Technological Unemployment, Redistribution and Social Welfare

Abstract

We develop and analyze a dynamic general equilibrium model of perpetual economic growth with endogenous automation. Our model uses a task-based framework, where robotic capital can replace human workers that perform routine and manual tasks. Using a calibrated version of the model for the United States, we explore the effects of robot taxes, income redistribution, and minimum wage policies in maximizing utilitarian social welfare. The novelty of our paper lies in two features of the model: First, we explicitly examine whether a minimum wage is binding for workers facing automation risk within a dynamic general equilibrium model. Second, instead of assuming that households earn both capital and labor income, we incorporate a class structure in which workers facing automation risk do not own capital. Theoretically, we show that the model has a unique balanced growth path, and we provide closed-form solutions for threshold tax rates and minimum wage levels that fully block automation. Quantitatively, we present a local analysis of policies around the benchmark balanced growth path, and we implement global searches for optimal policy mixes. These analyses show that a benevolent and fully informed government can maximize utilitarian social welfare in the long run with zero technological unemployment and without a binding minimum wage. Specifically, optimal policies feature perfectly egalitarian consumption outcomes as a result of redistribution through high capital income tax rates. Our results also show that, if income tax reforms are not feasible, a small robot tax rate can maximize welfare with positive technological unemployment.

Keywords: endogenous automation, fiscal policy, income redistribution, robot tax. JEL Codes: D33, E64, I31, I38

JEL Classification: D33, E64, I31, I38