2.10 - Public goods and free riders

Public, private and quasi-public goods

Goods are classified by who can be excluded and whether one person's use reduces what remains for others, rather than by who pays for them.

A private good is excludable and rival. A bakery can withhold a loaf from a non-payer, and a loaf eaten by one customer cannot also be eaten by another. A pure public good is non-excludable and non-rival over the relevant range: excluding non-payers is impractical, and another person's benefit does not reduce anyone else's. National defence for a territory is a usual example. Defence does not become less available to one resident because another resident benefits from the same protection.

A quasi-public good has some, but not all, public-good characteristics, or has them only under some conditions. A quiet untolled road can be hard to exclude users from and non-rival at low traffic. At congestion, an extra vehicle slows others; toll technology can also make exclusion possible. A broadcast can be non-rival yet made excludable by a subscription. Be explicit about the good, technology and congestion range rather than giving one permanent label.

An open fishery is a common-access resource, not a pure public good. State-funded healthcare contains rival appointments and can exclude users, so government funding does not make all healthcare a public good. Conversely a privately funded warning broadcast may have public-good properties.

The public-good characteristics

The public-good model separates the cost of creating a service from the cost of allowing one more person to benefit from a service already supplied.

Non-excludability means non-payers cannot practicably be kept from the benefit. Non-diminishability, or non-rivalry, means one person's enjoyment does not diminish another's. Non-rejectability describes a benefit that people receive without being able to opt out individually, such as territorial protection or cleaner ambient air in the area where they live. It does not mean a person must approve of the service, and it is not a claim that every commonly cited public or quasi-public good is impossible to avoid in every setting.

Zero marginal cost in the pure model means admitting an extra beneficiary costs no additional resources once the given service is provided. An extra resident can benefit from an existing defence system without another system being built for that resident. Providing more defence, maintaining equipment or expanding the protected territory can still be costly. Zero cost of an extra user is not zero total cost or zero marginal cost of expanding the amount or quality supplied.

These ideal characteristics explain why charging each user can be difficult and why charging a positive price may inefficiently exclude users from an already provided non-rival service when exclusion becomes possible. Yet someone must cover the service's creation and maintenance costs. Real services can have capacity limits, different geographical reach or partly rejectable benefits. State the assumptions before applying the pure model.

Why free riders weaken provision

A free rider receives a benefit without contributing to its funding when others pay to provide it.

Suppose an illustrative coastal community needs £1,000 to run a shared storm-warning broadcast. Each household can hear it once it is supplied. If households expect others to fund it, an individual may decline to pay while still enjoying the warning. Because excluding non-payers is impractical, the provider cannot reliably convert all beneficiaries' willingness to pay into revenue. Even when aggregate benefits exceed £1,000, voluntary contributions may be too small and the broadcast may not be supplied, or be supplied at too low a level.

Non-rivalry makes joint benefit possible; non-excludability makes the payment problem acute. The free rider problem can lead to market failure because a profitable sale to each beneficiary is difficult, not because public goods have no value or no demand. An individual small contribution may appear unlikely to determine whether the whole service is supplied, which strengthens the incentive to wait for others.

Voluntary donations, reputation, sponsorship and repeated cooperation can sometimes support provision. These possibilities mean private supply is not universally zero. However, donations may fluctuate, sponsors may choose services attractive to advertisers, and one contributor can benefit many non-payers. A public-good judgement should ask whether voluntary mechanisms actually deliver adequate quantity and quality, rather than assume they always fail or always work.

Funding and providing public goods

Compulsory tax funding can overcome free riding, but funding a service and producing it are separate decisions.

Government can collect contributions through taxation and provide the public good directly, or contract a private organisation to supply it. Taxation reduces reliance on voluntary individual payments and allows wider benefits to be considered. In the warning example, public funding could keep a reliable broadcast operating even when donations are low. The service should be expanded only while the additional collective benefit warrants the extra resource cost; a zero extra-user cost does not imply unlimited production.

Direct provision can support universal coverage and long-term coordination. Contracting may use specialist expertise and competition among suppliers to control costs. Its success depends on specifying quality and monitoring delivery: a cheaper contractor who omits hard-to-reach communities may create poor value. Government can also face weak demand information, administrative costs and a budget opportunity cost. Funding a warning system may mean forgoing a different valuable service; taxes can change incentives elsewhere.

For a quasi-public good, a toll or subscription can fund provision by making exclusion feasible, but may deter users even when spare capacity makes their additional resource cost negligible. Congestion charges are a different case: an extra user imposes a cost on others, so charging can improve allocation. Subsidised access, sponsorship or targeted public funding may combine strengths, depending on the actual characteristics.

An illustrative decision compares a £1,000 reliable warning service with a £600 volunteer service that stops at night. Tax funding has a strong case if dangerous storms are frequent at night and the reliable service's extra benefit exceeds £400 plus collection and incentive costs. If storms are rare and volunteers can cheaply cover nights, the cheaper alternative may yield better net benefits. Judge reliability, reach, collective benefit and opportunity cost; “public good” identifies a payment problem but does not prove every government project is worthwhile.

Quick recap

Public-good characteristics describe use and exclusion, not public ownership. Free riding can weaken voluntary funding. Tax funding can solve that payment problem, while the provider, quality, quantity and opportunity cost still require judgement.