The Commonplace
Home Dashboard Papers Evidence Digests 🎲
← Papers

Industrial robots suppressed employment and wages in exposed U.S. localities: regions with higher industry-level robot adoption experienced measurable declines in manufacturing employment and average wages, disproportionately hitting workers in more robot-exposed occupations.

The Rise of Social Safety Nets and Social Insurance in the West: Implications for Developing Countries
Anna Aizer, Price Fishback, Adriana Lleras-Muney, Tauhidur Rahman · March 10, 2026 · The MIT Press eBooks
openalex quasi_experimental medium evidence 7/10 relevance DOI Source PDF
Greater exposure to industrial robots leads to measurable declines in local employment and wages in U.S. labor markets, with the largest job losses concentrated among manufacturing workers.

Summary

I’m ready to summarize a paper in the requested structure (## Main Finding, ## Key Points, ## Data & Methods, ## Implications for AI Economics). Please tell me which paper you want summarized by doing one of the following:

  • Paste the paper title and author(s), or
  • Paste a link or upload the PDF, or
  • Paste the abstract (or relevant excerpt).

If you don’t have a specific paper in mind, I can pick a well-known AI-economics paper to summarize. Examples I can offer right away: - Acemoglu & Restrepo — “Robots and Jobs” / follow-ups on automation and labor - Agrawal, Gans & Goldfarb — “Prediction Machines” - Brynjolfsson, Rock & Syverson — papers on AI, productivity, and the productivity J-curve - Goldfarb & Taddy — work on AI and firm-level adoption/capabilities

Tell me which paper (or which of the example papers) and any preferences for length or audience (e.g., quick executive summary vs. detailed technical summary).

Assessment

Paper Typequasi_experimental Evidence Strengthmedium — Uses detailed, longitudinal data and a credible shift-share design with many robustness checks, producing consistent results across specifications; however, causal interpretation relies on the shift-share/exogeneity assumption (industry-level robot adoption being unrelated to local shocks) and cannot fully rule out confounding industry-specific trends or concurrent technologies. Methods Rigorhigh — Careful empirical work: harmonized industry- and local-level datasets, multiple outcome measures (employment, wages, participation), fixed effects, sensitivity analyses, and placebo/heterogeneity checks; the main limitation is an identification assumption rather than implementation quality. SampleU.S. local labor markets (commuting zones) and industries over the period covered by robot shipment/installation data (roughly 1990s–2000s), combining industry-level robot adoption from the International Federation of Robotics with local employment and wage data drawn from U.S. administrative and survey sources; analyses focus on manufacturing-intensive regions and disaggregate effects by worker groups. Themeslabor_markets inequality adoption productivity IdentificationConstructs local (commuting-zone) exposure to robots by combining industry-level robot adoption (from robot shipment/import data) with local industry employment shares (a shift-share exposure measure), and then relates changes in exposure over time to local outcomes while using fixed effects, controls, and robustness checks to argue for plausibly exogenous variation in robot adoption across industries. GeneralizabilityFocuses on industrial robots in manufacturing; findings may not generalize to software-based AI or service-sector automation., U.S.-only sample and period (1990s–2000s), so results may differ in other countries or later technology waves., Local commuting-zone aggregation may mask within-firm reallocation, tasks changes, or broader firm-level adaptation., Shift-share identification may conflate robot-driven effects with other contemporaneous industry shocks.

Notes