The result and its meaning
Japan’s first separately reported school-library e-book expenditure has a much more concentrated distribution across administrative accounting entries than the separate book-cost category. The five largest entries account for 55.61% of recorded e-book expenditure, compared with 26.06% for the five largest book-cost entries. The groups of five are selected independently for each category.
An important part of the evidence is missing from the regional tables: additional financing appears in broader national totals without the same regional breakdown. This analysis therefore tests whether that reporting difference could reverse the comparison. Under a stated model in which the broader totals add nonnegative expenditure to the same 67 accounting entries, even the lowest conservative e-book concentration exceeds the highest conservative book-cost concentration by at least 15.53 percentage points.
These findings describe published expenditure accounts for Japanese fiscal year 2024. They do not identify how many schools use e-books, how much children read, or which suppliers received money. The contribution is a reproducible account of where expenditure is recorded and how strongly its interpretation depends on the source’s financing boundary.
What the source records
The Ministry of Education, Culture, Sports, Science and Technology, MEXT, published the selected fiscal-year 2024 expenditure results on 19 December 2025 and corrected them on 23 March 2026. This study uses that corrected document throughout. Fiscal year 2024 is the reference period; the publication and correction dates are different events. A later publication date does not turn these accounts into spending for 2026.
The main analysis concerns the public elementary and junior-high school accounts specified in the source. MEXT includes compulsory-education schools and the first stage of secondary-education schools within these categories. The document’s separate special-support-school accounts are outside this study’s main totals. Combining the two sets without preserving their distinct coverage would create a different population.
Three categories are transcribed: school-library book costs, school-library e-book costs, and school-librarian placement expenditure. Book costs are kept under their source-defined name. The selected publication does not establish that every item in that category is a print-only product. E-book costs similarly do not reveal whether the transaction involved a content purchase, a recurring licence, a subscription, or another service arrangement. Librarian-placement expenditure is money, rather than a count of employees or full-time-equivalent positions.
All amounts are displayed by MEXT in thousands of nominal yen. The calculations retain those units. They are not adjusted for inflation, purchasing power, school enrolment, or the cost of particular titles.
The 67 entries are an accounting partition
The regional tables contain 47 prefectural entries and 20 separate designated-city entries. Prefectural entries exclude the designated-city accounts, while the source also includes prefectural-run spending. The resulting rows are separate administrative accounting entries; they should not be represented as 67 mutually exclusive physical territories or equal population areas.
This distinction matters when reading a familiar place name. A prefectural entry is not automatically the expenditure of every school physically inside that prefecture. A designated-city entry has its own accounting position. Adding a city’s value to an already inclusive geographical total elsewhere would risk double counting. Conversely, substituting a population denominator for the name on an accounting row would require a separate, justified population mapping.
The downloadable data preserve every original Japanese row name, its source order, the elementary and junior-high amounts, the reported total, the category and the source page. Each row’s two school-level amounts sum to its total. Across all 67 entries, the nine category-by-school-level column sums reconcile with the source totals. This makes the partition visible and checkable without treating the original PDF as an easily comparable map of student access.
The observed concentration contrast
In the narrower financing scope used for the regional tables, total e-book costs are 66.336 million yen. The five largest recorded entries are Chiba, Aichi, Gifu, Ibaraki and Tochigi prefectural accounts. Together they record 36.890 million yen, or 55.61% of the category’s published total.
The separate book-cost category records 12.576031 billion yen. Its five largest entries are the Tokyo, Osaka, Aichi, Chiba and Saitama prefectural accounts. Their combined 3.277425 billion yen represents 26.06% of that category. Aichi and Chiba occur in both lists, but the comparison does not require the lists to be identical. It asks how much of each category is contained in that category’s own five largest entries.
The concentration measure is the sum of the five largest values divided by the category total. It gives the same number of accounting entries a consistent role in the comparison. It does not give those entries equal population, equal numbers of schools or equal educational responsibilities. A higher share means that recorded expenditure is distributed less broadly across this particular accounting partition. It is not a finding that individual schools have unequal digital resources by the same percentage.
