Bargaining Before or After Communication?

The paper has benefited greatly from comments by an anonymous referee and numerous seminar participants as well as discussions with Abstract We ask how bargainers' incentives to communicate about more efficient widget designs depend on whether they negotiate price prior to, or after, fixing the traded design. We find three effects: (1) Since communication reveals information about preferences, bargainers with little power prefer to remain quiet prior to bargaining. (2) Later bargaining gives communicators a chance to share in joint gains from more efficient trades. (3) The revealed preference information enhances the efficiency of the bargaining process. The comparison might help explain why some contracts have more features left incomplete and throw some light on the nature of the employment relationship. 3 Most reasonable people would agree that bargaining is costly and by implication that rational agents can be expected to take steps to avoid or minimize these costs. It has, however, proven difficult to build a satisfactory theory of bargaining avoidance. The challenges include modeling the costs of bargaining, accounting for its advantages, and comparing to alternative trading institutions. We will here take a small step in this direction. In particular, we look at a context in which identification of efficient trades requires communication and ask whether bargaining should take place before or after this communication. The question is of obvious theoretical interest, but could also be mapped into recent theories of the firm by identifying " bargaining before communication " with employment relationships and " bargaining after communication " with arms length contracts. The costs of bargaining after communication are that the players may withhold certain types of information. We will call this the " bargaining power effect ". The essential idea is that communication carries a signal about a player's preferences and therefore reduces his expected share of gains from trade. Consistent with this argument Busse, Silva-Risso, and Zettelmeyer [2006], and Scott-Morton, Zettelmeyer and Silva-Risso [2005] show that better-informed bargainers get bigger shares of surplus, while Simester and Knez [2002] show that firms take steps to restrict communication between employees and outside trading partners. The costs of bargaining prior to communication are twofold. One straightforward problem is that the terms of trade can not be changed to reflect the content communicated. Players will thus only communicate information that is beneficial to them under the agreed upon terms of trade. With bargaining after communication, one would expect agreement …