When Schools Do More Than Teach (updated August 9, 2026)

When PISA scores fall, the familiar explanations point at the child—that Thai students do not read, cannot concentrate, are addicted to their phones, or come from inattentive families. Likewise, when violence erupts, society rushes to locate the “troubled child” and elevates individual pathology into the primary explanation. In social psychology this tendency is known as the fundamental attribution error: the inclination to explain behavior through a person’s inner traits while systematically underestimating the influence of context and structure.

Individual factors are real and should not be denied. But stopping the analysis there makes us miss the systemic question, which carries far greater explanatory power once a phenomenon appears at population scale.

This matters all the more because PISA does not chiefly measure the recall of textbook content; it assesses the capacity to apply knowledge—to read for meaning, evaluate information, reason scientifically, and solve problems in novel situations (OECD, 2023a). A sustained decline in scores therefore signals not merely that “children cannot memorize formulas,” but a limitation in the capacity to think and use knowledge— the foundational competence of citizens and workers in a knowledge economy (Hanushek & Woessmann, 2015).

Explaining the PISA 2022 results by the effects of COVID-19 alone risks a misdiagnosis, because OECD data show that Thailand’s downward trend began to appear around 2012–2015—several years before the pandemic (OECD, 2023b).

Between 2012 and 2022, Thailand’s average scores fell by roughly 30 points in both mathematics and science, and about 60 points in reading. Expressed in developmental terms, that decline in reading exceeds the progress a typical 15-year-old gains from a full year of schooling.² Moreover, compared with 2012, the proportion of students scoring below the baseline rose by 19 percentage points in mathematics, 32 in reading, and 19 in science (OECD, 2023b).

The policy implication is decisive: Thailand faces not merely “post-COVID learning loss,” but a structural learning decline—a deterioration of outcomes embedded in the system and accumulating across a decade. COVID is not the cause but an accelerant that made a chronic condition suddenly visible.

Counting strategic plans, curriculum revisions, teacher-development programs, assessment reforms, technology rollouts, and new policy announcements, one must concede that Thailand has almost never stopped reforming its education. The real question is therefore not “why no reform?” but “why do so many changes take so long to register in the classroom?”

Part of the answer lies in a power structure that remains centralized and steeply hierarchical:

In such a structure, policies, directives, indicators, evaluation forms, and projects flow downward, while information about children’s real lives—and the voices of the teachers closest to them—travel back up to policymakers far less easily. Lipsky’s classic account of the street-level bureaucracy holds that teachers and frontline staff are where policy is effectively re-created at the point of practice; yet when overwhelmed by demands amid scarce resources, they inevitably develop coping mechanisms that can distort the original intent of a policy (Lipsky, 2010).

Building on this, I propose the concept of the bureaucratic compression of learning. No one in the system need harbor ill intent; but when every layer of the chain of command sends down “just a little” more reporting, evaluation, and administrative work, what reaches the teacher becomes an enormous pile of paperwork—until time for the most essential task is squeezed out. That task is knowing the child.

A modern education system holds abundant data—test scores, attendance rates, instructional hours, teacher and school evaluations, budgets, and countless indicators. But there is another kind of data, harder to measure and perhaps more vital to a child’s life: which child is bullied regularly; which child has no one to eat lunch with; which child has begun to miss school in unusual patterns; which child is intensely angry; which child is struggling at home; which child is withdrawing from the group; which child is gifted yet bored; and which child does not know whom to turn to in distress.

These, too, are educational data—they simply do not fit neatly into a reporting field. For this reason, an act of violence should not lead only to questions about adding security guards or searching more schoolbags. Physical measures have their place, but research on school-violence prevention converges on a common point: effective prevention is a proactive system that begins long before the day of the event, not a reactive measure at the school gate (Thapa et al., 2013).

Schools therefore need systems that can see conflict, bullying, isolation, intimidation, behavioral change, and warning signs—together with channels through which children can report a problem without fear of being condemned or labeled a “troubled child.”

Many schools still treat bullying as a dispute between two children, or apply the simple formula, “shake hands and it’s over.” Yet in scholarly terms, bullying has three defining features that separate it from an ordinary quarrel: repetition, a power imbalance, and an intent to harm the dignity or safety of the other party (Olweus, 1993).

Understood this way, the remedy cannot be confined to punishing the aggressor or teaching the victim to “toughen up.” It requires changing the culture of the whole school into one that offers no reward for humiliating others and that builds the capacity to manage conflict without violence—consistent with the evidence from whole-school anti-bullying programs that have proven effective in many countries (Olweus, 1993).

Thailand tends to answer problems by adding posts, projects, or evaluation forms. But sometimes the more important answer is to give teachers their time back. A teacher burdened with heavy administrative work, projects from multiple agencies, evaluation paperwork, and duties unrelated to teaching inevitably has less time to observe children, talk with them, and design learning.

