How The World Works | Public Goods

A lighthouse does not shine only for the ship that paid.

Once the beam crosses the water, every ship in range can use it.

Your seeing the light does not make the beam dimmer for me. And unless the lighthouse can somehow identify every ship and block the beam from non-payers, exclusion is difficult.

This creates a strange economic problem.

The thing can be extremely valuable.

Yet the usual market question—“Who will pay for their own unit?”—does not fit very well, because there may not be a separate unit for each user.

This is the world of public goods.

The phrase is easy to misunderstand because ordinary speech uses “public good” to mean “something good for society.” Economics uses a narrower definition.

A public good is not defined by being morally admirable, government-funded or open to the public.

It is defined by two properties.

Quick Read

A pure public good is non-rival and non-excludable.

Non-rival means one person’s use does not substantially reduce what remains available for another person.

Non-excludable means it is difficult or costly to prevent people from benefiting once the good is provided.

OpenStax uses national defence as a classic example. If national defence protects a territory, one resident’s protection does not meaningfully remove protection from another, and it is difficult to defend the territory while selectively excluding a particular resident.

These properties create the free-rider problem.

If people can benefit without paying, each individual may prefer somebody else to fund the good. If enough people reason that way, a valuable good can be underprovided.

The central question is:

How do you fund and govern something that many people can benefit from at once when each person has an incentive to let somebody else pay?

The One-Sentence Answer

Public goods work by creating benefits that are hard to divide into individually saleable units, so ordinary pricing can underfund provision and societies develop alternative mechanisms such as taxation, membership, sponsorship, philanthropy, bundling, regulation and social norms.

The Two Properties

Start with rivalry.

A sandwich is rival. If I eat it, you cannot eat the same sandwich.

A downloadable mathematical theorem is mostly non-rival. My reading the theorem does not stop you reading it too.

Now excludability.

A cinema seat is excludable. A ticket and door control can prevent non-payers from entering.

A city-wide mosquito-control programme is harder to exclude people from. Once mosquitoes are reduced across an area, nearby residents can benefit whether or not each one personally paid.

Pure public goods combine low rivalry with difficult exclusion.

The Four-Quadrant Map

A useful way to read goods is to place them in a two-by-two map.

Private goods: rival and excludable.
Club goods: relatively non-rival until congestion, but excludable.
Common-pool resources: rival but difficult to exclude.
Public goods: non-rival and difficult to exclude.

This map is more useful than the everyday categories “public sector” and “private sector.”

A publicly funded hospital appointment is still rival because one appointment slot cannot serve unlimited patients at once. A private radio broadcast can have public-good characteristics because one listener does not prevent another listener from receiving the same signal.

Ownership and economic characteristics are different questions.

Publicly Funded Is Not the Same as Public Good

This distinction is worth repeating because it prevents an enormous amount of confusion.

A government may provide education, healthcare, buses, parks, museums and housing.

These services can have public benefits, positive externalities or policy reasons for public funding.

But many are not pure public goods.

A bus becomes crowded. A classroom has finite capacity. A hospital bed can serve one patient at a time. A museum gallery becomes congested.

The correct analysis starts from properties, not political labels.

The Public-Good Chain

shared benefit → difficult exclusion → incentive to free ride → weak private payment signal → underprovision risk → collective funding / institutional solution → widespread benefit

The phrase “underprovision risk” matters.

It does not mean markets never provide anything with public-good characteristics.

Radio has been funded by advertising. Open-source software can be funded by companies, foundations or contributors who receive indirect benefits. Research can be funded by universities, governments, philanthropy or firms seeking spillover advantages.

The mechanism simply means private willingness to pay can fail to reveal total social value.

The Free-Rider Problem

Imagine ten neighbours would each benefit $100 from a flood-warning system.

The system costs $500.

Total benefit is $1,000.

Socially, the project can be worthwhile.

But each neighbour may think: “If the others pay, I receive the warning anyway.”

If too many take that position, the system is not funded even though collective benefit exceeds cost.

This is the free-rider problem.

The problem is structural rather than necessarily moral.

