The answer to this question bears directly on explanations of why care provision is undervalued by the market.
Ask almost anyone what a public good is and you will get an answer along the lines of “something that benefits everyone” or “something we all need and shouldn’t have to pay for individually.” Clean air, public schools, vaccines, a healthy next generation. Ask an economist and you will get something narrower and stranger: a public good is a good that is non-rival (my use of it doesn’t reduce what’s left for you) and non-excludable (nobody can be kept from using it, or at least not cheaply). A lighthouse beam, national defense, the knowledge that the earth goes round the sun.
The two definitions overlap, but they are not the same, and the gap between them causes a lot of muddled argument. Many things that benefit everyone are perfectly rival and excludable. A hot lunch for a hungry child benefits the child, and arguably the rest of us, but it is a private good in the economist’s sense–once eaten, it’s gone, and it’s easy to refuse it to anyone who can’t pay. Conversely, plenty of things that fit the technical definition don’t benefit everyone equally, and some benefit hardly anyone at all. A broadcast in a language nobody speaks is non-rival and non-excludable. It is not, in any ordinary sense, a public good.
What everyday usage is reaching for is closer to what economists, when they bother to use the term, call a merit good: something a society decides its members ought to have, whether or not the market would supply it and whether or not people would buy it on their own. The interesting question is how the technical definition came to crowd out the normative one, and whether we can get the normative one back without losing what the technical one teaches.
Samuelson Trumps Musgrave
The strict definition is usually credited to Paul Samuelson, who, in 1954, set out the mathematics of goods that everyone consumes jointly. However, the non-excludability criterion came from Richard Musgrave, whose 1959 Theory of Public Finance systematically explored the idea that some goods can’t easily be fenced off.
Over the following decade or two the two properties were fused into a tidy two-by-two table that many economics students are asked to memorize: private goods (excludable and rival) in one cell, public goods (non-excludable and non-rival) in another cell, and two cells that represent other possibilities: club goods, where members of a group can exclude others but enjoy membership services that are non-rival, and common-pool resources such as fish in the sea, that are not easily excludable, but are rivalrous in consumption.
Keep in mind that all four combinations are framed in abstract terms; in the real world there’s seldom a bright line between the criteria. Some goods and services are simply more excludable and/or rival than others—those that have a per unit market price and an underlying value that consumers can easily assess.
Musgrave was more skeptical than Samuelson of the primacy of such “private” goods. He was a German émigré steeped in the Continental tradition of public finance, in which the state was not merely a fixer of market failures but an expression of collective purposes. Alongside “social goods” he introduced the category of “merit wants”: things like education, housing, and health care that a community decides to provide because it judges them worth having, even where individuals might choose otherwise and even where the market could in principle deliver them.
This normative strand of thinking was gradually squeezed out in favor of the technical two-by-two table, which was more compatible with the primacy of individual choice. Merit goods, by contrast, required someone to say that some preferences are better than others, which most economists tried to avoid. Samuelson’s 1954 article bewailed the impossibility of coming to any reasonable agreement about the definition of “social welfare.” His view proved influential.
However, more recent thinking challenges the strict distinction between the “technical” and the “normative.”
Publicness as a Choice
“Publicness” is not always a natural property of goods or services. Almost anything can be made excludable if we spend enough on fences, tolls, and passwords, and almost anything can be made non-excludable if we decide to keep it open. Land is about as rival and excludable as a good can be, yet many societies have held it in common. Knowledge is inherently non-rival, yet patent law makes it excludable on purpose. Some argue persuasively that most public goods are those which society has chosen to keep in the public domain, so the question of what belongs there is largely a political one.
There is a lot to be said for this argument. It recovers Musgrave’s point that the boundary between public and private is drawn, not discovered. But the distinction between public goods and merit goods remains useful. Rivalry and excludability are not just ideological labels. They describe real features of how goods behave, and those features have consequences that no amount of political will can wish away.
If a good is non-rival, then charging for it is wasteful: excluding someone costs society a benefit and saves it nothing. If a good is non-excludable, then nobody can make a living producing it, because nobody can be made to pay, and so it will be undersupplied unless it is financed collectively. These are engineering facts about the good, not verdicts about whether it is worthy.
The merit-good question (should everyone have this?) and the public-good question (can markets efficiently supply it?) are different questions, and a society trying to decide how to organize care provision needs to ask both. This is another way of saying that issues of both equity and efficiency come into play.
Stocks and Flows
A second issue buried in ordinary usage is handled badly even by the technical vocabulary.
