Optimal Uniform Capital Taxation in a Partially Automated Society

87 Pages Posted: 26 Nov 2019 Last revised: 13 Jan 2020

See all articles by Hideto Koizumi

Hideto Koizumi

University of Pennsylvania - Business & Public Policy Department; Hitotsubashi University

Date Written: October 22, 2019


A recent rapid-automation movement has been displacing routine labor and has sparked a series of discussion about taxation on automation such as a robot tax. However, the government's dilemma is that the planner may want to tax such physical capital that displaces routine labor--for example, industrial robots--for redistributive motives but does not want to tax other physical capital that increases such workers' productivities for instance, collaborative robots (cobots). This paper studies a novel setting of the optimal capital taxation on physical capital in which there is asymmetric information on both labor types and capital types between the planner and market. In particular, my model focuses on a two-by-two scenario where there are two types of labor (routine and non-routine labor) and two types of capital (displacing and reinstating capital). Despite asymmetric information, I find that the optimal uniform capital tax rate over different types of capital is strictly positive, as long as the solution is interior.

Keywords: Automation, Optimal Capital Taxation, Optimal Taxation, Robot Tax

JEL Classification: H21, H25

Suggested Citation

Koizumi, Hideto, Optimal Uniform Capital Taxation in a Partially Automated Society (October 22, 2019). Available at SSRN: https://ssrn.com/abstract=3486827 or http://dx.doi.org/10.2139/ssrn.3486827

Hideto Koizumi (Contact Author)

University of Pennsylvania - Business & Public Policy Department ( email )

3641 Locust Walk
Philadelphia, PA 19104-6372
United States

HOME PAGE: http://https://bepp.wharton.upenn.edu/profile/hkoizumi/

Hitotsubashi University ( email )

2-1 Naka Kunitachi-shi
Tokyo 186-8601

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics