So far in this chapter, we have looked at market failures in which markets end up with an inefficient level of output as a result of externalities or unclear property rights. There is another type of good for which markets can fail to deliver the socially optimal level of output. Public goods are goods (such as national defense, a fireworks display, or clean air) that are accessible to anyone who wants to consume them and that remain just as valuable to the consumer even as other people consume them. For example, when you watch a fireworks display from your backyard, it makes no difference if you are the only person watching the display or if all of your neighbors are watching the fireworks, too. And, the individuals putting on the fireworks display can’t stop people from watching it, even if they wanted to. As a result, public goods have distinctive properties that make it difficult for markets to deliver them at efficient levels. Note that a public good is not necessarily provided by a government, although governments often provide public goods because private markets may provide too few of them.

Public goods are similar in some ways to positive externalities: They can provide external benefits to individuals other than those who purchase them. For example, if you host a fireworks display, you may not consider the benefits received by your neighbors when making your purchases. This leads to an output that is smaller than is socially optimal because the full benefits of the fireworks display are not considered. Put simply, if the true social benefits were taken into account, there would be more (and perhaps more elaborate) fireworks displays.

Public goods have two important properties. One is that they are nonexcludable, which means all people who want to use the good have access to it. They cannot be prevented—excluded—from using the good. Nonexcludability happens when it is impossible or impractical for one party to monitor and limit the use of a good by any particular user. For the fireworks display in the example above, by nature it would be difficult for those putting on the display to limit its visibility only to a particular set of users. Therefore, it is nonexcludable.

The second property that public goods have is that they are nonrival, meaning that one person’s consumption of the good does not diminish another consumer’s enjoyment of the same good. (A different way to think of a good being nonrival is that the marginal cost of providing the good to yet another consumer is zero.) For example, a TV weather forecast is a nonrival good because the value of the forecast to a consumer is unaffected by the number of individuals also watching the same forecast. The fact that your neighbor watches the forecast (or hears it on the radio, or reads it online or in a newspaper) does not in any way diminish or eliminate the utility you derive from receiving the information. Any good that many people can consume independently without using it up or degrading it is a nonrival good. In contrast, if a consumer buys a rival (regular) good, then someone else cannot buy that exact good. For example, if you are eating a taco, another person cannot eat that same taco. Because the taco is also excludable (a person can be kept from consuming it), it is a private good. Table 17.1 shows the different types of goods that exist categorized by rivalry and excludability.

677

678

Before we can see why markets provide inefficient levels of public goods, we have to define what the efficient level of output is for these types of goods. It’s a little different from the efficiency conditions we’ve discussed before.

Efficiency generally occurs when a market produces the quantity for which the marginal cost of producing the good just equals the marginal benefit society receives from the good. For a competitive market without externalities, that quantity is also where quantity supplied equals quantity demanded. As we saw earlier in this chapter, in markets with externalities present, we need to be sure that all external costs and benefits are considered when determining the efficient output. Efficiency in these markets thus occurs at the output for which social demand (which measures social marginal benefit) is equal to social marginal cost.

With public goods, the marginal benefit from the good is not the benefit for just one consumer because more than one person can consume the good simultaneously. A public good is nonrival, so to value the good, we have to add the marginal benefits of everyone who consumes it. Put differently, a public good’s total marginal benefit curve is the vertical sum of the marginal benefit curves of all its many consumers. This total marginal benefit (MB_T) is what equals marginal cost when a public good is being provided at its optimal level. In equation form, the total marginal benefit is written

MB_T = ΣMB_i

The summation symbol Σ denotes that the good’s marginal benefits are added over all consumers (whom we’ve indexed with i) of that good.

There is no “sum of marginal costs” condition because the cost of producing a public good is just like the cost of producing any other good. While a public good is simultaneously consumed, there is nothing in its definition that implies it is simultaneously produced. Whether made by an individual, a firm, or a government, the marginal cost of a good is the same no matter if that good is a public or standard private good.

Figure 17.9 shows the public goods efficiency condition for a simple example in which the public good is consumed by two people. The marginal benefit curves of each person are illustrated in the figure as MB₁ and MB₂. The total marginal benefit of the public good, shown as curve MB_T, is the vertical sum of the individual marginal benefit curves. MC is the marginal cost of producing the good.

