Jump to ContentJump to Main Navigation
Paying for PollutionWhy a Carbon Tax is Good for America$

Gilbert E. Metcalf

Print publication date: 2019

Print ISBN-13: 9780190694197

Published to Oxford Scholarship Online: January 2019

DOI: 10.1093/oso/9780190694197.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 28 October 2020

Why Do Economists Like a Carbon Tax?

Why Do Economists Like a Carbon Tax?

(p.35) Chapter 3 Why Do Economists Like a Carbon Tax?
Paying for Pollution

Gilbert E. Metcalf

Oxford University Press

Abstract and Keywords

In non-technical language, the chapter explains why economists favor putting a price on pollution as the least expensive way to cut pollution to its socially optimal level. Using a pollution example from nineteenth-century British law made famous by the Chicago economist Ronald Coase, the chapter lays out the principles for good environmental policy and shows how a carbon tax fits with those principles. It introduces Arthur C. Pigou, an early twentieth-century economist who saw that a tax on pollution could use the power of the market to solve the pollution problem. It then demonstrates that carbon taxes in British Columbia and Sweden have not harmed their economies but have helped reduce carbon pollution.

Keywords:   Coase Theorem, Pigou, externality, pollution, British Columbia, Sweden

Given the costs of not reducing future carbon pollution, we need to do something to protect ourselves and future generations. It would be great if the efforts of individuals could achieve this goal, but in reality, governments will need to get involved. Why? To answer this question, let’s take a little detour through history to the nineteenth-century English court case Sturges v. Bridgman.

Ronald Coase and the Case of Sturges v. Bridgman

In the 1860s, a confectioner named Bridgman had a shop at 30 Wigmore Street just off Cavendish Square in London where he and his father had been producing candy for more than sixty years. He used two mortars, built into a brick wall in the kitchen at the back of the house, to pulverize loaf sugar and other ingredients with two large wooden pestles. In 1865, a doctor named Sturges bought a home at 85 Wimpole Street, an adjacent side-street, where he opened a medical practice. In 1873, Dr. Sturges built a consulting room at the end of his garden, using a common wall with Bridgman’s shop. That common wall also served as the back wall of Bridgman’s kitchen and the supporting structure for the candy maker’s mortars and pestles.

Not surprisingly, the noise and vibration from the pounding of the pestles interfered with the doctor’s practice. After complaining to Bridgman without success, he brought suit to restrain Bridgman from using the mortars and pestles. The candy maker responded that there had been no issue during the eight years before the new consulting room was built (p.36) and that the problem would have been avoided altogether had the doctor built the consulting room with a separate back wall rather than using the common wall.

This obscure court case is recounted in one of the most important and highly cited articles in economics and is a major reason its author, Ronald H. Coase, was awarded the Nobel Prize in Economics in 1991.1 Coase was an unusual economist who, by his own admission, was weak in mathematics and formal theory. The son of a post-office worker, Coase was raised in a decidedly unacademic household where the major interest was sports. Coase had “weak legs” (as he recounts in his Nobel Prize autobiography), necessitating his wearing leg braces and prompting him to turn to more cerebral pursuits. At age eleven, he was examined by a phrenologist who told him, “You are in possession of much intelligence, and you know it, though you may be inclined to underrate your abilities.” The phrenologist continued, “You will not float down, like a sickly fish, with the tide . . . you enjoy considerable mental vigor and are not a passive instrument in the hands of others. Though you can work with others and for others, where you see it to your advantage, you are more inclined to think and work for yourself. A little more determination would be to your advantage, however.” Whether or not Coase needed more determination, he clearly did not “float down like a sickly fish.”2

Coase recounts the court case of Sturges v. Bridgman to make an important point about externalities. An externality occurs when one person’s action has some impact on another person that is not reflected in the costs borne by the first person. That impact can be beneficial (a positive externality) or, as in this case, harmful (a negative externality). Here, the externality is the noise and vibration coming from Bridgman’s candy making, which has an adverse impact on the doctor who is not able to use his consulting room when the pestles are pounding away. The damages suffered by the poor doctor are not reflected in Bridgman’s costs of candy production.

Pollution is a classic negative externality, and carbon pollution is central to the problem of climate change. The burning of fossil fuels adversely impacts others through the accumulation of greenhouse gases in the atmosphere. In the case of burning fossil fuels, the costs of burning coal in an industrial boiler include the costs of mining and transporting coal to the factory. The costs, however, do not include the damages from the carbon dioxide emissions.

Key to the problem of pollution is the distinction between private costs and social costs. The costs to firms burning fossil fuels are private costs. Private costs reflect the costs borne by the firms as they decide whether and how much oil, natural gas, or coal to burn. These are costs that the firm (p.37) accounts for as it seeks the profit maximizing level of activity. Social costs, in contrast, are the total costs of burning fossil fuels and include both the private costs as well as the damages from the accumulated greenhouse gases in the atmosphere.3 The difference between private and social costs is the damage done by the polluting activity. Firms do not take into account the damages to others as they make their production decisions, including the amount of fossil fuels to consume. To be clear, this is not a value judgment on firm behavior. Firms cannot be expected to include in the costs of running their business the damages from greenhouse gas emissions, especially when those damages may impact people far away from the firm or people not yet born. Publicly owned firms have a fiduciary responsibility to maximize profits for their shareholders. They can’t, of course, fall back on that fiduciary responsibility to justify reckless or illegal behavior. But burning fossil fuels is not illegal. And if costs of burning fossil fuels don’t include the full social costs of their use, that’s a failure of public policy, not of the firm’s ethical and fiduciary obligations.

