上海财经大学经济学院：贝叶斯聚合，异质信念与价值观

 

　　不确定条件下的个人决策贝叶斯理论规定，个人断定一个实用功能，在事件的概率分布，并评估其期望效用的基础上，每一幕。我们试图推广这种模式，社会决策。我们所考虑的聚集规则要求社会的信仰和价值观分别为个人的信仰和价值观的仿射聚集。它是公知的，标准的帕累托条件是不与这些单独的聚合规则兼容。两个自然延伸帕累托条件的话建议。内Anscome-奥曼框架，我们表明，不变帕累托条件等价于分开的聚集(定理1和2)，而套头帕累托条件是等效于野蛮框架(定理3)内的单独聚合。

 

　　原文：Bayesian theory for individual decision-making under uncertainty prescribes that individuals posit a utility function, a probability　distribution over events and evaluate each act on the basis of its　expected utility. We attempt to generalize this model to social　decision-making. The aggregation rules we consider require society's　beliefs and values to be respectively the affine aggregation of　individuals beliefs and values. It is well-known that the standard　Pareto condition is not compatible with such separate aggregation　rules. Two natural extended Pareto condition are then suggested.　Within the Anscome-Aumann framework, we show that the Invariant Pareto　condition is equivalent to separate aggregation (Theorems 1 and 2)　while the Hedging Pareto condition is equivalent to separate　aggregation within the Savage framework (Theorem 3).