Vision 2035, Pillar 2: Reform Assessment

Vision 2035: Advancing English Proficiency for the Future of Japan
← Previous Article: Pillar 1: Time to Raise the Bar

日本語版はこちら.

Starting at the Finish Line: Why Japan’s Entrance Exams Must Lead the Way

In Japan, few events shape a student’s academic life more than the university entrance exam. What’s on the test determines what is taught not only in the final years of high school, but often all the way down into junior high. What this means, sadly, is that until the test changes, the system won’t. That’s why revising the entrance exam is the most critical domino that must fall if Vision 2035 is going to become a reality.

Right now, the Common Test for University Admissions measures English through reading and listening only. Speaking and writing — the “productive” skills that matter most in real-world communication — are absent. The result is predictable: schools devote the bulk of classroom time to passive skills, because that’s what gets students into university.

This isn’t the first time reform has been on the table. In the 2010s, policymakers explored adding a speaking component to the national exam, and pilot discussions were held. But plans never materialized. The concerns were familiar: scoring spoken answers requires trained human examiners, which raises issues of fairness, consistency, and cost. Even with rigorous training, human raters can differ in their judgments, and scaling this up to hundreds of thousands of students would be slow and extremely expensive. For these reasons, the proposals stalled.

Until recently, that seemed like the end of the story.

The Automated Shift

Advances in machine learning and automated scoring systems have begun to remove the very barriers that once made speaking and writing tests unworkable at scale. These systems can evaluate pronunciation, grammar, vocabulary, fluency, and coherence in seconds. Because they apply the same standards to every student, they reduce the inconsistencies inherent in human scoring. And after the initial investment in design and calibration, they can process results rapidly and at much lower cost.

This isn’t hypothetical. The TOEFL iBT already combines automated scoring with human raters for its speaking and writing sections. The GMAT’s Analytical Writing Assessment has incorporated automated scoring for years. The Duolingo English Test is scored entirely by machine learning systems, and its results are now accepted by thousands of universities worldwide.

Two Paths Forward

Japan’s national testing authority has a choice. It can develop its own automated speaking and writing component, tailored to the Common Test. Or it can recognize scores from established four-skill exams — TOEFL, IELTS, Duolingo, Eiken — and allow universities to count them toward admissions.

Either path would send a powerful signal: that productive skills matter.

And that signal shouldn’t stop at the university level. High school entrance exams also need to measure speaking and writing. If younger students know they will be tested on these skills to advance into competitive high schools, preparation will begin much earlier. This would accelerate reform across junior highs and even into elementary schools, embedding productive skills into the learning journey from the start.

Keeping it Fair

Access is the challenge that must be addressed head-on. If external exams are accepted, their fees should not become a barrier. Government subsidies or vouchers for low-income families could ensure every student has the same opportunity. Test centers — physical or online — should be available nationwide so geography doesn’t determine access.

The payoff would be significant. International studies — including OECD reports and the EF English Proficiency Index — consistently show a strong correlation between higher national English proficiency and positive economic outcomes, from innovation to global competitiveness. In other words, investing in modern, equitable language testing is also an investment in Japan’s future growth.

The First Domino

When the finish line includes speaking and writing, schools will prepare students for speaking and writing. High schools will adapt. Junior highs will adapt. Elementary schools will adapt. Curriculum, teacher training, and student priorities will all shift in response.

But it won’t happen in reverse. No amount of well-meaning reform at the lower grades will stick if the university and high school entrance exams remain the same.

Change the tests, and the rest will follow. That’s how Vision 2035 begins.

Next Article: Pillar 3: Flip the Model: Focus on Speaking →