Air pollution

11 Market successes and failures

11.1 Introduction

The pesticide chlordecone was used on banana plantations in the Caribbean islands of Guadeloupe and Martinique (both part of France) to kill the banana weevil. It was perfectly legal, and to the plantation owners it was an effective way of reducing costs and boosting the plantations’ profits.

As the chemical was washed off the land into rivers that flowed to the coast, it contaminated freshwater prawn farms, the mangrove swamps where crabs were caught, and what had been rich coastal spiny lobster fisheries. The livelihoods of fishing communities were destroyed and those who ate contaminated fish fell ill.

The fact that this pesticide was a grave danger to humans had been known since the time it was introduced, when workers in the US producing the chemical reported symptoms of neurological damage, leading to its prohibition in 1976. The French government received reports on contamination in Guadeloupe a few years later, but waited until 1990 to ban the substance, and were pressured by banana plantation owners to give them a special exemption until 1993.

Twenty years later, fishermen protesting the slow pace of French government assistance in addressing the fallout from the contamination demonstrated in the streets of Fort de France (the largest town in Martinique) and barricaded the port. Franck Nétri, a Gaudeloupean fisherman, worried: ‘I’ve been eating pesticide for 30 years. But what will happen to my grandchildren?’

He was right to worry. By 2012, the fraction of Martiniquean men suffering from prostate cancer was the highest in the world and almost twice that of the second-highest country, and the mortality rate was well over four times the world average. Neurological damage in children, including cognitive performance, had also been documented.

Let’s think of the chlordecone problem as a doctor would.

First, we diagnose the problem. The problem is that the actions of the banana plantation owners endanger the fishermen’s livelihood and health, but these costs of using the pesticide do not show up anywhere in the profit and loss calculations of the plantation owners. The price of pesticides—their cost as seen by the plantation owners—does not include the downstream costs imposed on the fishermen.

social dilemma
A situation in which actions, taken independently by individuals in pursuit of their own private objectives, may result in an outcome that is inferior to some other feasible outcome that could have occurred if people had acted together, rather than as individuals.
market failure
When markets allocate resources in a Pareto-inefficient way.

Next, we aim to devise a treatment. In some cases, the treatment is obvious. Chlordecone was simply banned in France and the US, and its use could have been vastly reduced if the plantation owners had been required (by law or by private agreement with those affected) to pay damages to the fishing communities for the harm inflicted by their pesticide.

The social dilemma associated with the use of chlordecone is termed a market failure as it arises from the buying and selling of pesticides and bananas on markets.

But notice that our suggested treatment of the problem was not simply to abandon using markets as a way of distributing bananas, pesticides, fish, and the other products making up the sad story of Martinique’s experience in this case. Our treatment was to harness the market through targeted policies to better serve social ends, not to abandon its use. The very fact that bananas—a crop that originated half the world away in the islands of the western Pacific Ocean—are grown in the Caribbean for consumption around the world is itself an example of a market success.

To understand why markets fail in cases like that of chlordecone, it is helpful to remember the conditions that are needed for markets to work well.

11.2 The market and other institutions

Markets are one way of organizing the production and distribution of goods and services. For example, you may hire a person on the labour market to take care of your child while you are at work. But your child’s caregiver could also be a relative who is not paid a wage. Or the infant could be cared for in a government daycare centre that is free as a matter of right to any citizen, or by the firm you work for, as part of your compensation package. Each of these are examples of different economic institutions—markets, families, governments, and firms—organizing some particular activity.

In Unit 2 you learned that our interactions with others can be represented as games; for this reason, the institutions that organize these interactions can be described as the rules of the game. In Unit 5 you saw that the way pirates interacted at sea was determined by the rules of the game laid out in the constitution of The Royal Rover.

In Section 11.1, we contrasted markets with firms, families, and governments. As economists, we can think of each situation as a different game with its own particular rules. In this sense, markets are just one of the ways that we organize our societies—you pay for what you get. But what about the others?

Just like a country’s constitution, we can take all the rules of the game that apply to us in our roles as parents, voters, employees, shoppers, and so on, and consider them to also have constitutions. The rules of the game of markets can be seen as a ‘constitution’, with firms, families, and governments making up the institutions that jointly organize how we interact with each other in producing and distributing our livelihoods.

Markets are essential economic institutions, not because they work perfectly (as we shall see in this unit), but because there are a great many aspects of production and distribution for which markets do better than the alternative institutions—governments, families, and firms.

11.3 Markets, specialization, and the division of labour

When you hear the word ‘market’ what word do you think of? ‘Competition’ probably comes to mind, and you would be right to associate the two words.

But ‘cooperation’ applies too. Why?

Because markets allow each of us, pursuing our private objectives, to work together, producing and distributing goods and services in a way that, while far from perfect, is in many cases better than the alternatives. Markets allow us to interact with people (for the most part complete strangers to us) in ways that we can mutually benefit from, specializing in the things we are relatively good at doing.

This is a more amazing accomplishment than it might at first seem.

Look around at the objects in your workspace. Do you know the people who made them? What about your clothing? Or anything else in sight from where you are sitting?

Now imagine that it is 1776, the year that Adam Smith wrote The Wealth of Nations. The same questions, asked anywhere in the world at that date, would have had different answers.

At that time, many families produced a wide array of goods for their own use, including crops, meat, clothing, even tools. Many of the things that you might have spotted in Adam Smith’s day would have been made by a member of the family or someone in the village. You would have made some objects yourself; others would have been made locally and purchased from the village market.

One of the changes that was underway during Adam Smith’s life, but has greatly accelerated since, is specialization in the production of goods and services. As Smith explained, we become better at producing things when we each focus on a limited range of activities. This is true for three reasons:

economies of scale
These occur when doubling all of the inputs to a production process more than doubles the output. The shape of a firm’s long-run average cost curve depends both on returns to scale in production and the effect of scale on the prices it pays for its inputs. Also known as: increasing returns to scale. See also: diseconomies of scale.

These are the advantages of working on a limited number of tasks or products. People do not typically produce the full range of goods and services that they use or consume in their daily lives. Instead, we specialize, some producing one good, others producing other goods, some working as welders, others as teachers or farmers. This is called the division of labour.

Adam Smith begins The Wealth of Nations with the following sentence:

The greatest improvement in the productive powers of labour, and the greater part of the skill, dexterity, and judgment with which it is anywhere directed, or applied, seem to have been the effects of the division of labour.1

But people do not specialize unless they have a way to acquire the other goods they need. For this reason, specialization and the resulting division of labour pose a problem for society: how are the goods and services to be distributed from the producer to the final user?

This result—the coordination of the division of labour—is accomplished in different ways, depending on a society’s institutions. In the course of history, the division of labour has been coordinated by means of direct government requisitioning and distribution, as was done in the US and many economies during the Second World War, or by gifts and voluntary sharing as we do in families today; it was even practised among unrelated members of a community by our hunter-gatherer ancestors. Today, in most countries, markets play an essential role in coordinating the division of labour.

Chapter 3 in Smith’s Wealth of Nations is titled ‘That the Division of Labour is Limited by the Extent of the Market’, in which he explains:

When the market is very small, no person can have any encour­agement to dedicate himself entirely to one employment, for want of the power to exchange all that surplus part of the produce of his own labour, which is over and above his own consumption, for such parts of the produce of other men’s labour as he has occasion for.

Markets accomplish an extraordinary result: they create unintended cooperation on a global scale. The people who produced the phone in your pocket did not know or care about you. They produced it because they are better at producing phones than you are, or were willing to work for lower wages than you are. You ended up with the phone because you paid the producers, allowing them to buy goods, also produced by total strangers to them.

11.4 The ‘magic of the market’: Prices are messages plus motivation

The key to how this process works can be expressed in a single sentence. When markets work well, prices send messages about the real scarcity of goods and services. The messages provide information that motivates people to take account of what is scarce and what is abundant, and as a result to produce, consume, invest, and innovate in ways that make the best use of an economy’s productive potential.

Prices coordinate specialization among complete strangers

If drought in the American Great Plains means that there is less wheat on the world market, the resulting increase in the price of bread sends a message to the shopper: ‘Consider putting potatoes or rice on the table tonight instead of bread.’ The shopper may know nothing about weather conditions in America and need not be the slightest bit concerned about consuming less of a good that has become scarcer. To respond to the message of the higher price in a way that makes the best use of a society’s available resources, the shopper needs to be concerned about just one thing: saving money. The shopper not only gets the message, but has a good reason to act on it.

It is this—the fact that prices combine information and a reason to act on the information—that allows the market system (many markets interlinked) to coordinate the division of labour through the exchange of goods among entire strangers, without centralized direction. Friedrich Hayek, who was part economist and part philosopher, suggested we think of the market as a giant information-processing machine that produces prices; the prices provide information that guides the economy, usually in desirable directions. The remarkable thing about this massive computational device is that it’s not really a machine at all. Nobody designed it, and nobody is at the controls. When it works well, we use phrases like ‘the magic’ of the market.

Are the prices sending the right messages?

But for this to be the case, the messages that prices send must convey the right information—how scarce a good really is. Think about what this means—the scarcity of a good is measured by its social marginal cost, that is, the total cost of having one more unit of it, including not only the cost of those producing and distributing it, but also the external effects imposed on others (for example, environmental damages).

You have seen many cases in the previous units in which the price of a good is not equal to its social marginal cost. The price of bananas in Martinique, for example, did not include the loss of life and livelihood inflicted on the downstream fishing community by the pesticides used on the plantations.

The price may fail to reflect the social marginal cost due to either:

When prices send the wrong messages, we ask whether some modification of how markets work could be introduced by public policies to improve economic outcomes, for example, taxing production processes that emit greenhouse gases or subsidizing basic research.

We now illustrate how prices can send the right message, and sometimes not, by two real world cases.

Great economists Friedrich Hayek

Friedrich Hayek The Great Depression of the 1930s ravaged the capitalist economies of Europe and North America, throwing a quarter of the workforce out of work in the US. During the same period, the centrally planned economy of the Soviet Union continued to grow rapidly under a succession of five-year plans. Even the arch-opponent of socialism, Joseph Schumpeter, had conceded: ‘Can socialism work? Of course it can. … There is nothing wrong with the pure logic of socialism.’2

Friedrich Hayek (1899–1992) disagreed. Born in Vienna, he was an Austrian (later British) economist and philosopher who believed that the government should play a minimal role in the running of society. He was against any efforts to redistribute income in the name of social justice. He was also an opponent of the policies advocated by John Maynard Keynes, designed to moderate the instability of the economy and the insecurity of employment.

Hayek’s book, The Road to Serfdom, was written against the backdrop of the Second World War, when economic planning was being used both by German and Japanese fascist governments, by the Soviet communist authorities, and by the British and American governments. He argued that well-intentioned planning would inevitably lead to a totalitarian outcome.3

His key idea about economics—that prices are messages—revolutionized how economists think about markets. Messages convey valuable information about how scarce a good is, information that is available only if prices are free to be determined by supply and demand, rather than by the decisions of planners. Hayek even wrote a comic book, which was distributed by General Motors, to explain how this mechanism was superior to planning.

