On average across OECD countries, lower secondary school provision has undergone a process of consolidation despite an increase in the student population. The number of institutions declined slightly by 0.10% per year between 2015 and 2024, while the number of students enrolled at the typical age grew by 0.96% per year over that period.
On average across OECD countries, there are 1.97 primary schools for every lower secondary school. The ratio is higher for public institutions, with 2.26 public primary schools per public lower secondary school.
On average across OECD countries, lower secondary schools enrol nearly twice as many students per grade as primary schools (64 students per grade compared to 34).
National averages often mask substantial internal variation in school sizes. Across the OECD, the median size of lower secondary institutions ranges from 16 to 166 students per grade. Although systems in Ireland and Italy display relatively homogeneous school sizes, in places like Korea and England (United Kingdom) provision is much more disparate with very large urban schools coexisting alongside smaller rural ones.
Education at a Glance 2026
In per cent
Note: Information is derived from system-level (aggregate) data.
1. Year of reference differs from 2024.
2. Changes in the typical age of enrolment following education system reforms affect the comparability of trend estimates over time.
For data, see Table B2.1. The data for this figure can be accessed via https://stat.link/n9wxcm.
Context
Educational institutions constitute one of the most visible and enduring components of education systems. As well as providing physical spaces for learning, they shape how students progress through education, how resources are allocated and how educational opportunities are distributed across populations and territories. Decisions regarding the organisation, the number and size of institutions therefore influence not only the efficiency of education systems but also students’ learning experiences and pathways (OECD, 2018[1]).
The organisation of school provision varies considerably across OECD countries. Some systems rely on highly integrated structures in which students remain within the same institution for their entire primary and secondary education, while others have distinct primary, lower secondary and upper secondary institutions. These institutional arrangements influence the timing of transitions between schools, the continuity of learning experiences and the extent to which students encounter specialised educational environments as they progress through the system (Perico E Santos, 2023[2]).
At the same time, education systems are operating in increasingly diverse demographic contexts. Changes in fertility patterns, internal mobility and migration have reshaped the demand for educational provision in many OECD countries, requiring policymakers to consider how educational resources and infrastructure can best respond to evolving population needs (OECD, 2018[1]; Di Cataldo and Romani, 2024[3]) (see Chapter 1, Teacher shortages and the attractiveness of the teaching profession). Beyond these demographic shifts, the resilience of educational infrastructure is increasingly being tested by environmental factors, including the rising exposure of students to extreme heat, which can disrupt learning and impact overall well-being (see Box B2.2).
Differences in the size and distribution of institutions further shape students’ educational experiences. Larger institutions may support broader curricular offerings, greater teacher specialisation and a wider range of extracurricular opportunities, whereas smaller institutions may facilitate more personalised learning environments and stronger connections between students and staff. Understanding how countries organise educational institutions across levels, how school networks evolve over time and how school size varies within systems provides important insights into the ways education systems balance accessibility, quality, continuity and efficiency (Ares Abalde, 2014[4]).
Other findings
Primary education (ISCED 1) starts at the age of 6 in most OECD and partner countries, although the typical entry age ranges from 5 in countries such as Australia, Ireland, New Zealand and the United Kingdom to 7 in Bulgaria, Croatia, Estonia, Finland, Latvia, Lithuania, Poland and Sweden. The total duration of the primary and secondary education cycle ranges from 11 years in Colombia, Costa Rica and Peru to 14 years in Iceland.
Between 2015 and 2024, nearly 43% of OECD small regions experienced a decline in their population aged 10-14. In Japan, Korea, the Netherlands and Portugal, this decline affected more than 85% of small regions.
The organisation of institutions shapes how students progress through compulsory education. In Poland and Slovenia, where integrated school structures cover both primary and lower secondary education, most students remain in the same institution throughout compulsory education, whereas students in France and the Netherlands typically transition to larger and more specialised schools.
Climate change is expected to increase students’ exposure to extreme heat in many countries. In Hungary, for example, the share of primary students exposed to more than 30 hot days in school per year is projected to rise from 19% to 96% by 2050 (relative to the 1983-2016 reference period), highlighting the need for investments in cooling and climate-resilient school infrastructure.
Organisation of primary and secondary education systems
The organisation of primary and secondary education reflects fundamental policy choices regarding how students progress through the education system. Primary education (ISCED 1) is designed to provide foundational literacy, numeracy and learning skills, while lower secondary education (ISCED 2) introduces more subject-specific instruction and prepares students for the broader curricular choices available at upper secondary level (ISCED 3) (OECD/Eurostat/UNESCO Institute for Statistics, 2015[5]). The duration and sequencing of these levels vary considerably across OECD countries, shaping both the timing of educational transitions and the structure of learner pathways.
