Optimal fiscal policy in the automated economy

Adding (1) the endogenous labor supply of workers, (2) fiscal policy instruments, and (3) monopolistic competition to Berg et al.’s (2018) general equilibrium model of automation, we study how automation (i.e., robots and artificial intelligence) affects the efficacy of redistribution policy. Using the consumption equivalent welfare gain developed by Domeij and Heathcote (2004) and assuming a 50 percent increase in robot-augmented technology shock, we derive the optimal tax rates for various tax policy instruments in the steady state of the model economy calibrated for the United States. We find that the optimal capital income tax rate is 20 percent. Another finding is that the zero tax rate on the wage income of unskilled workers is an optimal tax policy. We also find that the optimal tax rates on robots and consumption are dependent on the preference of the government. Finally, we find that the Pareto-efficient optimal tax system is characterized as a combination of a 15.9 percent rate on capital income tax and a zero tax rate on unskilled workers’ income. Our analysis contributes to the literature on optimal taxation in the automated age.