Public Goods, Merit Goods, and Value Pluralism

Why this matters in public value

Every budget round, somewhere, a committee is handed a spreadsheet that ranks a flood-defense scheme, a road-widening project, and a drug-treatment service on the same single score, usually a benefit-cost ratio, and is asked to fund the highest-scoring options first. The spreadsheet is useful, but it hides a category error. A flood-defense scheme protects a whole town whether or not any resident pays for it; a road-widening project mostly benefits the people who drive that road; a drug-treatment service is funded partly because society has judged that recovery matters more than the addicted person's own preference in the moment of craving. These are not three points on one curve — they are three different kinds of good, justified by three different kinds of reasoning, and collapsing them into a single number does not resolve the difference. It only hides it from the people who will have to defend the decision later.

This topic gives public managers, elected leaders, and social-sector executives the taxonomy needed to know which situation they are actually in, plus the vocabulary of value pluralism for the more uncomfortable cases where the situation resists a single metric altogether. Topic 1.3 — Market Failure and Government Failure (see Topic 1.3 — Market Failure and Government Failure) explains why markets under-supply certain goods and why collective provision is not automatically the cure; this topic goes one level deeper into which good is which, and into why the values used to judge the remedy are plural rather than singular. Getting the taxonomy wrong is not a technicality. A council that treats a congestible club good as if it were a pure public good will underprice it and watch it degrade; a health agency that treats a merit good as a purely personal choice will under-invest in it; a minister who treats liberty and security as fungible, tradeable at some fixed exchange rate, will eventually be unable to explain a decision to either the people who lost liberty or the people who lost security.

Core concepts

The starting point is a two-by-two matrix built on two properties economists use to classify goods: rivalry and excludability. A good is rivalrous if one person's consumption reduces what is left for someone else — a loaf of bread, a hospital bed, a seat on a train. A good is excludable if a provider can feasibly stop someone who has not paid, or is not authorized, from consuming it — a toll road with a barrier, a subscription channel with a paywall. Crossing these two properties produces four categories, and the taxonomy comes from the work of the American economist Paul Samuelson, whose 1954 formalization of the pure public good — non-excludable and non-rivalrous by definition — remains the founding statement of the field.

A private good is rivalrous and excludable: ordinary market goods, from bread to haircuts, which markets are generally well-suited to price and allocate. A club good is excludable but non-rivalrous, at least up to a capacity limit: a members' swimming pool, a toll motorway at off-peak hours, a subscription database. Providers can charge for access, so under-provision by the market is less of a problem than with a true public good, but club goods become rivalrous once congestion sets in, which is why capacity management is as much a public value question as pricing is. A common-pool resource is rivalrous but hard to exclude anyone from: an inshore fishery, an aquifer, an unmetered grazing common, urban road space at rush hour. Because it is costly to stop someone using it but easy for them to deplete it for everyone else, common-pool resources are exposed to what the ecologist Garrett Hardin popularized as the "tragedy of the commons" — individually rational use that is collectively ruinous. The political economist Elinor Ostrom showed, against the assumption that only privatization or state control could prevent that outcome, that communities repeatedly design their own durable governance institutions — monitoring, graduated sanctions, and locally adapted rules — that manage common-pool resources sustainably without either. Finally, a public good (economics) is both non-rivalrous and non-excludable: national defense, a lighthouse, an epidemiological early-warning system, or the rule of law. Because no one can practically be charged and one person's benefit does not diminish another's, markets systematically under-supply these goods, which is the classic "free-rider problem" — the rational strategy for any individual is to let others pay and consume anyway, so, left alone, too little gets produced or maintained by anyone.

