Pollution is a bad thing. Yet most pollution is a side effect of activities that provide us with good things: our air is polluted by power plants generating the electricity that lights our cities, and our rivers are damaged by fertilizer runoff from farms that grow our food. Why shouldn’t we accept a certain amount of pollution as the cost of a good life?

Actually, we do. Even highly committed environmentalists don’t think that we can or should completely eliminate pollution—even an environmentally conscious society would accept some pollution as the cost of producing useful goods and services. What environmentalists argue is that unless there is a strong and effective environmental policy, our society will generate too much pollution—too much of a bad thing. And the great majority of economists agree.

To see why, we need a framework that lets us think about how much pollution a society should have. We’ll then be able to see why a market economy, left to itself, will produce more pollution than it should. We’ll start by adopting the simplest framework to study the problem—assuming that the amount of pollution emitted by a polluter is directly observable and controllable.

How much pollution should society allow? We learned in Chapter 9 that “how much” decisions always involve comparing the marginal benefit from an additional unit of something with the marginal cost of that additional unit. The same is true of pollution.

The marginal social cost (MSC) of pollution is the additional cost imposed on society as a whole by an additional unit of pollution. For example, acid rain damages fisheries, crops, and forests, and each additional tonne of sulphur dioxide released into the atmosphere increases the damage.

The marginal social benefit (MSB) of pollution—the additional benefit to society from an additional unit of pollution—may seem like a confusing concept. What’s good about pollution? However, avoiding pollution requires using scarce resources that could have been used to produce other goods and services. For example, to reduce the quantity of sulphur dioxide they emit, power companies must either buy expensive low-sulphur coal or install special scrubbers to remove sulphur from their emissions.¹ The more sulphur dioxide they are allowed to emit, the lower these extra costs. Suppose we could calculate how much money the power industry would save if it were allowed to emit an additional tonne of sulphur dioxide. That savings would be the marginal benefit to society of emitting an extra tonne of sulphur dioxide.

The marginal social cost of pollution can be expressed as the sum of two terms (MSC = MPC + MEC): marginal private cost (MPC) of pollution—the additional cost imposed on the polluter if the polluter creates another unit of pollution—and marginal external cost (MEC) of pollution—the additional cost imposed on others if the polluter creates another unit of pollution. If, as is often the case, creating another unit of pollution imposes zero additional costs on the polluter (i.e., the pollution only harms others in society), then the marginal social cost is equal to the marginal external cost (MSC = MEC, as MPC = 0).

Similarly, the marginal social benefit is the sum of two terms (MSB = MPB + MEB): marginal private benefit (MPB) of pollution—the additional benefit the polluter receives if the polluter creates another unit of pollution—and marginal external benefit (MEB) of pollution—the additional benefit others receive if the polluter creates another unit of pollution. If creating another unit of pollution only creates additional benefits for the polluter, the marginal social benefit is equal to the marginal private benefit (MSB = MPB, as MEB = 0).

Using hypothetical numbers, Figure 16-1 shows how we can determine the socially optimal quantity of pollution—the quantity of pollution society would choose if all its costs and benefits were fully accounted for. The upward-sloping marginal social cost curve, MSC, shows how the marginal cost to society of an additional tonne of pollution emissions varies with the quantity of emissions. (An upward slope is likely because nature can often safely handle low levels of pollution but is increasingly harmed as pollution reaches high levels.) The marginal social benefit curve, MSB, is downward sloping because it is progressively harder, and therefore more expensive, to achieve a further reduction in pollution as the total amount of pollution falls—increasingly more expensive technology must be used. As a result, as total pollution falls, the cost savings to a polluter of being allowed to emit one more tonne rises.

The socially optimal quantity of pollution in this example isn’t zero. It’s Q_OPT, the quantity corresponding to point O, where MSB crosses MSC. At Q_OPT, the marginal social benefit from an additional tonne of emissions and its marginal social cost are equalized at $200.

But will a market economy, left to itself, arrive at the socially optimal quantity of pollution? No, it won’t.

Pollution yields both benefits and costs to society. But in a market economy without government intervention, those who benefit from pollution—like the owners of power companies—decide how much pollution occurs. They have no incentive to take into account the costs of pollution that they impose on others.

To see why, remember the nature of the benefits and costs from pollution. For polluters, the benefits take the form of monetary savings: by emitting an extra tonne of sulphur dioxide, any given polluter saves the cost of buying expensive, low-sulphur coal or installing pollution-control equipment. So the benefits of pollution accrue directly to the polluters. (This means MEB = 0, so MSB = MPB).

