Public Goods, Commons, and Institutional Choice

Part A treated externalities as effects that escape an agreed price because the relevant rights are missing, vague, or costly to enforce. Clear rights sometimes permit bargaining; when many parties are involved, taxes or tradable permits can make users face a cost they otherwise escape. Part B turns to two harder cases: benefits that many people can use at once, and resources that anyone can enter but each user can deplete.

The labels matter because the problems point in opposite directions. A public good may be undersupplied because non-payers cannot easily be excluded. A commons may be overused because no owner can protect its future value. Neither label selects a remedy automatically. The institutional question is how to create enough control, payment, and accountability to align choices with the costs they impose.

A Pure Public Good Is Both Nonrival and Nonexcludable

We have been talking about goods that one person consumes: if I eat an apple, you cannot. Those are private goods. A pure public good has two different features. Once it has been produced, one person’s use of it does not reduce the amount or worth available to anyone else. Economists call that nonrivalry. It is also difficult or prohibitively costly to keep non-payers from benefiting, which is nonexcludability. A mathematical theorem and some forms of national defense come close to both conditions.

Keep the two dimensions separate. An uncongested streaming program may be nonrival yet excludable behind a password. Fish in an open ocean can be difficult to exclude people from yet rival because every fish caught leaves fewer for others. A theater performance is shared but congestible because seats are limited. The labels describe resource characteristics, not whether government or a private firm happens to provide the good.

Nonrivalry is also a fact about consumption, not production. Writing the theorem, producing the broadcast, and building a defense all use scarce resources. What may be nearly costless is one more person’s use after the good exists, not its creation.

The public-good provision rule also changes how values are added. For a private good, market demand adds the quantities different buyers want at a common price. For a public good, everyone uses the same quantity, so the analyst adds each user’s marginal value at that quantity. If three viewers value one more program at $5, $3, and $2, the combined marginal value is $10. Producing it would create gains if its marginal cost were below $10, assuming those values and costs were known and all affected parties were represented. That last condition is difficult: a person who expects to receive the program anyway has reason to understate what it is worth.

The Free-Rider Problem Has Several Imperfect Institutional Responses

If a good is nonrival and hard to exclude people from, why pay when you can benefit whether or not you chip in? This is the free-rider problem. A supplier who cannot collect from beneficiaries may see only part of the benefit and provide less than users would collectively be willing to finance. Three remedies recur, all variations on one theme, finding a way to make the beneficiaries pay.

One response is tax finance. Compulsory collection can overcome voluntary free riding, but it does not reveal how much each person values the good or guarantee that the provider chooses the right quantity or production method. Elections, budgets, expert analysis, and administrative rules become substitute information and control systems. Mosquito control and flood protection have public-good features, but streets and police services are mixed cases that can become congested or locally excludable.

A second response is an excludable club. Members finance a shared facility and restrict access to contributors. Golf courses, neighborhood pools, and subscription databases can work this way. Clubs solve collection by sacrificing open access, and they still must manage congestion, membership rules, maintenance, and conflicts among members. Transferable memberships can give owners a capital stake in future quality, but boards, voting rules, reputation, and repeated interaction can also discipline stewardship.

A third response is a tie-in. Broadcast radio can be offered without a listener fee because advertisers pay for expected access to the audience. The listener pays indirectly through attention and the opportunity cost of interruptions, though no one can force a particular listener to watch or hear an ad. Newspapers, online services, sponsorships, donations, and bundled products use related arrangements. They are not literally free; they connect provision to some marketable complement, audience, reputation, or mission.

Open Access Creates a Commons Problem; Common Property Can Be Governed

A common-pool resource is rival in extraction but costly to exclude people from using: a fishery, pasture, forest, or irrigation system. If entry is open and no effective rules limit withdrawal, each user captures the current gain while sharing the future depletion with everyone else. Picture the park tree. Why do the apples get picked before they ripen? Not because anyone prefers sour fruit, but because a person who waits may find that somebody else picked them first. The incentive is to harvest too early and devote too many resources to the race.

This is an open-access problem, not a definition of all common property. A private owner can internalize future stock value, but private ownership is only one possible institution. A regulator might set seasons, gear rules, or catch limits. Fishers might receive transferable harvest shares or territorial rights. A stable community might create and enforce its own boundaries, monitoring, graduated penalties, and conflict-resolution rules. Elinor Ostrom’s field research showed that user associations sometimes manage forests, fisheries, and grazing lands successfully rather than choosing between privatization and central command. The result depends on whether users can define the group and resource, observe withdrawals, adapt rules to local conditions, and sanction violations. Her Nobel lecture therefore framed the issue as going beyond a simple choice between markets and states.

Stock an open-access lake with fish and the fish may be caught younger and with more boats, gear, and hours than the additional catch warrants. An owner who captures the asset’s future value has reason to wait. A fishing association can create a similar incentive if its members expect the rules to persist and can exclude outsiders and police one another. A quota can preserve the stock only if officials possess enough biological information and can monitor the catch. Each institution changes who holds the effective right, how restraint is rewarded, and what enforcement costs must be paid.

