A public good is non-rival and non-excludable. Non-rival: your consumption does not reduce mine. Non-excludable: the provider cannot practically exclude non-payers from enjoying it. National defence, clean air, lighthouses, open-source software, and basic research fit the definition. Once the good exists, everyone can use it; no one can be barred. That creates a funding problem. If you cannot exclude non-payers, you cannot charge them. If you cannot charge them, voluntary provision fails. The market underprovides or does not provide at all.
Paul Samuelson formalised the concept in the mid-1950s. His point was mathematical: for a pure public good, the efficient quantity is where the sum of everyone's marginal benefits equals marginal cost. But no private firm can capture that sum through voluntary payment, because each user has an incentive to free-ride — to enjoy the good without paying. So private provision stops short of the socially efficient level. The result is underinvestment in things society values: defence, disease eradication, climate stability, fundamental science, open protocols.
The strategic implication is twofold. First, when you are analysing why something valuable is underprovided, check whether it has public-good character. If benefits are non-rival and non-excludable, market failure is the default. Second, when you are building or funding such a good, you need a mechanism that does not rely on voluntary payment: government provision, mandatory contributions, bundling with an excludable private benefit, or philanthropy. Wikipedia survives on donations and institutional grants, not on charging readers. GPS is provided by the U.S. government; the private sector builds applications on top. The good is free at the point of use; the funding comes from elsewhere.
Many goods are only partly public. A road is non-rival until it congests, then it becomes rival. A software library is non-excludable if it is open-source, excludable if it is proprietary. Club goods are excludable but non-rival within the club (e.g. a paywalled journal). The more excludable or rival a good becomes, the more likely private provision works. The purer the public good, the more you need a non-market solution.
In business, the public-good lens helps explain underinvestment in shared infrastructure. Industry standards, open protocols, and certification schemes benefit all participants but are underprovided by any single firm because the benefit is diffuse and non-excludable. The firms that do invest often do so strategically: they fund the good to build adoption or lock-in and capture value through complements (products, services, or membership) that are excludable. The model does not say that public goods are never provided privately; it says they are underprovided unless a non-market mechanism or a strategic bundling makes provision rational for at least one actor.
Section 2
How to See It
Public goods show up wherever a valuable good or service benefits many people but no one can be charged for marginal use. Look for underprovision of something that would be worth more to society than its cost, or for funding that comes from taxes, donations, or bundling rather than from direct user payment.
Business
You're seeing Public Goods when a company invests in an industry standard or an open protocol that benefits the whole ecosystem. The standard is non-rival — every implementer can use it — and often non-excludable. The firm that develops it bears the cost; competitors free-ride. The strategic move is either to make the good excludable (license, certification) or to fund it as a loss leader and capture value elsewhere (ecosystem lock-in, complementary products).
Technology
You're seeing Public Goods when critical open-source infrastructure — Linux, OpenSSL, core libraries — is maintained by a small number of contributors while millions of users and enterprises depend on it. The code is non-rival and non-excludable. Voluntary contribution is insufficient; funding comes from foundations, corporate sponsors, or a single commercial steward. The sustainability question is always: who pays, and for how long?
Investing
You're seeing Public Goods when a company's moat depends on a shared resource — e.g. a talent pipeline, a standard, or a regulatory framework — that is underfunded because benefits are diffuse. The investment thesis must ask who funds that resource and whether that funding is stable. If the resource is a public good with no durable funder, the moat may be fragile.
Markets
You're seeing Public Goods when governments fund basic research, disease surveillance, or climate monitoring. The output is non-rival and non-excludable; private firms would underinvest because they cannot capture the full benefit. Public provision or public subsidy is the standard response. When that provision is cut or politicised, the good is underprovided and everyone bears the cost.
Section 3
How to Use It
Decision filter
"Before assuming the market will supply something valuable, ask: is it non-rival and non-excludable? If yes, expect underprovision unless there is a non-market funder — government, mandate, philanthropy, or bundling with a private good. Design or invest accordingly."
As a founder
Identify which parts of your value chain are public-good-like. Open standards, shared data, and ecosystem trust are often underprovided because benefits are diffuse. Either fund them yourself as a strategic cost (and capture value via lock-in or complements) or push for collective funding — consortia, industry levies, government — so the good is sustained. Do not assume voluntary contribution will be enough. If you build a platform that depends on a public good (e.g. an open protocol), secure a durable funding mechanism or you will inherit the underprovision problem.
