4. Information Exchanges among Firms and Their Welfare Implications (Part I): The Dual Relations between the Cournot and Bertrand Models

This long series of papers consist of three parts. Part I is concerned with the basic dual relations between the Cournot and Bertrand models. Part II begins to deal with the world of risk and uncertainty, with a discussion of the Cournot duopoly model with a common demand risk as a starting point. It then deals with other types of duopoly models with a common risk. Part III discusses more complicated problems such as private risks and oligopoly models. All these three parts taken together aim to carefully outline and critically evaluate the problem of information exchanges in oligopoly models, one of the most important topics in contemporary economics.

The true motivation of writing such survey papers is to strive for a synthesis of the economics of imperfect competition and the economics of imperfect information. The problem at issue is how and to what extent the information exchanges among firms influence the welfare of producers, consumers, and the whole society. It is seen in the paper that a definite answer to the problem really depends on the following many factors. (1) The type of competitors (Cournot-type or Bertrand-type), (2) the nature of risk (a common value or private values; demand risk or cost risk), (3) the degree and direction of physical and stochastic interdependence among firms, and (4) the number of firms. If any set of those factors is specified in a given oligopoly model, the welfare and policy implications may very systematically be derived by way of their decomposition into the following four effects. That is, (i) own variation effects, (ii) cross variation effects, (iii) own efficiency effects, and (iv) cross efficiency effect. In the real world, trade associations may be regarded as typical information exchange mechanisms. Many welfare implications obtained in the papers will shed a new light to the effectiveness and limitations of those trading groups.