Give Workers a Fighting Chance por Daron Acemoglu publicado Project Syndicate (2/2021).
“At a time when businesses are awash in technologies capable of replacing human labor, the US tax code is encouraging them to embrace excessive levels of automation – even to the point where it is no longer efficient to do so. Rarely before has the deck been stacked so fully against workers.
CAMBRIDGE – The first months of US President Joe Biden’s administration will be defined by the efforts to contain COVID-19 and deliver vaccinations on a mass scale. Over the medium term, however, the economy will determine the administration’s success. Here, Biden has indicated that tax reform will be a high priority, and he has released plans to address long-running fiscal problems such as federal government revenue shortfalls and the tax system’s loss of progressivity. But these proposals do not yet go far enough to address a major fault line in the tax code: the excessively favorable treatment of capital income (profits and returns on financial assets and savings)…”
La revolución no tiene por qué ser automatizada por Daron Acemoglu y Pascual Restrepo publicado por Project Syndicate (3/2019)
“BOSTON – La inteligencia artificial está transformando cada aspecto de nuestras vidas, sobre todo la economía. Por tratarse de una tecnología de uso general, las aplicaciones de IA son potencialmente infinitas. Si bien se la puede utilizar para automatizar tareas que anteriormente eran realizadas por personas, también puede hacer que la mano de obra humana sea más productiva, aumentando así la demanda laboral…”
Does the U.S. Tax Code Favor Automation? by Daron Acemoglu, Andrea Manera, Pascual Restrepo published by Brookings Papers (3/2020).
We argue that because the US system is distorted against labor and in favor of capital and has become more so in recent times, it has promoted levels of automation beyond what is socially desirable. Moving from the US tax system and level of automation in the 2010s to optimal taxation of factors and corresponding optimal level of automation would raise employment by 5.85% and the labor share by 0.53 percentage points. If moving to optimal policy is not feasible, more modest reforms can still increase employment by 1.35-2.31%. Interestingly, if only partial reforms are feasible, our theoretical framework and quantitative work show that it would not be desirable to increase taxation of capital per se (even though capital is lightly taxed in the US); rather, directly reducing the extent of automation would be much more e↵ective. This is because marginal automated tasks do not bring much productivity gains and displace workers, reducing employment. In contrast, increasing the capital intensity of already-automated tasks raises the demand for labor because of the complementarity between tasks. These conclusions are reinforced when technology and/or human capital investments are endogenous and respond to tax policies.