But Hayek did not think much of the theory of competitive equilibrium, in which all buyers and sellers are price-takers. ‘The modern theory of competitive equilibrium,’ he wrote, ‘assumes the situation to exist which a true explanation ought to account for as the effect of the competitive process.’4

In Hayek’s view, assuming a state of equilibrium (as Walras, one of the founders of the neoclassical school of economics, had done in developing general equilibrium theory) prevents us from analysing competition seriously. He defined competition as ‘the action of endeavouring to gain what another endeavours to gain at the same time.’ Hayek explained:

Now, how many of the devices adopted in ordinary life to that end would still be open to a seller in a market in which so-called ‘perfect competition’ prevails? I believe that the answer is exactly none. Advertising, undercutting, and improving (‘differentiating’) the goods or services produced are all excluded by definition—’perfect’ competition means indeed the absence of all competitive activities.5

The advantage of capitalism, to Hayek, is that it provides the right information to the right people. In 1945, he wrote:

Which of these systems [central planning or competition] is likely to be more efficient depends mainly on the question under which of them we can expect [to make fuller use] of the existing knowledge. This, in turn, depends on whether we are more likely to succeed in putting at the disposal of a single central authority all the knowledge which ought to be used but which is initially dispersed among many different individuals, or in conveying to the individuals such additional knowledge as they need in order to enable them to dovetail their plans with those of others.6

11.5 Prices as messages

Prices worked as messages on a global scale even before the transatlantic telegraph was introduced. Students of American history learn that the defeat of the southern Confederate states in the American Civil War ended the use of slaves in the production of cotton and other crops in that region. There is also an economics lesson in this story.

At the war’s outbreak on 12 April 1861, President Abraham Lincoln ordered the US Navy to blockade the ports of the Confederate states. To preserve the institution of slavery, these states had declared themselves independent of the US.

Lincoln’s blockade halts the export of cotton

As a result of the naval blockade, the export of US-grown raw cotton to the textile mills of Lancashire in England came to a virtual halt, eliminating three-quarters of the supply of this critical raw material. Sailing at night, a few blockade-running ships evaded Lincoln’s patrols, but 1,500 ships were destroyed or captured.

excess demand
A situation in which the quantity of a good demanded is greater than the quantity supplied at the current price. See also: excess supply.

We have seen in Unit 7 that the market price of a good, such as cotton, is determined by the interaction of supply and demand. In the case of raw cotton, the tiny quantities reaching England through the blockade were a dramatic reduction in supply. There was large excess demand—at the prevailing price, the quantity of raw cotton demanded exceeded the available supply. As a result, some sellers realized they could profit by raising the price. Eventually, cotton was sold at prices six times higher than before the war, keeping the lucky blockade runners in business. Consumption of cotton fell to half the prewar level, throwing hundreds of thousands of people who worked in cotton mills out of work.

British textile mill owners increase demand

Mill owners responded. For them, the price rise was an increase in their costs. Some firms failed and left the industry due to the reduction in their profits. Mill owners looked to India to find an alternative to US cotton, greatly increasing the demand for cotton there. The excess demand in the markets for Indian cotton gave some sellers an opportunity to profit by raising prices, resulting in increases in the prices of Indian cotton, which quickly rose to almost match the price of US cotton.

Farmers in India and Egypt switch to cotton

Responding to the higher income now obtainable from growing cotton, Indian farmers abandoned other crops and grew cotton instead. The same occurred wherever cotton could be grown, including Brazil. In Egypt, farmers who rushed to expand the production of cotton in response to the higher prices began employing slaves, captured (like the American slaves that Lincoln was fighting to free) in sub-Saharan Africa.

There was a problem. The only source of cotton that could come close to making up the shortfall from the US was in India. But Indian cotton differed from American cotton and required an entirely different kind of processing. Within months of the shift to Indian cotton, new machinery was developed to process it.

Mill owners introduce new machinery

As the demand for this new equipment soared, firms that made textile machinery, like Dobson and Barlow, saw profits take off. We know about this firm, because detailed sales records have survived. Dobson and Barlow responded by increasing production of these new machines and other equipment. No mill could afford to be left behind in the rush to retool, because if it didn’t, it could not use the new raw materials. The result was, in the words of Douglas Farnie, a historian who specialized in the history of cotton production: ‘such an extensive investment of capital that it amounted almost to the creation of a new industry.’7

A change in price was the message and the motivation

The lesson for economists—Lincoln ordered the blockade, but in what followed, the farmers and sellers who increased the price of cotton were not responding to orders. Neither were the mill owners, who cut back the output of textiles and laid off the mill workers, nor were the mill owners desperately searching for new sources of raw material. By ordering new machinery, the mill owners set off a boom in investment and new jobs.

All these decisions took place over a matter of months, made by millions of people, most of whom were total strangers to one another, each seeking to make the best of a totally new economic situation. American cotton was now scarcer, and people responded, from the cotton fields of Maharashtra in India, to the Nile delta, to Brazil, to the Lancashire mills.

To understand how the change in the price of cotton transformed the world cotton and textile production system, think about the prices determined by markets as messages. The increase in the price of US cotton shouted: ‘Find other sources, and find new technologies appropriate for their use.’ Similarly, when the price of petrol rises, the message to the car driver is: ‘Take the train’, which is passed on to the railway operator: ‘There are profits to be made by running more train services.’ When the price of electricity goes up, the firm or the family is being told: ‘Think about installing photovoltaic cells on the roof.’

In many cases (like the chain of events that began at Lincoln’s desk on 12 April 1861) the messages make sense, not only for individual firms and families, but also for society; if something has become more expensive, then it is likely that more people are demanding it, or the cost of producing it has risen, or both. By finding an alternative, the individual is saving money and conserving society’s resources. This is because, in some conditions, prices provide an accurate measure of the scarcity of a good or service.

See ‘Who’s in Charge?’, Chapter 1 of Paul Seabright’s book, for more detail on how market economies manage to organize complex trades among strangers.

Paul Seabright. 2010. The Company of Strangers: A Natural History of Economic Life (Revised Edition). Princeton, NJ: Princeton University Press.

Centrally planned economies (or firms)

In Unit 1 we discussed planned economies, which operated in the Soviet Union and other central and eastern European countries before the 1990s. In these economies, messages about how things are produced are sent deliberately by government experts. They decide what is produced and at what price it is sold. The same is true, as we saw in Unit 6, in large firms like General Motors, where managers (and not prices) determine who does what.

The amazing thing about prices determined by markets is that individuals do not send the messages, but rather the anonymous interaction of sometimes millions of people. And when conditions change—a cheaper way of producing bread, for example—nobody has to change the message (‘put bread instead of potatoes on the table tonight’). A price change results from a change in firms’ costs. The reduced price of bread says it all.

11.6 Putting motivation behind the price message

Fish and fishing are a major part of the life of the people of Kerala in India. Most of them eat fish at least once a day, and more than a million people are involved in the fishing industry. But before 1997, prices were high and fishing profits were limited due to a combination of waste and the bargaining power of fish merchants, who purchased the fishermen’s catch and sold it to consumers.

The importance of timing

When returning to port to sell their daily catch of sardines to the fish merchants, many fishermen found that the merchants already had as much fish as they needed that day. They refused to buy any more fish at any price. The price was effectively zero! Fishermen at these markets just dumped their worthless catch back into the sea.

A lucky few returned to the right port at the right time when demand exceeded supply, and they were rewarded by extraordinarily high prices.

On 14 January 1997, for example, 11 boatloads of fish were brought to the market at Badagara, which was found to be oversupplied; the catch was jettisoned. There was excess supply of 11 boatloads. But at fish markets within 15 km of Badagara, there was excess demand—15 buyers left the Chombala market unable to purchase fish at any price. The luck, or lack of it, of fishermen returning to the ports along the Kerala coast is illustrated in Figure 11.1.

Only seven of the 15 markets did not suffer either from over- or under-supply. In these seven villages (on the vertical line), prices ranged from Rs4 per kg in the market at Aarikkadi to more than Rs7 per kg in Kanhangad.

Figure 11.1 Bargaining power and prices in the Kerala wholesale fish market (14 January 1997). (Note: Two markets had the same outcome, with a price of Rs6.2 per kg.)

Robert Jensen. 2007. ‘The Digital Provide: Information (Technology), Market Performance, and Welfare in the South Indian Fisheries Sector.’ The Quarterly Journal of Economics 122 (3) (August): pp. 879–924.

When the fishermen had bargaining power because there was excess demand, they got much higher prices. In markets with neither excess demand nor excess supply, the average price was Rs5.9 per kg, shown by the horizontal dashed line. In markets with excess demand, the average was Rs9.3 per kg. The fishermen fortunate enough to put in at these markets obtained extraordinary profits, if we assume that the price in markets with neither excess demand nor supply was high enough to yield economic profits. Of course, on the following day they may have been the unlucky ones who found no buyers at all and would have to dump their catch into the sea.

Prices contained important messages: ‘Fish are scarce in Chombala, but not in Badagara’, but the fisherman did not receive them in time to motivate them to change what they did so as to adjust to surpluses in one market and excess demand in others.

Cell phones deliver the message on time

This all changed when the fishermen got cell phones. While still at sea, the returning fishermen phoned the beach fish markets and picked the one where the prices were highest that day. If they returned to a high-priced market, they would earn an economic rent (that is, income in excess of their next best alternative—returning to a market with no excess demand or, even worse, one with excess supply).

By gaining access to real-time market information on relative prices for fish, the fishermen could adjust their pattern of production (fishing) and distribution (the market they visit) to secure the highest returns.

A study of 15 beach markets along 225 km of the northern Kerala coast found that, once the fishermen had cell phones, differences in daily prices among the beach markets were cut to a quarter of their previous levels. No boats jettisoned their catches. Reduced waste and the elimination of the dealers’ bargaining power raised the profits of fishermen by 8%, at the same time as consumer prices fell by 4%.

Cell phones allowed the fishermen to become very effective rent seekers. Their rent-seeking activities changed how Kerala’s fish markets worked—virtually eliminating the periodic excess demand and supply—to the benefit of fishermen and consumers, but not of the fish dealers who had acted as middlemen.

This happened because the Kerala sardine fishermen could respond to price messages. The information given by the prices at different beach markets came at the right time, so they were motivated to land at the markets where fish was needed most.

Question 11.1 Choose the correct answer(s)

Figure 11.1 shows how bargaining power affected prices in Kerala beach markets on 14 January 1997. Based on this information, what can we conclude?

  • The higher the excess supply, the lower the price of fish.
  • The price of fish in all markets with excess demand is Rs9.3 per kg.
  • In cases of demand being equal to supply, the price that fishermen faced was the same in all markets.
  • The data demonstrates that buyers have bargaining power when there is excess supply.
  • The price of fish is zero in all markets with excess supply.
  • The average price in the markets with excess demand is Rs9.3, per kg but the higher the excess demand, the higher the price.
  • Even when demand equalled supply, fish were sold at different prices in different places.
  • When there is excess supply, the price is zero. The buyers have all the bargaining power, while sellers have none.

11.7 Market failure: External effects of pollution

But markets do not always work well. We return to take a closer look at the diagnosis and treatment of a case like the one in Section 11.1, of pesticides in Martinique and Guadeloupe. As we saw in Unit 1, private property is a key requirement for a market system. For something to be bought and sold, someone must claim the right to own it. You would not pay for something unless you believed that others would acknowledge (and if necessary protect) your right to keep it.

property rights
Legal protection of ownership, including the right to exclude others and to benefit from or sell the thing owned.

For a market to work effectively (or even to exist), other social institutions and social norms are required. Governments provide a system of laws and law enforcement that guarantees property rights and enforces contracts.

Many of the problems with markets that we investigate in this unit arise because of difficulties of guaranteeing property rights or writing appropriate contracts. There are goods—like clean rivers—that matter to people but cannot easily be bought and sold. There are ‘bads’ like second-hand smoke that a person can impose on others without paying the damages, as he might, for example, if the damage was to the other person’s car.

