Games and Economic Behavior最新文献

We formulate and study the requirement on an allocation rule that no agent should be able to benefit by artificially augmenting their endowment. This can be either through simply exaggeration or through a transfer of resources from outside of the current trading partners, resources that have to be returned after the rule is applied and the agent has received their assignment. We show that the Walrasian rule is not “augmentation-proof” even on standard domains. More seriously, no efficient selection from the individual-endowments lower bounds correspondence, or from the no-envy correspondence, or from the egalitarian-equivalent correspondence is augmentation-proof. These impossibilities hold even when preferences are homothetic, and even if the agent cannot augment their endowment by more than an arbitrarily small proportion of the resources they truly own.

It is well known that in the absence of transfers, stable matchings are positive assortative when agents' preferences are strictly monotonic in each other's types and types are commonly known. Instead, monotonicity is consistent with the existence of stable matchings that exhibit negative sorting when types on one side of the market are private information. This paper sheds light on the scope and meaning of this consistency by showing that within familiar monotonic markets we cannot conclude that a stable matching is negative, and not positive, assortative.

Many-player divide-the-dollar (DD) games have been a workhorse in the theoretical and experimental analysis of multilateral bargaining. If we deal with a loss or consider many-player divide-the-penalty (DP) games, the theoretical predictions are not simply those from DD games with the sign flipped. We show that the stationary stage-undominated equilibrium (SSUE) is no longer unique in payoffs. The most “egalitarian” equilibrium among the stationary equilibria is a mirror image of the essentially unique SSUE in the Baron–Ferejohn model. That equilibrium's allocations are sensitive to changes in parameters, while the most “unequal” equilibrium is less affected by such changes. Experimental evidence supports the most unequal equilibrium: Most of the approved proposals under a majority rule involve an extreme allocation of the loss to a few members. Other observations such as no delay, the proposer advantage, and the acceptance rate are also consistent with predictions based on the most unequal equilibrium.

We study how voting rules shape voter coordination in large three-candidate elections. We consider three rules, that differ on the number of candidates that voters can support: Plurality (one), Anti-Plurality (two) and Approval Voting (one or two). We show that the Condorcet winner is always elected at some equilibrium under Approval Voting, and that this rule provides better welfare guarantees than Plurality. We then numerically study a dynamic process of political tâtonnement which delivers rich insights. The Condorcet winner is virtually always elected under Approval Voting, but not under the other rules. The dominance of Approval Voting is robust to several alternative welfare criteria and the introduction of expressive voters.

We investigate a widely used mechanism to resolve disputes among business partners, known as Texas Shoot-Out, under Knightian uncertainty. For a non-degenerate range of intermediate valuations, an ambiguity-averse divider truthfully reveals their valuation in equilibrium. As a consequence, the resulting outcome is efficient, in contrast to the Bayesian case. With equal shares, both agents prefer ex-ante to be the chooser and would like to avoid triggering the mechanism in the first place.

In multi-battle team contests with pairwise battles, how battles are organized—sequentially or (partially) simultaneously—may affect the expected winner's total effort (WE), a natural objective in R&D races, elections, and sports. We focus on noise (modeled via the contest success function's discriminatory power) and across-team heterogeneity, abstracting from player-specific heterogeneity. With sufficient noise, we show that: (1) If teams are symmetric, all temporal structures yield the same WE; and (2) If teams are asymmetric, WE is maximized by a fully simultaneous contest and minimized by a fully sequential one. With no noise, we show that: (3) If teams are symmetric, WE is maximized by a fully sequential contest and minimized by a fully simultaneous one; and (4) If teams are asymmetric, neither the fully sequential nor the fully simultaneous temporal structures maximize or minimize WE. Our results use a novel technique that simplifies temporal structure comparisons: extractions and mergers.

In 1971, Wilson (1971) proved that “almost all” finite games have an odd number of mixed Nash equilibria. Since then, several other proofs have been given, but always for mixed extensions of finite games. In this paper, we present a new oddness theorem for large classes of polynomial payoff functions and semi-algebraic sets of strategies. Additionally, we provide some applications to recent models of games on networks such that Patacchini-Zenou's model about juvenile delinquency and conformism (Patacchini and Zenou, 2012), Calvó-Armengol-Patacchini-Zenou's model about social networks in education (Calvó-Armengol et al., 2009), Konig-Liu-Zenou's model about R&D networks (König et al., 2019), Helsley-Zenou's model about social networks and interactions in cities (Helsley and Zenou, 2014).

I study the impact identity change (assimilation) has on economic mobility. I experimentally assign people to different group identities, majority or minority, before they interact in a social coordination setting. In equilibrium, minority assimilators achieve economic mobility by integrating with the majority. In the experiment, assimilators are discriminated against and cannot integrate, if majority members encounter conformists (non-assimilators) in the minority. Thus, assimilators fail to attain economic mobility because those who maintain the status quo impose negative externalities on those who risk changing it.

We consider an auction design problem with private values, where the seller and bidders may enjoy heterogeneous priors about their (possibly correlated) valuations. Each bidder forms an (interim) belief about the others based on his own prior updated by observing his own value. If the seller faces uncertainty about the bidders' priors, even if he knows that the bidders' priors are within any given distance from his, he may find it worst-case optimal to propose a dominant-strategy auction mechanism.

Condorcet domains are sets of preference orders such that the majority relation corresponding to any profile of preferences from the domain is acyclic. The best known examples in economics are the single-peaked, the single-crossing, and the group separable domains. We survey the latest developments in the area since Monjardet's magisterial overview (2009), provide some new results and offer two conjectures concerning unsolved problems. The main goal of the presentation is to illuminate the rich internal structure of the class of maximal Condorcet domains. In an appendix, we present the complete classification of all maximal Condorcet domains on four alternatives obtained by Dittrich (2018).