Efficiency requires that the public good be provided at point A, where MB_T = MC, with a quantity of Q*_Pub. It’s important to realize that this is the quantity of the public good that both individuals consume. Because it is nonrival, they don’t have to split Q*_Pub between them. They consume Q*_Pub units of the good simultaneously.

Why isn’t this efficiency condition achieved in the private market? For the same reasons we saw in markets with externalities, the free market is going to encounter a problem providing the right level of public goods. The market facilitates private exchanges until a good’s private marginal cost equals individuals’ private marginal benefits. However, that is not where the combined marginal benefits of multiple consumers equal the good’s marginal cost. If individuals could buy the public good themselves at marginal cost, each would buy the quantity at which her own marginal benefit equals its marginal cost. Looking at Figure 17.9 again, Person 1 would buy Q₁, where MB₁ = MC. Similarly, Person 2 would buy Q₂.

679

Therefore, although both individuals would be willing to pay for some quantity of the public good, neither would by herself pay for the efficient quantity because her individual marginal benefit is less than the joint marginal benefit. The inefficiency arises in a way that is similar to how inefficiency arises with a positive externality: Individuals paying privately for a public good don’t consider everyone else’s benefit from the good. This is one reason why private markets fail to produce the efficient quantity of public goods. Left to their own devices, private markets don’t supply enough national defense, mosquito abatement, clean air, fireworks displays, and the like.

In addition to not taking into account the total marginal benefit of the good and thus having less of it available than is socially optimal, the free-rider problem is another source of inefficiency in the markets for public goods. A free-rider is someone who enjoys a good or service without paying for it. For example, electronics manufacturers, shoe manufacturers, and car manufacturers want people to be able to try out their products before purchase, and there are retailers that provide those services. In many cases, however, online retailers, who don’t have to provide (or pay for) free trials and other services, can offer lower prices than “brick-and-mortar” stores. The online retailers are free-riding on the services provided by the brick-and-mortar stores. The problem with this is that if enough people try out a product in a retail store and then go online to buy it at a lower price (a practice sometimes called “showrooming”), offline retailers will stop providing try-out services. The free market ends up generating too little of the try-out information (which benefits consumers and producers) because it has become a public good.

680

A common solution to the free-rider problem in the fireworks case, the online retailer case, and in paying for public goods more generally is for an entity like the government or the manufacturer to compel people to pay their share through a tax or a retailer fee. This is why the most expensive and broad-scoped public goods such as national defense, “rule-of-law” goods (e.g., court systems, police services), and national transportation infrastructure—items for which free-riding might be the most severe—are paid for through taxation.16

681

Other solutions are also possible. Beneficiaries of a public good can form an organization that compels members to pay their share of the public good’s costs. The tricky issue in such situations is convincing potential members to voluntarily form a group that will compel them to pay for something they’d rather not pay for at all. Condo owners, for example, often form a condo association to pay for maintaining a pool and other common areas because free-rider problems will develop among the owners if they don’t.

As economist Mancur Olson pointed out in 1965, the likelihood of solving the free-rider problem depends on how large the public good’s benefits are per potential member of the group that would enjoy the good.17 If each potential member’s gain from cooperative action—her benefit from having the efficient level of the public good available—would be large, it is more likely that a cooperative group will form. When the benefit-per-potential-member ratio is low, group-forming success is less likely. Because each member of the group only benefits a little from the public good, less is lost trying to free-ride. The incentive to free-ride makes it difficult to hold any possible group together. (This is a lot like the logic we discussed in Chapter 11 about why it’s easier for a small number of firms to form a stable cartel than it is for a large number of firms to do the same thing.)