In the absence of public policy to reduce pollution, markets will overproduce goods or services associated with pollution because the producer does not pay the full cost of its inputs. If the factory owner doesn’t need to pay for the costs of the carbon dioxide added to the atmosphere from burning coal, then he will make decisions about how much coal to burn relying on an artificially low coal price. The price is artificially low because it does not reflect the damages from the carbon dioxide emissions flowing from the factory’s smokestack. If the price reflected the true social cost of burning coal, then the factory would likely burn less coal. The higher price of coal provides the right incentive to factory managers to fine tune their equipment to reduce energy consumption, buy more energy efficient machines, and otherwise cut down on its coal consumption.

Let’s consider the issue of private and social costs for our two nineteenth-century London neighbors. For Bridgman, using the mortar and pestle to grind ingredients and make candy is essential for his business. Let’s assume that if he operates the machine for two hours a day, he can make a small amount of candy that will earn him, after all his costs for ingredients and labor, ₤50. That’s ₤25 for each hour of operation. If he operates more hours, he can produce more candy but he may need to lower the price a bit to sell it all. Table 3.1 shows Bridgman’s hourly profits for each hour of the day. Adding more hours of operation adds to his profits but at a diminishing rate. By the tenth hour, his profits go up only by ₤5 for that last hour of operation. Clearly it is profitable for Bridgman to operate his mortars and pestles ten hours a day. Presumably that’s what he had been doing before Dr. Sturges complained.

Table 3.1. Hypothetical costs of producing candy in Bridgman’s shop

Mortar and Pestle Operation (hours)

Additional Profit to Confectioner (Pounds Sterling)

Additional Damage to Doctor (Pounds Sterling)































Source: Author’s figures based on Sturges v. Bridgman as recounted in Coase (1960).

(p.38) But now let’s consider the doctor’s situation. If the machines operate for just two hours a day, he can rearrange his activities so that he is not inconvenienced and thus does not suffer any damages. But his damages (in the form of lost profits) increase as the machines are used for more of the day. Lost profits rise from zero to ₤15 per hour. Table 3.1 reports damages to Sturges in the final column.

Bridgman’s machines create pollution in the form of noise and vibration that interferes with the doctor’s practice. We typically think of pollution as some contaminant of air or water. But the vibration from the candy maker’s machines is also a form of pollution, albeit not a pollutant that would be regulated today by the US Environmental Protection Agency (EPA). The damages from that pollution are the loss of profits to the doctor. This simple example of the doctor and the candy maker illustrates a number of points that are important for thinking about negative externalities like pollution, the difference between private and social costs, and the socially optimal level of pollution.

Polluters do not have an incentive to voluntarily reduce their pollution. By operating his machinery for ten hours a day, Bridgman could earn profits of ₤50 + 45 + 16 + 15 = ₤126 a day. He clearly has no incentive to cut back on his use of the machines. Even reducing the use of his machines by one hour would reduce his profits by ₤5.

(p.39) Unrestricted pollution generally is not socially optimal. What is privately optimal is not necessarily socially optimal. Bridgman and Sturges are each focused on what is privately optimal for himself. Whatever maximizes each person’s profits is privately optimal. From society’s point of view, what matters is the welfare of everyone. In the case of Bridgman and Sturges, social optimality is achieved by maximizing the sum of profits to Sturges and Bridgman. Since no one else is harmed if Bridgman and Sturges arrange their affairs in a way that makes their joint profits as large as possible, everyone is made better off if we can make the two of them better off. Any increase in the sum total of their profits could be shared in a way that makes both of the two neighbors better off.

Dr. Sturges faces maximal inconvenience and loss of work when the machines are in constant use over the ten-hour work day. While Bridgman’s profits fall by ₤5 if he cuts back the use of his machines by one hour, a single hour of respite from the machines for Dr. Sturges would result in some very high-value work in his office that is worth ₤15. The gain to Sturges from a reduction in the use of the machines of just one hour is triple the loss in profits to Bridgman. Considering the profits of the two of them together, total profits rise by ₤10 if the machines are run for one hour less a day.4 Cutting back the use of Bridgman’s machines—and the consequent reduction in pollution—reduces Bridgman’s profits (a private loss) but increases the sum of their profits (a social gain).

Zero pollution generally is not optimal either. By suing Bridgman, Dr. Sturges hoped to shut the candy maker’s machinery down which means that no candy would be produced and Bridgman, his workers, and his customers would all suffer damages. But consider what happens if Bridgman were allowed to use his mortar and pestle two hours a day. He would earn ₤50 profits. And the loss to Dr. Sturges? Zero. He can rearrange his work in a way so he is not inconvenienced in any way for the two hours the machines are operating behind the common wall. The joint profits of the two businessmen would be increased if even two hours of candy making activity were allowed.

There is a socially optimal level of pollution. Operating the confectioner’s machines for five hours a day will maximize the sum of profits of the doctor and confectioner. The machines operating for two hours generated profits of ₤50 for the confectioner at no cost to the doctor. Operating three more hours adds ₤45 to the confectioner’s profits but costs the doctor ₤15. Total joint profits rise by ₤30 by expanding the hours to five. That’s as far as they should go. Adding one more hour of machine use earns the confectioner only ₤8 but loses the doctor ₤10. Total joint profits start to fall once we run the machines more than five hours.

(p.40) To determine the optimal amount of time for the machines to operate, we start at zero and increase the number of hours until the sum of profits for the two businessmen peak.5 Joint profits of Sturges and Bridgman are maximized at some intermediate level of pollution. Running the machines all day or shutting them down entirely leads to lower total profits for the two men. The optimal level of pollution, in general, will vary from case to case. If the damages from the first bit of pollution were higher than the benefits from that first bit of pollution, then it would be optimal to ban all pollution. Similarly, if the benefits from allowing the last bit of pollution were higher than the damages from that last little bit, then restricting pollution would not be socially desirable. For most pollutants—including greenhouse gas emissions—it is not socially desirable either to ban pollution entirely or to allow an unrestricted amount.

