Category: The Invisible Cage

On education, credentials, and the difference between human beings and human capital

On August 10, 2026, millions of families across Bangladesh waited for a number.

Not a name.

Not a story.

Not a measure of what a child had become.

A number.

The SSC results were published that morning. Across the country, students opened their phones, checked websites, waited for SMS messages, and sat quietly while parents stood nearby. Some families celebrated. Some cried. Some immediately began comparing.

The overall result was not particularly comforting. The pass rate across the 11 education boards was 62.25 percent. Only 116,676 students achieved GPA-5. The pass rate was the lowest in nineteen years.

For the students who passed with good results, there was relief. For those who did not, there was disappointment. And for many families, the next question had already begun.

Which college can you get into?

Which group will you take?

What was your GPA?

What did your friend get?

What did the neighbour’s son get?

The result had barely arrived before the next competition was already waiting.

I have seen this cycle before. Most of us have. We grow up inside it so completely that we stop noticing how strange it is.

A child starts school. Then comes homework. Then private tuition. Then coaching. Then another teacher. Then another exam. Then SSC. Then college admission. Then HSC. Then university admission coaching. Then university. Then another degree. Then job preparation. Then the job.

And somewhere along the way, we start calling the whole process education.

But I have started wondering about that word. Is this really education? Or have we slowly built something else and simply kept the old name?

The Number Becomes the Child

Look at what happens on result day. A student receives a GPA. The school receives a result sheet. The parents receive something they can compare. The institution receives something it can advertise.

A GPA-5 student becomes a success story. A school with many GPA-5 students becomes a “good school.” A school with few becomes a problem.

The number moves from the student’s marksheet to the school’s reputation. Then from the school’s reputation to the parent’s social standing. And finally, back to the child. The child carries all of it.

The number moves from the student’s marksheet to the school’s reputation. Then from the school’s reputation to the parent’s social standing. And finally, back to the child. The child carries all of it.

I do not think GPA-5 is meaningless. A student who can consistently study, understand material, prepare for an examination and perform under pressure has developed abilities that matter. But it tells us only some things. It does not tell us whether the child can think independently. Whether he can solve an unfamiliar problem. Whether she can work with people. Whether they can make a decision without someone telling them what to do. Whether they can create something. Whether they can handle failure. Whether they know what they actually want. Whether they can build a life.

The examination cannot measure all of that.

So why have we allowed one number to become such a powerful description of a person?

The Strange Success of a School

There is another part of the result season that has always interested me. Some institutions produce extraordinary results. Almost everyone passes. Large numbers receive GPA-5. Then another school, sometimes not very far away, produces a completely different picture. Students fail. GPA-5 becomes rare. Some institutions have no successful candidate at all.

In the 2025 SSC results, 984 institutions achieved a 100 percent pass rate, while 134 institutions recorded zero passes. The national pass rate that year was only 68.45 percent.

That contrast deserves attention. Not because every high-performing school is doing something dishonest. Many are doing excellent work. Some have better teachers. Some have better facilities. Some have stronger academic cultures. Some simply have students who have been preparing seriously for years.

But there is another question worth asking. Who gets to enter these classrooms in the first place? What happens before the examination? Do schools simply teach whoever comes through the door, or does the system gradually sort students according to performance, ability to pay, previous results, coaching, and family resources?

I do not want to make an accusation without evidence. But I think the question deserves to be asked. Because if an institution is allowed to select the strongest students, train them intensively, surround them with private tutors and coaching, and then present their examination results as proof of the institution’s educational quality, we need to be careful about what exactly that number is measuring.

Is it measuring the school? The student? The family? The money spent before the examination? Or all of them together?

We Have Learned to Confuse the Result With the Learning

If a child gets GPA-5, we say: excellent student. If a school produces hundreds of GPA-5 students, we say: excellent school. But what exactly has been proven? The student has demonstrated that they can perform very well in a particular examination system. That is valuable. It is not the same thing as proving that they have been deeply educated.

Even Bangladesh’s recent education reform discussions have started pointing toward this gap. A government review report noted a divergence between examination outcomes and independently measured learning, warning that credentials can rise while actual mastery remains modest.

That distinction matters. Education should not simply produce students who can answer questions. It should produce people who can ask better questions. Perhaps that is the part we have forgotten.

The Child Who Has No Time to Be a Child

I sometimes look at the schedule of a serious student today and wonder how much life remains outside it. School in the morning. School ends. Private tutor. Then coaching. Then another class. Then homework. Then preparation for the next test.

The parents are running too. Taking the child to school. Bringing the child home. Taking the child to coaching. Waiting outside. Paying the fees. Finding another teacher when the marks fall. Searching for the next institution when admission season arrives. The whole family begins orbiting around the child’s education.

The child is told that all of this is necessary. Because the competition is high. Because everyone else is doing it. Because a good GPA leads to a good college. A good college leads to a good university. A good university leads to a good job. A good job leads to a good life.

That is the promise. I believed some version of it myself. Most of us did. But I have lived long enough to see that the last part of the promise is not guaranteed. A certificate can open a door. It cannot tell you what to do after you enter.

So What Are We Actually Measuring?

This is the question I want to carry through the rest of this essay.

Maybe the problem is not that examinations exist. We need ways to assess learning. Maybe the problem is not that parents care about their children’s future. Of course they should. Maybe the problem is not even competition itself. Competition can push people to improve.

The problem may be what happens when the measurement becomes more important than the thing being measured. When getting the grade becomes more important than understanding. When entering the right institution becomes more important than knowing why you are going there. When finishing the syllabus becomes more important than curiosity. When memorising the right answer becomes more valuable than discovering a new one. And when a child spends so many years preparing to prove that he has learned something that he eventually forgets why he wanted to learn anything at all.

That is where I become uncomfortable. Because I have started to think that perhaps we are asking the wrong question. We keep asking: how well did the student perform? Perhaps we should also ask: what kind of person is this education helping us produce?

And that leads to a much harder question. How well a student can perform in an examination, and how well a human being can live a life, are these really the same thing?

I don’t think they are. And if they are not, then we need to ask what happened to the space between the two.

That is where this conversation begins.

The Race and Why It Feels This Way

I sometimes wonder when the race actually begins. There is no starting gun. No one announces that the competition has started. A child simply goes to school one morning. Then another class is added. Then homework. Then an exam. Then a private tutor. Then coaching. Then another exam. Before long, the child is no longer simply going to school. The child is preparing for the next thing. And the next thing is always waiting.

School. Private tutor. Kumon. Coaching. Guidebook. Homework. Model test. Exam. Result. Next exam. At some point the schedule becomes so full that nobody remembers when the child was supposed to simply be a child.

There is nothing wrong with teaching a child, or with extra help. Some children need it. Some enjoy it. The problem begins when more learning always means better learning. It doesn’t. A child can spend hours studying and still never develop a love for learning. A child can solve hundreds of questions and still struggle with an unfamiliar one. A child can memorise an entire chapter and have no idea how that knowledge connects to the world outside the book. And a child can become so busy preparing for the future that there is very little present left.

When does a child play without a reason? Climb a tree? Sit outside and watch the rain? Read a book because they want to, not because someone assigned it? Adults often see these things as wasted time. I am not sure they are. UNICEF describes play as a fundamental way children explore their environment, build knowledge, develop problem-solving abilities and express themselves, and as a source of the independence that comes from making their own choices. A child playing with other children is learning negotiation. A child building something and watching it fall apart is learning failure. A child getting bored may eventually learn how to create something to do. None of this improves the GPA. But it is part of becoming a person.

From Curiosity to Performance

A small child does not usually begin learning because someone promised a GPA. Why is the moon following us? Why does rain fall? Why can birds fly? The questions come naturally. There is no examination waiting at the end. The child simply wants to know.

Then school begins, and slowly the question changes. It becomes: will this come in the exam?

At first the child may enjoy getting an answer right. Then comes the first star, the first certificate, the first class position. Then the first comparison. Who came first? How much did you get? None of this is necessarily harmful. Competition can motivate. Recognition can feel good. The problem begins when performance becomes the main way a child learns to understand his own value. Then something subtle happens. The question is no longer “did I do well?” It becomes “am I good enough?” Those are not the same question.

A healthy education system lets a child fail an exam and think: I didn’t understand this yet. A highly competitive one slowly turns that into: maybe I am not good enough. Hear “you got 95, why not 100, your friend got 98, you used to be better” often enough, and a child learns something nobody explicitly taught him. My result matters. My result matters to my parents. My result determines how they see me. Eventually: my result determines how I see myself. That does not mean grades are bad. It means what the child learns to chase matters.

Fear Is a Powerful Teacher

Perhaps the most powerful force here is not ambition. It is fear. Parents are afraid their child will fall behind. Children are afraid their friends will move ahead. Because the competition is real, the fear does not feel irrational. A study of shadow education in Dhaka found this pattern directly: parents described strong academic competition as a reason for investing in private tutoring, while students reported pressure from parents, peers and neighbouring students because they did not want to fall behind, with the extra workload reducing time for sports and social activities. The child may not even want another coaching class. But if everyone else is attending, refusing feels dangerous. So the child goes. Not from curiosity. From fear of falling behind.