Reducing teachers’ load is therefore not only a matter of teacher welfare; it is at once a student-safety policy and a learning policy. A teacher who has time and knows her students is more likely to notice small changes before they become a full crisis. The caring relationship between teacher and student is thus not a soft secondary concern but a structural condition of both learning and safety (Noddings, 2005).

World Bank reports consistently show that Thailand’s educational outcomes are strongly associated with economic status and geography, with disparities between high- and low-socioeconomic-status schools more pronounced than the averages for the East Asia and Pacific region and for the OECD (World Bank, 2023). Children from disadvantaged families are more likely to drop out of the system, show lower learning outcomes, and have fewer chances of reaching higher education.

This makes the slogan “anyone who works hard will succeed” only partly true. One child may begin with a private room, books, a computer, internet access, parents who help with reading, and a well-resourced school, while another must share a phone with the family, help parents work, study in a school without a full complement of teachers, or travel far to attend—yet both are measured by the same ruler at the finish line.

Educational equity is therefore not “giving everyone the same” (equality) but giving each child what they need to develop their potential in a meaningful way (equity). In this sense, the aim of a just education is to expand each child’s capabilities—the substantive freedom to be and to do what they have reason to value—rather than merely to distribute resources equally in formal terms (Sen, 1999).

World Bank data make a pointed observation: although Thailand allocates a high share of its budget to education, the spending is “inefficient,” and per-student expenditure from Grade 1 to Grade 9 stands at roughly USD 27,271 (in PPP terms)— less than one-third of the OECD average. Crucially, compared with countries at a similar level of per-student spending, Thailand’s results are lower than expected (World Bank, 2023).

The problem cannot be solved by the slogan “increase the education budget” alone. One must go further and ask how the money is allocated: which schools receive too little; what share is absorbed by administrative structures; how much autonomy schools have over their resources; and whether each additional baht actually changes a child’s experience in the classroom—especially given the World Bank finding that investment in financial, human, and digital resources is especially low in disadvantaged and rural schools (World Bank, 2023).

Thailand still has many small schools that lack teachers, equipment, and resources, but merging schools by a single nationwide formula is not the answer, since some sit in remote areas and form vital community infrastructure. Policy must therefore differentiate by context—which schools should be networked, which should share teachers, and which must remain and receive additional funding—rather than deciding by headcount alone.

Invoking a foreign leader’s message for comparison does not imply that the education system of that country is a perfect model. It, too, faces inequality, school violence, cultural conflict, gaps in access to mental-health services, and severe disparities in school quality. What is interesting, then, is not one country outdoing another, but the language a state chooses to use with its children.

Such language positions the child as an agent—one who can think, choose, create, and participate in society. It is worth turning the question back to Thailand: how do we speak to our children? Do we teach them to obey more than to question? Do we prize hairstyles, uniforms, and outward discipline above curiosity? Do we award marks for correct answers while leaving little room for good questions?

Rules are necessary for living together. But when rules become the highest goal, a school may succeed at producing children who “stay in line” while failing to form human beings who can think when the world offers no line to follow.

For decades we have discussed education chiefly as an instrument for producing a workforce for the economy and the labour market. This “human capital” frame is important, and empirically supported: the quality of learning is associated with long-run economic growth (Hanushek & Woessmann, 2015). Yet the frame alone is insufficient, because schools do not produce only future workers—they are developing human beings in the present.

We should therefore move from the notion of Human Capital toward a broader frame: Human Capacity, grounded in the idea that the aim of development is to expand a person’s freedom and ability to live a life they have reason to value (Sen, 1999). A child should leave school able to read for understanding, not merely to decode; to ask questions, not merely to answer tests; to distinguish data, evidence, and opinion; to work with those who think differently; to cope with disappointment; to manage anger; to ask for help when needed; and to know their own worth without having to diminish anyone else.

The central lesson of several decades is that Thailand may need not “new programs” so much as a change to the operating system of its education. Studies of educational change converge on the point that technical reforms that never touch the culture and power relations of the system tend to end as changes on paper rather than in the classroom (Fullan, 2007; Sahlberg, 2011). The proposals below therefore aim to change the system’s logic, not to add to the volume of activity.

We could write the finest curriculum in the world, yet if, in the classroom, children fear to ask questions, teachers fear administrators, administrators fear the service area, the service area fears the ministry, and the ministry fears mistakes, innovation will scarcely occur. Learning depends on a measure of psychological safety—the shared belief that interpersonal risk-taking (asking a question, admitting one does not yet know, experimenting and failing) will not be punished or humiliated (Edmondson, 1999).

A system that punishes error too harshly produces people who try “not to be wrong” rather than people who try to discover something new. This may be one reason so many technical reforms fail to produce qualitative change: we changed the curriculum but not the relations of power in the classroom (Fullan, 2007).