Figure B2.2 illustrates the substantial differences in the age at which students start primary education and how long the different educational levels last. Although most OECD and partner countries begin primary education at the age of 6, students typically enter at age 5 in Australia, Ireland, New Zealand and the United Kingdom, and at age 7 in Bulgaria, Croatia, Estonia, Finland, Latvia, Lithuania, Poland and Sweden. The combined duration of primary and secondary education ranges from 11 years in Colombia, Costa Rica and Peru to 14 years in Iceland. It is important to note that the duration of primary and secondary education should not be equated with the duration of compulsory education, as some countries include some years of pre-primary education within compulsory schooling while others do not require attendance throughout the full upper secondary cycle (OECD, 2024[6]).
Education systems also vary in the number and timing of institutional transitions experienced by students. In integrated systems, students often progress through primary and lower secondary education within the same institution or school cluster, reducing the number of transitions between schools. In more segmented systems, students move between institutions at a younger age and may encounter several transitions during their compulsory schooling. Transitions represent important moments in students' educational trajectories because they often coincide with significant academic, social and developmental changes (Perico E Santos, 2023[2]). When poorly managed, transitions can increase the risk of disengagement, grade repetition or early leaving from education and training; when well supported, they can facilitate progression into programmes better aligned with students’ interests and aspirations (Perico E Santos, 2023[2]).
The organisation of educational pathways therefore involves balancing continuity and differentiation. Integrated structures may strengthen students’ sense of belonging and provide greater continuity in learning experiences by allowing them to remain within a familiar institutional environment for longer periods. Conversely, systems offering multiple pathways and earlier differentiation may facilitate specialisation and provide a wider range of educational opportunities, but they also require greater coordination to ensure permeability between programmes and equitable progression opportunities. Recent evidence suggests that successful education systems are characterised not by the absence of transitions, but by their ability to support students through them and maintain flexible pathways between programmes and levels of education (Varsik, 2025[7]).
For data, see Table B2.1. The data for this figure can be accessed via https://stat.link/n9wxcm.
Demographic change and the evolution of school provision
Demographic change is reshaping school provision across many OECD countries. Declining fertility rates, together with migration flows, have altered the size and geographical distribution of school-age populations, pressuring education systems to adapt infrastructure to shifting demand. These trends can play out very differently both across and within countries, with most countries experiencing growth in some regions and a decline in others (see Box B2.1). An increase in the student-age population may generate demand for new schools in metropolitan areas, while declining populations in remote areas can pose challenges to service delivery. However, although aggregate changes in student numbers and institutions can mask the impact of local responses to changing demographic patterns, they can still provide an important indication of the overall responsiveness of school networks and how countries are balancing efficiency, accessibility and long-term planning in the provision of education services (OECD, 2018[1]). The need for network responsiveness is further amplified by environmental stressors; for example, rising temperatures are prompting countries to reconsider school infrastructure and operational schedules to mitigate negative impacts on student health and cognitive performance (see Box B2.2).
Between 2015 and 2024, school provision changed slowly overall across the OECD, although the pace differed by level of education. On average across OECD countries, the number of primary institutions declined by 0.57% per year, while the number of students increased by 0.17% per year. At lower secondary level, the number of institutions remained broadly stable, declining by only 0.10% annually, despite student numbers increasing by 0.96% per year. Upper secondary education showed similar stability, with the number of institutions increasing by just 0.08% annually despite a 0.68% annual increase in enrolment. These global trends suggest that systems have sought to accommodate demographic change through adjustments within existing provision rather than through large-scale expansion of educational infrastructure (Table B2.1).
National aggregate changes in student numbers often mask diverse local responses and shifting demographic patterns within a country. A country may show overall stability or growth in student enrolment at the national level, while simultaneously experiencing a decline in its student-age populations in many of its regions (Figure B2.1 and Figure B2.3).
Between 2015 and 2024, almost 43% of small regions in the OECD saw a decrease in the number of 10-14 year-olds, with most countries experiencing growth in some regions and a decline in others. The decline was most widespread in Japan, Korea, the Netherlands and Portugal, where over 85% of small regions saw a decline in the number of 10-14 year-olds. In total, only seven countries did not experience an aggregate decline in any of their small regions (Czechia, Ireland, Luxembourg, New Zealand, the Slovak Republic, Slovenia and Sweden). Israel also recorded no aggregate decline, based on the latest available data for 2023.