Why the national financing boundary matters
The regional tables exclude costs financed by nationally transferred funds and similar sources specified in the document. MEXT also prints broader national amounts, in parentheses, which include expenditure financed through national and other sources. These broader figures are part of the same selected release; they are not a separate estimate assembled from another year.
For e-books, the broader total is 81.917 million yen. The difference from the narrower regional-table total is 15.581 million yen. That difference is 19.02% of the broader amount. For book costs, the broader total is 13.189743 billion yen and the difference is 613.712 million yen, or 4.65% of the broader amount.
The omitted financing share is therefore much larger for the e-book category. Simply carrying the regional percentage into the broader funding frame would conceal a material source limitation. Yet the selected document does not locate that additional amount across the same 67 rows. Assigning it to a favoured city, spreading it equally, or allocating it in proportion to existing expenditure would create information the source does not supply.
A sensitivity test without invented regional values
The test keeps every recorded regional amount unchanged and allows each category’s entire added financing amount to go anywhere among the same 67 accounting entries. The added amounts must be nonnegative and must sum to the difference between the broader and narrower national totals. No preferred allocation is chosen.
Let the observed regional amounts be x, with total T, and let S5 be the sum of their five largest values. Let M be the additional national financing amount. After adding any permitted regional allocation, the broader total is T + M. The broader top-five share cannot be smaller than S5 divided by T + M. The original five entries still contain at least their original amount, even if another set of five becomes larger.
The broader top-five share also cannot exceed (S5 + M) divided by T + M. For any proposed set of five entries, its original expenditure cannot exceed the original top-five sum. Its additional expenditure cannot exceed the entire national addition. Maximizing over every possible set of five therefore preserves this upper limit.
These are conservative mathematical limits under a nested-financing model. They are not observed regional values, fitted estimates, or confidence intervals. The identities of the five largest entries may change. The comparison remains valid without assuming that the same five entries stay at the top, or that the two expenditure categories have the same largest entries.
The concentration ordering survives that test
For e-books, the conservative broader-funding range for the top-five share is 45.03% to 64.06%. For the separate book-cost category it is 24.84% to 29.51%. These displayed coverage limits are rounded outward to two decimals. The lower e-book limit remains above the upper book-cost limit.
The separation is at least 15.53 percentage points under the stated assumptions, calculated from the unrounded limits rather than by subtracting the displayed rounded endpoints. This result is stronger than a comparison that ignores the additional financing. It shows that even an allocation deliberately chosen to reduce the e-book concentration and increase the book-cost concentration cannot reverse their ordering within this sensitivity model.
The finding does not establish the exact broader-funding shares. Both may lie anywhere permitted by the accounting information and the model. The intervals also do not measure sampling uncertainty or guard against source classification mistakes, inconsistent reporting, subsequent revisions or a different interpretation of financing coverage. Their purpose is narrower: to test whether the unlocated financing addition alone is large enough to overturn the concentration comparison.
The national digital share answers another question
E-book costs represent 0.525% of combined book-plus-e-book expenditure in the narrower financing frame. Under the broader national funding frame, the corresponding share is 0.617%. These are two explicitly defined accounting ratios, rather than a statistical interval around one unknown estimate.
The denominator contains the separate book and e-book categories. It excludes librarian-placement expenditure. Including staff costs would answer a different question and would reduce the percentage for reasons unrelated to the mix of the two recorded content-cost categories. Equally, dividing e-book costs by book costs alone would produce a ratio rather than a share of their combined total.
Neither percentage is an estimate of e-books’ share of all Japanese education spending, children’s reading, school holdings, library loans or the commercial publishing market. The source population and expense classifications determine what belongs in the denominator. Retaining those boundaries allows the small recorded national share and the strong accounting concentration to be understood together without implying a much broader market conclusion.
