Between 2013 and 2022, the ratio of children to teaching staff at pre-primary level fell across most countries, from 16:1 to 15:1 on average in OECD countries, due to fewer enrolled children and more teachers. In some countries, however, the ratio has increased due to rising child enrolment and teacher shortages.
On average, class sizes in public primary schools are slightly larger than those in private schools, at 21 students per class in public schools and 20 in private ones. Overall, class sizes have remained constant in both primary and lower secondary education across OECD countries between 2013 to 2022, with variations seen in individual countries.
The size of primary schools varies significantly both across and within countries. In general, however, across all countries, metropolitan regions tend to have a lower concentration of small primary schools, while non-metropolitan regions tend to have a higher one.
Education at a Glance 2024
Class sizes and student-teacher ratios are much-discussed aspects of education and are among the determinants of the demand for teachers, along with students’ instruction time (see Chapter D1), teachers’ working time and the division of teachers’ time between teaching and other duties (see Chapter D4). Together with teachers’ salaries (see Chapter D3) and instruction time, class sizes and student-teacher ratios also have a considerable impact on the level of current expenditure on education through teacher salary costs.
The ratio of students to teaching staff is an indicator of how resources for education are allocated. Smaller student-teacher ratios often have to be weighed against measures such as higher salaries for teachers, investment in their professional development, greater investment in teaching technology, or more widespread use of assistant teachers whose salaries are often considerably lower than those of teachers.
Smaller classes are often seen as beneficial, because they allow teachers to focus more on the needs of individual students and reduce the amount of class time needed to deal with disruptions. Yet, while there is some evidence that targeted and intensive implementation of smaller classes has shown promise in narrowing scholastic performance gaps (Bouguen, Grenet and Gurgand, 2017[1]), overall evidence of the effect of class size on student performance is mixed (OECD, 2016[2]). Changes in class size over time can also highlight potential imbalances in the supply of teachers compared to student demand. Some countries face difficulties in recruiting new teachers to respond to a growing student base, while others face the opposite problem of adjusting the overall number of teachers as enrolments decline (OECD, 2019[3]).
Schools are the central organisational unit of education systems and the environment in which students spend most of their time in education. They differ from each other in their size, organisation and available resources. This is reflected in basic school attributes, such as class sizes and the range of subjects that are offered, and affects important policy-relevant outcomes, such as learning outcomes, the costs of education and student well-being.
Policy makers face important decisions on how many schools to operate and where to locate them, particularly in rural contexts. These decisions involve trade-offs between accessibility, quality and the cost of providing education. Although the geography of a country affects where schools are located, there are still large differences in school sizes across and within countries, even in urban settings where accessibility to nearby schools is generally less of a concern due to higher population densities. Such variation indicates that countries differ in how they approach policy choices around school size.
Across OECD countries with available data in 2022, there were an average of 14 students per teacher at the primary level, falling to 13 students per teacher at the lower secondary level.
Median student-teacher ratios in primary schools in metropolitan regions are generally higher than those in non-metropolitan regions across 9 countries with available school-level data, but the distributions of student-teacher ratios exhibit a diverse pattern within and across countries.
Student-teacher ratios, class sizes and school sizes measure very different characteristics of the educational system. Student-teacher ratios provide information on the level of teaching resources available in a country relative to its student population and serve as a pivotal indicator reflecting the human resources allocated, whether directly or indirectly, to children's education. This ratio is of importance from both administrative and economic standpoints as it is closely related to the amount of money spent per student. In contrast, class sizes measure the average number of students that are grouped together in a classroom, which has greater significance from a psychological standpoint and is a more direct measure of the teaching resources brought to bear on a student’s development. Meanwhile, school sizes measure the theoretical size of a specified level of education per school. Given these differences, it is possible for countries with similar student-teacher ratios to have different class sizes, or for countries with similar class sizes have different school sizes. A school can have small average class sizes but a large number of classes, resulting in a large school in terms of the number of students enrolled.
Note that the indicators discussed in the main analysis are calculated at the national level, whereas those in the boxes are calculated based on the school-level data collected through the ad-hoc survey on primary schools.
Early childhood education and care (ECEC) profoundly influences children's educational, cognitive, behavioural and social development both in the short and long term. It plays a crucial role in bridging the academic achievement gap between children from disadvantaged backgrounds and their peers.
The staffing landscape in ECEC is diverse. Those who have regular and direct contact with children, and whose roles involve education and/or care, can generally be categorised into four main groups: classroom teachers, teachers’ aides, school-level management personnel, and staff providing pedagogical, health or social support. The qualifications and compensation structures for these staff members often vary based on their distinct roles.Typically, each OECD country has between 2 to 10 types of ECEC staff, which can be grouped into these 4 categories. For instance, in Austria's ECEC system, there are supportive specialists who are pedagogical staff and facilitate group activities. In Latvia, there are teacher assistants who oversee childcare and support teachers in facilitating learning. The Republic of Türkiye has counsellors, typically holding bachelor's degrees in psychological counseling and guidance, who operate in ECEC centres with high levels of enrolment.