Fear can make a person work very hard, which is why the system can look successful from the outside. The marks improve. The GPA rises. Everyone feels the pressure worked. But did the child become more curious, more independent, more capable of learning without supervision? Or simply more afraid of getting the next answer wrong? A 2023 systematic review of 52 studies on academic pressure and adolescent mental health found that 48 of them showed a positive association between academic pressure and at least one mental-health outcome, though the authors were careful to note the evidence was largely observational. We should not say exams cause mental illness. The evidence does not allow that simple a conclusion. But we can say that academic pressure is consistently associated with poorer mental-health outcomes across a substantial body of research. That should be enough to make us pause.

I do not think the parent is usually the villain here. It is easy to blame parents for pressuring, comparing, demanding higher marks. But they are inside the same system, and their own expectations are not the enemy: a large study of 9,654 adolescents found that higher parental and teacher expectations were associated with later academic performance and attainment. What matters is how those expectations are communicated. There is a difference between “I believe you can do this” and “if you don’t do this, you have failed me.” Between support and control.

What the Older Child Loses

Self-determination theory places autonomy, competence and relatedness at the centre of healthy motivation. In simple language, a child needs to feel: I can do something, I have some choice, I belong. That is very different from: I must get this mark, I must beat that person, I must enter that university. Education should gradually move a person from “someone tells me what to learn” to “I know how to learn” and eventually “I know what I want to learn.” That is development.

But our system often moves the opposite way. The older the child becomes, the higher the stakes: SSC, HSC, university admission, job examination. The child gains more knowledge but sometimes loses more freedom. Follow the syllabus, follow the teacher, follow the guidebook, follow the model answer, follow the admission strategy, follow the career path, and eventually, follow the job.

This is where research on achievement goals becomes relevant: performance-focused goals can undermine intrinsic motivation compared with mastery-focused goals. There is a difference between “I want to understand this” and “I want to prove I am better than others at this.” Both produce effort. They produce very different relationships with learning.

The habits learned in childhood do not disappear at graduation. The student who spent twenty years chasing marks may enter a workplace and start chasing appraisals. The one who learned to compare may compare salaries. The one who learned to seek approval may seek it from managers. We are not only deciding how children perform in school. We may be shaping how they behave for the rest of their lives.

A very young child asks a question because he genuinely wants to know. Nobody promised him a certificate. That impulse is precious, and education should protect it. Instead, somewhere along the way, the child stops asking “why?” and starts asking “will this come in the exam?” Learning becomes performance.

If a child learns to measure his worth by results, what happens when the results stop coming?

The Coaching Economy

A child’s education does not end at the school gate. In many families, that is where the spending really begins. School. Private tutor. Guidebook. Coaching. The exam. Admission coaching. University. Job preparation. Finally, the job. It looks like one continuous journey. Economically, it is a chain of markets.

Education Watch estimated that Bangladeshi families spent an average of Tk 13,882 a year on a primary-school child in 2022, and Tk 27,340 for a secondary-school child: not school fees, but private tutoring, coaching and commercial guidebooks. Families reportedly bear around 71 percent of the country’s overall education costs. By 2023, roughly 70 percent of primary students and 76.5 percent of secondary students had private tutors, and guidebook use had risen to around 92 and 93 percent respectively. A 2026 government review of secondary and higher-secondary education described private tutoring and coaching as a kind of “shadow assessment system” that families use to manage uncertainty around examination outcomes. That phrase stayed with me. The official system gives the exam. The unofficial one prepares the child for it. Families pay for both.

Coaching turns the same idea into a business model, one teacher, many students, one curriculum, one payment, then another batch. The Bangladesh Bureau of Statistics’ 2024 survey counted 6,587 registered coaching centres serving roughly 917,000 students, and unregistered centres are reportedly far more numerous still. Estimates of the total offline coaching market vary widely, from roughly Tk 25,000 crore to Tk 50,000 crore, there is no single audited figure, and I would not pretend there is. But the exact number may not be the important part. The important part is that the market is enormous, and it exists because families believe the formal system alone is not enough.

After SSC and HSC comes admission coaching, one 2025 investigation estimated Bangladesh’s university admission coaching market at around Tk 1,400 crore, with individual programmes commonly running Tk 15,000 to Tk 20,000, medical preparation higher still. One student interviewed for that report admitted coaching did not feel essential, but once everyone else joined, preparing alone felt risky. That sentence explains a lot. The market does not only sell education. Sometimes it sells reassurance.

And it does not end at university. Graduates prepare for BCS, bank jobs, government recruitment, teacher registration: another set of books, guides, question banks, coaching, mock interviews. One estimate put the annual market for competitive job-preparation books alone at Tk 1,500–2,000 crore, with BCS candidates commonly spending Tk 12,000–15,000 per person on books from preliminary through viva. The child who started with a schoolbook is now an adult buying a job-solution book. The names have changed. The pattern has not.

Nobody has to sit in a room and design this together. A coaching centre wants students. A parent wants security. A teacher wants income. Each person follows their own interest, and something larger emerges: the coaching centre says everyone is preparing, the parent thinks then my child should too, the child sees everyone else going and goes. The market grows on its own momentum. Imagine if only ten percent of students believed they needed private tutoring. The market would stay small. Once ninety percent believe it, not hiring a tutor starts to feel like the risk. Nobody is asking “do I actually need this?” anymore. They are asking “can I afford not to?” That is a very different question, and it is the one that keeps this economy growing even when the child is not necessarily becoming more educated. The market does not need education to end. It needs the next exam.

Nobody designs this together. It reinforces itself.

A System Teaching Itself

Shadow education has a precise meaning: fee-based instruction that exists outside the formal school system but follows its curriculum and prepares students for the same exams. What makes it a shadow rather than a separate thing is that it does not create its own curriculum. It follows yours. Without the official syllabus there would be no tutoring curriculum; without the examination, no exam coaching; without admission tests, no admission coaching. The shadow follows the shape of the main system exactly.

There is nothing wrong with a tutor. A child can genuinely struggle with a subject, and research on Bangladesh confirms private tutoring can help close specific learning gaps. The harder question is why extra help has become so widespread it no longer feels extra. If a child spends five or six hours at school and then needs another teacher to understand what was taught there, something is happening inside that classroom: too little time, too large a class, too heavy a curriculum, an examination system that rewards preparation ordinary teaching cannot provide. Probably some combination of all of it. But once the second system becomes almost compulsory, the tutor stops filling a gap and becomes part of the structure itself.

The incentives compound each other. A teacher earning less from classroom teaching than from private lessons to the same students has a tension no amount of good intention resolves. A parent is often not buying learning so much as buying insurance against what happens if everyone else’s child has a tutor and theirs doesn’t. And a child watching everyone else go to coaching starts asking not whether he needs it, but what happens if the others get ahead. None of this requires anyone to be evil. Millions of individually reasonable, defensive decisions add up to a system nobody chose on purpose.

There is a cost here beyond taka. A child has only so many hours, and the ones spent moving between tutor and coaching and guidebook homework come from somewhere, play, sleep, family conversation, sitting quietly, discovering an interest. We call the money an education expense. We rarely call the lost time one. And there is a further cost still: one Bangladeshi study found heavy reliance on tutoring can encourage rote learning, weaken critical thinking, and leave students dependent on being taught rather than able to learn on their own. A tutor can help a child understand something. But if the child comes to believe he cannot study without someone teaching him, the system has solved one problem while creating another, because the purpose of education should eventually be to make the teacher less necessary, not permanently necessary.

There is also inequality quietly built into all of this. One child gets school, private tutor, English coaching, admission coaching, a laptop, courses. Another gets the schoolbook. Both sit the same exam. We call the result merit. Sometimes it is. But sometimes it also reflects how much preparation a family could afford to buy. And education is supposed to reduce the importance of family background, not multiply it.

If a child needs another education system to succeed in the first education system, which system is actually doing the teaching? I don’t have a simple answer. Perhaps the school is failing, perhaps the exam system is, perhaps parents have become too anxious, perhaps all of it at once. But a child should not need to spend most of his waking hours being taught how to succeed at being taught. At some point education should step back, the teacher, the tutor, the parent, and the child should be able to stand alone with a question and say: I want to know.

If a child needs another education system to succeed in the first education system, which system is actually doing the teaching?

That, perhaps, is where education actually begins.

The Window

After everything above, I think we have reached the question underneath the other questions. Not how much education costs, or how many exams a child takes, or whether a degree leads to a job. Those matter, but they are secondary. The first question is simpler: what is education actually for?

