National defense protects every resident of the United States simultaneously. Adding one more person to the protected population costs essentially nothing (non-rival). No resident can be excluded from protection — the military cannot selectively defend some zip codes and leave others undefended (non-excludable). These two properties together make national defense a public good: markets cannot provide it efficiently because no private firm could charge for it without being undercut by free-riding. A private defense company that charged subscribers would be protecting non-payers just as much as payers — making the subscription model unworkable. The government must fund national defense through taxation rather than market pricing.
In plain terms
Goods are classified by two properties:
Rival: does one person's consumption reduce the amount available for others? A sandwich is rival; a radio broadcast is not. Excludable: can non-payers be prevented from using it? A concert inside a ticketed venue is excludable; a streetlight is not.
This creates the 2×2 typology:
| Rival | Non-rival | |
|---|---|---|
| Excludable | Private goods (food, clothing, cars) | Club goods (streaming services, toll roads, national parks with gates) |
| Non-excludable | Common resources (fisheries, congested roads, groundwater) | Pure public goods (national defense, basic research, broadcasting) |
Pure public goods are both non-excludable and non-rival. Markets fail to provide them because:
- Non-excludability makes free-riding rational — individuals can use the good without paying if they can't be excluded.
- Non-rivalry means additional users impose no cost — so setting a price above zero produces underutilization.
The Congressional Budget Office's analysis of public goods spending covers the major categories: defense, basic research, and public health infrastructure — all goods that markets systematically underprovide due to their non-excludable, non-rival character.
Why it works this way
The free-rider problem is decisive for pure public goods. When a good is non-excludable, every rational individual waits for others to fund it — capturing the benefit for free once it's provided. If everyone free-rides, the good is never provided, even though everyone would benefit from it and the total benefit would far exceed the cost. This is a coordination failure that the price mechanism cannot resolve without excludability.
Club goods (excludable, non-rival) can be provided by markets despite non-rivalry — subscription models work because exclusion is feasible. Streaming video, software, and club memberships are all club goods sold by private firms. Non-rivalry means the marginal cost of an additional user is near-zero, making pricing above marginal cost (monopoly pricing) the typical market outcome — which creates a different, smaller efficiency problem.
A real example
Epidemiological surveillance — monitoring disease spread across populations — is a near-pure public good. The CDC's National Notifiable Diseases Surveillance System provides information about disease incidence that benefits everyone who avoids infection based on it, without depleting the information for others (non-rival) and without being able to charge individual users for the knowledge (non-excludable). Private firms would massively underprovide this service; government provision is the standard solution.
Why it matters
The public-private goods typology is the organizing framework for identifying which activities markets can and cannot provide efficiently. Public goods provision is one of the core legitimate functions of government in economic analysis — not because governments are efficient providers, but because markets are particularly ill-suited to provide non-excludable goods at efficient levels. Understanding this distinction prevents two errors: expecting markets to provide what they structurally cannot, and expanding government provision beyond what market failure actually justifies.