Although classroom teachers are traditionally regarded as the core practitioners in ECEC, there is growing recognition of the invaluable contributions made by auxiliary staff. The research literature highlights that assistant teachers play a crucial role in children’s development by facilitating learning, bridging gaps and providing caring support in various scenarios. (Van Laere, Peeters and Vandenbroeck, 2012[4]; Figueras-Daniel and Li, 2021[5]; Mowrey and Farran, 2021[6]; Webster and De Boer, 2019[7]). Teachers also benefit from the availability of support from other staff. Centre-based ECEC teachers and home-based providers who experience supportive relationships with co-workers and supervisors report less stress and depression (Smith and Lawrence, 2019[8]). Therefore, the composition of ECEC staffing has significant policy implications, not just for optimising teaching teams to deliver high-quality education to young children but also for strategic human resource allocation within the ECEC sector.
Child-staff ratios and child-teacher ratios are important indicators of the resources devoted to education. Smaller ratios are often seen as beneficial because they allow staff to focus more on the needs of individual children and reduce the amount of time spent addressing disruptions. Regulating these ratios can therefore be used to improve the quality of early childhood education (ECE). On average in OECD countries, there are 15 children for every teacher working in pre-primary education in 2022, with wide variations across countries. The ratio of children to teaching staff, excluding teachers’ aides, ranges from 6 or fewer children per teacher in New Zealand to more than 30 in Colombia and the United Kingdom (Table D2.1).
Lower child-staff ratios are particularly important for high-quality interactions with children under 3 (COFACE, 2023[9]). With the exception of Indonesia, Lithuania, and Romania, the child-to-teacher ratio in early childhood educational development services (ISCED 01) is consistently lower than for pre-primary education (ISCED 02) across all OECD member, partner and/or accession countries. On average across OECD countries, there are 9 children for every teacher working in early childhood educational development services, ranging from 3 in New Zealand to 30 in the United Kingdom (Table D2.1).
Sensitive and responsive interactions between adults and children, enabled by lower child-staff ratios, bring benefits to both children and staff. Children develop personalised relationships with ECE staff in a stimulating environment, while staff benefit from good-quality working conditions, which are in turn linked to stable relationships between children and practitioners as well as low staff turnover rates (COFACE, 2023[9]).
Low child-staff ratios may offer opportunities for stronger partnerships between parents and ECE staff. Having fewer children to take care of during the day allows caregivers and teachers more time to discuss children’s activities and communicate and develop relationships with parents. This, in turn, can determine the quality of relationships between educators and children (COFACE, 2023[9]).
Some countries, such as Lithuania and the Netherlands, also make extensive use of teachers’ aides, as can be seen from the smaller ratios of children to contact staff compared to teaching staff. In most cases, early childhood development services and pre-primary education have similar shares of teachers’ aides among contact staff, with differences of less than 9 percentage points. In Chile, however, the share of teachers’ aides at pre-primary level is nearly twice that in early childhood development services, while the ratio of children to teaching staff (20:1) is well above the OECD average of 15:1 (Table D2.1).
Between 2013 and 2022, the ratio of children to teaching staff at pre-primary level fell across most countries, from 16:1 to 15:1 on average in OECD countries. In most of these cases, the drop is due to the number of teachers growing faster than the number of children enrolled in pre-primary education (Figure D2.1). In Belgium, Mexico, New Zealand and Romania, the number of teachers increased despite a fall in the number of children enrolled over the period. Conversely, in Italy, Japan and Korea, both the number of children enrolled and the number of teachers fell at pre-primary level, but the drop in the number of children was greater than the drop in the number of teachers. The common factor in both scenarios is the overall fall in the number of children enrolled, which may be partly attributed to a general reduction in the population of 0-6 year-olds during this period.
In contrast to the prevailing downward trend across OECD countries, the child-teacher ratio increased by an average of 14% per year in Saudi Arabia between 2013 and 2022. This was a combined effect of both an increase in the number of children enrolled in pre-primary educaion and a fall in the number of teachers (Figure D2.1). The increase in student enrolment can be largely attributed to increased participation among children of pre-primary age during the period (see Table B1.2 in Chapter B1). In parallel, Saudi Arabia faces a significant challenge due to a shortage of qualified kindergarten teachers (OECD, 2020[10]).