Education can mean knowledge, skill, character, curiosity, judgement, independence, the ability to work with others, employability. It can mean learning how to be a human being. A job is important, we need to earn, and a society needs people with useful skills, so there is nothing wrong with education helping someone become employable. But UNESCO’s Delors Report named four pillars of learning back in 1996: learning to know, learning to do, learning to live together, and learning to be. Learning to do is the one closest to employability, and even that pillar is broader than job preparation, it means the competence to handle a situation, work with others, turn an idea into something real. Learning to live together means understanding people who disagree with you, cooperating, managing conflict, what good is a brilliant result if a person cannot listen? And learning to be means developing autonomy, judgement and personal responsibility. A student can follow every instruction perfectly and still have very little autonomy. That, more than any of the others, is the pillar I think we have let go.

We measure what is easiest to measure, marks, GPA, rank, certificates, because these fit in a spreadsheet. Curiosity, judgement, courage, the ability to understand another person: these are much harder to count, and perhaps that is one reason they slowly disappear from the centre of education, not because they matter less but because they are harder to measure. Even UNESCO has acknowledged the problem, arguing more recently for a broader understanding of learning that includes social, emotional and ethical dimensions, while admitting how difficult these are to assess. What we can measure is not necessarily everything that matters.

The Cupboard and the Window

I keep returning to an image. A small child stands inside a room. Outside is an enormous world: forests, cities, languages, ideas, machines, art, failure, everything. Education gives the child a window. She looks through it and sees something she did not know existed. At first she needs someone to point: a teacher says look at this, a book introduces an idea, a story lets her enter someone else’s world. The window grows larger. Then, if it works, something beautiful happens. She starts choosing where to look herself. The window becomes a door. The teacher is no longer the only source of knowledge. She walks outside, gets lost, makes mistakes, discovers something.

But look at what we have built instead. A cupboard, full of textbooks, guidebooks, question banks, model tests, past papers: everything the child might need to pass the next examination. The child reads, memorises, underlines, marks possible questions, repeats the answers, and eventually knows the contents of that cupboard extremely well. The world outside the window is still waiting. Only nobody has given him the answers to that one.

A syllabus has to have boundaries, no school can teach everything, and I understand that a curriculum must select something. The problem begins when a child learns that only what is inside the curriculum is worth learning, and starts filtering the world through a single question: will this be in the exam? A very young child asks why is the sky blue. Later she may ask what causes the sky to look blue. That is still curiosity. Will this come in the exam is a different question entirely, it comes from strategy, not wonder. Strategy is not the enemy. But when it gradually replaces curiosity, something has been lost. The child stops asking what do I want to understand and starts asking what do I need to remember.

Real learning has no fixed endpoint: question, explore, fail, reflect, create, understand, and then another question. Exam learning has a clean one: read, memorise, reproduce, score, certificate, next chapter. The second process is efficient. Efficiency is not the same thing as education.

I want to be careful here, because memorisation is not the enemy. A doctor needs to remember things. A musician needs scales. Nobody thinks without some knowledge to think with. The problem is when memorisation becomes the destination rather than the foundation. Give two students who both reproduced the textbook perfectly a problem that has no model answer, and the difference shows immediately. One asks which chapter is this from. The other asks what is actually happening here. That difference rarely appears on the original exam. It matters enormously in life, because life does not hand you four options and an answer key. It asks what should we do, not what is the correct answer, and that requires judgement, not memory.

Failure means something different in each system. In an examination, failure is a negative outcome, lost marks, a lower rank. In real learning, failure is information: you build something, it breaks, you ask why, you try again differently. Our system has little patience for the second kind. A child who does poorly hears “why did you get less than him,” “what will happen to your future,” and the message lands clearly. Failure is not part of learning. Failure is a threat. So the child learns to avoid it, stays inside the answer she already knows, stops experimenting unless she’s sure. And creativity, almost by definition, needs the freedom to be wrong. A system built around standardised measurement naturally prefers the predictable answer to the interesting one, and a child absorbs that preference as a lesson about the world: predictability is safer.

Perhaps the most invisible lesson is this: that knowledge is something other people possess. The teacher knows, the textbook knows, the examiner knows, and the child’s job is to receive and reproduce it correctly. Real education should eventually reverse that relationship, until the child realises: I can find things out too. I can ask, investigate, disagree, be wrong, change my mind, learn without someone standing over me. That is intellectual independence, and it may be the most valuable thing education can give.

An A+ tells us a student performed extremely well on a particular assessment. It cannot tell us whether that person can think independently, deal with uncertainty, work without instructions, or recover from failure. Those questions are harder. They are also closer to the heart of what education is supposed to do.

I don’t want to blame the book. A book lets knowledge travel across generations; a child can sit in a small room and enter another century, and that is beautiful. The problem comes when the book becomes a wall instead of a window, when the syllabus becomes the boundary of curiosity, when the examination becomes the purpose of learning. Real learning says let’s find out. Exam learning says let’s remember. Real learning allows failure. Exam learning penalises it.

The problem comes when the book becomes a wall instead of a window, when the syllabus becomes the boundary of curiosity, when the examination becomes the purpose of learning.

What happens when we give a child enough books to fill a room, but not enough freedom to open the window?

How We Got Here

Until now I have been talking about the education system as if it simply appeared around us. It didn’t. The system has a history, and once I started looking into it, I found the popular version too simple. The Industrial Revolution did not invent the school, and factory owners did not sit down and design classrooms to produce obedient workers. Research on European education history shows countries like Prussia, France, Sweden and the Netherlands were already expanding public schooling for religious, political and nation-building reasons before industrialisation fully transformed their economies. Some societies, like China with its civil-service examination tradition, had sophisticated exam systems centuries earlier still.

What changed was scale. A small community can run on informal systems: a teacher who knows every child personally, a village that manages its own affairs. A large industrial, bureaucratic society cannot. Once societies needed millions of people who could read, write, calculate and operate inside increasingly complex institutions, education took on a much larger social role, and it began to systematise: standardised curricula, trained and certified teachers, regular attendance, defined age groups.

Once schooling became mass education, a new kind of question appeared: how do you measure millions of students against some common standard? That is where the examination enters the story in its modern form. In 1845, Horace Mann helped introduce large-scale written examinations into Boston’s public schools, largely because thousands of children could no longer be evaluated individually by committees using informal judgement. Written exams could be given to many students under the same conditions and compared. That is reasonable. A growing system needs some way to measure itself. But once learning could be measured, it could be ranked. Once ranked, compared. Once compared, sorted. A number is easier to compare than a person; a certificate easier to file than a story. Education began moving closer to bureaucracy, because bureaucracy is what large systems need to function. And standardisation, whatever else it makes possible, also reduces difference.

Once examinations were widely used, another possibility appeared: proving something about yourself to someone who does not know you personally. I passed this exam. I hold this degree. The credential became a portable signal, a shortcut institutions could use instead of asking what can this person actually do. That shift, asking what credential does this person have before asking what can they do, is the hinge the rest of this essay turns on, and once credentials mattered, people began competing for them.

The credential became a portable signal, a shortcut institutions could use instead of asking what can this person actually do.

This region entered the same history by a different door. Colonial rule brought modern schooling, examinations and institutional structures into Bengal; the Wood’s Despatch of 1854 became a foundation for public education in British India, and later administration grew more centralised through instruments like the Bengal Primary Education Act of 1930. So the system Bangladesh inherited is not home-grown from a single root. It is layers, colonial education, postcolonial nation-building, mass schooling, examination culture, university expansion, private coaching, credentials, labour-market competition, all still sitting on top of one another.

None of this happened because someone secretly designed the whole machine. Different institutions developed for reasonable reasons, then became connected, and the connections and incentives strengthened until the system began to reproduce itself. That is how systems become hard to question: nobody has to be running it; everyone inside it just has a reason to keep it running. The education system we inherited was built to solve the problems of mass society, bureaucracy and measurement, and it succeeded at much of that: it expanded literacy, expanded access, created professional knowledge. But a system can be successful at one historical task and inadequate for another. We placed our children inside a machine built for standardisation and asked it to also help them discover who they are. I am not sure it was built for that.

Once education became measurable, it became comparable. Once comparable, competitive. Once competitive, markets grew around it. And once it became tied to the labour market, one more idea became powerful: that the value of education could be measured by the economic value of the person it produces.

Capital, Credential, Employee

Capital. We use the word casually now, human capital, human capital development, human capital investment, and the phrase sounds almost natural. But stop for a moment. A child is born, we send him to school, he spends years learning, and we say we have invested in his human capital. The logic is simple: spend today, earn more tomorrow. There is real truth in it. Theodore Schultz helped establish this as serious economic thought in 1961, arguing that spending on education should be understood as investment in productive capacity. Gary Becker developed it systematically in his 1964 book, analysing education and training as investments and estimating their returns. Jacob Mincer connected schooling, experience and earnings empirically in 1974. Together this gave economists a powerful framework, and the OECD still describes education this way, noting that higher education is strongly associated with better employment and earnings.

