Working Paper: CEPR ID: DP14993
Authors: Michel Le Breton; Alexander Shapoval; Shlomo Weber
Abstract: In this paper we examine a game-theoretical generalization of the landscapetheory introduced by Axelrod and Bennett (1993). In their two-bloc settingeach player ranks the blocs on the basis of the sum of her individualevaluations of members of the group. We extend the Axelrod-Bennett settingby allowing an arbitrary number of blocs and expanding the set of possibledeviations to include multi-country gradual deviations. We show that aPareto optimal landscape equilibrium which is immune to profitable gradualdeviations always exists. We also indicate that while a landscapeequilibrium is a stronger concept than Nash equilibrium in pure strategies,it is weaker than strong Nash equilibrium.
Keywords: landscape theory; landscape equilibrium; blocs; gradual deviation; potential functions; hedonic games
JEL Codes: C72; D74
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Pareto optimal landscape equilibrium (D50) | immune to profitable gradual deviations (F12) |
| players' strategies (C72) | evaluations of coalitions (D79) |
| evaluations of coalitions (D79) | resulting equilibrium states (D50) |
| symmetry assumption of the proximity matrix (C10) | existence of a landscape equilibrium (C62) |
| players evaluate their coalition based on pairwise proximity coefficients (D79) | landscape equilibrium established (C62) |
| players' influence parameters (Z22) | evaluations of coalitions (D79) |
| landscape equilibrium reinforces the concept of Nash equilibrium (C62) | countries reach optimal outcome (O57) |
| strong Nash equilibria may not exist (C72) | complexity of deviations allowed in the model (C52) |