On average across OECD countries, there are 14 students for every teacher at primary level, ranging from 8:1 in Greece and Luxembourg to 24:1 in Mexico. On average, there are fewer students per teacher at lower secondary level (13:1) than at primary level (14:1). This reduction in the student-teacher ratio between primary and lower secondary education may be due to differences in annual instruction time (as instruction hours tend to increase with the education level, so does the number of teachers (OECD, 2023[11])) or from differences in teaching hours (teaching time falls with the level of education as teacher specialisation increases (see Chapter D4)). Only Chile, Colombia, Costa Rica, Luxembourg, Mexico, the Slovak Republic and the United States have a larger student to teacher ratio at lower secondary level than at primary level (Figure D2.2). Additionally, the student-teacher ratios exhibit regional variations within countries (Box D2.1).
As student-teacher ratios vary across countries, they also differ across subnational regions within countries. Among the eight OECD countries with available primary school-level data, the median student-teacher ratios in primary schools in metropolitan regions are generally higher than those in non-metropolitan regions (Figure D2.3). Notably, in Korea, the median in metropolitan regions is 7 students more per teacher than in non-metropolitan regions. In contrast, in the Slovak Republic, the difference is smaller, with metropolitan regions having 2 students more per teacher. Hungary and Lithuania show little difference in student-teacher ratios between metropolitan and non-metropolitan regions (Table D2.4).
Note: The numbers in square brackets after the country names [a;b] indicates the number of schools, where 'a' is the number of schools in metropolitan regions and ‘b’ the number in non-metropolitan regions.
Countries are ranked in descending order of the median ratio in all regions.
See Table D2.4. for data and under Chapter D2 Tables section for StatLink. For more information see Source section and Education at a Glance 2024 Sources, Methodologies and Technical Notes (https://doi.org/10.1787/e7d20315-en).
The distributions of student-teacher ratios fall within similar ranges across both metropolitan and non-metropolitan regions in Hungary and Poland (Figure D2.3). However, in Korea, the student-teacher ratio varies more widely in non-metropolitan primary schools than those in metropolitan regions. In non-metropolitan regions of Korea, 25% of schools have 4 or fewer students per teacher, while the schools in the top 25% have 14 or more students per teacher (Table D2.4).
Maintaining low student-teacher ratios in a large number of primary schools in rural areas can put considerable pressure on public resources, especially in systems where central funding for schools is tied to the number of students they serve. A low student-teacher ratio may require municipalities or other subnational levels of government to contribute more funds towards maintaining small local schools (OECD, 2018[12]). One primary contributing factor could be the high salary costs of teachers, which constitute the largest portion of education expenditures.
Low student-teacher ratios are crucial for fostering close teacher-student relationships, catering to individual learning needs effectively, and cultivating a supportive social environment conducive to student development (Ares Abalde, 2014[13]), which is particularly essential in disadvantaged rural areas. Nevertheless, the literature underscores unique challenges faced by small rural schools with low student-teacher ratios. Although the average number of students per teacher is lower, teachers often devote a significant amount of time to tasks other than teaching and class preparation due to the absence of additional administrative or professional support (Creagh et al., 2023[14]). Moreover, rural schools encounter difficulties in recruiting teachers for certain subjects and in preparing them with the necessary skills to teach effectively within rural contexts (Monk, 2007[15]). The evidence also highlights a low level of geographical mobility among teachers, with location preferences significantly influencing their job searches (Reininger, 2012[16]; Engel, Jacob and Curran, 2014[17]). This strong regional dimension of the teacher labour market can create supply shortages for schools in both rural and disadvantaged urban areas, which may further amplify regional disparities.
At the upper secondary level, the student-teacher ratio is about 13 students per teacher on average, which is slightly lower than in lower secondary education. However, at this level, the ratio can vary dramatically depending on the programme orientation. In about 40% of OECD countries with data, the ratio is greater in vocational programmes than in general ones. In Colombia, there are about 30 more students per teacher in vocational programmes than in general ones, while the difference is 9 more in Latvia and 6 more in Denmark. In other countries, such as Brazil and Mexico, the difference is reversed: there are over 9 more students per teacher in general programmes (Table D2.2). These differences could be attributed to a combination of the different nature of general and vocational programmes, and differences in country policies.
At tertiary level, there are on average 16 students per teaching staff member. The student-teacher ratio ranges from 4:1 in Luxembourg to 27:1 in Brazil and Indonesia (Table D2.2). The difference in student-teaching staff ratios across different levels of tertiary education also varies across countries with available data. These results should be interpreted with caution, however, as the student-teacher ratio remains a limited measure of the level of teaching resources at tertiary level, where research staff may make up a significant share of academic personnel. Moreover, the relatively low enrolment in short-cycle tertiary in some countries also limits comparability across tertiary levels.
At higher levels of education, students are often split into several different classes, depending on the subject area. This makes class sizes difficult to define and compare at these levels. Therefore, the indicator on class size is limited to primary and lower secondary education.
