French Education Sorts Where It Should Elevate
In 2022, only 7% of French students achieve excellence in mathematics [4]. The gap in results between a privileged student and a disadvantaged student reaches 113 points, one of the largest in the OECD [4]. Spending per student in primary education remains 9% below the OECD average [5].
And 67% of students are enrolled in institutions where the principal reports a shortage of teachers [4]. These four data points share a common root: a system where barely 10% of educational decisions are made at the institutional level [11]. The 2027 decision point is clear: transfer to institutions the authority to recruit teachers, or continue assigning teachers through national examination, a teacher the principal does not choose.
Twenty Years of Decline in Mathematics, Stabilization at Primary Level
France has participated in the PISA rankings since 2000. The series now covers more than a generation. In mathematics, the score of 15-year-olds was 511 in 2003 [1].
It still reached 495 in 2012 [1]. In 2022, it falls to 474 [1]. This is a drop of 21 points compared to 2018. The average for OECD countries loses 15 points over the same period [1].
In reading comprehension, the French score in 2022 is 474, compared to an OECD average of 476 [2]. The French decline since 2018 reaches 19 points. The OECD average loses 10 points [2].
The decline touches the entire distribution. In 2003, 15% of French students ranked among the highest performers in mathematics [1]. This was 13% in 2012, then 11% in 2018 [1].
In 2022, this figure falls to 7% [1]. In the same movement, the share of students struggling with mathematical literacy goes from 22.4% in 2012 to 29% in 2022 [4]. The decline strikes both the top and bottom of the distribution simultaneously.
Primary education presents a different picture. PIRLS measures reading comprehension among fourth-grade students. After fifteen years of continuous decline, France stabilizes its results in the 2021 cycle [3].
The vast majority of European Union countries show a statistically significant decline [3]. With a score of 514 points, France exceeds the international average of 500 points, but remains below the European average of 527 points [3]. The twenty-year trajectory: 525 in 2001, 520 in 2011, 511 in 2016, then 514 in 2021 [3]. A modest reversal after twenty years of decline.
These two series draw a system fracture. The fundamentals of primary education are recovering. Secondary education loses skills at a rate higher than the international average.
French School Inequality, Stable for Twenty Years and Deeper Than Elsewhere
The gap between results for privileged and disadvantaged students reaches 113 points in mathematics in 2022, compared to an OECD average of 93.5 points [4]. The most privileged students in France achieve better results than their OECD counterparts. The least privileged students achieve worse results than theirs [4].
This phenomenon is long-standing. All international comparisons for more than twenty years place France among countries where the gap in achievement between privileged and working-class backgrounds is most pronounced. Over the 2012-2022 period, the OECD notes no significant worsening [4]. The stability of this figure means that none of the successive policies pursued over a decade has touched the main mechanism.
The structure of secondary education aggravates the phenomenon. In 2022, 33% of students from disadvantaged backgrounds are enrolled in vocational tracks in France [4]. The OECD average is 17% [4].
Only 6% of the most privileged students take this path [4]. Students in these tracks obtain results 90 points lower in mathematics than those in general tracks. In the OECD average, this gap is 59 points [4].
National assessments over thirty years confirm a decline in performance in mathematics and French [8]. Approximately 30% of students in difficulty are concentrated in the least well-resourced tracks [8].
The social gap forms in primary education, widens there, and vocational track assignment at age 15 crystallizes it. The quality of teaching provided to disadvantaged students before assignment separates them from others is the decisive lever, more than late remedial programs [8].
High Total Spending That Underinvests Where Gaps Form
The weight of education spending in GDP reaches 6.8% in 2024 [7]. The state finances 55% of the total, local authorities 23% [7]. Average spending per student stands at 10,920 euros, from 9,080 euros for a primary student to 11,660 euros for a secondary student [7].
This volume of spending does not translate into comparable performance. Spending per student in primary education is 9% below the OECD average [5]. In the final cycle of secondary education, it is well above this average [5]. France underinvests at the level where gaps form. It overinvests where destinies are already determined.
The Council of Economic Analysis documents, in a 2025 note, the high impact of an increase in student-teacher ratios on academic performance [9]. The class divisions in first and second grade in priority education zones (REP+) go in this direction, but have only touched a limited fraction of the students involved [9].
The Treasury Department, in its December 2025 note, estimates that if France reached the level of the ten highest-performing OECD countries at PISA, productivity would ultimately increase by 2.7% [8]. Human capital is the main engine of productivity gains in economies at the technological frontier [8].
Sixty Percent Fewer Candidates for Teacher Exams in Twenty Years
The decline in performance is accompanied by a parallel rupture: the drying up of teacher recruitment. In 2005, there were 171,750 candidates for 32,049 open positions [8]. In 2023, only 71,075 candidates applied for 23,233 positions [8]. This is a drop of nearly 60% in applications in less than twenty years.
In 2024, of the 27,589 open positions, 3,185 remained vacant [8]. Applications for secondary education exams continue to decline. In 2025, a 4% decline represents nearly 3,000 fewer candidates than in 2024 [8]. The agrégation alone loses 1,600 candidates [8].
In 2024, 11.71% of positions in total remain unfilled [8]. In mathematics, this rate reaches 20% [8]. In classical literature, it rises to 36%. In German, to 54% [8].
The source of the problem is salary. Between 2015 and 2023, real salaries of mid-career French teachers increased by 1%, compared to 4% on average in the OECD [6]. The nominal increase of 15% essentially compensated for inflation [6]. The gap in annual salary for secondary education teachers with fifteen years of experience reaches 16% between France and the OECD average [6]. It takes an average of thirty-five years to reach the top of the salary scale in France, compared to twenty-six years on average in the OECD [6].
