Skip to content

← Economics

One More Plate

Chapter 6 · Buyers, Sellers and Firms

In this chapter 5 sections
  1. Utility, plate by plate
  2. Water and diamonds
  3. The last rupee
  4. What a bargain is worth
  5. Prices, fairness and habit

Prapti comes out of school at four, hungry, and crosses the road from Anshu’s tea shop to the momo stall. She is made up, like Anshu, but her evening is one most students know. The first plate is wonderful. The second is good. Halfway through the third she slows down, and a fourth would be a punishment. The momos came from the same steamer, at the same price, made by the same hands. What changed was Prapti.

That small evening holds most of this chapter. Prapti never decided between momo and no momo. She decided, plate by plate, whether one more was worth its price, and she stopped at the plate where it no longer was. Nearly every economic choice is made like that, at the edge. A farmer doesn’t choose between farming and not farming, but whether to sow one more terrace. Anshu doesn’t choose between opening her shop and closing it for good, but whether to buy one more liter of milk this morning, or stay open one more hour tonight. The edge of a choice is what economists call the margin, सीमान्त, and thinking about one more of something, or one less, is called thinking at the margin. Few real choices are all or nothing. Most are a little more or a little less.

Thinking at the margin has a useful side effect: it lets the past go. The momo already eaten and the money already paid are gone whatever Prapti does next, so they shouldn’t sway the next choice. Economists call costs like these sunk. A family that has paid a year’s fees for a course their son hates, and keeps paying “so as not to waste the first year”, is letting a sunk cost decide. The first year is spent either way. The only question is whether the next year is worth what it costs.

Utility, plate by plate

The satisfaction a person gets from a thing is called its utility, उपयोगिता. In economics the word doesn’t mean usefulness in the everyday sense. A cigarette gives utility to a smoker even while it harms him, and sunlight gives it for nothing. Utility belongs to the person, not the thing. Some people love tea and some can’t stand it. The same plate gives more to a hungry student than to a full one, and a woolen jacket gives a lot in Poush and almost nothing in Jestha.

The English philosopher Jeremy Bentham thought pleasure and pain could in principle be measured and added up. Economists who followed him imagined a unit for it, the util. Nobody has ever measured a util, and nobody needs to. It is a way of putting numbers on a feeling so that the feeling’s pattern can be seen.

Here is Prapti’s evening in utils. They’re made-up numbers, but the pattern is what every textbook draws. The first plate gives her 10 utils. The second adds 8, the third 5, the fourth 3, the fifth only 1, and a sixth adds nothing at all. Forced to eat a seventh, she would feel worse, a loss of 2.

Total utility (utils)0102030highestMarginal utility (utils)−405101018253341506−27Plates of momo
Figure 6.1 Prapti’s plates of momo in one sitting, in made-up utils. Each bar below is what one more plate adds, its marginal utility; the line above is the running total, her total utility. The total climbs while each plate still adds something, is highest where a plate adds nothing, and falls after that.

Each of those numbers is a marginal utility: what one more plate adds. Add them up and you get her total utility, the satisfaction of the whole meal, which after five plates is 10 + 8 + 5 + 3 + 1 = 27 utils. Going the other way, the marginal utility of any plate is how much the total changed when that plate was eaten. After two plates the total is 18; after three it is 23; so the third plate’s marginal utility is 23 − 18 = 5.

The two numbers move together in a fixed way. As long as each plate adds something, the total goes up, though more and more slowly. When a plate adds nothing, the total is as high as it will go, and economists call that point full satisfaction. Once a plate makes her worse off, the total starts to fall. The total is highest exactly where the marginal reaches zero.