That matrix explains under-provision on economic grounds alone. A second, distinct category explains why societies intervene even where a good is privately rivalrous and excludable and a market could, in principle, supply it: the merit good. The German-American economist Richard Musgrave (economist) introduced the term in 1957 for goods — his examples included school meals and public housing — that society judges people should consume more of than their own willingness to pay would produce, on grounds of benefit rather than market demand: primary education, vaccination, and public libraries are standard cases. Its mirror image is the "demerit good": a good, such as tobacco, unregulated gambling, or recreational drugs, that society judges people should consume less of than their preferences alone would produce, again independent of ordinary market failure. Merit and demerit goods are conceptually distinct from public goods: a vaccine is excludable and, at the level of a single dose, rivalrous, so it fails the public-good test on economic grounds, yet a government may still subsidize or mandate it, because vaccination also generates a public good — herd immunity — as a side effect, and because society has made a collective judgment about the good's value that goes beyond the sum of individual preferences. That is also what makes merit goods contested in a way pure public goods are not: subsidizing a lighthouse is rarely accused of paternalism, but subsidizing swimming lessons, or restricting sugary drinks, always risks the charge that the state is substituting its judgment for the citizen's own.

The deepest concept in this topic is not about goods at all, but about the values used to weigh them: value pluralism. The philosopher Isaiah Berlin argued that fundamental human values — liberty, equality, security, justice, efficiency — are each genuinely valid, frequently in conflict, and not reducible to a single common currency that would let a decision-maker calculate the "correct" trade-off the way a shopper compares two prices. More liberty for one group can mean less security for another; more equity in outcomes can cost efficiency in production; more of one legitimate public good can come only at the expense of another, with no formula that converts them onto one axis. This is the sense in which public value is plural, not singular: unlike a market price, which lets utterly different goods be compared because they are all expressed in the same unit of currency, public value routinely requires comparing things that have no common unit at all. Value pluralism does not mean "anything goes" or that trade-offs cannot be reasoned about — it means the reasoning has to happen in the open, through argument and legitimate process, rather than being smuggled into a spreadsheet as if it were arithmetic.

Best practices

1. Classify the good before choosing the instrument. Before deciding whether to provide, subsidize, regulate, or simply price a service, work out where it sits on the rivalry/excludability matrix — the right instrument for a club good (pricing plus capacity management) is rarely the right instrument for a common-pool resource (governance plus monitoring) or a public good (direct provision or guaranteed funding).
2. Separate the economic question from the values question. Ask "would the market under-supply this?" (a market-failure question, see Topic 1.3 — Market Failure and Government Failure) and "should we override individual preference here?" (a merit-good question) as two distinct questions, because the justification, the evidence needed, and the risk of backlash differ sharply between them.
3. Name the values in tension explicitly. When a decision pits liberty against security, or equity against efficiency, say so in the business case rather than folding both into one weighted score; a reader who cannot see the trade-off cannot hold the decision-maker accountable for how it was struck.
4. State the justification for every merit-good intervention. "People will be better off" is not sufficient on its own; be specific about whether the justification is externalities (herd immunity), information asymmetry (people underestimate long-run harm), intergenerational effects (children cannot consent to their own under-education), or a straightforward collective judgment about human flourishing — each carries a different burden of proof.
5. Watch congestion thresholds on club goods. A club good behaves like a public good until it does not; monitor usage against capacity and adjust pricing, rationing, or investment before congestion turns a non-rivalrous benefit into a rivalrous, degraded one.
6. Design common-pool governance instead of assuming only two options exist. Privatization and state control are not the only ways to prevent a tragedy of the commons; consider monitoring, graduated sanctions, and user-designed rules of the kind Elinor Ostrom documented, particularly where the resource is local and its users are identifiable.
7. Revisit merit- and demerit-good classifications periodically. Social consensus shifts — attitudes to gambling, cannabis, and even sugar have moved within a generation — so treat the classification as a live policy question with a review cycle, not a permanent settlement.
8. Use multi-criteria methods when a single score would mislead. Where values are genuinely incommensurable, use structured multi-criteria decision frameworks that keep each value visible in its own column, rather than forcing a common-denominator score that quietly embeds one contestable set of exchange rates.
9. Route irreducible value conflicts to the authorizing environment, not to the analyst. A technical team can inform a trade-off between liberty and security; it should rarely be left to resolve one alone, because that resolution is a legitimacy question as much as an analytical one (see Topic 1.2 — The Strategic Triangle: Legitimacy, Value, and Capacity).
10. Distinguish a public good from a good the public wants. Popularity is not the test; a well-attended festival is not thereby a public good, and an unpopular disease-surveillance system may be one of the purest public goods a government funds.

