Local Public Goods & Dynamic Geography
Public goods are in part defined by their non-excludability. Once provided, they are provided to all. To take one canonical example, if the government provides a national defense, everyone under its shield benefits, even those who do not pay for it. That’s the nature of the thing.
Another element is that it is notoriously difficult to determine how badly the public wants these goods. With national defense, hawks want more, doves want less, but that’s not even the half of it. We also have the free-riding knave, the narrowly self-interested man who expects others to pay for a public good like national defense. Since he can’t be excluded from the benefits–that is the nature of this pure public good, after all–why not let others foot the bill? So we have a complex problem. Without any sense of loyalty or trust, all have an incentive to act like knaves, since they stand to benefit even if they do not share the burden. There are hawks, there are doves, and we can’t even identify them. Instead, everyone acts like a knave and the public good never materializes. It is, as they say, a market failure. Of course we’d like to exclude the knaves, and provide what the public truly wants–neither too much, nor too little–but given the nature of the good, and our inability to discover the public’s true preferences, we can’t. So the government coerces everyone and eye-balls it.
But Charles Tiebout, in a classic 1956 paper, noted that some public goods allow for a modicum of excludability. There is a subset of public goods such that only those who are nearby receive the benefit. Yes, within that neighborhood, the good is non-excludable. But since the positive externalities are local, we can exclude by geography. National defense is one thing, municipal golf courses are another. (The culture of the 1950s hangs like a grey-flannel specter over Tiebout’s discussion.)
From this perspective, Tiebout discovered a solution to the problem of determining the demand for local public goods: we can think of local town governments as competitive suppliers. People are free to live in the town that suits them. Since each town offers its own package of public goods and matching tax burdens, citizens shop among them by deciding which town to live in. Each public good consumer can ask, Is this package worth it? Politicians can create ad campaigns. Those towns that provide what people want, to the extent that they want it, will thrive, while those that don’t will wither. And so though we are dealing with public goods, we have at last the approximation of a competitive market.
And what about the incentive to act like free-riding knaves? Mobility eliminates it. The towns that fail to produce the local public goods people truly want will empty. Those that do will fill.
Now let’s indulge in a little fantasy. Tiebout’s assumption was that people can move anywhere within a nation. But that assumption is stated too precisely. (For this blog!) What really matters is mobility, period. If people can move the land they own and live on, then the upshot of Tiebout’s competitive market for public goods still holds. (Instead of voting with your feet, you vote with your dirt.) In a world of dynamic geography, a world of viable seasteads, we should expect a competitive market in public goods. Right? Individual seasteads will detach and attach and reattach to other seasteads according to how well that association fulfills their needs. Various associations of seasteads (perhaps with democratic governments) will act as efficient suppliers of local public goods (despite their democratic governments). Or so Tiebout might say.
But Bryan Caplan offers a few objections to Tiebout’s model. Competition between non-profits is not the same as that between for-profit ventures, he says. First, the incentives are different: a politician doesn’t get paid more when the local economy flourishes. And second, politicians may have an easier time when things go poorly. For all you seasteaders out there, what is to be done? What would a seasteading Tiebout say to Caplan?
Comments are closed.
Let A Thousand Nations Bloom 
Jeffrey Friedman has a paper on non-profits titled There Is No Substitute For Profit And Loss. I discuss some related issues here.
Public goods blah blah… correcting non-excludability blah blah.
WAR.
Any consumer surplus gained from summing all those little static triangles that we learned about in Public Choice theory, get blown to bits the moment the bombs start falling on other people’s children.
TRILLIONS of man-hours wasted on the fires set as a result of the pissing contests between ‘representative governments’. TWO HUNDRED MILLION dead – almost all of them civilian.
And THIS is the mechanism you employ to correct market failure? PLEASE.
If the market under-expands, you don’t create an X-inefficient bureaucracy (which tends to grow of its own accord) to deal with it.
And also – ‘hawks’ and ‘doves’ are usually identifiable to the extent that the further one is from actual combat, the more hawks you find. (Until you get all the way to homo cheneyensis, who never saw a war he didn’t like, although he declined to participate in any of them).
Rothbard may have been a bit of a nut, but he could see that all the talk of correcting market failure by tax-transfers (using progressive taxes due to the diminishing marginal utility of money)… all lovely in theory.
THEN you start a bureaucracy and it all goes to hell.
Then government wants to get bigger and bigger and bigger – and it has all the armed thugs it needs to promote its own growth.
Before you know it, you’re suffering from the MASSIVE, NEGATIVE externality you get from having a gigantic metastatic tumour on your system.
One important point I’ve not seen discussed by seasteaders is the distinction between Common-Interest Developments and Multi-Tenant Income Properties. This distinction is used by Spencer MacCallum, and I believe Patri is influenced by him to some degree. I know that the purpose of seasteading is to allow experimentation in government, so forcing the original seasteads to be MTIPs would not be in line with the goals. Still, it seems this should be a big topic for discussion.
“Competition between non-profits is not the same as that between for-profit ventures…a politician doesn’t get paid more when the local economy flourishes.”
Does a seastead need to be a nonprofit venture? If so, why?
I discuss basically the same problem here.
I think there are some advantages to small-scale governance not captured by the Tiebout model: there’s more experimentation and imitation of successful institutions. It depends, though, on the degree to which voters and policy-makers have the incentive to choose the best institutions, which is limited. Politicians compete for votes. Voters vote expressively rather than instrumentally, but there is significant overlap between expressive and instrumental preferences.
The biggest virtue of seasteading is that it’s not bound to any single model of organization. If Tiebout doesn’t work with non-profit governments (small-scale democracies or whatever), for-profit governments will flourish. I know if I ever go to live on a seastead, I’ll want it to be one where the rule-makers are competing for my money rather than my vote. Voluntarily entering a democratic government puts you in the prisoner’s dilemma of rational ignorance/irrationality.
Great post, nicely summarizes the problem! I should note a another problem with Tiebout competition, also from Caplan in this paper:
http://www.gmu.edu/departments/economics/bcaplan/tieburb.doc
Essentially, he argues that taxes and bad policies generally are capitalized into land values, mitigating the beneficial effects of inter-governmental competition.
The “upshot,” as Caplan might say, is that seasteading may be a way of getting around this problem: if land is just as mobile as other forms of capital, competition between governments will be strengthened. Or in the more pessimistic case, seasteading should at least create *more* mobility between governments, even if the mobility is not perfect, thus providing additional constraints on government predation.