Chapter 14 Cleaning Up the Air and Using Up the Oil

Smokers, Nonsmokers, and the Coase Theorem

Cigarettes are sold and smoked almost everywhere. Yet in most countries around the world, you are not able to smoke when and where you please. Governments around the world place limitations on who can buy cigarettes, where they can be bought, and where they can be consumed. From an economic point of view, governments are deliberately restricting the ability of individuals to engage in voluntary transactions.There are many different ways in which governments intervene in market transactions. Chapter 12 "Barriers to Trade and the Underground Economy" contains more discussion. Why do governments restrict an individual’s ability to smoke where and when that person wants?

Mixed messages.

Our answer to this question begins by imagining two people who must share an office. One is a nonsmoker who dislikes the smell of cigarette smoke, while the other likes to smoke while working. On the way to work one day, the smoker purchases a pack of cigarettes that she plans to smoke at work. We can reasonably deduce from this that her valuation of these cigarettes is greater than the price she has to pay. She gets buyer surplusA measure of how much the buyer gains from a transaction, equal to the buyer’s valuation minus the price. from the purchase. To be concrete, suppose a pack of 20 cigarettes costs $4 and her valuation of a pack of cigarettes is $10. Her surplus is then $6.

We can also reasonably assume that the seller’s cost is less than the price, otherwise he would not choose to make the sale. He gets seller surplusA measure of how much the seller gains from a transaction, equal to the price minus the seller’s valuation.. For example, if his wholesale price for a pack of cigarettes is $2, then he earns $2 surplus (= $4 − $2) on every pack that he sells. The total surplus from the sale is the buyer surplus plus the seller surplus—that is, $8. So far so good.

The problems begin when the worker smokes her pack of cigarettes in the office. She obtains her $6 worth of enjoyment. However, a third party has now been affected by her decision to consume cigarettes: her office mate. The office mate dislikes the smell of smoke and may even face health risks from second-hand smoke. Thus even though the smoker and the store that sold the cigarettes are both better off, the office mate has been made worse off.

We should not automatically assume that the best thing is to ban smoking in the office just because the office mate is adversely affected. We need to know how much the nonsmoker is inconvenienced. Suppose the most the nonsmoker would be willing to pay for a smoke-free office is $2 per day. In this case, the $6 gain to the smoker exceeds the $2 loss to the nonsmoker. It seems like it should be easy enough for the two individuals to find a way for everyone to be happy. For example, imagine that the smoker agreed to pay the nonsmoker $4 a day for the right to smoke in the office. Both would then get $2 surplus per day.

On the other hand, suppose a smoke-free office is worth $10 per day to the nonsmoker. In this case, his valuation of clean air ($10) exceeds the smoker’s gain ($6). The smoker would be unwilling to pay the nonsmoker enough to compensate for dirtying the air in the office. It would be better not to allow smoking in the office.

We have assumed here that the default situation is that the office should be smoke-free. In the language of economics, the nonsmoker owns the property rightsAn individual’s (or institution’s) legal right to make all decisions regarding the use of a particular resource. to the clean air in the office. Property rights over a resource mean that, by law, the owner can make all decisions regarding the use of the resource. Because of this, the smoker must pay the nonsmoker compensation if she wishes to be allowed to smoke in the office.

We could imagine the opposite situation, where the smoker starts off with the right to smoke in the office. Would we expect a different result from their negotiations? If the smoke-free office was worth only $2 to the nonsmoker, then he would not be willing to pay enough to persuade his office mate not to smoke: the most he would pay is $2, which is less than the smoker’s surplus. If, on the other hand, the nonsmoker valued the smoke-free office at $10, then he values a smoke-free office more than the smoker values smoking in the office. The nonsmoker could pay the smoker not to smoke. For example, imagine he pays her $8 per day to not smoke. He pays $8 for the clean air, which is worth $10 to him, so he gets $2 of surplus. The smoker receives $8, which exceeds the surplus she would get from smoking in the office. Again, they would both be happy.

