Preference Heterogeneity and Optimal Capital Income Taxation

by Mikhail Golosov, Maxim Troshkin, Aleh Tsyvinski & Matthew Weinzierl

Overview — Professor Matthew Weinzierl and coauthors test the idea that savings, which is concentrated among highly skilled workers, ought to be taxed as part of an optimal tax policy. They find that the welfare gains from these taxes would be negligible. Key concepts include:

Is taxing capital income (the return to saving) a good idea when highly skilled people value that income more than others?
A baseline simulation finds that the optimal capital income taxes are modest--only 2 percent on average and 4.5 percent on high earners.
Welfare gains from these optimal capital income taxes would be negligible.

Author Abstract

We examine a prominent justification for capital income taxation: goods preferred by those with high ability ought to be taxed. In an environment where commodity taxes are allowed to be nonlinear functions of income and consumption, we derive an analytical expression that reveals the forces determining optimal commodity taxation. We then calibrate the model to evidence on the relationship between skills and preferences and extensively examine the quantitative case for taxes on future consumption (saving). In our baseline case of a unit intertemporal elasticity, optimal capital income tax rates are 2 percent on average and 4.5 percent on high earners. We find that the intertemporal elasticity of substitution has a substantial effect on optimal capital taxation. If the intertemporal elasticity is one-third, optimal capital income tax rates rise to 15 percent on average and 23 percent on high earners; if the intertemporal elasticity is two, optimal rates fall to 0.6 percent on average and 1.6 percent on high earners. Nevertheless, in all cases that we consider the welfare gains of using optimal capital taxes are small.