The Reality After Japan's Digital Textbook 'Formalization': How the Three-Option System Entrenches Regional Inequality

On April 7, 2026, the Japanese government approved a bill amending the School Education Act and related legislation. According to Nikkei, the amendment formally positions digital textbooks as official textbooks on equal standing with print, with the law taking effect in April 2027 and full application to elementary schools beginning in academic year 2030. The bill simultaneously adopts a framework allowing each board of education to choose among three options: paper-only, paper-and-digital combined, or digital-only. At a press conference, Minister of Education Matsumoto stated, "We will not pursue uniform rollout" and "We currently do not permit children in the fourth grade of elementary school or below."

The cabinet decision's headline reads "formalization of digital textbooks" — but this article deliberately focuses on the transition period before the law takes effect. As of May 2026, roughly one and a half months after the cabinet decision, preparations for which of the three options to select are already underway in schools, and the structural preconditions constraining that selection have begun to emerge.

The first thing to grasp is the disparity in coverage across school types. According to the MEXT Survey on the State of Educational Informatization in Schools (FY2023 tabulation, as of March 1, 2024), learner-use digital textbook coverage stands at 99.8% for elementary schools, 99.8% for middle schools, just 11.5% for senior high schools, 37.8% for special-needs schools, and 88.2% as the overall average. While compulsory education has essentially completed its digital textbook infrastructure through the GIGA School Initiative's one-device-per-student deployment, nearly 90% of high schools have yet to equip themselves. Even if the three-option system is introduced, the reality for high schools is that the overwhelming majority cannot select "digital-only" or "paper-and-digital combined," effectively forcing them into a paper-only outcome.

Interprefectural disparities are also significant. According to MEXT's "On the Current Situation Regarding Digital Textbooks" (September 2024 version), survey data recorded a maximum 2.5-fold gap in learner-use digital textbook penetration rates across prefectures — with Wakayama Prefecture highest and Aichi Prefecture lowest — reflecting the direct influence of fiscal capacity index and board-of-education policy. Subject-by-subject usage frequency also varies: in the Ministry of Education's large-scale survey on the effects and impact of learner-use digital textbooks, more than 60% of teachers at elementary and middle schools providing digital textbooks reported using them in class "roughly once every four lessons or more," while approximately 40% used them less than once a week. Even where devices are provisioned, there remains a wide gap in the density of classroom implementation.

In short, at the time of the cabinet decision, digital textbook coverage, usage patterns, and teacher training systems already carry regional and school-type disparities — and the three-option system may function as a mechanism that institutionally endorses and entrenches those disparities. This article organizes this question as the "reality of the four-year transition period."

South Korea's AIDT Collapse After One Semester — A Definitive International Failure Case

The most recent and definitive international case for evaluating the three-option system is South Korea's AIDT (AI Digital Textbook). In March 2025, South Korea became the first country in the world to introduce AI digital textbooks across three subjects — English, mathematics, and information technology — targeting third and fourth graders in elementary school, first graders in middle school, and first graders in high school. The Ministry of Education (the South Korean equivalent of MEXT) touted them as "the world's first AI digital textbooks."

Anomalies emerged immediately after launch. According to Kyoiku Shimbun, two months after launch the daily connection rate hovered below 10%, with Daegu City's peak connection rate reaching only 11% and the rate among middle and high school students in Sejong City falling to a catastrophic 0.3%–0.5%. Teachers reverted to printed supplementary materials; students stopped opening the AIDT. The situation — infrastructure deployed but not used in classrooms — spread widely.

The matter was resolved in August 2025. According to AFPBB, on August 4, 2025, the South Korean National Assembly passed an amendment to the Elementary and Secondary Education Act, downgrading the AIDT from "textbook" to "educational material." There is no precedent anywhere in the world for a digitally deployed educational material, introduced as a textbook, losing its official textbook status within a single semester (five months). The fact that a system hailed as "world-first" at launch was legislatively downgraded demonstrates the magnitude of risks inherent in textbook formalization.

