The Case of Landsburg
[Part of a week-long series on neoclassical economics]
Steven Landsburg’s The Armchair Economist (The Free Press, 1993) is an absolutely delightful book. It is filled with compelling arguments that will challenge any reader's opinions.
Unfortunately, in several instances Landsburg falls into the familiar neoclassical trap of relying on unrealistic assumptions to obtain his 'surprising' results. If the assumptions were more in accordance with reality, it is not merely that the analysis would be rendered intractable, but that the conclusion itself would be reversed. This essay will analyze two such instances of this phenomenon.
A Tale of Two Cities
In his chapter, "The Indifference Principle: Who Cares if the Air Is Clean?" Landsburg explains that "all economic gains accrue to the owner of fixed resources" (33, emphasis removed). He applies this Principle to Bush’s Clean Air Act:
[The Act] was expected to cost…approximately $25 billion per year. If that estimate is correct, the cost to the average American family of four is about $400 per year in the form of lower profits, lower wages, and higher prices for consumer goods. On the other hand, clean air is a great benefit, which uncritical observers expected would be shared by everyone who breathes, which is to say everyone. But the ability to breathe is not a fixed resource. Universal skills do not ordinarily reap great rewards. If breathers do not benefit from clean air, then who does? Theory tells us to look for the owners of fixed resources. The most obvious candidates are urban landowners, who are able to charge higher rents after the smog lifts. (34)
The "uncritical observers" will no doubt object to this conclusion. Landsburg anticipates this and thus offers a "fable" (so named to defuse the knee-jerk rejection of "models") to solidify his argument. In fairness to Landsburg, we quote extensively:
Somewhere in the heart of the Rust Belt are two small cities: Cleanstown and Grimyville. All of the activities of daily life—shopping, working, going to the park—are equally pleasant in both cities, with one exception: breathing. The Grimyville Steel company accounts for that. No Grimyvillian ever wakes up and fills his lungs with the crisp morning air that Cleanstowners take for granted….Life expectancy is ten years lower in Grimyville than in Cleanstown. Why would anyone live in Grimyville? For one reason: it’s cheaper. A house that rents for $10,000 a year in Cleanstown can be had for $5,000 in Grimyville. That $5,000 difference is just enough to keep folks in Grimyville. If it weren’t, people would leave Grimyville and rents would fall even further. Young people deciding where to settle are indifferent between the two towns. They like the atmosphere in Cleanstown, but they like the housing prices in Grimyville.
Last week, the Grimyville Council passed a Clean Air Act that requires Grimyville Steel to adopt extensive antipollution measures. Soon the air in Grimyville will be as pure as the purest air in Cleanstown. And when that happens, the rents in Grimyville will rise to Cleanstown levels.
Eventually renters in Grimyville will be living in a clone of Cleanstown. Is this an improvement for them? Evidently not, because if they’d wanted to live in Cleanstown, they could have moved there long ago. Those young people deciding where to settle gain nothing from the Clean Air Act. Earlier they had a choice between Cleanstown and Grimyville, and they were indifferent. Now they have a choice between two Cleanstowns. They’re no worse off than they were before, but no better off either.
The only people who stand to gain from this entire affair are the property owners of Grimyville, who can now command higher rents than they did before. The Clean Air Act is equivalent to a tax on Grimyville Steel with the proceeds distributed entirely to Grimyville landowners. (34-35)
Landsburg is quick to concede that his fable rests on several unrealistic assumptions, the most obvious being that people's tastes are identical. (E.g. if some people have an unusual tolerance for smog, they will actually be hurt by the Clean Air Act.) But Landsburg does not discuss his implicit assumptions on the number of people living in the two cities, and the (identical) preferences concerning population density. Unless these assumptions are chosen very carefully (and probably unrealistically), Landsburg's conclusion falls away.
Suppose there are 100 people stranded on a tropical island. The island is separated into two distinct regions: On the northern half-—dubbed Sunnytown--there is a lovely beach, allowing the islanders to become bronzed beauties. On the southern side of the island--named Shadyville--dense foliage prevents sunlight from reaching the inhabitants, who are doomed to an Irish-like existence of mediocre melanin.
Now our 100 inhabitants were stranded when their plane from the Mont Pelerin Society crashed into the sea. All staunch libertarians, they have instituted a system of private property on their tropical island. Consequently, a small group owns Shadyville, and a small group owns Sunnytown. The rest pick fruit as rental payments to their landlords.
As it stands right now, 25 people live in Shadyville, and 75 people live in Sunnytown. The reason everyone doesn't live in the sunny north is that it is too small to support 100 people; its owners would raise rents until people moved back to the south. The rents in the south are just low enough to make people indifferent between living in the south or north.
