Gender role models and early-career decisions

ABSTRACT

This paper analyzes the link between the subject choices of German students in upper secondary school and teacher gender when these choices are taken. Our results corroborate the hypothesis that teacher gender matters in this regard, and they indicate that girls respond more strongly than boys to same-sex role models. This is in particular the case in maths: Girls who have a female maths teacher in grade 10 are significantly more likely to choose maths as an advanced course in upper secondary school than girls who are taught by a male teacher. We do not find a corresponding same-sex teacher effect in maths for boys. In contrast, both girls’ and boys’ choices of German as an advanced course in upper secondary school are, at least to some extent, influenced by teacher gender in grade 10. However, also this same-sex teacher effect tends to be smaller for boys than for girls.

KEYWORDS:

• Gender
• education
• STEM
• subject choices

JEL CLASSIFICATION:

• I21
• J16
• J24

Acknowlegement

This paper uses data from the National Educational Panel Study (NEPS): Starting Cohort 4 – 9th Grade, doi:10.5157/NEPS:SC4:11.0.0. From 2008 to 2013, NEPS data were collected as part of the Framework Programme for the Promotion of Empirical Educational Research funded by the German Federal Ministry of Education and Research (BMBF). As of 2014, the NEPS survey has been carried out by the Leibniz Institute for Educational Trajectories (LIfBi) at the University of Bamberg in cooperation with a nationwide network.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ A documentation can be found at DOI: 10.5157/NEPS:SC4:11.0.0.

² Maths: ‘I get good grades in math’ (1 = strongly disagree, 5 = strongly agree), German: ‘I am a hopeless case’ (1 = strongly agree, 4 = strongly disagree)”.

³ These were ‘Duties in life: men should earn money.’ (1: totally disagree, 4: totally agree), ‘Men are more capable in some jobs.’, ‘Men and women have the same duties in house care work’, ‘There should be equal quotas in politics’, and ‘Men and women have equal control over technical devices’. We calculate the arithmetic mean across these statements so that the final variable is still between one and four. Note that Cronbach’s Alpha reliabilities (Cronbach 1951) are high (0.76).

⁴ p-values of Student’s t-tests: 0.734 (students’ gender, maths), 0.869 (grades, maths), 0.899 (students’ gender, German), and 0.003 (grades, German).

⁵ Unfortunately, this information is only available for a subsample.

⁶ Probit and logit models yield very similar results.

⁷ When we also control for gender composition in class in a reduced sample in column (4), our point estimates further increases and corresponds to an increased likelihood of 10.4 percentage points.

⁸ It should be noted that the same-sex teacher effect for girls become smaller in magnitude and statistically insignificant when we include gender ratio in class in column (4). However, also the number of observations is substantially lower in this regression.