发表机构
Lingnan College, Sun Yat-sen University（中山大学岭南学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文在严格凸体边界上，用对角一致性刻画诚实选择，证明策略proof性导致固定独裁，而内部结果允许加权平均妥协。
AI 中文摘要
集体规则通常必须在受约束的边界上选择公共结果。我们在有限维数至少为二的任意紧致全维严格凸体的整个边界上，用对角一致性刻画了诚实选择。在线性支撑偏好的全域上，普通的个体策略proof性和对角一致性迫使规则在每个配置下选择某个固定参与者的偏好点。总体是任意固定的有限非空集合。不假设帕累托条件、边界光滑性，或规则的连续性或可测性。如果允许内部结果，则两个或更多参与者可以使用偏好点的正固定加权平均，在同一类型域上进行诚实的非独裁妥协。证明从两个参与者的激励不等式导出正则性，跨报告传输一个共同的支撑测度界，并使每个凸化的可达菜单成为可行体的闵可夫斯基和项。边界生成则迫使固定控制。该结果扩展了球形公共物品定位问题，并识别了边界值选择在消除诚实妥协中的作用。
英文摘要
A collective rule must often choose a public outcome on a constrained boundary. We characterize truthful choice with diagonal unanimity on the entire boundary of any compact full-dimensional strictly convex body in finite dimension at least two. On the full domain of linear support preferences, ordinary individual strategy-proofness and diagonal unanimity force the rule to select one fixed participant's preferred point at every profile. The population is any fixed finite nonempty set. No Pareto condition, boundary smoothness, or continuity or measurability of the rule is assumed. If interior outcomes are allowed instead, two or more participants can use positive fixed weighted averages of preferred points for truthful non-dictatorial compromise on the same type domain. The proof derives regularity from the two agents' incentive inequalities, transports a common support-measure bound across reports, and makes every convexified attainable menu a Minkowski summand of the feasible body. Boundary generation then forces fixed control. The result extends the spherical public-good location question and identifies the role of boundary-valued choice in eliminating truthful compromise.
Comments28 pages, including appendices