Did the judges take these economic principles into account when they passed down their ruling in Sturges v. Bridgman? Alas, no! The judges weighed the candy maker’s rights to continue operating as he (and his father) had for some sixty years against the doctor’s rights to use his property in peace. In the end, they sided with the doctor and ruled that Bridgman must stop operating his machinery.

Coase’s analysis went a step further, however, to deal with the fact that the doctor was awarded the right to a quiet workplace environment. This right to a quiet work environment is a property right and the assignment of this right to the doctor created an opportunity for Sturges and Bridgman to negotiate among themselves to ensure the socially optimal level of pollution occurs. This is Coase’s key insight, enshrined as the Coase Theorem (although Coase himself never set down the rigorous conditions that would elevate his insight to the level of a theorem).

By winning the lawsuit, Dr. Sturges had been granted the right to be free from the nuisance of Bridgman’s machinery. Coase observed that this is a right that the doctor would be free to exchange (in part or the whole). Assuming Bridgman and Sturges are focused only on the profitability of their respective businesses, a deal could be struck that would make them both better off. Bridgman could come to Sturges and say, “My dear Dr. Sturges. Shutting my machinery down entirely is very costly to me. What if I were to pay you ₤35 daily for the right to run my machinery five hours a day?” Sturges thinks to himself, “Hmmm, I will be slightly inconvenienced to the tune of ₤15 but my neighbor is willing to pay me more than double that amount. We definitely can make a deal to allow him to use his machines for five hours in return for a payment.” After some negotiation, they settle on a payment and Bridgman goes back to his candy making albeit only for five hours a day.

(p.41) Coase’s analysis made clear that we cannot predict what the actual payment would be between the two businessmen. The best we can do is put bounds on it. Since the doctor’s profits fall by ₤15 when the machines operate for five hours a day, ₤15 is the minimum payment that would be acceptable to him. For Bridgman, his profits equal ₤80 when he can operate the machines for five hours a day. If he can’t operate the machines, his profits are zero. So Bridgman would be willing to pay some amount up to ₤80 for the right to use his machines. The actual payment will be between ₤15 and ₤80 with the precise amount depending on the relative bargaining skills of the two men.

Coase’s analysis goes further. Consider if Dr. Sturges had lost the lawsuit and Bridgman continued to use his machines for ten hours a day. After stewing for a few days and drinking many glasses of port, Sturges could approach Bridgman saying, “Mr. Bridgman. You have won the lawsuit and have the right to use your machines as you wish. I will pay you ₤40 a day if you cut back your use of them from ten to five hours. What say you, sir?” Bridgman is a reasonable man and thinks, “Well, my profits will go down by ₤5 an hour for each of the first three hours I cut back and an additional ₤8 pounds an hour for the next two hours. That’s a loss for me of ₤31. But my neighbor is offering me ₤40. We definitely can make a deal where I cut back my use of the machines to five hours in return for a payment.” After some negotiation, they settle on a payment and Bridgman goes back to his candy making albeit only using his machines for five hours a day.

As above, we can only put bounds on the payment. Since Bridgman’s profits will fall by ₤31 per day if he reduces his machine use to five hours a day, he will only entertain offers of at least ₤31. Sturges’s profits go up by ₤65 when the machine use is curtailed. Recall that for the first three hours of respite from the pounding, Sturges could see high-value patients that would earn him profits of ₤15 per hour. That’s ₤45 total. For the next two hours of peace and quiet, he could earn ₤10 per hour. That’s an additional ₤20 for a grand total of ₤65. The most Dr. Sturges would pay for five hours of peace and quiet during the day is ₤65. So whatever deal is struck will include a payment from Sturges to Bridgman between ₤31 and ₤65 a day.

For Coase, the key to ensuring the optimal level of pollution is the clear establishment of property rights. When Sturges wins the lawsuit, he has been granted rights to a quiet and peaceful office space and Bridgman must make an overture to the doctor to obtain permission to use his machinery. If Bridgman wins the lawsuit, he has been granted rights to the free and unfettered use of his machines and it is Dr. Sturges who must make an overture to obtain a few hours of peace and quiet during which he can run his medical practice.

(p.42) In both scenarios, the doctor and candy maker settle on the level of machine use that maximizes their joint profits. Coase argued in his famous paper that this is precisely what would happen. Let’s assume that after the trial, Bridgman makes an offer to Sturges that allows him to use his machines for four hours a day and Sturges accepts the offer. After making that agreement, it is easy to see that there is an opportunity for a mutually beneficial trade that they are passing up. Sturges might come to Bridgman and say, “I understand that if you were allowed to use your machines for one more hour a day, your profits would go up by ₤15. My profits would fall by ₤5. Pay me an additional ₤10 and I’ll let you use your machines for five hours a day instead of four.” Since Bridgman’s profits (after the payment) will rise by ₤5, he agrees and Bridgman uses his machines for five hours a day.

At any number of hours of operation other than five, there is some payment between the two neighbors that would increase each of their profits if the machine use can be raised or lowered to five hours a day. But the same is not true when the machines are operating for five hours a day. Increasing to six hours would raise the confectioner’s profits by ₤8 but cost the doctor ₤10 in lost profits. There is no payment that is acceptable to both parties to increase the machine use from five to six hours. Similarly, reducing from five to four is not mutually beneficial. Doing so lowers Bridgman’s profits by ₤15 but gains the doctor only ₤5. Again, there is no payment acceptable to both parties to decrease the machine use from five to four hours.