What South Korea's failure reveals is this: if implementation prerequisites — teacher training, device stability, and learning experience design — are not in place before the institutional framework is erected, textbook formalization becomes counterproductive. It is certain that Japan's adoption of the "three-option system" reflects this South Korean lesson, and Minister Matsumoto's statement of "no uniform rollout" can be read as absorbing this lesson. Yet even with optional selection, if the prerequisite conditions for selection exist as regional disparities, the same problem is liable to recur in a different form.

Teacher ICT Competency — A 37-Percentage-Point Training Completion Gap

The regional pathway through which the three-option system entrenches inequality is most clearly illustrated by regional disparities in teachers' ICT instructional competency. According to the MEXT FY2023 Survey Results on the State of Educational Informatization in Schools, the national averages show 90.7% of teachers able to use ICT for lesson preparation, research, assessment, and administrative tasks; 82.2% able to use ICT in instruction; and 83.1% able to guide students in ICT use — averages in the 80th percentile range depending on the indicator.

The figures alone suggest that the situation is "broadly manageable," but the distribution behind the average demands attention. The national average ICT instructional competency training completion rate in FY2024 is 71.9%, but Gifu Prefecture, the highest, achieves 95.8%, while Gunma Prefecture, the lowest, reaches only 58.8% — a gap of 37.0 percentage points. Teacher training systems are operated by each board of education using its own budget, personnel, and priorities, meaning fiscal capacity and education administration policy are directly reflected in completion rates.

The natural prediction under the three-option system is this: municipalities with higher training completion rates are more likely to select "paper-and-digital combined" or "digital-only," while those with lower rates are more likely to be forced into "paper-only." If teachers cannot effectively use digital textbooks in instruction, even provisioned devices go unused — replicating the South Korean model of "deployed but unused" — and municipalities will default to paper-only as the safe choice. Conversely, municipalities with well-developed training systems will be positioned to maximize the benefits of digital textbooks. The result is a structural entrenchment of qualitatively different learning experiences for children within the same compulsory education framework, differentiated by region.

Sweden's "Return to Print" — The Gap Between Narrative and Reality

Alongside South Korea, Sweden's "return to print" is frequently cited as an international case. But this case involves a substantial gap between the widely circulated narrative of "Sweden returned to print because digital technology caused academic decline" and the actual policy intent.

According to the analysis by Nikkei BP and JETRO, two factors drove Sweden's pivot to paper-based instruction in August 2023. First, Sweden had no clear definition of "textbook" — publishers freely distributed digital educational materials without robust quality assurance frameworks. Second, students' screen time on smartphones and tablets had increased substantially, with declines in reading time and academic performance proceeding simultaneously. The policy focused on two goals: "providing textbooks (print) equitably through subsidies" and "limiting screen time to increase reading time." It was not a rejection of digital educational materials per se.

Japan's digital textbooks are quality-assured through the textbook screening system, which fundamentally distinguishes the Japanese case from Sweden's. Equating freely distributed, unscreened digital materials with screened digital textbooks is analytically imprecise as a policy argument. That said, the risk of "excessive screen time" is a legitimate concern common to Japan as well, and it is the more substantive challenge in discussions of the three-option system. Reporting that reduces Sweden to a "return to print = digital failure" narrative obscures this substance and warrants caution.

On the academic side, Noriko Arai's 『AI vs.教科書が読めない子どもたち』 (AI vs. the Unreadable Children) (Noriko Arai, Toyo Keizai Inc., 2018) demonstrates through empirical data that Japanese students' reading comprehension skills were already critically deficient even before digitalization, providing an important counterpoint to narratives that position digital textbooks as a "solution to the reading comprehension crisis."

The Ministry of Education's Empirical Research: Achievements and Evidence Gaps

This brings us to the problem of the absence of RCT (randomized controlled trial) in Japan. The Ministry of Education's Empirical Research on the Effects and Impact of Learner-Use Digital Textbooks was conducted from FY2019 through FY2021 on students in fifth grade through ninth grade (third year of middle school) in English and other subjects. A large-scale questionnaire survey was added in FY2024.