Thus if the smoker “owns” the clean air, the nonsmoker must pay the smoker if he wants a smoke-free office. If the nonsmoker has the property rights, it is the smoker who must pay. In either case, the basic outcome will be the same: there will be smoking in the office if the smoker’s valuation exceeds the nonsmoker’s valuation; there will be no smoking if the nonsmoker’s valuation exceeds the smoker’s valuation. But the property rights are valuable. It is the owner of the property rights—whoever that may be—who gets compensation from the other.

As long as they know who has the property rights, it seems likely that the two individuals will be able to come to an agreement that benefits them both. You might imagine, though, that they would find it far harder to come to an agreement if it was ambiguous who had the property rights in the first place. The smoker would likely claim the right to smoke in the office, while the nonsmoker would assert his right to clean air. If they could not settle this basic question, it is unlikely that they would be able to reach a more complicated agreement involving compensation payments.

Let us imagine a further twist. Suppose the nonsmoker has property rights but values clean air at only $7 per day. This is still greater than the smoker’s surplus, so, as before, we expect that they would agree to a smoke-free office. But the nonsmoker’s valuation for clean air is less than the total surplus of the smoker and the store that sold her the cigarettes (recall that the smoker gets $6 surplus and the store gets $2 surplus). If the storekeeper, the smoker, and the nonsmoker all got together, they should again be able to find an arrangement that benefits everyone. For example, the storekeeper and the smoker could jointly give the nonsmoker $7 and still have $1 of surplus to bargain over.

It seems perfectly reasonable to imagine that two people who share an office could come to a mutually beneficial agreement about smoking in the office. It seems much more far-fetched, though, to imagine that they would come to an agreement together with the storekeeper who sold the cigarettes. Economists say that the difference between the two cases is due to transaction costsThe costs of making and enforcing agreements.—the costs of making and enforcing agreements.

We began this section by observing that a transaction may affect not only the buyer and the seller but also third parties. When this is the case, we cannot be sure that trade benefits everyone. Even if the buyer and the seller are both made better off, third parties may be made worse off. However, if property rights are clearly established and transaction costs are low, then we can expect that private negotiations could solve these problems. This idea was first articulated by the Nobel prize–winning economist Ronald Coase (http://www.coase.org) and is called the Coase theoremIf property rights are clearly established and transaction costs are low, private bargaining will lead to efficient outcomes.: if property rights are clearly established and transaction costs are low, private bargaining will lead to efficient outcomes.

It is notable that in reality, we do not see office workers buying and selling the right to smoke in an office. Instead, blanket bans on smoking have been enacted throughout the United States and in many other countries throughout the world. Why do we get this government response? One argument for these bans is that smoking poses health risks. In this case, antismoking campaigns are based on an idea that individuals are not always capable of making good choices for themselves.Chapter 5 "Life Decisions" has more to say about whether individuals are good judges of their own actions, particularly when making decisions with long-term consequences. But another reason is a recognition of the transaction costs involved in these private negotiations. Even if you think two coworkers could reach an agreement, imagine an office with 10 people—perhaps there are 3 smokers and 7 nonsmokers—all of whom place a different valuation on clean air in the office. If we knew everybody’s true valuations, then in theory it would be possible to create a system of payments that made everybody better off. In practice, however, people might lie about their valuations. Finding the right system of payments would be very hard indeed and would take a lot of time and effort.

Perhaps you can now see the parallel with Mexico City. The major pollution problem is the emissions from the cars that people drive. Individual residents of Mexico City make decisions to buy gasoline and drive. These transactions create value for the drivers and the sellers of gasoline—but third parties are adversely affected. The Coase theorem can work when the parties involved are easily identifiable and small in number. In contrast, it is impossible to imagine the 20 million residents of Mexico City all meeting and coming to some kind of private agreement to limit their collective driving behavior.

Social Dilemma

Air pollution and second-hand smoke have a common structure, which we explain in this section. Once you understand these common elements, you will probably be able to think of many other examples. We show the interactions between an individual and the rest of society in Table 14.1 "The Payoffs in a Social Dilemma Game". This is called a social dilemmaA situation where individually rational choices lead to an outcome that is bad for society as a whole. game.