For a generation of economists schooled in the tradition that government intervention is the only way to address pollution, Coase’s argument that private parties could efficiently deal with pollution on their own left them dumbstruck. Coase included the argument in a paper about the Federal Communications Commission that he submitted to the Journal of Law and Economics, the premier academic journal focused on the application of economic principles to legal analysis published by scholars at the University of Chicago. The economists at Chicago—one of the top economics departments in the world—could not believe that “so fine an economist could make such an obvious mistake,” in the words of George Stigler, himself a winner of the Nobel Prize in Economics. In a sign of how economics was done in the old days (the late 1950s), they invited Coase to visit Chicago and discuss his paper over dinner at the home of Aaron Director, editor of the journal and himself a renowned economist. As Stigler recounted later, “In the course of two hours of argument, the vote went from twenty against and one for Coase, to twenty-one for Coase.” Coase was asked to elaborate his arguments in a paper that became “The Problem of Social Cost.” That was not the end of it. Coase was subsequently (p.43) invited to join the Department of Economics at Chicago where he spent the rest of his illustrious career.6

Coase’s insight is that excessive pollution results from a failure to assign clear and enforceable property rights. Assigning and protecting property rights is a key role of government. But that can be easier said than done at times. And so, there is one more insight from his theorem that is critical for thinking about greenhouse gas emissions.

The Coase Theorem requires low barriers to negotiation. A key condition for Coasian bargaining to lead to an efficient outcome is that it can’t be too difficult for the affected parties to work out a deal among themselves. Bridgman and Sturges were neighbors and knew each other. Only the two of them were affected by the use (or non-use) of the machines. Moreover, each knows (or has reason to believe) the change in profits for the other party as machine use is changed and that any offer made by either of the parties is credible and will be upheld.

The Coase Theorem is often dismissed as just the sort of writing that would come from economists at the University of Chicago, the bastion of conservative economic thinking and free-market principles—think Milton Friedman, another Nobel Laureate, self-avowed libertarian, and author of the wildly popular book Free To Choose—a manifesto of sorts in the Reagan era with its call for less government intervention in the economy.

That dismissive argument is unfair to Coase. Coase was responding to the conventional view among economists at the time: If there is an externality, then government regulation is called for. Coase’s view was more nuanced. Clearly, the sort of bargaining imagined between Bridgman and Sturges is utterly unrealistic for a situation where billions of people are responsible for pollution that in turn affects billions of others. Coase acknowledged as much in his paper. He wrote that in the “standard case of a [pollutant] which may affect a vast number of people engaged in a wide variety of activities,” private negotiations to optimally reduce the pollutant are impossible. In such a case, Coase argued, the solution may well be government regulation or a tax.

The notion of applying Coasian bargaining at the individual level to reduce greenhouse gas emissions is a fanciful idea. How would we do it? Consider a neighboring natural gas fired power plant generating electricity. With whom would the plant owners negotiate? Climate change has global impacts; it would be impossible to coordinate dealmaking between the world’s population and this plant. Moreover, because carbon remains in the atmosphere for centuries, many of the damages will be suffered by future generations. Who is going to negotiate for them? An obvious answer is the government. But this would be very cumbersome, especially (p.44) since there are nearly two thousand natural gas and about four hundred coal fired power plants in the United States. There are also another two hundred and sixty million vehicles on the road burning gasoline or diesel fuel.7 There must be a better way to reduce emissions optimally. There is. And for that we can thank another great twentieth-century British economist, Arthur Pigou.

Using Prices to Internalize the Externality

Born in 1877 on the Isle of Wight, Arthur Cecil Pigou entered King’s College, Cambridge, in 1896 where he took a first class in the history tripos—the undergraduate exams given at Cambridge—followed by a first in the moral sciences tripos in 1900. He came to economics rather late, having first submitted a thesis on “Browning as a Religious Teacher” in an unsuccessful bid for a fellowship at King’s College. He had better luck in economics: he began lecturing in economics at King’s in 1901 and in 1908 took the chair of political economy previously held by the great economist Alfred Marshall, arguably the developer of modern neoclassical economics. Pigou’s reputation was not assured at first. In a short biography, David Collard (2004) writes that Pigou’s “publications record at the time [of his appointment to the chair] was respectable though not substantial; he was certainly appointed very much on promise rather than performance.” Collard notes that Pigou “loved practical jokes . . . and often affected a silly high-pitched voice. He was an enthusiastic rock-climber and led parties of undergraduates both in the Alps and in the Lake District.” But he had a serious side as well. A committed pacifist, he volunteered as an ambulance driver in World War I and frequently put himself in danger on the front lines.

A brilliant debater and highly organized lecturer, Pigou made his mark with his 1920 book The Economics of Welfare, in which he emphasized the critical distinction between private and social benefits from which arises the modern idea of an externality. But Pigou went beyond articulating the idea of an externality. He came up with an elegant solution to the problem of externalities that harnessed the power of markets.

Pigou had the following insight. Pollution creates a cost for someone that is not paid by the polluter. That cost is the damages caused by pollution and the adaptation expenses to cope with the impacts of climate change. If pollution creates a cost, argues Pigou, then government should impose a tax on the polluting activity. Return to the example of Bridgman and Sturges and imagine that Pigou was the judge in the case. Given the facts, (p.45) he might have ruled as follows: “It is neither fair to subject Dr. Sturges to unrestricted pollution nor is it fair to deprive Mr. Bridgman of his right to operate his business. I will set a tax on Bridgman’s use of his machines that balances his profits from using the machines against Dr. Sturges’s loss of profits from their use. I will set a tax of ₤10 per hour of machine operation. Revenues from the tax will be used to lower the city property tax.”

