Streetwise Professor

Holman Jenkins’s column in today’s WSJ gives well-deserved props to my thesis advisor, Lester Telser. Jenkins rightly recognizes that Lester’s work on core theory is the key to understanding the airline industry’s chronic financial difficulties. The theory predicts that when demand is divisible, but costs are not, (a) it is often optimal to operate with excess capacity, but (b) when there is excess capacity, competitive pressures push prices to marginal costs that are not sufficient to cover even the avoidable costs (e.g., fuel) of operation, let alone make a contribution to capital costs. This fits the airline industry and its checkered history to a “T.”

Jenkins also makes the analogy between airlines and ocean shipping, the subject of my thesis; I showed that core theory can explain the pattern of adoption and non-adoption of collusive arrangements in the shipping market. He doesn’t mention my work, but he does cite the roughly contemporaneous research of Bil Sjostrom (a Chicago guy, though now in Ireland), which is a good thing. (In an email, Holman said he had tried to find an online version of my thesis. LOL! Good luck with that, since it was written BDE–Before the Digital Era.)

Core theory is a potentially valuable tool that provides numerous insights on developments in a variety of industries. It is also a tool that mainstream IO economists shun like the plague. Noncooperative game theory rules, and although I understand that it is in theory desirable to model everything as a non-cooperative game, this comes at too high a cost. Core theory makes shortcuts regarding how coalitions form and the process of negotiating deals within coalitions, and noncooperative game theorists snort derisively at this. But ALL theories make shortcuts. The issue is whether the shortcuts allow you to come to insights that are not possible if you don’t make them. Full noncooperative treatment of multi-agent bargaining situations are extremely difficult–so guess what, noncooperative game theorists make shortcuts! Sad thing is, many of these lead to observationally vacuous theories. In contrast, as the body of Lester’s work (with assists from Sjostrom, Bill Sharkey, George Bittlingmayer, and me) shows, core theory is anything but observationally vacuous. It actually helps explain the real world. What a concept!

I gave up (for the most part–a few things are still in inventory) working on core related stuff because it was too hard for a junior person to swim against the professional tide. The immediate reaction of most mainstream IO economists to a mention of core theory was an “are you nuts” look, and an effort to change the subject. What a shame, because, as the airline and shipping and other examples show, core theory often has far more interesting things to say than the Nth tweaking of noncooperative oligopoly models. In particular, it provides a ready method for exploring and exploiting the implications of cross sectional and time series variations in cost and demand structures for industrial structure. People in business know that cost structures matter a lot. Unfortunately, traditional noncooperative game theory-based IO almost completely ignores this issue.

Perhaps Jenkins’s article is a harbinger of a renewed interest in how costs and demand affect market structure. Here’s hoping. And if it is, core theory is the most promising way to examine this issue.

And, I am deeply appreciative to Holman for giving Lester Telser’s research some well-warranted visibility. Lester is one of the best economists of the last 50 years, but his iconoclasm and originality have, sadly, meant that he hasn’t received the recognition he deserves.