Bargaining set theory and stability in coalition governments

Abstract

Various bargaining set theories are compared as predictors of coalition government portfolio distribution. While the kernel and B 1-bargaining set are known to exist in voting games with side payments, it is argued that the kernel, and thus B 1, are poor predictors. The B 2-bargaining set, a subset of B 1, when it exists is shown to be a good payoff predictor in a fractionalized and depolarized parliamentary situation (Finland: 1945ndash;1971). Moreover this predictor provides some explanation for the formation of surplus (winning but not minimal) coalitions.