, the proportion of women among authors remains lower than the contemporary proportion of women in each subfield.
Explaining the Gender Book-Publication Gap
As in journal articles, a gender gap in book publishing exists in political science: women publish relatively fewer books than their numbers in the discipline might predict. Moreover, although the gender book-publication gap has narrowed over time, it remains larger than for articles in top journals by about 5%. Most of women's relative gains have come as solo authors, but they remain far less likely to collaborate on books than on articles. Within subfields, the gender book-publication gap is most significant and has remained unchanged in political theory. It has narrowed somewhat over time in IR and American politics and mostly in comparative politics.
To better understand the roots of the book-publication gender gap, we conducted informal telephone and Internet audio interviews with political science acquisitions editors from one commercial press and seven university presses (three men and five women). These professionals had at least 175 combined years of editorial experience. Our discussions with them suggested several reasons why a book gender gap may exist.
First, in contrast to the attention focused on the gender balance of article publishing in recent years, which in turn raised attention among journal editors (Brown and Samuels Reference Brown and Samuels 2018), our interviews revealed that the gender balance of book authors is not yet a topic of wide concern among book editors. One editor affirmed that "Nobody ever raises the issue; I don't even know who I'd raise the issue with. We just assume we are doing alright." Another stated that "There has never been a discussion in five years at [our press] about finding more women. Maybe we need to do more of that, be more proactive." Even at a press where the issue of gender balance has been discussed in recent years, the editor could not confirm that anything had actually changed: "I don't really know—I think my author list is pretty balanced…[but] it's possible that there are patterns I just don't pick up."
Second, even for editors who stated that gender balance is a concern at their press, book editors lack the type of information that journal editors possess that could help identify a gender gap. It is not only that books are submitted and evaluated differently from journal articles. The issue is that every editor we spoke to confirmed that presses keep no records at all of information relevant to the editorial process—including the number of submissions, rejections, and acceptances or any type of author-demographic information. Given this, book editors could provide only rough estimates of the number of submissions per year or the gender balance of published books. (Two had counted the number of books they had published in the previous year and the proportion of women authors immediately before the interview.) Given the relative lack of attention to the issue and their lack of information, editors must rely on guesswork to track the balance by gender (or rank or ethnicity) of published books—whether the press receives a few dozen or several hundred submissions per year.
Third, book editors' professional incentives also may slow responses to demands for attention to a gender gap in publishing. The editorial process for book publication is not only less transparent than the (already opaque) article publication process; book editors also face relatively fewer incentives to improve transparency. They are not nominated and hired by teams of professional colleagues as are many journal editors, they do not have to publish annual editors' reports, and they do not answer to editorial boards of professional colleagues. Their professional incentives also differ from those of journal editors, who typically serve relatively brief terms and then return to their academic positions in which they must continue to produce journal articles and book manuscripts.
Fourth, because they are not colleagues, book editors do not participate as directly in APSA or other professional association task forces or in other activities focused on bringing attention and conducting research on issues related to women's professional advancement.
Factors specific to the book-publishing process also may contribute to the gender gap in book publications in political science.