Jesper Fries defends his PhD thesis at the Department of Economics

Candidate:

Jesper Fries, Department of Economics, University of Copenhagen

Title:

Essays in Education Economics - Class Size, Peers and Teacher Assessments

Supervisors:

  • Mette Ejrnæs, Professor, Department of Economics, University of Copenhagen
  • Miriam Wüst, Associate Professor, Department of Economics, University of Copenhagen

Assessment Committee:

    • Mette Foged, Associate Professor, Department of Economics, University of Copenhagen
    • Hans Henrik Sievertsen, Professor MSO, VIVE
    • Ludger Wößmann, Professor, Department of Economics, Ludwig-Maximilians-University Munich

    Summary:

    This thesis consists of three self-contained chapters, all within the field of Education Economics. All chapters are empirical papers about students and teachers in Danish public schools from grade 0 to grade 9. All papers exploit data on either student wellbeing, test scores, absence rates, or final GPA. Two of the papers also use administrative data. 

    Chapter 1: Class Size Revisited and Student Well-Being

    Extensive literature exists on the effect of class size on academic performance. This paper extends this work by examining the causal effect of class size on student well-being and perceived school quality in the Danish public school system. First, in line with recent research, I show that comparisons of different school-cohorts locally around thresholds induced by a class size cap can be misleading even in a public school system. Second, I use variation in average class size within school-cohorts generated by a maximum class size rule of 28 students. Exploiting yearly enrollment changes that cause cohorts to move above and below the 28-student threshold, I find that increasing class size lowers students’ well-being and perceived school quality. The effects are particularly strong and sizeable for students’ perception of the learning environment.

    Chapter 2: Peer Effects on Well-Being and Performance in Elementary School

    With Isabel Skak Olufsen

    We study the effect of having resourceful peers on pupils’ well-being, absence rate, and academic performance in elementary school. To estimate the causal effect, we compare pupils in the same school who started school in different years and, consequently, were exposed to different shares of high socioeconomic status (SES) peers. Our study is made possible by a unique combination of administrative, survey, and test data on pupils in Danish public schools. We find that a higher share of high-SES peers increases wellbeing and performance, while the share of high-SES peers does not affect pupils’ absence rates. Low-SES pupils have greater performance gains than high-SES pupils, while the positive effect on well-being is constant across pupils’ SES. The effect is also larger for pupils whose parents do not cohabit and boys. Overall, our results show that high-SES peers positively affect the learning environment in elementary school. This implies that educational inequalities may increase due to the growing concentration of high-SES pupils in certain schools.

    Chapter 3: You Remind Me of My Child - The Impact of Parenting Daughters on Gender Difference in Teacher Assessments

    With Magnus Lindgaard Nielsen and Jeppe Søndergaard Johansen

    This paper assesses whether gender differences in teacher assessments are linked to teachers’ gender attitudes. Using Danish administrative data on 9th-grade students in public schools, we compute the difference between teacher-assigned grades and national exam grades in Danish and Math. We show that female students are assessed more favorably than male students in both subjects, reaffirming findings from other studies. Although male teachers generally grade more leniently, teacher sex does not affect the gender difference in assessments. Next, we examine whether the sex of a teacher’s firstborn child, a plausibly exogenous shock to gender attitudes, affects the gender difference in assessment. Our findings do not suggest that teachers’ gender attitudes affect student assessments. However, we are reluctant to make strong conclusions based on this analysis, as we ultimately cannot determine whether our empirical strategy captures a shift in gender attitudes. Thus, the documented gender difference in assessment could be related to both differences in student behavior and gender-biased behavior among teachers. 

    An electronic copy of the dissertation can be requested here: lema@econ.ku.dk