The Four-Day Week Around the World · Four-Day School Week Resource Center

The four-day week beyond the United States

The four-day school week is often described as an American oddity. It isn’t — but the American version really is one of a kind. France runs the world’s largest four-day week, in about nine of every ten of its communes, and argues about it in the language of children’s sleep rather than school budgets. Canada runs the world’s most American one: a couple dozen rural schools, adopted for bus routes and staffing. The United Arab Emirates decreed one overnight. And Japan, Korea, and China each cut a school day at national scale decades ago, leaving behind the best evidence anywhere on what removing a day actually does.

No other country combines what the United States does: local district choice, rural cost-and-recruitment logic, annual hours held roughly steady by lengthening the remaining days, at a scale of roughly 900 districts and more than 2,100 schools across 26 states. This page is a tour of the alternatives — and of what each one knows that the others don’t.

France: la semaine de quatre jours

France is Missouri’s mirror image: the same schedule, adopted for opposite reasons, and argued about on entirely different grounds.

The Darcos decree of May 2008 abolished Saturday-morning classes and put every French primary school on four days — 24 hours of instruction across Monday, Tuesday, Thursday, and Friday. Nearly every school moved within two years. A 2013 reform under Vincent Peillon pushed schools back to 4.5 days, restoring Wednesday mornings and leaving communes to fund the afternoon activities that replaced the lost class time. Mayors and parents disliked it, and a June 2017 decree let communes opt back out. They stampeded: 43 percent of communes were back on four days that autumn, 87 percent within a year, and about 93 percent by 2023. The holdouts are the big cities — Paris, Toulouse, Lille.

The result is the developed world’s most extreme school calendar: roughly 144 instruction days a year, packed into six-hour days. That number is the single most misunderstood figure in the international comparison, and it is worth the detour below.

What France knows that America doesn’t.

France ran the experiment America never ran. A half-century of French chronobiology — Testu’s work on children’s daily performance rhythms, the 2001 INSERM collective expert review, the Académie nationale de médecine’s 2010 position — concludes that the four-day week desynchronizes young children. Under a four-day schedule the normal daily attention curve inverts, with depressed vigilance on Monday and Tuesday; under 4.5 or five days that effect is largely confined to Mondays. The Académie’s 2010 verdict was blunt: the four-day week “n’est pas favorable à l’enfant.”

This body of work is essentially uncited in the American four-day debate. It is the strongest child-development argument against the schedule anywhere, and it comes from the only country running the schedule at national scale.

What America knows that France doesn’t.

France never measured achievement. When the Sénat examined the school-rhythms reforms in 2017, the head of DEPP — the French education ministry’s research arm, roughly its DESE equivalent — told senators that the way the 2013 reform had been rolled out made scientific evaluation impossible, because there was no comparison group. The inspectorate called it the astonishing paradox of a reform that gave itself no means of evaluating itself.

So the American literature’s estimates of achievement, teacher retention, and juvenile crime have no French counterpart, and the French chronobiology has no American counterpart. Each country ran half the experiment.

The politics rhyme.

French teachers strike for the four-day week — Rennes did so in May 2025. Communes prefer it because the 4.5-day version left them paying for Wednesday-morning childcare and afternoon programming, on the order of €160 net per child per year. And the sharpest French critique is one American parents will recognize: a free weekday advantages families with money and flexible schedules, and lands hardest on working mothers.

What happens next.

In May 2025 the president empaneled a citizens’ convention on children’s time — 133 randomly selected citizens plus 20 children. In November 2025 it adopted, by a large majority, a recommendation to move French primary schools back to five days, Monday through Friday, on explicitly chronobiological grounds: “les chronobiologistes sont unanimes.” The government has not acted. In February 2026 the education minister ruled out any change for the coming school year; in April 2026 he called the school-week question “typiquement un sujet de présidentielle 2027” — a matter for the 2027 presidential campaign. Paris launched its own convention in May 2026.

France is, in other words, seriously considering abolishing the schedule that American districts are adopting — and for reasons American districts have never studied.

When countries cut a school day

Three countries have cut a school day at national scale, and only there can you watch what happens when a day disappears and a private tutoring market stands ready to absorb it.

Japan, 1992–2002. Japan went from six days to five in stages — one Saturday off a month in 1992, two by 1995, all of them by 2002 — and paired the change with a roughly 30 percent cut in curriculum content, the reform known as yutori. It is the best-studied day reduction anywhere, and the finding is distributional rather than average: the gap in study time between children of less- and more-educated parents widened sharply, because advantaged families bought the lost Saturday back at juku cram schools and other families could not. Household spending on supplementary education rose, and rose most among higher-income households. When Japan’s international test scores fell — mathematics down 23 points on PISA between 2000 and 2003 — the country reversed the curriculum cut rather than the calendar: hours were repacked into five days, and optional Saturday classes were allowed back from 2014. The share of public elementary schools holding any Saturday classes roughly doubled in two years, from about 9 percent to about 17 percent. Parents backed the extra Saturdays at 86 percent; teachers at 38 percent. Missouri superintendents will recognize that ratio — inverted.

