Japan Lost 215 Primary Schools in a Year, and the Average School Still Got Smaller — From 315.7 Pupils to 312.4

Primary school staff

30.5→22.6（7.9 fewer）

Lower secondary staff

37.8→27.5（10.3 fewer）

Primary school transport cost（258 cases）

3,663k yen→9,191k yen（×2.51）

Lower secondary transport cost（112 cases）

5,284k yen→16,261k yen（×3.08）

In the same survey, schools running a bus service rose from 211 before the mergers to 494 after. Over three years, 651 mergers turned 1,617 schools into 694.

On 26 December 2025 the Ministry of Education, Culture, Sports, Science and Technology published the final figures of the FY2025 School Basic Survey, taken as of 1 May 2025.

Primary schools numbered 18,607, down 215 on the previous year[1]. Enrolment stood at 5,812,375, a fall of 129,358 and the lowest on record[1].

Behind that net 215, public schools fell by 215, one national school closed and one private school opened[1]. What is disappearing is the public sector.

Dividing one figure by the other changes the picture. Pupils per school come to 312.4 in FY2025. The year before, 5,941,733 across 18,822 schools gave 315.7. Closing 215 schools still left the average school smaller than it had been.

Lower secondary follows the same shape. 9,827 schools, down 55, with 3,105,297 pupils, a fall of 35,835 and the lowest on record[1]. Per school, 317.9 down to 316.0.

One category grew. Compulsory education schools, which combine the primary and lower secondary stages, rose by 23 to 261, with a record 86,924 enrolled, up 7,113[1]. Folding the two stages into one institution has become one way mergers land.

The merger criteria sit in a 1958 ordinance

Article 4, paragraph 1 of the Enforcement Order of the Act on the National Treasury Share of Facility Costs for Compulsory Education Schools gives two conditions for appropriate scale. The first is class count: roughly 12 to 18 classes. The second is commuting distance: roughly within 4 kilometres for primary schools, and within 6 kilometres for lower secondary and compulsory education schools.

That ordinance was promulgated on 27 June 1958. A figure premised on walking or cycling decides whether a merger qualifies for national treasury support.

Paragraph 3 allows an exception: where the minister judges it appropriate given educational effect, transport conditions and other circumstances, mergers beyond those distances also qualify. In practice, school buses have carried mergers well past them.

A 2015 guide added a measure of time

So the guide issued on 27 January 2015 added a second yardstick. Surveying municipal criteria, the ministry found that among municipalities setting a standard for travel time by public transport, roughly one hour was the most common[2], and adopted that as a provisional benchmark.

The guide also reports an analysis of past mergers. Travel time from the furthest point after a merger ranged from under 10 minutes to 75 minutes, with over 90% within an hour[2]. Read the other way, close to one in ten exceeds an hour.

What shape the mergers take

In a ministry census of every merger over three years, 651 mergers turned 1,617 schools into 694[3].

By form, 458 were primary-to-primary (1,153 schools into 475), 164 lower-secondary-to-lower-secondary (370 into 171), and 29 involved setting up a compulsory education school or an integrated primary-lower secondary school (94 into 48)[3]. Primary mergers alone removed 678 schools.

Where the merged school sits is also settled. 88% use the site of one of the pre-merger schools; 12% use a different site[3]. Nearly nine in ten consolidate into one of the existing buildings.

From first discussion to opening, 44% take four years or more

Mergers take time. From the first discussion to opening, 2 years accounted for 15%, 3 years 16%, 4 years 14%, 5 years 12%, 6 years 10% and 7 years 8%[3].

Those listed come to 75%. At least 44% took four years or more. Only 15% were done in two.

Pupil numbers fall every year. The school size at the start of discussion is not the size at opening. A merger completed over four years can arrive at a scale that already needs merging again.

Municipalities carry 60% of the building cost

Where the merged school required new construction, primary schools averaged 2,023,308 thousand yen in total project cost, of which 1,252,610 thousand fell on the municipality; lower secondary averaged 1,977,970 thousand, of which 1,301,405 thousand fell on the municipality[3].

That is 61.9% for primary and 65.8% for lower secondary. Six-tenths of a two-billion-yen project is carried by the municipality that decided on the merger. Of the facility work involved, 19% was new construction, 19% renovation and 9% both[3]: close to half involved building work of some kind.