Exercise 11.1 Property rights and contracts in Madagascar

Marcel Fafchamps and Bart Minten, two economists, studied grain markets in Madagascar in 1997, where the legal institutions for enforcing property rights and contracts were weak. Despite this, they found that theft and breach of contract were rare. The grain traders avoided theft by keeping their stocks very low, and if necessary, sleeping in the grain stores. They refrained from employing additional workers for fear of employee-related theft. When transporting their goods, they paid protection money and travelled in convoy. Most transactions were paid in cash. Trust was established through repeated interaction with the same traders.8

  1. Do these findings suggest that strong legal institutions are not necessary for markets to work?
  2. Consider some market transactions in which you have been involved. Could these markets work in the absence of a legal framework, and how would they be different if they did?
  3. Can you think of any examples in which repeated interaction helps to facilitate market transactions?
  4. Why might repeated interaction be important even when a legal framework is present?
market failure
When markets allocate resources in a Pareto-inefficient way.

When markets allocate resources in a Pareto-inefficient way, we describe this as a market failure. We encountered one cause of market failure in Unit 7—a firm producing a differentiated good chooses its price and output level such that the price is greater than the marginal cost. In contrast, we know from later in Unit 7 that a competitive market allocation maximizes the total surplus of the producers and consumers and is Pareto efficient, as long as no one else is affected by the production and consumption of the good.

external cost
A negative external effect: that is, the negative effect of production, consumption, or other economic decisions on another person or party, which is not specified as a liability in a contract. Also known as: external diseconomy. See also: external effect.
external effect
When a person’s action confers a benefit or cost on some other individual, and this effect is not taken account of by the person in deciding to take the action. It is external because it is not included in the decision-making process of the person taking the action. Positive effects refer to benefits, and negative effects to costs, that are experienced by others. A person breathing second-hand smoke from someone else’s cigarette is a negative external effect. Enjoying your neighbour’s beautiful garden is a positive external effect. Also known as: externality. See also: incomplete contract, market failure, external benefit, external cost.
marginal private cost (MPC)
The cost for the producer of producing an additional unit of a good, not taking into account any costs its production imposes on others. See also: marginal external cost (MEC), marginal social cost (MSC).
marginal social cost (MSC)
The cost of producing an additional unit of a good, taking into account both the cost for the producer and the costs incurred by others affected by the good’s production. Marginal social cost is the sum of the marginal private cost and the marginal external cost.

But others are often affected by the production or consumption of the good. The market allocation of the good will not be Pareto efficient if the decisions of producers and consumers affect others in ways that they do not adequately consider. This is a social dilemma, and in this case, a cause of market failure. Recall the social dilemmas we studied in Unit 2 such as the overuse of antibiotics. When we analyse gains from trade in such cases, we must consider, not only the consumer and producer surplus, but also the costs or benefits experienced by parties who are neither buyers nor sellers. For example, the superbug that emerges because of the sale and overuse of an antibiotic may kill someone who had no part in its sale or purchase.

In this unit, we focus on social dilemmas in markets, and hence, on market failure.

We will analyse the gains from trade in a case in which the production of a good creates an external cost—pollution. Our example is based on the real-world case of the plantations’ use of the pesticide chlordecone to control the banana weevil in Guadeloupe and Martinique. To simplify the analysis, we are focusing solely on the adverse effects of the pesticide on the fishing industry, and are setting aside the health impact on the fishing community.

A thought experiment

To see why this is called an external effect (you will also see this called an ‘externality’ in some economics books), imagine that the same company owned the banana plantations and fisheries; the company hired fishermen and sold what they caught for profit. The owners of the company would decide on the level of banana pesticide to use, taking account of its downstream effects. They would trade-off the profits from the banana part of their business against the losses from the fisheries.

But this was not the case in Martinique and Guadeloupe, the islands we described in Section 11.1. The plantations owned the profits from banana production, which were increased by using a pesticide. The fisherman ‘owned’ the losses from fishing. The pollution effect of the pesticide was external to the people making the decision on its use. Joint ownership of the plantations and fisheries would have internalized this effect, but the plantations and fisheries were under separate ownership.

Separate ownership and external effects

To model the implications of this kind of external effect, Figure 11.2 shows the marginal costs of growing bananas on an imaginary Caribbean island where a fictional pesticide called Weevokil is used. The marginal cost of producing bananas for the growers is labelled as the marginal private cost (MPC). It slopes upward because the cost of an additional tonne of bananas increases as the land is more intensively used, requiring more Weevokil. By contrast, the marginal private cost curve in the case of the production of Spanish language courses in Unit 7 was flat.

Use the analysis in Figure 11.2 to compare the MPC of producing bananas with the marginal social cost (MSC), which includes the costs borne by fishermen whose waters are contaminated by Weevokil.

Figure 11.2 Marginal costs of banana production using Weevokil.

The marginal private cost

The purple line is the marginal cost for the growers—the marginal private cost (MPC) of banana production. It slopes upward because the cost of producing an additional tonne increases as the land is more intensively used, requiring more Weevokil.

The marginal external cost (MEC)

The orange line shows the marginal cost imposed by the banana growers on fishermen—the marginal external cost. This is the cost of the reduction in quantity and quality of fish caused by each additional tonne of bananas.

The marginal social cost

Adding together the MPC and the MEC, we get the full marginal cost of banana production—the marginal social cost (MSC). This is the green line in the diagram.

The total external cost

The shaded area in the figure shows the total costs imposed on fishermen by plantations using Weevokil. It is the sum of the differences between the marginal social cost and the marginal private cost at each level of production.

A Pareto-inefficient outcome

You can see in Figure 11.2 that the marginal social cost of banana production is higher than the marginal private cost. To focus on the essentials, we will consider a case in which the wholesale market for bananas is competitive, and the market price is $400 per tonne. If the banana plantation owners wish to maximize their profit, we know that they will choose their output so that price is equal to their marginal cost—that is, the marginal private cost. Figure 11.3 shows that their total output is 80,000 tonnes of bananas (point A). Although 80,000 tonnes maximizes profits for banana producers, this does not include the cost imposed on the fishing industry, so it is not a Pareto-efficient outcome.

Figure 11.3 The plantations’ choice of banana output.

To see this, think about what would happen if the plantations were to produce less. The fishermen would benefit but the plantation owners would lose. Therefore, it appears that producing 80,000 tonnes must be Pareto efficient. But let’s imagine that the fishermen could persuade the plantation owners to produce one tonne less.

The fishermen pay the plantation owners to reduce production

If the fishermen paid the plantation owners any amount between just greater than zero and just less than $270, both groups would be better off with 79,999 tonnes of bananas.

marginal external cost (MEC)
The cost of producing an additional unit of a good that is incurred by anyone other than the producer of the good. See also: marginal private cost, marginal social cost.

What about another payment to get the plantation owners to produce 79,998 tonnes instead? You can see that, because the marginal external cost imposed on the fishermen is still much higher than the surplus received by the plantation owners on the next tonne (the difference between the price and the MPC), such a payment would also make both parties better off.

By how much could the fishermen persuade the plantations to reduce production? Look at the point in Figure 11.3 at which the price of bananas is equal to the marginal social cost. At this point, 38,000 tonnes of bananas are produced. If the payments by the fishermen to the plantation owners resulted in them producing just 38,000 tonnes, then the fishermen could no longer benefit by making further payments in return for reduced output. If production were lowered further, the loss to the plantation owners (the difference between price and marginal cost) would be greater than the gain to the fishermen (the difference between private and social cost, shaded). At this point, the maximum payment the fishermen would be willing to make would not be enough to induce the plantations to cut production further. The Pareto-efficient level of banana output is, therefore, 38,000 tonnes.

The possibility of a Pareto-efficient outcome without government intervention

To summarize:

In general, pollutants like Weevokil have negative external effects, sometimes called environmental spillovers. They bring private benefits to those who decide to use them, but by damaging the environment—water resources, in this case—they impose external costs on other firms or on households that rely on environmental resources. For society as a whole, this is a market failure; compared with the Pareto-efficient allocation, the pollutant is overused, and too much of the associated good (bananas, in our example) is produced.

The features of this case of market failure are summarized in Figure 11.4. In the following sections, we summarize other examples of market failure in a similar table. At the end of this unit, we bring all the examples together in Figure 11.15 so that you can compare them.

Decision How it affects others Cost or benefit Market failure (misallocation of resources) Terms applied to this type of market failure
A firm uses a pesticide that runs off into waterways Downstream damage Private benefit, external cost Overuse of pesticide and overproduction of the crop for which it is used Negative external effect, environmental spillover

Figure 11.4 Market failure: Water pollution.

Question 11.2 Choose the correct answer(s)

A factory is situated next to a dormitory for nurses who work night shifts. The factory produces 120 humanoid robots a day. The production process is rather noisy, and the nurses often complain that their sleep is disturbed. Based on this information, which of the following statements are correct?

  • The marginal private cost is the factory’s total cost of producing 120 robots a day.
  • The marginal social cost is the noise cost incurred by the nurses from production of an additional robot.
  • The marginal external cost is the cost to the factory, plus the noise cost incurred by the nurses, when an additional robot is produced.
  • The total external cost is the total costs per day imposed on the nurses by the factory’s production.
  • The MPC is the factory’s cost of producing one additional robot, not the total cost.
  • The MSC is the sum of the costs to the factory and the nurses of producing the additional robot, MSC = MPC + MEC.
  • The MEC is the noise cost incurred by the nurses when an additional robot is produced.
  • The total external cost is the sum of the marginal external costs of production of all the robots. This is the total cost imposed on the nurses.

11.8 External effects and private bargaining

To demonstrate that the market allocation of bananas (producing 80,000 tonnes, using Weevokil) is not Pareto efficient, we showed that the fishermen could pay the plantation owners to produce fewer bananas, and both would be better off.

Does this suggest a remedy for this market failure that might be implemented in the real world?

More real-world cases of external effects and remedies can be found in Unit 20 of The Economy. See, for example, Section 20.3 on cost-benefit analysis of climate change abatement policies and Section 20.5 on cap and trade environmental policies. Unit 21 of The Economy, Section 21.7 covers the design of patent policy.

It does. The fishermen and the plantation owners could negotiate a private bargain. Solutions of this type are often called Coasean bargaining, after Ronald Coase who pioneered the idea that private bargaining might be preferable to dealing with external effects by governmental intervention. He argued that the two parties to the exchange often have more of the inform­ation necessary to implement an efficient outcome than does the government.

Great economists Ronald Coase

Ronald Coase You have already met Ronald Coase (1910–2013). He was featured in Unit 6 for his representation of the firm as a political organization. He is also known for his idea that private bargaining could address market failures, in some cases doing this more effectively than government policies.

He explained that, when one party is engaged in an activity that has the incidental effect of causing damage to another, a negotiated settlement between the two may result in a Pareto-efficient allocation of resources. He used the 1879 legal case of Sturges v Bridgman in the UK to illustrate his argument. The case concerned Bridgman, a confectioner (candy-maker) who for many years had been using machinery that generated noise and vibration. This caused no external effects until his neighbour Sturges built a consulting room on the boundary of his property, close to the confectioner’s kitchen. The courts granted the doctor an injunction that prevented Bridgman from using his machinery.

Coase pointed out that, once the doctor’s right to prevent the use of the machinery had been established, the two sides could modify the outcome. The doctor would be willing to waive his right to stop the noise in return for a compensation payment. And the confectioner would be willing to pay if the value of his annoying activities exceeded the costs that they imposed on the doctor.

Also, the court’s decision in favour of Sturges rather than Bridgman would make no difference to whether Bridgman continued to use his machinery. If the confectioner had been granted the right to use it, the doctor could have paid him to stop. But he would have been willing to do this if, and only if, the costs to him were greater than the confectioner’s profits gained by using the machinery.

In other words, private bargaining would ensure that the machinery was used if, and only if, its use, along with a compensation payment, made both better off. Private bargaining would ensure Pareto efficiency. Bargaining gives the confectioner an incentive to take into account not only the marginal private costs of using the machine to produce candy, but also the external costs imposed on the doctor. That is, the confectioner takes account of the entire social cost. To the confectioner, the cost of using the annoying machinery during the doctor’s visiting hours would now send the right message. Private bargaining could be a substitute for legal liability. It ensures that those harmed are compensated, and that those who could inflict harm would make efforts to avoid harmful behaviour.