That each extra plate adds less than the one before is so common that it has a name: the law of diminishing marginal utility, सीमान्त उपयोगिता ह्रास नियम. A German civil servant, Hermann Heinrich Gossen, set it out in a book in 1854, and it is sometimes called Gossen’s first law. Gossen had a high opinion of his work; he opened the book by claiming for himself honors in economics like those of Copernicus in astronomy. Almost nobody read it. The English economist who later reached the same ideas found out about Gossen only in 1878, when a friend tracked down a copy, and he couldn’t read German. The man who made the law famous was Alfred Marshall, whose Principles of Economics of 1890 became the textbook of the age. The extra benefit a person gets from a little more of a thing, Marshall wrote, gets smaller with every increase in what he already has.

The law comes with conditions, and they explain the exceptions people raise. Nepal’s textbooks list eight cases where they don’t hold and the law fails, and for four of them, rare things, roads and schools, books and music, and habits, they say each extra unit pleases more than the last. Each is better seen as a condition broken: the thing or the person has changed. The law is about the same thing, in sensible units, eaten or used one after another, while the person’s tastes and income stay the same. A ripe mango after a sour one isn’t one more of the same thing. A plate eaten a week later isn’t one more plate of the same meal; Prapti will be hungry again by then. Sipping water by the spoonful and then by the glass changes the unit. A stamp collector’s next stamp may delight him more than the last, but often because it completes a set, which makes it a different good from the stamps he has. A smoker whose habit grows with each cigarette has changed while smoking, and so has a reader who comes to love books, for whom each new book is a different good anyway; the same song played ten times in one evening wears thin like anything else. Roads and schools, the textbooks’ other example, serve many people at once, and a road that joins others into a network is worth more than one alone; but for one family, a second road to the same village still does less than the first. Where the conditions hold, the law holds very widely.

Shops use it every day. Because the next kilo of rice is worth less to a family than the last, a seller who wants them to buy more must offer it for less, which is why bulk buying comes with a discount. A holiday is best in its first days, which is why most are short. And the idea reaches money itself. Marshall gave the example of two clerks, one earning £100 a year and one £300, deciding whether to pay twopence for a ride when it rains. The poorer clerk walks through a heavier downpour before he spends it, because to him twopence means more. The richer a person becomes, the less an extra rupee is worth to him. That is one of the old arguments for taxing larger incomes at higher rates: the same rupee taken costs a rich person less satisfaction than a poor one. Chapter 18 comes back to it.

Measuring utility in numbers is called the cardinal idea, सङ्ख्यात्मक उपयोगिता. Many economists came to think the numbers were more than anyone could know. A person can say that the first plate pleased her more than the second, and that she would rather have momos than chowmein today, without being able to say by how much. The view that only this ranking can be known, and that it is enough, is the ordinal idea, क्रमवाचक उपयोगिता. For seeing why a fourth plate gets refused, either will do.

Water and diamonds

Marginal utility settled one of the oldest puzzles in economics. Adam Smith put it in 1776. Nothing, he wrote, is more useful than water, yet it will buy hardly anything. A diamond has hardly any use, yet a great quantity of other goods can be had for it. Smith concluded that a thing’s value in use and its value in exchange, its price, were two separate matters, and for nearly a century the puzzle stood.

The answer is that a price is set at the margin. The whole of the world’s water is worth more than all its diamonds; people can live without diamonds and can’t live a week without water. But nobody buys the whole of the world’s water. A person buys one more bucket, and where water runs in every stream, one more bucket adds very little to what she already has. Diamonds are so rare that even a rich person has few, so one more adds a lot. Water’s total utility is enormous and its marginal utility is tiny. A diamond’s is the other way round. Price follows the marginal.

The same water, of course, gets dear where it is scarce. A jar of water costs nothing at a spring in the hills and costs money in Kathmandu in the dry months, when the taps run short and tankers sell it by the load. The water hasn’t changed. The last bucket has.

Much the same answer arrived three times at once. In 1871 William Stanley Jevons in England and Carl Menger in Vienna published it, each without knowing of the other, and in 1874 Léon Walras did the same in Switzerland. Economists call the change the marginal revolution. Gossen, it turned out, had seen it first, and Jevons said so, generously, in the preface to his second edition: a great share of the credit, he wrote, must go to Gossen.

