Peter John Lambert (London School of Economics and Political Science)
Tito Boeri, Bernardo Mottironi and Paolo Naticchioni
This paper examines the causal impact of the post-pandemic transition to remote work on labor market power, connecting to an extensive literature which highlights distaste for commuting as a key source of rent extraction for local employers. Our analysis combines three sources of microdata: (i) millions of job vacancy postings to derive an index of remote work using a frontier large language model; (ii) worker-level administrative records on the workplace and occupational group; and (iii) firm-level production data to estimate wage markdowns. Exploiting the heterogeneity in shifts to remote work across occupations --- instrumented with shifts observed abroad --- and the heterogeneity in occupation mixes across local labor markets, we implement a shift-share design and identify a substantial reduction in employer rents. Our counterfactual analysis quantifies the impact of the 2019-2023 transition to remote work in one-fifth of the deviation from perfect competition.