Teacher workload cannot be reduced to the number of lessons shown on a timetable. A teacher may spend 18 hours in class yet work well beyond 40 hours once preparation, assessment, student support, meetings, administration and professional learning are counted. International comparisons therefore need several measures: annual statutory teaching time, contractual working time, time required at school and hours actually worked. The latest OECD evidence shows large differences between education systems, but it also shows why a simple league table can mislead. Countries assign different duties to teachers, count time differently and organise the school year in different ways.
The Main Numbers
- 55 education systems participated in OECD TALIS 2024, with responses from about 280,000 teachers and school leaders.
- Full-time lower secondary teachers reported 55.1 hours per week in Japan, compared with 29.7 hours in Kosovo among the selected systems shown below.
- Across OECD systems, classroom teaching represented about 43% of full-time teachers’ task-based workload; planning accounted for 14%, marking for 9% and administration for 6%.
- Official OECD rules for 2022/23 required an average of 773 teaching hours a year at primary level and 706 at lower secondary level.
What Counts As Teacher Working Time?
International datasets describe several different clocks. Statutory teaching time counts scheduled contact with a class, converted into 60-minute hours. Total statutory working time covers all hours expected under policy or contract, whether completed at school or elsewhere. Actual working time records what teachers really report or what administrative systems observe. These measures answer different questions and should not be placed in the same column without a clear label.
| Measure | What It Includes | Typical Unit | Main Limitation |
|---|---|---|---|
| Statutory teaching time | Scheduled contact with a class or group of students | Hours per year | Excludes preparation, marking and most other duties |
| Total statutory working time | All contractually expected work at school and elsewhere | Hours per year or week | May differ from actual time worked |
| Required presence at school | Teaching and non-teaching time that must occur on site | Hours per year or week | Does not capture work completed at home |
| Self-reported actual workload | All job-related tasks recalled for a recent week | Hours per week | Depends on recall and survey interpretation |
| Task share | Proportion of reported time assigned to each activity | Percentage | Task estimates may sum to more than the separately reported total |
| Work intensity | Frequency of high-speed work, pressure or tight deadlines | Survey scale or percentage | Measures pressure rather than duration |
The distinction matters because low classroom hours do not automatically mean a light job. A system may deliberately reserve time for joint planning, curriculum design, counselling or extracurricular activity. Another may set a high teaching load but assign fewer formal duties outside class. What does a short timetable reveal if preparation and assessment occur elsewhere? The visible timetable is only the face of the clock; much of the mechanism sits behind it.
The Latest International Evidence
OECD TALIS 2024, published in October 2025 and revised in December 2025, supplies the newest broad survey evidence on teachers’ actual working patterns. About 280,000 educators in 55 education systems took part. The central workload measure asks full-time lower secondary teachers how many hours they spent on all job-related tasks during their most recent complete calendar week.[a]
This is a self-reported weekly measure, not a payroll record and not a legal working-time limit. It includes work beyond classroom instruction, including planning, assessment, teamwork, administration and communication. TALIS also asks about individual tasks, but the OECD warns that the sum of task estimates can exceed the separately reported weekly total because respondents estimate each item independently.
Selected Weekly Totals From TALIS 2024
| Education System | Hours per Week | Position Among 55 Systems | Reading Note |
|---|---|---|---|
| Japan | 55.1 | 1 | Highest reported weekly total |
| New Zealand | 47.5 | 2 | Estimate carries a higher non-response caution |
| Singapore | 47.3 | 3 | Lower teaching hours, extensive non-teaching work |
| Shanghai, China | 46.9 | Joint 4 | Large planning and marking component |
| Alberta, Canada | 46.9 | Joint 4 | Subnational result; higher non-response caution |
| Italy | 32.7 | 49 | Formal non-teaching duties are not fully represented by a national total-hours rule |
| Türkiye | 31.1 | 51 | Lower reported planning and marking time |
| Kosovo | 29.7 | 53 | Among the lowest reported weekly totals |
The table is a selection, not a complete ranking. It is designed to show the spread and the different ways workload is composed. Results refer to lower secondary teachers employed full time. They should not be treated as estimates for every teacher, every school level or every part of a federal country. Alberta, for example, represents one Canadian province rather than Canada as a whole.
Between 2018 and 2024, full-time teachers in participating OECD systems reported working almost half an hour more per week on average. The movement was uneven. Korea rose by about nine hours, Brazil by nearly five, Japan fell by about four and Kazakhstan fell by more than 14. Roughly half of the systems with comparable data recorded no statistically clear change.