At primary level, the average class in OECD countries has 21 students. There are fewer than 25 students per class in all the countries with available data, with the exception of Chile, Israel, Japan, and the United Kingdom. At lower secondary level, the average class in OECD countries has 23 students. Among all countries with available data, the number ranges from fewer than 20 students per class in Croatia, Estonia, Finland, Latvia, Luxembourg and Poland to 32 students or more per class in Chile, and Japan (Table D2.3).
The number of students per class tends to increase from primary to lower secondary education, which is the opposite trend to that seen with student-teacher ratios. In Costa Rica, the number of students per class increases by 15 students. On the other hand, in Australia and, to a lesser extent, Hungary and the United Kingdom, it falls between these two levels of education (Table D2.3) between 2013 and 2022, class sizes remained constant at primary and lower secondary levels on average across OECD countries, but this average masks considerable changes in individual countries. At primary level, class sizes fell by 3 students in Brazil on average and increased by 4 students in Mexico over this period. At lower secondary level, some changes have been even more striking, with average class size falling by 7 students in Korea and increasing by 4 in the United Kingdom (Table D2.3.).
Students in primary and lower secondary education are often assigned to their neighbourhood school. However, in recent decades, reforms in many countries have tended to give parents (and students) greater choice, enabling families to choose a school that meets their child’s education needs or family preferences. When choosing between public and private institutions, parents evaluate a multitude of factors to determine the best educational environment for their children. Class size is one important factor as it allows for more individualised attention, better support and tailored instruction, enhancing the learning experience. However, other criteria such as the quality of education, school facilities and extracurricular activities are equally important. Practical considerations, such as location and cost, also play vital roles in the decision-making process, alongside the school’s reputation and parental involvement.
At primary level, the average class size stands at 21 students in public institutions, while private institutions tend to have slightly smaller class sizes, of around 20 students per class across OECD countries in 2022 (Figure D2.4). This pattern persists at lower secondary level, where public institutions typically have 23 students per class compared to 21 students in private institutions (Table D2.2). On average across OECD countries, class size remained constant in both public and private lower secondary institutions between 2013 and 2022 (Table D2.3). This again masks more substantial changes in individual countries. For example, Estonia had some of the smallest average class sizes in both public and private institutions in 2013 and, although they still remain below the OECD average in 2022, they have increased over the period. Other countries, such as Korea, which had the highest average class size in 2013, saw class sizes fall by 2022, both for public and private institutions (Table D2.3).
Across countries with available data, the difference in average class sizes between government-dependent private institutions and independent private institutions remains consistent across primary and lower secondary education. In Chile, Costa Rica, France, Peru and the United Kingdom, independent private institutions maintain an average class size at least 9 students smaller than that of government-dependent private institutions at lower secondary in 2022 (Table D2.3).
Both across and within countries, primary schools vary widely in size, as measured by the number of students enrolled per primary grade. Notably, Australia, Korea and the United States stand out with the largest median primary school sizes. In Korea, the median primary school has 52 students per grade, with 5% of primary schools having 197 or more students per grade. Although the number of such large primary schools is limited, they are attended by a large number of students. Conversely, countries including Brazil, Costa Rica, Croatia, Estonia and Ireland show a notable peak in the density of schools with fewer students per grade, indicating a large number of smaller schools. In Costa Rica and Croatia, 25% of primary schools have 3 or fewer students per grade (Figure D2.3). These variations across and within countries can be attributed to factors such as school size policies, population density and resource allocation (Box D2.3).
Countries also differ in terms of how much school sizes vary within the country. In Belgium, Luxembourg and the Netherlands school sizes are concentrated within a narrow range. For example, in the Netherlands, the interquartile range spans from 15 to 36 students per grade (i.e. 50% of all schools fall within this range while 25% of schools have fewer students and 25% of schools have more students). In contrast, the interquartile range spans from 15 to 62 students per grade in Australia and 10 to 105 students per grade in Korea (Figure D2.5)).
Research indicates school size has a mixed impact on academic achievement, with both small and large schools having different advantages (Scheerens, Hendriks and Luyten, 2014[18]; Ares Abalde, 2014[13]). School size interacts with some meditating factors such as socio-economic background, school climate, curricular offerings, extracurricular activities, self-esteem, student-teacher relationships and home-school relationships, all influencing student outcomes (McCathern, 2004[19]). Some evidence points towards an inverted U-shaped relationship between school size and achievement, suggesting there is an optimal size for maximizing student performance (Borland and Howsen, 2003[20]; Slate and Jones, 2005[21]). However, the impact varies by grade and subgroup, with larger schools being potentially more detrimental in higher grades, particularly for socio-economically disadvantaged students and those with learning disabilities (Egalite and Kisida, 2016[22]; Gershenson and Langbein, 2015[23]).