Korea, 2005–2012. Korea also went from six days to five in stages, following its five-day work week, and — crucially — moved the Saturday hours into weekdays rather than cutting them. That is the American four-day design, which makes Korea the closest structural analog anywhere with real causal evidence. The best study of it tracked how children actually spent the freed Saturday: about half of it went to sleep. Parents cut paid work to be home. Average private-tutoring time did not rise. Even so, the authors concluded the policy “could worsen the intergenerational transmission of inequality,” because what gains the tutoring market did see concentrated among the wealthy.

China, 1995. China’s move to a five-day school week was largely nominal. Weekend instruction migrated into make-up classes and, later, into an enormous commercial tutoring sector — the one Beijing’s 2021 “double reduction” crackdown set out to dismantle. Remove official days and unofficial ones grow wherever demand exists.

Why this matters for American research. The distributional damage in Asia depended on families’ ability to buy the time back. Rural America mostly has no juku sector — which may be exactly why average four-day achievement effects in the United States look modest in rural districts and worse in non-rural ones, where the ability to substitute differs most. Nobody has tested that mechanism directly here. It is an open question, and an international framing no American paper has used.

Canada: the closest thing to an American four-day week

If you want to see the American model running outside the United States, look north — not across the Atlantic.

British Columbia’s Boundary district, around Grand Forks, put all its schools on four days in 2001 while facing a deficit, and has never gone back; its 2026–27 calendar runs 153 instruction days with Fridays closed, and the district has long credited the schedule with saving about C$200,000 a year in busing and heating. Gulf Islands, also in B.C., runs a four-day calendar. In Saskatchewan, Prairie South has had five rural schools on the schedule for more than fifteen years, and this January added its first urban one, in Moose Jaw. Alberta’s cases are one-offs adopted by community vote: Granum voted 94.4 percent in favor in 2024, and Byemoor moves to four days in September 2026 after enrollment fell below a funding threshold, lengthening its school day by 51 minutes.

The enabling condition is the same one operating in most American four-day states. Alberta and British Columbia regulate instructional hours, not days — 950 hours in the lower grades and 1,000 in the upper ones — so a district may rearrange the week as long as the hours hold. Ontario effectively requires five days, and blocked a francophone board’s two-school pilot in 2023 on exactly that basis.

Two things are worth noticing. First, where Canadian districts have run the numbers, the savings resemble American ones in kind but not in source: a division-wide proposal in Alberta’s Livingstone Range, rejected in February 2025, projected C$556,100 in savings — about 1 percent of the budget — of which roughly C$513,000 was staffing. Busing looked like the second-biggest line at about C$103,000 until the division’s transportation contracts were accounted for, at which point the net transportation savings came to about C$1,100. Staffing drives the money, not buses. Second, and more striking: at perhaps two dozen schools nationally, Canada has produced almost no research on the schedule. The only academic study we could locate is a single master’s project from Simon Fraser University, whose author notes the same absence. Everything Missouri learns transfers there. Nothing there yet tests what Missouri believes.

A new wave is forming. A Saskatchewan division launches a two-year opt-in pilot in 2026–27, requiring 70 percent parent support at each school, and community-led proposals in a neighboring division go to parent votes in November 2026. If those land, Canada’s total could double — and the whole country would still have fewer schools on the four-day week than Missouri has districts on it.

Decreed overnight, and still holding out

The United Arab Emirates decreed it overnight. In December 2021 the federation announced that from January 1, 2022, government offices and public schools would run four and a half days — Monday through Thursday plus a half-day Friday — with a Saturday–Sunday weekend. The emirate of Sharjah went further, to a full four-day week, and let private schools choose how to make up the time: longer days, up to three hours a week of e-learning, or up to seven extra days on the calendar. That makes Sharjah the only jurisdiction we have been able to find anywhere that imposed a four-day school week wholesale, on every school at once.

What came back was approval, not evidence. A February 2023 survey by Sharjah’s private-education authority, covering about 7,000 staff, more than 31,000 families, and 127 schools, reported large majorities perceiving improvements in work-life balance, staff motivation, student relationships — and even academic achievement. Those are perceptions, not test scores. No peer-reviewed evaluation exists. It is a natural experiment waiting for a researcher.

Israel is the holdout debating the move everyone else already made. Israel is the only OECD country still running a six-day school week, and it has the most instruction days in the OECD — 214 a year in primary school, against an average of 186. Israeli press reporting of ministry figures puts primary instruction at about 941 hours a year, well above the OECD average of 804. In February 2026 the finance ministry floated a pilot to drop Fridays in kindergarten and elementary school, framed in terms any Missouri levy campaign would recognize: roughly 30 fewer school days a year, about 8 percent fewer teaching positions needed, and 1.7 billion shekels saved. The education ministry denied the report the same day, and the teachers’ union chief drew his own line in public. Watching Israel argue six days versus five, with no causal evidence to hand, is watching Japan’s 1992 decision replay in real time.

