The welfare economics of macroeconomics and chooser-dependent, non-expected utility preferences: A Senian critique with an application to the costs of the business cycle

The advent of representative agent, microfounded macroeconomics brought with it new techniques for analyzing the welfare effects of macroeconomic policy. The agent's utility function becomes a measuring rod for the costs and benefits of policies that alter the path of consumption over time. Robert E. Lucas used this method in his Presidential Address to the AEA, when he claimed that the welfare costs of the business cycle in the United States equaled .05% of consumption. Applying Amartya Sen's chooser-dependent preferences to this class of welfare measurements, I will suggest that in carrying out public duties, policymakers must use different sorts of reasoning than optimizing consumers, and that people have different attitudes toward risk that is imposed than risk that is voluntarily taken on. Hence, different objective functions are appropriate in each case. These conceptual arguments lead to new expressions for the welfare effects of macroeconomic policies for use when preferences are chooser-dependent and not of the expected value, discounted utility form and to caveats for the estimation and calibration of utility functions used in such welfare analysis.