Students in smaller schools are more engaged in extracurricular activities and have a stronger sense of belonging. Close student-teacher relationships in these schools help address individual needs and reduce dropout rates (Bard, Gardener and Wieland, 2006[24]; Slate and Jones, 2005[21]; Ares Abalde, 2014[13]; Cotton, 1996[25]). Teachers in small schools experience greater co-operation and parental involvement (Cotton, 1996[25]; Slate and Jones, 2005[21]; Leithwood and Jantzi, 2009[26]; Nguyen, Schmidt and Murray, 2007[27]; Duncombe and Yinger, 2001[28]).
Conversely, larger schools are more cost-efficient, benefiting from economies of scale, which reduce expenses while maintaining or enhancing efficacy (Ares Abalde, 2014[13]). This cost efficiency follows a U-shaped relationship, where initial economies of scale become diseconomies above a critical size (Slate and Jones, 2005[21]). Expenditure per student is highest in the smallest schools (Falch, Rønning and Strøm, 2008[29]; Andrews, Duncombe and Yinger, 2002[30]), which may be due to fixed costs associated with administrative work and constructing, running and maintaining school facilities. Larger schools also offer a more diverse curriculum and avoid multi-grade teaching (Ares Abalde, 2014[13]). They also allow for greater investment in administrative staff, reducing the administrative burden on teachers and allowing them to focus more on teaching (Bradley and Taylor, 1998[31]).
1. As per the Australian Education Act 2013, a primary school is defined as small if it has between 15 and 200 students.
Countries are listed in alphabetical order.
See Table D2.4 for data and under Chapter D2 Tables section for StatLink. For more information see Source section and Education at a Glance 2024 Sources, Methodologies and Technical Notes (https://doi.org/10.1787/e7d20315-en).
As Box D2.2 shows, primary schools vary significantly in size across countries. This variation is also found within countries with regional differences Figure D2.6 illustrates the share of small primary schools in subnational regions in countries with available school-level data. A small school is defined as having 21 or fewer students per grade at primary level, which is the average class size in primary schools across OECD countries (Table D2.3).
The distribution of small primary schools across the 27 countries with data available shows a highly diverse pattern. In the United States, for example, less that 10% of primary schools in southeastern and southwestern regions are small, while central northern regions have a higher share. In France, large primary schools are concentrated in and around Paris, as well as the southern coastal regions, such as Bouches-du-Rhône. These regions are typically metropolitan. Conversely, in inland rural regions that are remote from metropolitan areas, such as Cantal, Creuse and Lozère, the share of small schools exceeds 90%. In contrast, the share of small schools is lower even in very sparsely populated regions in the northern part of the Nordic countries, as well as in northern Australia. This may be explained by settlement patterns, as population in those regions may be concentrated in mid-sized towns rather than small villages (Figure D2.6).
A general trend across all countries is that metropolitan regions tend to have a smaller share of small primary schools, while non-metropolitan regions tend to have a larger share. This pattern is also evident in Table D2.4, which shows that the median size of primary schools in metropolitan areas is larger than in non-metropolitan areas within each country with available school-level data. As discussed in Box D2.2, small rural schools face challenges due to their inefficiency related to their small size, including increased administrative workload for teachers and a less comprehensive educational offering. These limitations extend beyond academic issues to include supplementary services for struggling students and those with special learning needs in a rural context (OECD, 2018[12]).
Countries have developed school size policies based on their individual historical, geographical and socio-economic contexts. For instance, since the 1980s, Korea's education policy has aimed to maximise the efficiency of its education investments, leading to significant pressure to merge and close schools in rural regions. Schools with fewer than 180 students were advised to close, but many opted to become hub schools, linking two to four schools together, with one taking the lead in managing educational programmes and facilities (Ares Abalde, 2014[13]). This policy has softened since 2004, as the government has shifted its focus from maximising efficiency to improving the quality of education in rural areas. However, consolidation policies remain controversial: proponents argue that they ensure educational equity and resource efficiency through economies of scale and broader curricula. Opponents point to the negative impact on rural communities, where schools play crucial economic and social roles, symbolising autonomy, vitality, integration and tradition (Nitta, Holley and Wrobel, 2010[32]).
Despite their smaller average size, retaining rural schools is essential to ensuring every student has equitable access to educational resources. Additionally, since rural school networks are often less dense than those in urban or suburban areas, rural families have fewer options when choosing schools for their children. Although consolidation is often proposed to increase the size, resources and educational reach of the remaining rural schools, such measures may further reduce school diversity and limit parents’ ability to choose between multiple providers or programmes (OECD, 2018[12]; Gronberg et al., 2015[33]). Beyond their educational role, small rural schools serve additional important functions by helping to attract and retain residents in areas that often face challenges related to population decline.
1. As per the Australian Education Act 2013, a primary school is defined as small if it has between 15 and 200 students.
See under Chapter D2 Tables section for StatLink. For more information see Source section and Education at a Glance 2024 Sources, Methodologies and Technical Notes (https://doi.org/10.1787/e7d20315-en).
