Case supporting the view. A pure public good is both non-rivalrous and non-excludable. Non-excludability means that private firms cannot charge users who refuse to pay, and non-rivalry means that charging would be socially wasteful even if it were possible (an extra user adds no cost). Each consumer therefore has an incentive to free-ride on others' contributions. If every consumer reasons that way, the sum of voluntary contributions falls short of the sum of true valuations, and a profit-seeking firm cannot cover its production costs. The pure public good is under-produced, or not produced at all, in a free market. State provision funded through compulsory taxation sidesteps the free-rider problem — the state can finance and provide the good regardless of individual willingness to pay. On this logic the public-good classification is a clean economic case for government provision.
Case against the view — qualifications. Three considerations complicate the step from 'public good' to 'government provides'.
First, opportunity cost. Every dollar of public financing has a foregone alternative use (healthcare, education, debt reduction). A public good is a candidate for state provision but is not automatically worth providing — the social benefit must exceed the social opportunity cost of the resources used. A low-value public good (an under-used recreational park) may fail that test even though it has public-good characteristics.
Second, private provision is sometimes feasible. Many goods are 'quasi-public' — partially excludable or mildly rivalrous — and private firms have found contractual innovations to supply them (toll roads with electronic charging, subscription satellite TV, member-funded clubs). The classification 'public good' may therefore over-state the market-failure case for some real-world goods.
Third, government failure. State provision is not automatically efficient. Political capture, poor targeting, procurement inefficiency and bureaucratic cost can leave state-provided outcomes worse than imperfect market provision. The decision between the two depends on comparing imperfect alternatives — not comparing state provision to an idealised market or the market to an idealised state.
Fourth, 'provision' is not a single act. The state can finance a public good through tax without directly producing it — using competitive contracts with private firms to deliver the service (defence procurement, flood engineering, satellite launches). Recognising this distinction weakens the claim that classification immediately dictates WHO produces.
Reasoned judgement. Classifying a good as a public good establishes that the free market will under-provide it — this is the robust implication of non-excludability and non-rivalry. It does NOT, by itself, establish that government provision is the right remedy in every case. The classification is a necessary signal of market failure, not a sufficient condition for state production. State provision is justified when (i) the social value of the good is large relative to the opportunity cost of public funds, (ii) private contractual solutions are unavailable or inferior, and (iii) government failure risks are manageable. Pure public goods with large social benefit — national defence, basic flood defence, lighthouse-type services, non-excludable public health programmes — meet these conditions and are canonical cases for state provision. For impure public goods or those with modest benefits, the classification by itself is not decisive — the full comparison of imperfect state and imperfect market alternatives is required.