Questions to discuss with your team

Which of our services are genuinely non-excludable and non-rivalrous, and which have we been treating as public goods out of habit? Many services described internally as "public goods" are, on inspection, club goods with unmanaged congestion, or common-pool resources with no governance at all. Walk through your five largest services against the rivalry/excludability matrix and be honest about where each one sits. The answer changes what the right fix looks like: a congested club good needs pricing or capacity investment, not a bigger subsidy. Getting the category wrong wastes money and does not fix the underlying problem.

Where are we substituting a single cost-benefit score for a decision that is really a value trade-off? Look at the last three contested budget decisions your organization made and ask whether the recommendation paper showed the underlying value conflict or hid it inside one composite number. If a benefit-cost ratio was doing work that a genuine liberty-versus-security or equity-versus-efficiency argument should have done, that is worth naming now, before the next contested decision arrives. Cost-benefit analysis (see Topic 2.2 — Social Cost-Benefit and Cost-Effectiveness Analysis) is a legitimate tool for many decisions, but not a universal solvent for all of them. Decide, as a team, where its limits are for your portfolio.

What is our justification for each merit-good intervention we run, and whose preference are we overriding? For every program that nudges, subsidizes, or restricts behavior beyond what a pure market-failure argument would justify, be able to state the specific reason — externality, information gap, protection of future selves or of children, or a collective judgment about flourishing. If the team cannot articulate the reason cleanly, the program is vulnerable to the charge of unexamined paternalism, and probably deserves review regardless of whether anyone has yet raised the charge.

Are any of our common-pool resources at risk of a tragedy of the commons, and do we have governance for them? Identify any shared, depletable resource your organization is responsible for — road space, a fishery, an aquifer, even a shared digital commons like open data or unmoderated public space — and check whether it has real governance (monitoring, enforceable rules, a feedback loop) or is simply assumed to look after itself. Ostrom's research suggests durable local governance is achievable well short of privatization or heavy-handed state control; discuss whether your organization has actually built that governance or has just avoided the question.

When our core values conflict, how do we make that conflict visible to the people who authorize our work, rather than resolving it quietly inside the analysis? Public value depends on legitimacy and support as much as on the analysis itself (see Topic 1.2 — The Strategic Triangle: Legitimacy, Value, and Capacity). Discuss a recent decision where liberty, equity, efficiency, or security genuinely pulled in different directions, and whether your governing body or authorizing environment was given the real trade-off to weigh in on, or was handed a conclusion that had already made the trade-off for them without saying so.

How do we revisit merit- and demerit-good classifications as social consensus shifts, and who owns that review? Classifications that were settled a decade ago — around gambling, recreational drugs, screen time, or diet — may no longer match current evidence or public attitudes. Agree who is responsible for periodically re-opening these classifications, on what cycle, and with what evidence threshold, so that the organization is not caught defending a position it adopted for reasons that no longer hold.

In practice: a public value example

A fictional coastal town, Port Avon, illustrates how the taxonomy plays out in a single council's portfolio. Its harbor authority manages four things that look superficially similar — "assets by the water" — but that behave as four different kinds of good, and its finance director had, for several years, ranked all four on the same benefit-cost spreadsheet before a new head of strategy arrived and asked the team to reclassify each one first.