None of that is wrong. The relationship is real. But something changes when we look at a five-year-old this way. She is curious, she plays, she gets bored, she discovers something, nobody calls her an investment. Then school begins, and the vocabulary shifts: education investment, skill development, future earning potential, human capital. The child has not changed. The language describing her has. And language shapes what a society learns to value. A child fascinated by plants, or history, for no productive reason at all, is that curiosity worth less because it may never show up in a salary? Economic analysis is very good at answering what education produces economically. It is much less capable of answering what education produces in a human being. What is difficult to measure can slowly become what a society forgets to value. Even the OECD, in the middle of discussing financial returns, explicitly notes that education produces broader outcomes beyond earnings, productivity, health, well-being, so even within economics, education resists being reduced entirely to salary.

There is a difference, easy to miss, between saying “we need educated people because the economy needs skilled workers” and saying “the purpose of education is to produce skilled workers for the economy.” The first gives the economy a legitimate place in education. The second gives it final authority. That distinction matters because labour markets change, a skill goes obsolete, a technology replaces a task, and education built primarily around today’s labour market can become outdated quickly. Education that teaches someone how to learn lets the person adapt regardless. Perhaps the most valuable form of human capital is not a particular skill but the capacity to keep learning.

Signal Versus Capability

Here is where a second, related idea enters: not everything a degree does is teach. Imagine an employer with five hundred CVs and no way to know any of these people personally. She needs information, so she looks for something visible, a degree, a university name, a GPA, and uses it as a signal. This is the problem Michael Spence examined in his 1973 paper “Job Market Signaling”: employers make hiring decisions under incomplete information, and observable credentials fill the gap. A degree can therefore be both things at once: genuine knowledge and skill, developed through years of study, and simultaneously a signal that lets a stranger take a chance on you. The trouble starts when the signal becomes more important than the capability it was supposed to indicate.

And once a credential works as a signal, everyone rationally wants one. More people get the degree. The degree becomes common. The employer needs a new way to sort five hundred CVs, so requirements rise: a master’s, a certification, more experience. In 1979 the sociologist Randall Collins described exactly this cycle in *The Credential Society*: expanding educational attainment pushes job requirements upward even when the underlying work has not become more demanding. A job that once needed a high-school education starts asking for a bachelor’s, not because the work changed but because the applicant pool did. The screening mechanism inflates. The work stays roughly the same.

I don’t want to overstate this into “degrees are useless.” A medical degree is not a piece of paper. An engineering qualification involves real technical knowledge. The honest answer to whether a degree is capability or signal is that it can be both, and the OECD’s own 2026 work on skills-first labour markets makes the same point, arguing that qualifications remain important but cannot capture the full range of what a person can do, and pointing toward hiring approaches that ask what can you actually do rather than only what degree do you hold. That the labour market itself is starting to say this is a significant admission. The problem is not signalling, markets need information. The problem is what happens when people spend years acquiring credentials primarily to earn the right to demonstrate skills they may already have.

This is where the theory becomes painfully familiar in Bangladesh. A young person finishes school, then college, then university, then perhaps a master’s, then professional courses. And the job advertisement still says experience required. So he takes another course. The CV lengthens. The person may still be unemployed. The IMF’s 2026 Bangladesh assessment found that unemployment among highly educated youth reached roughly 25 percent in 2024, and that rising education levels have not translated into movement toward higher-productivity sectors. We keep producing more educated people. Education alone is not producing enough good opportunities to absorb them. That is not a uniquely Bangladeshi phenomenon, but it is an especially sharp version of it.

The Employee Factory

Now put the pipeline together and look at the whole shape of it. School produces a result. University produces a degree. Coaching optimises exam performance. Job preparation produces employment. The employer asks for productivity. Every stage prepares the person for the next one, and the whole thing is coherent, almost too coherent. Society genuinely needs employees, doctors, teachers, engineers, administrators. I am not arguing that becoming an employee is a failure. The question is different: is becoming an employee the final purpose of education?

Because look at what the pipeline rewards. The student who gives the expected answer, follows the syllabus, performs well on the exam, enters the prestigious university, rewarded at every stage. Where, in that sequence, does someone learn to say “I disagree,” or “there might be another way”? Where is the researcher, who has to sit with “I don’t know” long enough to find out? Where is the entrepreneur, who gets no model answer and no teacher confirming she’s right, only a market that says yes or no? Where is the artist, for whom there is no marking scheme? None of this is an argument that every child should start a business, a society needs employees and researchers and teachers far more than it needs everyone chasing a startup. It is an argument that a healthy education system should give a child the capacity to create, to question, to solve an unfamiliar problem, regardless of which of those paths they eventually choose. The deeper thing missing from the pipeline is not entrepreneurship specifically. It is agency: the felt sense that you can decide, initiate, and take responsibility for what happens next.

The employee is not the problem. An employee can think independently, challenge management, build new things, live a deeply meaningful life. The problem is when the entire system optimises for employability while neglecting everything that makes employment, or any life, meaningful in the first place: the capacity to think, to judge, to create, to cooperate, to take responsibility. Those qualities make a good employee. They also make a good entrepreneur, researcher, artist, parent, citizen.

Children enter school. They are taught a standard curriculum, sit standardised exams, receive scores, are ranked, move to the next level, collect credentials, enter the labour market, and the successful ones are hired. The metaphor is uncomfortable because children are not products and human beings are not units of production. But the structure can feel remarkably similar to one. If the final product is an employee, the system is doing reasonably well. If the final product is supposed to be an educated human being, the evaluation gets much harder: can this person think, learn independently, live with uncertainty, question, understand other people, decide what a good life means? We measure the first set of outcomes obsessively. We barely measure the second. Perhaps that is exactly why the second keeps disappearing from view.

If the system produces someone who can perform a job but has no idea how to think, create, choose or live beyond that job: have we produced something useful to the labour market? Or have we produced an educated human being? What happens when the labour market itself no longer needs all the people the education system was built to produce?

Bangladesh as the Proof

All of this can sound abstract discussed globally. Bangladesh makes it visible, because the entire chain sits almost in front of us: school, private tutor, guidebook, coaching, SSC, HSC, admission coaching, university, degree, job preparation, employment. This is not a theoretical model. It is a familiar Bangladeshi life, and the numbers show how deep the dependence runs, Education Watch’s 2023 data put more than three-quarters of primary and secondary students on private tutors or coaching, with 92 and 93 percent of primary and secondary students respectively depending on commercial guidebooks. Of an average family’s roughly Tk 13,882 annual education spending on a primary-school child, a large share went not to the school itself but to tutors, coaching and guidebooks, the shadow system around the school, not the school.

Then comes the sorting. Around 1.86 million students registered for SSC and equivalent exams in 2026. That is not simply an examination. It is a national sorting mechanism, and it does not end the competition, it changes its form. College becomes HSC. HSC becomes university admission, and here the competition turns sharply positional. It is no longer enough to say I want to study; the question becomes where, Dhaka University, BUET, a medical college, IBA, because the institution’s name has become part of the signal, and with limited seats and enormous demand, another market appears around it. A 2025 investigation estimated Bangladesh’s admission-coaching industry at around Tk 14 billion a year, individual programmes commonly running Tk 15,000–20,000. One student told the reporters that preparing alone felt risky once everyone else had joined coaching, which is close to a summary of the whole system. Competition itself creates the demand.

Then the degree meets the labour market, and this is where the story turns genuinely uncomfortable. The World Bank reported that more than 650,000 higher-education graduates enter Bangladesh’s labour market every year, and that less than half find employment within one or two years, linking the gap partly to a mismatch between what institutions teach and what employers need. The headline unemployment number hides the real shape of the problem: Bangladesh’s overall unemployment rate in 2024 was a modest 3.69 percent, but youth unemployment stood at 8.49 percent, unemployment among highly educated people at 8.7 percent, and among highly educated youth at roughly 25 percent, worse still for highly educated women, at close to 17 percent against 6.4 percent for highly educated men. The deeper problem is not that Bangladesh has unemployment. It is that Bangladesh is producing educated people faster than its economy is creating suitable opportunities for them.

There is a second problem beneath unemployment: mismatch. The IMF’s 2026 analysis found that between 2015 and 2024, average education levels rose across sectors while employment moved the wrong way: agricultural employment actually grew as a share of the workforce, especially among young workers, while manufacturing and services shrank. People are becoming more educated. The economy is not necessarily moving them into more productive work to match. Education → skills → productivity → income is the promised chain. In Bangladesh it is not running smoothly, because modern, skill-intensive industries remain underdeveloped while agriculture continues absorbing a large share of employment despite low productivity. You can produce more graduates. If the economy does not produce more productive jobs to match, those graduates compete for the same limited positions anyway.

And when the job finally arrives, it may not pay well. A 2025 World Bank poverty and equity assessment found that around half of young people aged 15–29 work in low-paying jobs, suggesting a significant skills mismatch, with one in four educated young women left without a job at all. This is where the promise starts breaking apart for a lot of young Bangladeshis. They did what they were told, studied, got the degree, learned English, built the CV, competed, and the reward does not always match the investment. The competition did not end when the job arrived. It simply changed its name: who got the highest mark becomes who got GPA-5 becomes who got into a good university becomes who got the better job becomes who got promoted, who earns more. The person who thought education would eventually produce freedom discovers it may only have prepared them for the next competition.