In Mexico City, it is very possible that people would agree that they would prefer a situation where everybody drives less. Yet this is not what happens. To see why, look at Table 14.1 "The Payoffs in a Social Dilemma Game" and imagine you are a car owner in Mexico City. You have to decide how to get to work—by driving or taking public transportation. There are two rows in the table. One is labeled “You Drive,” and the other is labeled “You Take Public Transportation.” The rows represent your possible choices. The columns refer to everybody else’s choices. Everyone else who owns a car similarly chooses between driving and taking public transportation. To keep things simple, we suppose that everyone else makes the same choice. If everybody chooses to drive, then the air is polluted. If everybody chooses to take public transportation, then the air is clean. The current situation in Mexico City is that you and all other car owners are driving to work. You enjoy the convenience of driving rather than taking public transportation, but you suffer from the polluted air.

The numbers in the table refer to your payoffs from the different possible combinations. As in our smoking example, we can think of these as the valuations per day that you place on different outcomes. What matters is how these different possibilities compare with the status quo, where you and everyone else drive. We therefore begin by setting your payoff at the status quo (the number in the top left cell) at $0. Suppose also that the following are true.

• You value clean air rather than dirty air at $2 per day—that is, you would be willing to pay $2 per day to have clean air rather than dirty air.
• After taking into account the relative costs and inconveniences of driving versus public transportation, you value driving compared to public transportation at $1 per day—that is, you would need to be compensated $1 per day to make you just as happy to take public transportation rather than drive.

Based on these conditions, we can calculate the payoffs in the other three cells of the table:

• The top right cell is the case where you drive and everyone else takes public transportation. Compared to the status quo, you are better off because you get to enjoy clean air. This is worth $2 to you, so your payoff is $2.
• The bottom left cell is the case where you take public transportation and everyone else drives. You still have to breathe polluted air, and you no longer get the convenience of driving (which is worth $1). Compared to the status quo, you are worse off. Your payoff is −$1.
• The bottom right cell combines the two previous cases. In this case, you and everyone else take public transportation. You get the benefit of clean air (worth $2), but you lose the benefit of driving (worth $1). Your payoff is $2 − $1 = $1.

What would you do in this situation? Suppose you think that everyone else is going to drive. You are better off if you drive (payoff is $0) rather than take public transportation (payoff is −$1). What if everyone else takes public transportation? Then you still prefer to drive (payoff is $2) rather than take public transportation (payoff is $1). We conclude you will drive regardless of what others in society choose to do.

Here is the crux of the problem: the situation looks the same to everybody else as it does to you. As you evaluate these relative payoffs and choose to drive, so too does everyone else. We therefore expect that everyone will follow their individual incentives and choose to drive. We end up in the top left cell, where your payoff is $0.

What is striking, though, is that you would prefer the outcome where everybody—including you—uses public transportation. Your payoff in the bottom right cell is $1, which is better than the current situation. Everyone else would prefer this outcome as well. Society ends up in a bad situation, with everybody driving, even though everyone agrees that there is a better option out there. This is the essence of the social dilemma.

You are one of many people. Although you may value clean air, you are powerless as an individual to keep it clean. And because you are only one person, your decision has a tiny effect on the overall quality of the air. Thus you choose to drive, without paying attention to your effect on the environment. But because everybody makes the same decision, the cumulative effect is that there is a lot of air pollution.

The social dilemma is also sometimes known as the “tragedy of the commons.” This refers to the time when cattle farmers had access to common grazing land. Because every farmer had the right to graze his cattle on this land, no one was in a position to ensure that the land was well managed. Every farmer paid attention only to the health of his own cattle and did not worry about the effect of his cattle on the overall quality of the grazing land. Because every farmer made the same decision, the result was overgrazing, which destroyed the land for everyone.

How do we solve problems such as this? To avoid the bad outcome of the social dilemma, we must find some way of changing the payoffs of the game. We have already seen that, if transaction costs are low, people may be able to negotiate privately. In the case of Mexico City smog, however, such negotiation is impractical. In this case, one possible solution is for the government to alter the payoffs. For example, in the run-up to the 2008 Olympics in Beijing, the Chinese government wanted to improve air quality in the city. It therefore allowed cars to drive in the city only every other day (whether a car was permitted on a given day depended on whether the last digit of the license number was odd or even). If you tried to drive on the wrong day, you might have to pay a large fine, so the payoff to driving changed. We have more to say about government policy later in the chapter.

