Negotiation On Rugged Landscapes: Explaining The Hidden Complexity That Drives Impasses

Authors:

Abstract: Why do negotiations fail even when all parties benefit from reaching agreement? Even when the ZOPA is wide—typically an easy negotiation— impasses can result from socio-cognitive factors such as biases, poor information, and perceptions of bad faith. However, we show that even when negotiators are working collaboratively and sharing information, negotiations can result in impasse due to the inherent computational complexity of finding an agreeable solution.

This paper formalizes the idea of “negotiation as search” and shows that negotiations can be characterized by features well-known in computational search models. We show that negotiation can be represented as a “rugged landscape” where it is possible to get fixated on “local optima” i.e. be unable to achieve movement despite the existence of better (e.g. agreement-inducing) but undiscovered alternatives. We test a range of hypothetical search strategies, showing that the computational complexity of negotiation holds across several behavioral assumptions and offer testable hypotheses.

Track: NEG

Keywords: Negotiations, Multi-party Multi-issue Negotiations, Negotiation Process, Search, Complexity, Impasse, Computational Model