Some of the more industrious Shadyvillians decide that it would be to their benefit to chop down most of the trees near their shore, creating a replica of Sunnytown. But Steven, one of the free marketeers stranded on the island, finds their analysis to be a bit shady. He points out that, right now, everyone is indifferent between the south and the north. After cutting down the trees, the only difference in the regions will be rental price; people will move from Sunnytown into (the now inaptly named) Shadyville, causing rents to fall in the former and rise in the latter. Since the south is no bigger than the north, it cannot be that everybody moves to Shadyville; once again, an equilibrium will be reached in which renters are indifferent between their place of residency. The efforts to cut down the trees will be completely wasted, benefiting only the owners of Shadyville.
Bob (who went along merely to get the free airline food) objects to Steve's analysis. It's true that the residents of Shadyville will be no better off, relative to the people of Sunnytown. But that’s completely irrelevant, for there are no egalitarians on the island. The only important question is: Are the residents of Shadyville better off than they themselves were before?
To this, the answer must be yes. Before, a Shadyvillian thought his lot was equivalent to living in Sunnytown with 74 other people. But now, in the new equilibrium in which 50 people live in each area, the Shadyvillian can live in sunshine and share his beach space with only 49 other people. This is an option that no one had before, and it is an unambiguous improvement. Depending on our assumptions, the landlords might capture most of this surplus, but this outcome has nothing to do with the Indifference Principle.
Now even this immanent critique does not 'explode' Landsburg’s analysis; when such concerns as the above are raised, the neoclassical usually responds that his model employs a continuum of agents. (For the mathematically disadvantaged: A "continuum" means a heluva lot of people—count to infinity, rinse and repeat.) In any event, the renters (who pick fruit for the landlords) would probably be considered to own the "fixed resource" of labor, and thus the above is not a true counterexample to the Indifference Principle. Nonetheless, it seems a bit much to criticize "uncritical observers" for not employing trains of thought which start off with the assumption of an uncountably infinite population.
The second instance of Landsburg's reliance on false assumptions to reach surprising conclusions is in relation to his discussion of cartels:
Modern technology offers women a variety of innovative but costly ways to attract men. These innovations include everything from new methods of birth control to silicone breast implants. The costs to women include not only out-of-pocket expenses but a variety of health risks.
It can be advantageous for women to withhold such products from the marketplace. In doing so, they act like Ford, General Motors, and Chrysler agreeing to stifle a new automotive technology that would serve their customers well. In ordinary circumstances, each of the Big Three would be left wondering who was going to violate the agreement first. But if they can arrange to have the innovation outlawed, auto executives can sleep more soundly at night.
Likewise, women cannot simply agree among themselves to avoid dangerous methods of birth control or cosmetic surgery. Aside from the logistical problems…cheating would be uncontrollable. The only hope is to ban the products, and feminist organizations have exerted substantial effort in this direction...
...Any woman who wanted silicone breast implants and could be assured of having the only implants in America would be happy. Given the reality—-that if implants are legal her competitors will acquire them too-—she might prefer an absolute ban.
The best argument for keeping new technologies legal is not that they benefit manufacturers but that they benefit their customers. Analogously, the best argument for keeping cosmetic breast implants legal is not that they guarantee freedom for women but that they gratify men. The economically correct argument is the most politically incorrect argument imaginable.
A careful cost-benefit analysis would probably conclude that breast implants should be legal, because the benefits to men exceed the costs to women. (171-172)
Now this analysis is absolutely delicious, and the world is clearly a better place for Landsburg’s invention of it. Unfortunately, it is probably incorrect. The choice of breast implants was perhaps the worst possible for Landsburg’s case, for a man’s preferences in this arena clearly are influenced by Veblen (envy) effects. Landsburg wishes to keep the model simple and so assumes that a man only cares about the absolute size of his mate’s endowment, but this assumption is clearly contrary to fact.
If a device became available which could make any female look like Buffy the Vampire Slayer, but reduced her life expectancy to forty years, "in the long run" men (and women) would most likely be worse off. In the limit, if every woman availed herself of this option, then all women would look identical. It is true that those men who lived through the transition period would perhaps be elated at the innovation, but it is an open question whether the "subjective happiness" (however the neoclassical wishes to define the concept) of the next generation of men would be higher or lower. This ambiguous gain for men is balanced against the clear loss to women.
(It must be emphasized that the relevant question is not: Would the average man prefer to swap his current mate with a short-lived supermodel? Rather, the question is, Would the average man prefer to swap the current female population with another? His present evaluation of Buffy's looks are not reliable, since ‘she won’t look so good after all’ in the new equilibrium.)
Thus it appears that this too is an example of a ‘surprising’ result which falls away after a correction of assumptions. In closing, we note that Landsburg has unwittingly slipped into the real world: Since he assumes men can benefit from breast implants, it must be that their population is finite, and thus that their biological endowments are "fixed resources." For otherwise, no individual woman could be more attractive than any other; augmentation in cleavage would necessarily be offset by the amount of nagging or some other factor.