The last rupee

Prapti doesn’t only eat momos. She has Rs. 300 of pocket money for snacks this week, and two things she likes: a plate of momo at Rs. 100 and a glass of tea at Anshu’s for Rs. 25. Both prices are made up, and so are her utils for a week of snacks, counted on a scale of their own, not the evening’s. How should she spend it?

Not by buying only what she likes best. All on momos buys three plates and no tea, and the third pleases her much less than the first. All on tea buys twelve glasses, and by the twelfth she is sick of it. Somewhere between, the last rupee spent on momos does as much good as the last rupee spent on tea, and that is where she should stop moving money from one to the other.

Try it

Rs. 300 for momos and tea

Move rupees from tea to momos and watch her week’s total.

Utils for every Rs. 10012345Momo, Rs. 100Tea, Rs. 25boughtnot bought
Plates of momo
Glasses of tea
8
Her week’s total
97 utils
Last plate, per Rs. 10
4.0 utils
Last glass, per Rs. 10
1.2 utils

The last glass gives about 1 util for every Rs. 10, a second plate 3.2. Move money to momos.

With one plate of momo, the last glass of tea gives her 3 utils for its Rs. 25, about 1 util for every Rs. 10, while a second plate would give 32 utils for Rs. 100, 3.2 for every Rs. 10. Money is doing much more on the momo side, so she should move some there. With three plates and no tea, the move has gone too far: the first glass of tea would now do far more than the third plate does. Two plates and four glasses is best, 111 utils in all. There the last plate gives 32 utils for Rs. 100 and the last glass 8 for Rs. 25, the same 3.2 utils for every Rs. 10. Nothing can be moved without losing.

Marshall gave the same rule with a housewife sharing a year’s yarn between socks and vests. She has it right, he wrote, when the last bundle turned into socks does as much good as the last bundle turned into vests: anything that can be put to several uses is best shared so that it has the same marginal utility in each.

In Nepal’s textbooks this is the law of substitution, प्रतिस्थापनको नियम, and elsewhere the law of equi-marginal utility, the equal-margin rule. It says a person gets the most from a budget by spending it so that the last rupee on each thing brings the same satisfaction. Nobody does the sums at a momo stall. But people do move money towards what pleases them more for the price, and away from what has started to pall, and that is all the law claims.

It also says what a price change does. If momos rise to Rs. 150, each plate does less for every rupee, and the balance tips towards tea. With her utils, one plate and six glasses is now best: Prapti buys fewer plates. That is why, for nearly everything, people buy less of a thing when its price goes up, the beginning of demand, which Chapter 7 takes up.

What a bargain is worth

Marshall took tea for his example. Imagine, he wrote, a man who would pay 20 shillings a pound for tea rather than have none at all, and buy just one pound a year at that price. At 14 shillings he would buy a second pound, at 10 a third, at 6 a fourth, at 4 a fifth, at 3 a sixth. In fact tea costs 2 shillings, and he buys seven pounds.

Shillings051015201234567Pounds of teaSurplus: 45sPaid: 14s
Figure 6.2 Marshall’s tea buyer. Each bar is the most he would pay for one more pound of tea. He pays 2 shillings for each (the line); the green part above it is what he gets without paying for it.

Every one of those pounds is worth at least as much to him as he would have paid for it: the first 20 shillings, the second 14, and so on down to the seventh, worth just the 2 he paid. Together the seven are worth at least 59 shillings to him. He pays 14. The other 45 are a gain he gets for nothing, because the price of the last pound is the price of them all.

Marshall called the difference between what a person would be willing to pay and what she actually pays the consumer’s surplus, उपभोक्ताको बचत. It is why buying something feels like a small win. A man who has been on a cold bus since dawn might have paid Rs. 60 for Anshu’s first glass of tea. He pays Rs. 25, the price everyone pays, and keeps the rest of the value.