The rest of Europe

UNITED KINGDOM

Budget pressure in 2019 pushed a number of English schools to trim the week — most visibly Bellfield Junior in Birmingham, which moved to a 4.5-day week that September to save about £50,000 a year, with about a dozen Birmingham schools finishing early on Fridays. That episode is over. The Department for Education set a minimum expectation of a 32.5-hour school week in its 2022 schools white paper, issued guidance in July 2023, and told schools to comply by September 2024 at the latest.

The live British story now is a different one, and American districts should watch it: compression on the teacher side, not the pupil side. Since September 2024 the Dixons Academies Trust has run a nine-day fortnight across 17 schools — staff get one day in ten off, pupil hours unchanged. An independent evaluation funded by the Education Endowment Foundation found staff wellbeing improved and pupils were largely unaffected, while cautioning that workload did not fall as much as intended and that support for early-career teachers suffered. The trust reports a large drop in staff turnover. That is a way of buying the recruitment benefit of a four-day week without taking a day from students — and it may be the version that arrives in American districts next.

Read the DfE guidance → Read the EEF evaluation →

BELGIUM AND THE NETHERLANDS

Both countries have long-standing Wednesday half-days, and both have seen individual schools drop to four days — but as emergency responses to teacher shortages, not as policy. Dutch rules cap four-day weeks at seven per year for most primary grades; a structurally four-day school is not permitted. When you see a European four-day school in the news, check whether it is a choice or a staffing failure. Usually it is the second.

A CAUTION ON ICELAND

Iceland’s widely publicized “four-day week” refers to a shorter work week, not a school week. We went back through the record again in 2026 to be certain: there is no Icelandic four-day school week, and no shortening of the pupil week anywhere in the Icelandic trials. If you see Iceland cited in a four-day school week argument, the citation is wrong.

The opposite direction

While the United States adds four-day districts, much of Latin America spent two decades doing the reverse. Chile’s full-day school reform, begun in 1997, extended the school day by about 30 percent; evaluations find modest achievement gains concentrated among rural and public-school students, along with later childbearing and higher earnings in adulthood. Colombia’s single-shift reform since 2015 and Mexico’s full-time schools program, which ended in 2022, show the same shape: small positive effects, largest in the poorest schools. Brazil’s version is the notable null.

The direction of travel proves nothing on its own. But it is worth knowing that while American districts debate selling a day, much of the hemisphere has been buying hours — and finding that the children who gain most are the ones with the least at home.

The bigger picture: days versus hours

Internationally, how you count school matters as much as how many days it has — and the four-day week is exactly where the counting breaks down.

A correction to this page. We previously said France has among the fewest instructional days in the OECD, around 160. That figure needs fixing, and the fix is more interesting than the error. The 162-day number the OECD published for years was built on the 4.5-day statutory week. From its 2023 edition the OECD has reported France at 180 days instead, without explaining the change. Neither number describes what most French children actually attend. A four-day commune runs 36 weeks of four days — 144 instruction days a year, the fewest in the developed world — inside about 864 annual hours, close to the OECD average. That 144 appears in no OECD table. It appears in France’s own 2011 national report on school rhythms, which stated it plainly: French primary schooling runs 144 days a year against an OECD average of 180.

The comparison below is the one that actually frames the American debate.

Instruction days per year, primary education

Published national calendars, with the actual four-day calendar used by most French communes shown separately

Israel

214

Japan

202

OECD average

186

United States

180

France — as the OECD has reported it since 2023

180

France — four-day communes, actual

144

How to read this. The OECD counts a country’s statutory calendar, not what individual schools run. That is why France appears twice: once as the OECD reports it, and once as most French primary children actually attend. The American four-day week creates the same gap in the American data — a district on four days is not running the 180-day year the state reports.

See the data

Instruction days per year, primary education
Jurisdiction	Instruction days per year (primary)
Israel	214
Japan	202
OECD average	186
United States	180
France — statutory basis, as published by the OECD since 2023	180
France — four-day communes, actual	144

Source: OECD, Education at a Glance 2025, Indicator D1, for all figures except the last. France’s four-day figure is 36 school weeks × 4 days, the number stated in France’s 2011 national report on school rhythms; it appears in no OECD table.

The United States is already the compressed model. American primary schools run 180 days — fewer than the OECD average of 186 — at about 974 hours, far above the OECD average of 804. Long days inside a short year. France, at 864 hours across 144 days, shows where that road ends. The four-day week does not move the United States toward the international mainstream; it moves it further from it, in the direction France is now debating abandoning.

Does the length of the school year decide how much children learn? Not on its own. The OECD’s own comparison of cumulative instruction hours against PISA mathematics scores is weak, and the causal research converges on three propositions rather than one headline: instructional time does matter; its value per additional hour is modest and depends heavily on whether the classroom is working; and its distribution matters more than its average, because time removed is taken hardest from the students with the least support outside school.

That is the international version of the American finding that it is the hours, not the days, that move achievement — which is why a four-day week that genuinely protects annual instructional hours is a different policy from one that does not.