What falls is staff

Averaged over two-school mergers, primary schools go from 30.5 staff to 22.6, lower secondary from 37.8 to 27.5[3]: 7.9 and 10.3 fewer.

Against the structure behind a shortage of 4,317 teachers（このサイトの記事）, mergers push the other way. The same authority reports a shortage while reducing posts.

What rises is the cost of getting pupils there

Over the same averages, the cost of bringing pupils in from a distance rose from 3,663 thousand yen to 9,191 thousand for primary, and from 5,284 thousand to 16,261 thousand for lower secondary[3]: 2.5 times and 3.1 times.

Schools running a bus service went from 211 before the mergers to 494 after[3], 227 of them by purchase. Average purchase cost was 18,348 thousand yen for primary (162 cases) and 18,896 thousand for lower secondary (56 cases)[3].

Staff costs fall; transport costs rise

The two sit on the same page of the same survey, yet they rarely appear together when a merger is argued. The falling side is presented as a saving; the rising side goes into a separate line as an incidental expense. Both come out of one decision.

What the ranking of requested support shows

Asked what support they wanted from central government, municipalities named extra staffing allocations at 79%, facility subsidies at 74%, school bus subsidies at 60%, reference material for judging school scale at 42% and collections of good practice at 40%[3].

The top two concern what happens inside the school; getting there comes third. Even among those actually carrying out mergers, transport ranks third. At 60%, though, it is not small either.

Two standards, distance and time, still sit side by side

Four kilometres on foot, one hour on a bus. The first is a 1958 ordinance and decides eligibility for national treasury support. The second is a 2015 guide with no binding force. A merger the distance rule would block, the time benchmark lets through.

Paragraph 3 of the ordinance wrote the exception in from the start, so on the face of the statute there is no contradiction. But because the benchmark has no binding force, how far to stretch is left to the municipality.

The one-hour figure rests on a distribution, not on what a child can bear

Two things support it: one hour was the commonest municipal standard, and over 90% of past mergers already fell inside it.

Both are grounded in what is already the case. Neither is worked back from the burden on a child. The guide itself notes that longer travel can bring reduced fitness and less time for study at home. It also lists workarounds: audio materials on the bus, a librarian reading aloud, dropping pupils a set distance from the gate to preserve some walking. The difficulties are acknowledged, and the benchmark is still taken from the distribution.

Not merging is also being chosen

Merging is not treated as the only route. In the same survey, 31% of municipalities are actively working on drawing out the advantages of small schools that cannot be merged, and a further 48% are working on it[3], through year-round local studies drawing on community expertise at 82%, securing basic academic attainment through close instruction at 74%, and deliberately rotating every pupil through various roles at 61%[3].

On mitigating the drawbacks, 24% are actively working and 53% working[3], with year-round joint activities between primary and lower secondary at 52%, shared school administration across several schools at 49%, and joint school events across the year at 48%[3]. Close to eight in ten municipalities are running small schools without merging them.

How far is too far is a question of consent, not statistics

That mergers are not keeping pace with the fall in pupil numbers is what the opening division shows. Push the merger count higher and the per-school figure recovers, but the distribution of travel times stretches to the right.

And 44% of mergers take four years or more. The scale at the decision is not the scale at opening. Raising the pace to catch up lengthens each journey, and the numbers fall again in the interval between deciding and opening.

The consensus problem set out in designing participatory policy processes（このサイトの記事） is exactly where communities have to choose between those two numbers. How far is too far is not a statistical question.

Further Reading

• 『学習と協働: 学校統廃合をめぐる住民・行政関係の過程』（外部サイト、新しいタブで開きます） (Yasuhito Tanma, Toyokan Shuppansha). A study tracing what happens between residents and administrators through the course of a merger. Useful for the consensus-building that sits outside the figures.

References

FY2025 School Basic Statistics (School Basic Survey), Final Figures — Ministry of Education, Culture, Sports, Science and Technology (2025). Ministry of Education, Culture, Sports, Science and Technology

Guide to Appropriate Scale and Placement of Public Primary and Lower Secondary Schools — Ministry of Education, Culture, Sports, Science and Technology (2015). Ministry of Education, Culture, Sports, Science and Technology

Current State and Issues in Primary, Lower Secondary and High School Mergers (briefing to the Depopulation Policy Round Table) — Ministry of Education, Culture, Sports, Science and Technology, Elementary and Secondary Education Bureau (2017). Ministry of Internal Affairs and Communications