Private Benefits and Social Costs

When people face a social dilemma, the actions that are the best for all individuals lead to an outcome that is bad for everyone. Much of the rest of the chapter addresses why this happens.

We begin by remembering our theory of how people make consumption choices. In general, people consume a good up to the point where their marginal valuationThe maximum amount an individual would be willing to pay to obtain one extra unit of that good. from the last unit of that good equals the price of that unit. This is a specific statement of a more general principle for decision making: “consume until marginal benefit equals marginal costThe extra cost of producing an additional unit of output, which is equal to the change in cost divided by the change in quantity..”

In the social dilemma of Table 14.1 "The Payoffs in a Social Dilemma Game", you had two choices: driving or not driving. Let us now expand that to think about a situation where you are deciding how much to drive. Driving your own car brings private benefits (that is, benefits that are obtained only by you), such as comfort and convenience. Driving also brings private costs, such as the costs of gasoline and maintenance of your vehicle. If this were all that was going on, there would be no problem: you would drive up to the point where your marginal (private) benefit from driving was equal to your marginal (private) cost of driving.

When you drive, though, you also impose costs on other people. Your decision to drive one more mile has a marginal social costThe cost to society of consuming or producing one more unit of a good or a service. as well as a marginal private cost. Marginal social cost is the cost to society of consuming or producing one more unit of a good or a service. By “society,” we simply mean “you and everybody else.”

When you choose to drive your car, you contribute to air pollution. This is a cost to the rest of society. However, you have no incentive to worry about this. You care only about the extra cost to you. The same is true when the smoker smokes one more cigarette in the office: She imposes a cost on her office mate, but—unless they make an agreement otherwise—she does not pay this cost. She takes into account the marginal private cost to her (that is, how much she must pay for one more cigarette), but she ignores the cost to other people.

Ignoring for a moment some tricky questions about how to measure the cost of your actions to society, we can set out a principle for how much you should consume if your concern were the overall well-being of society: “consume until marginal benefit equals marginal social cost.” The marginal social cost of your driving is the extra cost if you drive more. It is the cost both to you and the cost that you impose on others in society. There can be many components to this cost. One component is the marginal private cost to you. In addition, you pollute the air, contributing to public health problems. You add greenhouse gases to the atmosphere, contributing to the risks of climate change. You make the roads more congested, thus wasting the time of other drivers on the roads. You cause wear and tear on the roads, which will ultimately be paid for by taxes on all drivers.

You might be puzzled at this point. If you drive one extra mile, does that really have any appreciable cost on society? The extra emissions from your driving obviously have a tiny effect on pollution and greenhouse gases. The wear and tear you impose on the roads is minimal. How can these minuscule effects possibly matter? The first and more obvious answer to this question is that the quality of the atmosphere and the roads is affected by everybody’s decisions—not only yours. Hundreds of thousands of cars around the world are polluting the atmosphere and damaging the roads. The second, more subtle, answer is that though your individual influence on air and road quality is very small, you are affecting a very large number of people. If you drive an extra mile in Mexico City, you are affecting the air that is breathed by 20 million people. The marginal social cost of your driving includes the effect on every single one of these people. Likewise, one more mile of driving may add only a tiny amount to the greenhouse gases in the atmosphere, but that tiny effect must be added up over the entire population of the world.

As a result, the marginal social cost of your driving is greater than your marginal private cost. Figure 14.3 "A Divergence between Marginal Private Cost and Marginal Social Cost" shows the implications of this. Think of consumption in this diagram as referring to the amount of driving you do. The marginal private cost of your driving is the cost of fuel, depreciation of your car, and so on. But the marginal social cost also includes the pollution of the air and the congestion of the roads. Because the marginal social cost is greater than the marginal (private) cost, you will drive too much, from the perspective of society as a whole. This is exactly what we saw in the social dilemma. People choose to drive and pollute the air, even though all members of society could be happier if everyone were to take public transportation and generate less pollution.