The harbor's 40 berths were being treated as a public amenity funded from general taxation, priced far below demand, with a permanent waiting list. Reclassified, they were a textbook club good: excludable (a berth can be assigned or withheld) and non-rivalrous only up to the 40-berth limit. The fix was not a bigger subsidy but a waiting-list auction for a portion of berths and a modest fee rise, which cleared the backlog and funded maintenance the general budget had never covered. The inshore mussel beds just offshore were a common-pool resource: any boat could dredge them, and catch had fallen for three seasons running, a slow-motion tragedy of the commons. Rather than close the beds or auction exclusive rights to one operator, the council convened the dozen local boats that actually used them and, drawing on the same logic Elinor Ostrom documented in fisheries elsewhere, agreed a self-monitored quota with graduated penalties for over-catching, enforced by the fishers' own association rather than by the council. The town's storm-surge early-warning siren system was a genuine public good: non-excludable (every resident hears it, whether or not their household has paid the harbor levy) and non-rivalrous (one household's warning does not reduce another's). It had been chronically under-funded precisely because no individual household's willingness to pay reflected its true collective value, and the strategy team's main recommendation was to fund it centrally and stop trying to attribute its cost to individual properties at all. Finally, a subsidized children's swimming-lesson program was a merit good: swimming lessons are rivalrous and excludable, so a market could supply them, but the council judged that under-provision of a basic drowning-prevention skill among low-income families was a harm worth overriding ordinary willingness-to-pay for, particularly given the town's tidal currents. The strategy team documented the justification explicitly — child safety and an externality argument about emergency-service costs, not simply "swimming is good for you" — so that the subsidy could be defended on its actual merits rather than on a vague appeal to wellbeing.

The point of the exercise was not that any one instrument was wrong before reclassification, but that ranking a pricing problem, a commons-governance problem, a chronic-underfunding problem, and a paternalism-justification problem on one benefit-cost table had obscured which lever actually needed pulling in each case.

Four sector lenses

Local government

Local authorities carry the largest share of everyday club goods and common-pool resources: car parks, allotments, leisure centers, local fisheries, shared green space, and road capacity are all classic cases of congestion and depletion risk that a national ministry rarely has to manage at street level. The practical skill at this level is less abstract classification and more operational: pricing and capacity management for club goods, and designing enforceable local rules for common-pool resources before congestion or depletion becomes politically visible.

National government

National governments are the primary funders of large-scale public goods — defense, monetary and financial stability, disease surveillance, the justice system — and of nationally delivered merit goods such as compulsory education and immunization programs. The distinctive challenge here is scale and universality: a national merit-good program must justify itself to a much more diverse population, with more genuinely competing value claims about liberty and paternalism, than a local pilot ever has to.

Social sector and nonprofit

Nonprofits and non-governmental organizations (NGOs) are frequently the delivery arm for merit goods — literacy programs, addiction recovery, maternal health outreach — without the state's coercive authority to mandate or tax for them. Their legitimacy to encourage a merit good or discourage a demerit good rests on mission, trust, and voluntary relationship rather than law, which means the values-pluralism problem shows up as a persuasion and consent problem: a charity that pushes too hard on what it has decided is good for its beneficiaries risks the same paternalism critique as a government agency, with none of the government's democratic mandate to fall back on.

Multilateral and international

At the international level, many of the largest public goods are genuinely global — climate stability, pandemic preparedness, a stable international financial system, an open and functioning internet — and no single state can be excluded from the benefit or compelled to pay its share, producing free-riding among nations rather than individuals. Multilateral bodies exist substantially to solve this coordination problem, negotiating burden-sharing arrangements where market and national incentives alone would under-supply the good; the climate case specifically is developed further in Topic 4.3 — Climate, Sustainability, and Environmental Public Value.

Common failure modes

• False commensurability. Forcing every decision through a single monetized or weighted score, even when the underlying values are genuinely incommensurable, produces a spuriously precise number that hides the real argument and collapses under challenge once someone asks what exchange rate was assumed between, say, a life saved and a mile of road built.
• Treating a club good as a pure public good. Under-pricing a congestible asset because it "feels" public leads to over-use, waiting lists, and a maintenance backlog, because the revenue that should fund capacity was never collected.
• Unexamined merit-good paternalism. Subsidizing or restricting a good without a stated, specific justification invites — and deserves — the charge that an organization is substituting its preferences for citizens' own; the backlash often lands hardest on programs that had a perfectly good justification but never wrote it down.
• Commons neglect until collapse. Treating a shared, depletable resource as if it would look after itself, with no monitoring or enforceable rules, until the fishery, aquifer, or shared digital resource is visibly degraded and only heavy-handed intervention remains.
• Demerit-good bans without addressing the merit-good gap. Prohibiting a demerit good (an addictive substance, for instance) while doing nothing to expand the merit good that would reduce demand for it (treatment capacity, for instance) reliably pushes the behavior underground rather than reducing it.
• Silent value substitution. Letting one value quietly stand in for another across an entire portfolio — for example treating "efficiency" as if it were the whole of "public value" — until equity or legitimacy concerns resurface as a crisis rather than a routine trade-off.

