Every few years, the release of results from the Programme for International Student Assessment (PISA) triggers another round of debate about education. Countries are ranked against one another, common features of high-performing systems are identified, and people ask what their own education systems are getting wrong. Some see evidence for the effectiveness of exam-oriented education. Others see a crisis of educational equity. Still others argue that standardized tests are inherently inadequate measures of educational value.

In September 2026, the Organisation for Economic Co-operation and Development (OECD) released the results of PISA 2025. The findings were troubling: average scores in reading, mathematics, and science across OECD countries were at or near their lowest levels in the history of the assessment. Compared with 2015, average reading scores had fallen by 28 points and mathematics by 22 points, while science had also experienced a longer-term decline.

These trends deserve serious attention. But what, exactly, should we make of them?

Are students becoming less capable of learning? Have schools lost their educational effectiveness? Are digital media changing the way people think? Or is the broader system through which we organize education facing a more fundamental problem?

Rankings make it easy to reach conclusions. They do not necessarily help us understand the underlying issues.

1. What Does a Decline in Scores Mean?

PISA is valuable partly because it attempts to measure whether 15-year-old students can use their knowledge of reading, mathematics, and science to understand problems, interpret evidence, analyze information, and participate in modern society. It is not simply a test of how much textbook content students can recall.

A sustained decline in scores therefore cannot be dismissed as an inconsequential statistical fluctuation. Reading ability, in particular, deserves attention. Students need to evaluate the reliability of information, integrate evidence from different sources, distinguish facts from opinions, and form judgments when confronted with conflicting claims. These abilities affect how people understand the world, not merely how they perform in school.

Yet we should be careful about interpreting the results. Falling scores tell us that student performance on a particular assessment has changed. They do not, by themselves, explain why. Family environments, teaching methods, learning motivation, digital media, and broader social changes may all play a role. Blaming smartphones, artificial intelligence, or a particular teaching method would simply replace an unresolved question with a familiar explanation.

The more important questions are these: Why have scores declined across multiple education systems over an extended period? Do these changes share common causes? If so, do those causes originate within schools, or in the wider social environment in which schools operate?

PISA alone cannot answer these questions. But it gives us reason to consider whether the challenges facing education extend beyond classroom techniques.

2. Do High Scores Mean Better Education?

PISA encourages an intuitive conclusion: if countries can be compared by their scores, some must be doing better than others. We should identify what successful systems have in common and replicate their practices.

But this approach overlooks a fundamental question: What, exactly, are we comparing?

The four Chinese regions participating in PISA—Beijing, Shanghai, Jiangsu, and Zhejiang—as well as Singapore have performed strongly in the 2025 assessment. Both have achieved results well above the OECD average, particularly in mathematics and science.

These results demonstrate strong performance in the competencies measured by PISA. But can we conclude that exam-oriented education is more effective? That centralized curricula work better? That the more centralized the governance of education, the higher the student achievement?

No.

First, education systems with similar results can have very different institutional arrangements. Japan, South Korea, and Estonia, for example, have also performed strongly, yet they cannot simply be classified as instances of one common educational model. Second, high scores may reflect a combination of factors, including foundational knowledge, teacher expertise, family investment, curriculum design, and time spent learning. Country rankings alone cannot tell us how much each factor contributes.

More importantly, PISA measures specific learning outcomes, not the full value of education.

A student who can solve standardized test problems proficiently has demonstrated relevant knowledge and skills. But can we infer from that performance that the student is equally capable of identifying new questions, exploring the unknown, completing long-term open-ended projects, or making judgments in situations without predetermined answers?

These abilities can be investigated through other forms of assessment, but they cannot be read directly from PISA scores.

We should therefore avoid two opposite mistakes. PISA is not merely a test of memorization, but neither should its results be equated with overall educational quality. It is a useful measuring instrument, not a measure of everything that matters.

3. Where Does Educational Authority Actually Lie?

If we move beyond rankings and ask how high performance is produced, we encounter a more fundamental question: Who defines educational success?

Educational authority does not necessarily mean political authority. It can also operate in a highly marketized society. The issue is who gets to decide what knowledge is worth learning, what counts as success, and which educational experiences are recognized by society.

A system may have a common curriculum, shared foundational standards, and rigorous assessment. None of these arrangements is inherently harmful. Without a common foundation of knowledge, students may struggle to advance their learning. Without public standards, it may also be difficult to ensure educational quality.

The problem arises when common standards stop functioning as a foundation and become the only accepted route to success.

Suppose a society treats entrance examinations, academic credentials, and admission to a small number of elite institutions as the primary measures of individual ability. Even if schools compete fiercely with one another, that competition may not produce genuinely diverse forms of education. Schools may simply adopt different strategies to compete for the same credentials.

Schools may be public or private, government-run or market-based, traditional or technologically advanced. But if everyone must pass through the same narrow gateway to gain social recognition, how much have these differences really changed?

This is where we need to reconsider educational authority.

The key question is not merely who sets the rules. It is how strongly those rules constrain individual choices, and whether people have realistic alternatives.

4. Can Marketization Create Educational Freedom?

Does an increase in the number of schools mean that students have more meaningful educational choices? Does the ability of parents to pay for a school mean that students genuinely possess educational freedom?

Not necessarily.

Marketization can enable different institutions to experiment with curricula, teaching methods, and educational goals. It may allow needs neglected by a uniform public system to be addressed. This is one way market mechanisms can contribute to diversity.

