RESEARCh
RESEARCh
with Shubhdeep Deb, Jan Eeckhout and Larry Warren
Econometrica. Vol 92 (3), 603-636, May 2024. DOI. Draft
Comments: Gianluca Violante, John Van Reenan, Response
with Shubhdeep Deb, Jan Eeckhout and Larry Warren
Journal of the European Economic Association. 20 (6), 2181-2225, December 2022. DOI. Draft.
with Nicolò Dalvit and Joanne Tan
[LATEST VERSION] | [SSRN] | [Submitted]
Abstract: I develop and estimate a quantitative general equilibrium model of oligopsonistic labour markets with endogenous task-based automation, and provide conditions for a unique equilibrium and global identification of its parameters. Relative to an economy without wage-setting power, more productive firms employ too few workers and automate too many tasks. These distortions are larger when labour's comparative advantage rises less steeply across the tasks it still performs. The same schedule determines whether cheaper automation strengthens or weakens firms' wage-setting power. In line with the theory, French manufacturing firms with greater labour market power hold more automation capital. Across three specifications of labour's comparative advantage, removing labour market power raises welfare by 6.3–9.0% of consumption. A benchmark without automation understates this gain by 10–24%. Cheaper automation delivers almost all its gains to firm owners and raises the welfare cost of labour market power by 3–4% of its initial level.
with Zsofia Barany and Christian Siegal
[Draft] | [CEPR Discussion Paper, DP21700] | [VoxEU]
Abstract: A prominent explanation for declining labor shares is automation. We test this mechanism directly at the firm level, drawing on a central prediction of the task-based model of production: firms that automate more should see larger declines in routine employment and labor share. Using administrative data from France between 1994 and 2019, we find no support for this prediction. Changes in routine employment share are either uncorrelated or even negatively correlated with changes in labor share within firms. Our evidence shows that automation is not driving labor share dynamics in France.
I was honored with the "Best Young Labor Economist" prize by the European Association of Labour Economist for this paper.
A small Twitter thread on the paper!
Lo Bello and Pesaresi