How does Bridgman react to this tax? Bridgman’s costs of using the machines have just risen by ₤10 for each hour of operation. Bridgman now does a mental calculation: he earns ₤25 in profits for each of his first two hours of machine use. Even after paying the tax, he will earn a profit. His hourly profit for the next three hours of use is ₤15. He still earns a profit by expanding his machine use from two to five hours. But if he runs the machines another hour, he’ll earn an additional ₤8. This won’t be enough to cover the ₤10 tax he’ll have to pay, so it’s not worth running the machines the extra sixth hour. Rather, he runs the machines for five hours and pays a total tax of ₤50. The noise and vibration pollution from the pounding mortars and pestles has been reduced to the level that maximizes the total profits of the two businessmen.

Applying a tax in Sturges v. Bridgman is a bit silly. But it illustrates how a tax can be used to reduce pollution to the socially optimal level. For greenhouse gas emissions in general and carbon dioxide emissions in particular, a carbon tax can nudge polluters to the optimal level of emissions by confronting fossil fuel users with the full social costs of burning these fuels. Those include the private costs of mining coal or drilling for natural gas or oil, processing and transporting the fuels, storing them, and finally burning them. The full social cost also includes the damages to our planet from burning those fuels and adding more and more greenhouse gases to the atmosphere.

Part of the appeal of a carbon tax is that it uses markets to solve the pollution problem. Rather than mandating new carbon-free technologies or subsidizing clean-burning fuels, a carbon tax acts as a price on pollution that steers millions of people toward activities and purchases that reduce our carbon footprint. As the carbon tax raises the cost of burning fossil fuels, people react by buying more fuel-efficient cars and investing in more home insulation and higher efficiency furnaces, among other things. Factories purchase more fuel-efficient boilers and source less energy-intensive intermediate goods used in production, along with a host of other activities.

Conservatives often talk about the power of markets and the way that Adam Smith’s “invisible hand” leads us to efficient outcomes. Where pollution is involved, Pigou’s insight ensures that the invisible hand has a green thumb!

(p.46) Measuring the Costs of a Carbon Tax

A carbon tax has costs. That should not be surprising. After all, the payoff to the carbon tax is a cleaner environment and reduced damages to future generations. We should not expect to get those benefits for free. Those costs include investments in new energy-efficient technologies, shifts in modes of transportation (e.g., to electric or hydrogen-powered motor vehicles), potentially higher fuel costs, and so on. But, it is important to emphasize that the cost to society of the carbon tax is not the tax payment itself. The tax itself is a transfer from one group of people to another group. A tax applied to the Bridgman v. Sturges case illustrates the point.

Imagine the government levies Pigou’s tax on Bridgman for every hour of machine use and uses the revenue to lower London’s property tax. The ₤50 collected from Bridgman benefits all London property tax payers, including Bridgman, by lowering their aggregate taxes by the same amount. This is a pure redistribution from one group to another. While the ₤50 is clearly a cost to Bridgman, it is not a cost to London taxpayers as a group since the net amount of taxes collected is unchanged.

The economic cost of the tax is the reduction in Bridgman’s profits as he curtails the use of his machines five hours a day, a loss of profits equal to ₤31.8 What about the benefits to Dr. Sturges in being able to use his examining room. With the tax in place and the machine use reduced from ten to five hours a day, the doctor’s profits increase by ₤65, more than outweighing the loss of profits to Bridgman. The increase in the doctor’s profits is the benefit of the tax. What this example shows is that the net benefit (benefit minus cost) of the tax is positive. The tax has no aggregate social costs at all! If it is optimal to limit pollution, then society will be better off by doing so. That is, net benefits to society will rise. That is the whole point of imposing the tax.

Why do we say that taxes are costly then? For one, taxes are clearly costly to the taxpayer. Consider how Bridgman would feel if subjected to a Pigouvian tax to get him to reduce his machine use to the socially optimal amount of time. Not only does he pay ₤50 in taxes, his profits fall by ₤31 relative to when he operated for a full ten-hour workday. It is cold comfort to Bridgman that the ₤50 tax bill itself makes London property taxpayers better off (even if he himself also benefits in a small way from a lower property tax). He cares about his taxes, not any tax reduction enjoyed by others. Plus, he might argue, the tax reduction for others is miniscule in comparison to the tax he has to pay. All this is true but the fact remains that overall tax collections are unchanged: Bridgman’s taxes went up while taxes for others went down. It is easy to forget this. People paying a tax (p.47) tend to be especially vocal in their opposition to it, while those who benefit by only a small amount are not likely to speak up—it’s too small a sum to get exercised about.9

Accounting for how tax proceeds are used is especially important for a carbon tax. A carbon tax of $50 per ton of carbon dioxide would raise approximately $200 billion annually.10 If simply given back to every man, woman, and child in the United States through an income tax cut, a family of four would receive $2,500 as a “carbon dividend.” One could justify such a tax rebate as a Coasian property rights payment to citizens for the use of the atmosphere to store accumulating greenhouse gases. Such a rebate would more than offset the higher costs of goods and services due to the carbon tax for some 70 percent of US residents. While a potentially politically appealing bargain, the rebate is spread across millions of people while the impact of the tax is concentrated on a much narrower group of people, those who own, or work in, energy-intensive industries including, for example, the roughly fifty thousand workers employed in coal mining.

That a Pigouvian tax applied to Bridgman’s candy business improves societal welfare is clear in our simple example because we can easily account for the costs (lost profits to Bridgman) and benefits (increased profits to Sturges) from reducing pollution. It is much more difficult to sum up all the costs and benefits of a carbon tax. While the costs of reducing emissions are reasonably straightforward to calculate, it is difficult to measure the full benefits. For one thing, many of the benefits—avoided permafrost melt in the Arctic region, avoided sea-level rise around the world—are difficult to monetize. What is the value of the loss of polar bears, for example, if the Arctic region is fundamentally changed? Will polar bears go extinct? What about the disruptions of sea-level rise? To the extent that sea-level rise leads to shifts in living patterns and mass migrations, it is extraordinarily difficult to measure the dollar value of these disruptions. Moreover, the bulk of benefits from reducing emissions will go to unborn generations. This complicates the task of measuring the benefits of reducing greenhouse gas emissions now.