A person can be generous and still recognise that voluntary contribution is fragile when thousands or millions of users are involved.

Public Goods and Collective Action

Public-good provision is a collective-action problem.

Everyone can prefer the shared outcome while each person has an incentive to minimise their own contribution.

This is why cooperation alone does not exhaust the explanation.

The system also needs a mechanism for contribution.

Taxes can make contribution compulsory. Membership can make access conditional. Social norms can reward contributors. Matching grants can make individual contributions more effective. Philanthropists can fund provision because they value the shared outcome directly.

Different public goods can support different provision architectures.

National Defence: The Textbook Case

National defence is frequently used because its public-good properties are unusually clear.

Once a territory is defended, residents generally share the benefit.

Protecting one resident does not normally reduce protection for another in the same way that one hospital bed serves only one patient.

Excluding individual non-payers from territorial defence would also be extremely difficult.

This makes direct individual purchase awkward.

Tax-funded collective provision becomes a practical solution.

Knowledge Has Public-Good Properties

A theorem can be used by millions of people without being consumed.

A scientific discovery can improve many later technologies.

An algorithm can be copied at low marginal cost.

Knowledge therefore has strong non-rival characteristics.

But knowledge is not always non-excludable.

Paywalls, secrecy, patents, trade secrets and access controls can create exclusion.

This is a beautiful example of how technology and law can move a good between quadrants.

The underlying information remains non-rival while institutions alter excludability.

Why Basic Research Is Often Publicly Supported

Basic research can generate benefits that spill far beyond the original researcher or firm.

If no organisation can capture enough of the total benefit, private investment may be lower than socially desirable.

Public grants, university systems and foundations therefore help finance knowledge whose benefits may be broad, uncertain and delayed.

This connects public goods with Externalities.

A positive externality describes benefits spilling outside the transaction.

A public good is the more specific case where non-rivalry and difficult exclusion make individual charging especially difficult.

Clean Air Is More Complicated Than It Looks

Clean air is often described as a public good.

The intuition is strong: one person breathing clean air does not normally prevent another from doing so, and excluding individuals from cleaner ambient air is difficult.

But environmental quality also depends on a limited absorptive capacity for pollution.

That introduces common-pool and externality dimensions.

The world does not always fit one clean textbook box.

The categories are lenses for identifying mechanisms, not prisons for reality.

Street Lighting

A streetlight illuminates an area rather than selling one beam to one pedestrian.

My using the light usually does not reduce your ability to use it.

Charging every passer-by individually would be cumbersome.

Collective provision therefore makes sense.

Yet even this simple example contains boundaries.

The physical lamp has capacity and maintenance cost. Congested streets create other rival services around it. Electricity is not free.

Public-good analysis tells us about the consumption characteristic of illumination, not every cost in the system.

Radio: A Private Business Around a Public-Good Signal

OpenStax uses radio to show that markets can sometimes fund non-rival, difficult-to-exclude services indirectly.

A conventional broadcast reaches many listeners simultaneously.

The station cannot easily charge each listener for receiving the signal.

Advertising becomes the revenue mechanism.

The listener pays not with a direct fee but with attention that advertisers value.

Technology can also change the good’s excludability. Subscription streaming encrypts access and turns a previously open broadcast into a club-like service.

Technology Changes Excludability

Excludability is not always a permanent property.

A road can be difficult to charge for until electronic tolling exists.

A broadcast can be open until encryption creates access control.

Digital content can be copied easily but placed behind authentication.

Technology therefore changes the institutional menu.

This does not mean every good should be made excludable.

It means provision models can change when exclusion becomes cheaper.

Congestion Can Turn a Public-Like Good Into a Rival One

A quiet park has low rivalry.

Your presence does not meaningfully reduce my enjoyment.

At extreme crowding, another visitor does reduce space, quiet and access.

The same physical facility changes economic character with load.

This is why categories such as club good, congestible public good and common resource become useful.

Rivalry can be state-dependent.

This connects public goods directly to Nonlinearity and to density.

