Axiomatization of the Shapley value on minimum cost spanning tree games (Q700116)

\(N\) agents must be connected, not necessarily directly, among themselves and to a root \(0\), for example because they want cable a network in their houses. So one has to build a network, formally a \textit{connected graph}, which is a set of edges \((i,j)\), \(i,j\in N\cup \{0\}\), such that every pair of nodes is connected by a path; and building the edge \((i,j)\) costs \(c_{ij}\). Consider networks with minimum total building costs. How to share these costs among the agents; that is, what share \(\Phi_i(C)\) of the total cost should agent \(i\) bear given the cost matrix \(C=(c_{ij})\)? This is the question of the paper under review. The game-theoretical approach to this problem associates to the cost matrix \(C\) a transferable utility cooperative game, where the cost of each coalition \(S\subseteq N\) are defined to be the minimum costs of connecting \(S\cup \{0\}\). Once the cooperative games is set up one can use solution concepts of cooperative game theory to solve the given cost allocation problem, like the core or Shapley value. The present paper characterizes the Shapley Value of the game associated to the network building problem through a simple set of axioms based on the latter problem's primitive, i.e.\ the matrix \(C\).