The conclusions are relevant for publishers: ‘do you want to aim for tradition, reflection, or emancipation’? Do we see ourselves reflecting today’s situation of women and minorities, or do we see ourselves as catalysts of change, as inspiration for a society that is more diverse and more inclusive, and that will install a sense of belonging for all, irrespective of gender, ethnicity, sexuality and the other lenses of Diversity & Inclusion?

This also links to a broader discussion taking place at the Educational Publishers Forum addressing the value of (educational) publishing. Educational publishers stand for local solutions, i.e. have a local industry that is publishing text books that represent local society – in terms of gender, but also other lenses of D&I (here is a great example from Canada).

Over to Prof. Mesman:



In 2019 the director of the trade organization of Dutch educational publishers, expressed interest in connecting to my research program on gender and ethnic socialization. My team and I had already conducted a small-scale study on gender and ethnic representation in children’s books and expanding this research into schoolbooks was a great opportunity for us to broaden our research horizon, and contribute to publishers’ insights about these issues. Research shows that exposure to underrepresentation and stereotyping can lead to educational and vocational pathways that are based on gender or ethnicity, rather than on individual talents and interests. Such mechanisms go against nationally formulated goals of education.



After several meetings, we decided to start with the analysis of schoolbooks for math and Dutch for the first year of secondary school, across all educational levels, which amounted to 33 books (and some additional digital materials). Math and Dutch are both mandatory subjects and are stereotypically associated with being more masculine (math) or more feminine (Dutch). The first year was chosen as a time when students begin to get a sense of where their talents lie in classic high-school subjects, which might put them on specific paths when it comes to deciding which subjects and studies to select later on. The publishers associated with the trade organization all expressed their willingness to share their books with us for analysis.



Our team trained a large group of students to code every single character in each of the books, note their gender, ethnicity, and activities or profession. This led to a set of about 13,000 characters for analysis. The main results were as follows:

  • There was systematic underrepresentation of female characters, ranging from 30.2% to 50.1% across the 33 books, with an average of 41%.
  • Similarly, characters of color (with a non-Western/non-White background) were underrepresented in the text of the books, with a range of 1.9% to 17.6%, and an average of 9%, compared to national Dutch statistics reporting 13.4% of people in the Netherlands belonging to this group.
  • When only looking at the pictures in the books, ethnic representation was on average quite close to the national statistics, and in several books clearly exceeded them.

Regarding the activities and professions of the characters, we noted some evidence of stereotyping:

  • Among characters with a profession, women and people of color were underrepresented, even when taking their already lower total representation into account as a baseline.
  • For women, underrepresentation was also found among characters who were scientists, professional athletes, celebrities, and among characters performing technical activities.
  • Women were overrepresented among characters performing household tasks, and those with a (grand)parental role.
  • For persons of color, the average societal status of their professions was significantly lower than the average profession status of white characters.
  • Overrepresentation of persons of color was found among characters who were professional athletes and celebrities.
  • There was no or little gender/ethnic stereotyping regarding intellectual activities.
  • Finally, representation of LGBT+ characters was completely absent in all books.



The publication of the reports was covered in many newspapers, on many websites, and in several radio and television programs. Of course, it is great if your research gets a lot of media attention, but the downside is that the general public also gets to respond, and it is mostly the ones that do not like what they read or hear who make themselves heard. I collected thousands of responses to the media coverage about the project.  The majority of those were negative.

A lot of people found this type of research to be a waste of money and suggested that I should do something more useful with my time. Many negative responses were also due to misunderstanding about the research because of selective reporting by the media. There were also suggestions that I was trying to dictate the content of schoolbooks, whereas all I did was uncover what was in them right now and have tried to be as neutral as possible in all my dealings with the media. However, complete neutrality is of course impossible. The simple fact that I conducted this research means that I think this is an important topic, even if I am reporting on it in a relatively neutral manner. This also means that people who find the entire topics of gender and ethnicity a waste of time, will never like this type of research, regardless of how it is presented.

The most amusing responses were those in which the readers assumed I was a man (referring to the professor who did the research as ‘he’ or ‘him’), and one in which the reader assumed that this male professor was probably married to a woman of color. A great example of the way in which stereotypes influence people’s perceptions: because often all we see are white male professors, people are unlikely to realize that there are also female professors of color.



Our findings were also shared with each of the publishers involved in the study, with tailored reports about their own specific materials. Their responses have been very positive across the board. What was most striking is that the publishers were unaware of these patterns. They appreciated our efforts to show them what was actually in their books when it comes to gender and ethnicity, even though they did not consciously put it there, and even though they were often unpleasantly surprised. This is also the main contribution of the project in my view: showing publishers that unconscious biases play a role when there are no explicit policies or checks in place, and they were all very much receptive to this message.



Of course, researchers do not dictate what needs to be in schoolbooks. This is why each of my presentations to publishers ended with the key question for them to answer: do you want to aim for tradition, reflection, or emancipation? Tradition was not a serious option for any publisher. Most did like the idea of reflecting society in terms of representation: roughly 50% women in books, and a representation of characters of color that is similar to national statistics. However, when reflection is the goal, underrepresentation of women among scientists in the books, or lower-status jobs for persons of color would also be justified, because this reflects the actual situation in our society. If publishers want to contribute to emancipation that goes beyond the current situation, character representation might actually be different than a reflection of our society. It is up to the publishers to have these challenging discussions and formulate their explicit vision to guide the contents of the next editions of their books.