Pia Schober; Silke Büchau; Marie-Fleur Philipp

University of Tübingen

This study explores how gender equality and recognition of gender diversity at institutional and normative levels of societies relate to the prevalence of more versus less polarised gender self-concepts. Responding to widespread criticism of binary measures of gender, we conceptualise gender as a multidimensional social structure and draw on gradational measures of femininity and masculinity in the European Social Survey (2023, Round 11). Gradational measures of femininity and masculinity grant individuals more flexibility in selecting an accurate and comprehensive self-definition, while also allowing for more variation for quantitative analyses.

Our analysis includes 40,156 individuals nested in 194 European regions (across 24 countries). We construct a continuous gender polarisation score and a binary variable of non-conforming gender identification, which comprises trans and non-binary individuals as well as cisgender women who feel more masculine than feminine and cisgender men who feel more feminine than masculine.

On average gender polarisation is lowest in the Nordic countries and highest in Eastern European countries. The prevalence of non-conforming gender identification is highest in the Nordic countries (>12%) and lowest in Poland, Germany and Greece (3-5%).

We explain respondents’ gradational gender identity with individual-level variables as well as regional-level gender norms and institutions in the European regions. Multilevel regression models show that parental college education and respondents’ higher education relate to a less polarised gender identification. Less polarised and non-conforming gender identifications are more prevalent in regions with a higher share of women in political representation and in regions with more egalitarian gender norms.

Gunnar Otte; Gianluca Thorn

Johannes Gutenberg University Mainz

Although gender comprises various facets, hardly any measuring instruments have been established in survey research beyond the dichotomous measurement of “male” and “female”. In current research, the focus is on how third genders can be surveyed, but this leaves the majority of respondents unaffected. Against this background, Westbrook and Saperstein suggested that gender identity should be measured in a gradational way, i.e. people of all genders should be asked how masculine and feminine they consider themselves to be. We applied this instrument in the third wave of the panel study “Cultural education and cultural participation in Germany” in 2025. The gender entered on the birth certificate was queried as well. We first examine the quality of this new measurement approach. Secondly, we investigate the explanatory power of gradational gender identity in comparison with binary gender affiliation for a broad set of outcome variables of cultural consumption at home and outside the home, as well as one’s own creative cultural activities. Our general hypothesis is that gender identity produces stronger effects than binary gender categorization. We expect that cultural activities which are more typical for women than for men (e.g., attending dance shows and reading), will be practiced more strongly with an increasingly feminine identity. Conversely, cultural activities with masculine connotations (e.g., attending rock concerts and visiting technology museums) should be practiced more often as masculine identity increases. In addition, the interaction between both gender measures is examined in multivariate regression models to obtain more differentiated results.

Anica Waldendorf

Nuffield College, University of Oxford

How does a new behaviour spread? Numerous studies have investigated diffusion processes and revealed that factors such as network topology and characteristics are central in determining the course of diffusion. Less attention is paid to what is diffusing (i.e. characteristics of that behaviour) in favour of identifying patterns of diffusion. These studies hinge on a key assumption: that adoption is absolute. Either the behaviour is adopted or not. I assume that what is diffusing matters and waive the assumption that the behaviour is absolute. Instead, I propose that there is a class of behaviours where adoption is not absolute but flexible, which makes adoption transient. This in turn could hamper or enable adoption. Flexible adoption could lower the threshold for adoption, making diffusion more likely. Yet it could also create instability as any spread can be reversed before a critical mass is reached. I use gender-inclusive language as a case-study to build the concept of transient adoption and delineate implications this has for diffusion processes. I draw on two waves of qualitative interviews with integrated ego-centric network data for 21 cases. In a second step, I use these insights to set up agent-based models where adoption is not absolute but transient, to test whether relaxing the assumption of absolute adoption is a barrier or enabler for diffusion.

Mark Lutter; Naomi Pech; Lisa Wunsch

University of Wuppertal

Scholarship allocation plays a crucial role in shaping of academic careers. While prior research has examined social biases such as class and gender, less is known about the effects of using vs. not using gendered language in professional contexts. We report findings from a large-scale between-subjects survey experiment with over 10,000 responses, in which academics at German universities evaluated motivation letters for fictitious scholarship applications. The letters varied by applicant gender, class background, and use of gendered language. Respondents assessed style, form, spelling/grammar, persuasiveness, recommendation likelihood, and perceived success. Results indicate that the use of gendered language generally leads to more favorable evaluations, particularly in terms of recommendation likelihood and perceived success. This is also partly true for the rating variables, especially for the ratings on the letter’s form and persuasiveness. This discrimination effect is especially strong among post-docs, but not so strong among pre-docs and even absent among professors. Furthermore, applicant gender and class background matter; female applicants receive more favorable ratings than their male counterparts, while those from affluent backgrounds are less likely to be recommended but are perceived as more likely to succeed. Overall, male students, those from higher class backgrounds, or those who do not use gendered language face lower recommendation rates and less favorable letter ratings.