Many organizations think that being inclusive is simply an issue of hiring members of underrepresented groups. But people hired on that basis are not going to stick around if they feel isolated or that they stand out, or that they’ve been hired just because they’re “diverse”. One very important element in inclusivity is representation; people want to see others like them, and also want to see those other people being respected and valued.

Part of a new study by a group of US researchers looks at the effects of representation in a place that isn’t often examined: the readings that students are assigned in university courses. There has been plenty of discussion over the past few decades about “the canon” in various academic fields and what determines whether a work is a “classic”  that all students should be familiar with. The researchers investigated whether the gender balance of assigned readings in a political-science course – the number of readings written by men and the number of readings written by women – affected female and male students’ self-efficacy: their confidence in their own ability to succeed. The study looked at readings in Ph.D.-level courses, on the basis that students in these programs are being trained to be professors and researchers. If students don’t see themselves represented in their future profession, they may decide to seek other opportunities.

The researchers recruited 300 political science Ph.D. students from across the US, and gave each participant one of four different course outlines for one fictional course (‘New Research Methods in Social Sciences’). The reading list was identical in each of the four versions of the outline, but the first names of the authors of the assigned readings were changed, so that the percentage of readings by female authors varied from 10% to 30%. Some of the assigned readings were real articles or books, so that the participants would be less likely to suspect that the course did not actually exist. The participants were also asked general questions about their attitudes toward “curricular diversity” and to rate how strongly they felt about their own potential in their Ph.D. program.

The results of the data analysis were surprising. When there was an increase in the number of assigned readings by female authors, it had no effect on female students’ self-efficacy, but it reduced male students’ self-efficacy. In other words, more readings by female authors made male students feel they were less likely to be academically successful.

Students who were less supportive of “curricular diversity” – those who gave lower ratings to items like “Professors should pay attention to gender balance when they write syllabi [course outlines]” – also displayed lower self-efficacy than students who agreed more strongly with such statements. The participants’ views on “curricular diversity” had more of an impact on their self-efficacy than the number of female authors in the course outline they received.

The researchers identify two important themes in these findings. One is that the people “who [benefit] when Ph.D. programs adopt diversity initiatives [are] women and students who already believe that diversity is important”. The other is the presence of “male backlash”: “some male students, notably ones with low support for curricular diversity, had lower expectations about academic success when confronted with increased gender diversity in syllabi”.

This study is based in a specific environment, and only looks at one aspect of that environment; a student’s experience in a course is shaped by a lot more than the assigned readings. And in a real-life course, a student’s attitudes could be changed by things like discussions in the classroom with the instructor and other students.

But the results of the study are still interesting to think about in the context of other kinds of organizations. If, as the researchers suggest, gender equality is seen by some as a “zero-sum game” – more importance for women = less importance for men – that has significant implications for workplaces.

We might be tempted to think that with more women in organizational leadership positions, workplace attitudes toward gender equality have changed. But many studies have indicated that isn’t true – that around 30% of men see gender equality in the workplace as reducing opportunities for them.  “Toxic masculinity” can also negatively affect gender equality in workplaces, and most workplace diversity training is stunningly ineffective in discouraging that kind of behaviour.

Changing those attitudes and behaviours is a huge challenge, but the results of this study suggest two places to begin. One is realizing that representation is important, and that acknowledging backlash is also important. If some students or employees feel threatened by representations of diversity, emphasizing that inclusion means including everyone may help to counteract the zero-sum perceptions. (However, that emphasis will only be effective  if the organization doesn’t just say that everyone is included, but actively works to ensure that they are.) The other is that representation matters everywhere. The gender balance of authors on the reading list of a single university course may not seem to be important on its own, but if small things are not identified and addressed, big things will never change.