Surplus Maximization and Optimality

Expected consumer's surplus rarely represents preferences over price lotteries. Still, I give sufficient conditions for policies which maximize aggregate expected surplus to be interim Pareto Optimal. Besides two standard partial equilibrium conditions, I assume that feasible prices satisfy a single-crossing property; and each consumer's indirect utility satisfies increasing differences in the price and income. I use the result to extend well-known welfare conclusions beyond the knife-edge quasilinear utility case. Since increasing differences puts no upper bound on risk aversion, the result is useful for applications in which risk aversion is important.