Annual Teaching Hours Tell A Different Story
The OECD’s latest published cross-country rules for teacher time refer mainly to the 2022/23 school year. Based on official regulations or agreements, public-school teachers were required to teach an average of 1,007 hours a year in pre-primary education, 773 in primary, 706 in general lower secondary and 679 in general upper secondary education.[b]
| Education Level | Average Teaching Hours per Year | What the Number Represents |
|---|---|---|
| Pre-Primary | 1,007 | Net contact time with children or groups |
| Primary | 773 | Scheduled classroom teaching by a full-time teacher |
| Lower Secondary | 706 | General-programme classroom teaching |
| Upper Secondary | 679 | General-programme classroom teaching |
Required teaching time generally falls as students move through the system. Primary teachers taught about 9% more hours than lower secondary teachers on average. At lower secondary level, the country range ran from fewer than 600 hours in Bulgaria, Finland, Korea, Poland, Romania and Türkiye to more than 1,000 in Chile and Costa Rica. Some reported values are minimum requirements, others are typical or maximum values, so even this standardised measure needs its country notes.
Annual teaching hours cannot be divided mechanically by 52 to create a weekly workload. School systems differ in the number of teaching weeks, teaching days, holidays, examination periods and professional development days. OECD data show that some teachers must work during student holidays, while others complete most statutory duties within the instructional year. Calendar design changes the meaning of the same annual total.
How Teachers Divide Their Working Time
TALIS 2024 found that teaching was the largest single activity in every participating system, yet it usually occupied less than half of total task-based time. Across OECD systems, teaching accounted for about 43%. Only Finland and the French Community of Belgium reached at least 50%. In Azerbaijan, Japan, Shanghai, Singapore and Uzbekistan, the share was below one-third.
| Activity | Share of Task-Based Time | Examples |
|---|---|---|
| Teaching | 43% | Classroom instruction and direct teaching |
| Lesson Planning | 14% | Preparation, adaptation and resource design |
| Marking and Correcting | 9% | Assessment, feedback and grading |
| Administration | 6% | Records, forms, reporting and routine procedures |
| Other Tasks | 28% | Teamwork, counselling, parent contact, management, professional learning and extracurricular duties |
The 28% residual category is not spare time. It contains activities that can be central to school quality: student counselling, collaboration, communication with families, professional learning, school management and extracurricular supervision. Two teachers with the same weekly total can therefore have very different jobs. One may teach many classes with limited planning time; another may teach fewer classes while carrying a substantial role in curriculum design or student support.
Experience Changes the Task Mix
Novice teachers spent more than half an hour extra each week on lesson planning than experienced teachers on average. In 12 systems, the difference approached one to two hours. Marking showed a different pattern: where experience-related differences existed, experienced teachers usually spent more time assessing work. Experienced teachers also reported more administrative time in all 19 systems where a measurable experience gap appeared.
Teaching loads can also differ by career stage. In 11 systems, novice teachers taught more weekly hours than colleagues with over ten years of experience. The gap reached about three hours or more in Bahrain, New Zealand and Türkiye. Six systems showed the reverse pattern. These differences matter because a beginning teacher may need additional preparation and mentoring even when assigned the same number of classes.
Three Models Of Workload Allocation
Iceland: A Defined Annual Envelope
Iceland’s baseline collective agreement model sets 1,800 annual hours, or 40 hours per week. The OECD describes an allocation of about 641 hours for teaching, 395 for preparation, 150 for professional development and 614 for other work. Special assignments are meant to be balanced by reducing time elsewhere. Schools can adjust the split according to subject, class composition, assessment demands and the need for co-operation.
Portugal: Teaching, School Duties and Autonomous Work
Portugal defines a 35-hour full-time week made up of 22 teaching hours, 2.5 hours for other school activities and 10.5 hours for autonomous work such as preparation and assessment. Teaching hours can fall with experience, while time for other duties rises. School leaders also receive credit hours that can reduce selected teachers’ classroom load for projects, mentoring or co-ordination.
Singapore: Fewer Lessons, More Structured Professional Work
Singapore illustrates why contact hours and total workload can move in opposite directions. Full-time lower secondary teachers reported 47.3 total hours per week, while teachers reported only 17.7 teaching hours. Schools commonly reserve timetable space for collaborative planning, curriculum and assessment design, training and professional learning. The resulting workload is high, but much of it occurs outside direct instruction.