Early childhood education (ECE): ECEC services in adherence with the criteria defined in the ISCED 2011 classification (see ISCED 01 and 02 definitions) are considered early childhood education programmes and are therefore referred to as ECE in this chapter. Therefore, the term of ECE excludes the programmes that do not meet the ISCED 2011 criteria.
Teachers’ aides and teaching/research assistants include personnel or students who support teachers in providing instruction to students.
Teaching staff refers to personnel directly involved in teaching to students. The classification includes classroom teachers, special-education teachers and other teachers who work with a whole class of students in a classroom, in small groups in a resource room, or in one-to-one teaching situations inside or outside a regular class. At the tertiary level, academic staff include personnel whose primary assignment is instruction or research, or both. Teaching staff also include departmental chairs whose duties include some teaching, but exclude non-professional personnel who support teachers in providing instruction to students, such as teachers’ aides and other paraprofessional personnel.
Class size is defined as the number of students who are following a common course of study, based on the highest number of common courses (usually compulsory studies), and excluding teaching in subgroups.
School size refers to the number of students enrolled in a given level of education per grade.
The ratio of students to teaching staff is obtained by dividing the number of full-time equivalent students at a given level of education by the number of full-time equivalent teachers at that level and in similar types of institutions. At tertiary level, the student-teacher ratio is calculated using data on academic staff instead of teachers.
For the ratio of students to teachers to be meaningful, consistent coverage of personnel and enrolment data are needed. For instance, if teachers in religious schools are not reported in the personnel data, then students in those schools must also be excluded.
Class size is calculated by dividing the number of students enrolled by the number of classes. In order to ensure comparability among countries, special needs programmes are excluded. Data include only regular programmes at primary and lower secondary levels of education, and exclude teaching in subgroups outside the regular classroom setting.
In Education at a Glance 2024, school size analysis is restricted to primary level of education. School size is calculated by dividing the number of enrolled students in primary level by the theoretical duration of primary education in each country. For instance, the theoretical duration of primary education is 5 years in France. Therefore, the size of each school in the French dataset is calculated by dividing the number of enrolled students in primary education by 5, resulting in the number of enrolled students per grade in primary education. In cases where countries provide the information on the number of grades offered by each school, the school size (students per grade) is calculated by dividing the total number of students at primary level by the number of grades provided by the school. For example, if a school offers only grades 1 and 2 at the primary level, whereas the theoretical primary education duration in the country is 6 grades, the school size is calculated by dividing the total number of students enrolled in these two grades by 2.
In order to ensure comparability among countries, schools with zero students enrolled in primary education, schools that cannot be assigned to any TL3 region based on the OECD classification (Fadic et al., 2019[34]) and students enrolled in special education are excluded from the analysis. When the number of enrolled students is reported as "1-4" or "<5", the value is replaced with "2.5".
Following the OECD territorial grid and regional typologies, small (TL3) regions are classified based on their access to functional urban areas (FUAs) with over 50 000 inhabitants. According to this classification, TL3 regions fall into three groups: 1) metropolitan regions; 2) regions near a medium-sized or large FUA; and 3) regions far from a medium-sized or large FUA (Fadic et al., 2019[34]). In Education at a Glance 2024, these regions have been consolidated into two types of region: metropolitan and non-metropolitan (comprising regions near a medium-sized or large FUA and regions far from a medium-sized or large FUA). For further information, please refer to OECD Regions and Cities database at http://oe.cd/geostats.
Data refer to the reference year 2022 (school year 2021/22) and are based on the UNESCO-UIS/OECD/Eurostat data collection on education statistics administered by the OECD in 2024. For more information see Education at a Glance 2024 Sources, Methodologies and Technical Notes (https://doi.org/10.1787/e7d20315-en).
Data from Argentina, the People’s Republic of China, India, Indonesia, Saudi Arabia and South Africa are from the UNESCO Institute of Statistics (UIS).
Data on school size and student-teacher ratio at subnational level refer to the reference year 2023 (school year 2022/23) and were collected through a special survey undertaken in 2024. Data for some countries may have a different reference year. For more information see Education at a Glance 2024 Sources, Methodologies and Technical Notes (https://doi.org/10.1787/e7d20315-en).
References
[30] 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.
[13] 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.
[24] Bard, J., C. Gardener and R. Wieland (2006), “Rural school consolidation: History, research summary, conclusions, and recommendations”, The Rural Educator, Vol. 27/2, pp. 40-48, https://doi.org/10.35608/ruraled.v27i2.497.
[20] Borland, M. and R. Howsen (2003), “An examination of the effect of elementary school size on student academic achievement”, International Review of Education, Vol. 49/5, pp. 463-474, https://doi.org/10.1023/A:1026348922511.
[1] Bouguen, A., J. Grenet and M. Gurgand (2017), “Does class size influence student achievement?”, HAL Open Science, https://shs.hal.science/halshs-02522747/document.
