主观期望效用偏好下的匿名且防策略投票
Anonymous and Strategy-Proof Voting under Subjective Expected Utility Preferences
AI总结：
本文在主观期望效用框架下刻画了同时满足匿名性、防策略性和范围一致性的社会选择函数，证明其可分解为常值或四种基本类型之一的二元投票规则。
摘要：
我们研究不确定性下受约束社会选择模型中的三个公理，其中（i）主体对行动具有主观期望效用偏好，且（ii）不同的自然状态具有（可能）不同的可用结果集。匿名性要求主体的姓名或标签在用于选择社会行动的机制中绝不发挥作用。防策略性要求在每个主体参与的直接显示博弈中，报告其真实偏好是（弱）占优策略。范围一致性本质上要求：只要一个可行行动被报告为每个投票者在机制范围内最偏好的行动，该行动就必须被社会选中。我们首先证明，每个满足这三个公理的社会选择函数都可以分解为若干投票规则的乘积，这些投票规则要么是常值规则，要么是二元规则（在每个状态下总是产生两个预先指定结果之一）。我们描述了二元因子的四种基本类型：其中三种类型是本文献中首次提出的，并利用了投票者的主观信念。我们的刻画结果随后表明，一个社会选择函数是匿名的、防策略的且范围一致的，当且仅当其规范分解中的每个二元因子都属于这四种基本类型之一。
英文摘要：
We study three axioms in the model of constrained social choice under uncertainty where (i) agents have subjective expected utility preferences over acts and (ii) different states of nature have (possibly) different sets of available outcomes. Anonymity says that agents' names or labels should never play a role in the mechanism used to select the social act. Strategy-proofness requires that reporting one's true preferences be a (weakly) dominant strategy for each agent in the associated direct revelation game. Range unanimity essentially says that a feasible act must be selected by society whenever it is reported as every voter's favorite act within the range of the mechanism. We first show that every social choice function satisfying these three axioms can be factored as a product of voting rules that are either constant or binary (always yielding one of two pre-specified outcomes in each state). We describe four basic types of binary factors: three of these types are novel to this literature and exploit the voters' subjective beliefs. Our characterization result then states that a social choice function is anonymous, strategy-proof and range-unanimous if and only if every binary factor (in its canonical factorization) is of one of these four basic types.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