Public Goods and Common-Pool Resources Are Mirror Problems

These two concepts are often confused because both can involve difficult exclusion.

The difference is rivalry.

A fishery is rival: the fish I catch is gone for you.

National defence is largely non-rival: my protection does not remove yours.

The common-pool problem is overuse.

The public-good problem is underprovision.

This is one of the cleanest conceptual pairs in economics.

See How The World Works | Common-Pool Resources.

Why Government Appears So Often

Government can solve one part of the free-rider problem through taxation.

Instead of asking every citizen for a voluntary contribution to national defence, flood control or disease surveillance, government can require collective funding.

This can make provision possible.

It does not automatically determine the correct quantity, quality or design.

Government faces its own information, incentive and accountability problems.

The public-goods argument explains why markets may underprovide.

It does not prove every government provision is efficient.

The Preference-Revelation Problem

How much public good should be provided?

This is harder than asking how much people buy at a market price.

If people believe stating a high willingness to pay will increase their tax burden, they may understate value.

If they believe somebody else will pay, they may exaggerate desired provision.

Public decision-making therefore needs alternative methods: voting, cost-benefit analysis, surveys, expert assessment, revealed behaviour and political processes.

No method is perfect.

The Marginal Social Benefit Problem

For a private good, one more unit is consumed by one person.

For a public good, one more unit of provision can benefit many people at the same time.

Economic analysis therefore aggregates the marginal willingness to pay across beneficiaries rather than treating one person’s valuation as the whole demand.

This is a conceptual reason public-good demand works differently from ordinary private-good demand.

Public Goods and Opportunity Cost

Collective provision still consumes scarce resources.

A flood barrier uses capital and land.

Defence uses people, equipment and budgets.

Research grants use tax revenue or philanthropic capital.

Calling something a public good does not make it free.

Provision has opportunity cost.

See How The World Works | Opportunity Cost.

Public Goods and Common Knowledge

Contribution can depend on what people believe others will contribute.

If a community publicly commits to funding a shared project, contribution expectations can become common knowledge.

Visible contribution by respected participants can change social norms.

Conversely, public evidence that many people are free riding can weaken willingness to cooperate.

Common knowledge does not eliminate free riding, but it changes the strategic environment.

See How The World Works | Common Knowledge.

Public Goods and Information Asymmetry

People may disagree about whether the public good is needed because they hold different information.

Experts know flood models. Residents know local water behaviour. Security agencies know threat intelligence. Citizens know their own priorities.

Provision therefore needs information aggregation as well as financing.

See How The World Works | Information Asymmetry.

Public Goods and Scale

Some public goods become more efficient at large scale because fixed costs can be spread widely.

A satellite-based weather system can serve millions of users. A national time standard can coordinate an entire economy. A disease-surveillance platform can become more informative as coverage broadens.

But scale can also weaken local accountability and increase administrative complexity.

The next article develops scale directly.

Public Goods and Density

Density changes the value and cost of public goods.

One streetlight serves more pedestrians in a dense area. One transit station serves more people. One drainage failure can affect more property. Disease-control measures can have greater benefit where interactions are more frequent.

The same public infrastructure can therefore produce different benefit per square kilometre depending on density.

Public Goods and Fixed Costs

Many public goods require large upfront investment.

A weather radar, flood barrier, satellite, scientific instrument or national digital identity system can have large setup cost and relatively low marginal cost for additional users.

This cost shape strengthens the case for scale but can also create barriers to alternative provision.

Education Contains Public-Good Layers

A seat in a classroom is rival.

But knowledge created and shared through education can be non-rival.

An educated population can create broad social benefits through productivity, civic capability, public health and knowledge spillovers.

This is why education policy should not be analysed as if it were a pure public good or pure private good.

It is a bundle containing rival services, private benefits, positive externalities and public-good-like knowledge effects.

Open Educational Resources

Digital educational material can have public-good characteristics once created.

One student downloading a lesson does not consume the lesson for another.

The marginal cost of another reader can be tiny.

But creation, editing, hosting, verification and maintenance still cost time and money.

