A comment on issue-by-issue negotiations (Q1359011)

This note demonstrates that, in the two issue, incomplete information bargaining problem of \textit{M. Bac} and \textit{H. Raff} [Games Econ. Behav. 13, No. 1, 125-134 (1996; Zbl 0848.90134)], the use of issue-by-issue bargaining by the high discount factor type arises not because of any special signaling value associated with this procedure but purely because it is the least cost signal. Specifically, assumptions on the size of discount factors maintained by Bac and Raff imply that signaling would arise even if an issue-by-issue bargaining procedure were not available. When such a procedure is available, it is only because agents obtain the surplus from an issue as soon as agreement is reached on it that issue-by-issue bargaining is the preferred signaling strategy. The note also argues that both theorems in Bac and Raff are incorrect as stated, but that the key theorem (Theorem 2) on signaling via issue-by-issue bargaining can be rescued. Finally, the note calls attention to papers by Busch and Horstmann in which issue-by-issue bargaining does serve a special signaling role, in that it is the only means by which signaling can (and does) occur.