Up to the early 1920s the classical theory of price included two main models, pure competition and monopoly.

Duopoly models were considered as intellectual exercises rather than real-world situations. The general model of economic behaviour from Marshall to Knight was pure competition.

In the late 1920s economists became in­creasingly dissatisfied with the use of pure competition as an analytical model of business behaviour. It was obvious that pure competition could not explain several empirical facts. The assumption of a homogeneous product, in the theory of competition, did not fit the real world.

Furthermore advertising and other selling activities, practices widely used by businessmen, could not be explained by pure competition. Finally, firms expanded their output with falling costs, without however growing infinitely large, as the pure competition model would predict in the event of continuously decreasing costs.


It was in particular this last fact of falling costs that produced dissatisfaction and caused a widespread reaction against pure competition theories. The dissatisfaction gave rise to a long series of arguments and the publication of numerous articles which form the ‘Great Cost Controversy of the 1920s’.

The earliest summary of the Cost Controversy is to be found in Piero Sraffa’s article. Sraffa pointed out that the falling- cost dilemma of the classical theory could be resolved theoretically in various ways by the introduction of a falling-demand curve for the individual firm; by adopting a general equilibrium approach in which shifts of costs induced by external economies of scale (to the firm and the industry) could be adequately incorporated; or by introducing a U-shaped selling-cost curve into the model.

Of these solutions Sraffa adopted the first, that is, he argued that a model in which the individual-demand curve is negatively sloping is more operational and theoretically more plausible. The same line was adopted independently by E. Chamberlin and by Joan Robinson, in works both published in 1933.

It should be noted that although both writers arrive at the same solution for the firm and market equilibrium, their analytical approach and methodology differ considerably.



The basic assumptions of Chamberlin’s large-group model are the same as those of pure competition with the exception of the homogeneous product.

We may summaries them as follows:

1. There is a large number of sellers and buyers in the ‘group’.


2. The products of the sellers are differentiated, yet they are close substitutes of one another.

3. There is free entry and exit of firms in the group.

4. The goal of the firm is profit maximization, both in the short run and in the long run.

5. The prices of factors and technology are given.

6. The firm is assumed to behave as if it knew its demand and cost curves with certainty.

7. The long run consists of a number of identical short-run periods, which are assumed to be independent of one another in the sense that decisions in one period do not affect future periods and are not affected by past actions. The optimum decision for any one period is the optimum decision for any other period. Thus, by assumption, maximization of short-run profits implies maximization of long-run profits.

8. Finally Chamberlin makes the ‘heroic’ assumption that both demand and cost curves for all ‘products’ are uniform throughout the group. This requires that consumers’ preferences be evenly distributed among the different sellers, and that differences between the products be such as not to give rise to differences in costs. Chamberlin makes these assumptions in order to be able to show the equilibrium of the firm and the ‘group’ on the same diagram.

The above ‘heroic’ assumptions lead to a model which is very re­strictive, since it precludes the inclusion in the ‘group’ of similar products which have different costs of production. Chamberlin himself recognizes that the ‘heroic’ assump­tions are unrealistic and he relaxes them at a later stage.