None of this means the system is broken beyond use, that would be too easy a conclusion. It produces doctors, engineers, teachers, researchers, entrepreneurs, and millions of people have genuinely improved their lives through it. But producing real gains is not the same as producing the best possible outcome, and the gap that matters most runs all the way from education to learning, credential to capability, employment to a decent life, and finally, the one I care about most: from becoming employable to becoming free.

Perhaps we should stop asking only how many children are passing SSC, or how many are entering university, and start asking how many can think independently, solve an unfamiliar problem, find work that uses what they learned, earn enough to live with dignity. And perhaps the hardest question of all: after spending most of their childhood and youth preparing for the future, how many young Bangladeshis actually feel prepared to live the life they wanted?

What happens when an education system becomes very good at producing employable people, but the economy is not creating enough good employment for them? Perhaps that explains something visible all around us: why so many educated young people are exhausted before their working lives have even properly begun.

University, Market, Mismatch

After the admission battle, one question becomes surprisingly important: not what will I learn, but where will I study. Somewhere along the way, the university itself became a credential, and the name on it became a signal. Two students studying the same subject, one from a widely recognised institution and one from a less recognised one, may have similar knowledge and similar practical skill, but an employer who knows neither of them personally will read the name as information regardless. Dhaka University illustrates the point well. It carries enormous national prestige, having produced generations of politicians, academics, writers and public intellectuals, and it remains the highest-ranked university in Bangladesh, holding a place in the world’s top 600 in the QS World University Rankings for the third year running. But a ranking is not a universal measurement of educational quality, it is a score built from a particular set of indicators, weighted a particular way. It cannot tell us whether a student discovers a passion, reads beyond the syllabus, or learns to think independently in the four years they spend there. A university can be excellent for one student and a poor fit for another, regardless of its number.

Yet Bangladeshi students often talk about universities and subjects the way they talk about GPA, as a hierarchy. This university is better, that subject has more value, this degree gets better jobs. And once a hierarchy becomes socially accepted, competition around it intensifies, admission coaching grows, and the university starts to function less like a place of learning and more like a status signal. The same thing happens one level down, with subjects. Medicine and engineering carry a visible, well-understood professional identity, you study it, you become that thing. Computer science has surged because the economy visibly created new roles for it. Business is attractive because it is flexible. Meanwhile history, philosophy, literature, and pure science get asked the question “what job will you get with that,” repeated often enough that it starts to reshape what a whole generation considers worth knowing. That is a real cost. Markets are good at rewarding what they currently need. They are not good at predicting what a society will need in twenty years, the researcher on an obscure problem, the writer who shapes how a generation thinks, may never have a job advertisement waiting for them, and their value may only become visible decades later. I don’t think the answer is to ignore the labour market; a student needs to understand it. But there should be a second question sitting alongside “what pays,” what do I want to become good at, because if you dislike the subject, market demand alone will not carry you through a career, and if you genuinely love something, capability itself can create opportunities no subject ranking predicted.

After the Job Arrives

Suppose the degree is finished, the CV is ready, and the phone finally rings: you are selected. For a moment, all those mornings and tutors and coaching centres feel worth it. Then a quieter question arrives. Now what?

The salary comes, and then rent, transport, food, family expenses, sometimes a loan, come with it, and a number that looked reasonable on paper starts to feel different in practice, not simply a personal budgeting failure, but a reflection of an economy where labour-income growth has weakened while living costs have risen. Someone can have a real job, a payslip, an office ID, and still struggle to build a life, because unemployment is not the only way work can fail a person. There is underemployment, low wages, insecure conditions, Bangladesh’s informal sector alone accounts for a large share of employment and is marked by low productivity and low pay. The more revealing question is not do you have a job, but what kind of life does that job allow you to build.

Career growth, too, is not automatic. Positions above may already be filled, salary increases may be small, promotion may depend on relationships rather than transparent performance, and recent reporting in Bangladesh has surfaced worker complaints about favouritism, long hours without overtime pay, and limited transparency, alongside genuinely well-run organisations that do none of this. Once employees believe effort alone is not enough, the question they ask silently changes from “how can I get better” to “who do I need to know.” And the competition that started with who got the highest mark in Class Three does not really end with the job. It changes clothing, who got promoted, who became manager first, who earns more, while quietly consuming the one resource that cannot be recovered: time. Long hours without adequate compensation remain a persistent concern in Bangladeshi workplaces, and the cost shows up everywhere outside the office, family, health, the ability to simply think.

There is a distinction worth holding onto here: a job is an economic arrangement, a livelihood is a way of living, and the two overlap without being identical. A person can earn enough to pay bills and still have no savings, no growth, no time, no ability to leave a bad situation, technically employed, not necessarily secure. And Bangladesh’s very large labour force means an employer often has more bargaining power than the worker, because someone else is waiting for the same position. The child was told: get educated so you have choices. What the labour market can actually deliver is sometimes closer to: get a job so you can survive. Those are not the same promise, and choice requires alternatives that a tight labour market does not always provide. Perhaps this is why building capability that does not depend entirely on one employer, the ability to solve problems, sell, communicate, manage money, learn independently, adapt. Matters as much as any single credential. Not because everyone should quit and start a business, but because that capacity is a form of independence in itself.

The Queue

Put credential inflation next to job scarcity and the picture sharpens. When five hundred people compete for one position, credentials become filters rather than proof, degree, university, GPA, experience, certification, each one narrowing the list before anyone asks what the person can actually do. More graduates leads to more competition leads to more credentials leads to higher requirements leads to more education leads to more graduates. The system feeds itself. And even when someone wins that race, the job at the end of it may still pay poorly, still sit below what their education prepared them for, the World Bank has flagged skills mismatch as a defining feature of the youth labour market, with roughly half of young workers in low-paying jobs. The person is educated. The person is employed. That is not the same as being well paid, or using their education, or building a secure future.

Unemployment is visible, you don’t have a job, and everyone can see the problem. Underemployment is quieter. You have a job, a salary, a LinkedIn profile that says employed, while your education, skills and creativity sit mostly unused. Society assumes this is fine, because on paper it looks fine. This is the more subtle waste, and the country pays for it too, years of family spending, years of study, years of teaching, all producing a graduate whose capability the economy is not actually using.

None of this means education is useless, or that universities should disappear, or that a degree has no value. It means education and economic development have to grow together, or a country ends up with a generation highly prepared for opportunities that do not exist in sufficient numbers. Somewhere in that gap sits a young graduate asking a question the school never prepared them to answer: if I did everything the system asked of me, why is the life I was promised still out of reach?

What if the problem was never that a particular person failed to win the race, but that the race itself was never designed to give everyone a good life?

The Family Divide

We often talk about “the child” as if every child starts from the same place: same school, same textbook, same examination. They don’t. Imagine three ten-year-olds sitting in the same classroom. After the school bell rings, their lives separate completely.

The first child goes home to a full schedule: private tutor, coaching, English class, homework supervision, model tests, parents who check every result and compare it with the neighbour’s. This is real love, and real fear: parents who know the competition is real and believe that if they don’t do everything possible, their child will fall behind. But at some point involvement can tip into over-management, the child’s entire day scheduled from morning to sleep, and research on shadow education in Dhaka found exactly this pattern: parental investment in tutoring associated with high expectations, competition, financial pressure, and students losing time for sports and social life. The parent is trying to give the child more opportunity. The child may be receiving less childhood.

The second child goes home to parents who are busy surviving. They may love the child just as deeply, but one or both parents work long hours, there may be several children, and money that could go to a tutor goes to rent and food instead. This is not a failure of caring. It is a resource constraint: of money, but also of time, and of information a parent who never attended university simply does not have about how admissions or careers work. Bangladeshi research consistently finds a widening wealth-related learning gap across the primary years, and parental education and household economic status are significant predictors of children’s basic competencies. The starting line is not equal.

The third child goes home to a family shop, an office, a farm, a professional practice. This child watches adults negotiate, price, sell, fail, recover. That is not tutoring. It is exposure: a curriculum nobody wrote down, running alongside the official one. A child who has watched a parent negotiate with suppliers has learned negotiation. A child who has seen a family business survive a bad year has learned risk, and a child who has seen one fail has learned something more valuable still: that failure has consequences, and that people recover from it. This child has seen someone build something, and does not experience employment as the only imaginable future.

Capital Beyond Money

The sociologist Pierre Bourdieu’s framework is useful here: alongside economic capital there is cultural capital (knowledge, language, familiarity with institutions) and social capital (relationships, networks, people who can provide information or access). These reinforce each other, money buys tutoring, education provides knowledge, networks provide information, information leads to opportunity, opportunity creates more income, and the cycle continues into the next generation. A parent doesn’t need to hand a child money to give them an advantage. “I know someone in this field,” “come with me to the office,” “try building this yourself,” these are cultural and social capital delivered in ordinary sentences. A Bangladeshi study of 553 students, parents and teachers across secondary schools found family social capital had a stronger effect on students’ educational aspirations than school social capital did. The family does not only fund education. It shapes what a child believes is possible: you cannot easily aspire to a life you have never seen anyone live.

