A characterization of convex games by means of bargaining sets (Q956580)

The aim of the paper is to characterize the classical convexity notion for cooperative games with transferable utility function [see \textit{L. S. Shapley}, Int. J. Game Theory 1, 11--26 (1971; Zbl 0222.90054); \textit{T. Ichiishi}, J. Econ. Theory 25, 283--286 (1981; Zbl 0478.90092); \textit{R. J. Weber}, The Shapley value. Essays in honor of Lloyd S. Shapley, 101--119 (1989; Zbl 0707.90100)] by means of the Mas-Colell and the Davis-Maschler bargaining set [see \textit{A. Mas-Colell}, J. Math. Econ. 18, No. 2, 129--139 (1989; Zbl 0676.90011); \textit{M. Davis} and \textit{M. Maschler}, Bull. Am. Math. Soc. 69, 106--108 (1963; Zbl 0114.12501)]. The authors introduce the concept of max-Weber set defined as the convex hull of the max-marginal worth vectors. A max-marginal vector is a vector of payoffs assigned in such a way that a player obtains the maximum of his/her marginal contribution to a subset of predecessors (according to some ordering). There are two main results. First, it is proven that for a cooperative game with a non-empty core, convexity is characterized by the inclusion of the Weber set in the Davis and Maschler bargaining set or, alternatively, in the Mas-Colell bargaining set. The second result gives a general characterization of convexity under the condition that the corresponding bargaining set is equal to the max-Weber set.