Echoes of Henry George in Modern Analysis: a comment on three applications

American Journal of Economics and Sociology, The, Nov, 2004 by Randall G. Holcombe

The three preceding papers all use the ideas of Henry George to analyze important policy issues. Each paper is interesting in its own right and as a group they show that, even a century after he wrote, Henry George's ideas are taken seriously as a foundation for policy analysis. This comment considers each paper individually, and concludes with some observations regarding the contemporary application of George's ideas.

Optimal City Size

THE EASIEST OF THE PAPERS TO COMMENT ON is Richard Arnott's paper on optimal city size. Professor Arnott asks whether the Henry George Theorem offers a practical guide to determining optimal city size and concludes that it does not. I agree completely. Despite Arnott's negative conclusion, his paper offers the reader much. It has a good review of both the Theorem and the related literature, with a special focus on Kanemoto, Ohkawara, and Suzuki (1996), who try to use the Henry George Theorem to estimate optimal city sizes in Japan.

Professor Arnott accepts the Henry George Theorem in principle, and after discussing it concludes that "the generalized Henry George Theorem holds very generally." Fitting existing data to the theory is what gives rise to the difficulties. A primary problem, which Arnott notes, is that agglomeration economies having nothing to do with local public goods may be the most substantial determinant of land rents. Arnott does not consider public choice problems, which may stand in the way of applying the Theorem at least as much as agglomeration economies. Cities are assumed in the Theorem to produce public goods with the tax revenues they collect, but in the real world public goods are not homogeneous--and what local governments produce are often private goods paid for at public expense. There are also likely to be inefficiencies in production because of bureaucratic suppliers, interest group politics, and rent seeking that further complicate the relationship between property taxes and public goods. This observation simply reinforces Professor Arnott's conclusion. Even if the problems he cites in his article were solved, there are more problems that hinder the use of the Henry George Theorem as a practical guide to optimal city size.

The Winner's Curse

PROFESSOR TIDEMAN OFFERS THE INTRIGUING HYPOTHESIS that land speculation tends to keep land undeveloped for too long because owners of undeveloped land suffer Dora a winner's curse. Economic theory, Tideman argues, is based on the assumption of perfect foresight, but when foresight is not perfect, the highest bidder for a piece of property will be the person with the most overly optimistic forecast of the property's value. Landowners systematically overestimate the appreciation potential of their land and, as a result, leave it undeveloped for speculative reasons rather than developing it.

This is an interesting idea, to be sure, but there are a number of reasons why it must be considered a conjecture rather than a solidly supported conclusion. First, the winner's curse was originally formulated to apply to auctions, but land is not typically sold at auction. It is sold on the market like other resources, and unless the real estate market is especially thin, the sale of other properties helps provide information to potential buyers about the value of a particular piece of property,. One of the things markets do, as Hayek (1945) noted, is to generate information about the value of goods and services. The buyer of any good always is the person who places the highest value on the item purchased, but would we then conclude that anybody who buys something has always overpaid because the buyer outbid others to make the purchase? Are people who buy cars the victims of a winner's curse? Would we conclude that someone eating a meal in a restaurant has overpaid because that person demonstrated the highest willingness to pay? To make a convincing argument, Professor Tideman first needs to show that land markets resemble auctions more than markets for other goods and services. Then there is the question of whether high bidders really do overpay. Even if the theory of the winner's curse is true for auctions (and that is open to question), the theory may not apply to land purchases.

A second problem with the idea is that people can develop and use land even as they are holding it for speculative purposes. There are several reasons why development might even enhance the appreciation potential of land. It could reveal the property's potential to generate revenue, for example, if development demonstrated the rents that tenants would be willing to pay. Perhaps more significantly, environmental and growth management regulation may present obstacles to development, and the resulting uncertainty about whether--or how intensively--a parcel could be developed may lower the value of a piece of property. If the property is already developed, those obstacles would already be cleared and the property would be worth more. Even if the winner's curse holds for land, more needs to be done to show that this also implies inefficient patterns of development.