Christian Møller Dahl
University of Southern Denmark

Nick Ford
Lund University

Kristin Ranestad
University of Oslo

Paul Sharp
University of Southern Denmark

Christian Emil Westermann
University of Southern Denmark

---

Among educational policymakers and practitioners, a central concern relates to observed differences between male and female students. In general, female students perform better in graded assessments than their male peers. Moreover, male and female students exhibit different propensities to select into different subjects and courses, feeding into gender disparities across a range of occupations.

We offer a historical lens on these contemporary trends, by examining Norwegian high school records over a period of more than a century. The high school grade lists contain detailed individual-level data on graduates, including parental information, school attended, program chosen, and grades in individual subjects. Our data series commences in 1813, with the opening of the University of Oslo — the first year the high school exam (examen artium) was held in Norway. From 1882, girls were permitted to take the exam and access university education.

To what extent can differences in educational attainment and performance be observed between male and female students in historic settings?

The earliest female cohorts principally enrolled in the natural sciences (Real) programme at high school. However, as the number of female students increased, most participated in the humanities tracks (the Latin and English lines). By contrast, male students to a substantially greater extent selected into the natural sciences line.

With respect to grades, we find no persistent differences between male and female students during the period we study. While the first female students were exceptionally talented, achieving on average up to a whole grade higher than males, within a generation this advantage had dissipated. Our findings hold when controlling for differences in study programmes, schools, and household backgrounds.