IESET.
Conditions › Conditions favoring intervention

Public goods non excludable

Pure public goods are non-rival (my consumption does not reduce yours) and non-excludable (cannot prevent free-riding). The canonical Samuelson (1954) analysis shows that private voluntary provision underprovides these goods because rational agents free-ride on others' contributions. Examples include national defence, basic legal and property-rights infrastructure, most scientific knowledge, clean air, flood defence, public health surveillance, and open-access mapping and statistics. Compulsory tax-finance with centralised provision is the near-universal institutional solution, and works across regime types because the underlying market failure is so fundamental.

confidence: highConditions favoring interventionentry added 2026-04-29public_goods_non_excludable

Institutional features that make the model work

›Tax finance with compulsion
Mandatory taxation overcomes the free-rider problem by forcing contributions. The political legitimacy of tax collection is the precondition for public-good provision.
›Centralised provision or procurement
National defence, basic legal infrastructure, and flood defence are typically provided directly by the state. Scientific research and statistics can be provided through autonomous public agencies (ONS, BLS, Eurostat) or publicly-funded independent institutions.
›Constitutional or legislative scope limits
Clear constitutional allocation of which goods qualify as public (defence, courts, police, roads) vs merely state- provided prevents open-ended expansion of the public-good rationale to cover rent-seeking.
›Open access distribution
Outputs — legal judgments, public-health advisories, weather forecasts, government statistics, much publicly-funded research — are provided free at the point of use, consistent with their non-rival nature.

Supporting cases

national_defence_universal_state_provision

Every functioning state provides national defence through compulsory tax finance. The stylised absence of any counter-example across regime types is itself the strongest evidence of the free-rider mechanism's empirical force.

public_statistics_systems

National statistical offices (ONS, BLS, INSEE, Statistics Canada, Eurostat) provide open-access economic and demographic data that private markets have never replicated at comparable coverage. Private data firms (Bloomberg, Refinitiv) provide complementary commercial layers but depend on public statistical foundations.

weather_forecasting_public_infrastructure

NOAA, Met Office, ECMWF operate publicly-funded weather observation and forecasting networks that underpin private weather services. Data is largely open; commercial layering is downstream of public infrastructure.

basic_legal_and_police_infrastructure

Property rights enforcement, contract law, and basic policing are near-universally state-provided. The few historical experiments with purely private provision (feudal law, private policing in weak-state settings) produce substantially worse outcomes.

Disconfirming cases

lighthouse_coase_critique

Coase (1974) documented that some British lighthouses were historically privately funded through port dues — showing that ingenious exclusion mechanisms can sometimes be built. A real partial counter-example that limits claims of absolute non-excludability, though the scale of modern navigation infrastructure remains publicly provided.

open_source_software_as_private_club_good

Open-source software production through voluntary contribution and company-sponsored developers has produced a large corpus of non-rival, near-non-excludable code without direct state finance. Partial counter-example: some information public goods can be produced via reputational and reciprocal incentives.

club_goods_and_tolled_roads

Toll roads, gated communities, and cable TV illustrate the club-good case where exclusion is feasible but non-rivalry still holds. These do not count as pure public goods and can be privately provided under the right institutional conditions.

What this condition is NOT

  • An argument that everything the state currently provides is a public good
  • A claim that healthcare, education, or housing are pure public goods in the Samuelson sense — they are rival and excludable
  • A refutation of Coase's observation that some apparent public goods admit private provision
  • A general argument for state ownership of information or media
  • A claim that the boundary between public and club goods is always clear — many cases are genuinely mixed

Policy implications

Core public goods — defence, basic legal and policing infrastructure, statistics, weather, public-health surveillance, open scientific outputs — should be tax-financed and provided either directly by the state or through autonomous public agencies. Club goods (toll roads, pay-TV, many infrastructure services) should not be classed as public goods and can be provided privately under standard contract. The policy error to avoid is inflation of the public-good category to justify state provision of services that are in fact excludable and rival.

Framework position

Pure non-excludable, non-rival public goods are the least contested case for state provision. The free-rider mechanism is mathematically clean, empirically confirmed by the near-universal pattern of state provision of defence and basic legal infrastructure across regime types, and admits a straightforward institutional solution via tax finance. The framework endorses state provision of the narrow set of genuinely non-excludable non-rival goods while flagging that the rationale is often stretched to cover club goods and ordinary services, and that such stretching undermines rather than strengthens the case for state provision where it genuinely belongs. The policy discipline is to maintain the category carefully rather than expand it.