None of this means family background is destiny. A poor child can become a brilliant scientist; a wealthy child can waste every opportunity; a highly educated family can create its own kind of pressure. The honest claim is narrower and still uncomfortable: family resources, money, time, education, networks, exposure, shape the opportunities available to a child, and those differences show up in outcomes we then call merit. A GPA is visible. The private tutor behind it is not. A parent’s education does not appear on the certificate. A family network is not printed beside the degree.

Return to the window from earlier. Not every child gets the same window. One has a large window. Another has a small one, or one covered by curtains. Another grows up in a house where someone has already opened the door and shown them what’s on the other side. That is the family divide, and it explains why simply telling children to compete harder does not make the system fairer, because before the race begins, some children have already been given better maps of the world.

How much of what we call merit is individual achievement, and how much is inherited advantage: money, time, knowledge and networks that were purchased or transferred before the exam ever began?

Alternatives and the Question of Agency

We have spent a long time examining the formal system, and there is an easy mistake waiting at the end of that road: concluding that school itself is the problem. It isn’t. The problem is not that formal education exists. The problem is that it has become so dominant we have stopped recognising the other kind of learning that has always run alongside it.

A child in a family shop watches a parent negotiate with a customer, calculate a cost, absorb a complaint, solve it. Nobody hands her a textbook. Nobody says today we are learning negotiation. But she is learning how people behave, how money moves, how mistakes get handled. An apprentice with an experienced mechanic learns by holding the tool wrong, being corrected, trying again, learning through participation, a model that predates universities by thousands of years, and the ILO still defines quality apprenticeship as exactly this: on-the-job and off-the-job learning combined into structured competence, not a replacement for education but a different shape of it. Bangladesh’s informal economy, the World Bank puts it at around 85 percent of employment and 43 percent of GDP, already runs enormous amounts of this kind of learning, tailor to apprentice, mechanic to trainee, largely without a certificate at the end of it, but not without capability.

What real work provides that an exam cannot is immediate, unambiguous feedback. A wrong exam answer costs marks. A wrong price costs money and teaches something in the same instant. Action, result, reflection, adjustment, better action, that loop is one of the more powerful ways humans learn, and it is largely missing from a system where the feedback usually arrives as a number weeks later. Mentorship works the same way, compressing years of trial and error into transferred judgement rather than transferred information, which is exactly why judgement is so hard to teach from a textbook. And self-learning has become more possible than ever, a motivated person can now learn programming, design, accounting, a language, largely on their own, though this has real limits; medicine, engineering and rigorous research still need formal structure, and unsupervised beginners can absorb bad information as easily as good.

None of this argues for tearing down the school. Formal education solves a problem informal learning cannot: scale. You cannot apprentice a whole generation to individual experts, or assume every family runs a business that teaches negotiation. School gives a society a shared foundation, reading, mathematics, scientific reasoning, access to teachers and libraries that not every family could provide alone. The actual danger is not school. It is treating school as the only legitimate form of learning, believing that if it isn’t taught in a classroom it isn’t real, that if it doesn’t produce a certificate it has no value. The real alternative was never school-or-no-school. It is school plus life: theory plus practice, teacher plus mentor, credential plus capability.

Agency, Not Just Entrepreneurship

We have used the word entrepreneur several times in this essay, and I want to be precise about what I mean by it, because the word is usually heard too narrowly, as business owner, risk-taker, someone who starts a company. That is one form of it. The deeper idea, and the one I actually care about, is agency: the capacity to look at a situation and think I can do something about this, rather than waiting for someone else to. The OECD describes entrepreneurial people in almost exactly these terms, initiative, learning through trial and error, relying on their own judgement: and explicitly includes people who exercise this inside existing jobs, not only people who start new ones.

The deeper idea, and the one I actually care about, is agency: the capacity to look at a situation and think I can do something about this, rather than waiting for someone else to.

Give two students the same broken water pump. One asks, “sir, what should I do?” The other asks why it’s broken, who might know how to fix it, whether they could try it themselves. Neither is necessarily smarter. One has started exercising agency, moving from someone should solve this to perhaps I can help. The exam habitually trains the opposite reflex: find the correct answer the examiner wants, write it, receive marks, move on. Life mostly arrives with no question at all, just a problem, no four options, no marking scheme, you have to decide what the problem even is before you can decide what to do about it.

I want to be careful here, echoing an important caution from the World Bank’s own review of over 230 entrepreneurship programme evaluations: this kind of education can build entrepreneurial mindsets and capabilities, but it is not a silver bullet for unemployment, and outcomes depend heavily on context. Teaching business-plan templates to every child will not, by itself, solve anything, because agency is not a chapter you can examine. It is a practice, built through actually choosing, building, failing, revising, defending an idea and changing your mind, not through memorising the vocabulary of starting a company.

And agency does not require leaving employment to exist. A factory worker who finds a better way to cut waste, a teacher who invents a new way to explain something, a nurse who redesigns a broken workflow: none of them own a business, all of them are exercising exactly the capacity I’m describing. The opposite of this person is someone highly educated, with excellent grades and multiple degrees, who nonetheless waits: someone else will handle it, that’s how we’ve always done it, someone else is probably already doing it. That person may be extremely employable and have very little agency, and the two things are not the same at all.

Agency also means the ability to say no, to ask is this the right career for me, do I actually want this degree, is this salary worth the hours I’m giving away, is there another way to live. A child trained for years to follow the next assigned step can grow into an adult afraid to step off the path even when the path stops making sense. This is why entrepreneurship belongs in the education conversation at all, not because every child should found a company, but because every child deserves the chance to develop the capacity to act rather than only react, whether they end up an employee, a researcher, an artist, or eventually someone who starts something of their own.

Everyone does not need to become an entrepreneur. But everyone deserves the opportunity to develop agency, the difference between being prepared for the next test and being prepared to participate in a world that rarely comes with four options.

The AI Question

For most of modern history, education had a fairly simple economic promise: learn more, become more skilled, become more productive, earn more. The Industrial Revolution replaced human muscle with machines, and work moved rather than disappeared, new occupations, new skills, a permanently changed relationship between physical effort and production. The computer did something different: it began processing information for us, calculating, sorting, retrieving: and office work changed accordingly. Now comes AI, and the question is bigger again. What happens when the cognitive work education spent decades preparing us to perform can increasingly be done, assisted, or accelerated by a machine?

The evidence does not support a simple “AI will replace us” story. The ILO’s 2025 analysis estimated that roughly one in four workers globally are in occupations with some exposure to generative AI, but concluded that most jobs are more likely to be transformed than eliminated outright, because human input remains necessary for most of them. The sharper question is not will AI replace humans, but which parts of human work no longer need a human in the same way.

This is where education has a real problem, because it has spent decades rewarding exactly the abilities machines are becoming cheap at: remembering information, applying known formulas, producing a polished essay, solving a routine problem. UNESCO has already flagged this directly, arguing that education needs to reconsider what should be learned and how it should be assessed, since generative AI can already write an organised essay and pass some knowledge-based exams. The traditional exam made sense when a student had to carry knowledge inside their own head because access to it was limited. Today knowledge is nearly everywhere, and AI can retrieve and arrange it for us. So the important question shifts from can you remember the answer to can you tell which question is worth asking in the first place. Ask an AI for a business plan, an explanation of inflation, working code, a ten-point marketing list: you’ll get one. What becomes valuable is not producing the answer but choosing the right problem, judging the quality of what came back, knowing what’s missing, knowing when the machine is simply wrong.

There is a pattern worth naming: as machines get better at lower-level cognitive tasks, human value seems to move upward, from information toward judgement, from calculation toward interpretation, from answering toward asking. This doesn’t make the underlying knowledge useless; a doctor still needs medical knowledge to catch an AI-assisted error, an accountant still needs to recognise a wrong analysis. AI doesn’t remove the need for knowledge. It changes how that knowledge gets used. It also sharpens the credential problem from earlier, two graduates with identical degrees, one who uses AI to investigate and verify and build, the other who simply asks it to do everything, will have identical paper and very different capability. The World Economic Forum’s 2025 Future of Jobs analysis expects nearly 40 percent of workers’ core skills to change by 2030, with employers valuing both technical fluency and distinctly human capacities, creative thinking, resilience, collaboration. The future worker will likely need both, not one instead of the other.

Where the essay says human value seems to be moving.

So yes, teach AI. But teaching a child to prompt a chatbot well is not the deep answer. They need to ask is this correct, where did it come from, what assumption is hidden in here, what does the machine not understand, what should I actually decide. UNESCO’s competency framework for students takes this same human-centred approach, ethics and judgement alongside technique, which is a far more interesting goal than teaching better prompts. For centuries literacy meant can you read, then can you calculate, then can you use a computer. The next version may be: can you think effectively alongside a machine without handing your thinking over to it.

