A public good isn’t one that’s made by the state’s enterprises. There are good and bad reasons for the state to hold a large fraction of the productive capital in an economy but that’s a separate concern. The state may also turn out to be the best provider for some public goods but that’s a consequence, not their definition. A public good is also not the same as a good whose provision is a moral necessity. Sometimes the two are aligned, but they’re not the same. Providing food is a moral necessity, but usually it’s handled as a private good. Public transport isn’t a pressing moral necessity, but arguably it’s best handled as a public good. I propose a marginal definition of public goods:
A good is marginally public at a certain point in its demand curve if lowering the price or other barriers to consumption brings a marginal benefit to society. The good is marginally private otherwise.
In everyday terms, a good is a public good if society wants it cheaper so that more of it is consumed. The additional utility of consumption, minus the production costs and any externalities such as resource use, pollution, etc. yield net positive utility in the eyes of society. Conversely a good is private when society wants its price, taxation, etc. to rise, either to act as barriers to consumption or to compensate the public for cost externalities. Private goods are produced because they provide utility to someone but society sees marginal consumption as a net loss of utility, perhaps because of subjective utility, waste, or cost externalities. The same good can be marginally public through part of its demand curve and marginally private through another part. Typically it’s marginally public at the cheap side of the curve when wider access is desirable, and it’s marginally private at the expensive end when the good becomes specialized or elitist. When we say that something is a public or a private good without the marginal qualification we’re making a generalization as to how it is for a relevant part of its demand curve. The best example is health. Health is marginally public for the entire middle of its demand curve. When people are ill, they want to be treated to get better, pretty much universally, and they don’t want to be treated too much or for no reason. Health becomes marginally private at low acuity level, where people may get over-medicated, and at high acuity level where resources spent on patients with desperate prognosis could do more good elsewhere. Having said that, preventive medicine and insurance can make each end of the curve marginally public. Other examples:
- Fossil fuel energy is marginally private for most price points because using more than necessary depletes resources or damages the environment in a way detrimental to society. However at very high prices it becomes marginally public: It needs to be subsidized or rationed enough to enable essential services.
- Education is marginally public up to school level because it’s worth living in an educated society whether or not you are a kid or have kids. It continues to be marginally public at college level because equal opportunity access to life skills is good for society, and because a skilled society is a better society. Education may become marginally private at the research level, where too much abstract research could arguably be a waste.
- Food is marginally private for most price points because there’s a lot of individual preference and because overusing food is a waste. However at times of famine, for destitute people, or otherwise at the low end of the demand curve, food becomes marginally public.
- Music, software, Google, and other information goods are marginally public from the bottom of the demand curve, where they’re free and aimed at the mass market, up to the point where they become specialist services able to command a high price. Information goods tend to be misperceived as private goods, and therefore mispriced too high causing underuse or price evasion. Lowering the price or monetizing indirectly would increase net benefit to society. Google has it right and the publishers usually have it wrong.
- City transportation is marginally public at the lower part of the demand curve. Buses, subways, trams etc. should be free because having easy transportation is valuable to all (even rich people, since it allows poor people to get around) and because the transaction cost of having fares, fumbling for change in your pocket, etc. is a waste. Transport becomes marginally private at the taxi or car level, so cities should collect fees from private road users.
- Communication bandwidth is marginally public for low usage levels and becomes marginally private as you want to do more high-end things, host sites, have mobile video chat, etc. Phone companies are fleecing the public by treating it as a private good throughout, but within their offering they try to approximate the pubilc/private inflection by charging a flat fee that covers low usage levels. For historical reasons, land bandwidth has more public economics and mobile bandwidth positioned itself as more private.
When considering what’s marginally public you have to take account of the marginal change of utility to society with respect to pricing (or any other barrier to consumption), as well as the marginal change in the ability to finance the good. Reducing the price, dropping controls, or making it free increases use, and if that increases utility to society it’s probably a public good. If reducing the price makes the good too hard to finance, so that too little is produced and society suffers, it’s probably a private good. But you need to consider both direct and indirect ways to finance the good. For example health, transport, or music take real cost to produce. One way to finance them is to charge consumers. Other ways include taxation, charitable sponsorship, and indirect monetization such as advertising. The good is marginally public if there’s some way to finance it that makes it so, not necessarily the current way it’s financed. The textbook treatment of public goods tends to focus on the supply side and ignore the demand side. But the demand side effect of pricing is important. If you make access difficult, adversarial, and expensive you’re pretty much guaranteed to reduce demand. You may think that’s inevitable if you believe a-priori that your product is a private good. But if in fact it’s a public good, you’re doing a disservice to society and should think of ways to monetize it as a public good.