In the game theory section we showed that a natural outcome for Harold and William is the Nash equilibrium. Yet there are other outcomes, although not sharing an equilibrium property, giving a better result in terms of utility payoffs. For example, think about the jointly improving Pareto optimal outcomes. In axiomatic bargaining theory the Pareto optimal outcomes were analysed in the utility set but no attention was paid to how to reach them. In negotiation analysis, such questions are addressed and methods to generate Pareto optimal outcomes are analysed.
The term negotiation analysis was first introduced by Raiffa (1982) who gave a comprehensive description about the field and the possible methods by integrating elements from game theory and multiple criteria decision analysis. In this section basic concepts in negotiation analysis are introduced.
In Harold’s and William’s problem, it is very easy to identify the negotiating parties. Nevertheless, it is not always easy to recognize different parties.
For instance, consider a water system regulation policy making. Usually there are many stakeholders involved in such a policy making. Suppose that it is possible to identify some interest groups such as power producers, farmers, environmentalists, recreational users and fishermen and every stakeholder belongs to some group. Nevertheless, it is not at all clear if these groups can act as parties in negotiation since it is well possible that the members within a group may not agree among themselves in such a situation. Thus there emerges pressure to splinter the groups to smaller ones.
In the case of Harold and William the issues are easily identified: the number of goats they choose. Nevertheless, they may include other issues, like money, to their negotiation or link completely separate negotiations together establishing the possibility to achieve better outcomes. In principle, the issues are chosen so that they describe sufficiently well the underlying negotiation situation. Note that the issue selection can be considered as a negotiation problem of its own.
Often the negotiating parties ask a third party to intervene the negotiation. His or her role is to suggest agreements or to offer facilities, like the communication forms, for the parties. One type of intervenor is a mediator that is a neutral party gathering some confidential information from the parties, making suggestions for them and assisting them to find a jointly accepted agreement. Usually a mediator is a person using a software to support the negotiation, but also a software alone can take the role of a mediator. In the latter case the parties use the supporting software by themselves.
Totally another type of intervenor is an arbitrator that analyses the problem and, unlike the mediator, dictates the solution for the parties.
Here we present the main types of negotiation procedures as classified by Raiffa (1982):
<![if !supportLists]>1. <![endif]>Whether the value functions of the parties are elicited or not.
<![if !supportLists]>2. <![endif]>Whether the parties
<![if !supportLists]>· <![endif]>make concessions or
<![if !supportLists]>· <![endif]>seek for joint improvements from a reference outcome.
that step by step seek for joint improvements are called single negotiation text (SNT) methods. This type of negotiation was developed
by Fisher and Ury (1981) and it was first applied in
below presents the
In concession based methods, the parties start the negotiation from separate positions and they are supposed to make concessions from their current positions until they finally reach each other.
For instance, suppose that Harold and William both insist, that the other is not allowed to graze his goats in the green at all. So, they are in the different initial positions and they do not have a tentative agreement as in the SNT-procedure. They make concessions by giving in more and more to the other party until they finally shake hands. The figure below shows an example about how they might proceed in the utility set.
As a result of concessions, they reach an agreement which might not be Pareto optimal. Hence, there still exists room for joint improvements and they could apply SNT-method.
In value function based methods, the parties’ utility, or value functions as they are called in the negotiation analysis context, functions are elicited. This makes it possible to construct all Pareto optimal outcomes and to let the parties finally choose among these. One possibility then is to select a particular outcome by using the axiomatic bargaining theory.
Since the elicitation of the parties’ value functions may be a formidable task a good alternative is to use interactive methods. They are used to prevent the elicitation of the value functions as a whole. In interactive methods, only local preference information from the parties is required. In practise, a mediator could ask the parties only a few relatively simple questions. For instance, the mediator might ask them to compare some outcomes.
<![endif]>Raiffa, H., J.
Richardson and D. Metcalfe (2002). Negotiation
Analysis: The Science and Art of Collaborative Decision Making. Belknap
<![endif]>Raiffa, H. (1982). Art and Science
<![if !supportLists]>· <![endif]>Fisher, R., and W. Ury (1981). Getting to Yes. Arrow.