On the equivalence of optimal mechanisms with loss and disappointment aversion
研究了在准线性机制设计框架下，允许代理人具有损失厌恶或失望厌恶偏好时，两类偏好模型的最优机制在广泛问题中具有等价性，表现出显著的稳健性。
We consider a standard, quasi-linear mechanism design setting in which agents’ outcomes consist of a binary part and a transfer, thus encompassing applications such as auctions, bilateral trade or public good provision. We augment preferences by allowing for loss aversion (Kőszegi and Rabin, 2007) and disappointment aversion (Bell, 1985; Loomes and Sugden, 1986). While the preferences induced by these models only have a trivial intersection given by classical expected utility (Masatlioglu and Raymond, 2016), we show that the optimal mechanisms for the two types of preferences are equivalent across a broad range of problems and thus display a remarkable robustness.