Providing equal opportunities is a key higher education policy concern all over Europe (EC 2020). This includes tackling inequalities along certain axes such as migration or social background and gender. Considering gender, the gap between women and men in education has strongly decreased and/or reversed in many postmodern societies (Hadjar & Buchmann, 2016). However, this refers only partly to German higher education. Among upper secondary school leavers, females are less likely to opt for higher education than males (Lörz & Schindler, 2011). Furthermore, despite their overall better achievement, women are underrepresented among scholarship recipients (Haas & van de Werfhorst, 2017). 

While the main pillar of financial support for students in German higher education consists of a public need-based funding scheme, a smaller share of students receives a scholarship. These public scholarship schemes particularly aim at supporting talented and committed students through monetary funding, but also with the provision of networks, mentoring and workshops.

There are two main programmes: first, the scholarships for the talented are administered through foundations that are loosely linked to religious denominations and political parties, and second, the recently introduced “Deutschland-Stipendium” (Germany stipend) which is partly funded through local private sponsors and higher education institutions. As the share of funded students is relatively small, competition for scholarships is strong and the selection process is demanding. Some programmes accept candidatures via recommendation by teachers or professors, other programmes have extensive application procedures. Thus, these scholarships are perceived as a signal of distinct eligibility and talent.

Given that females perform better in higher education, the female underrepresentation among scholarship recipients – with excellence rather than need as main criterion – seems implausible. This paper sheds light on this finding by examining a set of potential mechanisms responsible for this outcome: first, males might be advantaged in receiving scholarships as they are more likely to be eligible for a scholarship, i.e. overall, they fulfil the programme requirements better than females, which are – in addition to excellent achievement – more soft criteria such as societal engagement, commitment to serve society or motivation. Second, as prior studies have shown that male students act more career-oriented than females, they may perceive such scholarships particularly as a signal of distinction helping them in realizing future career goals (Jonsson, 1999). Third, as it has been found that an information dissymmetry is responsible for students’ disadvantage regarding certain educational decisions (Herber, 2018; Plank & Jordan, 2001), it could also be that this gender gap might be due to an information deficit in relation to these scholarship programmes. Fourth, we hypothesize that male students have better chances to receive scholarships as they are more critical towards their own achievement (Lörz & Schindler, 2011), consequently making them less likely to feel eligible for a scholarship or less self-confident during the selection process.

The female underrepresentation may be also related to the gender-specific field of study choice. However, while females are still underrepresented in STEM fields (Lörz, 2011), this hypothesis would only seem valid if students in STEM fields or other male dominated fields would have a significantly higher proportion of scholarship recipients – an argument that seems incoherent in the German case. In contrast, we hypothesize that female students in male-dominated subjects might be advantaged in receiving a scholarship as professors may assume that they deserve particular support in serving as female role models. In contrast, we further hypothesize that male students in female-dominated fields might be overrepresented among scholarship recipients as they perceive scholarships as a strategy to compensate the perceived penalty of studying in a female-dominated subject – which yield on average lower returns to education than male-dominated ones.

Our analysis is based on the starting cohort 5 “first-year students” of the National Education Panel Study (NEPS; Blossfeld et al, 2011). This is a representative, large-scale dataset, which follows students that enrolled in the academic year 2010/2011 in German higher education. Students are regularly surveyed on their higher education progress, employment situation, attitudes and many other themes. We define scholarship recipients as students who received some financial contribution from a scholarship on at least two points in time (e.g. over a period of more than a year) in order to exclude students that receive a studying abroad grant (e.g. Erasmus scholarships). Reverse causality bias can be minimized through the nature of panel data. For several reasons, we excluded students studying at the more vocationally oriented universities of applied sciences. Our modelling strategy is as follows: we employ logistic regression models to estimate the likelihood to be a scholarship recipient. Second, we add successively more explanatory variables to the model and decompose the female disadvantage in being a scholarship recipient into several effect variables using the KHB decomposition method for binary regression models (Breen et al, 2013). However, the low number of scholarship recipients in our sample threatens the validity of our results – a problem known as small sample bias of rare events. Thus, we are considering sensitivity checks using methods such as bias-corrected logistic regression or Penalized Maximum Likelihood Estimation.

Our descriptive results confirm previous findings by yielding a gender gap in the likelihood to be a scholarship recipient. Our preliminary logistic regression results are as follows: relevant factors explaining the likelihood to receive a scholarship are regular societal engagement, engagement in higher education organizations, university-entrance GPA as well as current GPA. Students in humanities are less likely to receive scholarships compared to students in other subjects, but this effect is not significant across all field of studies. When controlling for field of study, achievement and engagement, females are significantly less likely to receive a scholarship pointing to the fact that lacking eligibility does not explain the gender gap. Adding further explanatory variables to the model, the direction of the effect persists but remains only weakly statistically significant. While regular discussion on social and political issues as measure for students’ interest in societal issues and students’ self-assessment of achievement in relation to their actual achievement does not explain whether students receive a scholarship, going to information sessions on financial aid significantly helps in receiving a scholarship. Regarding the relationship of field of study and gender, we find that male engineering, law, economics and social sciences students are significantly more likely to receive a scholarship compared to the reference group of female humanities students, while females in these respective fields are not. In science subjects, both male and female students are significantly more likely to receive a scholarship compared to female students studying humanities. Summing up, while we identified some explanatory factors on the likelihood to receive a scholarship and confirmed the gender gap among scholarship recipients, the results should be considered as preliminary. Next steps include the decomposition of effects, adjusting for the complex survey design and attrition and rare occurrence bias.

Blossfeld, H.-P., et al. (2011). "The National Educational Panel Study: need, main features, and research potential." Zeitschrift für Erziehungswissenschaft 14(2): 5-17. Breen, R., et al. (2013). "Total, direct, and indirect effects in logit and probit models." Sociological Methods & Research 42(2): 164-191. European Commission 2020: About higher education policy. Webpage https://ec.europa.eu/education/policies/higher-education/about-higher-education-policy_en, last accessed on 23/01/2020 Haas, C. and H. Van De Werfhorst (2017). "Ahead of the pack? Explaining the unequal distribution of scholarships in Germany." British Journal of Sociology of Education 38(5): 705-720. Hadjar, A. and C. Buchmann (2016). "Education systems and gender inequalities in educational attainment." Education systems and inequalities. International comparisons: 159-184. Herber, S. P. (2018). "The role of information in the application for highly selective scholarships: Evidence from a randomized field experiment." Economics of Education Review 62: 287-301. Jonsson, J. O. (1999). "Explaining sex differences in educational choice an empirical assessment of a rational choice model." European Sociological Review 15(4): 391-404. Lörz, M. and S. Schindler (2011). “Geschlechtsspezifische Unterschiede beim Übergang ins Studium. Geschlechtsspezifische Bildungsungleichheiten“, Springer: 99-122. Plank, S. B. and W. J. Jordan (2001). "Effects of information, guidance, and actions on postsecondary destinations: A study of talent loss." American Educational Research Journal 38(4): 947-979.