But competition can also revolve around existing definitions of success. If parents prioritize admission rates, schools prioritize rankings, and students prioritize entry into elite universities, competition may push institutions toward increasingly similar strategies.

In that case, the market produces more competitors for the same credentials, not greater diversity in educational goals.

OECD analyses of PISA 2022 offer a relevant caution. After accounting for the socioeconomic backgrounds of students and schools, the association between school competition and higher mathematics performance was no longer statistically significant in most education systems. Competition, in other words, cannot simply be assumed to improve educational quality.

Choice itself can also become a scarce resource.

Some families can afford tuition, transportation, and additional learning costs. They may have access to better information and the flexibility to change schools when an arrangement proves unsuitable. Other families face constraints involving income, location, time, and information. The same education system may offer meaningful choices to the first group while providing only nominal choices to the second.

The important questions, then, are not simply how many kinds of schools a society has, but how many socially recognized pathways to success it offers. It is not merely whether parents can choose a school, but whether students can move between different learning pathways without sacrificing their future opportunities.

If a student does not thrive in a mainstream school, can they continue developing through vocational education, project-based learning, alternative curricula, or other routes? Can these experiences lead to recognized qualifications and social opportunities? Must the student pay a high financial price to leave a pathway that does not suit them?

These questions are closer to the substance of educational freedom than the public-private school ratio.

5. Educational Equity Means More Than Making Everyone Take the Same Test

PISA 2025 also reports an important finding: across OECD countries, students' socioeconomic status accounts for approximately 12% of the variation in science performance. Family background remains significantly associated with learning outcomes.

This does not mean that family background determines an individual's future. It does, however, suggest that education systems have not eliminated the influence of family resources on student achievement.

We must also avoid a different mistake: assuming that differences in test scores and social stratification are the same thing.

A widening achievement gap means that student performance is becoming more dispersed. A stronger association between family background and achievement means that educational outcomes are becoming more dependent on socioeconomic origins. Declining intergenerational mobility means that a person's education, occupation, or income is increasingly associated with the social position of their parents. These phenomena may be related, but none is a substitute for the others.

One finding deserves particular attention. Between 2022 and 2025, socioeconomic achievement gaps narrowed on average across OECD countries. Yet in reading, mathematics, and science, this narrowing was driven largely by declining performance among advantaged students, rather than sustained improvements among disadvantaged students.

If advantaged students fall behind while disadvantaged students make little progress, it would be premature to call the result an improvement in educational equity.

The real questions of equity are these: Do students from different backgrounds have comparable opportunities to acquire a solid foundation of knowledge? Can they change their developmental trajectories through learning and effort? When one form of education does not suit them, do they have other viable routes?

If an education system can make everyone participate in the same competition but cannot provide different pathways for different individuals, formal uniformity does not necessarily amount to substantive fairness.

Conversely, diversity does not mean allowing every student to choose arbitrarily. Students need foundational knowledge. They also need exposure to possibilities they may not yet recognize, along with the capacity to take responsibility for their choices. Genuine educational freedom requires both choice and the conditions that make choice meaningful.

6. PISA May Ultimately Force Us to Reconsider the Purpose of Education

PISA can tell us how students perform in reading, mathematics, and science. It can reveal how learning outcomes are distributed and help us investigate the relationship between family background and achievement.

But it cannot independently answer a more fundamental question: What kind of education is worth pursuing?

What do we want education to develop in people? Someone who can rapidly solve standardized problems, or someone who can formulate questions, evaluate evidence, understand others, and continue learning throughout life?

These capabilities are not mutually exclusive. Foundational knowledge, rigorous training, and independent exploration can reinforce one another. The problem arises when educational resources and social recognition become concentrated around a narrow set of measurable indicators. Do other important abilities gradually lose their place?

If schools continually reduce other learning activities to improve test scores, we should investigate the long-term costs. If market competition leads schools to adopt increasingly similar strategies for university admission, we should ask what the competition is actually creating. If family resources determine who can access better educational choices, we should reconsider what we mean by educational freedom.

PISA scores cannot directly answer these questions. But their inability to measure the full range of educational value is no reason to ignore them.

At a time when artificial intelligence is changing access to knowledge, information processing, and work, we need to distinguish between two things: whether students possess necessary knowledge, and whether they can independently judge, create, and act in an environment where information and answers can be obtained rapidly.

The first remains essential. Without foundational knowledge, talk of creativity can easily become empty rhetoric. But if education ultimately consists only of abilities that can be measured, compared, and ranked, are we developing people who can understand and change the world, or people who are exceptionally good at adapting to existing evaluation systems?

This is not an argument against examinations, nor a denial of the value of standardized assessment. On the contrary, because reliable evaluation matters, we should be clear about its limits.

PISA deserves attention because it warns us that many students are struggling to develop foundational learning capabilities. But it also reminds us that scores are not the whole of education. We need to know not only how well students are learning, but why they are learning as they are; not only what the average performance is, but who is being left behind; and not only what an education system produces, but how much room it leaves individuals to choose their own futures.

Ultimately, education must confront a question far older than standardized testing: What should human beings gain from education?

If we have not answered that question, then however much we improve examinations, curricula, and teaching technologies, we may simply be making competition more sophisticated within a set of standards that we have never seriously examined.

Whether PISA scores rise or fall, they cannot answer that question for us.

Sources