In countries like Finland and Lithuania, sustained growth in student numbers at the national level has been accompanied with substantial institutional reorganisation to address declining student numbers in specific regions, with 21% of regions in Finland and 70% in Lithuania experiencing a fall in their student-age populations. These differences underscore how demographic pressure is rarely uniform, requiring policymakers to balance the need for new infrastructure in growing hubs with the challenges of consolidation in depopulating areas.
In per cent
Note: Small regions refer to Territorial Level 3 (TL3) regions, which are subnational administrative or statistical units below the national level. See Definitions section for details.
1. Year of reference differs from 2024: 2023 for Costa Rica and Israel; 2022 for Colombia, Germany, Latvia, Mexico, Norway and the United Kingdom; and 2020 for Portugal.
Source: OECD (n.d.[8]), OECD Regions, Cities and Local Areas database http://oe.cd/geostats.
The data for this figure can be accessed via https://stat.link/n9wxcm.
OECD countries have followed markedly different pathways in their efforts to align school provision with local demographic realities and changing settlement patterns. In Lithuania, the number of institutions declined substantially across primary and lower secondary education, despite sustained growth in student numbers at the national level. The optimisation of the Lithuanian school network is part of broader reforms aimed at improving efficiency and educational quality in response to a projected sharp decline in the population (Shewbridge et al., 2016[9]). Finland provides an example of institutional adaptation driven by local demographic changes, even as aggregate student populations continue to grow. Faced with declining student populations, some Finnish municipalities have progressively consolidated school provision and reorganised educational provision to maintain services (Table B2.1) (OECD, 2025[10]). However, these developments have not been confined to areas experiencing demographic decline, as longer-term trends towards larger school units and broader structural changes have also shaped educational provision across the country.
The strongest evidence of institutional adaptation can be observed at lower secondary level, where several countries reduced the number of institutions despite stable or growing student populations (Figure B2.1). While Lithuania recorded the largest decline in lower secondary institutions (-2.94% annually), neighbouring Poland also experienced a substantial reduction (-2.38% annually), largely reflecting the 2017 education reform that phased out lower secondary schools (gimnazjum) and replaced the previous 6+3 structure with an eight-year primary cycle (OECD, 2025[11]). Elsewhere, countries have increasingly relied on administrative restructuring rather than simple school closures. Portugal's school clustering reform (agrupamentos de escolas), for example, consolidated governance and administration across multiple institutions while maintaining educational provision in different locations. Such approaches illustrate how school systems can adapt to demographic change through organisational innovation, allowing them to improve co-ordination and resource management without necessarily reducing access to educational services (World Bank, 2025[12]).
School closures can have important implications beyond the education sector. Evidence from Finland suggests rural communities have experienced stronger population decline following the loss of local schools (Lehtonen, 2021[13]). The adaptation of education systems must therefore balance efficiency and educational quality with wider territorial considerations about maintaining the accessibility of services, and the attractiveness and competitiveness of rural or remote regions.
The distribution of institutions across education levels
The number of primary schools per lower secondary school provides an indication of how education systems organise progression between these levels. In most OECD countries, lower secondary provision is delivered through a smaller number of institutions serving larger catchment areas than primary education. In some systems, both levels are provided within integrated institutions. These schools are counted once at each level, which contributes to similar numbers of primary and lower secondary institutions. These organisational choices influence the number of transitions students experience, the use of educational infrastructure and the allocation of teaching resources across levels (OECD, 2018[1]).
Note: This figure combines system-level and institution-level data. When both sources were available, institution-level data were used. OECD and EU averages were recalculated accordingly and differ from those reported in Tables B2.2 and B2.3.
1. Year of reference differs from 2024.
For data, see Table B2.2. and Table B2.3. The data for this figure can be accessed via https://stat.link/n9wxcm.
On average across OECD countries, there are 1.97 primary schools for every lower secondary school (Figure B2.4). This indicates that lower secondary education is generally organised through a smaller number of institutions serving larger catchment areas than primary education. Such structures reflect both the increasing complexity of educational provision as students progress through the system and the greater capacity of older students to travel longer distances to school using a wider range of transport options. Lower secondary education typically requires a broader curriculum, greater teacher specialisation and access to specialised facilities, such as science laboratories or technical equipment, which can be more efficiently provided through larger institutions serving multiple primary school catchment areas (OECD, 2018[1]).