The gap between private costs and social costs means that too much driving is undertaken, from the perspective of society as a whole. The outcome is inefficient because people only have an incentive to take account of the private costs of their actions.

Figure 14.3 A Divergence between Marginal Private Cost and Marginal Social Cost

The gap between marginal private cost and marginal social cost is a measure of the impact that an individual has on the rest of society.

Figure 14.3 "A Divergence between Marginal Private Cost and Marginal Social Cost" also illustrates another point. Economic analysis tells us that there is “too much” pollution, from a social point of view. This observation probably comes as no surprise. Economic analysis also makes it clear, though, that it is possible to have too little pollution as well as too much. There is an optimal amount of driving for each individual, to be found where marginal social cost and marginal benefit are equal. At this amount of driving, there will be some pollution: the optimal amount of pollution for society as a whole. If we were to ban driving altogether, we would have less pollution, but we would also lose all the benefits from driving.

Our discussion here has been about decisions of consumers, but firms also are sources of pollution. Firms use trucks and other vehicles that, like cars, impose costs on the rest of society. Some firms pollute the air or the water. Exactly the same principles still apply. Any individual firm has no incentive to take into account the costs that it imposes on the rest of society. As a result, firms pollute too much from a social point of view.

Solutions to Externality Problems

The definition of an externality makes it clear that the fundamental problem is one of behavior—actions by a firm or a household. The behavior reflects a difference between private costs or benefits and social costs or benefits. These observations also point us to a solution. We need to change incentives so as to align private costs or benefits and social costs or benefits. For example, if the private marginal cost of pollution to a firm were somehow equal to the social marginal cost, then a firm acting in its own self-interest would produce the socially optimal amount of pollution. The challenge for policymakers is to find a way to adjust the incentives so that the firm takes into account social marginal costs in addition to private marginal costs.

From this perspective, inefficiency arises because there are no market signals that force the polluter to take into account how its actions are affecting others. The goal of government policy in the presence of externalities is to provide incentives for firms and households to internalize their effects on others. These policies include direct restrictions on what people can do (for example, banning smoking in public buildings), taxes and subsidies that affect prices in an economy, and the introduction of markets that force polluters to pay for the right to pollute. Because externalities involve a divergence between private costs and social costs (or private benefits and social benefits), the goal in all cases is to adjust the incentives so that the actor internalizes the externality.

Difficulties with Environmental Policy

Our discussion of externalities makes it seem that it is straightforward for policymakers to design effective environmental regulation. Government regulators simply need to calculate the difference between marginal private costs and marginal social costs and between marginal private benefits and marginal social benefits. Then they can put in place the appropriate taxes, subsidies, or both.

In practice, a major difficulty is knowing how to place values on externalities. Environmental policies to combat air or water pollution require the government to monitor the amount of emissions effectively and accurately. If emissions cannot be monitored, then tax or permit schemes are impossible to implement. Effective environmental policies also require the government to measure the damage incurred by the victims of the pollution.

As with many economic policies, questions also arise concerning the distribution of resources. In an environmental context, the key point of debate is often whether compensation should and will be paid. If a factory pollutes the river running through a town, imposing negative externalities on the town’s residents, then is it enough to adjust the incentives so that there is the “right” amount of pollution—that is, so that the marginal benefit to the firm equals the marginal social cost? Or should the firm also be required to compensate the residents of the town? This is again a question of property rights, for we are really asking who has the initial right to the clean water in the river. It is also closely related to the question of whether pollution permits should be given away (implying that firms have the property rights) or sold at auction.

Whenever government steps in and enacts policies to affect behavior, it also must worry about whether there will be perverse responses to those incentives. As an example, the disposal of solid waste is a significant environmental problem in many countries. It is in part an economic problem: people do not usually pay directly for the removal of their garbage, so they do not have an incentive to recycle or avoid waste in other ways. A solution—favored by many economists and adopted by some municipalities—is to charge people a fee per bag for garbage removal. A study by Don Fullerton and Tom Kinneman revealed that this policy can bring its own problems.Don Fullerton and Thomas Kinneman, “Household Responses to Pricing Garbage by the Bag,” American Economic Review 86, no. 4 (1996): 971–84. When such a scheme was introduced in Charlottesville, Virginia, citizens responded by putting much more garbage in each bag than they used to (Fullerton and Kinneman called this the “Charlottesville stomp”), so the reduction in the waste stream was lower than the authorities had anticipated. More seriously, some people also responded by dumping their trash illegally.

