Gender inequalities in higher education: extent, development and mechanisms of gender differences in enrolment and field of study choice

Abstract

In the course of educational expansion, ‘gender differences in access to higher education have decreased substantially in many European countries. In Germany women are currently over-represented in upper secondary education and more frequently attain a general qu1alification for university entrance. Despite those advantages, women still enrol in higher education less frequently. Even where they apply to higher education, their field of study choices differ from those of men. Drawing on recent data on upper secondary graduates, we seek to understand which factors account for both phenomena. Our findings suggest that the growing gender gap in the transition to higher education can be largely ascribed to differential perceptions of labour market outcomes. In addition, the more risk-averse assessment of study costs and success probabilities are more important for women's decision not to enter higher education. Furthermore, the gender gap in the choice of a technical field can be explained by differences in the subjectively perceived comparative advantage in technical skills and by differences in educational motivations and conceptions of life and career plans. These results lead to the conclusion that gendered educational choices are the consequence of effects that have their provenance much earlier in life.

Keywords:

• gender inequality
• higher education
• enrolment decision
• field of study

Notes

1. An overview of the type of training which is required for a specific occupation can be found at http://berufenet.arbeitsagentur.de/berufe/index.jsp.

2. The weights recalibrate the data according to the marginal distributions of GPA, gender, type of school, type of upper secondary degree and federal state.

3. The German grading scale runs from: 1 (excellent) to 6 (insufficient), which might appear counterintuitive to many readers. Hence, we work with an inverted and standardised scale, so that positive values indicate above-average performance and negative values below-average performance.

4. This item is potentially problematic with respect to the causal direction. Since 46% of the respondents are already enrolled in higher education at the time of the interview, their answers might be biased through ex post rationalisation. Hence, our measurement might overestimate its effect on study intentions.

5. The fact that these items were collected in the second wave makes them prone to the same problems as described in note 4, i.e. their effects might be overestimated.

6. A detailed description of the method can be found in Fairlie (2005).

7. The contribution of this variable might be overestimated through ex post rationalisation (cf. note 4). We ran some sensitivity checks excluding those respondents who are in post-secondary education already. The results did not reveal any serious bias.