There is also a quieter, structural worry: AI may not need to replace an entire profession to disrupt how people enter it. Careers have traditionally started with simple tasks that build toward harder ones: and if a machine now handles many of the simple tasks, where does a junior analyst, programmer or accountant get the early experience that used to be the training ground? The ILO notes clerical roles show particularly high exposure to this kind of change. The question is not only will there be jobs, but where will people learn to become good at them.

If AI increasingly handles the routine, education needs to become less built around the routine, more project, discussion, research, creation, ethics, judgement, practical responsibility, not because machines can’t touch these things, but because these activities require a human to exercise judgement rather than reproduce information. The window metaphor gets more relevant here, not less. A child does not need a bigger wall of information; they already live inside more information than any generation before them. They need a better window: a way to see what matters, what’s true, what’s worth creating, what’s worth refusing. AI can supply information. It should not be the thing deciding what kind of life is worth living.

This connects directly back to agency. Give two people identical access to the same AI. One says tell me what to do. The other says here is a problem I care about, help me explore it. The technology is identical. The outcome is not. Education cannot only teach students to use powerful tools: it has to teach them when and why, because a person without agency may let the tool become the decision-maker.

There is a real danger in turning all of this into just another race, learn English, then learn coding, then learn AI, then collect an AI certificate, then join an AI coaching centre. Same machine, different label. Technology moves too fast for that anyway; a tool learned today may not exist in five years. What lasts is learning capacity itself, the ability to learn, unlearn, question, adapt.

For decades a parent could reasonably ask a child what will you become, doctor, engineer, teacher, and the answer pointed at a stable profession. In a labour market this unsettled, the more durable question may be what kind of person will you become: someone who can learn, solve, create, work with people, use technology without depending on it, adapt when the profession itself changes. That person survives a shifting job market better than someone trained narrowly for one role that may not exist in the same form in fifteen years.

If machines can increasingly perform the cognitive labour our education system was built to supply, the biggest mistake would be to keep designing education entirely around the labour market. Perhaps the purpose was never only to make yourself useful to the economy, perhaps it was to learn enough about the world to decide what kind of life is worth living, what problems are worth solving, and what kind of world is worth building.

What “Fake” Actually Means

We have travelled a long way, school, tutoring, coaching, exams, GPA, university, degrees, job preparation, employment, career competition, now AI, and at every stage found something genuinely useful sitting next to something increasingly uncomfortable. The system keeps confusing education with the machinery built around it. That brings me to a word I have deliberately avoided until now.

Fake.

Not the education itself. A child learning mathematics is not fake. A teacher explaining science is not fake. Knowledge is real, curiosity is real, and education can genuinely transform a person. When I call the system fake, I am talking about something narrower and more specific: its claim about what it is doing. The student reads the prescribed books, answers the prescribed questions, receives marks, receives a certificate, repeats this several times, becomes employable, and we look at the whole sequence and say this person has been educated. But have we actually established that, or have we established that the person can follow instructions, remember information, perform under examination conditions, and compete successfully? Those abilities matter. They are not the whole of education. “I passed the examination,” “I learned something,” and “I became a different person because I learned something” are three different sentences, describing performance, learning and development respectively, and real education should be primarily concerned with the third.

A credential is a representation of learning, not learning itself. It has value, it reduces uncertainty for an employer, signals commitment, opens doors. The problem starts when the representation becomes more important than the thing it represents, and students begin optimising for the certificate rather than for understanding. The same slide happens with examinations: a student can memorise an answer, reproduce it perfectly, score full marks, and forget most of it within six months, while another student struggles to reproduce the textbook line but understands the underlying idea deeply. The exam usually cannot tell the two apart, because it is built to measure what can be measured, and once a whole system organises itself around that, people start teaching to the measurement, and the measurement quietly becomes the mission. Eventually the chain reaches its final destination, employability, and the circle closes: school produces results, university produces degrees, coaching produces performance, the employer produces productivity. Everything has one destination. But where, in that equation, is the human being, the parent, the friend, the reader, the person who sits beside a river and watches the evening, none of which increases anyone’s GDP but all of which is part of a life?

So this is the actual thesis, stated as precisely as I can state it: the system becomes fake when it presents credential production as education, examination performance as learning, and employability as the ultimate measure of human development. That is a claim about a substitution, not an accusation against schools, teachers, degrees or universities, and not an argument that everyone should become an entrepreneur. A map is useful. A map is not the territory. A GPA is useful. A GPA is not the student. The instrument became the destination somewhere along the way, without anyone deciding it should.

I want to be fair here too. The system does produce brilliant doctors, engineers, writers, researchers, entrepreneurs. But did the system create all of that ability, or did the person survive the system and keep learning despite it, through family, through books, through failure, through their own stubborn curiosity? The people who build unusual things rarely fit neatly inside the frame; they read past the syllabus, start things, fail, teach themselves, keep going, and their formal education is part of that story but rarely the whole of it. That the human capacity to learn keeps escaping the frame is not an argument for tearing the frame down. It’s an argument for shrinking its authority, letting school be school, a degree be a degree, useful and important, without asking either one to define a human being.

Real Education

If this is not enough, what should education actually become? Not another rigid model, there is no curriculum that predicts a child’s whole future, but a different direction. Return to the window one more time: a wall says this is the world, a window says look, there is more. Education should make a child’s world larger, not smaller, and everything in this section is really one idea wearing different names.

Curiosity comes first, because a child’s why, how, what happens if is not a distraction from education, it is the beginning of it, and a good education answers a question by creating another one rather than closing the subject. Critical thinking follows close behind: not just what is the answer, but how do we know, what’s the evidence, what would change my mind, increasingly essential in a world full of misinformation and generative AI producing confident, plausible text. Then capability: the movement from I know this to I can use this, from understanding economics to understanding why a real business is struggling. Creativity, not only art, but the habit of asking is there another way rather than only what answer does the teacher want. Character, which is often the most absent from the conversation: can this person be trusted, take responsibility, admit they were wrong, resist a shortcut that harms someone else. Knowledge makes a person capable; character determines how that capability gets used. Agency, discussed already, the movement from tell me what to do toward here’s the problem I found and what I tried. Practical competence: how to manage money, read a contract, solve an everyday problem, none of it prestigious, all of it life-easing. And underneath all of it, the ability to learn: because nobody knows exactly what the next fifteen years will require, and a person who was only given one skill has a problem when that skill goes out of date, while a person who was taught how to learn can simply begin again.

I don’t want to be idealistic about money. Food, housing, healthcare, family responsibilities, these matter, and the ability to earn is legitimately part of a good education. But there’s a difference between teaching someone how to earn and teaching them that earning is the only reason to learn, and that difference is exactly the one this whole essay has been circling. Alongside it sits something even more neglected: the ability to live, to build relationships, handle disappointment, find meaning without a promotion attached to it, stop measuring yourself against someone else’s result. A person can be highly educated by every conventional measure and still have no idea how to do this. That is not a contradiction. It is one of the clearest signs that education and human development are not automatically the same thing.

None of this makes real education anti-business, or anti-exam, or anti-degree. A person with capability, judgement, creativity and agency is also, incidentally, a better employee, not an obedient one, but one who can identify a problem, improve a system, learn quickly, work with other people. The difference is that this kind of education does not reduce the person to the worker it also, usefully, produces.

So perhaps the honest final test of an education is not what grade did you get, or even what job did you get, but something harder: after all these years, has your world become larger? Can you see more, question more, create more, choose more freely, take more responsibility? If yes, something real happened, whatever the certificate says. If no, the certificate may be the only thing that happened.

We did not necessarily build a fake education system. We built a real system for producing credentials, sorting people and preparing labour: and then quietly started calling that whole process “education.” Once that distinction is visible, the only honest next question is the one underneath everything this essay has asked: what kind of human beings are we actually producing?

The Final Question

We have followed the journey from the beginning. A child enters school. Then the examination. The result. GPA. Coaching. Another examination. University admission. The degree. Job preparation. The job. Promotion. Another promotion. More responsibility. Retirement. And somewhere inside this long journey, we keep telling the child: this is for your future. Perhaps it is. But there is one question we rarely ask. What happens to the present?

The pipeline the essay has been tracing, start to end.

A child is told: study now, play later. Prepare now, enjoy later. Get good results now, life will be better later. Get into a good university, then life will begin. Get a good job, then you will be secure. Earn more, then you can enjoy life. And suddenly twenty or thirty years have passed. The child has become an adult. The adult has become an employee. And the person is still waiting for life to begin.

Perhaps the race never actually ends. The school competition becomes the university competition, becomes the job competition, becomes the promotion competition, becomes the salary competition, becomes the lifestyle competition, who bought a better house, whose children study at a better school, who retired with more money. The race simply changes its name. Compare, compete, perform, advance, repeat.