Other Kinds of Externalities

Our discussion here has been about externalities arising from either the production or the consumption of goods and services. However, the idea of externality is used much more broadly in economics because it is a very helpful way of diagnosing inefficiencies.

One example comes from the working of labor markets. In general, both firms and workers spend time and resources trying to make a good “match.” Firms have human resource departments that spend resources advertising jobs, going through résumés, interviewing job applicants, and so on. Workers spend time preparing their résumés, interviewing at different firms, and so on. Both are willing to do this because a good match can be very beneficial: the firm gets a highly productive worker for which it is willing to pay a good wage.Chapter 9 "Growing Jobs" discusses search and matching in the labor market in more detail.

Now, the more effort the worker and the firm expend, the more likely they are to come up with a good match. And though each benefits individually from this effort, some of the benefit also flows to the other side of the market. In other words, if a worker tries harder to make a good match, this bestows a positive externality on a firm, and if a worker tries harder to make a good match, this bestows a positive externality on the worker.

As another example, is it a good thing if a new fast-food restaurant opens in your town? To answer this, we can think about the externalities that arise because of the decision to enter. When a new restaurant opens, it is providing a product that is similar but not identical to some of the offerings already in the market. It thus steals business from existing restaurants and reduces their profits. We can think of the entry of a new restaurant as imposing a negative externality on existing restaurants. Conversely, the entry of a new restaurant bestows a positive externality on consumers because they benefit from the increased choice. Thus the entry of a new restaurant leads to both positive and negative externalities.

We said previously that if an action has positive externalities, then there will be too little of that action from a social point of view. If an action imposes negative externalities, there will be too much of that action. In this example, the action is the entry of a new firm or product. If the positive externalities are more important, then there is too little entry, from a social point of view. If the negative externalities are more important, then there is too much entry. In general, we cannot draw any conclusions about whether there are too many firms (or products) or too few.

Nonrenewable Resources and the Hotelling Rule

Oil is a nonrenewable (exhaustible) resourceA resource that does not regenerate over time.—that is, a resource that does not regenerate over time. Obviously, if we keep using up a nonrenewable resource, we will eventually run out of it, which is why it is called exhaustible. We most often hear this concern voiced about oil. For example, as oil prices increased to record levels in 2008, there was a great deal of discussion in the media about “peak oil”—the point at which world oil production is at its maximum.

Given that total oil resources are limited, we will eventually go beyond the peak oil point. Many commentators believe that we are at or close to this point already and have expressed concern about the implications of this for the world’s economy. Consider, for example, the following quotation from a 2005 report:

When economists read a quotation like this, however, they typically think that something is missing. The quotation says that, at some point “world oil production will no longer satisfy demand.” Economists respond that this does not make sense: the price of oil will adjust to ensure that supply equals demand. Similarly, they would say that it is very unclear what “oil demand is expected to grow 50 percent” means. Does this mean a shift in demand (a 50 percent increase in the quantity demanded at every price) or an increase in the equilibrium quantity (the quantity purchased will be 50 percent higher than it is now)? Fundamentally, economists would say that we cannot talk sensibly about the oil market without discussing what will happen to the price of oil. We turn to this question next.

Does the Hotelling Rule Work in Practice?

The Hotelling rule is based on a very simple arbitrage idea, so it is highly compelling. Yet it is sometimes difficult to observe the rule in operation in the data for oil or for any other nonrenewable resource. Figure 14.5 "The Price of Oil (in 2008 Dollars)" shows what has happened to the price of oil in the last 60 years or so. The prices are in 2008 dollars. Oil prices were reasonably steady in the 1950s and the 1960s. The 1970s are sometimes called the “oil shock decade,” and from the graph you can see why. The price of oil jumped to a level that is equivalent to over $100 a barrel in 2008 dollars. Then oil prices fell again and were relatively low in the 1980s and 1990s. The early years of the 21st century saw another big increase in oil prices, with the price reaching record levels in 2008.

