This paper considers the impact of endogenous human capital accumulation on optimal tax policy in a life cycle model. Including endogenous human capital accumulation, either through learning-by-doing or learning-or-doing, is analytically shown to create a motive for the government to use age-dependent labor income taxes. If the government cannot condition taxes on age, then it is optimal to use a tax on capital in order to mimic such taxes. Quantitatively, this work finds that introducing learning-by-doing or learning-or-doing increases the optimal tax on capital by forty or four percent, respectively. Overall, the optimal tax on capital is thirty five percent larger in the model with learning-by-doing compared to the model with learning-or-doing implying that how human capital accumulates is of significant importance when determining the optimal tax policy.
JEL: E24, E62, H21.Key Words: Optimal Taxation, Capital Taxation, Human Capital.In their seminal works, Chamley (1986) and Judd (1985) determine that it is not optimal to tax capital in an infinitely-lived agent model. In such a model, taxing capital income is equivalent to an ever increasing tax on future consumption, thus implying an exponentially increasing distortion between the marginal rate of substitution and the marginal rate of transformation. In contrast, in a life cycle model agents live for a finite number of periods so the distortion imposed by a capital tax is bounded and may not necessarily be bigger than the distortions caused by other taxes. Atkeson et al. (1999), Erosa and Gervais (2002), and Garriga (2001) demonstrate in simplified life cycle models that if the government cannot condition labor income taxes on age, then it will generally tax capital in order to mimic an age-dependent tax. 1 The * E-mail: william.peterman@gmail.com. Views expressed on this site are my own and do not reflect the view of the Federal Reserve System or its staff. For extensive discussions and helpful comments, I thank Irina Telyukova, Valerie Ramey, Roger Gordon, and Scott Borger, as well as seminar participants at University of California at San Diego, Madrid Macroeconomic Workshop, the Federal Reserve Board of Governors, the Federal Reserve Bank of Philadelphia, the Eastern Economics Association Conference, the Missouri Economics Conference, the Midwestern Macroeconomics Conference, and the Conference in Computing in Economics and Finance. A previous version of this paper was distributed under the title "The Effect of Learning-by-Doing on Optimal Taxation" 1 Atkeson et al. (1999) demonstrate a related result. They show conditions under which the optimal tax on capital is zero if agedependent taxes on labor income are allowed. Gervais (2010) demonstrates that a progressive labor income tax can also mimicking age-dependent taxes on labor income.