Reading zero entries responsibly
Twenty-seven of the 67 entries record positive e-book costs; 40 contain zero in the displayed thousand-yen category. That distinction is useful when locating the source of the published amount. It cannot establish that 40 areas have no e-books or that their schools have no digital reading access.
A school might obtain access through an arrangement recorded outside the selected category or financing frame. The document does not supply the transaction detail needed to decide whether that explanation applies to any particular zero. Nor does an accounting entry reveal the distribution of spending among individual schools. The correct public statement is about recorded entries in the defined category, without filling the explanatory gap with an assumed procurement model.
The same restraint applies to designated cities. Their narrow-scope accounts contain only 0.084% of recorded e-book expenditure, compared with 22.87% of book costs. This is a marked difference between accounting distributions. It is not evidence that city pupils lack digital resources, that a municipal authority bought no digital content from any budget, or that a publisher faces an unserved customer base.
How the analysis can be used
For a publishing or education reporter, the defensible story is that first-year separately recorded e-book costs are concentrated among administrative accounting entries in the published accounts, and that the ordering survives a transparent test of the larger unlocated financing component. The exact accounting scope belongs near the statistic, including in a social graphic or short caption.
For librarians and public-data researchers, the full table helps locate the entries behind a national total and distinguish categories that are easily combined in a headline. The data can support a request for a broader regional funding breakdown or a more precise description of the e-book expense category. It cannot identify which schools, titles, suppliers or contracts should be investigated without additional primary records.
Authors and publishers can use the study to understand the limits of institutional-spending statistics. An aggregate public-library or school-library cost is not an author’s royalty receipt, a retail sale, or a verified sales lead. A well-defined concentration result can inform a market discussion while leaving those commercial outcomes unestimated. No recipient list, procurement recommendation or estimate of future sales is inferred here.
Reproduction and source control
The source package identifies the exact corrected PDF and its checksum. The extraction recipe reads the three regional expense tables, checks all 67 row identities and source-order sequences, verifies each elementary-plus-junior-high row sum, and checks the nine national column sums. It rejects an unexpected source file rather than silently accepting a changed release.
The portable calculation recipe works from the pinned selected-value files. It recomputes national shares, independently selected top-five sums, geographic expenditure shares and the financing-sensitivity limits. Machine-readable outputs retain more precision than the article’s rounded percentages. Public text rounds only at display, so rounded figure labels need not sum with the precision of the underlying values.
A future update should preserve this snapshot, collect the new official release and recheck every category and accounting boundary before comparing years. MEXT first collected e-book expenditure in fiscal year 2024. This study therefore provides no prior-year e-book growth rate and does not replace an unavailable earlier observation with zero. A trend requires a genuinely comparable later observation and a review of any revisions to the first-year data.
Limits and the next useful evidence
The study’s unit is a published accounting entry and its outcome is nominal expenditure. There is no adjustment for unequal administrative scale or a claim about spending per school or pupil. The analysis does not merge the expenditure tables with older staffing or library-inventory surveys, whose dates and units answer different questions.
The most useful extension would be a regional breakdown of the broader financing totals using the same accounting partition, accompanied by a detailed category definition. That would replace sensitivity limits with an observed broader distribution. School-level provision, title availability and use would require their own records. Combining those outcomes would then require a documented join and a separate research design, rather than an assumption that expenditure automatically measures access.
The present result remains useful within those limits: the stronger concentration of recorded e-book expenditure survives a substantial financing-scope difference without inventing where the extra money went. The downloadable source values, calculations and figures make that conclusion open to inspection and reuse.
Source and reuse
Source: MEXT, fiscal-year 2024 school-library-related expenditure survey, corrected 23 March 2026, pages 1 and 3–5, accessed 31 October 2026. Selected numerical data were transcribed and independently analysed by Alisson Bento. This article, its calculations and its figures were not produced or endorsed by MEXT. Preserve the source and adaptation credit when reusing the accompanying materials; the original source and current MEXT reuse terms are linked with the downloads.
