Essays in normative macroeconomics

This thesis is divided into two main parts. The first provides a novel analysis of dynamic optimal taxation under the assumption that individuals in an economy have ‘hidden’ idiosyncratic productivity levels. Specifically, it shows how to derive a complete set of optimality conditions characterising the solution to a problem of this kind. The method relies on constructing perturbations to the consumption-output allocations of agents in a manner that preserves all relevant incentive compatibility restrictions. We are able to use it to generalise the ‘inverse Euler condition’ to cases in which preferences are non-separable between consumption and labour supply, and to prove a number of novel results about optimal income and savings tax wedges. The second main part investigates a more general problem. When policymakers are constrained in their present choices by expectations of future outcomes a well-known time-inconsistency problem hinders optimal decision-making: the preferences of policymakers who exist at different points in time are not in agreement with one another, because of differences in the constraints faced by each. We present a new approach to determining policy in this setting, based on asking: What policy would be chosen by a decisionmaker who did not know the time period in which their choice was to be implemented? This is akin to designing institutions from behind a Rawlsian ‘veil of ignorance’. The theory is used to obtain qualitative policy prescriptions across a number of environments; these policies have several appealing properties that we outline.EThOS - Electronic Theses Online ServiceGBUnited Kingdo