[31] Bradley, S. and J. Taylor (1998), “The effect of school size on exam performance in secondary schools”, Oxford Bulletin of Economics and Statistics, Vol. 60/3, pp. 291-324, https://doi.org/10.1111/1468-0084.00102.
[9] COFACE (2023), “High-quality early childhood education and care: low children-to-staff ratio as a primary driver for children’s well-being and families’ engagement”, COFACE Thematic Note, COFACE Families Europe, https://coface-eu.org/wp-content/uploads/2023/12/COFACE_ECEC_ThematicNote-MAY-2023.pdf.
[25] Cotton, K. (1996), “School size, school climate, and student performance”, School Improvement Research Series (SIRS), No. 20, http://www.nwrel.org/scdp/sirs/10/c020.html (accessed on 5 March 2024).
[14] Creagh, S. et al. (2023), “Workload, work intensification and time poverty for teachers and school leaders: a systematic research synthesis”, Educational Review, pp. 1-20, https://doi.org/10.1080/00131911.2023.2196607.
[28] Duncombe, W. and J. Yinger (2001), “Does school district consolidation cut costs?”, Working Paper, No. 122, Center for Policy Research, Syracuse University, https://surface.syr.edu/cpr/122.
[22] Egalite, A. and B. Kisida (2016), “School size and student achievement: a longitudinal analysis”, School Effectiveness and School Improvement, Vol. 27/3, pp. 406-417, https://doi.org/10.1080/09243453.2016.1190385.
[17] Engel, M., B. Jacob and F. Curran (2014), “New evidence on teacher labor supply”, American Educational Research Journal, Vol. 51/1, pp. 36-72, https://doi.org/10.3102/0002831213503031.
[34] Fadic, M. et al. (2019), “Classifying small (TL3) regions based on metropolitan population, low density and remoteness”, OECD Regional Development Working Papers, No. 2019/06, OECD Publishing, Paris, https://doi.org/10.1787/b902cc00-en.
[29] Falch, T., M. Rønning and B. Strøm (2008), “A cost model of schools: School size, school structure and student composition”, in Governance and Performance of Education Systems, Springer Netherlands, Dordrecht, https://doi.org/10.1007/978-1-4020-6446-3_11.
[5] Figueras-Daniel, A. and Z. Li (2021), “Evidence of support for dual language learners in a study of bilingual staffing patterns using the Classroom Assessment of Supports for Emergent Bilingual Acquisition (CASEBA)”, Early Childhood Research Quarterly, Vol. 54, pp. 271-285, https://doi.org/10.1016/j.ecresq.2020.09.011.
[23] Gershenson, S. and L. Langbein (2015), “The effect of primary school size on academic achievement”, Educational Evaluation and Policy Analysis, Vol. 37, pp. 135-155, https://doi.org/10.3102/0162373715576075.
[33] Gronberg, T. et al. (2015), “School district consolidation: Market concentration and the scale‐efficiency tradeoff”, Southern Economic Journal, Vol. 82/2, pp. 580-597, https://doi.org/10.1002/soej.12029.
[26] 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.
[19] McCathern, D. (2004), The relationship between prek-5 and k-5 elementary school size and student achievement of grade 5 students on the MAT7 in South Carolina for the school years 1996-97 and 1997-98., Doctoral dissertation, University of South Carolina, Columbia.
[15] Monk, D. (2007), “Recruiting and retaining high-quality teachers in rural areas”, The Future of Children, Vol. 17/1, pp. 155-174, https://doi.org/10.1353/foc.2007.0009.
[6] Mowrey, S. and D. Farran (2021), “Teaching teams: The roles of lead teachers, teaching assistants, and curriculum demand in pre-kindergarten instruction”, Early Childhood Education Journal, Vol. 50/8, pp. 1329-1341, https://doi.org/10.1007/s10643-021-01261-7.
[27] Nguyen, H., M. Schmidt and C. Murray (2007), “Does school size matter? A social capital perspective”, Review Of Educational Policy Literature, Simon Fraser Unversity.
[32] Nitta, K., M. Holley and S. Wrobel (2010), “A phenomenological study of rural school consolidation”, Journal of Research in Rural Education, Vol. 25/2, pp. 1-19, https://jrre.psu.edu/sites/default/files/2019-08/25-2.pdf.
[11] OECD (2023), “How much time do students spend in the classroom?”, in Education at a Glance 2023: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/8008b7a1-en.
[10] OECD (2020), Education in Saudi Arabia, Reviews of National Policies for Education, OECD Publishing, Paris, https://doi.org/10.1787/76df15a2-en.
[3] OECD (2019), Working and Learning Together: Rethinking Human Resource Policies for Schools, OECD Reviews of School Resources, OECD Publishing, Paris, https://doi.org/10.1787/b7aaf050-en.