These models do not form a quality ranking. They show three ways of making non-classroom work visible: a yearly time budget, a weekly division between settings, or scheduled professional activity. A useful comparison asks whether assigned duties have protected time, not only how many lessons appear on the schedule.
Country Patterns Behind The Totals
Japan
Japan remained the highest-workload system in TALIS 2024 at 55.1 hours per week, even after a decline of about four hours from 2018. Teaching occupied less than one-third of task-based time. Full-time lower secondary teachers reported 5.2 hours of administration per week, among the highest values recorded. School activity extends beyond lessons through planning, student guidance, meetings and extracurricular responsibilities.
New Zealand and Alberta
New Zealand and Alberta reported 47.5 and 46.9 hours respectively. Both estimates carry an OECD warning about a higher risk of non-response bias, so their precise rank should be read cautiously. The numbers still show that a recent full working week can extend well beyond classroom contact in these systems. Alberta’s result should remain labelled as provincial.
Finland
Finland shows a different composition. Teaching represented at least half of reported task-based time, while planning averaged 4.8 hours, marking 3.3 and administration 1.5 hours per week. Its statutory lower secondary teaching requirement was below 600 annual hours in the OECD’s 2022/23 comparison. A high teaching share does not mean unusually high classroom hours; it can also reflect a smaller reported administrative component.
Italy, Türkiye and Kosovo
Italy, Türkiye and Kosovo appeared near the lower end of the TALIS weekly distribution, at 32.7, 31.1 and 29.7 hours. These figures should not be translated into claims about effort or professionalism. Italy, for example, requires scheduled collegial duties but does not define a complete national statutory annual total in the same way as many other systems. Türkiye reported comparatively low planning and marking hours. National duty definitions help explain part of the difference.
Readers comparing specific systems can place workload alongside governance, curriculum, school structure and assessment through the country comparison library.[f] Workload numbers become more informative when the institutional setting is visible.
Workload Pressure Is Not The Same As Hours
TALIS separates duration from perceived pressure. Across OECD education systems, 52% of lower secondary teachers said excessive administrative work caused stress “quite a bit” or “a lot.” Excessive marking was reported by 40%, too much lesson preparation by 35% and too many lessons by 31%. Administrative work represented only 6% of task-based time on average, yet it produced the highest workload-related stress response.
| Workload Demand | OECD Average | Interpretation |
|---|---|---|
| Too Much Administrative Work | 52% | Most frequently reported workload stressor |
| Too Much Marking | 40% | No participating system reported a share below 20% |
| Too Much Lesson Preparation | 35% | More often reported by novices in many systems |
| Too Many Lessons to Teach | 31% | Perception does not always track teaching hours directly |
Why can a relatively small task create the most stress? Fragmentation, deadlines, duplication and low control can make one administrative hour feel more demanding than one planned teaching hour. TALIS found that additional time spent on marking, administration and parent communication had a stronger negative association with teacher well-being than comparable increases in some other tasks. This is an association, not proof that the task alone caused the change.
OECD comparisons with similarly qualified workers add another layer. In 13 of 26 participating systems, teachers reported fewer weekly hours than comparable professionals; 11 showed no clear difference. Only Japan and Singapore recorded more hours for teachers. Work intensity was often similar across occupations, although Chilean teachers reported high-speed or deadline pressure more frequently than their counterparts.
The wider labour-market context also matters. The International Labour Organization reports that more than one-third of workers worldwide regularly work over 48 hours a week, while about one-fifth work fewer than 35.[e] Teacher hours should therefore be compared with national working-time norms as well as with teachers abroad.
Why International Rankings Can Mislead
- School level changes the result. Primary teachers usually have more annual contact time than secondary teachers, while subject teachers may carry heavier marking loads.
- Full-time is nationally defined. A full contract may use weekly hours, annual hours, teaching periods or a percentage of a standard post.
- A lesson is not always 60 minutes. OECD statutory data convert periods to clock hours, but national documents may publish 40-, 45- or 50-minute lessons.
- Survey time and contract time differ. TALIS records a recent actual week; OECD statutory tables describe official requirements.
- Federal systems vary internally. Provinces, states, regions and school boards may set different calendars and duty rules.
- Work completed at home is easy to miss. Presence requirements capture on-site time but may exclude evening planning and marking.
- School calendars differ. The same annual total can be concentrated into fewer weeks or spread across a longer year.
- Task boundaries are cultural and administrative. Counselling, clubs, lunch supervision and data entry may belong to teachers in one system and specialist staff in another.
- Part-time employment changes averages. Comparisons limited to full-time teachers do not describe systems where part-time posts are common.