Differences between the public and private sectors reveal greater variation in how educational provision is organised in some systems. In most OECD countries and economies, the number of primary schools per lower secondary school is broadly similar for both sectors, but the ratio of primary to lower secondary schools is slightly higher for public institutions than private ones in many countries. This pattern suggests that public school networks more frequently prioritise local access to primary education through a larger number of primary institutions, while private providers are more likely to concentrate provision across fewer institutions or operate schools spanning multiple education levels. The contrast is particularly pronounced in England (United Kingdom), where the ratio reaches 8.6:1 among public institutions but only 2.5:1 among private institutions. This difference partly reflects the prevalence of "all-through schools" and other vertically integrated models within the private sector (Chapman and Muijs, 2014[14]). Overall, these differences suggest that public and private sectors pursue distinct organisational approaches, balancing accessibility, institutional integration and the efficient use of educational resources in different ways.
Differences in the number of primary schools for every lower secondary school can highlight important differences in how education systems organise the transition between these levels of education. Countries such as Hungary (0.94), Lithuania (0.92), Poland (1.01) and Slovenia (1.00) have almost equal numbers of primary and lower secondary institutions, whereas the ratio exceeds four primary institutions per lower secondary institution in Costa Rica (4.62), France (5.03), Ireland (4.27) and the Netherlands (4.71) (Figure B2.4).
These contrasting patterns reflect different approaches to organising compulsory education. In some countries, including Poland and Slovenia, primary and lower secondary education are frequently provided within the same institution, allowing students to remain in the same school throughout compulsory education. In Slovenia, where the unified basic school (osnovna šola) spans both levels, students generally progress without changing schools and remain in the same institution until the end of basic education. However, similar ratios do not necessarily imply similar institutional arrangements. A low ratio may reflect integrated schools spanning multiple education levels, but it may also result from separate primary and lower secondary institutions operating at similar scales within the same school network. Conversely, high ratios generally indicate that lower secondary education is organised through a smaller number of institutions serving larger catchment areas than primary education.
Average school size across education levels
The average size of educational institutions varies substantially across OECD countries and education levels, reflecting different approaches to organising educational provision. Some systems operate relatively small institutions throughout compulsory education, while others concentrate students into larger schools, particularly at lower secondary level. These differences influence not only how resources are allocated, but also the range of educational opportunities available to students, including access to specialised subjects, extracurricular activities and support services (OECD, 2018[1]).
At primary level, institutions enrol an average of 34 students per grade across OECD countries (Figure B2.5) although average sizes range from fewer than 20 students per grade in Costa Rica and Switzerland to more than 50 in Colombia, Israel, Korea and Türkiye. Smaller primary institutions often reflect the geographical distribution of the population and the need to provide local access to education in sparsely populated areas, while in some systems they may also reflect a preference for smaller learning communities during the early years of schooling. Research suggests that smaller schools can facilitate closer relationships between students and teachers, greater parental engagement, and stronger connections to the local community, all of which may contribute to students’ sense of belonging and well-being (OECD, 2024[15]).
Institutional size increases substantially at lower secondary level
On average across OECD countries and economies, lower secondary schools enrol 64 students per grade, nearly twice the OECD average of 34 students per grade at primary level. Particularly large institutions on average are found in Luxembourg (145), the Netherlands (162) and England (United Kingdom) (149). These larger institutions often expose students to a wider range of educational opportunities than would be feasible in smaller settings. As students progress through education, schools are increasingly expected to offer optional subjects, extracurricular activities, academic support services and guidance programmes tailored to diverse interests and aspirations. Larger institutions may therefore provide students with greater flexibility in shaping their learning pathways and exploring future educational and career options (OECD, 2018[1]; Ares Abalde, 2014[4]).
The relationship between school size and educational quality is not straightforward. Although larger institutions may benefit from economies of scale and a broader educational offer, research suggests that the advantages of increasing size diminish beyond a certain point. Very large institutions may face greater organisational complexity and can make it more difficult to maintain personalised learning environments and strong student-teacher relationships (Ares Abalde, 2014[4]). This helps explain why some countries continue to operate relatively small lower secondary institutions despite the potential efficiency gains associated with larger schools.
Differences between primary and lower secondary school size also reveal distinct models of educational organisation. In the Netherlands, lower secondary institutions are several times larger than primary schools, indicating a system in which students move from local primary provision into larger specialised institutions serving wider catchment areas. Conversely, countries such as Poland and Slovenia display relatively similar average sizes across both levels of education. This pattern is consistent with more integrated institutional structures in which students remain within the same school or educational complex for a longer period, resulting in greater continuity in cohort size and educational provision across levels (OECD, 2018[1]).
