This paper investigates the impact of guilt aversion on the equilibria of the class of symmetric games that are characterized by the same Nash equilibrium structure as the Hawk–Dove game: two asymmetric strict pure Nash equilibria and a completely mixed-strategy equilibrium. We classify these games, which we refer to as generalized Hawk–Dove games, into two subclasses, Type 1 and Type 2, based on players’ preferences over deviations toward symmetric profiles. We characterize the best reply correspondences and equilibria under guilt aversion, extending the analysis to cases involving ambiguity in second-order beliefs. We show that the resulting outcomes are sensitive to guilt sensitivity parameters and players’ attitudes toward ambiguity. For Type 1 games, when guilt sensitivity exceeds a specific threshold, a new symmetric equilibrium emerges, while ambiguity under pessimism makes this symmetric profile the unique equilibrium. Under optimism, guilt aversion is neutralized, and the equilibrium set coincides with that of the classical game. For Type 2 games, guilt aversion only affects the mixed equilibrium, but under pessimism, an infinite set of mixed equilibria appears, whereas optimism restricts the set of equilibria just to the two pure asymmetric profiles.
Guilt aversion and ambiguity in generalized Hawk-Dove games / De Marco, G., Romaniello, M., Roviello, A.. - In: DECISIONS IN ECONOMICS AND FINANCE. - ISSN 1593-8883. - (2026). [10.1007/s10203-026-00567-8]
Guilt aversion and ambiguity in generalized Hawk-Dove games
Roviello, Alba
2026
Abstract
This paper investigates the impact of guilt aversion on the equilibria of the class of symmetric games that are characterized by the same Nash equilibrium structure as the Hawk–Dove game: two asymmetric strict pure Nash equilibria and a completely mixed-strategy equilibrium. We classify these games, which we refer to as generalized Hawk–Dove games, into two subclasses, Type 1 and Type 2, based on players’ preferences over deviations toward symmetric profiles. We characterize the best reply correspondences and equilibria under guilt aversion, extending the analysis to cases involving ambiguity in second-order beliefs. We show that the resulting outcomes are sensitive to guilt sensitivity parameters and players’ attitudes toward ambiguity. For Type 1 games, when guilt sensitivity exceeds a specific threshold, a new symmetric equilibrium emerges, while ambiguity under pessimism makes this symmetric profile the unique equilibrium. Under optimism, guilt aversion is neutralized, and the equilibrium set coincides with that of the classical game. For Type 2 games, guilt aversion only affects the mixed equilibrium, but under pessimism, an infinite set of mixed equilibria appears, whereas optimism restricts the set of equilibria just to the two pure asymmetric profiles.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