This is a recurring public-goods pattern:

expensive to create, cheap for additional people to use

Digital Public Goods

Software, standards, public datasets and open protocols can generate broad value because replication is cheap.

Yet digital goods also require governance.

Who maintains the code?

Who fixes vulnerabilities?

Who pays for servers?

Who decides standards?

The information itself may be non-rival while maintainer attention is rival.

The public-good layer can sit on top of a common-pool resource of human attention.

Public Goods Can Be Local, National or Global

Not every public good serves the whole planet.

Street lighting can be local.

National defence can be national.

Basic scientific knowledge can be global.

The geographic scope affects governance.

A neighbourhood can coordinate one small shared facility. Global climate monitoring requires international institutions and distributed contribution.

Global Public Goods

Some benefits cross borders.

Pandemic surveillance, scientific knowledge, climate stability, space-object tracking and certain forms of international security can produce benefits beyond one state.

This makes provision harder because there is no single global government with ordinary taxation authority over all beneficiaries.

International treaties, organisations, voluntary contributions, burden-sharing rules and clubs become part of the solution space.

Club Goods: Make Exclusion Possible

One response to free riding is to create membership.

A private park, subscription network or members’ facility can remain relatively non-rival below congestion while excluding non-members.

This turns a public-good-like service into a club good.

The gain is easier funding.

The cost is exclusion of people who might otherwise benefit at low marginal cost.

Institutional design decides where that trade-off is acceptable.

Bundling

Another solution is to bundle the public good with something excludable.

Radio bundles broadcasting with advertising. A shopping centre provides decorative public space because the space attracts customers to excludable shops. A platform provides free tools because they increase demand for paid services.

The public-good component is funded indirectly by another revenue stream.

Philanthropy

Some people value the public outcome directly enough to contribute without receiving exclusive access.

Foundations fund science, museums, open knowledge, cultural preservation and public health.

Philanthropy can fill important gaps.

It can also create governance questions because donors may shape priorities differently from the broader public.

Funding solves one problem and can introduce another.

Social Norms and Contribution

People are not perfect calculating machines.

They contribute because of identity, reciprocity, reputation, fairness and community norms.

Visible contribution can increase cooperation.

Visible free riding can reduce it.

Public-good provision therefore sits partly inside culture and trust, not only price theory.

The Assurance Problem

Sometimes people are willing to contribute if they know enough others will too.

A resident may gladly pay $50 for a shared playground if the project will definitely be built.

They may refuse if they fear their $50 will be wasted because the project will not reach the funding threshold.

Crowdfunding platforms use assurance contracts, thresholds and refunds to reduce this problem.

Common knowledge of the funding condition becomes part of the mechanism.

Public Goods and Distribution

The benefit of a public good can be non-rival and still unevenly distributed.

A flood barrier may protect some neighbourhoods more than others. A national scientific database may be used more by highly connected institutions. Street lighting benefits those who travel at night more than those who do not.

Public-good status does not mean equal benefit.

See How The World Works | Distributions.

The Overprovision Risk

Public-goods theory usually highlights underprovision.

But collective institutions can also overprovide if political incentives reward visible projects, cost estimates are weak or users do not bear marginal cost.

“Public good” is not a blank cheque.

The correct quantity still depends on marginal social benefit, marginal social cost and opportunity cost.

The Quality Problem

A public good can be provided badly.

A siren that nobody can hear, a flood map nobody trusts, a public dataset full of errors or a defence system that cannot respond does not become valuable merely because the category is correct.

Provision is not receipt.

The system must work for the receiver.

The Public-Goods Audit

Define the service. What benefit is actually being produced?
Test rivalry. Does one user reduce what remains for another?
Test exclusion. Can non-payers realistically be prevented from benefiting?
Separate ownership from properties. Is the good government-provided, privately provided or mixed?
Find the free-rider problem. Can people benefit while letting others pay?
Measure social benefit. Who benefits and by how much?
Check distribution. Are benefits concentrated even if consumption is non-rival?
Check marginal cost. What does one more user or one more unit of provision cost?
Check congestion. Does rivalry appear at high use?
Check technology. Can exclusion be made cheaper?
Compare funding mechanisms. Tax, membership, advertising, philanthropy, bundling, voluntary contribution?
Check governance. Who decides quantity and quality?
Check information. How are preferences and needs revealed?
Check opportunity cost. What alternative use of resources is displaced?
Check second-order effects. How does provision alter behaviour and development?
Check receipt. Does the shared benefit actually reach intended users?