We talk about goods as if they were single things, but many of the things we care about most are really two things: a stock that exists at a point in time and a flow of services that the stock yields over time. A forest is a stock. The carbon it absorbs each year, the water it filters, the flood it prevents, are flows. A fishery is a stock; the annual catch is a flow. A power grid is a stock; electricity is a flow.
The distinction matters because a stock and its flow can sit in different cells of the two-by-two table. A forest, considered as a stock of timber and land, is rival and excludable: cut it down and it is gone, purchase the land it sits on and it is yours. That is why forests, grazing lands, and fisheries appear in the textbooks as “common-pool resources,” a category the political scientist Elinor Ostrom spent her career studying, rather than as public goods. Their defining problem is that use subtracts from the stock. But the carbon uptake of a standing forest is a textbook public good. My share of a stable climate does not reduce yours, and there is no way to charge Bangladesh for the benefit of trees standing in Brazil.
So when someone says “forests are a public good,” they are half right and half wrong, and which half depends on whether they mean the stock or the flow. The stock is a private or common-pool asset whose owner bears the cost of maintaining it. The flow is a public good whose beneficiaries cannot be billed. This is exactly the combination that produces underprovision: the costs of maintaining the stock are concentrated and appropriable, the benefits of the flow are diffuse and not.
The owner of the forest can sell the timber but cannot sell the climate regulation, so the marketable commodity takes priority. Ecological economists, following Nicholas Georgescu-Roegen and Herman Daly, have long distinguished “funds” that yield services from “stocks” that are used up, and analysts of ecosystem services now routinely separate ecosystem structure from the services it generates. But neither the public-finance literature nor everyday usage has fully caught up with this distinction.
Notice, too, that the flow can change character as the stock is drawn down. Three hundred years ago the atmosphere’s capacity to absorb carbon was, for all practical purposes, non-rival. Now one country’s emissions leave less room for another’s. A public good has become a congestible one, and the stock behind it has become scarce. Stocks and flows are not fixed categories any more than public and private are. But the relationship between them is where the action is.
Care as a Stock-and-Flow Problem
All of this applies with particular force to the work of caring for people, and it helps explain why arguments that care provision is a public good often fail to persuade economists.
Consider the raising of children. Parents, and disproportionately mothers, supply the time, attention, money, and forgone earnings that go into producing a capable, healthy, trustworthy adult. Those inputs are rival and excludable in the most literal way: an hour spent with one child is usually not spent with another, and it is not spent in employment. The stock being built up, the child’s developing capabilities, is likewise a particular person whose upbringing is a particular family’s cost. In the economist’s vocabulary, none of this looks like a public good.
But the flow that the stock generates is another matter. A generation of adults who can read, reason, cooperate, obey the law, pay taxes, care for the old, and raise the next generation in turn yields benefits that are non-rival and non-excludable in the classic sense. My enjoyment of living in a society of competent, trustworthy people does not diminish yours, and there is no way to exclude the childless from it. Pension systems are financed by other people’s children. Public health depends on other people’s children being vaccinated and educated. The very possibility of markets depends on a supply of people socialized to keep promises. These are flows, and they are public goods in exactly the sense Samuelson would recognize.
The same structure holds for care of the sick, the disabled, and the elderly. The labor is rival and largely unpaid or underpaid; the flow of benefits, a society in which people can count on being cared for when they are vulnerable, is something everyone consumes whether they contribute or not.
Putting the stock-flow distinction at the center clarifies what is wrong with the two familiar ways of talking about care provision. Those who say care provision is a public good and leave it there invite the reply that it obviously isn’t: children are rival, care hours are rival, and families choose to have them. Those who say care is a private choice whose costs belong to the chooser ignore the flow entirely and treat the free ride on other people’s caregiving as if it were nobody’s business. Both are looking at only half of the good.
Care is a private cost that produces a public flow. That is neither a merit-good claim about what parents deserve nor a loose claim that care “benefits everyone.” It is a technical claim, in the economist’s own terms, about why markets and families left to themselves will supply less of it than a society wants, and why the people who supply it bear costs that the beneficiaries never repay. The forest owner who cannot sell the climate regulation is in the same position as the mother who cannot bill the pension system for the taxpayer she raised. In both cases the remedy is the same: the public, which enjoys the flow, should help pay for the stock.
Musgrave’s merit goods were an attempt to say that a society may legitimately decide certain things ought to be provided. Samuelson’s public goods were an attempt to say why certain things won’t be provided unless it does. We lost something when the second definition swallowed the first. But we also lose something if we let the first swallow the second. The point of distinguishing stocks from flows is to hold on to both: to see that a good can be worth having and structurally undersupplied, and that the people who build the stock deserve a share of what the flow returns.