As an investor
When a company's strategy relies on a shared good — a standard, a talent pool, a regulatory environment — ask who pays for it and whether that is stable. Businesses that sit atop underfunded public goods are exposed to collapse or political capture of the funding. Conversely, businesses that provide public goods in a way that is bundled with excludable value (e.g. open core with enterprise features) can capture enough to sustain the good and the enterprise.
As a decision-maker
Use the public-good lens to prioritise what only collective action can fix. Things that are non-rival and non-excludable will not be supplied efficiently by the market. That justifies regulation, taxation, or public provision — but also clarifies where private actors can add value by making the good partly excludable (club goods, licensing) or by bundling it with something they can charge for.
Common misapplication: Calling everything that is "good for society" a public good. The term is technical: non-rival and non-excludable. Many socially valuable goods are excludable (education, healthcare in many systems) or rival (housing, bandwidth). The funding and policy implications differ.
Second misapplication: Assuming that because something is a public good, government must provide it. Government is one solution. Others include mandated private contribution (e.g. R&D levies), philanthropy, and bundling with private goods. The right mechanism depends on the good and the political context.
Netflix's massive content spend produces a form of public good: a large catalogue of originals and licensed titles that defines the streaming category. The content is non-rival (one viewer's stream does not reduce another's) but Netflix makes it excludable via subscription. By funding the "good" — high-quality series and films — and controlling access, Netflix captures value through membership rather than per-use payment. The strategic lesson: when the underlying asset has public-good character, bundle it with an excludable relationship (subscription, account) so you can fund provision.
Bezos treated infrastructure as a shared good that was underprovided: reliable, scalable compute and storage. AWS made that infrastructure excludable — you pay for usage — while preserving some non-rivalry (Amazon's scale spreads fixed cost across many users). The move turned a category that could have been a public-good problem (who builds and maintains the backbone?) into a paid service. The same logic applied to fulfilment and logistics: make a shared capability excludable and charge for it, so provision is sustainable.
Section 6
Visual Explanation
Public goods — Non-rival (your use doesn't reduce mine) and non-excludable (provider can't bar non-payers). Voluntary payment fails; provision requires taxation, mandate, or bundling with a private good.
Section 7
Connected Models
Public goods sit at the intersection of collective action, externalities, and funding. The models below either explain why provision fails (free-rider, collective action), extend to related resource problems (tragedy of the commons), or describe policy and incentive responses (externalities, marginal analysis).
Reinforces
Free-rider Problem
When a good is non-excludable, rational users do not pay; they free-ride. The free-rider problem is the behavioural consequence of public-good structure. Public goods theory explains why the good is underprovided; free-rider theory explains why voluntary contribution fails. Both point to the same fix: change the structure so that contribution is mandatory, observable, or tied to a private benefit.
Reinforces
Collective Action Problem
Collective action problems arise when individuals must coordinate to produce a shared benefit. Public goods are a central case: the benefit is shared and non-excludable, so each individual prefers others to pay. Olson's logic — large groups underprovide unless selective incentives or coercion apply — is the same as Samuelson's: voluntary payment cannot fund the efficient quantity.
Tension
Tragedy of the Commons
In the tragedy of the commons, individuals over-extract from a shared resource (rival, often non-excludable). In the public good case, individuals under-contribute to producing a shared benefit (non-rival, non-excludable). One is a problem of too much use; the other of too little provision. Both stem from misaligned individual incentives; solutions often involve property rights, quotas, or mandatory contribution.
Tension
Marginal Cost/Benefit
Section 8
One Key Quote
"[E]ach individual's consumption of such a good leads to no subtraction from any other individual's consumption of that good."
— Paul Samuelson, 'The Pure Theory of Public Expenditure' (1954)
Samuelson is defining non-rivalry — the first pillar of a public good. Your use of national defence or clean air does not reduce mine. That property, combined with non-excludability, is why voluntary exchange cannot supply the efficient quantity. The quote is the cleanest statement of what makes a public good different from a private one. Once you see it, you see why markets stop short and why provision requires a different mechanism.
Section 9
Analyst's Take
Faster Than Normal — Editorial View
Public goods are underprovided by definition. If a good is non-rival and non-excludable, the market will not supply the efficient amount. That is not a market failure in the sense of a correctable glitch; it is a structural feature. The strategic question is always: who funds it, and how? Governments fund defence and basic research. Philanthropy and foundations fund some open-source and public health. Bundling funds many shared resources — you pay for the excludable part (subscription, license) and the non-excludable part comes along.