It’s important to realize that the benefit per member of a public good might not be closely correlated with the total benefit of the good. It’s quite possible that a public good that would offer an enormous total benefit may not end up being provided because the size of its potential group is also really large, making the average benefit per person too small to overcome free-riding. At the same time, a public good with a much smaller total benefit might be provided if it benefits only a select group who enjoy a large benefit per member. This effect can explain why special interest groups successfully lobby for programs that jointly benefit their members despite imposing a much larger collective cost on everyone else. Each member of the special interest group enjoys large gains by paying lobbyists to put forward the group’s agenda and encourage legislators to vote favorably for its programs. The lobbying can be thought of as a public good for the members. Even though the program imposes a very large collective cost on everyone outside the special interest group, the cost borne by any one person is too small to overcome free-riding, and no one mounts a concerted effort to stop the legislation.

A common resource (a type of public good that is rival but nonexcludable) isn’t a pure public good, but it presents a similar kind of inefficiency called the tragedy of the commons. The concept of the “tragedy” comes from the dilemma that common resources create: Because everyone has free access, the resource is used more intensively than it would if it were privately owned. This leads to a decline in the value of the resource for everyone.

A reservoir of water, public forests, public airwaves, and even public restrooms are all examples of common resources. The key element is that the common resource is nonexcludable: Anyone can use or take from it, and those who use it cannot be monitored easily. Because access to the resource is shared by all comers, common resources are also called common-pool or common-property resources.

We can think of the tragedy of the commons as a problem caused by a negative externality. No single user takes into account the negative externality she imposes on others by using the common resource, leading everyone to consume too much of it. The externality involved with a common resource arises because of the combination of open access and depletion through use. When deciding how much of the common resource to consume, everyone considers only her own cost of use. But, this use depletes the resource for all other users. Because individuals don’t consider the cost they impose on others when making their decisions about how much of the common resource to consume, they end up using too much of the resource. And, because everyone accessing the resource creates this same externality with her own use, the total utilization of the resource is above the socially optimal amount. This outcome is analogous to the market quantity being higher than the efficient level in our earlier negative externality examples. Without controls, too much water is extracted from aquifers, fishing grounds are overfished, too many trees are cut on public lands, too many communications devices jam the airways, and so on.

682

683

The Petrified Forest National Park in Arizona, for example, is a surreal landscape full of fallen ancient trees that have been petrified into rock. It is a common resource for everyone to enjoy, yet many visitors steal small pieces of petrified wood to keep as a souvenir. The parks service estimates that it loses about 12 tons per year in theft. Each thief thinks, “I can take just a small piece . . . it won’t make any difference,” and doesn’t take into account that removal of that one piece reduces the value of the park for others. Without controls—park ranger patrols, heavy fines and punishments, and so on—this attitude would easily lead to a dwindling of the resource. This happened to the original Plymouth Rock, in Massachusetts, the supposed landing spot of the Mayflower immigrants in 1620. So many people chipped off pieces of the rock that by 1835 it had dwindled to less than half its original size and had to be moved behind a fence to prevent it from disappearing completely.

In some cases, the consequences of this overuse can be severe: Instead of preserving a sustainable fishing stock, for example, overfishing can easily drive an entire population of fish to extinction. It’s not that fishers prefer to drive their prey to extinction; in fact, they prefer to have a steady supply available. But, the overfishing happens because of the externality and the open access to the fishing stock.

Remedies for the Tragedy of the Commons Because the tragedy of the commons is just a special form of negative externality, any of the solutions we described in the previous section can be used to try to fix it, and they are, in fact, employed in many actual situations. For example, governments use price-based mechanisms (Pigouvian taxes) to charge individuals for the external damage they do to the common resource, or they charge a fee for each unit of the common resource an individual consumes (such as an entrance fee to a national park). Governments also set quantity restrictions (quotas) on the rate at which common resources can be extracted, such as limits on the amount of fishing, hunting, or logging, for example. These interventions can be used by governments or other organized public or private regulatory bodies. In fact, as with other types of public goods, users of common resources have incentives to join together to self-regulate use of the resource.

Another way to fix the tragedy of the commons externality involves defining property rights and facilitating negotiation among those who share the common resource. For the airwaves, for example, the government auctions off spectrum and says that whoever buys the spectrum is the only one allowed to broadcast at that frequency. Granting sole control of one part of the resource to one party eliminates the negative externality. The person in control gets all the benefit and pays all the cost of using the resource, so she doesn’t have an incentive to overuse it.