Note: This figure combines system-level and institution-level data. When both sources were available, institution-level data were used. OECD and EU averages were recalculated accordingly and differ from those reported in Tables B2.1 and B2.3.
1. Year of reference differs from 2024.
For data, see Table B2.1 and Table B2.3. The data for this figure can be accessed via https://stat.link/n9wxcm.
Rather than indicating that there is a single optimal model, these differences illustrate the trade-offs between larger institutions offering more educational opportunities and smaller institutions that can provide more personalised learning environments. Maintaining very small schools can also increase per-student costs, particularly where fixed infrastructure and staffing costs serve a limited number of students. The wide variation observed across OECD countries suggests that school size reflects broader policy choices regarding educational priorities, institutional organisation and the balance between diversity of provision and proximity to learners.
Variation in school size within countries
National averages provide useful information about the overall size of institutions, but they often conceal substantial variation within education systems. Although the median lower secondary school size (based on institution-level data) is 72 students per grade (Table B2.3), the spread of school sizes reveals that many students are educated in much smaller settings. Across the OECD, the middle 50% of lower secondary institutions range from 46 students per grade at the 25th percentile to 111 students per grade at the 75th percentile. However, in many countries, the 25th percentile is significantly lower, highlighting that a large number of very small schools coexist with larger institutions (Figure B2.6. In some countries, such as Bulgaria, Estonia, Latvia, Norway, Peru and Poland, lower secondary schools at the 25th percentile enrol only between 10 and 20 students per grade. Maintaining such small institutions is often a strategic necessity to ensure accessibility in low-density or remote areas, particularly where long travel distances could otherwise limit participation in education (OECD, 2024[16]). However, it also presents significant challenges for education systems. Operating a network of very small schools is generally more costly on a per-student basis due to the limited scope for economies of scale in infrastructure, administration and staffing (Andrews, Duncombe and Yinger, 2002[17]; Leithwood and Jantzi, 2009[18]). The small scale may also limit the educational opportunities available to students, as these institutions often lack the critical mass of staff and students required to offer a broad range of optional subjects, specialised facilities such as science laboratories or diverse extracurricular activities. As a result, students within the same education system may experience markedly different learning environments depending on the size of the institution they attend (OECD, 2024[16]).
Number of students per grade at the 25th, 50th (median) and 75th percentiles
Note: Information is derived from institution-level (school register) data.
1. Year of reference differs from 2024.
2. Average number of students per grade is estimated from enrolment by education level divided by the typical number of grades in that level, as information on the grades offered by individual schools is not available.
For data, see Table B2.3. The data for this figure can be accessed via https://stat.link/n9wxcm.
The degree of variation in school size differs considerably across education systems. Countries such as Poland and the Slovak Republic display relatively narrow distributions, indicating that students are likely to attend schools of broadly similar size regardless of where they live. Such homogeneity suggests that educational provision is organised in a relatively consistent manner across the country, with students experiencing comparable institutional environments. Conversely, other countries exhibit substantial disparities in school size. In England (United Kingdom), the 25th percentile corresponds to 49 students per grade while the 75th percentile reaches 215 students per grade. Korea shows a similarly wide distribution, ranging from 29 to 195 students per grade. These large spreads indicate that students within the same education system may experience very different schooling environments depending on the institution they attend. While some students are educated in relatively small schools characterised by closer-knit cohorts, others attend much larger institutions serving several times as many students per grade. Such differences in scale may reflect differences in settlement patterns, population density and school network organisation (Figure B2.6)
The coexistence of small and large institutions within the same education system does not necessarily indicate unequal provision but rather reflects the diversity of educational contexts in which students learn. The policy challenge is therefore not to eliminate variation in school size, but to ensure that students have access to high-quality learning environments and comparable educational opportunities regardless of the type of institution they attend (Ares Abalde, 2014[4])
Heat can affect students’ health, well-being and educational outcomes. Several studies have linked high temperatures with lower performance on cognitive and curricular tests as well as lower school engagement. One study using Programme for International Student Assessment (PISA) data across multiple countries found that higher temperatures reduce test performances among school-age children. Each additional day above 26.7°C during the three years preceding an exam lowered PISA test scores by 0.18% of a standard deviation (Park, Behrer and Goodman, 2021[19]). At the extremes, high temperatures can also lead to school closures, such as the 2025 heatwave that forced authorities to cancel classes in France (FRANCE 24, 2025[20]) and an extreme heat episode in Mexico in 2023 that led to the school closures and restrictions on outdoor exercise (Unicef, 2023[21]).