- Self-reporting introduces uncertainty. Recall, interpretation and non-response can affect estimates, especially at the edges of a ranking.
A sound comparison therefore uses matched populations, matched time units and matched definitions. Lower secondary full-time survey results should be compared with the same population elsewhere. Annual contact time should be compared with annual contact time. When those rules are followed, the data explain working arrangements instead of merely producing a rank.
Workload, Staffing and Retention
Teacher workload also connects to staffing. UNESCO estimates that the world needs 44 million additional primary and secondary teachers by 2030 to reach universal schooling. About 15 million of those posts are needed in sub-Saharan Africa. More than half of the global requirement arises from replacing teachers expected to leave the profession, not only from expanding enrolment.[c]
Shortages and workload can reinforce one another. Vacancies may enlarge classes, increase cover duties and reduce the specialist support available for administration or student needs. Heavier duties can then weaken retention. The relationship is not automatic: salary, employment security, leadership, professional autonomy and local labour markets also shape career decisions. Still, time allocation is part of workforce capacity, not simply an individual scheduling issue.
Class size also changes what one teaching hour contains. A lesson with more students can produce more marking, parent communication and differentiation even if contact time stays fixed. The number of subjects or grade levels taught matters too. Preparing five different courses usually requires more time than repeating one course across several groups. International datasets rarely compress all of these features into a single workload number.
The 2026 European Update
In June 2026, the European Commission published a new analysis based mainly on TALIS 2024. It covered 23 EU member states at lower secondary level and reported that about 90% of teachers were satisfied with their profession and current work environment, while around 70% were satisfied with employment terms other than salary. Supportive leadership, collaboration, induction, mentoring and professional autonomy were associated with stronger well-being and retention.[d]
The same report placed administrative workload, classroom discipline and student behaviour among the demands most consistently associated with poorer well-being and a stronger intention to leave. The EU teaching workforce is also ageing: about 40% of teachers are aged 50 or above, compared with 8% under 30. This age structure makes the organisation of work relevant to both retention and replacement planning.
Digital Work and Artificial Intelligence
Digital systems can remove or create work. Shared resources, automated routines and well-designed data systems may reduce repetition. Multiple platforms, duplicate reporting and constant communication can add small tasks throughout the day. TALIS 2024 found that around one-third of teachers across OECD systems had used artificial intelligence in their work, while 29% reported a strong professional learning need in this area.
Time effects depend on the task and the rules around it. Preparing a first draft of a lesson resource may become faster, but reviewing accuracy, adapting material, protecting student data and redesigning assessment can require additional work. Technology changes the composition of workload before it necessarily changes the total. This is one reason time-use surveys need to keep pace with changing classroom practice.
How To Read The Comparison Responsibly
The strongest international reading uses a small set of linked indicators rather than one headline number. Weekly actual hours describe duration. Annual teaching hours describe contact requirements. Task shares show where time goes. Stress responses describe perceived pressure. Class size, staffing and contract rules explain part of the setting.
- Use TALIS 2024 for recent self-reported weekly work and task patterns among lower secondary teachers.
- Use the OECD annual indicator report for official teaching and working-time requirements by education level.
- Check whether a result is national or subnational.
- Keep primary, lower secondary and upper secondary results separate.
- Read confidence warnings, response notes and country definitions before interpreting small rank differences.
- Compare workload with student numbers, course variety, support staffing and school-calendar length.
No single country number defines an ideal workload. The more useful question is whether the system provides enough paid and protected time for every required task. International evidence shows that teaching usually occupies less than half of teachers’ working time, non-teaching duties vary widely, and administrative demands create more stress than their average time share might suggest. That combination explains why a timetable alone cannot describe a teacher’s working week.
Sources and Data Notes
[a] Results from TALIS 2024: OECD report presenting the 2024 teacher and school-leader survey, workload tables, task shares and methodological notes.
[b] How Much Time Do Teachers Spend Teaching and Working?: OECD definitions and 2022/23 statutory teaching and working-time comparisons.
[c] Global Report on Teachers: UNESCO and the International Task Force on Teachers for Education 2030 on global recruitment and replacement needs.
[d] The Teaching Profession in the EU: European Commission analysis released in June 2026 using TALIS 2024, Eurydice and related European data.
[e] Working Time and Work-Life Balance Around the World: International Labour Organization evidence on weekly hours and working-time arrangements across occupations.
[f] Comparisons Library: Education by Country pages for placing workload data beside governance, structure, curriculum and system indicators.