The tuition economy

The school day ends at two. The second one starts at four. Between them there is a gap long enough for a bus ride and a plate of something, and in most Indian cities the coaching market lives there: two rooms above a shop, a banner listing last year's ranks, a generator on the pavement for the power cuts.

It is tempting to read this as enthusiasm — parents who care, children who work hard. That reading explains none of the prices. A tuition class is a business with a cost structure, a labour supply and a demand curve, and its demand curve only moves one way. What follows describes it as a market rather than as a habit.

Three conditions, and all three are present

The first is that the seats are scarce and that a written score decides who gets them. Gross enrolment in higher education is about 28 per cent of the 18-to-23 age group — a queue roughly three and a half times the number of places — and the institutions whose names carry weight are a smaller subset again. Once allocation runs on rank, the score stops being a measure of learning and becomes a currency with an exchange rate.

The second is that school quality is uneven and — this matters more — that it is unobservable at the moment of choosing. A parent can see a building, a board, a uniform and a bus; a parent cannot see whether the person at the front of the room can explain why a fraction grows when its denominator shrinks. Where a product's quality cannot be inspected, buyers reach for a proxy, and a coaching class sells itself as exactly that proxy.

The third is a willingness to pay for a small number of marks. Because admission is a rank and not a threshold, the value of ten extra marks is not constant: in the middle of the distribution it is worth very little, at the boundary a year of a family's income. That convexity is the reason the price is inelastic. Nobody is buying the average mark. They are buying the mark that moves them across a line.

How a batch is priced

The unit of sale is a seat in a room for a year, not an hour of teaching, and that sets the whole price schedule. The seller's costs — rent, power, a teacher, a photocopier — are nearly all fixed; the only variable is how many chairs are filled. So the seller wants a full hall, and the buyer, who is paying for attention, wants a small one. The product range is that negotiation.

There are three tiers. The hall takes sixty to a hundred, runs on a projector and a microphone, and charges least per head. The mid batch takes twenty-five to thirty and is most of the market. The top tier takes ten or fifteen and is taught by a name that parents recognise from last year's results list. Same syllabus, same hours, several times the fee. What is bought at the top is not more teaching. It is a filter — a batch small enough that the others in it were also selected — plus a name the family can repeat.

That is why coaching firms are built around people rather than subjects, and price the name. The name is the asset, and a name can walk out and take the product along.

Why the teacher teaches twice

The labour supply is the same pool that staffs the schools. A government school post is one of the better-paid jobs open to a graduate in most of the country; a private school post pays a fraction of it, because its fee is set by what parents can pay. A teacher in the second category has an obvious arbitrage: the same subject, taught in the evening, paid in cash, at an hourly rate the school cannot match.

This is not a scandal, it is a timetable. The hours a school most needs — the ones after the syllabus is finished and the questions start — are the hours that pay best elsewhere. So the tutoring is not independent of the school. The tutor is often the schoolteacher; where not, the tutor is frequently the person who sat the school's recruitment test and did not get in. Either way, two markets are pricing one skill, and one of them is paying more.

A second effect gets less attention. Because tuition is paid per batch and per result, it rewards a particular kind of teaching — fast, exam-shaped, repeatable — and it rewards it in cash. The slower work is left to the school: the child who has fallen behind, the language the child does not speak at home, the class of fifty. The two systems are not competing for the same job. They are competing for the same hours.

What a parent can see

Teaching quality is not observable afterwards either. A parent who sat through a class could form an opinion about the explanation, but almost none do; they see a receipt and a test score. The score is the only legible output, and it is written in the units the admission system uses.

So the feedback loop runs on rank. If the rank improves, the class worked. If it does not, the parent cannot tell whether the teaching failed, the child did not work, or the paper was harder than the last one — and the safest response is to buy a more expensive batch, which is also a visible statement of how seriously the family is taking the year.

This is why so many centres run their own weekly tests, rank them and send the results home. The test is the instrument that makes the effect visible in a number the customer already trusts — and it produces the next sale.

The seats do not multiply

Here is the arithmetic nothing escapes. The number of seats is set by the institutions that own them. Tuition does not create one. If every family in the queue doubles its spending, the queue is the same length and the seats are the same number; what changes is the score at which the last seat is filled. The cutoff is not a standard but a residue: whatever the marginal family was willing to pay for, converted into marks.

The private return is therefore real and the social return is zero. A family that buys tuition can improve its position; all families buying tuition improves nobody's position and raises everybody's cost. That is a positional good, and positional goods have a characteristic price path: spending climbs to the limit of the household budget, then a little further, because the alternative is the fall in rank that comes from standing still.

The costs land in three places. In the household budget, where education is one of the few lines defended when income falls. In the child's day, which runs from seven to nine with school as its shorter half. And in the distribution of seats, which becomes more closely correlated with income without a single rule being changed. The coaching market does not have to break the meritocratic rule to make it less meaningful. It only has to sell the preparation for it.

Which is why the claim that tuition improves education gets the causality backwards. It improves one family's position in a fixed queue, and it charges that family for a service whose value depends on other families not buying it. The two variables that would change the total are the number of seats and the rule that allocates them. Neither is set in the coaching market — and that is precisely why the market can grow for a long time without producing the thing it advertises.

This is a piece in Everyday Life, part of About India.