Talking about progressivism and its roots in urban governance, I’m going to go ahead and say that air conditioning is not a public good. Not that the systems and proposals in this article are wrongheaded – far from it! What I’m criticizing is the stultified limits on our political language.

Air conditioning is not a public good. A public good is non-excludable and non-rivalrous – that is to say, one cannot be prevented from consuming it and its consumption by one person does not affect the supply available to be consumed by others. Now, these things can be effective rather than absolute; defense, in many ways a quintessential public good, has some excludable and rivalrous properties, but in general the entirety of the United States benefitted roughly equally from the nuclear deterrence of the Cold War.

As we all learned in Econ 101, public goods are those things even economists agree the government should provide! The problem with that thinking is three-fold – one, public goods are extremely rare, and, in most cases, it is possible to argue that even the seemingly-clearest cases of public goods are not, in fact, public goods; secondly, as the late, great Elinor Ostrom pointed out, it is entirely possible for voluntary non-governmental institutions to emerge for governance of public goods (and their cousins, common-pool resources); lastly, the restriction of government provision to public goods, even as a heuristic or guidepost, is totally wrong to begin with.

Consider, for example, roads. Now, it is totally possible to imagine specific situations in which one would want totally private roads – not roads built or operated by private firms like the new HOT lanes on the Capital Beltway, but roads that are a solely private enterprise from start to finish. For example, if a company is drilling for oil and builds a refinery ten miles away and wants to pave a road between them, there’s no obvious reason for taxes to pay for that road. But in general most American roads are public roads, at least on a fundamental level if not publically operated and maintained top to bottom (though many, perhaps the majority, are), and that makes intuitive and economic sense, even though roads are both excludable and rivalrous.

The reason is network effects and positive externalities; that is, actions, institutions, and infrastructure that provide benefits to society above-and-beyond those that flow to the direct users, as well as those systems whose benefits to everyone increase as more and more people join the system. In the case of roads, the quicker transportation of goods and services as well as the new possibilities for travel create benefits to society well beyond the benefits experienced by those who actually drive on the roads.

In the case of district cooling, we’re talking about a massive and complex investment whose benefits are over a very long time horizon and cannot necessarily be captured in sufficient amounts by the cost-bearers to justify private investment. Certainly we may see individual small institutions make a self-interested investment in district cooling; but large-scale mass investment would bring benefits to the energy market and to the environment that simply won’t happen otherwise.

And the thing is – the modern world, thanks in large part to urbanization, is riddled with positive externalities and network effects. They are omnipresent. And when that is the case, and when transaction costs are high (and they are high), a lot of assumptions about markets and utility maximization have to go out the window. Which is not to advocate for command-and-control economies or socialism; just to say that reliance on simple models that were first developed to understand largely-agricultural economies is a poor way to go about assessing public policy in the 21st century.