Everyone Fails Somewhere
How Asia avoids grade inflation, and why their model doesn't work either.
A few weeks ago, I published an article about grade inflation and top marks around the world, which got quite a bit of traction. The journalist Héctor García Barnés even covered it in ElConfidencial. Well, there’s more to say:
On May 20, 2026, Harvard’s Faculty of Arts and Sciences formally voted to cap A grades at a maximum of 20% of students per course, with an additional four discretionary top marks per class. The result was 458 votes in favor and 201 against, and the measure will take effect in the 2027-2028 academic year, with a review after three years. It is the most severe grading restriction adopted by an elite American university in decades, and it comes at a moment when more than 60% of Harvard’s undergraduate grades are already A’s, compared to just a quarter twenty years ago.
Harvard is not the first to try. Princeton did it in 2004 with a 35% cap, and abandoned it ten years later because, according to the rector, it had become a source of stress. Grades immediately returned to their previous levels. The institution that most shaped Anglo-Saxon academic culture throughout the twentieth century is now admitting that its grades have stopped meaning anything.
The natural question is: where do grades still mean something? There is more to it, though: at the end of my previous article, I already suggested that this grade inflation is not entirely global, and that there are a handful of universities holding out against it.
They are Asian universities.
Where grades still make a distinction
In Singapore, the National University (NUS) and Nanyang Technological University (NTU) use the bell curve as an explicit grading moderation mechanism to prevent both inflation and deflation. When a professor at SMU decided to give an A to all 169 students in a business module, the institution immediately reviewed the grades. The professor himself declared that the guidelines were “actually rules,” and that regardless of how well students performed, the number of top marks was strictly limited.
In South Korea, many universities enforce grade distribution policies, especially in compulsory courses, capping the percentage of students who can receive an A or a B, with the explicit aim of maintaining academic rigor and preventing inflation. In Japan, grade inflation is generally less prevalent, making high grades harder to achieve. In China, the Gaokao is the only real filter: more than nine million students compete for fewer than seven million university places. What counts is what counts. At China’s most prestigious institution, Peking University, only 1 in every 4,000 applicants gains admission.
The difference is not one of resources or pedagogy. It is philosophical. In Asia, a grade is still a signal of position in a competition. In the West, it has become an entitlement earned simply by showing up.
Why not us
This has a cultural explanation that is uncomfortable to say out loud: in the West, we have stopped believing in academic meritocracy as a legitimate principle.
It is not that we say so openly. But the system has spent decades making decisions that lead exactly in that direction. Every reform that reduces the formal demands of an exam, every student satisfaction survey that affects the career of the professor who fails the most, points to the same place: grades must not hurt. Grades must include.
In Asia, that logic has not taken hold with the same intensity, because academic meritocracy retains social legitimacy as a genuine lever of mobility. In China, the Gaokao was restored in 1977 precisely as an antidote to the political arbitrariness of the Maoist period. The exam is perceived as fair by students from rural backgrounds because it depends on effort, not on special equipment or extraordinary opportunities: access to textbooks is enough. The grade, in that context, is the most neutral thing the system can offer.
So how do they avoid it?
The uncomfortable answer is that Asia has not solved grade inflation through cultural virtue. It has contained it through concrete mechanisms, and each of those mechanisms carries its own cost.
In South Korea, the relative evaluation system with fixed quotas did not emerge from some ancient tradition of rigor. It arose in 1997, after the financial crisis, because grade inflation had become a serious problem and the government began withdrawing subsidies from universities with grades that were too high. In other words, the same tendency we see in the West appeared there too, and the response was regulatory intervention. The result is a system where SKY universities cap the percentage of students who can receive an A at between 30 and 35% of those enrolled per course. An excellent student can receive a B+ simply because the A quota was already filled by classmates who scored half a point higher. The grade does not certify that you have mastered the subject: it certifies that you are better than the person sitting next to you. And when the system relaxes even temporarily, grades rise on their own: during the pandemic, many Korean universities switched temporarily to absolute grading because the curve did not work well in remote classes, and that window was enough for inflation to reappear.
In China, the internal pressure is so intense that Peking University itself launched a pilot in 2024 in its School of Life Sciences to abandon the numerical GPA ranking and switch to simple letter grades, precisely because students were trapped in unnecessary competition over fractions of a point that the system could no longer sustain. Grade deflation solves inflation, but produces its own pathology: a race of exhaustion over thousandths of a point among people who were already the elite before they even arrived.
Japan is perhaps the only case where the problem does not exist in an acute form, but for a very specific reason: in the Japanese labor market, what university you attended matters a great deal, and what grade you received matters very little. If the grade is not the relevant signal, no one has an incentive to inflate it. The problem disappears because the credential that counts is a different one.
The Asian model does not work either
This is where the article would end if everything above were simply admirable. But as Lucy Crehan explains in her book Cleverlands, it is not.
The academic pressure of the Korean system has contributed to one of the highest youth suicide rates among OECD countries, with a teenage rate that reached 7.9 per 100,000 inhabitants in 2023, particularly linked to education-related stress. In 2022, 20.3% of South Korean high school students had experienced suicidal ideation due to anxiety about their future and career. In China, the response of the younger generation to the system is not activism but exhaustion: the growing interest in tangping, “lying flat,” and in the concept of neijuan or “involution,” reflects an escalating social competition for ever smaller gains and a possible crisis of faith in meritocracy itself. As we discussed in this blog, a growing number of Chinese graduates were finding no employment commensurate with their qualifications, following an intense government push to funnel all students into higher education.
The system that demands produces individuals who, once past the filter, find no reason to keep running. The system that gives things away produces individuals who carry a diploma that taught them nothing. Both models have a built-in failure, and neither seems uncomfortable with it.
What Harvard has understood, belatedly, is that an institution that gives top marks to 84% of its students is not evaluating: it is managing public relations. As professor Alisha Holland, the driving force behind the reform, declared, the goal is to “restore public confidence in universities and ensure that our internal standards mean something.”
Even so, the measure faces resistance from students who consider it unfair. That resistance is the clearest symptom that the problem is not one of institutional design but of social expectations. A grade that selects is perceived as an aggression when the cultural norm assumes that selecting is discriminating.
Both the Korean student who studies sixteen hours a day to avoid falling to the ninth rank in their cohort, and the Western student who arrives at university expecting to be confirmed as already excellent, share the same anxiety: that their worth depends on what a piece of paper says.
That is the problem no grading system can solve on its own.
We’ll see.