I’ve been reading Brett Frischman’s “An Economic Theory of Infrastructure and Commons Management”, which develops a general theory of infrastructure management and then applies this theory to (among other things) the network neutrality debate. I’ll likely have more to say about it once I’ve finished digesting it, but I wanted to note one offhand comment that I found interesting. On page 924, Frischmann writs:
A list of familiar examples [of "infrastructure"] includes: (transportation systems such as highway systems, railways, airline systems, and ports, (2) communications systems, such as telephone networks and postal services; (3) governance systems such as court systems; and (4) basic public services and facilities, such as schools, sewers, and water systems.
Which of these things isn’t like the others? Let me first quote his definition of “infrastructure resources”:
(1) The resource may be consumed nonrivalrously;
(2) Social demand for the resource is driven primarily by downstream productive activity that requires the resource as an input; and
(3) The resource may be used as an input into a wide range of goods and services, including private goods, public goods, and nonmarket goods.
As far as I can see, schools don’t fit any of these criteria. While there are certainly some under-attended schools who could be called non-rivalrous in some sense, there is typically a trade-off between the number of students in a classroom and the quality of instruction. And I suppose you could say that in general, education is a pre-condition for higher future earnings, but if we interpret this definition that broadly, almost everything is “infrastructure.” People couldn’t be productive citizens without food, so does that make farmers, supermarkets, and restaurants “infrastructure?”
I don’t mean to pick on Prof. Frischmann, because I think there are actually lots of people who would agree with him about this. But I do think that casually lumping schools in with roads in courthouses is a reflection of politics driving public policy conclusions rather than the other way around. A lot of people have strong attachment to the idea that the state should operate schools. And with that assumption as a given, they seem to reason backwards to the conclusion that schools must be basic infrastructure like roads, airports, and courts.
I think this assumption is harmful for a variety of reasons. The obvious libertarian reason is that I think we’d be better off if the government didn’t run schools, and so I’m resistant to classifying them in ways that makes the argument for education monopoly sound more plausible. But the more subtle reason, I think, is that it leads people to misconceive education in ways that are harmful to the quality of the education. Infrastructure is impersonal, almost by definition. The highway engineers at the Department of Transportation only worry about getting people from Point A to Point B without accidents or congestion. This is a straightforward engineering challenge, and they don’t spend too much time worrying about any given car, or trying to think of ways to make driving more interesting, emotionally satisfying, or inspiring. Infrastructure owners are almost always focused on aspects of the service they can provide that can be easily captured in statistics: number of miles traveled, number of tons shipped, number of cases decided.
This is a disastrous way to run a school. Every child is different, and good education absolutely requires teachers that take an individual interest in their children. Children are not self-motivated free agents who can pursue their own goals the way a driver finds his own route on the highway. They need personal relationships and individual guidance as much as they need to be force-fed a series of multiplication tables, grammatical rules, and historical names and dates.
Thinking about schools as infrastructure encourages us to focus on incidental aspects of education—have the walls been painted recently, do they have the latest textbooks and computers, do the school busses run on time—and ignore the far more important question of whether schools are recruiting talented teachers and supporting them properly. This mindset also leads to grotesque educational reforms like No Child Left Behind that pretends to improve education by reducing all educational achievement to a series of (easily fudged) test results and then declaring victory when those numbers go up. Collecting statistics and making 5-year plans are the way you run a highway system, not a school.
Schools are not infrastructure. They’re not really even public goods in more than the trivial sense that almost every productive activity in society generates benefits for third parties. But they are immensely important, and I think our schools would be a lot better if we stopped thinking of them as impersonal infrastructure that’s properly management by massive, impersonal bureaucracies.
Update: Frischmann doesn’t really mention schools much elsewhere in the paper, but he does discuss them further in footnote 170:
education could be provided exclusively by the private sector. However, this would leave many children without access to education and cause a subsequent host of social problems when these children do not have the necessary skills to become productive members of society. See id. Education is a good whose social merit has been recognized, and therefore both the public and private sectors often provide it to insure more widespread consumption.
Note, though, that this is merely an argument for government subsidies to education. It tells us nothing about how schooling should be organized and operated. For example, there’s no reason to think that a society that had no public schools but provided every parent with a voucher sufficient to send their child to a good private school would be any worse educated than our present society; to the contrary, I would expect it to be better educated. This is the approach we take with food (food stamps) and housing (section 8 vouchers). Food, housing, and education are all extremely important, and we should be anxious to ensure that they’re widely available. But they’re not infrastructure, and there are no good reasons for the state to provide them directly.