Since 2018, France is, after Cambodia, the country where the number of students affected by teacher shortages has increased the most [4]. Positions remain vacant. Untrained contract workers replace permanent staff [8]. Teaching quality degrades first and foremost in the least attractive institutions, often the most disadvantaged [8]. One-time bonuses do not address this mechanism: it is effective compensation in mid and late career that determines vocations [6].
Ten Percent of Decisions at the Institutional Level: The Institutional Lock
In France, only 10% of decisions made in education are made at the institutional level, and barely 2% with complete autonomy [11]. Since the 1980s, the vast majority of OECD countries have increased the degree of autonomy of their schools, in pedagogy, financial management, and human resources. France has remained apart from this movement [11].
Three forms of autonomy are associated with performance in international data: personnel management, pedagogy, and evaluation systems [11]. Budget autonomy alone does not generate results. Personnel management autonomy and pedagogical autonomy are decisive [11].
An education system that evolves only under external constraint reproduces the same bottlenecks as industries defended by established corporate interests. Carl Benedikt Frey develops this diagnosis in How Progress Ends (2025) [12]: centralization protects position rents, national teacher assignment, uniform programs, absence of school-level evaluation, to the detriment of overall performance.
Transfer Recruitment to Institutions and Make Results Public
The highest-performing countries have, over the past twenty years, separated two powers that the French central state still combines: deciding who teaches where, and deciding what happens in the classroom [11]. These two monopolies coexist nowhere with strong school performance [11].
The 2027 decision is as follows: transfer to institutions, under academic and not ministerial oversight, the authority to recruit permanent teachers and define pedagogical methods. The national assignment of teachers through a single exam ceases to be the rule. Each institution publishes its results each year, adjusted for the social profile of its students.
When barely 10% of educational decisions are made at the institutional level, the principal manages an engine room whose levers he does not hold [11]. And 67% of French students are enrolled in institutions where the principal reports a teacher shortage [4]. These two realities are connected: a system that cannot recruit locally cannot respond locally to teaching vacancies.
The counterpart is real. Giving institutions the power to recruit means accepting that attractive institutions more easily capture good teachers. Without transparency of results adjusted for social profile, this autonomy widens inequalities instead of reducing them. A system that decentralizes without evaluating reproduces position rents at the local level rather than the national level.
The mandatory publication of results by institution, adjusted for the social profile of students, is therefore the condition, not the accessory. It makes it impossible for an institution to declare itself excellent without anyone being able to verify it. It imposes political discomfort that unions and inspection bodies have so far blocked. This is what this decision requires accepting.
Sources
[1] DEPP, Bernigole V., Fernandez A., Loi M., Salles F., “PISA 2022: France is not an exception to the widespread decline in mathematical literacy performance in the OECD,” Information Note No. 23.48, December 2023, https://www.education.gouv.fr/media/159228/download (accessed 04/09/2026).
[2] DEPP, Bret A. et al., “PISA 2022: Scientific literacy, reading comprehension and student life,” Information Note No. 23.49, December 2023, https://www.education.gouv.fr/depp/pisa-programme-international-pour-le-suivi-des-acquis-des-eleves-41558 (accessed 04/09/2026).
[3] DEPP, Conceicao P., Desclaux J., Lacroix A., “PIRLS 2021: France stabilizes its results contrary to most other European countries in decline,” Information Note No. 23.21, May 2023, https://www.education.gouv.fr/depp/pirls-2021-la-france-stabilise-ses-resultats-contrairement-aux-autres-pays-europeens-majoritairement-452301 (accessed 04/09/2026).
[4] OECD, PISA 2022 Results (Volume I): Learning and Equity in Education, December 5, 2023, https://www.oecd.org/fr/publications/2023/12/pisa-2022-results-volume-i_76772a36.html (accessed 04/09/2026).
[5] OECD, Education at a Glance 2024, September 2024, https://www.oecd.org/en/publications/education-at-a-glance-2024_c00cad36-en.html (accessed 04/09/2026).
[6] OECD, Education at a Glance 2025, September 2025, https://www.oecd.org/en/publications/2025/09/education-at-a-glance-2025_c58fc9ae.html (accessed 04/09/2026).
[7] DEPP, Ministry of National Education, “In 2024, 197.1 billion euros devoted to education, or 6.8% of GDP,” Information Note No. 25-52, September 2025, https://www.education.gouv.fr/depp/en-2024-1971-milliards-d-euros-consacres-l-education-soit-68-du-pib-452853 (accessed 04/09/2026).
[8] Treasury Department, Barde V. and Leclair B., “The performance of the French education system: what economic stakes?,” Trésor-Éco Letter No. 377, December 2025, https://www.tresor.economie.gouv.fr/Articles/2025/12/02/la-performance-du-systeme-educatif-en-france-quels-enjeux-economiques (accessed 04/09/2026).
[9] CAE, Grenet J. and Landais C., “Education: how to better direct public spending,” Notes from the Council of Economic Analysis No. 84, 2025, https://www.cae-eco.fr/Education-comment-mieux-orienter-la-depense-publique (accessed 04/09/2026).
[10] France Stratégie, “2017/2027: What autonomy for schools?,” analysis note, 2017, https://www.strategie-plan.gouv.fr/publications/20172027-autonomie-etablissements-scolaires-actions-critiques (accessed 04/09/2026).
[11] OECD / Strategic Analysis Centre, “Lessons from research on the effect of school autonomy,” Analysis Note No. 316, January 2013, https://strategie.archives-spm.fr/cas/content/autonomie-etablissements-scolaires-NA316.html (accessed 04/09/2026).
[12] Carl Benedikt Frey, How Progress Ends: Technology, Innovation, and the Fate of Nations, Princeton University Press, 2025, https://press.princeton.edu/books/hardcover/9780691233079/how-progress-ends (accessed 04/09/2026).