This is also why a good’s classification can change with congestion. An empty park, road, or beach may be effectively nonrival; once crowded, one more user imposes delay or reduces everyone else’s space. Admission prices, reservations, time limits, access rules, or expanded capacity may respond to that congestion. The relevant comparison is among feasible arrangements, not between a perfect market and a costless government.

Who Bears the Cost Is a Political Question, Not Just an Economic One

I want to close by widening the lens from allocation to politics, since every remedy we have discussed lands its costs on some identifiable group, and the one a society actually adopts depends on who can capture the gains and who can be made to bear the costs. A ban, a tax, a permit market, public provision: none is free, and choosing among them is as much a political contest as an economic calculation.

Decision-makers respond to different forms of discipline. A private owner who is the residual claimant gains from reducing cost or moving an asset to a more valuable use, but profit does not automatically capture pollution, monopoly effects, distributional goals, or benefits that cannot be sold. A government agency or nonprofit cannot distribute a residual to an owner and may pursue a broad mission, but it is not uncontrolled: statutes, budgets, boards, donors, voters, audits, professional norms, and managerial reputation can reward or punish performance. These controls are imperfect, as are prices and ownership. The comparative question is which errors each arrangement is likely to reveal, correct, or conceal.

The Postal Service illustrates the danger of forcing a one-cause story onto an institution. It is expected to finance most operations from postal revenue while also meeting legally specified service obligations. The GAO’s current assessment says USPS has lost money in almost every fiscal year since 2007 and calls its business model unsustainable, but attributes the problem to a combination of declining mail volume, rising costs, statutory requirements, and tension between service expectations and available revenue. That record can motivate questions about governance and cost control; it cannot, by itself, prove that the absence of a private residual claimant caused the losses or that privatization would preserve the same service obligations at lower social cost.

Political organization is another mechanism to examine. When a policy gives a large per-person benefit to a small group and spreads a small per-person cost across millions, beneficiaries may have a stronger incentive to acquire information and organize. That pattern does not prove that every concentrated group wins or that its policy lacks a public purpose. Opposition groups, institutions, ideas, media attention, and electoral rules also matter. It does explain why the distribution of stakes belongs in the analysis rather than being inferred from the policy’s stated goal.

There is a rhetorical hazard here too. Because so many goods carry some public-good character, “this is a public good” can support a genuine shared-benefit argument or a bid for subsidy. The label alone cannot reveal the motive or select the remedy. Nor does private provision disprove the public-good feature: advertising, subscriptions, patents, donations, and clubs may create enough exclusion or complementary revenue to finance some shared benefits.

The same caution applies to redistribution. Below-cost state tuition benefits enrolled students, but whether it transfers resources from richer to poorer households depends on who attends, who pays taxes, and how education changes later earnings. A tariff may protect workers and owners in one industry while raising prices for consumers, including lower-income households. A state monopoly may pursue control, revenue, or both. The relevant question is not the program’s label but who ultimately receives the benefit and who bears the full cost.

Consider a hypothetical town-owned airport beside developable land. Housing could expose residents to noise and create pressure to restrict flights. Zoning could prohibit homes, disclosure rules could require buyers to acknowledge the noise, the airport could purchase an easement, or law could specify which party bears the nuisance. Each arrangement defines rights and shifts costs. Disclosure helps only if buyers understand the information and later restrictions remain credible; zoning requires officials to judge incompatible uses; purchase requires financing and a bargaining process. Economics can compare these consequences without pretending that efficiency alone settles the distribution of rights.

The durable method is to ask the same questions of every proposal. Is the good rival or nonrival, and can users be excluded? Who has relevant information? Who captures the future gain from restraint? What must be monitored and enforced? How can errors be corrected? Then trace the political incidence: who has the incentive and ability to choose the rule, and who ultimately pays? That is comparative institutional analysis rather than a presumption for either markets or government.

T9 will take the residual-claimant mechanism used here and develop it into a full theory of firms, monitoring, and organizational governance.

For Further Reading

Want to explore the source material? This lecture draws on the following chapters from two books by Armen A. Alchian and William R. Allen:

• Universal Economics (Liberty Fund, 2018): Ch. 13, “Markets and Property Rights”.
• Exchange and Production, 3rd ed. (Wadsworth, 1983): Ch. 5, “Information Costs and Achievement of Exchanges”; Ch. 18, “The Domestic and Political Economies”.

• Full UE 13 and pinned E&P 5/18 sections read; Ostrom Nobel materials and current GAO USPS assessment checked externally.
• Pure public good now requires nonrivalry and nonexcludability; mixed goods keep the dimensions separate.
• Open access is distinguished from common property. Private rights, regulation, transferable rights, and user governance are compared conditionally.
• Government/nonprofit and private controls are stated symmetrically; categorical ownerlessness, concentrated-interest, and USPS causal claims were qualified or retired. The socialism paragraph was retired and the airport case made explicitly hypothetical.
• Three objectives/takeaways and all four registered Core/coverage anchors preserved.
• Canonical words 3,121 -> 2,218; budget 3,210 -> 2,307, retaining 89 headroom.
• Local closeout: default/Core/coverage lint PASS; extensions 19/19 PASS; migration 320/0; fixtures 17/17; policy-claim YAML PASS. —>