This is perhaps the saddest part. The child who once wanted to draw something may have stopped drawing. The one who loved reading stories may have stopped reading. The one who asked endless questions may have learned to stop asking: not because anyone deliberately destroyed those interests, but because life became too crowded. School, tutor, coaching, exam, university, job, bills, family, responsibilities. The window was always there. There was never enough time to look through it.

Perhaps this is what the whole system has become exceptionally good at: preparing. For the next examination, for university, for employment, for promotion, for the future. But preparation is supposed to lead somewhere. At some point we are supposed to live, and if the preparation consumes most of the life it was supposed to prepare us for, what exactly have we achieved?

From a very young age we ask children what will you become, doctor, engineer, accountant, government officer. We ask them to choose an occupation before they’ve had enough time to understand themselves. Perhaps we should sometimes ask something else: what are you curious about, what makes you feel alive, what kind of person do you want to become. The first question asks for a career. The second asks about a human being.

We have made education very good at sorting people: GPA, class position, admission score, CGPA, job title, salary, promotion. The numbers are useful; they help institutions decide and students understand performance. But when every important question becomes a ranking question, we start to believe the person at the top is somehow better. Human beings are not examination papers. A person who earns less may be a wonderful parent. Someone with an ordinary degree may be deeply knowledgeable. Someone with a huge salary may be miserable. Someone with a perfect academic record may still be unable to make an independent decision. The measurements are real. They are also incomplete.

We have become very good at measuring what a person produces. We are much less comfortable measuring who the person becomes. Curiosity, wisdom, courage, judgement, kindness, independence, a sense of meaning, the ability to enjoy an ordinary afternoon, the ability to sit with someone who is suffering, the ability to look at a problem and say I don’t know, let me find out, these are harder to measure, which may be exactly why institutions quietly stop trying. Not because they don’t matter. Because they resist a number. And yet at the end of a life, nobody remembers every examination passed or every GPA. The certificate may still hang on the wall. The life itself will have been lived somewhere else, in conversations, relationships, work that mattered, things created, people helped, mistakes made, ordinary mornings, quiet evenings, moments that never appeared on a CV.

This is not an argument against ambition. Wanting to become a doctor is beautiful. Wanting to build something is beautiful. Wanting a better life for your family is deeply meaningful. Competition is not evil. Discipline is not evil. Formal education is not evil. The problem begins when one definition of success becomes so dominant that every other form of a good life starts looking like failure. A successful person may earn a great deal. They may also, or instead, be someone who understands deeply, raises a healthy family, helps others, learns throughout life, has time for the people they love, and can live without constantly comparing themselves to someone else. That kind of success will never appear in a university ranking. It is still success. Perhaps the more important kind.

So the deepest question was never really “is Bangladesh’s education system good or bad.” It is what education is supposed to do to a human being: produce a worker, a credential, an economically useful person, or help someone understand the world, discover themselves, and decide what kind of life they want to live. Perhaps it can do all of this. But when one of them becomes dominant, the others quietly disappear.

I don’t want to end with a list of reforms, or ten solutions, or another perfect model. Perhaps what we need first is to look honestly at the person standing at the end of the system: after twelve years of school, examinations, coaching, university, degrees, job preparation, years of competition and years of employment. Who is that person? Are they curious? Independent? Can they create, think, make decisions, learn without being told what to learn, earn without becoming completely dependent, build something, love, rest, live? Or have we simply made them very good at being employable?

And perhaps the question sitting underneath this entire essay is finally the simplest one: are we creating human beings, or are we creating employable human capital? Human capital asks how useful are you. Education should also ask who are you. Human capital asks what can you produce. Education should ask what can you understand, create and become. Human capital asks what is your economic value. Education should remind us that a human being has value beyond the economy.

The child enters the system with questions. We give the child answers. The child enters with curiosity. We give the child examinations. The child enters with a world to explore. We give the child a syllabus. And perhaps somewhere along the way, we forget the most important thing.

The child is not a project. The child is a person.

The child is not a project. The child is a person.

Perhaps the greatest failure of education is not that children learn too little. Perhaps it is that they spend so much of their lives learning what to become, that they never get enough time to discover who they are.

Sources & Further Reading

This essay draws on Bangladesh education data, institutional reports, academic research, and established theories of education, human capital, credentialism, and labour markets. Statistical and historical claims are sourced where appropriate. The broader conclusions and interpretations are the author’s own.

Bangladesh Education

• SSC pass rate falls to 62.25%, GPA-5 holders also decline

  The Daily Star, 2026

  Source for the 2026 SSC pass-rate and GPA-5 figures that open the essay.

  View source →

• SSC, equivalent results published: Pass rate 64.05pc, 106,009 obtain GPA-5

  Prothom Alo, 2026

  Confirms the 11-board average pass rate and GPA-5 total cited for 2026.

  View source →

• SSC pass rate plunges to a 16-year low

  The Daily Star, 2025

  Source for the 2025 pass rate (68.45%) used as the year-on-year comparison.

  View source →

• SSC results: 669 educational institutions record 100pc pass rate

  Prothom Alo, 2025

  Source for the 984 full-pass / 134 zero-pass institution contrast discussed early in the essay.

  View source →

• QS Rankings: DU secures place among world’s top 600 univs

  The Business Standard, 2026 (reporting QS World University Rankings 2027)

  Source for Dhaka University’s ranking, discussed in the section on university prestige.

  View source →

Education, Tutoring & Shadow Education

• Education Watch Report 2023: Post-Pandemic Resilience and Sustainability

  Education Watch / CAMPE, 2023

  Institutional source for private tutoring, guidebook use, and household education spending.

  View source →

• Education costs soar: Private tuition, guidebook, notebook blamed

  New Age, 2023 (reporting Education Watch 2023 findings)

  Reports the average household education-spending figures cited in the essay.

  View source →

• Students’ dependence on guides, coaching rises in Bangladesh

  New Age, 2023 (reporting Education Watch 2022 draft findings)

  Source for the private-tutoring and commercial-guidebook usage figures cited in the essay.

  View source →

• Private Tutoring in Bangladesh: Evolution, Expansion, and Policy Responses

  International Journal of Comparative Education and Development, 2022

  Academic source on the historical growth of Bangladesh’s private tutoring and coaching sector.

  View source →

• Guidebooks, pvt tuition increase out-of-pocket costs

  The Daily Star, 2024, citing the UNESCO Global Education Monitoring Report 2022

  Source for the finding that families bear 71 percent of Bangladesh’s total education expenditure.

  View source →

Human Capital & Credentialism

• Investment in Human Capital

  Theodore W. Schultz, American Economic Review, 1961

  Foundational source for the human-capital theory discussed in the essay.

  View source →

• Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education

  Gary S. Becker, National Bureau of Economic Research, 1964

  Foundational source for the human-capital theory discussed in the essay.

  View source →

• Schooling, Experience, and Earnings

  Jacob Mincer, National Bureau of Economic Research, 1974

  Source for the empirical link between education, experience and earnings.

  View source →

• Job Market Signaling

  Michael Spence, Quarterly Journal of Economics, 1973

  Source for the essay’s discussion of credentials functioning as signals rather than pure capability.

  View source →

• The Credential Society: An Historical Sociology of Education and Stratification

  Randall Collins, 1979 (Legacy Edition, Columbia University Press)

  Source for the essay’s discussion of credential inflation.

  View source →

Education & the Purpose of Learning

• Learning: The Treasure Within

  Jacques Delors et al., UNESCO, 1996

  Source for the four pillars of learning discussed in the essay.

  View source →

Education, Skills & Labour Markets

• Voices of Bangladeshi Youth: Preparing for the World of Work

  World Bank Blogs, citing a 2019 World Bank study

  Source for the finding that over 650,000 graduates enter Bangladesh’s labour market yearly and fewer than half find work within one to two years.

  View source →

• Bangladesh: 2025 Article IV Consultation

  International Monetary Fund, 2026

  Source for the youth-unemployment and education-to-labour-market mismatch figures discussed in the essay.

  View source →

• Job Creation Key to Equality and Faster Poverty Reduction in Bangladesh

  World Bank, 2025 (Bangladesh Poverty and Equity Assessment 2025)

  Source for the finding that half of young Bangladeshis aged 15 to 29 work in low-paying jobs.

  View source →

• Working to Raise the Prospects of Bangladesh’s Youth and Migrants

  World Bank, 2025

  Source for the finding that Bangladesh’s informal sector accounts for around 85 percent of employment.

  View source →

AI & the Future of Work

• Generative AI and Jobs: A Refined Global Index of Occupational Exposure

  ILO Working Paper 140, 2025

  Source for the finding that roughly one in four workers globally have some occupational exposure to generative AI.

  View source →

• The Future of Jobs Report 2025

  World Economic Forum, 2025

  Source for the finding that employers expect close to 40 percent of workers’ core skills to change by 2030.

  View source →

• AI Competency Framework for Students

  UNESCO, 2024

  Source for the essay’s discussion of how education might respond to generative AI, emphasising judgement and ethics alongside technical skill.

  View source →