Figure 14.5 The Price of Oil (in 2008 Dollars)

Oil prices were relatively stable in the 1950s and 1960s, rose rapidly in the 1970s, were low from the mid 1980s until about 2002, and have since increased rapidly.

Source: Federal Reserve Bank of St. Louis FRED database: Spot price of West Texas crude oil (http://research.stlouisfed.org/fred2/data/OILPRICE.txt) deflated by the Consumer Price Index (http://research.stlouisfed.org/fred2/data/CPIAUCSL.txt).

It is not easy to reconcile this figure with the Hotelling rule. We do not see the price of oil increase steadily, as the Hotelling rule seems to suggest. The problem is that the complications that we mentioned in Section 14 "Complications" are quite significant in practice. Over the last several decades, we have seen technological improvements, the discovery of new oil fields, political instability in the Middle East and other oil producing regions of the world, price-fixing by the oil-producing countries, and so on. Most of the time, it seems as if these variables are swamping the pattern we expect from the Hotelling rule.

If we look at other nonrenewable resources, it is likewise difficult to see Hotelling effects. In large part this is because technological improvements have often made resources less valuable, even as they became scarcer. For example, copper became significantly less valuable when scientists developed techniques for transmitting information along fiber-optic cables rather than along copper wire.

For all of these reasons, you should not think of the Hotelling rule as literally describing what will happen to the price of nonrenewable resources in the real world. Instead, you should think of it as explaining one component of the price. If the exhaustion point of the resource is a long way in the future, the Hotelling rule may not play a big role in explaining the price. Other factors that shift the demand and supply curves may explain most of the price variation. Nevertheless, most economists are confident that the Hotelling rule does contribute to resource price changes, and as a resource gets closer to exhaustion, the Hotelling rule will play a bigger and bigger role. James Hamilton, an economist who is an expert on oil, put it as follows in late 2006:

My own view is that, for most of the past century … the resource exhaustion was judged to be sufficiently far off as to be ignored.…I do not infer that the next decade will necessarily be like the previous century. Certainly declining production from U.S. oil reservoirs set in long ago.…

I am not at all prepared to dismiss the hypothesis that [the Hotelling rule has] indeed started to make a contribution to oil prices over the last five years, and will become more apparent over the next five. For example, the announced intention of OPEC producers to cut back production as the price goes below $60 might be most naturally interpreted from that perspective—producers don’t see it as being in their interests to sell for less, given what the oil will be worth in the future.See James Hamilton, “Is Peak Oil Irrelevant?,” Econbrowser, October 24, 2006, accessed January 31, 2011, http://www.econbrowser.com/archives/2006/10/is_peak_oil_irr.html.

Excludable and Nonexcludable Resources

The Hotelling rule tells us that we expect the price of a nonrenewable resource to rise as we use it up. As its price increases, households and firms have an incentive to substitute other goods for the resource. As the price of oil increases, people switch to other forms of transportation and more fuel-efficient vehicles. More generally, an increase in the price of oil makes other forms of energy—such as wind, solar, or nuclear power—more attractive.

Nothing in this description suggests any failure of the market mechanism. Although we have explained what will happen as we start to run out of a resource, we have not given any reason to suggest that we will use up the resource too quickly. In fact, we have even pointed to one reason we might be using up oil too slowly: if oil producers have market power, they have an incentive to limit the supply to the market and increase the price. The mere fact of using up a nonrenewable resource does not mean that the market is not efficient. To understand the difference between using oil and polluting the atmosphere, we need a new distinction.

Go back once more to our example of a smoker in an office. The smoker actually consumes two things: she consumes cigarettes and she consumes the clean air in the office by turning it into dirty air (economists call this “joint consumption”). Clean air, like a cigarette, is a “good.” But it is a good with a special property. Under most circumstances, we cannot allow some people access to clean air while denying others access to clean air. Clean air is nonexcludableNot having the possibility that we can selectively allow or deny access..