When your strategy depends on a public good, secure the funding. Do not assume the community or the ecosystem will chip in. They have the wrong incentives. If you are the steward of an open standard or a shared dataset, either make it excludable enough to charge (club good, enterprise tier) or lock in a durable funder — grant, mandate, or strategic sponsor. Sustainability is a design choice, not a hope.
The line between public and private is often blurry. Many goods are partly rival (congestible) or partly excludable (freemium). The more you can make the good excludable, the more you can fund it through the market. The strategic move is often to create a hybrid: a public-like core that builds adoption and an excludable layer that captures value. Open core, closed enterprise. Free tier, paid tier.
Use the lens to explain underprovision. When something valuable is chronically underfunded — basic research, pandemic preparedness, climate monitoring — ask whether it has public-good character. If it does, the fix is not to shame people into paying; it is to create a funding mechanism that does not rely on voluntary contribution. That may be government, regulation, or a consortium with mandatory dues.
Club goods and toll goods are the escape hatch. When you can make a good excludable (members only, paywall, license), you can fund it through the market even if it remains non-rival. That is why streaming services, paywalled research, and enterprise software work: the good is non-rival in use but excludable at the gate. The strategic move for many "public-good-like" assets is to add just enough excludability to charge, while keeping the non-rival core that drives adoption and value.
Section 10
Test Yourself
Is this mental model at work here?
Scenario 1
A government funds a nationwide disease surveillance system. Data is published openly. Hospitals, pharma, and insurers use it; none pay for it directly.
Scenario 2
A startup opens its API and documentation to everyone. It makes money from enterprise support and managed hosting.
Scenario 3
A city builds a new park. It charges an entry fee to cover maintenance.
Scenario 4
An open-source project has 10 million users and three maintainers. One maintainer asks for donations; few respond.
Section 11
Summary & Further Reading
Summary: A public good is non-rival (your use does not reduce mine) and non-excludable (the provider cannot bar non-payers). Voluntary payment fails because everyone can free-ride. The market underprovides. Provision requires a non-market mechanism: government, mandate, philanthropy, or bundling with an excludable good. Use the model to explain why certain valuable goods are chronically underfunded and to design sustainable funding when you build or depend on shared, non-excludable value. When your strategy relies on a public good, secure a durable funder or make part of the good excludable so you can charge for it.
The foundational definition. Samuelson defines public goods and shows that efficient provision requires Σ(MB) = MC, which voluntary payment cannot achieve when exclusion is impossible. Short and precise; the starting point for any serious treatment of public goods and market failure.
Olson extends the logic to groups: why large groups underprovide collective goods and how selective incentives or coercion can fix it. Essential for understanding when and how public goods get funded. The small-group exception and the role of leadership and selective benefits are key for strategy.
Accessible treatment of public goods, externalities, and the role of government in correcting market failure. Good for connecting theory to policy and real-world examples. Covers impure public goods, club goods, and the empirical evidence on underprovision.
Stiglitz covers public goods, asymmetric information, and government failure. Useful for the boundary between when government should provide and when other mechanisms work better. Includes the critique of government provision and the case for mixed solutions.
Ostrom shows how some common-pool resources are successfully managed without government or privatisation. Complements public goods theory by showing that collective provision can work under certain institutional conditions — clear boundaries, monitoring, graduated sanctions, and conflict resolution. Relevant when the "good" is a shared resource rather than a pure public good.
Efficient provision of a public good requires Σ(MB) = MC. Marginal analysis still applies — produce until the sum of marginal benefits equals marginal cost — but the market cannot reveal or aggregate those benefits through price, because exclusion is impossible. So marginal logic defines the target; it does not deliver it without a non-market mechanism.
Leads-to
Positive & Negative Externalities
Public goods are an extreme form of positive externality: the entire benefit spills over to non-payers. Subsidising or providing public goods is the standard response to positive externalities that are large and non-excludable. The same logic — align private incentive with social benefit — underlies both; public goods are the case where the spillover is total.
Leads-to
Incentives
Funding public goods requires changing incentives: mandatory payment (taxes), selective incentives (rewards for contributors), or bundling with a private good so that paying for the private part funds the public part. The design of those incentives determines whether the good is provided and at what level.