Whether the courts decided in favour of Sturges (the doctor) or Bridgman (the confectioner) made no difference from the standpoint of Pareto efficiency. As long as the court clearly established who had the right to do what, so that the two could bargain, the result would be efficient. But the legal decision did matter for the distribution of income between the two. Because the court decided in favour of Sturges, Bridgeman would have to pay Sturges for the right to use the machinery. Had it gone the other way, Sturges might have paid Bridgeman to stop using the machinery.

To summarize:

  • The court establishes the initial property rights: In this case, Bridgman’s right to make a noise or Sturges’ right to quiet.
  • This leads to a Pareto-efficient outcome: As long as private bargaining exhausts all the potential mutual gains, the result would (by definition) be Pareto efficient, regardless of which party owned the initial rights.
  • Is this fair? We might object that the court’s decision resulted in an unfair distribution of profits, but however one evaluates this concern (or if, like Coase, one puts ‘questions of equity aside’), the outcome would be Pareto efficient.
transaction costs
Costs that impede the bargaining process or the agreement of a contract. They include costs of acquiring information about the good to be traded, and costs of enforcing a contract.

Coase emphasized that his model could not be directly applied to most situations because of the costs of bargaining and other impediments that prevent the parties from exploiting all possible mutual gains. Costs of bargaining, sometimes called transaction costs, may prevent Pareto efficiency. If the confectioner cannot find out how badly the noise affects the doctor, the doctor has an incentive to overstate the costs to get a better deal. Establishing each party’s actual costs and benefits is part of the cost of the transaction, and this cost might be too high to make a bargain possible.

However difficult it might be for the doctor and the confectioner to acquire this information, it is often even more difficult for a government to gather the information necessary to directly impose an efficient and fair solution, beyond simply clarifying the property rights in question.

Coase’s analysis suggests that a lack of clear property rights, and other impediments leading to high transaction costs, may stand in the way of using bargaining to resolve external effects. But with a clear legal framework in which one side initially owned the rights to produce (or to prevent production of) the external effect, there might be no need for government policies to address the market failure.

Until now you have probably thought about property rights as referring to goods and services that are typically bought and sold in markets, like food, flights, or houses. Coase’s approach suggests that we could think of other rights—in his example, the right to make a noise or to have a quiet work environment—as goods that can be bargained over and traded in return for money.

Could private bargaining solve the pesticide problem?

reservation option
A person’s next best alternative among all options in a particular transaction. Also known as: fallback option. See also: reservation price.

Let’s see how a private bargain might solve the pesticide problem. Initially, it is not illegal to use Weevokil; the allocation of property rights is such that the plantation owners have the right to use it, and choose to produce 80,000 tonnes of bananas. This allocation and the associated incomes and environmental effects represent the reservation options of the plantation owners and fishermen. This is what they will get if they do not come to some agreement.

For the fishermen and the plantation owners to negotiate effectively, they would each have to be organized so that a single person (or body) could make agreements on behalf of the entire group. Let’s imagine that a representative of an association of fishermen sits down to bargain with a representative of an association of banana growers. To keep things simple, we assume that there are no feasible alternatives to Weevokil, so they bargain only over the output of bananas.

Both sides should recognize that they could gain from an agreement to reduce output to the Pareto-efficient level. In Figure 11.5, the situation before bargaining begins is point A, and the Pareto-efficient quantity is 38,000 tonnes. The total shaded area shows the gain for the fishermen (from cleaner water) if output is reduced from 80,000 to 38,000. But reducing banana production leads to lower profits for the plantation owners. Use the analysis in Figure 11.5 to see that the fall in profit is smaller than the gain for the fishermen, so there is a net social gain that they could agree to share.

Figure 11.5 The gains from bargaining.

The status quo at point A

The situation before bargaining is represented by point A, and the Pareto-efficient quantity of bananas is 38,000 tonnes. The total shaded area shows the gain for fishermen if output is reduced from 80,000 to 38,000 (that is, the reduction in the fishermen’s costs).

Lost profit

Reducing output from 80,000 to 38,000 tonnes reduces the profits of plantations. The lost profit is equal to the loss of producer surplus, shown by the blue area.

The net social gain

The net social gain is the gain for the fishermen minus the loss for the plantation owners, shown by the remaining green area.

Since the gain to the fishermen would be greater than the loss to the plantation owners, the fishermen would be willing to pay the banana growers to reduce output to 38,000 tonnes if they had the funds to do so.

minimum acceptable offer
In the ultimatum game, the smallest offer by the Proposer that will not be rejected by the Responder. Generally applied in bargaining situations to mean the least favourable offer that would be accepted.

The minimum acceptable offer from the fishermen depends on what the plantations get in the existing situation, which is their reservation profit (shown by the blue area labelled ‘loss of profit’). If plantation owners agreed to this minimum payment to compensate them for their loss of profit, the fishing industry would achieve a net gain equal to the net social gain, while plantations would be no better (and no worse) off.

The maximum the fishing industry would pay is determined by their reservation option (also known as their fallback option), as in the case of the plantation owners. It is the sum of the blue and green areas. In this case, the plantation owners would get all the net social gain, while the fishermen would be no better off. As in the examples of bargaining in Unit 5, the compensation the plantation owners and fishermen agree on between these maximum and minimum levels is determined by the bargaining power of the two groups.

You may think it unfair that the fishermen need to pay for a reduction in pollution. At the Pareto-efficient level of banana production, the fishing industry is still suffering from pollution (shown by the fact that the MSC is above the MPC), and it must pay to stop the pollution getting worse. This happens because we have assumed that the plantation owners have a legal right to use Weevokil.

An alternative legal framework could give the fishermen a right to clean water. If that were the case, the plantation owners wishing to use Weevokil could propose a bargain in which they paid the fishermen to give up some of their right to clean water to allow the Pareto-efficient level of banana production. This would be a much more favourable outcome for the fishermen. In principle, the bargaining process would result in a Pareto-efficient allocation independently of whether the initial rights were granted to the plantations (right to pollute) or to the fishermen (right to unpolluted water). But the two cases differ dramatically in who gains and who loses when the market failure is solved. This limitation, among others pointed out in a UMassEconomics video and discussed below, make Coase’s proposal difficult to implement in practice.

Why private bargains may not work

As Coase acknowledged, practical obstacles to bargaining may prevent the achievement of Pareto efficiency:

The pesticide example illustrates that, although correcting market failures through bargaining may not require direct government intervention, it does require a legal framework for enforcing contracts to ensure that all parties stick to the bargains they make. Even with this framework, the problems of collective action, missing information, and enforcement of what are inevitably complex contracts make it unlikely that Coasean bargaining alone can address market failures.

The polluter pays principle

According to the Rio Declaration on Environment and Development, issued in 1992 by the United Nations:

National authorities should endeavour to promote the internalization of environmental costs and the use of economic instruments, taking into account the approach that the polluter should, in principle, bear the cost of pollution, with due regard to the public interest and without distorting international trade and investment.

polluter pays principle
A guide to environmental policy according to which those who impose negative environmental effects on others should be made to pay for the damages they impose, through taxation or other means. See also: external cost.

Several of the approaches we describe in this unit are consistent with this principle, called the polluter pays principle. Either option—giving the fisherman a right to clean water or enforcing compensation—means that the plantations will have to pay at least as much as the costs incurred by the fishing industry. A tax also means that the polluter pays, although the payment goes to the government rather than the fishing industry. The same abatement could be accomplished by providing the plantation owners with a subsidy for the use of an alternative technology that resulted in a lower level of pollution.

The firm’s view of these two policies may be that the tax is the ‘stick’ and the subsidy the ‘carrot’. The tax, which reflects the polluter pays principle, lowers the profits of the firm. A subsidy raises the firm’s profits. Whether the carrot or the stick is the right policy depends on the feasibility and cost of implementing the subsidy compared to the tax, and whether raising or lowering the income of the target of the policy is desired on fairness grounds.

Seen in this light, the polluter pays principle is not always a good guide to the best policy. Think of a large city in a low-income country in which much of the cooking is still done over wood fires, generating high levels of airborne particulate matter and causing asthma and other respiratory illnesses:

Exercise 11.2 Bargaining power

In the example of plantation owners and fishermen, explain some factors that might affect the bargaining power of these parties.

marginal private benefit (MPB)
The benefit (in terms of profit, or utility) of producing or consuming an additional unit of a good for the individual who decides to produce or consume it, not taking into account any benefit received by others. 
marginal social benefit (MSB)
The benefit (in terms of utility) of producing or consuming an additional unit of a good, taking into account both the benefit to the individual who decides to produce or consume it, and the benefit to anyone else affected by the decision.

Exercise 11.3 A positive external effect

Imagine a beekeeper, who produces honey and sells it at a constant price per kilogram.

  1. Draw a diagram with the quantity of honey on the horizontal axis, showing the marginal cost of honey production as an upward-sloping line, and the price of honey as a horizontal line. Show the amount of honey that the profit-maximizing beekeeper produces.
  2. For the beekeeper, the marginal private benefit (MPB) of producing a kilogram of honey is equal to the price. But since the bees benefit a neighbouring farmer by helping to pollinate her crops, honey production has a positive external effect. Draw a line on your diagram to represent the marginal social benefit (MSB) of honey production. Show the quantity of honey that would be Pareto efficient. How does it compare with the quantity chosen by the beekeeper?
  3. Explain how the farmer and beekeeper could both be made better off through bargaining.

Question 11.3 Choose the correct answer(s)

The graph depicts the MPC and MSC of production by the robot factory introduced in Question 11.2.

Figure 11.6 Robot factory production.

The robot market is competitive and the market price is $340. Currently, the factory is producing an output of 120, but 80 would be Pareto efficient. Which of the following statements are correct?

  • To reduce output to 80, the factory’s minimum acceptable payment would be $1,600.
  • The maximum that the nurses are willing to pay to induce the factory to reduce the output to 80 is $2,400.
  • The factory would not reduce its output to 80 unless it received at least $4,000.
  • The net social gain from the output reduction to 80 depends on the amount paid by the nurses to the factory.
  • The minimum acceptable payment is the loss of surplus, which is the green shaded area = 0.5 × (340 − 260) × (120 − 80) = $1,600.
  • The maximum that the nurses would pay is the total gain (of reduction in noise cost) associated with the fall in output. This is the sum of the green triangle and the pink triangle, which is [0.5 × (340 − 260) × (120 − 80)] + [0.5 × (460 − 340) × (120 − 80)] = $1,600 + $2,400 = $4,000.
  • The maximum the nurses would pay is $4,000 (the whole shaded area). The minimum acceptable payment is the green shaded area.
  • The net social gain is the difference between the nurses’ reduction in the noise cost and the loss of profit for the factory. This is the purple shaded area. It is not affected by the payment, which only determines the distribution of the net social gain.

Question 11.4 Choose the correct answer(s)

Consider the situation in which the noise of a factory’s production affects nurses in the dormitory next door. If there are no transaction costs to impede Coasean bargaining, which of the following statements are correct?

  • Whether the final output level is Pareto efficient depends on who has the initial property rights.
  • The nurses would be better off in the bargained allocation if they initially had a right to undisturbed sleep than they would if the factory has the right to make noise.
  • If the factory has the right to make noise, it will prefer not to bargain with the nurses.
  • If the nurses have the initial rights, they will obtain all the net social gain from robot production.
  • The final allocation is Pareto efficient, regardless of whether the factory has the right to make noise or the nurses have the right to undisturbed sleep. This is the main result of Coasean bargaining.
  • If the factory has the initial rights, the nurses will have to pay for a reduction in noise. If the nurses have the initial rights, they will gain both a noise reduction and a payment from the factory.
  • The factory can be made better off by obtaining a payment from the nurses in return for reducing output.
  • This is the maximum that the nurses can obtain, but whether they will do so depends on their bargaining power.