Drivers in Mexico City likewise consume clean air: driving leads to less clean air and more dirty air. Like the smoker in the office building and the car drivers in Mexico City, polluting firms consume clean air as well. In this instance, we can think of clean air as being another input into the production process. Air pollution is a process by which the nonexcludable resource that we call “clean air” is consumed by households and firms. This is a somewhat unfamiliar way to talk about environmental pollution, but it makes sense once you think about it. At the end of the production process, there is less clean air than before because the firm has made some air dirty. The polluting firm uses clean air in its production process.

Renewable Resources

Oil is a nonrenewable resource. New oil reserves are not going to be created (at least over any period of time relevant to human beings). Over time, the stock of a nonrenewable resource can only decrease, never increase. By contrast, a renewable resourceA resource that regenerates over time. is one that regenerates over time.

Whether the stock of a renewable resource grows, shrinks, or stays constant depends on the balance between how fast we use up the resource and how quickly it regenerates. If a resource is depleted at a rate faster than it regenerates, the stock will decrease. If a resource regenerates faster than it is depleted, the stock will increase. Examples of renewable resources are stocks of fish in the ocean and forests. Clean air and water are also examples of renewable resources: if the source of pollution disappears, then a polluted river or polluted air will tend to improve naturally over time. Of course, this process can take quite a while.

In some cases resources may be renewable only above a certain threshold. For example, we can overfish a particular species of fish to the point of extinction. So although a renewable resource will normally naturally regenerate over time, it can turn into a nonrenewable resource if the stock falls below a critical level.

As with nonrenewable resources, what matters most for efficiency is whether a resource is excludable or nonexcludable. If a renewable resource is excludable, then a modified version of the Hotelling rule applies. The arbitrage condition in this case takes into account the fact that the resource stock gets larger as time goes by. If the resource is nonexcludable, however, then all the problems we saw earlier in the chapter come into play. The marginal social cost of using the resource will typically exceed the marginal private cost, leading to overconsumption of the resource.

Accumulable Resources

Nonrenewable environmental resources get used up over time. With good stewardship, which usually requires well-established property rights, the stocks of renewable resources can stay constant or even increase. Environmental resources, however, typically cannot grow without limit. There is a limit to how many fish there can be in the ocean or how clean the air can be.

Economists also identify a third class of resources, called accumulable resourcesA resource that can be increased without limit over time through investment.. Such resources can increase (more or less) without limit. The most important examples are physical capital (factories, machines, etc.), human capitalThe skills and knowledge that are embodied within workers. (the skills and know-how of workers), and general human knowledge.

Even though there is no reason to think that the depletion of a nonrenewable resource is necessarily a source of economic inefficiency, it does not follow that the depletion of natural resources is without cost for an economy. For example, as oil becomes more expensive, it costs more to produce other goods and services that use energy as an input. From an economic perspective, this is a lot like a worsening of technology.

Crucially, though, we are accumulating resources such as physical capital, human capital, and knowledge at the same time that we are running down our stocks of oil, coal, and other natural resources. In the last couple of centuries, the economy of the world has grown substantially because we have been able to accumulate resources at a rate that far outpaced our depletion of the natural environment. For much of that time, we had natural resources in abundance, so their price stayed low. At the same time, we increased our technological know-how at unprecedented rates.

There is no guarantee, from economics or anywhere else, that this state of affairs will continue. We will continue to run down our stocks of nonrenewable resources. If technological advance (and other accumulation) fails to keep pace, then we might well see the prices of these goods increase. More generally, the fact that we have seen our economies growing in the past is no guarantee that they will continue to grow in the future.

Technological optimists point to the rapid growth of our knowledge and expect this to continue. If they are right, then technological advance will keep our living standards growing even though we are depleting our stocks of resources. Technological pessimists observe that the last two centuries are an anomaly if we look at the broad sweep of human history. If the rate at which we accumulate knowledge, human capital, and other accumulable resources were to decrease significantly, then the drag on the economy from declining resources would begin to seem substantial. Economics does not allow us to predict the future, and we do not know who is right.