However, the research design for these studies centers on "documenting usage patterns" and "surveying teacher and student responses." No published RCT rigorously comparing groups that used digital textbooks against groups that did not — to verify the causal effect on academic achievement — exists. As an indirectly related data point, OECD PISA 2022 shows a correlation in which students who use ICT devices for learning score higher on average than those who do not — but this reflects the relationship between ICT environments broadly and academic achievement, not the causal effect of digital textbooks specifically.

In other words, the claim that "introducing digital textbooks will improve academic achievement" is not, as of now, supported by Japan's empirical evidence base. Under the three-option system, the achievement gaps that emerge between municipalities selecting "digital-only" and those selecting "paper-only" may constitute the first large-scale natural experiment to measure causal effects during the four-year transition period. The risk of proceeding with formalization in this evidence vacuum is concentrated in precisely this dimension.

The Three-Layer Mechanism by Which the Three-Option System Entrenches Regional Inequality

The school-type disparities, intermunicipal disparities, and teacher training disparities examined above are not independent problems. Through the institutional design of the three-option system, they operate in concert to produce a three-layer structure that entrenches inequality.

The first layer is school-type disparity. Under the premise of high school coverage at 11.5% versus elementary and middle school coverage at 99.8%, the three-option system produces a massive number of high schools that can realistically only select "paper-only." The standard scenario becomes one in which children educated in a "paper-and-digital combined" or "digital-only" environment in elementary and middle school revert to "paper-only" upon entering high school. This is a structure that severs the continuity of learning experience at the school level transition.

The second layer is intermunicipal disparity. Implementing three-option selection in a context where prefectural coverage rates vary by up to 2.5-fold and training completion rates diverge by 37 percentage points will drive polarization: municipalities with high fiscal capacity and well-developed teacher training systems will move toward digital options, while those with lower capacity and training will select paper-only. Intermunicipal disparities manifest directly as disparities in children's learning opportunities.

The third layer is individual teacher disparity. The fact that 82.2% of teachers nationally can use ICT in instruction means, conversely, that the remaining 17.8% struggle to do so. Even within the same municipality and the same school, the intensity of digital textbook use will vary depending on the individual teacher. When the three-option system operates at the municipal level, a third layer of misalignment arises between the selection result and classroom implementation.

Each of these three layers is problematic in isolation, but their combination amplifies their impact. In the worst-case scenario — which can play out at the regional level — a municipality with low fiscal capacity and low training completion rates selects "paper-only," its high schools also operate paper-only due to low coverage, and the classroom teachers have not attended ICT instructional competency training. Conversely, municipalities with high fiscal capacity and high training rates will advance digitalization across all three layers, with children learning in a digital textbook environment from elementary school through high school. Under the same compulsory education and high school education frameworks, qualitatively different learning experiences are structurally generated for children based on where they happen to live.

"Optional Selection" as a Mechanism for Dispersing Responsibility

The three-option system is presented as "accommodation of regional circumstances" and "self-determination for local governments." Yet from a systems design perspective, it is also a mechanism through which the national government disperses responsibility — avoiding the risk of nationwide uniform rollout failure while transferring the consequences of selection outcomes to municipalities.

From the national government's perspective, the risk of failing through uniform rollout — as in South Korea's AIDT — is avoided. Since municipalities made their own selections, the outcomes become municipal responsibility. But from the municipalities' perspective, their authority and fiscal resources to establish the prerequisite conditions for selection (teacher ICT competency, fiscal capacity, device coverage) are limited — and municipalities that are structurally "forced into paper-only" will emerge. From the children's perspective, the structure in which learning experience is determined by place of birth becomes entrenched.

The original purpose of an optional system is to enable flexible responses suited to regional characteristics and children's individual needs. But when the "space to choose" is already constrained by prerequisite conditions, optional systems become mechanisms that endorse rather than accommodate inequality. How much the bill's three-option system acknowledges this transformation — and what complementary measures are embedded — is the critical question.

