Friday, April 23, 2004

Getting the "Law of One Price" Wrong

Matthew Yglesias has been putting up posts on the Law of One Price, in which both he and the great majority of his commenters make very clear that they don't get what this economic principle is really about.

In a way, I can't really blame them for their error, as the mistake is a natural assumption of assuming that the Law of One Price really does make the claim that commodities everywhere will converge to a single price, thanks to free trade. In fact, the Law of One Price says something considerably weaker and far less controversial: that under a free trade regime, commodity prices will converge so that there are no profit opportunities for would be arbitrageurs. One implication of this weaker statement is that price differences will not vanish in as far as they reflect transportation costs, quality differences and so on. In particular one shouldn't expect even prices for materials like crude oil and gold to converge to a single global value, only that whatever price differences remain at equilibrium will not be great enough to present profit opportunities for those willing to, say, buy low quality Saudi oil instead of Nigeria's "Bonny Light" premium variety.

As for the original claim by Yglesias that his boss Robert Kuttner was wrong to say the following:

One of the most fundamental laws of economics is the law of one price. If Exxon is selling gas for $1.60 a gallon and Gulf tries to sell it for $2.60, everyone will go to Exxon.

I have to say that for once, Kuttner is on the side of the righteous, which is a rarity indeed on matters economic. Yglesias' example of two gas stations, one charging $2.00 and offering no queues, while the other offers $1.50 and long waiting times, only makes sense if the two stations' proprietors are so blind that both fail to see that they're leaving cash on the table by not altering their prices; and in equilibrium, the price that must entail at both establishments will be identical, and will lie somewhere between $1.50 and $2.00 - assuming that differences in service quality and costs of transportation to both establishment are insignificant, assumptions that are likely to be violated in practice, ergo the existence in real life of small variations in price between stations. The "time value of money" argument Matthew tries to make simply won't work here.

One more thing I find worthy of comment is the arrogance displayed by some (though, in fairness, by no means all) of the commenters in their dismissals of the Law of One Price as just so much worthless economic dogma being pushed by Economic quacks beholden to a free-market orthodoxy, an argument that is all the more ridiculous in that this is one of the single notions most copiously supported by empirical data. Not all criticisms of standard economic theory are meritless, but is it so unreasonable to insist that one actually has to understand a theory to point out its supposed shortcomings? It is precisely in their eagerness to skip the years of hard slog and get straight to the bitching and moaning that most opponents of standard economic wisdom come undone.