Abstract ABSTRACT: Cushing 1977 has suggested that one approach to avoiding the social choice paradoxes discussed in Demski 1974 might be to assume some degree of homogeneity in the beliefs and preference structures of individuals. In particular, he has speculated that there may well exist a set of assumptions about the nature of individual beliefs and tastes which are more appealing than complete homogeneity, but less rigid than complete heterogeneity, and which might lead to a theory of social choice offering greater promise of resolving financial reporting issues. This paper shows that the joint assumptions of homogeneous beliefs and quadratic utility functions are sufficient to rule out the social choice paradoxes in the case of social choice over lotteries. Unfortunately, this result does not extend to the case of social choice over information systems. The final section shows that there exists no class of twice continuously differentiable utility functions, with different measures of local absolute risk aversion for some level of payoff, which in combination with the assumption of homogeneous beliefs, is sufficient to rule out Condorcet's majority voting paradox.
Martin Walker (Sun,) studied this question.