Five entries hold most recorded e-book costs
Japan · FY2024 public elementary/junior-high accounts
Full chart and source
Scroll horizontally to inspect the chart. ↔
Chart and values
Scroll horizontally to see all columns.
| Source category | Narrow total (¥000) | Narrow top-five sum (¥000) | Narrow top-five share (%) | Broader national total (¥000) | Added financing (¥000) | Conservative broader lower (%) | Conservative broader upper (%) |
|---|---|---|---|---|---|---|---|
| Book costs (source-defined) | 12,576,031 | 3,277,425 | 26.061 | 13,189,743 | 613,712 | 24.848 | 29.502 |
| E-book costs | 66,336 | 36,890 | 55.611 | 81,917 | 15,581 | 45.033 | 64.054 |
Each category selects its own five largest accounting entries. Amounts are displayed thousand yen. Broader regional values are unobserved; the bounds assume nonnegative additions within the same 67-account partition. Bounds displayed in tables and figures are rounded outward; the minimum separation uses unrounded values. Calculated percentages rounded to three decimals; source monetary values retain displayed units.
The concentration ordering survives broader financing
Japan · FY2024 · conditional deterministic sensitivity test
Full chart and source
Scroll horizontally to inspect the chart. ↔
Chart and values
Scroll horizontally to see all columns.
| Source category | Narrow total (¥000) | Narrow top-five sum (¥000) | Narrow top-five share (%) | Broader national total (¥000) | Added financing (¥000) | Conservative broader lower (%) | Conservative broader upper (%) |
|---|---|---|---|---|---|---|---|
| Book costs (source-defined) | 12,576,031 | 3,277,425 | 26.061 | 13,189,743 | 613,712 | 24.848 | 29.502 |
| E-book costs | 66,336 | 36,890 | 55.611 | 81,917 | 15,581 | 45.033 | 64.054 |
Each category selects its own five largest accounting entries. Amounts are displayed thousand yen. Broader regional values are unobserved; the bounds assume nonnegative additions within the same 67-account partition. Bounds displayed in tables and figures are rounded outward; the minimum separation uses unrounded values. Calculated percentages rounded to three decimals; source monetary values retain displayed units.
Data & downloads
Scroll horizontally to see all columns.
| Source category | Narrow total (¥000) | Narrow top-five sum (¥000) | Narrow top-five share (%) | Broader national total (¥000) | Added financing (¥000) | Conservative broader lower (%) | Conservative broader upper (%) |
|---|---|---|---|---|---|---|---|
| Book costs (source-defined) | 12,576,031 | 3,277,425 | 26.061 | 13,189,743 | 613,712 | 24.848 | 29.502 |
| E-book costs | 66,336 | 36,890 | 55.611 | 81,917 | 15,581 | 45.033 | 64.054 |
Each category selects its own five largest accounting entries. Amounts are displayed thousand yen. Broader regional values are unobserved; the bounds assume nonnegative additions within the same 67-account partition. Bounds displayed in tables and figures are rounded outward; the minimum separation uses unrounded values. Calculated percentages rounded to three decimals; source monetary values retain displayed units.
Scroll horizontally to see all columns.