[12] OECD (2018), “Adapting the school network to changing needs in urban, rural and remote areas”, in Responsive School Systems: Connecting Facilities, Sectors and Programmes for Student Success, OECD Publishing, Paris, https://doi.org/10.1787/9789264306707-7-en.
[2] OECD (2016), PISA 2015 Results (Volume II): Policies and Practices for Successful Schools, PISA, OECD Publishing, Paris, https://doi.org/10.1787/9789264267510-en.
[16] Reininger, M. (2012), “Hometown disadvantage? It depends on where you’re from”, Educational Evaluation and Policy Analysis, Vol. 34/2, pp. 127-145, https://doi.org/10.3102/0162373711420864.
[18] Scheerens, J., M. Hendriks and H. Luyten (2014), “School size effects: Review and conceptual analysis”, in School Size Effects Revisited, Springer, https://doi.org/10.1007/978-3-319-06814-5_2.
[21] Slate, J. and C. Jones (2005), “Effects of school size: A review of the literature with recommendations”, Essays in Education, Vol. 13/1, https://openriver.winona.edu/eie/vol13/iss1/12 (accessed on 5 March 2024).
[8] Smith, S. and S. Lawrence (2019), “Early care and education teacher well-being: Associations with children’s experience, outcomes, and workplace conditions: A research-to-policy brief”, Child Care & Early Education Research Connections, National Center for Children in Poverty, https://doi.org/10.7916/d8-ngw9-n011.
[4] Van Laere, K., J. Peeters and M. Vandenbroeck (2012), “The Education and Care Divide: the role of the early childhood workforce in 15 European countries”, European Journal of Education, Vol. 47/4, pp. 527-541, https://doi.org/10.1111/ejed.12006.
[7] Webster, R. and A. De Boer (2019), “Teaching assistants: Their role in the inclusion, education and achievement of pupils with special educational needs”, European Journal of Special Needs Education, Vol. 34/3, pp. 404-407, https://doi.org/10.1080/08856257.2019.1615746.
Cut-off date for the data: 14 June 2024. Any updates on data along with more breakdowns can also be found at Education and Skills-OECD Data Explorer (database) http://data-explorer.oecd.org/s/5q.
Ratio of children to staff in full-time equivalents, by type of ECE service (public and private institutions)
Note: See under Chapter D2 Tables for StatLink and Box D2.4 for the notes related to this table.
Calculations based on full-time equivalents
Note: See under Chapter D2 Tables for StatLink and Box D2.4 for the notes related to this table.
Calculations based on number of students and number of classes
Note: See under Chapter D2 Tables for StatLink and Box D2.4 for the notes related to this table.
Average number of students enrolled per grade of primary education and student-teacher ratio at school level
Note: See under Chapter D2 Tables for StatLink and Box D2.4 for the notes related to this table.
Table D2.1 Ratio of children to staff in early childhood education (ECE), by level (2013 and 2022)
1. Year of reference differs from 2013: 2014 for Bulgaria, Croatia, Denmark and the United Kingdom; 2015 for Hungary, Italy, India and Romania; and 2016 for Colombia and France.
2. Children to teaching staff is overestimated as it includes part of 3 to 5-year-old children enrolled in early childhood educational development programmes.
3. Excludes data from independent private institutions (and government-dependent private institutions for teachers' aides).
4. Public institutions only.
5. Year of reference differs from 2022: 2018 for Indonesia; and 2021 for Peru.
Table D2.2. Ratio of students to teaching staff in educational institutions, by level of education (2022)
1. Data for short-cycle tertiary refer to the Flemish Community only.
2. The student teacher ratio is overestimated as it considers only teachers who teach vocational subjects, excluding those who teacher general ones.
3. Public institutions only.
4. Upper secondary vocational programmes include vocational programmes at the lower secondary, bachelor's and master's levels.
5. Year of reference differs from 2022: 2018 for Indonesia; and 2021 for Peru and South Africa.
Table D2.3. Average class size, by level of education and type of institution (2013 and 2022)
1. Year of reference differs from 2013: 2014 for Hungary and Sweden; and 2015 for Costa Rica and Switzerland.
Table D2.4. Distribution of primary school size (students per grade) and student-teacher ratio, by region type and percentile
1. Data do not include combined schools which provide both primary and secondary level of education.
2. Data represent only the schools in the Flemish Community of Belgium.
3. Data represent only public schools.
4. Data represent only publicly funded schools.
See Definitions and Methodology sections and Education at a Glance 2024 Sources Methodologies and Technical Notes (https://doi.org/10.1787/e7d20315-en) for more information.
Data and more breakdowns are available on the OECD Data Explorer (http://data-explorer.oecd.org/s/5q).
Please refer to the Reader's Guide for information concerning symbols for missing data and abbreviations.