11.9 External effects: Government policies and income distribution

Suppose in the case of our Weevokil example that Coasean bargaining proves to be impractical, and that the fisherman and plantation owners cannot resolve the Weevokil problem privately. We continue to assume that it is not possible to grow bananas without using Weevokil.

What can the government do to achieve a reduction in the output of bananas to the level that takes into account the costs for the fishermen? There are three ways this might be done:

Each of these policies has different distributional implications for the fisherman and plantation owners.


The government could cap total banana output at 38,000 tonnes, the Pareto-efficient amount. This looks like a straightforward solution. On the other hand, if the plantations differ in size and output, it may be difficult to determine and enforce the right quota for each one.

This policy would reduce the costs of pollution for the fishermen, but it would lower the plantation owners’ profits. They would lose their surplus on each tonne of bananas between 38,000 and 80,000.

The distributional effect of this policy is to shift income from the plantations to the fishermen.


Figure 11.7 shows the MPC and MSC curves again. At the Pareto-efficient quantity (38,000 tonnes), the MSC is $400 and the MPC is $295. The price is $400. If the government puts a tax on each tonne of bananas produced, equal to $400 – $295 = $105 (the marginal external cost), then the after-tax price received by plantations is $295. Now, if plantation owners maximize their profit, they will choose the point where the after-tax price equals the marginal private cost and produce 38,000 tonnes, the Pareto-efficient quantity. Use the analysis in Figure 11.7 to see how this policy works.

Figure 11.7 Using a tax to achieve Pareto efficiency.

The marginal external cost

At the Pareto-efficient quantity, 38,000 tonnes, the MPC is $295. The MSC is $400. Therefore, the marginal external cost is MSC – MPC = $105.

Tax = MSC − MPC

If the government puts a tax on each tonne of bananas produced equal to $105, the marginal external cost, then the after-tax price received by plantations is $295.

The after-tax price is $295

To maximize profit, the plantation owners will choose their output so that the MPC is equal to the after-tax price. They will choose point P1 and produce 38,000 tonnes.

Pigouvian tax
A tax levied on activities that generate negative external effects so as to correct an inefficient market outcome. See also: external effect, Pigouvian subsidy.
external benefit
A positive external effect: that is, a positive effect of a production, consumption, or other economic decision on another person or people that is not specified as a benefit in a contract. Also known as: external economy. See also: external effect.

The tax corrects the message conveyed by the price of bananas, so that the plantations face the full marginal social cost of their decisions and choose to produce less. When the plantations are producing 38,000 tonnes of bananas, the tax is exactly equal to the cost imposed on the fishermen. This approach is known as a Pigouvian tax, after the economist Arthur Pigou who advocated it. It also works in the case of a positive external effect; the marginal social benefit of a decision is greater than the marginal private benefit (MPB), this becomes a Pigouvian subsidy, which can ensure that the decision-maker takes this external benefit into account.

The distributional effects of taxation are different from those of regulation. The costs of pollution for fishermen are reduced by the same amount, but the reduction in banana profits is greater, since the plantation owners pay taxes as well as reducing output, and the government receives tax revenue.


The government could require the plantation owners to pay compensation for costs imposed on the fishermen. The compensation required for each tonne of bananas is equal to the difference between the MSC and the MPC, which is the distance between the green and purple lines in Figure 11.8. Once compensation is included, the marginal cost of each tonne of bananas is the MPC plus the compensation, which is equal to the MSC. Now the plantation owners will maximize profit by choosing point P₂ in Figure 11.8 and produce 38,000 tonnes. The shaded area shows the total compensation paid. The fishermen are fully compensated for pollution, and the plantation owners’ profits are equal to the true social surplus of banana production.

Figure 11.8 The plantation owners compensate the fishermen.

The effect of this policy on the plantation owners’ profits is similar to the effect of the tax, but the fishermen do better because they, rather than the government, receive payment from the plantations.

Diagnosis and treatment in the case of chlordecone

When we identified 38,000 tonnes as the Pareto-efficient level of output in our model, we assumed that growing bananas inevitably involves Weevokil pollution. Our diagnosis was that too many bananas were being produced, and we looked at policies for reducing production.

The correct diagnosis

In real life, in Guadeloupe and Martinique, there were alternatives to chlordecone. Therefore the problem was caused by the use of chlordecone, not the production of bananas.

The market failure occurred because the price of chlordecone did not incorporate the costs that its use inflicted on the fishermen, and so it sent the wrong message to the firm. Its low price said: ‘Use this chemical, it will save you money and raise profits. However, if its price had included the full external costs of its use, it might have been high enough to have said: ‘Think about the downstream damage and look for an alternative way to grow bananas.’

The best treatment

In this situation, a policy of requiring the plantation owners to compensate the fishermen would have given them the incentive to find production methods that caused less pollution and could, in principle, have achieved an efficient outcome.

But the other two policies would not do so. Rather than taxing or regulating banana production, it would be better to regulate or tax the sale or the use of chlordecone, to motivate plantations to find the best alternative to intensive chlordecone use.

In theory, if the tax on a unit of chlordecone was equal to its marginal external cost, the price of chlordecone for the plantations would be equal to its marginal social cost, which would send the right message about the choice of pest control method. The plantation owners could then choose the best production method, taking into account the high cost of chlordecone. This would involve reducing the use of chlordecone or switching to a different pesticide, and would determine their profit-maximizing output. As with the banana tax, the profits of the plantation owners and the pollution costs for the fishermen would fall, but the outcome would be better for the plantations, and possibly the fishermen also, if chlordecone were taxed instead of bananas.

What actually happened?

Unfortunately, none of these remedies was used for 20 years, and the people of Guadeloupe and Martinique are still living with the consequences.

In 1993, the government finally recognized that the marginal social cost of chlordecone use was so high that it should be banned altogether.

Chlordecone was first listed as carcinogenic in 1979. It was obvious that the external costs were much higher than in our case of Weevokil, damaging the health of islanders as well as the livelihood of fishermen. In fact, the marginal social cost of any bananas produced with the aid of chlordecone was higher than their market price, justifying an outright ban on its use.

The pollution turned out to be much worse than anyone realized at the time, and is likely to persist in the soil for 700 years. In 2013, fishermen in Martinique barricaded the port of Fort de France with their boats until the French government agreed to allocate $2.6 million in aid.

Limits to the success of tax, regulation, and compensation remedies

There are limits to how well governments can implement Pigouvian taxes, regulation and compensation—often for the same reasons as for Coasean bargaining:

Great economists Arthur Pigou

Arthur Pigou Arthur Pigou (1877–1959) was a pioneer in using economics for the good of society, which is why he is sometimes seen as the founder of welfare economics. He won awards in history, languages, and moral sciences (there was no dedicated economics degree at the time) during his studies at the University of Cambridge. He became a protégé of Alfred Marshall. Pigou was an outgoing and lively person when young, but his experiences as a conscientious objector and ambulance driver during the First World War, as well as anxieties over his own health, turned him into a recluse who hid in his office except for lectures and walks.

His book Wealth and Welfare9 was described by Joseph Schumpeter as ‘the greatest venture in labour economics ever undertaken by a man who was primarily a theorist’, and provided the foundation for his later work, The Economics of Welfare.10 Together, these works built up a relationship between a nation’s economy and the welfare of its people. Pigou focused on happiness and wellbeing. He recognized that concepts such as political freedom and relative status were important.

Pigou believed that the reallocation of resources was necessary when the interests of a private firm or individual diverged from the interests of society, causing what we would today call external effects. He suggested that taxation could solve the problem; Pigouvian taxes are intended to ensure that producers face the true social costs of their decisions.

Pigouvian taxes were largely unrecognized until the 1960s, but they have become a major policy tool for reducing pollution and environmental damage.

Now we can extend the table we started creating in Section 11.7 (Figure 11.4). Look at the fifth column in Figure 11.9—it adds the possible remedies for the problem of negative external effects.

Decision How it affects others Cost or benefit Market failure (misallocation of resources) Possible remedies Terms applied to this type of market failure
A firm uses a pesticide that runs off into waterways Downstream damage Private benefit, external cost Overuse of pesticide and overproduction of the crop for which it is used Taxes, quotas, bans, bargaining, common ownership of all affected assets Negative external effect, environmental spillover

Figure 11.9 Water pollution market failure, with remedies.

Exercise 11.4 Pigouvian subsidy

Consider the beekeeper and neighbouring farmer in Exercise 11.3.

  1. Why might they be unable to bargain successfully to achieve a Pareto-efficient outcome in practice? Use the diagram you drew to show how the government might improve the situation by subsidizing honey production.
  2. Describe the distributional effects of this subsidy, and compare it to the Pareto-efficient bargaining outcome.

Exercise 11.5 Comparing policies

Consider the three policies of regulation, taxation, and compensation arrangements discussed above. Evaluate the strengths and weaknesses of each policy from the standpoint of Pareto efficiency and fairness.

Question 11.5 Choose the correct answer(s)

Look back at Figure 11.6, which showed the MPC and MSC of robot production for the factory situated next to a dormitory for nurses who work nightshifts.

The market for robots is competitive and the market price is $340. The initial output is 120, but the government uses a Pigouvian tax to reduce this to the efficient level of 80. Which of the following statements are correct?

  • Under the Pigouvian tax, the factory’s surplus is $6,400.
  • The required Pigouvian tax is $120 per robot.
  • The nurses are at least as well off as they would be under Coasean bargaining.
  • The nurses obtain no benefit from the imposition of the Pigouvian tax.
  • The Pigouvian tax lowers the after-tax price to $260. The factory’s surplus is the area above the MPC line and below the lowest horizontal line for a price of $260 = 0.5 × 80 × (260 − 100) = $6,400.
  • The Pigouvian tax must lower the after-tax price from $340 to $260, so it is $80.
  • Under a Coasean bargain that involved a payment from the factory to the nurses, as well as a reduction in noise, the nurses would be better off.
  • The nurses do not receive a payment, but they benefit from the noise reduction.

11.10 Property rights, contracts, and market failures

In taking an action to maximize profits (choosing the level of banana production or the choice of pesticide), the plantation owners did not take account of the external costs they imposed on the fishermen. And they had no reason to take account of them—they had the right to pollute the fisheries and it cost them nothing to do so.

The same is true for the overuse of antibiotics analysed in Unit 2. A self-interested person has no reason to use antibiotics sparingly, because the superbug that may be created will probably infect someone else.

If the prices of chlordecone and the antibiotic were high enough, there would be no overuse. But the prices of these goods were based on the costs of production and excluded costs that their use would inflict on others. As you have seen, the private cost to the user (how much he paid to acquire the good) fell short of the social cost for this reason.

Another example: when automobile fuel costs are low, more people decide to drive to work rather than taking the train. The message conveyed by the low price (‘It’s not very costly to use your car’) does not include the environmental costs of deciding to drive. The effects on the decision-maker are termed private costs and benefits, while the total effects, including those inflicted or enjoyed by others, are social costs and benefits.

external diseconomy
A negative effect of a production, consumption, or other economic decision, that is not specified as a liability in a contract. Also known as: external cost, negative externality. See also: external effect.
external economy
A positive effect of a production, consumption, or other economic decision, that is not specified as a benefit in a contract. Also known as: external benefit, positive externality. See also: external effect.

Costs inflicted on others (such as pollution and congestion that are worse because you drive to work) are termed external diseconomies or negative external effects, while uncompensated benefits conferred on others are external economies or positive external effects.

