An Approach to Communication Equilibria

Mediated contracts and mechanism design

“…This, in particular, allows correlation between random draws of recommendations and random draws of transfer schedules. As a consequence, one can, for any mechanism in the framework of Rahman and 5 See also Forges (1986) and Myerson (1986), (1991). 6 When the action set of the agents is finite but transfers are non-countable, the definition of a mechanism in Rahman and Obara (2010) is, from a measure theoretical viewpoint, simpler than Myerson's view, because…”

“…1 The authors clearly show that, in team problems with balanced budgets, mediated contracts outperform standard, non-mediated ones and even enable a virtual implementation of the first best. What remains less clear is how their concept of mediation relates to other concepts of mediation in earlier work on mechanism design (e.g., Myerson 1986, 1991, Forges 1986). This note tries to clarify this link by contrasting mediated contracts to the mechanisms of Myerson (1982), who extends the revelation principle to settings with moral hazard.…”

“…For any pair of recommendations the output is shared equally, except for the recommendation (1,0) and (0, 1). For these two special cases, the agent who is given the recommendation to work efficiently (e r = 1) receives 1/ε times the output as a transfer, whereas the agent who is told to shirk (e r = 0) receives the (negative) transfer (ε − 1)y/ε.…”

Uniform Equilibrium: More than Two Players

“…The most general device receives at every stage some private message from each player and has perfect recall (communication device, Forges (1986Forges ( , 1988, Myerson (1986), Mertens (1994)). The most restrictive device bases its choice only on the currect state (and not even on past signals) (Nowak and Raghavan (1992)).…”

Cheap talk in games with incomplete information

“…The following list mentions only the work that is closely related to ours and two papers which deal with applications. The notion of a communication equilibrium was first studied by Forges [8] and Myerson [15]. Aumann and Hart [4] use the notion of a bi-martingale 4 to characterize the set of equilibrium payoffs in a cheap talk extension of a two-player game in which only one player has private information.…”