We Need a Theory of Publicly Provided Goods

Let's start with property rights...

In one of my previous posts, I explained what I believe to be the problems with Econ 101. Among those problems is the way that we teach things like public goods and common resources. The typical Econ 101 course discusses these concepts using a classification system. While this makes the presentation easy to understand with respect to some of the issues, it often creates confusion among the students.

It is not hard to understand why students get confused. For example, public goods are defined as goods that are non-excludable (we cannot stop people from benefiting from the provision of the good) and non-rivalrous (one person’s benefit from the provision of the good does not affect anyone else’s benefit). Textbooks then give examples like national defense. The entity providing national defense provides a benefit to all citizens and no one citizen’s benefit precludes others from benefiting. That seems straightforward enough.

The exact opposite of this is a private good, which is both excludable and rivalrous. Herein lies the source of confusion for students.

When students look around at the real world, what they see are things like municipal fire departments and golf courses. These are publicly provided goods that clearly belong within the classification of private goods.

This doesn’t just reveal a problem with Econ 101, it reveals a broader problem. We need a theory of publicly provided goods. So, this week I am going to be more speculative. I think that the consistent public provision of certain types of goods (e.g., fire departments) calls out for explanation. In this post I am going to explore a couple possible explanations: a more effective provision by the government and incomplete private contracts. Hopefully the post will inspire some conversation about the public provision of certain goods.


To examine the argument about the effectiveness of public provision, I think we need to think about this is terms of property rights.

Let’s start with something that isn’t likely to generate an emotional response: golf courses. Most golf courses are privately owned and operated. However, some golf courses are owned and operated by the city. What is the difference between the public and private golf courses?

Think about this in terms of property rights, or the right to the exclusive use of a resource. The private course is owned by an individual or a group. The owner(s) can exclude people from using the golf course. However, they can grant a temporary use of the golf course by collecting greens fees from those who want to play golf. Along with the ability to exclude others from using the golf course, the owner of the course has the ability to exchange his or her ownership stake in the course to someone else. The property right is transferable.

The public course is owned and operated by the city. The city has the right to exclude people from using the course and has the ability to transfer that right to someone else by selling the course.

Why should we think that these golf courses operate any differently? No doubt some would argue that these courses would have different objectives. The private course operates to earn a profit. The municipal course perhaps operates to provide greater access (this could be through lower prices or something like limiting discrimination). However, there is no a priori reason to believe that these courses have different motives. The city might see the profit of a public course as a way for the city to raise revenue. Private courses that offer lower prices might emerge through competition.

One might also be tempted to suggest that the difference between the two courses relates to the separation of ownership and control. However, this argument might be more nuanced than one imagines.

For example, it is tempting to say that the city course will be mismanaged. The people who operate the course on a day-to-day basis are responsible for financial performance. The financial performance ultimately affects the city’s taxpayers. The person in charge of the daily operations of the course might or might not be a taxpayer. Even if he or she is a taxpayer, the share of any losses is quite small in terms of tax liability. The share of any profits is paid indirectly through other city services that are provided by the revenue the course generates.

Nonetheless, this is not that much different from a private course owned by a group of shareholders and operated by a manager. The person in charge of daily operations might or might not be a shareholder, yet is responsible for the financial performance of the course. Even if the manager is a shareholder, he or she will earn only a fraction of the profits through ownership and only bear a fraction of the losses.

In each case, there is a principal-agent problem. In each case, a bad manager can be voted out. Shareholders can vote to remove the manager. Taxpayers can vote for representatives who promise to provide new management.

The real difference between the golf courses is in the ability to transfer shares. A shareholder who is unhappy with the private course can sell his or her shares to a new shareholder or be bought out by existing shareholders. A taxpayer does not have this ability. The taxpayer cannot sell his or her share of the public golf course to anyone. The only way to opt out would be to move out of the city or pressure the local government to sell the golf course to a private firm (there might be no direct financial reward to the taxpayers).

The reason that this difference matters is that it shapes the incentives of decision-makers. When it comes to private firms, people have the ability to choose to be a shareholder and to choose the number of firms for which they want to own shares. Related to this point, shareholders also have a choice about whether to be active or passive shareholders. In addition, the ability to transfer shares has important implications for the firm. Share prices reflect expectations about the future performance of firm. A falling price makes it harder to raise funding. A rising price makes it easier for a firm to raise funds. Similarly, the exchangeability of shares provides a means by which people can transfer risk. Activist shareholders can accumulate larger stakes in the firm in order to acquire a greater say in the conduct of the firm.

Taxpayers have something similar to the voting rights of a shareholder. However, an individual taxpayer does not have the same ability to transfer shares or even to determine how many projects for which they are a stakeholder. The larger the number of projects undertaken by the city, the weaker the incentive is for the taxpayer to pay attention to a particular project. Activists who care about publicly provided goods face the problem of one-person, one-vote. Nonetheless, these activists can form interest groups to give greater voice to their perspective. Finally, if the private value of something like a municipal golf course is greater than its current value, the decision-makers in city government could always sell the course for its capitalized private value. Thus, there is a financial incentive to sell an underperforming asset.