Because of the difficulty in measuring the benefits of reducing emissions, economists have focused considerable attention on the question of whether there are benefits to carbon taxes other than emission reductions. This has given rise to the so-called Double Dividend literature. The idea of a double dividend is straightforward. The first dividend from a carbon tax is the environmental improvement from reducing pollution. The second dividend is the revenue from the carbon tax that can be used to lower income tax rates. Our income tax discourages work and savings by individuals, encourages people to take out mortgages to buy homes (or subsidizes larger (p.48) mortgages to buy larger homes), and affects firms’ investment decisions. These are examples of what economists call tax distortions. A vast literature exists on the shortcomings of our tax system and how it distorts the behavior of taxpayers as they engage in various activities to reduce their tax burden.11 Having a tax system that discourages activities we like (e.g., jobs, saving, and investment) means there is the possibility of a second benefit (or dividend) from a carbon tax. If the carbon tax revenue is used to lower a distorting tax, then we can offset some of the costs of our tax system. In fact, if the tax system is especially messed up, then it is possible that you could get a strong double dividend, where the reductions in costs by lowering especially distorting taxes are greater than the costs to firms from the carbon tax. Then a carbon tax makes society better off even when we ignore the environmental benefits. Alas, while an appealing idea, there is little evidence to support the idea of a strong double dividend from a carbon tax.12

While the strong double dividend might not exist, it is certainly true that using the carbon tax revenue to lower those taxes that are especially distorting could contribute to stronger economic growth. The most distorting taxes, in general, are taxes on capital income (e.g., the corporate income tax and taxes on dividends and capital gains). For the largest possible double dividend, we’d use carbon tax revenue to cut taxes on capital income. But cutting taxes on capital income disproportionately benefits high income people. From a fairness point of view, we may feel that carbon tax revenue should benefit a broader group of individuals than those at the top of the income and wealth heap. This illustrates the political trade-off between economic efficiency and fairness. It is worth noting, moreover, that any argument for lowering the US corporate income tax has become even less persuasive after the 2017 tax cut lowered the top corporate rate from 35 to 21 percent, its lowest level since 1937.13

From Theory to Practice: Carbon Taxes in Action

It is one thing to advocate a carbon tax in theory. It’s another to see how it operates in practice. Carbon taxes have been used by countries and subnational governments for more than twenty-five years. In 2017, sixteen national or subnational carbon taxes were in effect with another three added in 2018.14 There have been two waves of carbon tax enactments: a Scandinavian wave starting in the early 1990s saw carbon taxes legislated in Denmark, Finland, Norway, and Sweden, among other countries. By 2000, seven countries had a carbon tax. A second wave in the mid-2000s saw (p.49) carbon taxes put in place in Switzerland, Iceland, Ireland, Japan, Mexico, and Portugal. In addition, the Canadian provinces of British Columbia and Alberta have enacted carbon taxes.

The carbon taxes in Sweden and British Columbia are worth a closer examination as they reveal quite a bit about how a carbon tax might affect the US economy. Sweden enacted its tax in 1991 with a “standard” rate that applied to households and the commercial sector. Industry and agriculture were taxed at a lower rate, given concerns about international competition. The general tax rate in 2018 is just under $130 per metric ton of carbon dioxide, and the industry discount has been eliminated.15

A 2016 report by the Heritage Foundation that analyzed a carbon tax in the United States at one-third the rate paid by Swedish industrial firms argued that the tax would have “devastating economic costs.”16 Has Sweden been devastated by its carbon tax? Hardly. Prior to enacting the tax, Sweden’s economy grew between 1961 and 1990 at an annual rate of 2.9 percent while the US economy grew at 3.7 percent. The gap between the Swedish and US growth rates fell after the tax was enacted. Sweden’s GDP grew between 1991 and 2015 at an average annual rate of 2.1 percent, just a bit lower than the US growth rate of 2.4 percent. This is not to argue that Sweden closed the gap between its growth rate and the US growth rate because of its carbon tax. But it is hard to argue that the country has been devastated by its carbon tax. In fact, Sweden’s growth rate has exceeded the US growth rate since 2000.17 Sweden’s carbon tax accounts for 1.6 percent of total tax collections—equivalent to two-thirds of 1 percent of GDP. Based on its share of GDP, Sweden’s carbon tax revenue would be comparable to a US carbon tax that raised $120 billion annually.

Why has Sweden thrived despite having the highest carbon tax rate in the world? Part of the success of the Swedish carbon tax is its use of revenue. From its beginnings, the carbon tax was part of a broader tax reform that reduced labor taxes through increases in taxes on energy use, including the carbon tax. This green tax shift contributed to economic growth and a decoupling of economic growth and emissions. Decoupling refers to the breakdown of the historic link between economic growth and carbon emissions. As economies grow, so do emissions. But as countries get richer, the demand for environmental quality grows and the growth–emissions link can be broken.18 The link has clearly been broken in Sweden’s case. Figure 3.1 shows that while real GDP grew by three-quarters between 1990 and 2016, Sweden’s emissions fell by one-quarter.

Why Do Economists Like a Carbon Tax?

Figure 3.1. Decoupling of economic growth and emissions in Sweden, 1990–2016

Source: World Bank and Statistics Sweden

Sweden is just one example. An example closer to home is the Canadian province of British Columbia, where the carbon tax is closer in structure to what the United States might consider. Sweden’s carbon tax applied only to (p.50) fossil fuels used by the residential sector plus industrial firms not subject to carbon pricing by other means.

