Abstract:
The literature on the growth effects of flat rate taxes focuses entirely on models with infinite horizons and constant returns to private inputs in human capital accumulation. In contrast, the traditional human capital literature assumes finite lifetimes and, based on microeconomic evidence, diminishing returns. Since human capital is not heritable, the infinite horizon framework cannot, in general, be justified as a reduced form of an overlapping generations model; it represents an approximation of unknown accuracy. This paper shows that abstracting from finite lives and diminishing returns makes an important difference for the growth effect of taxation, even if there is an operative bequest motive. Except in a knife-edge case, the growth effects of taxation become small (and may go to zero) as lifetimes get long, suggesting that infinite horizons may be a terrible approximation for long, finite horizons.