By 2050, primary school students will be exposed to more days with high temperatures at school (Figure B2.7). Even under a middle-of-the-road climate scenario, by 2050 Bulgaria and Chile are projected to join Colombia, Costa Rica and the United States on the list of countries where over one-quarter of primary school students attend schools experiencing more than 60 hot days a year. The effects will be different across OECD countries. In Hungary, the share of students exposed to more than 30 hot days per year will increase more than fourfold, from 19% during the 1983-2016 reference period to 96% in 2050, with the typical student facing 49 hot days in a year. The impact is less marked in countries like Denmark, Ireland, Latvia and Lithuania, where no primary school students will be exposed to more than 30 hot days per year. By 2050, even under the high-emissions scenario, more than a half of Danish primary school students will experience 1-5 hot days in school per year and those in Ireland will not experience any (OECD, 2026[22]).
Countries currently implement preventative measures to protect children from episodes of extremely high temperatures at school while ensuring the quality of their learning. In the Australian state of Queensland, the Department of Education recommends schools to modify or suspend outdoor activities, make cooler indoor spaces at school more accessible, ensure students stay hydrated and use simple cooling techniques to reduce heat stress (Queensland Government, 2024[23]). In Spain, Madrid’s Action Plan for Heat Waves foresees adjusting timetables, for example by moving classes into the cooler early morning hours and scheduling physical education classes during cooler hours (Comunidad de Madrid, 2026[24]). Although school closures can help manage the effects of extreme heat on student well-being, they risk aggravating the learning loss from high temperatures. Therefore, such measures should remain a last resort unless accompanied by compensatory measures like pre-planned contingency days. This is currently a common practice in the US state of Rhode Island where the school calendar must include four contingency days in the events extreme heat, extreme cold/snow, and health and safety concerns like illness or building conditions. (Rhode Island Department of Education, 2025[25]).
Share of primary students enrolled in schools exposed to days of 30°C or over
Note: Schools are grouped according to the number of hot days experienced in the calendar year. The share of students is determined using each school’s enrolment numbers for 2024.
Source: (OECD, 2026[26])) Using climate projections to assess increasing student exposure to high temperatures, https://doi.org/10.1787/98a7f132-en.
Countries also adapt school infrastructure to cope with extreme temperatures. Investment in heating, ventilation and air conditioning systems is a common way for governments to provide thermal comfort for children in schools, although the cost can be a major challenge. For example, Japan successfully equipped almost all regular classrooms in public primary and lower secondary schools with air conditioning in 2022 (Government of Japan, 2024[27]; Akesaka and Shigeoka, 2025[28]). This investment was partly financed through public-private-partnerships in which local governments entered long-term contracts with private companies that installed and maintained school air conditioning systems in return for service fees (Sekartaji et al., 2023[29]). Efforts have also been made in European countries including France, where the government aims to upgrade the infrastructure of 40 000 primary schools during the next 10 years with an initial investment of EUR 2 billion. The plan involves energy upgrades, insulation and transitioning to geothermal systems, among other improvements (Cerema, 2024[30]).
Educational institutions are entities that have the provision of instructional services to individuals, or related clinical, community, or research services, as their sole or main purpose.
Integrated structures (or basic education schools) are institutional frameworks where students typically progress through ISCED levels 1 and 2 within a single institutional cycle or on a shared school site. In contrast, segmented systems involve institutional transitions between education levels, sometimes as early as age 10.
OECD small regions refer to Territorial Level 3 (TL3) regions, which are subnational administrative or statistical units below the national level. The OECD Territorial Classification distinguishes between two regional levels: large regions (Territorial Level 2, TL2) and small regions (Territorial Level 3, TL3). These regions are used to support internationally comparable analyses of demographic, economic and social developments within countries. Further information on the OECD regional classification and regional statistics is available in the OECD Regions and Cities Database (OECD, n.d.[8]).
Average school size is measured as the number of students per grade. This is calculated by dividing the total number of students enrolled at a given level of education by the number of grades included in that level within the institution.
The analysis in this chapter draws on two complementary data sources that provide information on the number and characteristics of educational institutions.
System-level (aggregate) data are based on the OECD-INES ad-hoc data collection on the number of educational institutions with ICT services, basic hygiene facilities and adapted infrastructure by level of education – all programmes (general and vocational), conducted in November 2025.
Institution-level (school register) data are based on the OECD-INES Data Collection on Schools 2026, conducted in January 2026. These data are derived from national school registers and provide information at the individual institution level.
