Collaborative Research: Communication, Cartels and Collusion

Maintaining healthy competition among firms is key to the well-being of the economy. Cartels stifle such competition, leading to higher prices and restricted choices for consumers. Indeed, the purpose of antitrust law is to ensure that this does not happen. Case documents from recent cartel prosecutions in the US and Europe point to the important role that regular and frequent---sometimes even weekly---communication plays in sustaining high prices. Economic theory, on the other hand, is more or less silent on whether or not communication facilitates collusion. In this project, the PIs propose to study both the kinds of environments and the kinds of communication which are harmful to consumers. For instance, firms frequently exchange sales figures claiming that this is actually beneficial to the economy---it allows firms to ascertain market conditions and improve investment decisions. The PIs' preliminary research indicates that the exchange of sales figures can also be detrimental---it allows firms to monitor each other to see if cartel members are complying with the agreement. This suggests that society would benefit if limits were placed on such communication---say, if sales figures were exchanged only yearly rather than quarterly. This project will seek to develop a better theoretical understanding of when and how such a policy would improve welfare.

Examining the role of communication in cartels would represent both a significant intellectual advance in our understanding of collusion, and a practical advance in guiding government and business policy.

Specifically, this project is driven by two major questions. First, how does inter-firm communication affect competition? What forms of communication are detrimental to society and what forms are not? Second, how does communication facilitate cooperation in long-term relationships in general? Work by Awaya and Krishna has addressed the first question in situations in which market demand is rather variable. They showed that in this environment, absent any communication, firms are not able to successfully collude because detecting departures from any cartel agreement is difficult. But if firms can communicate, by exchanging even unverifiable sales data, then they can sustain high prices and profits. The fact that even unverifiable data can be used to foster collusion is surprising. The main objective of the first part of the project is to explore the extent to which this finding extends to other kinds of communication---in particular, the exchange of aggregate data via trade associations which the courts deem to be legal.

The second part of the project concerns the extension of these ideas to other contexts. It asks whether communication facilitates cooperation in general long-term relationships, specifically, general repeated games with private monitoring. Preliminary research indicates that this is not always the case. The project will seek to identify and classify situations in which communication has a positive benefit to the participants. It will develop methods for bounding the set of equilibrium payoffs in repeated games without communication. It will then seek to identify what forms of communication can overcome the bounds and further, what forms are most beneficial.

The questions addressed have direct implications for antitrust policy. For instance, it is an important policy as to what kinds of inter-firm communication are potentially detrimental to society.