A two-period model on optimal taxation with learning incentives (Q1072921)

Consider a situation where individuals can freely choose how much of their time they want to spend for labour of a certain skill and for leisure respectively. Then it is well-known that a lump-sum tax on abilities is the best tax-system because it distorts the individual labor decisions as little as possible. But from a dynamic viewpoint one may argue that the formation of abilities will be affected by such a tax system. The aim of the paper is to analyse these effects, and to derive the optimal lump-sum tax under a two-period utilitarian welfare function under the basic assumption that the learning incentives for the workers depend on the differences of utility levels attainable for different skill levels. It is shown by nonlinear programming techniques and sensitivity analysis that the optimal tax distribution depends strongly on the underlying learning behavior.