Market failure: public goods

• Define non-rivalrous and non-excludable, and use the two characteristics to classify a good as private, public, common pool or quasi-public.
• Explain the free rider problem as a chain of reasoning that ends in under-provision or a missing market.
• Explain how direct provision and contracting out each correct the failure, and what differs between them.
• Evaluate each response for taxpayers, users, firms and workers, naming the conditions under which it works well.

A market supplies a good when a firm can charge each buyer for the units that buyer uses. Two characteristics make this possible. A good is Rivalrous when one person's consumption of a unit leaves less for everyone else: a sandwich eaten by one person cannot be eaten by another. A good is Excludable when the seller can prevent people who do not pay from consuming it: the shop keeps the sandwich until it is paid for. A good with both characteristics is a Private good, and markets supply private goods.

A Public good has neither characteristic. It is Non-rivalrous: one person's use does not reduce the amount or the quality available to anyone else, so once the good exists the cost of serving one more user is zero. It is also Non-excludable: once the good is provided, nobody can be prevented from benefiting, whether or not they have paid. Both conditions are needed. A good with only one of them is not a public good.

The standard examples are national defence, street lighting, flood defences, lighthouses and public health measures that control infectious disease. Take a flood barrier. It protects every building behind it, including those whose owners paid nothing, so it is non-excludable. Protecting one more household does not weaken the protection of the others, so it is non-rivalrous.

Two contrasts matter most. The first is with a Common pool resource. Both are non-excludable, but a common pool resource is rivalrous: every fish caught by one boat is a fish that no other boat can catch. The second is with a Merit good. Health care and education are often provided by governments, but they are not public goods. A hospital bed used by one patient is not available to another, and a clinic or a school can turn away people who do not pay, so both are rivalrous and excludable. Private firms do supply them. They are merit goods because consumers buy less than the socially optimum quantity (topic 2.8). A public good is defined by its two characteristics, not by who provides it.

A public good is not a free good either. Building and running a flood barrier uses land, labour and capital that could have produced something else, so it has an Opportunity cost. "Free to the user" describes the price, not the cost.

Many real goods have the two characteristics only in part. A Quasi-public good is non-rivalrous or non-excludable to some degree or some of the time. A road is non-rivalrous while traffic is light and becomes rivalrous when it is congested. Technology and law can also change a good's classification: a toll gate makes a road excludable, and encryption makes a broadcast signal excludable. A good is classified by the two characteristics it has as it is actually supplied, not by its name.

A Free rider is a person who benefits from a good without paying towards it. The free rider problem is what follows when a good is non-excludable: people can free ride, so a firm cannot earn revenue from the good, so the market does not supply it. The reasoning has five links.

1. The good is non-excludable, so a person receives the benefit whether or not they pay.
2. One person's payment makes almost no difference to whether the good is provided. The choice that is best for each individual is therefore to pay nothing, to understate how much the good is worth to them, and to rely on others paying.
3. If most people reason in this way, a private firm cannot collect enough revenue to cover its costs. This is true even when the total benefit of the good to society is greater than the cost of producing it.
4. With no prospect of profit, firms do not supply the good. It is not provided at all, or only the small amount that donations or a few high-value users will pay for.
5. Resources are under-allocated to a good that people want and value. The free market has failed to produce the socially optimum quantity; where nothing is produced the case is called a missing market.

The result is a Market failure of under-provision or non-provision. It is not over-consumption. Because the good is non-rivalrous, a free rider takes nothing away from other users and imposes no cost on them. The harm done by free riding is that the good is never built, or is built too small.

Free riding also hides demand. In an ordinary market the price consumers pay shows how much they value the good. For a public good, each person has a reason to conceal that value, so no market price forms and a firm cannot observe the demand it would be serving. This matters again when a government has to decide how much to provide.

Non-rivalry gives a second, separate reason for failure. An extra user costs nothing to serve, so Allocative efficiency requires a price of zero: any positive price would turn away people who value the good at more than the zero cost of supplying them. A private firm cannot cover the cost of creating the good while charging nothing. So even where exclusion could be arranged, a market price would not give the efficient outcome.

Compare this with a common pool resource. Both are non-excludable. A common pool resource is rivalrous, so the failure there is overuse of something that already exists: each user reduces what is left for the others. A public good is non-rivalrous, so the failure is that something valuable is not produced.

The argument assumes that people act in their own narrow interest and that the group is large. In a small group, such as neighbours sharing the cost of a private lane, people know one another and can agree to contribute. In a city or a country that kind of agreement cannot be enforced, and voluntary payment falls far short of what the good is worth.