What this distinction makes clear is that the incentive structure differs for privately provided goods and publicly provided goods. Nonetheless, some of the same feedback mechanisms are operable in either case. The main difference seems to be that the feedback mechanisms are stronger for privately provided goods than for publicly provided goods since there is an extra layer of decision-makers. In addition, the selection mechanism of profit and loss tends to push unprofitable private firms out of the market. The private golf course that is subject to losses over time will be forced to shut down, or sold to new owners, possibly to reallocate the land for a more profitable purpose. The public course can continue to operate at a loss with the cost borne by the taxpayers. This can continue until public pressure leads to a change. However, this requires political organization and the salience of the issue among the population.

This discussion of property rights emphasizes the differences in the effectiveness of private and public provision. While it is important to understand and compare the effectiveness, this does not seem like it can explain why certain types of goods or services are publicly provided. It certainly does not seem to explain why a city might own and operate a golf course. But perhaps this is simply government folly and I’ve picked a bad example.

Still, other examples of publicly provided goods similarly cast doubt on the effectiveness argument. Cities often own and operate fire departments. It is not obvious, at least to me, that the city would be more effective at providing these services than a private fire department. Nevertheless, if we shift our focus to the fire department, I think that we can make some progress.


One characteristic of a golf course is that greens fees are sold on a spot market. A potential player shows up at the course and pays the price for a round a golf immediately before playing his or her round.

Now, imagine that fire services are privately provided and sold in a spot market. In this case, there is a difficult problem because you would need to purchase fire services when your house is on fire. In this circumstance, you do not have the ability to shop around for the best price or the best or the fastest fire department. In addition, when the fire department shows up, they have all of the negotiating power. As a result, a hold-up problem might occur in which the fire department gets you to agree to pay your maximum willingness to pay in exchange for putting out the fire.

One possible way to solve this problem is through contracting. In theory, it is possible that property owners could sign a contract with a fire department of their choosing. However, there are two problems here. First, the contract will likely be incomplete. The cost of putting out a fire is not uniform across all possible property. The price of providing fire services might depend on the type of the fire, the extent of the fire, how much of the property is damaged, and the type of property. It would be difficult, if not impossible, to price every possible contingency in advance. In addition, changes in property values and/or characteristics over time might lead to disputes over pricing.

The second problem is that the contract does not necessarily solve the hold-up problem. Even if all prices were agreed to, there is always a chance that the fire department chooses not to honor the contract in the moment that the house is burning or that there could be a dispute in real-time about the appropriate price for putting out the fire. Even if this was a clear violation of the contract, the property owner would have to pay the maximum willingness to pay to extinguish the fire and then pursue a legal battle to recover some of that payment as a violation of the contract. What’s more, the fire department would not even have to subject the property owner to a hold-up problem. It could simply bill the property owner for the maximum willingness to pay with a demonstrated commitment to collect that amount through the legal system.

From society’s perspective this prospect is wasteful since the total cost to society (including legal costs) would be greater than the value of the services rendered. Backward induction from this outcome would likely result in underinvestment in property.

Of course, one might argue that private fire departments care about their reputation. This does complicate matters, but it fails to solve the last period problem. If a fire department is planning on shutting down or there is a high enough payout to induce a hold-up followed by a shut down, reputation will not be a constraint.

The fire department could also operate on a different business model. It could conceivably collect monthly payments from customers in exchange for a promise of providing fire extinguishing services at some future date. However, in this case, one might have people who attempt to free ride by not paying monthly payments, but still attempting to acquire services in the event of an emergency. The government could make payments compulsory. However, by the same logic the city could create its own fire department and make payments compulsory through taxation. Furthermore, it is unclear whether this arrangement would solve the hold-up problem. The monthly payment might entail a certain level of services (or be interpreted as such by the fire department). A hold-up problem could occur if there is a particularly risky fire above and beyond the standard level of fire services offered and implied by the contract.

This idea for why certain goods are publicly provided is due to Earl Thompson. He argued that a moderate level of transaction costs would lead to the sort of outcome that I just described. If the transaction costs of initial negotiations are sufficiently high, this makes contracting difficult and imperfect. Yet, if the transactions costs associated with conflict ex post are sufficiently low, then conflict will occur. Thus, when these costs are moderate, something resembling a hold-up problem will occur. In these scenarios, the public provision of the good or service could prevent underinvestment through a public commitment to avoid such problems.

Thompson’s paper is provocative and difficult. Yet it attempts to rationalize why certain goods are consistently publicly provided. He argues that this theory can not only explain the public provision of police and fire services, but also zoning, workman’s compensation laws, and elements of early antitrust law.

Of course, no matter what one thinks of Thompson’s theory, it does not present an explanation for public golf courses. Perhaps public golf courses require a public choice argument. Nonetheless, I think that the consistent public provision of particular goods and services calls out for a price theoretic explanation.

While this post has been quite long and speculative, I think that this topic deserves greater attention. In short, we need a theory of publicly provided goods.