Enrollment in primary education has been steadily declining since the mid-2010s, while secondary education continued to grow between 2015 and 2025. According to the Institute for Public Policy (IPP), maintaining current teacher staffing levels would reduce average class size from 22.4 to 18.2 students, a gain that, translated into academic results, represents an estimated economic benefit of 4.5 billion euros. The choice to be made is straightforward to state: reallocate these resources or eliminate them. The study compares scenarios for state management of teachers without establishing which scenario local authorities would most often adopt.
The Essentials
- Maintaining teaching positions despite demographic decline would allow average class size to shrink from 22.4 to 18.2 students, with an estimated economic gain of 4.5 billion euros (IPP, June 2025).
- Territorial disparities are massive: the Versailles academy enrolls 25 times more students than the Corsica academy (DEPP, Geography of Schools 2026).
- Between 2015 and 2024, primary enrollment fell by 8% while teacher numbers rose by 1%; demographics create the opportunity, but choices about how to distribute teachers—requiring national and territorial coordination—determine whether this opportunity improves or fails to improve educational oversight.
- Without an explicit national strategy, demographic decline risks amplifying fractures between academies rather than bridging them.
Student Population Loss as an Unexpected Resource
France is losing students. Since 2017, births have generally declined, with an exception in 2021. The National Evaluation, Planning and Performance Department (DEPP) projects declining school enrollments until 2035; the IPP relies in particular on demographic scenarios from INSEE (the French national statistics institute). Classrooms are gradually emptying in many academies, and the student-to-teacher ratio is mechanically easing.
The IPP study was published in June 2025 and models notably the maintenance of classroom teachers at their 2024 levels in each department. Result: average class size would drop from 22.4 to 18.2 students. The cited literature indicates measurable effects of class size reductions, particularly for the most vulnerable children, without validating a general threshold set at twenty students.
The IPP puts the gain at 4.5 billion euros. This figure translates the expected effect on future earnings of beneficiary students, based on empirical estimates of class size effects on academic achievement and the return on skills in future earnings. It is an estimate based on debatable assumptions, not a certainty. Maintaining staffing levels, however, implies a budgetary reallocation: the report estimates that this strategy forgoes roughly 3.4 billion euros in annual savings expected by 2034.
The answer, for the moment, is: most of them.
How Academies Actually Manage Staffing When Students Leave
When enrollments fall, accounting logic takes hold. Fewer students means less overall teaching time allocation, fewer positions opened for transfer, sometimes class or school closures. The ministry and regional education authorities calculate ratios, prefectures make decisions, and unions negotiate at the margins. The overall result: between 2015 and 2024, primary enrollment fell by 8% while the number of teachers rose by 1%, showing that resources are not following the demographic decline proportionally.
This mechanism is not the product of ill intent. It stems from management rules that have their own logic: if a territory loses students, its needs appear to diminish. Maintaining positions would require an active decision to redeploy resources, backed by an explicit pedagogical vision. Such a decision is rare.
Data from DEPP in Geography of Schools 2026 illuminate the scale of existing disparities, even before this opportunity is seized or missed. The Versailles academy enrolls 25 times more students than the Corsica academy. These two territories do not have the same needs, the same constraints, or the same geography. But they fall under the same national administrative framework for school zoning; allocation is not formally reduced to student numbers alone and can incorporate territorial, social, and transportation criteria.
Demographic decline strikes unevenly and in highly territorial patterns. It affects notably the north and east, but is not limited to rural areas, small towns, and academies in the center. Changes vary sharply by academy and department, without a general opposition between metropolitan areas or coastal regions on one hand and other territories on the other. A proportional adjustment of teachers to demographic decline would maintain territorial gaps in class size rather than reducing them. Families with lower mobility, often less affluent, end up in academies that are losing both children and teachers.
Class Splitting: What We Already Know About Class Size Effects
France is not discovering this debate. Since 2017, the class-splitting program for first- and second-grade students in priority education areas has provided a large-scale experimental ground. Tens of thousands of students in priority network (REP and REP+) zones have benefited from classes reduced to twelve or fifteen students. DEPP evaluations show positive short-term effects in French and mathematics, more marked for students with greater difficulties; results at entry to sixth grade are not statistically significant in REP+.
This precedent matters. It validates the pedagogical intuition underlying the demographic analysis: reducing class size produces real, measurable effects that persist beyond the year in question. Class reduction tends to produce greater gains for students from disadvantaged backgrounds. This is precisely the profile of academies losing students: often territories with socioeconomic indicators below the national average.
Class splitting cost money. Class splitting mobilized additional positions and sometimes required new classrooms or facility adaptations. Demographic decline offers something different: teachers currently in place can be redeployed. Available teaching resources can be reallocated. Maintaining staffing levels would require redeploying teachers between territories to reduce inequalities.