We can understand why market failures are common by thinking about how they could be avoided.

How could the cost of driving to work accurately reflect all the costs incurred by anyone, not just the private costs made by the decision-maker? The most obvious (if impractical) way would be to require the driver to pay everyone affected by the resultant environmental damage (or traffic congestion) an amount exactly equal to the damage inflicted. This is, of course, impossible to do, but it sets a standard of what should be done or approximated if the ‘price of driving to work’ is to send the correct message to the decision-maker.

Something like this approach applies if you drive recklessly on the way to work, skid off the road, and crash into somebody’s house. Tort law (the law of damages) in most countries would require you to pay for repairing the damage to the house. You are held liable for the damages so that you would pay the cost you had inflicted on another. Knowing this, you might think twice about driving to work (or at least slow down a bit when you are late). It will change your behaviour and the allocation of resources.

But while tort law in most countries covers some kinds of harm inflicted on others (reckless driving), other important external effects of driving your car (such as adding to air pollution or congestion) are not necessarily covered. Here are two further examples:

Legal systems also fail to provide compensation for the benefits that one’s actions confer on others:

External effects arise because of incomplete contracts and missing markets

incomplete contract
A contract that does not specify, in an enforceable way, every aspect of the exchange that affects the interests of parties to the exchange (or of any others affected by the exchange).

Market failures occur in these examples because the external benefits and costs of a person’s actions are not owned by anyone. Think about waste. If you redecorate your house and you tear up the floor or knock down a wall, you own the debris and have to dispose of it, even if you need to pay someone to take it away. But this is not the case with fumes from the incinerator or loud music at night. You do not have a contract with the incinerator company specifying at what price you are willing to accept fumes, or with your neighbour about the price of the right to play music after 10 p.m. In these cases, economists say that we have ‘incomplete, missing, or unenforceable property rights’—or, simply, incomplete contracts.

missing market
A market in which there is some kind of exchange that, if implemented, would be mutually beneficial. This does not occur due to asymmetric or non-verifiable information.

We saw an important example of an incomplete contract in Unit 6. In the employment relationship, the employer can pay for the worker’s time, but the contract cannot specify how much effort the worker must make. Likewise, the external effects of a person’s actions are effects that are not governed by contracts. Another way to express the problem is to say that there is no market within which these external effects can be compensated. Economists also use the term missing markets to describe problems like this.

In the case of Weevokil pollution:

Why don’t countries just rewrite their laws to reward people for the benefits they confer on others, and make economic actors pay for the costs they inflict on others?

verifiable information
Information that can be used to enforce a contract.

In Unit 6, we reviewed the reasons why the kinds of contracts that would enforce these objectives are incomplete or unenforceable. These are that:

You can see in our example that it would not be possible to write a complete set of contracts in which each individual fisherman could receive compens­ation from each plantation for the effects of its individual decisions.

For these and other reasons, it is impractical in most cases to use tort law to make people liable for the costs they inflict on others. And it is equally infeasible to use the legal system to compensate people for the beneficial effects they have on others, for example, to pay those who keep beautiful gardens an amount equal to the pleasure this confers on those who pass their house. A court would have to know how much that pleasure was worth to each passerby.

In each of the examples in this section (incinerator, loud music, training, irrigation, and climate change), uncompensated external costs and benefits occur for the same reason, as shown in Figure 11.10.

Information that is of concern to someone other than the decision-maker is non-verifiable or asymmetric
There can be no contract or property rights ensuring that external effects are compensated.
Some of the social costs or benefits of the decision-maker’s actions are not included (or are not sufficiently important) in the decision-making process.

Figure 11.10 How external costs and benefits occur.

Exercise 11.6 Incomplete contracts

In each of the five cases above (incinerator, loud music, training, irrigation, and climate change):

  1. Explain why the external effects are not (and possibly cannot be) covered by a complete contract.
  2. What critical piece(s) of information required for a complete contract are asymmetric or non-verifiable?

11.11 Public goods, common pool resources, and market failure

Public goods and non-rivalry

An extreme form of external effect occurs in the case of public goods. Recall from the experiments described in Unit 2 that the defining character­istic of a public good is that, if it is available to one person, it can be available to everyone at no additional cost. An example is a view of the setting sun; your enjoyment of the sunset does not deprive anyone else of their enjoy­ment. Another is weather forecasting (if I can go online to find out if it’s likely to rain today, so can you and everybody else, at no additional cost).

The knowledge assembled in CORE’s Economy, Society, and Public Policy (and other works) is a public good: access to it by another user does not diminish the availability of it to other users. And because we wanted it to be available to everyone at its marginal cost (which is zero) we have made the material open access online using a Creative Commons licence.

A weather forecast is a public good, or the online version of the book you are reading, but so is knowledge more generally. You can use your knowledge of a recipe for a cake or the rules of multiplication without affecting the ability of others to use the same knowledge.

non-rival good
A good that, if available to anyone, is available to everyone at no additional cost. See also: rival good, non-excludable public good.
public bad
The negative equivalent of a public good. It is non-rival in the sense that a given individual’s consumption of the public bad does not diminish others’ consumption of it.
asymmetric information
Information that is relevant to all the parties in an economic interaction, but is known by some and not by others. See also: adverse selection, moral hazard.

In these cases, once the good is available at all, the marginal cost of making it available to additional people is zero. Goods with this characteristic are also called non-rival goods because potential users are not in competition with each other (rivals) for the good.

Of course, this is a good thing, but it results in a market failure. Think about a firm considering investing in research. If other firms can freely access and use the knowledge that the research of another firm produces, then the firm’s research is providing an external benefit to all other firms. In this case the benefits of the research (including the positive external effects) may far exceed the costs to the particular firm. But because the firm does not capture these benefits (as profits) it will invest less in research than would make sense if all of the benefits were considered.

Sometimes economists refer to ‘public bads’, which are just a negative kind of public good: something that people dislike, which, if it is experi­enced by anyone it is ‘available’ to all. Exposure to an adverse climate change falls under this heading.

Artificial scarcity: Excluding users from a non-rival good

Ownership rights over the use and distribution of an original work.
artificially scarce good
A public good for which it is possible to exclude some people from enjoying. Also known as: club good. See also: public good.
common-pool resource
A rival good that one cannot prevent others from enjoying. Also known as: common property resource. See also: rival good.

For some public goods, it is possible to exclude additional users, even though the additional cost of using it is zero. Examples are satellite TV, the information in a copyrighted book, or a film shown in an uncrowded cinema; it costs no more if an additional viewer is there, but the owner can nonetheless require that anyone who wants to see the film must pay. The same goes for an uncrowded road on which toll gates have been erected. Drivers can be excluded (unless they pay the toll) even though the marginal cost of an additional traveller is zero.

Public goods for which it is feasible to exclude others are sometimes called artificially scarce goods or club goods, because they function like joining a private club—when the golf course is not crowded, adding one more member costs the golf club nothing, but the club will still charge a membership fee. (Some economists exclude club goods from the category of public goods, defining a public good as not only non-rival, but also non-excludable).

Common-pool resources: Rival but non-excludable goods

The opposite of artificially scarce or club goods (that are non-rival but excludable) are common pool resources which are rival, but not excludable. An example is fisheries that are open to all. What one fisherman catches cannot be caught by anyone else, but anyone who wants to fish can do so. We can also think of busy public roads as a common-pool resource. Anyone who chooses to use them may do so, but each user makes the road more congested and slows down the journeys of others.

Non-rivalry, non-excludability and market failures

Figure 11.11 shows four distinct categories of goods, including private goods. These are both rival and excludable, things like portions of a cake, plots of land, flights to your dream vacation spot, and many of the products we have already studied like one-on-one language tutorials, and Cheerios.

Both the extent of rivalry or excludability in goods is a matter of degree. For some kinds of goods, the cost of additional users is not literally zero (which is what pure non-rivalry would require), but instead very small. An example is a medical drug that cost millions in research funds to create the first pill, but only cents per application to make treatments available to additional users once created. The knowledge that makes the drug possible is the major cost, and it is a public good; the pill in which the knowledge is embodied and delivered to the patient is both rival and excludable, and hence private.

  Rival Non-rival
Excludable Private goods (food, clothes, houses, most forms of medical care) Public goods that are artificially scarce (subscription TV, uncongested toll roads, knowledge subject to intellectual property rights)
Non-excludable Common-pool resources (fish stocks in a lake, common grazing land) Non-excludable public goods and bads (view of a lunar eclipse, public broadcasts, rules of arithmetic or calculus, national defence, noise and air pollution)

Figure 11.11 Private goods and public goods.

As can be seen from the examples, whether a good is private or public depends not only on the nature of the good itself, but also on legal and other institutions:

As you can see from the example in Figure 11.11, governments play an important role in the provision of public goods. Some important private goods, too, are produced and allocated by governments, for example, in many countries, medical care.

Most private goods are allocated in markets. Where external effects are absent or minor, the results may be Pareto efficient.

But for the other three kinds of good, markets are either not possible or very likely to fail. There are two reasons:

Market failure in the case of public goods is closely related to the problems discussed earlier in this unit—external effects, absent property rights, and incomplete contracts. We analysed Weevokil pollution as a problem in which the decisions of banana plantation owners imposed a negative external effect on fisherman. The private cost of using Weevokil was below the social cost, so the pesticide was overused. But we can also interpret the plantations as contributing to a public bad, suffered by all the fishermen.

The user of a common-pool resource imposes an external cost on other users. By driving your car on a busy road, for example, you contribute to the congestion experienced by other drivers.

Thus, any of the examples of non-private goods introduced in this section can be described using the framework we set up in Section 11.9 to summarize cases of market failure. See the table in Figure 11.12.

Decision How it affects others Cost or benefit Market failure (misallocation of resources) Possible policy remedies Terms applied to this type of market failure
You take an international flight Increase in global carbon emissions Private benefit, external cost Overuse of air travel Taxes, quotas Public bad, negative external effect
You travel to work by car Congestion for other road users Private cost, external cost Overuse of cars Tolls, quotas, subsidized public transport, tax on petrol Common-pool resource, negative external effect
A firm invests in R&D Other firms can exploit the innovation Private cost, external benefit Too little R&D Publicly funded research, subsidies for R&D, patents Public good, positive external effect

Figure 11.12 Examples of market failure, with remedies.

Exercise 11.7 Rivalry and excludability

For each of the following goods or bads, decide whether they are rival and whether they are excludable, and explain your answer. If you think the answer depends on factors not specified here, explain how.

  1. a free public lecture held at a university lecture theatre
  2. noise produced by aircraft around an international airport
  3. a public park
  4. a forest used by local people to collect firewood
  5. seats in a theatre to watch a musical
  6. this e-text, Economy, Society, and Public Policy
  7. bicycles available to the public to hire to travel around a city.

Question 11.6 Choose the correct answer(s)

Which of the following statements are correct?

  • Some public goods are rival.
  • A public good must be non-excludable.
  • A good cannot be rival and non-excludable.
  • If a good is non-rival, then the cost of an additional person consuming it is zero.
  • All public goods are non-rival by definition.
  • For some public goods it is possible to exclude additional users even though the cost of use is zero, for example, subscription TV. Such goods are called artificially scarce.
  • For example, common grazing land is rival but non-excludable. Such goods are called common-pool resources.
  • A good is non-rival if its use by one person does not reduce its availability to others, so it can be made available to another person without cost.

