Principles of Automated Negotiation by Shaheen Fatima Sarit Kraus & Michael Wooldridge

Author:Shaheen Fatima, Sarit Kraus & Michael Wooldridge [Shaheen Fatima, Sarit Kraus & Wooldridge, Michael]
Language: eng
Format: epub
Publisher: Cambridge University Press
Published: 2014-10-19T22:00:00+00:00

7.5Historical notes and further reading

The existing literature on agendas can broadly be divided into two types: those that treat the agenda as an endogenous variable and those that treat it as an exogenous variable. The difference between these two is that, for the former, an agenda is selected during the process of negotiation over the issues. For the latter, an agenda is decided first, and then the parties negotiate over the issues on the agenda. Our focus in this chapter was on the latter category.

Research on endogenous agendas includes Bac and Raff (1996), Busch and Horstman (1997), Inderst (2000), In and Serrano (2003), Fatima et al. (2004). Bac and Raff (1996) and Busch and Horstman (1997) dealt with two divisible issues in the context of an imperfect-information setting where one of the two players is uncertain about the other’s discount factor. For this setting, the authors studied the properties of the resulting equilibrium. Meanwhile, Inderst (2000) focused on multiple divisible issues and potentially infinite discrete time periods. Here, the parties are allowed to make an offer on any subset of a given set of issues but can only accept/reject a perfect offer. In this sense, an agenda is selected endogenously. The authors showed existence and uniqueness conditions for equilibria under the perfect-information assumption. A generalisation of these results to a larger class of utility functions was given by In and Serrano (2003) (still under the perfect-information assumption). Then, for a negotiation model based on reservation prices, conditions for existence of equilibria were determined by Fatima et al. (2004) for endogenous agendas.

Research on exogenous agendas includes the work of Fershtman (1990), who focused on bargaining over two divisible pies using the sequential procedure. For the perfect-information setting, he took the set of issues as given, and showed how the order in which the issues are negotiated affects the outcome of a negotiation. Although the influence of the agenda on the outcome of a negotiation was recognised, nothing was said about how to actually find an optimal agenda.

A slightly different approach (from the above ones) was taken by Busch and Horstman (1997). They considered the perfect-information setting and extended the model of Rubinstein (1985) by adding a preliminary period in which the agents bargain over the agenda. The outcome of this stage is then used as the agenda for negotiating over the issues.

1 The classification for the case where b is the last mover will be analogous.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.