As part of a broader package of tax reforms, British Columbia enacted a broad-based carbon tax in 2008 starting at $10 (Canadian) per metric ton of carbon dioxide and increasing by $5 per year to its current $35 in Canadian dollars (as of 2018). That’s equivalent to $27 in US dollars.19 The tax is scheduled to increase by $5 (Canadian) per year until it reaches $50 per ton in 2021. The tax is a broad-based tax on the carbon emissions of all hydrocarbon fuels burned in the province. Given the existing federal and provincial taxes already in place, the carbon tax raised the overall excise tax on gasoline by roughly one-fifth.

The tax collects over $1 billion annually—over 5 percent of provincial tax collections—and all the revenue is returned to businesses and households through a combination of tax-rate reductions, grants to businesses and households, and other business tax breaks. Worried that the new carbon tax would disproportionately affect low-income households, policymakers included several elements in the tax reform to offset adverse impacts on them. One element was a low-income climate action tax credit of $115.50 (p.51) per adult plus $34.50 per child, which reduced taxes by $300 for a low-income family of four. In addition, tax rates in the lowest two tax brackets were reduced by five percentage points. Also, in the first year of the carbon tax, there was a one-time “climate action dividend” of $100 for every BC resident. This equal-sized dividend represents a greater share of the disposable income of low-income households than of higher income households.

Meanwhile, business tax rates were cut. The tax rate for small businesses, for example, was cut from 4.5 percent to 2.5 percent in 2008. As the carbon tax rate rose from $10 to $20, there was more carbon tax revenue to rebate, much of which was channeled to businesses in the form of new business tax credits.

British Columbia’s carefully constructed policy package to return tax revenue to its residents and businesses balanced concerns about distributional impacts and economic growth. Targeting tax cuts to low-income households ensured the burden of the tax wouldn’t fall disproportionately on those households. And the focus on small business emphasized the importance of supporting economic growth.

The evidence strongly supports the conclusion that neither economic growth nor employment were adversely affected by the tax. One analysis comparing economic growth in British Columbia to that of other provinces finds that growth was unaffected by the tax after controlling for differences across the Canadian provinces. Another study found that overall employment increased given the growth in the service sectors stimulated by the increase in consumption by BC residents upon receiving carbon rebates.20 While jobs were shed in energy-intensive sectors, those lost jobs were more than offset with new jobs in sectors that use little energy (services, for example). This is precisely what the carbon tax’s price signal is expected to do: redirect jobs away from carbon polluting sectors to nonpolluting sectors. The reallocation of new jobs in British Columbia is an example of the power of prices to reshape an economy.

Balancing the carbon tax with decreases in personal and corporate income tax rates helped counteract any negative economic impacts of the tax. The evidence for the tax’s impact on reducing the use of fossil fuels is even more clear-cut. British Columbia experienced a sharp drop in fossil fuel consumption at a time when per capita fuel consumption was rising in the rest of Canada.21

The evolution of public attitude toward the tax has been documented by the Canadian political scientist Kathryn Harrison. She shows that attitudes toward the carbon tax have swung sharply from majority opposition to support since the middle of 2011. Harrison (2013) notes a few factors that may have contributed to this swing of opinion including 1) growing acceptance (p.52) that the tax was “here to stay”; 2) less media attention on the tax; and 3) a growing recognition that eliminating the tax would create a budget shortfall of roughly 3 percent that would be difficult to make up. She also notes that over time the rebated tax revenue has shifted from rebates to people (and especially low-income households) to tax reductions that favor “more specific, and presumably more attentive, subpopulations” such as business groups receiving special tax breaks. This suggests a coalition is emerging to support maintaining the tax.

British Columbia’s carbon tax is a case study of a revenue neutral tax reform where the revenue is used to lower other taxes so that overall tax collections are unchanged. It was a politically savvy approach since BC residents couldn’t argue that the tax would simply increase their overall tax burden. In returning the revenue in a visible and transparent way, policymakers undercut a potential opposition to the tax.

British Columbia’s experience with a carbon tax informed Canada’s roll-out of a national carbon policy. In October 2016, the Canadian federal government extended carbon pricing to all Canadian provinces. Each province is free to establish a carbon pricing approach as it sees fit so long as it satisfies criteria laid out in the federal pricing system. Specifically, pricing has to apply broadly and be at or above a federal benchmark of $10 per metric ton (Canadian) in 2017; the benchmark rises by $10 a year until it hits $50 in 2022. Provinces that don’t adhere to federal guidelines are subject to a federal carbon levy so that carbon pollution is priced in every province.22 The federal approach honors Canada’s commitment to its federal structure and provincial rights while ensuring national action is taken and consistent pricing achieved across Canadian provinces.

Taxing carbon pollution is an idea solidly grounded in economic theory. A tax on carbon pollution uses the power of markets to shift behavior, encourage energy savings, incentivize clean energy innovation, and raise money that we can use to good purpose elsewhere in the federal budget. Despite all this, it’s clear that people have a visceral dislike of taxes. This dislike can be traced to various factors, including the complexity of the current tax code and a sense that others are not paying their fair share of taxes.23 The question, however, is how best to achieve the goal of reducing our carbon emissions and avoiding the damages from those on-going emissions. The best policy is the policy that balances simplicity, efficiency, and fairness. Pointing to the problems with taxes without looking at the drawbacks of the alternatives is unfair to tax policy. Or as Winston Churchill noted in another context, “[m]any forms of government have been tried . . . Indeed, it has been said that democracy is the worst form of Government except for all those other forms that have been tried . . .”24


(1.) Coase’s 1960 paper, “The Problem of Social Cost,” has over 30,000 citations on Google Scholar, more than double the citations of the top-ranked papers by other Nobel laureates such as Joseph Stiglitz, Paul Krugman, and Paul Samuelson. Douglas Allen has a precis of the case at www.sfu.ca/~allen/Sturges v Bridgman.doc, accessed May 9, 2017.