11.12 Missing markets: Insurance and lemons

asymmetric information
Information that is relevant to all the parties in an economic interaction, but is known by some and not by others. See also: adverse selection, moral hazard.
asymmetric information
Information that is relevant to all the parties in an economic interaction, but is known by some and not by others. See also: adverse selection, moral hazard.
hidden actions (problem of)
This occurs when some action taken by one party to an exchange is not known or cannot be verified by the other. For example, the employer cannot know (or cannot verify) how hard the worker she has employed is actually working. Also known as: moral hazard. See also: hidden attributes (problem of).
moral hazard
This term originated in the insurance industry to express the problem that insurers face, namely, the person with home insurance may take less care to avoid fires or other damages to his home, thereby increasing the risk above what it would be in absence of insurance. This term now refers to any situation in which one party to an interaction is deciding on an action that affects the profits or wellbeing of the other but which the affected party cannot control by means of a contract, often because the affected party does not have adequate information on the action. It is also referred to as the ‘hidden actions’ problem. See also: hidden actions (problem of), incomplete contract, too big to fail.
hidden attributes (problem of)
This occurs when some attribute of the person engaging in an exchange (or the product or service being provided) is not known to the other parties. Example: an individual purchasing health insurance knows her own health status, but the insurance company does not. Also known as: adverse selection. See also: hidden actions (problem of).
adverse selection
The problem faced by parties to an exchange in which the terms offered by one party will cause some exchange partners to drop out. Example: The problem of asymmetric information in insurance. If the price is sufficiently high, the only people who will seek to purchase medical insurance are people who know they are ill (but the insurer does not). This will lead to further price increases to cover costs. Also referred to as the ‘hidden attributes’ problem (the state of already being ill is the hidden attribute), to distinguish it from the ‘hidden actions’ problem of moral hazard. See also: incomplete contract, moral hazard, asymmetric information.

We know that a common reason for contracts to be incomplete is that information about an important aspect of the interaction is unavailable, or unverifiable. Information is often asymmetric—that is, one party knows something relevant to the transaction that the other doesn’t know.

One form of asymmetric information is a hidden action. In Unit 6 we studied the case of the employee whose choice of how hard to work is hidden from the employer. This causes a problem known as moral hazard. There is a conflict of interest because the employee would prefer not to work as hard as the employer would like, and work effort cannot be specified in the contract. We also saw in Unit 6 how the employer’s response (paying a wage above the reservation level) led to involuntary unemployment, which is a Pareto-inefficient outcome in the labour market.

In this section, we introduce a second form of asymmetric information, that of hidden attributes. When you want to purchase a used car, for example, the seller knows the quality of the vehicle. You do not. This attribute of the car is hidden from the prospective buyer. Hidden attributes can cause a problem known as adverse selection.

Hidden actions and moral hazard

The problem of hidden action occurs when some action taken by one party to an exchange is not known or cannot be verified by the other. For example, the employer cannot know (or cannot verify) how hard the worker is actually working.

The term moral hazard originated in the insurance industry to express the problem that insurers face, namely, the person with home insurance may take less care to avoid fires or other damages to his home, thereby increasing the risk above what it would be in absence of insurance. This term now refers to any situation in which one party to an interaction is deciding on an action that affects the profits or wellbeing of the other but which the affected party cannot control by means of a contract, often because the affected party does not have adequate information on the action. It is also referred to as the hidden actions problem.

Hidden attributes and adverse selection

The problem of hidden attributes occurs when some attribute of the person engaging in an exchange (or the product or service being provided) is not known to the other parties. An example is that the individual purchasing health insurance knows her own health status, but the insurance company does not.

The term adverse selection refers to the problem faced by parties to an exchange in which the terms offered by one party will cause some exchange partners to drop out. An example is the problem of asymmetric information in insurance: if the price is sufficiently high, the only people who seek to purchase medical insurance are people who know they are ill (but the insurer does not). This will lead to further price increases to cover costs. Also referred to as the ‘hidden attributes’ problem (the state of already being ill is the hidden attribute), to distinguish it from the ‘hidden actions’ problem of moral hazard.

Hidden attributes and adverse selection

George Akerlof, an economist, was the first to analyse what came to be called the ‘lemons problem’ in 1970. (In addition to being a fruit, a ‘lemon’ is something which after being purchased turns out to be faulty.) Initially his paper on the subject was rejected by two economics journals for being trivial. Another returned it, saying that it was incorrect. Thirty-one years later, he was awarded the Nobel Prize for this work.

Akerlof and co-author Robert Shiller give a simple explanation of the so-called market for lemons in this book: George A. Akerlof and Robert J. Shiller. 2015. Phishing for Phools: The Economics of Manipulation and Deception. Princeton, NJ: Princeton University Press. Watch Robert Shiller’s TED talk on the book.

A famous example of how hidden attributes may result in a market failure is known as the market for lemons. The model describes a used-car market:

If prospective buyers were able to observe the quality of every car, then buyers would approach each seller and bargain over the price, and by the end of the day all the cars (except for the entirely worthless one) would be sold at a price somewhere between their true value and half the true value. The market would have assured that all mutually beneficial trades would take place.

But, on any day, there is a problem—potential buyers have no information about the quality of any car that is for sale. All they know is the price at which cars sold the previous day. The most that prospective buyers are willing to pay for a car is the average value of the cars sold the day before.

Now suppose that ten cars had been offered on the market the day before. We use a proof by contradiction to show that one by one, the sellers of the highest quality cars will drop out of the market, until there is no market for used cars. Consider the market today, as shown in Figure 11.13:

Yesterday, all the cars (as we assumed at the start) were put on the market and sold. The average value of these cars was $4,500, so the most a buyer is willing to pay today is $4,500.
At the beginning of the day, each prospective seller considers selling their car, expecting a price of $4,500 at the most. Most of the owners are happy, because it is more than half the true value of their cars. But one owner isn’t pleased. The owner of the best car will not sell unless the price exceeds half the value of his car—more than $4,500.
Prospective buyers will not pay this price. Therefore, the owner of the best car will not offer it for sale today. No one with a car worth $9,000 is willing to participate in this market.
The rest of the cars will sell today; their value averages $4,000.
Tomorrow, buyers will know the average value of the cars sold today. Therefore, buyers will decide they are willing to pay at most $4,000 for a car tomorrow.
The owner of tomorrow’s highest-quality car (the one worth $8,000) will know this, and know that she will not get her minimum price, which is greater than $4,000. Tomorrow, she will not offer her car for sale.
As a result, the average quality of cars sold on the market tomorrow will be $3,500, which means the owner of the third-best car will not put his car up for sale the day after tomorrow.
And so it goes on, until, at some point next week, only the owner of a lemon worth $1,000 and a totally worthless car will remain in that day’s market.
If cars of these two values had sold the previous day, then, the next day, buyers would be willing to pay at most $500 for a car.
Knowing this, the owner of the car worth $1,000 will decide she would rather keep her car.
The only car on the market will be worth nothing. Cars that remain in this market are lemons, because only the owner of a worthless car would be prepared to offer that car for sale.

Figure 11.13 The market for lemons.

Economists call processes like this adverse selection because the prevailing price selects which cars are left in the market. If any cars are traded, they will be the lower quality ones. The selection of cars is adverse for buyers. In the example above, there are no cars left at all—the market disappears altogether.

Adverse selection in the market for health insurance

The market for lemons is a well-known term in economics, but the lemons problem—that is, the problem of hidden attributes—is not restricted to the used-car market.

Another important example is health insurance. Imagine that you were born into a population in which you do not know whether you would have a serious health problem or develop such a problem later in life, or perhaps be entirely healthy until old age. There is a health insurance policy available covering any medical services you may need, and the premium is the same for everyone; it is set according to the average expected medical costs of people in the population. For the insurance company, the premiums will cover the total expected payout, assuming everyone signs up. Would you buy this health insurance policy?

In this situation, most people would be happy to purchase the policy, because serious illness imposes high costs that are often impossible for an average family to pay. The costs of protecting you and your family from a financial catastrophe (or the possibility that you can’t afford healthcare when you need it) are worth the insurance premium.

The assumption that you do not know anything about your health status in this thought experiment is unrealistic. It is another use of John Rawls’ veil of ignorance, discussed in Unit 3. Thinking about this problem as an impartial observer highlights the importance of the veil of ignorance assumption.

Though everyone would have bought insurance if they did not know about their future health status, the situation changes dramatically if we can choose whether to buy health insurance without the veil of ignorance, that is, knowing our health status. In this situation, information is asymmetric. Look at the situation from the standpoint of the insurance company:

This is another example of a missing market: many people will be uninsured. It is a market that could exist, but only if health information were symmetrical and verifiable (ignoring for the moment the problem of whether everyone would want to share their health data). It could provide benefits to both insurance company owners and people who wanted to insure themselves. Not having such a market is Pareto inefficient.

To address the problem of adverse selection due to asymmetric information, and the resulting missing markets for health insurance, many countries have adopted policies of compulsory enrolment in private insurance programs or universal tax-financed coverage, so that the healthy and the lucky pay the medical costs of the unhealthy and the unlucky.

Moral hazard in the insurance market

Hidden attributes are not the only problem facing insurers, whether private or governmental. There is also a problem of hidden actions. Buying an insurance policy may make the buyer more likely to take exactly the risks that are now insured. For example, a person who has purchased full coverage against damage or theft of his car may take less care in driving or locking the car than someone who has not purchased insurance.

Insurers typically place limits on the insurance they sell. For example, coverage may not apply (or may be more expensive) if someone other than the insured is driving, or if the car is usually parked in a place where a lot of cars are stolen. These provisions can be written into an insurance contract.

But the insurer cannot enforce a contract about how fast you drive or whether you drive after having had a drink. These are the actions that are hidden from the insurer because of the asymmetric information—you know these facts, but the insurance company does not.

principal–agent relationship
This is an asymmetrical relationship in which one party (the principal) benefits from some action or attribute of the other party (the agent) about which the principal’s information is not sufficient to enforce in a complete contract. See also: incomplete contract. Also known as: principal–agent problem.

This is a problem of moral hazard, similar to the one of labour effort. They are both principal–agent problems—the agent (an insured person, or employee) chooses an action (how careful to be, or how hard to work) that matters to the principal (the insurance company, or the employer), but the action cannot be included in the contract because it is not verifiable.

Though seemingly very different from the problem of chlordecone pollution or from public goods and common-pool resources described in the previous section, these moral hazard problems are similar in an important respect. In every case, someone makes a decision that has external costs or benefits for someone else. In other words, costs or benefits that are uncompensated. For example, in the moral hazard case, the insured person (the agent) decides how much care to take. Taking care has an external benefit for the insurer (principal) but is costly for the agent; consequently we have a market failure—the level of care chosen is too low.

Therefore, these problems of moral hazard (and also the adverse selection problems described earlier in this section) can be placed within the framework of external effects and market failure we are using throughout this unit. The problems arising from asymmetric information are summarized in the table in Figure 11.14.

Decision How it affects others Cost or benefit Market failure (misallocation of resources) Possible remedies Terms applied to this type of market failure
An employee on a fixed wage decides how hard to work Hard work raises employer’s profits Private cost, external benefit Too little effort, wage above reservation wage, unemployment More effective monitoring, performance-related pay, reduced conflict of interest between employer and employee Incomplete labour contract, hidden action, moral hazard
Someone who knows he has a serious health problem buys insurance Loss for insurance company Private benefit, external cost Too little insurance offered, insurance premiums too high Mandatory purchase of health insurance, public provision, mandatory health information sharing Missing markets, adverse selection
Someone who has purchased car insurance decides how carefully to drive Prudent driving contributes to insurance company’s profits Private cost, external benefit Too little insurance offered, insurance premiums too high Installing driver-monitoring devices Missing markets, moral hazard

Figure 11.14 Asymmetric information, market failures, with remedies.