| Original source account name | Administrative accounting scope | Book costs (¥000) | E-book costs (¥000) | Librarian-placement costs (¥000) | E-book share of book-plus-e-book costs (%) |
|---|---|---|---|---|---|
| 北海道 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 321,873 | 3,145 | 347,589 | 0.968 |
| 青森県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 102,107 | 0 | 36,070 | 0.000 |
| 岩手県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 125,644 | 0 | 165,685 | 0.000 |
| 宮城県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 85,020 | 0 | 333,180 | 0.000 |
| 秋田県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 65,857 | 0 | 166,363 | 0.000 |
| 山形県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 120,683 | 0 | 230,746 | 0.000 |
| 福島県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 207,854 | 1,504 | 519,144 | 0.718 |
| 茨城県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 302,002 | 4,864 | 579,842 | 1.585 |
| 栃木県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 181,866 | 4,562 | 684,516 | 2.447 |
| 群馬県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 200,128 | 0 | 424,683 | 0.000 |
| 埼玉県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 444,455 | 1,849 | 508,558 | 0.414 |
| 千葉県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 466,149 | 13,187 | 867,468 | 2.751 |
| 東京都 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 1,370,279 | 2,139 | 1,762,324 | 0.156 |
| 神奈川県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 270,295 | 1,070 | 539,521 | 0.394 |
| 新潟県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 141,016 | 0 | 154,410 | 0.000 |
| 富山県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 95,512 | 0 | 300,687 | 0.000 |
| 石川県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 135,028 | 3,980 | 715,073 | 2.863 |
| 福井県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 57,462 | 0 | 104,900 | 0.000 |
| 山梨県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 147,331 | 113 | 818,443 | 0.077 |
| 長野県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 310,359 | 719 | 935,964 | 0.231 |
| 岐阜県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 216,462 | 6,829 | 414,520 | 3.058 |
| 静岡県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 191,354 | 0 | 343,743 | 0.000 |
| 愛知県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 474,609 | 7,448 | 392,497 | 1.545 |
| 三重県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 160,082 | 1,965 | 107,716 | 1.213 |
| 滋賀県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 129,790 | 0 | 177,873 | 0.000 |
| 京都府 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 96,918 | 530 | 206,915 | 0.544 |
| 大阪府 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 521,933 | 3,667 | 1,335,711 | 0.698 |
| 兵庫県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 313,166 | 1,989 | 463,767 | 0.631 |
| 奈良県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 83,391 | 819 | 129,595 | 0.973 |
| 和歌山県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 111,995 | 1,917 | 185,493 | 1.683 |
| 鳥取県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 81,856 | 40 | 498,507 | 0.049 |
| 島根県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 66,813 | 0 | 304,840 | 0.000 |
| 岡山県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 146,502 | 583 | 820,509 | 0.396 |
| 広島県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 111,686 | 0 | 274,645 | 0.000 |
| 山口県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 152,639 | 0 | 311,788 | 0.000 |
| 徳島県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 56,072 | 0 | 22,934 | 0.000 |
| 香川県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 79,078 | 0 | 378,041 | 0.000 |
| 愛媛県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 121,534 | 542 | 29,250 | 0.444 |
| 高知県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 91,368 | 0 | 245,296 | 0.000 |
| 福岡県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 300,055 | 1,453 | 1,374,395 | 0.482 |
| 佐賀県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 89,507 | 297 | 624,941 | 0.331 |
| 長崎県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 174,020 | 585 | 339,596 | 0.335 |
| 熊本県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 131,298 | 484 | 458,885 | 0.367 |
| 大分県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 140,348 | 0 | 555,140 | 0.000 |
| 宮崎県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 89,956 | 0 | 224,612 | 0.000 |
| 鹿児島県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 215,976 | 0 | 1,373,353 | 0.000 |
| 沖縄県 | Prefectural account (excluding designated-city accounts; prefectural-run spending included) | 200,621 | 0 | 1,398,267 | 0.000 |
| 札幌市 | Designated-city account | 158,029 | 0 | 103,545 | 0.000 |
| 仙台市 | Designated-city account | 121,676 | 0 | 183,263 | 0.000 |
| さいたま市 | Designated-city account | 102,934 | 56 | 288,388 | 0.054 |
| 千葉市 | Designated-city account | 71,386 | 0 | 292,609 | 0.000 |
| 横浜市 | Designated-city account | 389,109 | 0 | 1,059,841 | 0.000 |
| 川崎市 | Designated-city account | 148,332 | 0 | 83,242 | 0.000 |
| 相模原市 | Designated-city account | 24,536 | 0 | 63,177 | 0.000 |
| 新潟市 | Designated-city account | 71,404 | 0 | 446,593 | 0.000 |
| 静岡市 | Designated-city account | 66,470 | 0 | 109,844 | 0.000 |
| 浜松市 | Designated-city account | 103,474 | 0 | 73,565 | 0.000 |
| 名古屋市 | Designated-city account | 301,501 | 0 | 147,875 | 0.000 |
| 京都市 | Designated-city account | 140,003 | 0 | 207,645 | 0.000 |
| 大阪市 | Designated-city account | 302,558 | 0 | 339,654 | 0.000 |
| 堺市 | Designated-city account | 119,206 | 0 | 66,456 | 0.000 |
| 神戸市 | Designated-city account | 187,256 | 0 | 372,753 | 0.000 |
| 岡山市 | Designated-city account | 79,352 | 0 | 537,706 | 0.000 |
| 広島市 | Designated-city account | 73,074 | 0 | 27,631 | 0.000 |
| 福岡市 | Designated-city account | 214,995 | 0 | 52,610 | 0.000 |
| 北九州市 | Designated-city account | 143,897 | 0 | 170,440 | 0.000 |
| 熊本市 | Designated-city account | 56,890 | 0 | 213,413 | 0.000 |
Corrected MEXT FY2024 values; narrower funding scope. Original Japanese source names preserved. The accounting partition is not a set of mutually exclusive physical territories. E-book zeros do not measure access or adoption. Calculated percentages rounded to three decimals; source monetary values retain displayed units.