(2.) Coase recounts his life story in the autobiography he provided to the Nobel Committee after receiving the 1991 Nobel Prize in Economics, http://www.nobelprize.org/nobel_prizes/economic-sciences/laureates/1991/coase-bio.html, accessed May 11, 2017.

(3.) Burning fossil fuels also creates local pollutants such as ozone and particulate matter that have damaging health effects. Those are ignored here. But any policy to reduce carbon pollution would have the additional benefit of reducing local pollutants.

(4.) The statement that something is not optimal suggests there is something we are trying to optimize. In this example it is the joint profits of the confectioner and doctor. But this statement holds regardless of what we are trying to optimize, whether it is the sum of profits in society, consumer well-being, or some other societal goal.

(5.) This is an example of marginal analysis which looks at the effects on joint profits of an increase or decrease of one unit of production (a marginal change).

(6.) This anecdote is recounted in Coase’s obituary written by Lyons (2013).

(7.) Power plant statistics for 2015 from the US Energy Information Administration (EIA)’s Electric Power Annual, Table 4.1, https://www.eia.gov/electricity/annual/. Vehicle statistics from the Bureau of Transportation Statistics publication National Transportation Statistics, Table 1.11, https://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/national_transportation_statistics/index.html, both accessed May 13, 2017.

(8.) But the cost could be lower. Bridgman might, for example, invest in new sound and vibration proofing along his back wall and install quieter machinery to grind his ingredients and so avoid having to pay a tax at all. In that case, he could continue to profitably operate ten hours a day. Let’s assume the cost of the investments (converted to a daily cost) is ₤10. Then the cost of addressing this noise pollution is ₤10 rather than ₤31 loss in profits.

(9.) Mancur Olson developed this idea in his famous book The Logic of Collective Action. It explains how it can be difficult to enact policies with diffuse benefits and concentrated costs. Since benefits are diffuse, those benefiting from the policy have little incentive to engage in the political process to support the policy while those who incur the costs have a large incentive to defeat the policy. This (p.148) speaks to the importance of addressing the costs to sectors such as coal mining if a carbon tax is to be politically successful.

(10.) The revenue estimate is taken from the US Department of the Treasury (Treasury) working paper written by Horowitz et al. (2017).

(11.) Books that do a good job of that include, among many others, Slemrod and Bakija (2017) and Gale (2018).

(12.) The double dividend literature is reviewed by Goulder (1995), Fullerton and Metcalf (1998), and Bovenberg (1999), among others.

(13.) Historic corporate income tax rates are reported at https://taxfoundation.org/federal-corporate-income-tax-rates-income-years-1909-2012/, accessed May 13, 2017.

(14.) Existing and planned carbon tax regimes are summarized in World Bank Group (2016).

(15.) A metric ton is 1,000 kilograms or roughly 2,200 pounds. Tax rates from Swedish Ministry of Environment, “Twenty Years of Carbon Pricing in Sweden” and presentation by Ulrika Raab, Senior Advisor of the Swedish Energy Agency at the PMR Technical Workshop on Carbon Tax: Design and Implementation in Practice on March 22, 2017. Swedish krona converted to US dollars using an exchange rate (SK 1 = USD 0.11) as of May 2018.

(16.) The Heritage Report written by Dayaaratna et al. (2016).

(17.) GDP growth data from the World Bank DataBank, http://data.worldbank.org/indicator/NY.GDP.MKTP.KD.ZG, accessed May 8, 2018.

(18.) The idea that pollution initially increases with a country’s income but then peaks and begins to decline is known as the Environmental Kuznets Curve and was first posited by Grossman and Krueger (1995).

(19.) All currency conversions to US dollars (CD1 = USD 0.78) use exchange rates as of late May 2018. Information about the tax rate taken from https://www2.gov.bc.ca/gov/content/environment/climate-change/planning-and-action/carbon-tax, accessed May 23, 2018.

(20.) Metcalf (2016) carries out a statistical analysis of the tax’s effect on economic growth in the province. Yamazaki (2017) shows that while energy-intensive sectors shed workers, employment overall grew given the more-than-offsetting growth in sectors not so reliant on energy.

(21.) See the studies by Elgie and McClay (2013), Elgie (2014), and Rivers and Schaufele (2015). Rivers and Schaufele argue that the drop in fossil fuel consumption is greater than one would expect given the typical response of energy demand to changes in fuel prices and conclude that the salience of the tax is affecting behavior. Tax salience has been shown to have significant effects in studies of other taxes as well. Murray and Rivers (2015) argue that the tax has cut BC emissions by between 5 and 15 percent. A recent study by Ahmadi (2016), however, finds no statistical support for the tax having reduced emissions. His is a stringent and conservative test that is biased towards finding no evidence of an impact on emissions.

(22.) As detailed in a technical document published by Environment and Climate Change Canada (2017), the levy would be paid by fuel distributors and natural gas retailers. The levy on industry is on emissions above a floor based on a sector-specific emissions standard (tons of carbon dioxide equivalent per unit of production) times a firm’s output. This is to provide some competitiveness protection to firms while maintaining the incentive at the margin to reduce emissions.

(23.) Complexity and fairness were the top issues identified by taxpayers in a survey by the Pew Research Center (2017).

(24.) Churchill speech in the House of Commons on Nov. 11, 1947, http://hansard.millbanksystems.com/commons/1947/nov/11/parliament-bill#S5CV0444P0_19471111_HOC_312, accessed May 13, 2017.