Exercise 11.8 Hidden attributes

  1. Identify the hidden attributes in the following markets and how they may impede market participants from exploiting all the possible mutual gains from exchange:
    1. a second-hand good being sold on eBay
    2. Craigslist or a similar online platform
    3. renting apartments through Airbnb
    4. restaurants of varying quality.
  2. Explain how the following may facilitate mutually beneficial exchanges, even in the presence of hidden attributes:
    1. electronic ratings shared among past and prospective buyers and sellers
    2. exchanges among friends, and friends of friends
    3. trust and social preferences
    4. intermediate buyers and sellers, such as used-car dealers.

Question 11.7 Choose the correct answer(s)

There are ten cars on the market, of which six are good quality cars worth $9,000 to buyers, and the others are lemons, worth zero. There are many potential buyers who do not know the quality of each car, but they know the proportion of good quality cars, and are willing to pay the average value. All sellers are happy to accept a price at least half the value of their car. Based on this information, which of the following statements are correct?

  • The buyers are willing to pay at most $4,500.
  • Only the lemons will be sold in this market.
  • All cars will be sold at a price of $5,400.
  • All cars will be sold at a price of $4,500.
  • The average value to the buyers is ($9,000 × 6 + $0 × 4)/10 = $5,400. This is the highest price that the buyers are willing to pay.
  • There are many buyers willing to pay the average value, which is $5,400. At this price, all sellers are willing to sell; so all cars will be sold.
  • There are many buyers willing to pay the average value, which is $5,400. At this price, all sellers are willing to sell, so all cars will be sold. Competition between buyers will ensure that the price will be the most they are willing to pay ($5,400).
  • All sellers will sell if the price is $4,500 or more. There are many buyers willing to pay the average value, which is $5,400. Therefore, all cars will be sold, but competition between buyers will drive up the price beyond $4,500.

Question 11.8 Choose the correct answer(s)

In which of the following cases is there an adverse selection problem?

  • a motor insurance market, in which the insurers do not know how carefully the insured people drive
  • a health insurance market, in which the insurers do not know whether or not the applicants for insurance are habitual smokers
  • online sales of nutritional supplements, when consumers cannot tell whether their contents are as claimed by sellers
  • a firm that employs home-workers, but cannot observe how hard they are working
  • This is a hidden action (moral hazard) problem.
  • Smoking is a hidden attribute as smokers have higher risks. If the premium was set for people of average risk, non-smokers would be less likely to apply for insurance.
  • The quality of the product is a hidden attribute. If the price was equal to the marginal cost of average quality products, only the producers of poor quality or fake products would want to sell.
  • This is a hidden action (moral hazard) problem.

11.13 Market failure and government policy

Figure 11.15 brings together the examples we have seen in which markets fail to allocate resources efficiently. At first sight they seem different from each other, but in each one, we can identify an external benefit or cost that a decision-maker does not consider. The table in Figure 11.16 shows that the fundamental source of market failure is an information problem—some important aspect of an interaction that cannot be observed by one of the parties, or cannot be verified by a court.

Possible remedies for market failures

The table in Figure 11.15 also shows some possible remedies. Governments play an important role in the economy in their attempts to diminish the inefficiencies associated with many kinds of market failure. However, the same information problems can hamper a government seeking to use taxes, subsidies, or prohibitions to improve on the market outcome. For example, the French government eventually decided to ban the use of chlordecone rather than collect the information necessary to devise a tax on banana production or provide compensation to the fisheries.

Sometimes, a combination of remedies is the best way to cope with these information problems and resulting market failures. An example is car insurance. In many countries, third-party insurance (covering damage to others) is compulsory to avoid the adverse selection problem that would occur if only the accident-prone drivers purchased insurance. To address the moral hazard problem of hidden actions, insurers sometimes require the installation of on-board monitoring devices so that prudent driving habits can be an enforceable part of the insurance contract.

Decision How it affects others Cost or benefit Market failure (misallocation of resources) Possible remedies Terms applied to this type of market failure
A firm uses a pesticide that runs off into waterways Downstream damage Private benefit, external cost Overuse of pesticide and overproduction of the crop for which it is used Taxes, quotas, bans, bargaining, common ownership of all affected assets Negative external effect, environmental spillover
You take an international flight Increase in global carbon emissions Private benefit, external cost Overuse of air travel Taxes, quotas Public bad, negative external effect
You travel to work by car Congestion for other road users Private cost, external cost Overuse of cars Tolls, quotas, subsidized public transport Common-pool resource, negative external effect
An employee on a fixed wage decides how hard to work Hard work raises employer’s profits Private cost, external benefit Too little effort, wage above reservation wage, unemployment More effective monitoring, performance-related pay, reduced conflict of interest between employer and employee Incomplete labour contract, hidden action, moral hazard
Someone who knows he has a serious health problem buys insurance Loss for insurance company Private benefit, external cost Too little insurance offered, insurance premiums too high Mandatory purchase of health insurance, public provision, mandatory health information sharing Missing markets, adverse selection
Someone who has purchased car insurance decides how carefully to drive Prudent driving contributes to insurance company’s profits Private cost, external benefit Too little insurance offered, insurance premiums too high Installing driver-monitoring devices Missing markets, moral hazard
Borrower devotes insufficient prudence or effort to the project in which the loan is invested Project more likely to fail, resulting in non-repayment of loan Private benefit, external cost Excessive risk, too few loans issued to poor borrowers Redistribute wealth, common responsibility for repayment of loans (Grameen Bank) Moral hazard, credit market exclusion
Bank that is ‘too big to fail’ makes risky loans Taxpayers bear costs if bank fails Private benefit, external cost Excessively risky lending Regulation of banking practices Moral hazard
A firm producing a differentiated good Price is too high for some potential buyers Private benefit, external cost Too low a quantity sold Competition policy Imperfect competition

Figure 11.15 Market failures with remedies.

Question Answer
Why do market failures happen? People, guided only by market prices, do not take account of the full effect of their actions on others
Why is the full effect of their actions on others not taken into account? There are external benefits and costs that are not compensated by payments
Why are some benefits or costs not compensated? No markets exist in which they can be traded
Why not? And why can’t private bargaining and payments solve the problem? The required property rights and contracts cannot be enforced by courts of law
What prevents property rights and contracts from being enforceable? Asymmetric or non-verifiable information

Figure 11.16 Market failures and information problems.

Implementing the remedies

In this unit, we have diagnosed market failures and used economic models to propose remedies. Although in some cases private bargaining may resolve the market failure, in many cases it is the government that must implement the remedy. This raises the question of the conditions under which governments actually implement policies to rectify market failures.

This is the subject of Unit 12, in which we consider the way both markets and hierarchies—including governments—resolve problems of resource allocation. And we also explain why, even in democratic societies, governments often fail to adopt policies that will make economic outcomes more efficient and fair.

Exercise 11.9 Market failure

Construct a table like the one in Figure 11.15 to analyse the possible market failures associated with the decisions below. In each case, identify which markets or contracts are missing or incomplete.

  1. You inoculate your child with a costly vaccination against an infectious disease.
  2. You use money that you borrow from the bank to invest in a highly risky project.
  3. A fishing fleet moves from the overfished coastal waters of its own country to international waters.
  4. A city airport increases its number of passenger flights by allowing nighttime departures.
  5. You contribute to a Wikipedia page.
  6. A government invests in research in nuclear fusion.

11.14 Conclusion

Markets facilitate the division of labour by providing a means for people who are specialized in one type of production to acquire the full range of goods and services that make up their livelihood. Ideally, market prices provide information about how scarce a good or service is. The message that this information sends motivates people to produce goods that are scarce because others value them. It also motivates people to limit their use of scarce goods. The combination of the message and the motivation provided by prices can sometimes produce efficient economic outcomes.

But markets often do not match this ideal, and as a result outcomes that are Pareto inefficient (that is, market failures) may occur. In some cases, a thing that is scarce⁠—like clean air or water⁠—does not have a price, and so it is treated by users as if it were not scarce. In other cases, prices fail to reflect the entire social costs of the good (such as the environmental damage resulting from the use of fossil fuels).

Where prices are either entirely missing or fail to reflect true social costs, the decision-maker does not fully account for the costs or benefits of their actions for others (external effects). Where external effects are present, we say that the contract governing the transaction is incomplete (or even entirely absent). Contracts are incomplete because the information that would be necessary to write and enforce a contract is either not known by at least one of the parties (asymmetric information) or cannot be used in court (for example, non-verifiable information).

Asymmetric information comes in two forms: hidden actions (such as how hard an employee is working, considered in Units 6 and 8), which causes a problem known as moral hazard, and hidden attributes (such as whether the person seeking to buy health insurance is already ill), which causes a prob­lem called adverse selection. In the case of adverse selection, a mutually beneficial market for the good (namely the insurance) could exist if information were symmetrical and verifiable, but otherwise there is a missing markets problem.

Where market failures exist, private Coasean bargaining between parties involved or public policies such as a Pigouvian tax (or subsidy) can improve outcomes. When seeking to use taxes, subsidies, or prohibitions to improve market outcomes governments face many of the same asymmetric information problems that confront private economic actors. Governments may fail to address market failures for another reason: powerful groups may benefit from the status quo. We address these limits to governmental policies in the next unit.

11.15 Doing Economics: Measuring willingness to pay for climate change mitigation

In this unit, we discussed external effects and how they arise, along with ways to mitigate them. When designing policies to reduce carbon emissions or air pollution, or to save an endangered species or preserve biodiversity, economists face the problem that markets for environmental amenities are missing. How can the value to people of the abatement of environmental damage be calculated and set against the cost of implementing any abatement? The issue of how to measure willingness to pay (WTP) of non-market goods is important for policymaking, since governments use the economic value of these goods as a major factor in deciding what quantity of these goods to provide (or whether to provide the good at all). Incorrectly estimating the WTP may result in inefficiencies, such as providing too much or too little of the good in question.

In Doing Economics Empirical Project 11, we will look at climate change mitigation as an example. Since tackling climate change may entail short-term costs such as decreased production, governments may want to know how much their citizens are willing to pay to reduce carbon emissions as a method of mitigating climate change. Using survey data, we will compare WTP (mean and median) under each method and assess whether WTP responses differ according to question type.

Go to Doing Economics Empirical Project 11 to work on this project.

Learning objectives

In this project you will:

  • compare survey measures of willingness to pay
  • construct indices to measure attitudes or opinions
  • use Cronbach’s alpha to assess indices for internal consistency
  • practise re-coding and creating new variables.

11.16 References

  1. Adam Smith. (1776) 2003. An Inquiry into the Nature and Causes of the Wealth of Nations. New York, NY: Random House Publishing Group. 

  2. Joseph A. Schumpeter. (1943) 2003. Capitalism, Socialism and Democracy. pp. 167–72. Routledge. 

  3. Friedrich A. Hayek. 1944. The Road to Serfdom. Chicago, Il: University of Chicago Press. 

  4. Hayek, Friedrich A. 1948. ‘The Meaning of Competition’, in Individualism and Economic Order. Chicago, Il: University of Chicago Press. 

  5. Hayek, Friedrich A. 1948. ‘The Meaning of Competition’, in Individualism and Economic Order. Chicago, Il: University of Chicago Press. 

  6. Hayek, Friedrich A. 1945. ‘The Use of Knowledge in Society’, reprinted in Friedrich A. Hayek. 1948. Individualism and Economic Order. Chicago, Il: University of Chicago Press. 

  7. Douglas A. Farnie. 1979. The English Cotton Industry and the World Market: 1815–1896. Oxford: Oxford University Press. 

  8. Marcel Fafchamps and Bart Minten. 1999. ‘Relationships and Traders in Madagascar’. Journal of Development Studies 35 (6) (August): pp. 1–35. 

  9. Arthur Pigou. 1912. Wealth and Welfare. London: Macmillan & Co. 

  10. Arthur Pigou. 1920. The Economics of Welfare. London: Macmillan & Co.