Welfare Effects of Automation with Redistributive Taxes

Abstract

Automation shifts income from workers to capital owners and raises income inequality. Can we turn automation into a welfare gain by adjusting the tax and transfer system or is it desirable to restrict automation? We study this question in a setting where capital can move across borders and governments set capital taxes strategically. We derive a sufficient statistics formula for the welfare effects of automation under optimal taxes and assess it quantitatively, showing that negative welfare effects of automation are plausible. Yet, seizing the implied welfare gains requires globally coordinated restrictions on automation, as unilateral policies prove ineffective. The optimal global policy, however, is a cooperative capital tax policy and we show that the welfare gains from such cooperation rise with automation. Joint paper with Hayato Kato)