Welfare without Happiness

Abstract

Normative analysis asks how we (as a society) should make trade-offs between individuals. Behavioral welfare economics extends the scope of this analysis to a single individual, asking how the individual should trade off potential motives. We argue that any faith in economists’ (behavioral or otherwise) abilities to resolve such problems in moral philosophy cannot be based on past accomplishments of welfare economics, and that the best way to understand welfare economics is to view it as a part of positive economics with no normative content. We distinguish three types of welfare analysis. We refer to them as welfare I, II, and III. Welfare I identifies Pareto efficient outcomes within a particular model and relates those to equilibrium outcomes. Welfare I includes results such as the first theorem of welfare economics or the Myerson and Satterthwaite theorem on two-sided bargaining—results that show when an efficient outcome must, or cannot, be attained in equilibrium. Such results are used to explain the persistence of certain economic institutions (such as competitive markets) or derive limitations on what can be achieved within any institutional framework (such as bargaining or auctions). Welfare II posits an objective function for a policymaker and evaluates different policies or institutions according to that objective function. In other words, welfare II analyzes the implications of a policymaker’s preferences. Policymakers often pursue sophisticated objectives that cannot be captured by the standard functional forms used to describe simple, selfinterested economic agents. Moreover, evidence regarding policymakers’ preferences is typically limited. Hence, discussions of policymaker