When the Public-Goods Lens Fails

The lens fails when “public good” becomes praise.

A good can be socially admirable and still be rival and excludable.

It fails when every government service is called a public good.

It fails when non-rivalry is assumed to mean zero cost. Creation and maintenance can be expensive.

It fails when the free-rider problem is treated as proof that only government can provide the good. Markets, clubs, philanthropy and social institutions sometimes produce workable alternatives.

And it fails when the category is used to avoid asking how much provision is actually worthwhile.

How Public Goods Connect to the Rest of the World

Scarcity: provision still consumes limited resources even when use is non-rival.
Opportunity cost: every collectively funded good displaces another use of resources.
Externalities: public goods often arise where benefits spill widely beyond the payer.
Common-pool resources: both can be hard to exclude from, but rivalry points in opposite directions.
Common knowledge: contribution and coordination depend on what people expect others to do.
Trust: voluntary contribution is easier when people expect others to cooperate.
Information asymmetry: users and experts may know different things about need and quality.
Scale: large user bases can spread fixed provision costs.
Density: concentration changes benefit per area and congestion.
Fixed costs: many public goods are expensive to establish and cheap to extend to another user.
Distribution: non-rival benefit can still be unevenly received.
Second-order effects: public infrastructure can reshape later private behaviour.

Questions a Reader Can Now Ask

Is the good actually non-rival?
Is exclusion genuinely difficult?
Is this publicly funded service being confused with a public good?
What free-rider incentive exists?
Could technology change excludability?
Does congestion make the good rival at high load?
Who benefits most?
Who pays?
What provision mechanism fits the scale?
How is the desired quantity decided?
What is the opportunity cost of funding it?
Could a club, sponsorship or bundled model work?
Does the public signal create enough common knowledge for coordination?
Does the good actually reach the receiver at useful quality?

Frequently Asked Questions

What is a public good?

A pure public good is non-rival and non-excludable: one person’s use does not materially reduce another’s, and excluding non-payers is difficult or costly.

Is education a public good?

Education is not a pure public good. Classroom places and teacher time are rival, but education creates broad positive externalities and knowledge that can have non-rival characteristics.

What is the free-rider problem?

It occurs when people can benefit without paying and therefore have an incentive to let others fund provision. If enough people do so, a valuable good can be underprovided.

Does government have to provide every public good?

No. Governments are important providers, but advertising, philanthropy, clubs, bundling, norms and other institutional arrangements can sometimes support provision.

What is the most important distinction?

Do not confuse public goods with common-pool resources. Public goods are non-rival and risk underprovision; common-pool resources are rival and risk overuse.

Research Basis and Further Reading

OpenStax, “Public Goods”, for non-excludability, non-rivalry, free riding and alternative provision mechanisms.
OpenStax, Chapter 13 Summary, for the core distinction between public goods and other categories.
How The World Works | Common-Pool Resources, for the rival mirror problem of difficult exclusion.

The Larger Idea

Some of civilisation’s most valuable things are awkward to sell one person at a time.

A safe coastline.

A weather warning.

A body of scientific knowledge.

A defended territory.

A reliable standard that lets strangers coordinate.

Once they exist, their value spills widely.

That is precisely what makes them difficult.

If everyone can benefit, everyone can hope somebody else pays.

So civilisation invents institutions around the gap between value and payment.

Taxes.

Grants.

Membership.

Sponsorship.

Philanthropy.

Norms.

Public law.

The mechanism is not mysterious.

When benefit can spread farther than the invoice, society needs a way to finance the distance.

How The World Works | Public Goods — Why Everyone Can Benefit Even When Nobody Wants to Pay Alone