This is an opportunity without equivalent in the recent history of French education policy, and it is structural, not temporary.
Why Municipalities Close Classes Rather Than Reduce Their Size
Primary schools operate according to its own logic. Mayors make decisions on class openings and closures according to a budgetary and political calculation. Closing a school in a declining village symbolically reinforces that decline. Maintaining an open school with reduced enrollment creates operating costs that rural municipalities struggle to bear.
Local elected officials face two imperatives that are difficult to reconcile: maintaining a structuring public service for the territory and managing municipal finances under pressure. Academic authorities locally determine class openings and closures; decisions concerning schools and institutions fall to the competent local authorities.
This fragmented governance model between the state, academies, and municipalities produces incoherent decisions at the national level. One academy can close positions in declining zones while another maintains overcrowded classes due to recruitment shortfalls. The absence of fully structured national reallocation leaves each level to decide in isolation; however, ministry coordination mechanisms for territorial coordination do exist.
Teachers themselves cannot be redeployed easily from one academy to another. Inter-academy transfer exists but remains limited in volume and attractiveness. A position eliminated in Creuse does not automatically benefit an overcrowded class in Seine-Saint-Denis. The administrative geography of the education system is rigid, and this rigidity is one of the main obstacles to intelligent reallocation of resources.
The Window Closes Before 2035: What Each Trajectory Implies
Demographic decline is not permanent. INSEE projects population and fertility; DEPP publishes school enrollment projections to 2035. Effects on school-age generations depend first on fertility assumptions. The window of opportunity is therefore limited in time. Decisions made between 2026 and 2030—whether to maintain or cut positions, reallocate or not—will lock in trajectories visible in educational data through 2050.
Two scenarios emerge, following the IPP’s analytical logic.
In the second scenario of the IPP study, the state maintains teachers at their 2024 levels, which mechanically reduces class size. Teacher numbers are maintained at their 2024 levels in each department, without reallocation between departments.
The third scenario models reallocation of primary teachers between departments to reduce class size inequalities, concentrating resources where inequalities are most marked. This scenario requires a decision at the highest political level and the ability to hold that course against annual budgetary pressures.
In the first scenario, teacher numbers fall in each department to maintain constant average class size. The student-to-teacher ratio depends on both student numbers, teacher numbers, and oversight policies; it cannot be attributed to teacher adjustments alone in each department. The scenario of diminishing teachers maintains existing territorial disparities rather than correcting them.
And when births tick up slightly after 2030 without primary school enrollments rebounding by 2034, urgent recruitment will be needed in a context of tension over teaching vocations, having lost a generation of experience in the territories that needed it most.
This second scenario is the status quo. It requires no active decision, only the absence of one. This is precisely why it is most likely if no national impetus is given in coming years.
The signals to watch are clear: the evolution of positions opened for exams in demographically declining academies, supervision rates by academy published annually by DEPP, and school zoning decisions in rural zones. These indicators will reveal by 2027 or 2028 in which direction the system is heading.
Choices Being Made Now, in Regional Education Authorities
The IPP does not recommend a universal school policy. Its study identifies an opportunity and calculates what it is worth. Political translation belongs to the actors: the Ministry of Education, regional authorities, prefects, mayors, and ultimately Parliament, which votes employment caps.
Several levers exist. Freezing position cuts in demographically declining academies is most direct. It can be decided by circular, without legislative reform. Prioritizing reduced classes in REP areas and territories with below-median household income would extend the class-splitting logic with existing resources. A pact between the state and municipalities to stabilize the rural school network in exchange for measured pedagogical objectives could align incentives instead of leaving them diverge.
The question of teacher mobility deserves particular attention. Experiments in voluntary mobility with incentives, zone premiums, housing, career advantages could allow resources to be redeployed where they are lacking, without imposing unwanted transfers that discourage vocations. Comparable mechanisms exist in other OECD countries, notably Nordic countries, and their evaluations give encouraging results for attracting positions in difficult zones.
None of these levers is spectacular. None requires an orientation law or large-scale structural reform. These are management choices, made in regional authorities and ministries, that do not make headlines but that build the educational trajectories of millions of children for the next twenty years.
Demographics has opened a door. What the institution does with it depends on decisions being made now, often quietly, in budgetary arbitrations that never mention the 4.5 billion euros the IPP has put on the table.
Sources
- Institute for Public Policy, Class Size and Territorial Inequalities: What Strategy in the Face of Demographic Decline? (June 2025)
- DEPP, Geography of Schools, 2026 edition (National Evaluation, Planning and Performance Department, Ministry of Education)
- INSEE, Demographic Projections 2035 (Demographic Dashboards, Annual Report)
- DEPP, Evaluations of Class Splitting for First and Second Grade in Priority Education Areas (2019–2023 publications)