Evidence & method
How we calculated it
- Transcribe all 67 administrative accounting entries for book, e-book and librarian-placement costs from the corrected MEXT FY2024 release. Preserve school-level columns, source order, category and source page.
- Independently select the five largest entries per category and divide their sum by that category’s total; use the same 67-account partition without converting it to school or pupil exposure.
- Compare narrow and broader national financing totals. Allow each channel’s nonnegative added financing to be allocated arbitrarily across the same accounts. Bound the broader top-five share by S5/(T+M) and (S5+M)/(T+M).
- Verify 603 source cells, row sums and all nine column totals; rerun the portable recipe from pinned selected values. Percentages are rounded for display only; no sampling interval or first-year e-book trend is estimated.
What the data cannot tell us
- These are separate administrative accounting entries, not mutually exclusive physical territories, individual schools, pupils or suppliers. Prefectural-run spending is included and designated-city accounts are separated.
- The book and e-book categories retain source definitions. No print-only classification, purchase/licence model, digital textbook coverage or actual title availability is established.
- Detailed regional values exclude the financing specified by MEXT. Broader national totals do not identify broader regional allocations; sensitivity limits assume nested nonnegative additions to the same accounts.
- Zero displayed expenditure is not evidence of absent digital access or adoption. Nominal expenditure is not a count of books, loans, readers, FTE or author royalties.
- E-book costs were first collected for FY2024. There is no prior-year e-book growth estimate, inflation adjustment, per-pupil normalization or causal finding. Reporting/classification/revision uncertainty is outside the financing sensitivity test.
Original sources
- MEXT: FY2024 school-library-related expenditure survey (corrected 23 March 2026)
Corrected source summary and regional tables, pages 1 and 3–5. Nominal fiscal-year 2024 costs in source-defined categories; financing scopes preserved.
Accessed October 31, 2026
- MEXT website reuse terms (amended 17 August 2026)
MEXT reuse terms, amended 17 August 2026: source attribution and adaptation credit; no implied government production or endorsement. Numerical data/simple tables are outside the stated copyright conditions.
Accessed October 31, 2026
Cite & reuse this study
Alisson Bento. Japan’s school-library e-book spending: a financing-scope sensitivity test. 2026. https://alissonbento.com/research/japan-school-library-ebook-expenditure-concentration-2024/
Source: MEXT, FY2024 school-library-related expenditure survey, corrected 23 March 2026, pages 1 and 3–5. Selected data adapted and independently analysed by Alisson Bento. MEXT did not produce or endorse this analysis. Original source: https://www.mext.go.jp/content/20260323-mxt_chisui01-100002176.pdf — MEXT reuse terms, amended 17 August 2026: source attribution and adaptation credit; no implied government production or endorsement. Numerical data/simple tables are outside the stated copyright conditions.
Corrections or editorial questions