The Asymmetry of Cost Burdens — National Treasury Coverage Limited by Subject

The cost burden structure also accelerates regional inequality entrenchment. According to MEXT's "On the Current Situation Regarding Digital Textbooks" (April 2025 version), the current system allocates cost burdens very differently by subject. English in fifth grade through ninth grade is 100% nationally funded; mathematics in the same grades is 50% nationally funded; all other subjects are municipality-funded. According to Britannica Japan, learner-use digital textbooks are priced at 200 to 2,000 yen per student per subject depending on the publisher (totaling 2,000 to 20,000 yen per student across all subjects), while teacher-use digital textbooks range from 70,000 to 90,000 yen per grade level per subject.

The amended bill indicates a policy direction of "including digital textbooks in the same free distribution scheme as print textbooks," but the specific scope and ratio of national treasury coverage is still being worked out in preparation for the April 2027 implementation. During the transition period until then, the structure in which fiscal capacity disparities directly determine adoption rates for nationally unfunded subjects continues to operate. Municipalities with lower fiscal capacity will be compelled to forgo digitalization of non-covered subjects, resulting in simultaneous two-track development: subjects with advanced digitalization alongside subjects that remain paper-only. Next GIGA device replacement subsidies are estimated at approximately 55,000 yen per device, but the funding for device provisioning and content provisioning are structured separately — and the gap between municipalities that can secure both and those that cannot will widen further.

The sister article Digital Textbooks and Their Legal Status: The Significance of the Cabinet Decision organizes the institutional design dimensions and significance of the cabinet decision as a counterpart to this article. Reading this article (on ground reality) alongside that one creates a three-dimensional picture of the transition period's structure.

How to Design the "Four-Year Transition Period"

To summarize: as of May 2026, Japan's digital textbook policy simultaneously embodies an institutional milestone — the April 7 cabinet decision — and an implementation starting point. The approximately four years until full elementary school application in academic year 2030 represent both the final opportunity for regional inequality to become entrenched under the three-option system, and the opportunity for policy design that corrects those disparities.

What is required is a set of complementary measures that maintain optional selection while raising the floor of "prerequisite conditions for selection." Specifically: priority allocation of training programs to municipalities with low training completion rates; separate fiscal sources for high schools to raise high school coverage rates; gradual expansion of nationally funded subjects; and linking device replacement subsidies with content provisioning subsidies. Without these measures, operating the optional system alone risks transforming the three-option system into a mechanism that endorses inequality.

The evidence gap problem must also be addressed in parallel. Elevating the Ministry of Education's empirical research from "documenting usage patterns" to "causally verifying academic achievement impact" — and using the regional variation that will naturally emerge under the three-option framework during the transition period to design quasi-RCT studies that measure causal effects — will provide the foundation for refining the system post-2030.

South Korea's AIDT collapse after one semester demonstrated how dramatically classrooms can respond when textbook formalization outpaces implementation readiness. At the same time, one contributing factor to that failure was the decision to mandate rather than offer optional adoption. Japan's three-option system partially incorporates this lesson — but if the prerequisite conditions for selection are not established, the same problem will be encountered in a different form. Formalization is a starting point, not an endpoint. The design of the four-year transition period will determine whether digital textbooks become a tool for correcting regional inequality or a mechanism for entrenching it.

Reading alongside the sister articles 350,000 School Refusals: The Tectonic Shift in Japan's Education Structure on the surge in school refusals and The Blind Spots of Private High School Tuition-Free Policy on the structural blind spots of tuition-free policy reveals that Japanese education policy as a whole carries a large implementation gap between "those who design the system" and "those who receive it in the field." Digital textbooks exist within that same structure — and whether the design of implementation conditions can successfully eliminate regional disparities will determine the policy's ultimate success or failure.

Related Articles

References

Reference Books

• 『AI vs.教科書が読めない子どもたち』 (AI vs. the Unreadable Children) (Noriko Arai, Toyo Keizai Inc., 2018): Demonstrates through empirical data that Japanese students' reading comprehension skills were critically deficient even before digitalization. A significant counterpoint to narratives positioning digital textbooks as a solution to reading comprehension problems.