Short‑term measured productivity jumps can be mechanically inflated by non‑AI forces — for example, removing lower‑productivity immigrant workers from the labor force or surges in capital utilization from front‑loaded AI and data‑center investment. That makes it hard to attribute single‑year productivity revisions to AI without decomposing demographic and capital‑utilization effects.
— If policymakers misattribute productivity gains to AI when they actually reflect compositional shifts or investment timing, they may adopt the wrong labor, immigration, and industrial policies.
Tyler Cowen
2026.03.06
75% relevant
The article reports a causal estimate that immigration increases local innovation and wages (a productivity proxy) and that post‑1965 immigration may have raised U.S. innovation and wages by ~5%, which directly links immigration flows to productivity and investment outcomes in the matched idea.
Noah Smith
2026.02.27
100% relevant
Noah Smith cites Martha Gimbel’s critique that immigration policy and elevated capital utilization (investment in AI infrastructure) could explain 2025 productivity revisions instead of genuine worker augmentation by AI.
← Back to All Ideas