A Generic Negotiation Game

Steven J. Brams

    Research output: Contribution to journalArticle

    Abstract

    A generic Negotiation Game (NG) is used to model a conflict between two parties seeking to resolve their differences and reach a settlement. NG is a 2 — 2 non-constant-sum symmetric game that is ‘generic’ in the sense that its payoffs, which are assumed to be cardinal, are only incompletely specified. Consequently, conclusions drawn about it are applicable to all games with which it is consistent, including such well-known games as Prisoners' Dilemma, Chicken, Deadlock and Stag Hunt. NG may be in exactly one of four mutually exclusive ‘states’, with different pure-strategy Nash equilibria associated with each state: (1) conflict alone; (2) cooperation alone; (3) a combination of conflict and cooperation; and (4) two win-lose outcomes, each of which favors a different player. The most potent force driving NG toward state (2) is reducing the difference between winning and losing; increasing the rewards of cooperation and the penalties of conflict are not as helpful unless they occur in tandem. Normative implications of these results for encouraging negotiated settlements are discussed.

    Original languageEnglish (US)
    Pages (from-to)53-66
    Number of pages14
    JournalJournal of Theoretical Politics
    Volume4
    Issue number1
    DOIs
    StatePublished - Jan 1992

      Fingerprint

    Keywords

    • Nash equilibrium
    • game theory
    • incomplete information
    • inducement
    • negotiation

    ASJC Scopus subject areas

    • Sociology and Political Science

    Cite this