A paper idea in Stigler (1964) on Oligopoly

Next week, I am teaching collusive agreements in my price theory class. I decided to take a different approach to the discussion than the one usually found in textbook. The approach consists in showing how economic thought on a topic has evolved over time. For collusion, I decided to discuss George Stigler’s 1964 article on the theory of oligopoly published in the Journal of Political Economy.

Simply put, Stigler proposes a simple approach for stating how collusive agreements can break apart by asking how much extra sales a firm can obtain by cutting its prices without being detected by other firms. Stigler argued that detection got easier as the number of buyers increased or as concentration increased. He also argued that detection became harder if buyers do not repeat purchases and if there is growth in the market through the addition of new customers as firms are not able to detect whether the growth of other firms is due to new customers or because old customers are purchasing its wares. Detection also became harder with a greater number of sellers but he also argued that this was of equal (or maybe lesser) importance than low repeat-sales rates or the arrival of new customers into the market.

This is pretty standard price theory and it is well executed. After postulating the theory, Stigler throws the empirical kitchen sink to see if, broadly speaking, his point is confirmed. One interesting regression is from table 5 in the article (which is illustrated below). That regression estimated rates for a line of advertising in newspapers market (i.e., cities) conditional on circulation in 1939 (its a cross-section of 53 markets). The regression itself is uninteresting to Stigler as he wants to consider the residuals. Why? Because he could classify the residuals by the structure of the market (with only one newspapers or with two newspapers. The idea is that more newspapers should be marked with lower rates as collusive agreements tend to be harder to enforce. Stigler thought this confirmed his idea that “that the number of buyers, the proportion of new buyers, and the relative sizes of firms are as important as the number of rivals” (p. 56).

While looking at Stigler’s regression, I thought that there might an interesting economic history paper to write. Notice that the source of the data used is cited below the table. Retracing that source and checking if (because there are clearly volumes of the Market and Newspaper Statistics) a panel can be constructed could allow for something interesting to be done. Indeed, a panel allows to directly test for the new customers’ hypothesis by adding a population growth variable. This advantage compounds that of increasing the number of observations. Both of those advantages could allow to test the relative importance of the mechanisms highlighted by Stigler.

A paper of this kind, I believe, would be immensely interesting. It is always worth engaging with important theoretical articles on their own terms. As Stigler set this test as one of his illustration, a paper that extends his test would engage Stigler on his own term and could provide a usefully contained discussion of the evolution of the theory of oligopoly. I honestly could see this published in journals like History of Political Economy or Journal of the History of Economic Thought or journals of economic history such as Cliometrica, European Review of Economic History or Explorations in Economic History.

One thought on “A paper idea in Stigler (1964) on Oligopoly

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s