Maturity model

Checklist

• Every major service has been placed on the rivalry/excludability matrix, with the classification documented.
• Club goods have a pricing or capacity-management plan that responds to congestion before it becomes a waiting list.
• Common-pool resources have identified monitoring and enforceable rules, not an assumption of self-regulation.
• Every merit-good subsidy or demerit-good restriction has a written justification distinct from "it is good for people".
• Merit- and demerit-good classifications are on a review cycle with a named owner.
• Business cases and board papers name value conflicts explicitly rather than only presenting a composite score.
• Multi-criteria or structured qualitative methods are available as an alternative to forced monetization.
• Genuinely plural value trade-offs are routed to the authorizing environment, not resolved solely by analysts.
• Public-good funding is not made contingent on individual willingness to pay where that would systematically under-fund it.
• Cross-border or cross-jurisdiction public goods have an identified coordination mechanism to reduce free-riding.

Key sources

• Paul A. Samuelson, "The Pure Theory of Public Expenditure", The Review of Economics and Statistics, 1954 — the founding formalization of the public good.
• Richard A. Musgrave, The Theory of Public Finance, McGraw-Hill, 1959, and his earlier 1957 introduction of the merit-good concept.
• Elinor Ostrom, Governing the Commons: The Evolution of Institutions for Collective Action, Cambridge University Press, 1990.
• Isaiah Berlin, "Two Concepts of Liberty" (1958) and subsequent essays developing value pluralism.
• Mark H. Moore, Creating Public Value: Strategic Management in Government, Harvard University Press, 1995 — for the strategic-triangle framing this topic builds on.
• OECD and World Bank publications on global public goods, listed in full below.

References

1. Public good (economics) — Wikipedia — https://en.wikipedia.org/wiki/Public_good_(economics)
2. Club good — Wikipedia — https://en.wikipedia.org/wiki/Club_good
3. Common-pool resource — Wikipedia — https://en.wikipedia.org/wiki/Common-pool_resource
4. Merit good — Wikipedia — https://en.wikipedia.org/wiki/Merit_good
5. Demerit good — Wikipedia — https://en.wikipedia.org/wiki/Demerit_good
6. Value pluralism — Wikipedia — https://en.wikipedia.org/wiki/Value_pluralism
7. Isaiah Berlin — Wikipedia — https://en.wikipedia.org/wiki/Isaiah_Berlin
8. Paul Samuelson — Wikipedia — https://en.wikipedia.org/wiki/Paul_Samuelson
9. Richard Musgrave (economist) — Wikipedia — https://en.wikipedia.org/wiki/Richard_Musgrave_(economist)
10. Tragedy of the commons — Wikipedia — https://en.wikipedia.org/wiki/Tragedy_of_the_commons
11. Elinor Ostrom — Wikipedia — https://en.wikipedia.org/wiki/Elinor_Ostrom
12. Free-rider problem — Wikipedia — https://en.wikipedia.org/wiki/Free-rider_problem
13. Development co-operation and the provision of global public goods — OECD — https://www.oecd.org/content/dam/oecd/en/publications/reports/2023/05/development-co-operation-and-the-provision-of-global-public-goods_e28bebd8/aff8cba9-en.pdf
14. International Cooperation and Global Public Goods — World Bank — https://www.worldbank.org/en/programs/knowledge-for-change/brief/Int-Coop-Global-Goods