Skip to content

Institute of Public Policies

Home » Publications » Policy Brief » Girls falling behind in maths from Year 1 onwards: a widespread trend in society

Girls falling behind in maths from Year 1 onwards: a widespread trend in society

Thomas Breda, Joyce Sultan Parraud, Lola Touitou
January 2024
The purpose of this note is to document the gender-based differences in academic attainment that emerge from the start of primary school and how these vary across different institutional and family contexts.
Reference IPP Policy Brief No. 101

Quote this publication

T. Breda, J. Sultan Parraud, L. Touitou. Girls falling behind in maths from Year 1 onwards: a pervasive trend in society. January 2024. IPP Policy Brief 101
Download

Presentation

Girls are at the same level as boys in maths at the start of Year 1 (CP), but begin to fall behind by the middle of this first year of primary school, whilst they maintain a lead over boys in French. Girls’ decline in maths continues, so that by the time they start the first year of primary school (CE1), their average rank or centile (where pupils are ranked in centiles according to their performance, with 100 being the highest value) falls from the 50th centile in CP to the 44th centile by the end of CE1.

Drawing on rich and comprehensive panel data combined with information on parents’ socio-occupational status, the teaching staff at each school, and the general characteristics of the schools, this paper seeks to characterise girls who drop out and to measure the respective contributions of the family environment and schooling (the school and teachers) in explaining the emergence of the performance gaps observed in CP and CE1.

The authors show that girls’ academic underachievement occurs throughout the country, in all types of schools and across all family backgrounds. In other words, no school context (state school, private school, school within a priority education network – REP – or alternative education) nor any family background (parents from affluent backgrounds, those in scientific professions or same-sex parent families) can prevent the emergence of a gap in mathematics performance to the detriment of girls at a very early stage of schooling.

 

Key Results

  • This report is based on the standardised national évaluations administered by the DEPP at the start of Year 1, halfway through Year 1 and at the start of Year 2 to more than 2.5 million pupils
    enrolled in schools in France between 2018 and 2022.
  • Whilst non-existent at the start of Year 1, a gap in favour of boys emerges and widens in mathematics during the course of Year 1. This decline in girls’ performance is observed across
    all the cohorts evaluated and for the majority of the evaluated tasks (addition, reading numbers, problem-solving, etc.). Girls, on the other hand, have an advantage over
    boys in French, which remains broadly stable throughout the first year of primary school.
  • The decline is most pronounced among the highest-performing girls at the start of Year 1 (those in the top 1 per cent at the start of Year 1). On average, these girls fall by nearly 7 places
    by the start of Year 2 compared with boys in the same initial percentile.
  • The widening of the gap in mathematics between boys and girls is observed across all social categories and family structures, and throughout the country.
  • The gap between girls and boys is smaller in classes comprising mainly girls, or when the teacher is a woman rather than a man, and
    when the school is located in a ‘Priority Education Network Plus’ (REP+) area. However, these characteristics linked to the school environment account for only a small proportion of the overall gap, which suggests that the trend is common to society as a whole.

Method and Data

Data matching

The report is based on the compulsory national standardised évaluations administered to more than 2.5 million pupils in CP and CE1 who were enrolled in schools in France between 2018 and 2022.

Information on school characteristics is drawn from the Primary Education Management and Self-Evaluation Tool (APAE 1D). In addition to school type (state, private, REP, REP+), information is provided on pupil numbers (average number of pupils per class, proportion of girls, the school’s IPS score) and on teachers (proportion of women, average age, length of service at the school).

The variables relating to the teacher’s gender, the school’s teaching approach or whether the school is a faith school were derived from text fields in APAE 1D or from the Repères evaluations (class name for the teacher’s gender, ‘private under contract’ status and school name for the school’s teaching approach and whether it is a faith school).

Finally, data on secondary school pupil numbers for pupils entering Year 7 in 2023–2024 provide information on the socio-occupational categories (PCS) of parents and family types of pupils in the first cohort of Year 1 (2018–2019). Based on the parents’ PCS, the authors reconstruct the individual IPS of the legal guardians and the cross-referenced IPS (family), using the DEPP transition tables. The authors also reconstruct family types based on the child’s legal guardians. Where a child has only one declared legal guardian, they are classified as part of a ‘single-parent family’. ‘Other family’ refers to children brought up by someone other than their parents (a relative other than their parents, a carer, a foster carer, a child welfare worker, etc.).

Partners

Chaire Femmes et Sciences de la Fondation Paris-Dauphine

This report is the result of a partnership with the Women and Science Chair.

This research was also made possible by access to data from the Directorate for Evaluation, Forecasting and Performance (DEPP) of the Ministry of National Education and Youth. This project received support from the ‘Innovations, Data and Experimentation in Education (IDEE)’, funded by the French National Research Agency (ANR) as part of the ‘Investissements d'avenir’ programme, reference number: ANR-21-ESRE-0034.

Last modified: July 21, 2026