Imagine someone is building you a skyscraper. Beautiful, glass, “smart”. There is just one catch: the foundation has been poured from sand, the rebar was forgotten, and the cracks have been painted over so that it looks respectable at the presentation.
That is roughly what education looks like today in many countries of East Asia and the Pacific, according to a major World Bank review of foundational skills. The region is richer, more technologically advanced and more self-confident, and yet children are often weak at the very basics without which neither “digital”, nor “AI”, nor the “knowledge economy” will ever get off the ground.
This is not about “complicated integrals”. It is about the foundation: being able to read with understanding and being able to do basic arithmetic.
To keep this concrete, take Thailand. According to international estimates, the share of ten-year-olds who cannot read and understand a simple text is about 23%. The World Bank calls this “learning poverty”, and the figure predates the pandemic. And this is happening in a country that, objectively, does not look poor by the standards of the region.
Why the problem doesn’t solve itself
There is a popular hope: “let’s wait, the generation will adapt”, “digital kids will figure it out themselves”, “the internet will fix everything”. But basic skills do not grow like that.
Reading with comprehension and arithmetic are not “content” that you can simply “consume”. They are mechanisms that have to be trained. If a child in primary school has not learnt to link letters into meaning, and numbers into understandable operations, catching up becomes more and more costly later on. Because new material builds on the old.
It is like trying to learn physics without having mastered fractions. The words seem clear enough, but the brain keeps tripping over simple steps, and the energy goes into survival rather than understanding.
The pandemic exposed the hidden architecture of inequality
In calm years, a system can look “normal on average”. But once a mass disruption hit, it became clear what was load-bearing and what was scenery.
Families with resources more often had access to stable internet, devices and a decent form of remote learning. Families without resources often simply dropped out: some children studied “on a phone, with whatever was left over”, others stopped taking part altogether. And when children come back to class carrying different “inner baggage”, a school that cannot quickly help those who have fallen behind turns into a mechanism that locks the gap in.
The problem here is not only poverty. The problem is that a system has to be able to deal with deviations from the norm. And often it can’t.
“Give us money” is not always the right answer
Money matters, but education has an unpleasant property: you can spend a lot and get weak results if the spending is badly designed.
First, in a number of countries the education budget really is lower than in countries with comparable incomes. That is an obvious place to improve: without resources it is hard to demand quality.
But even when the money is there, it can be distributed so that strong schools become stronger while weak ones stay chronically short of resources. Rich neighbourhoods get more good teachers, more support and more “opportunities”; poor ones get less. In the end the system “averages out” on paper, but inside it the gap between children’s paths only grows.
Thailand is a good example here of how “not being poor” does not guarantee “a foundation”. According to the World Bank, public spending on education was about 2.5% of GDP in 2023, down from more than 4% ten years earlier. So the share is not just modest, it is shrinking. But money alone will not fix this: if distribution within the country is set up so that weak areas get less support, the overall average will hide the real failures.
The paradox is that education is not just an “investment”; it is also the engineering of distribution. If the weak link is underfunded, it will limit the whole chain.
The central figure is the teacher, but not in the romantic sense
When people say “it all depends on the teacher”, they usually mean personal heroism: get an inspiring teacher and everything will take off. It is a beautiful myth, but at the level of a country it is useless.
A normal system cannot rely on thousands of people heroically making up for its mistakes every day. It has to make it possible for the average teacher to teach well, not “despite” the system but “thanks to” it.
Here it helps to think like an engineer: there are inputs, there is a process, there is feedback.
A system that wants to raise basic skills at scale usually runs up against three things.
Who we let into the profession
If the filter at the entrance to the profession is weak, the system pays for it for decades afterwards: a weak start is hard to make up for. This is not about “elitism” but about something simple: teaching is a profession that needs specific competences. And they need to be tested.
How we strengthen those already working
Most teachers are already in schools. You cannot “reset” them and replace them. So you have to be able to raise the quality of teaching among the existing workforce, quickly and at scale.
And here is the key point: one-off courses and nice presentations usually have a weak effect. Because the reality of the classroom quickly “washes away” new ideas. What works is not “attended a training session” but “changed everyday practice”.
Why it pays a teacher to do the job well
A system always shapes behaviour through incentives. If effort makes no difference to a career path, if a good result cannot be told apart from a box-ticking one, if there is no feedback, then over time even motivated people start saving their energy.
This is not about “lazy teachers”. It is about the fact that the human mind adapts to the rules of its environment.
What helps right in the classroom
There are a few tools that look boring, and that is exactly why they work: they do not require superpowers, they reduce variability and even out quality.
Structured lesson plans
This is an idea that meets resistance: “what do you mean, a lesson from a script, I’m not a robot”. But the point is not to turn the teacher into a robot. The point is to build good methods into the daily routine.
If a teacher lacks a strong grounding in teaching methods, a structured lesson is like rails: they lead to the result and stop things sliding into chaos. Over time, some of those “rails” turn into the teacher’s own skill.
Mentoring and coaching
If a teacher is changing their practice, they need feedback: what worked, what didn’t, where they simplified too much, where they overloaded the lesson. Without support, most changes roll back. Not out of malice, but because routine is stronger.
A mentor works here as “external quality control”, but a gentle kind: not to punish, but to help a new norm stick.
Transparent incentives
Incentives are a fraught subject, because they are easy to get wrong. Measure results crudely and teaching to the test begins. Reward only individuals and you get competition instead of cooperation. But if the design is careful (clear metrics, team elements, a link to professional growth), incentives can genuinely make effort more consistent.
It is important to understand: the goal is not to “force”. The goal is to make good work visible and to give it consequences.
Why reforms often break down not on methods but on power
And now we come to the most “grown-up” part: education is a huge system for distributing resources and influence. Appointments, transfers, posts, bonuses are all levers. And where those levers are built into political or informal mechanisms, any attempt to make the system more transparent starts to meet resistance.
That is why many reforms look the same: on paper everything is right, at the start everyone is happy, after a year or two people start going through the motions, and everything goes back to “the way it was”.
This is not because “people are bad”. It is because the system has its own internal logic of self-preservation. And if a reformer does not take that logic into account, their decisions will be “digested” and turned into scenery.
What a successful repair looks like
Successful reforms usually involve no magic, just a boring sequence of steps.
A strong team and political stability: a reform needs time, otherwise it will not have the chance to take hold.
A coalition of support: teachers, parents and administrators need to understand what exactly is changing and why. Otherwise the system will defend itself.
Gradual roll-out with adjustments. A big bang rarely works in social systems: there are too many unknowns.
And sometimes, honest compromises that reduce resistance. Not because anyone “loves” compromises, but because otherwise the changes simply do not survive long enough to produce results.
The main point
To put it very simply: a modern economy can dream as much as it likes about digital hubs, artificial intelligence and high technology. But all of it rests on human capital. And human capital does not start with “AI courses”. It starts with a ten-year-old who reads and understands what they have read, and does sums with confidence.
Thailand is a telling marker: when roughly a quarter of children cannot read with understanding by the age of ten, this is no longer “a problem for individual schools” but a constraint on the country’s whole trajectory, from labour productivity to people’s ability to retrain.
And repairing the foundation is not an inspiring slogan or a “new strategy”. It is engineering work: even out the quality of teaching, give teachers tools, build in support and set up incentives so that the system produces results consistently rather than by chance.
References:
World Bank. Fixing the Foundation: Teachers and Basic Education in East Asia and Pacific (Washington, DC: World Bank, 2023).
The numbers behind Thailand’s digital ambitions: BOI investment in data centres and electronics.
.C56YFDQF_Zj6Uaw.webp)