The two sources complement one another. System-level data provide broad coverage and support trend analyses over time, while institution-level data offer a more granular picture of the organisation of school provision and the distribution of schools within countries. Where both sources are available, the results are presented separately to ensure transparency regarding differences in data collection methods and coverage.
References
[28] Akesaka, M. and H. Shigeoka (2025), “Hotter days, wider gap: The distributional impact of heat on student achievement”, RIETI Discussion Paper Series, No. 25-E-024, The Research Institute of Economy, Trade and Industry, https://www.rieti.go.jp/jp/publications/dp/25e024.pdf.
[17] Andrews, M., W. Duncombe and J. Yinger (2002), “Revisiting economies of size in American education: Are we any closer to a consensus?”, Economics of Education Review, Vol. 21/3, pp. 245-262, https://doi.org/10.1016/s0272-7757(01)00006-1.
[4] Ares Abalde, M. (2014), “School size policies: A literature review”, OECD Education Working Papers, No. 106, OECD Publishing, Paris, https://doi.org/10.1787/5jxt472ddkjl-en.
[30] Cerema (2024), Rénovation des écoles : un guide à destination des élus locaux, https://www.cerema.fr/fr/actualites/renovation-ecoles-guide-elus-locaux (accessed on 18 November 2025).
[14] Chapman, C. and D. Muijs (2014), “Does school-to-school collaboration promote school improvement? A study of the impact of school federations on student outcomes”, School Effectiveness and School Improvement, Vol. 25/3, pp. 351-393, https://doi.org/10.1080/09243453.2013.840319.
[24] Comunidad de Madrid (2026), Plan de Actuación ante Episodios de Altas Temperaturas, Comunidad de Madrid, https://www.comunidad.madrid/docs/2026-05/web-plan-contra-el-calor_cm_2026.vf_.pdf.
[3] Di Cataldo, M. and G. Romani (2024), “Rational cuts? The local impact of closing undersized schools”, Regional Science and Urban Economics, Vol. 109, https://doi.org/10.1016/j.regsciurbeco.2024.104057.
[20] FRANCE 24 (2025), France shuts schools as temperatures peak amid Europe-wide heatwave, https://www.france24.com/en/environment/20250701-france-braces-for-peak-temperatures-amid-europe-wide-heatwave.
[27] Government of Japan (2024), Installation of air conditioning equipment in public school facilities, Ministry of Education, Culture, Sports, Science, and Technology, https://www.mext.go.jp/a_menu/shotou/zyosei/mext_01278.html (accessed on 7 October 2024).
[13] Lehtonen, O. (2021), “Primary school closures and population development – is school vitality an investment in the attractiveness of the (rural) communities or not?”, Journal of Rural Studies, Vol. 82, pp. 138-147, https://doi.org/10.1016/j.jrurstud.2021.01.011.
[18] Leithwood, K. and D. Jantzi (2009), “A review of empirical evidence about school size effects: A policy perspective”, Review of Educational Research, Vol. 79/1, pp. 464-490, https://doi.org/10.3102/0034654308326158.
[26] OECD (2026), Using climate projections to assess increasing student exposure to high temperatures, OECD Publishing, Paris, https://doi.org/10.1787/98a7f132-en.
[22] OECD (2026), “What are the likely impacts of rising temperatures on students and how are countries adapting?”, Education Indicators in Focus, OECD Publishing, Paris, https://doi.org/10.1787/960244d5-en.
[11] OECD (2025), Education at a Glance 2025: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/1c0d9c79-en.
[10] OECD (2025), “Shrinking smartly and sustainably in Finland: Multi-level governance, subnational finance and service delivery”, OECD Regional Development Papers, No. 166, OECD Publishing, Paris, https://doi.org/10.1787/4fb53cc8-en.
[6] OECD (2024), Education at a Glance 2024: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/c00cad36-en.
[16] OECD (2024), “How are demographic changes affecting education systems?”, Education Indicators in Focus, No. 87, OECD Publishing, Paris, https://doi.org/10.1787/158d4c5c-en.
[15] OECD (2024), PISA 2022 Results (Volume III): Creative Minds, Creative Schools, PISA, OECD Publishing, Paris, https://doi.org/10.1787/765ee8c2-en.
[1] OECD (2018), Responsive School Systems: Connecting Facilities, Sectors and Programmes for Student Success, OECD Reviews of School Resources, OECD Publishing, Paris, https://doi.org/10.1787/9789264306707-en.
