Price communications in a multimarket context: An experimental investigation

This experiment assesses the effects of nonbinding price communications in a multi-market, posted-offer environment. In half of the ten sessions, three symmetric sellers continuously submitted nonbinding prices for two minutes prior to posting final binding prices. In the remaining sessions sellers posted only binding prices. Competitive prices were observed infrequently in either treatment, but prices were persistently higher when communications were possible. The way that communications affect performance is unclear. With or without communications, high prices appear to be more a consequence of some sellers supporting the defections of others, than of a developed language of conspiracy.We wish to thank without implicating Will Gillespie, participants in seminars at the Antitrust Division of the Department of Justice and at Virginia Commonwealth University, and an anonymous referee. Support for this research was provided by the USC Zumberge Faculty Research and Innovation Fund, the VCU Grants-In-Aid Program, and the National Science Foundation (grant SBR 9319842). Data are archived at FTP address: fido.econlab.arizona.edu.