The costs of pollution, though, fall on people who have no say in the decision about how much pollution takes place: for example, about half of all airborne sulphur dioxide in eastern Canada comes from sources in the United States, especially coal burning power plants in and around the Tennessee Valley. People who fish in Canada do not control the decisions of power plants in Tennessee. (This means MPC = 0, so MSC = MEC > 0. A competitive market equilibrium will be where MPB = MPC = 0, which means too much pollution will be created.)

Figure 16-2 shows the result of this asymmetry between who reaps the benefits and who pays the costs. In a market economy without government intervention to protect the environment, only the benefits of pollution are taken into account in choosing the quantity of pollution. So the quantity of emissions won’t be the socially optimal quantity Q_OPT; it will be Q_MKT, the quantity at which the marginal social benefit of an additional tonne of pollution is zero, but the marginal social cost of that additional tonne is much larger—$400. The quantity of pollution in a market economy without government intervention will be higher than its socially optimal quantity. (The Pigouvian tax noted in Figure 16-2 will be explained shortly.)

The reason is that in the absence of government intervention, those who derive the benefits from pollution—in this case, the owners of power plants—don’t have to compensate those who bear the costs. So the marginal private cost of pollution to any given polluter is zero: polluters have no incentive to limit the amount of emissions. For example, before the U.S. Clean Air Act of 1970, power plants in the U.S. Midwest used the cheapest type of coal available, despite the fact that cheap coal generated more pollution, and they did nothing to scrub their emissions.

The environmental costs of pollution are the best-known and most important example of an external cost—an uncompensated cost that an individual or firm imposes on others. There are many other examples of external costs besides pollution. Another important, and certainly very familiar, external cost is traffic congestion—an individual who chooses to drive during rush hour increases congestion and so increases the travel time of other drivers.

We’ll see later in this chapter that there are also important examples of external benefits, benefits that individuals or firms confer on others without receiving compensation. External costs and benefits are jointly known as externalities with external costs called negative externalities and external benefits called positive externalities.

As we’ve already suggested, externalities can lead to individual decisions that are not optimal for society as a whole. Let’s take a closer look at why.

We have just shown that in the absence of government action, the quantity of pollution will be inefficient: polluters will pollute up to the point at which the marginal social benefit of pollution is zero, as shown by the pollution quantity, Q_MKT, in Figure 16-2. Recall that an outcome is efficient if no one could be made better off without making someone else worse off. In Chapter 4 we showed why the market equilibrium quantity in a perfectly competitive market is the efficient quantity of the good, the quantity that maximizes total surplus. Here, we can use a variation of that analysis to show how the presence of a negative externality upsets that result.

Because the marginal social benefit of pollution is zero at Q_MKT, reducing the quantity of pollution by one tonne would subtract very little from the total social benefit from pollution. In other words, the benefit to polluters from that last unit of pollution is very low—virtually zero. Meanwhile, the marginal social cost imposed on the rest of society of that last tonne of pollution at Q_MKT is quite high—$400. In other words, by reducing the quantity of pollution at Q_MKT by one tonne, the total social cost of pollution falls by $400, but total social benefit falls by virtually zero. So total surplus rises by approximately $400 if the quantity of pollution at Q_MKT is reduced by one tonne.

If the quantity of pollution is reduced further, there will be more gains in total surplus, though they will be smaller. For example, if the quantity of pollution is Q_H in Figure 16-2, the marginal social benefit of a tonne of pollution is $100, but the marginal social cost is still $300. In other words, reducing the quantity of pollution by one tonne leads to a net gain in total surplus of approximately $300 – $100 = $200. This tells us that Q_H is still an inefficiently high quantity of pollution. Only if the quantity of pollution is reduced to Q_OPT, where the marginal social cost and the marginal social benefit of an additional tonne of pollution are both $200, is the outcome efficient.

Pollution is a by-product of either the manufacture or consumption of goods and services—it is not a product that is directly produced. This implies that another way to study the social welfare impact of pollution is to look at pollution indirectly by investigating the social costs and benefits in the market for the good or service that causes the pollution by its manufacture or consumption.

Figure 16-3 depicts the market for electricity, which we assume is perfectly competitive. The demand curve represents the marginal social benefit of consuming electricity. Since pollution only imposes additional costs on the rest of society, the industry supply curve is the marginal private cost of producing electricity in perfect competition.