Direct provision means that the government supplies the good itself, through its own departments, agencies and employees, and pays for it out of tax revenue. Taxes are compulsory, so nobody can free ride on the cost: the payment that a firm could not collect voluntarily is collected by law. The good is then available to everyone without a charge at the point of use.

The difficult decision is how much to provide. In principle the right quantity is the one at which the benefit to society of the last unit equals its cost. In practice no market price reveals what people are willing to pay, so the government has to estimate the benefits, for example by valuing the property and lives a flood barrier would protect, and compare them with the cost. The final choice is made through the political process.

• Strength, the good exists: tax finance removes the free rider problem, so resources are allocated to a good the market would not supply.
• Strength, access: everyone benefits whatever their income, which supports equity.
• Strength, scale: goods with very high fixed costs and long lives, such as national defence or flood protection for a whole city, can be financed because a government can tax and can borrow against future tax revenue.
• Strength, control: the government decides the quality and can change the service without renegotiating a contract, and ministers answer to voters for the result.

• Limit, opportunity cost: tax revenue spent on the public good cannot be spent on health care or education, or left with households and firms.
• Limit, the wrong quantity: benefits are estimated, not revealed by prices, so the government may provide too much, which wastes resources, or too little, which leaves part of the failure uncorrected.
• Limit, political allocation: spending may go where it wins votes, not where the benefit is greatest compared with the cost.
• Limit, weak cost discipline: a government agency has no competitors and cannot go out of business, so it is under less pressure to keep costs down or to adopt better methods.

Time matters as well. The cost of building a public good falls on taxpayers now, while the benefits arrive over decades. In the long run the good also has to be maintained, and a government under pressure to cut spending may postpone maintenance because the effect is invisible until the asset fails.

Contracting out means that the government still pays for the good from tax revenue and still decides how much is provided, but pays a private firm to produce or deliver it. The free rider problem is solved exactly as it is under direct provision, by compulsory tax finance. What changes is who does the work. Typical contracts cover road building and maintenance, refuse collection, the upkeep of street lighting, defence equipment and the running of public facilities.

Contracts are usually awarded by competitive bidding: firms submit prices for doing the specified work for a fixed period, and the government chooses the lowest bid that meets the specification. Firms compete for the market instead of in it. A bidder that asks for more than its rivals loses the contract, so the price is pushed towards the lowest cost of doing the work. Once a firm holds the contract, every saving it makes adds to its profit, so the profit motive supplies the cost discipline that a government agency lacks.

• Strength, lower cost: where several firms bid, the same quantity of the public good is provided for less tax revenue, which lowers its opportunity cost.
• Strength, expertise: a specialist firm that does the same work for many clients has skills and equipment that an agency would have to build up from nothing.
• Strength, flexibility: when the contract ends, a poor contractor can be replaced, and the government does not have to keep a permanent workforce for work that is only needed occasionally.

• Limit, quality: the firm's profit is the contract payment minus its costs, so it gains by cutting costs. Where quality is hard to specify or to observe, cutting costs means cutting quality in ways the government cannot easily detect.
• Limit, contract costs: writing, awarding, monitoring and enforcing contracts uses resources, which reduces the saving.
• Limit, weak competition: if few firms bid, or the firm that holds the contract has an advantage at each renewal because it knows the work, the price stays above the competitive level and a private monopoly replaces a public one.
• Limit, contractor failure: the service cannot simply stop, so if the firm fails the government has to step in at short notice and the risk returns to taxpayers.
• Limit, accountability: contracts may be awarded through favouritism or corruption, and the public can hold a private firm to account only indirectly, through the government that hired it.

Different stakeholders are affected in different ways. Taxpayers gain if the contract price is lower than the cost of direct provision and quality holds. Users gain or lose only through quality, since the good stays free at the point of use in both cases. Firms gain a source of revenue and profit. Workers may lose: part of a contractor's lower cost can come from lower pay or fewer staff, which is a transfer from workers to taxpayers and shareholders and not a gain in efficiency.

• A public good is non-rivalrous (one person's use leaves no less for others) and non-excludable (non-payers cannot be kept out). Both are required.
• Non-excludable, so people free ride; firms earn no revenue; the good is not supplied; resources are under-allocated: a missing market.
• Direct provision: the government pays from taxes and produces the good itself.
• Contracting out: the government pays from taxes and decides the quantity; a private firm produces under contract.
• Both responses end free riding through compulsory taxation. They differ in cost and in control of quality, not in who pays.