The seller has a surplus too. Suppose milk, tea leaves, sugar and gas for one glass cost Anshu Rs. 10, so she would rather sell it for Rs. 10 than not at all. She gets Rs. 25. That Rs. 15 is her producer’s surplus, उत्पादकको बचत. Add the two together and you have the gain a sale creates for both sides at once. Neither had to be cheated for the other to win. Each traded something worth less to them for something worth more, which is why a voluntary trade, so long as both know what they are getting, leaves both better off.

Prices, fairness and habit

Everything so far rests on one plain assumption: people weigh what they gain against what they give up, and change their choices when the balance changes. Mostly it works. Rewards and penalties that push a choice one way or another are called incentives, and people respond to them, usually predictably.

Nepal tested one in 2005. A national program began paying women in cash to give birth in a government health facility, where a trained attendant could help if things went wrong. The payment was largest where getting to a facility was hardest: Rs. 1,500 in the mountain districts, Rs. 1,000 in the hills, Rs. 500 in the Tarai. Health workers got Rs. 300 for each birth they attended. It worked, modestly at first: women who had heard of the payment were about 17% more likely to give birth with a skilled attendant than similar women who hadn’t, and money slow to reach districts and a scheme hard to explain held it back.

But money isn’t the only thing people weigh, and sometimes it pulls the wrong way. In 1998 two economists studied ten day-care centers in Haifa, in Israel. Parents often came late to collect their children, and a teacher had to wait with them. Six centers brought in a small fine, 10 shekels for coming 10 minutes late or more. Late arrivals didn’t fall. They rose, and within a few weeks they had almost doubled, while the four centers with no fine stayed the same. When the fine was taken away, the parents kept coming late. The researchers’ explanation: before the fine, coming late meant imposing on a teacher who stayed out of kindness; the fine turned it into a service with a price, and a cheap one, and removing the price didn’t bring the old feeling back.

Fairness weighs too. In the ultimatum game, first described by economists in 1982, one player is given a sum of money and offers the other a share of it. If the second accepts, both keep their parts. If she refuses, neither gets anything. A person thinking only of money should accept any offer, since something beats nothing. People don’t. In one study comparing farmers in Kenya with students in the United States, only 2 farmers in 100 accepted a tenth of the sum, and almost half turned down even three-tenths. They gave up real money to punish a split they thought unfair. Students accepted a tenth more often, a little over half the time, so how much fairness weighs differs from place to place; but in all fifteen societies the larger study covered, people gave up something to punish an unfair split.

And habit weighs, often more than money. When a large American company changed its savings plan in 1998 so that new workers were put in unless they chose to leave, instead of left out unless they chose to join, 86% of new workers ended up saving in it. Of those hired a little earlier, under the old rule, only 37% had joined at the same stage. Only the starting point had changed, and most people stayed where they were put. The economist Richard Thaler and the lawyer Cass Sunstein called a change like this a nudge, in a book of that name in 2008: arranging a choice so that people tend to pick well, without forbidding anything or changing its price.

Economists argue about what all this means. Some think these findings show that the careful chooser of the marginal models is a poor picture of real people, and that policy should be built on how people actually behave. Others answer that the standard model still predicts the direction of most choices well, that people facing big, repeated decisions with their own money learn to choose carefully, and that nudges can be used by those who design them for their own ends. What none of them doubt is that the margin matters: change what one more costs, or what one more is worth, and people change what they do.

So a choice is rarely all or nothing. It is one more plate, one more glass, one more hour, and what is already spent doesn’t count. Each extra unit of a thing usually gives less satisfaction than the one before, which is why total satisfaction peaks and why water, though precious, is cheap. A person gets the most from a budget when the last rupee on each thing does the same good, so a rise in price means buying less. Buyers and sellers both gain from a fair trade, by the difference between the price and what the thing is worth to each. And people answer to prices, but also to fairness, trust and habit, and a sensible rule allows for all of them.

Found a mistake? Please tell us.