The marginal social cost of electricity production is the sum of the marginal private cost of generating the electricity and the marginal external cost of pollution. That means that the gap between MSC and MPC in the diagram must equal MEC. MSC gets its upward slope partly from the upward-sloping supply curve—which is also MPC—and partly from an increasing marginal external cost of electricity production. MEC is increasing because the amount of pollution increases as electricity production increases. As Figure 16-3 shows, increasing the quantity of electricity produced causes the costs related to pollution to rise at a faster rate, even if the amount of pollution created does not rise so quickly. For example, doubling the level of pollution can lead to more than twice as much harm resulting from the pollution.

The perfectly competitive market for electricity is in equilibrium where supply intersects demand at price P_PC and quantity Q_PC. The resulting consumer surplus is the triangle above P_PC (below the demand curve). Likewise, the producer surplus is the triangle below P_PC (above the supply curve). But unlike the used books we examined in Chapter 4, in this market, the sum of the consumer and producer surplus is not the total benefit to society from the production and consumption of the good—as this sum ignores the impact of pollution. The total external cost imposed by electricity production is the area between MSC and the supply curve. The total economic surplus is equal to the sum of consumer surplus and producer surplus minus the total external cost.

The socially optimal level of electricity production occurs where MSB intersects MSC. Figure 16-3 reveals that the perfectly competitive equilibrium is inefficient; perfect competition is not socially optimal when there are externalities—it results in an overproduction of electricity and added pollution. At the socially optimal level of output, the consumer surplus is the triangle above P_OPT, the producer surplus is the area below P_OPT but above the supply curve, and the total external cost is the area between MSC and the supply curve. Thus the total economic surplus at the socially optimal level is greater than that obtained via perfect competition! The table in Figure 16-3 shows this. Perfect competition results in a deadweight loss equal to the shaded area labelled as E in Figure 16-3.

Can the private sector solve the problem of externalities without government intervention? Bear in mind that when an outcome is inefficient, there is potentially a deal that makes people better off. Why don’t individuals find a way to make that deal?

In an influential 1960 article, the economist and Nobel laureate Ronald Coase pointed out that in an ideal world the private sector could indeed deal with all externalities. According to the Coase theorem, even in the presence of externalities an economy can always reach an efficient solution provided that the costs of making a deal are sufficiently low. The costs of making a deal are known as transaction costs.

To get a sense of Coase’s argument, imagine two neighbours, Mick and Christina, who both like to barbecue in their backyards on summer afternoons. Mick likes to play golden oldies on his stereo while barbecuing, but this annoys Christina, who can’t stand that kind of music.

Who prevails? You might think that it depends on the legal rights involved in the case: if the law says that Mick has the right to play whatever music he wants, Christina just has to suffer; if the law says that Mick needs Christina’s consent to play music in his backyard, Mick has to live without his favourite music while barbecuing.

But as Coase pointed out, the outcome need not be determined by legal rights, because Christina and Mick can make a private deal. Even if Mick has the right to play his music, Christina could pay him not to. Even if Mick can’t play the music without an okay from Christina, he can offer to pay her to give that okay. These payments allow them to reach an efficient solution, regardless of who has the legal upper hand. If the benefit of the music to Mick exceeds its cost to Christina, the music will go on; if the benefit to Mick is less than the cost to Christina, there will be silence.

The implication of Coase’s analysis is that externalities need not lead to inefficiency because individuals have an incentive to make mutually beneficial deals—deals that lead them to take externalities into account when making decisions. When individuals do take externalities into account, economists say that they internalize the externality. If externalities are fully internalized, the outcome is efficient even without government intervention.

Why can’t individuals always internalize externalities? Our barbecue example implicitly assumes the transaction costs are low enough for Mick and Christina to be able to make a deal. In many situations involving externalities, however, transaction costs prevent individuals from making efficient deals. Examples of transaction costs include the following:

In some cases, people do find ways to reduce transaction costs, allowing them to internalize externalities. For example, a house with a junk-filled yard and peeling paint imposes a negative externality on the neighbouring houses, diminishing their value in the eyes of potential house buyers. Some people live in private communities that set rules for home maintenance and behaviour, making bargaining between neighbours unnecessary. But in many cases, transaction costs are too high to make it possible to deal with externalities through private action. For example, tens of millions of people are adversely affected by acid rain. It would be prohibitively expensive to try to make a deal among all those people and all those power companies.

When transaction costs prevent the private sector from dealing with externalities, it is time to look for government solutions. We turn to public policy in the next section.