[8] OECD (n.d.), OECD Regions and Cities Database, http://oe.cd/geostats.
[5] OECD/Eurostat/UNESCO Institute for Statistics (2015), ISCED 2011 Operational Manual: Guidelines for Classifying National Education Programmes and Related Qualifications, OECD Publishing, Paris, https://doi.org/10.1787/9789264228368-en.
[19] Park, R., A. Behrer and J. Goodman (2021), “Learning is inhibited by heat exposure, both internationally and within the United States”, Nature Human Behaviour, Vol. 5, pp. 19-27, https://doi.org/10.1038/s41562-020-00959-9.
[2] Perico E Santos, A. (2023), “Managing student transitions into upper secondary pathways”, OECD Education Working Papers, No. 289, OECD Publishing, Paris, https://doi.org/10.1787/663d6f7b-en.
[23] Queensland Government (2024), Managing excessive heat in schools, https://education.qld.gov.au/students/student-health-safety-wellbeing/student-health/managing-excessive-heat-schools.
[25] Rhode Island Department of Education (2025), Calendar Guidance: Guidance for Establishing School Calendars and Length of the School Day, Rhode Island Department of Education, https://ride.ri.gov/sites/g/files/xkgbur806/files/2025-04/SchoolCalendarGuidance_v250411.pdf.
[29] Sekartaji, D. et al. (2023), “Energy-use and indoor thermal performance in junior high school building after air-conditioning installation with the private finance initiative”, Buildings, Vol. 13/2, p. 455, https://doi.org/10.3390/buildings13020455.
[9] Shewbridge, C. et al. (2016), OECD Reviews of School Resources: Lithuania 2016, OECD Reviews of School Resources, OECD Publishing, Paris, https://doi.org/10.1787/9789264252547-en.
[21] Unicef (2023), Las olas de calor, agravadas por el cambio climático, afectan la garantía del derecho a la educación de la niñez y adolescencia en México, https://www.unicef.org/mexico/comunicados-prensa/las-olas-de-calor-agravadas-por-el-cambio-clim%C3%A1tico-afectan-la-garant%C3%ADa-del (accessed on 7 October 2025).
[7] Varsik, S. (2025), “Transitions through education and into the labour market: Exploring educational outcomes and equity gaps using international surveys”, OECD Education Working Papers, No. 324, OECD Publishing, Paris, https://doi.org/10.1787/18be3c52-en.
[12] World Bank (2025), Portugal’s School Clustering Reform:Lessons for Georgia for Improving Efficiency, Equity, and Quality through School-Network Restructuring, World Bank, Washington, DC, http://documents.worldbank.org/curated/en/099112525193042171.
The data for the figures and tables in this chapter can be downloaded via https://stat.link/n9wxcm.
To access further data and/or other education indicators, please visit the OECD Data Explorer: http://data-explorer.oecd.org/s/4s
Data cut-off for the print publication 17 June 2026. Please note that the Data Explorer contains the most recent data.
a – category not applicable; b – break in series; d – contains data from another column; m – missing data; x – contained in another column (indicated in brackets). For further control codes, see the Reader’s Guide.
For further methodological information, see Education at a Glance 2026: Sources, Methodologies and Technical Notes (https://doi.org/10.1787/dcf64a14-en)
Based on system-level (aggregate) data
1. Year of reference differs from 2015: 2016 for Colombia and Costa Rica; and 2017 for Lithuania.
2. Year of reference differs from 2024: 2023 for Costa Rica.
3. Changes in the typical age of enrolment following education system reforms affect the comparability of trend estimates over time.
The data for this table can be accessed via https://stat.link/n9wxcm.
Based on system-level (aggregate) data
1. Year of reference differs from 2015: 2016 for China, Colombia, Costa Rica, India, Indonesia, Peru and Saudi Arabia; and 2017 for Lithuania and Sweden.
2. Year of reference differs from 2024: 2023 for Costa Rica and South Africa.
The data for this table can be accessed via https://stat.link/n9wxcm.
Based on institution-level (school register) data
Note: Columns showing the average number of students per grade in primary and lower secondary schools are available for consultation online.
1. Year of reference differs from 2024: 2023 for Brazil, Bulgaria, Chile, Colombia, Costa Rica, Hungary and Spain; and 2025 for Estonia, Israel, Latvia, Norway, Peru, Poland, Slovenia and the Slovak Republic.
2. Average number of students per grade is estimated from enrolment by education level divided by the typical number of grades in that level, as information on the grades offered by individual schools is not available.
The data for this table can be accessed via https://stat.link/n9wxcm.