A speech delivered by Federal Reserve Board Governor Michael S. Barr at the Reykjavík Economic Conference on May 9, 2025. It was the first formal Federal Reserve Board speech dedicated to artificial intelligence and labor markets. Barr framed AI's labor-market impact through two contrasting scenarios, set out four indicators the Federal Reserve would track to gauge which scenario was unfolding, and noted implications for two monetary-policy reference rates. Barr emphasized substantial uncertainty about AI's trajectory and did not endorse either scenario.
Speaker: Gov. Michael S. Barr, Federal Reserve Board Venue: Reykjavík Economic Conference Date: May 9, 2025
Two scenarios
Barr structured the labor-market question around two scenarios.
In the first, incremental progress, AI primarily automates specific tasks within occupations rather than replacing entire jobs. Workers shift their responsibilities while remaining employed — Barr's examples included programmers using AI coding assistants and lawyers using AI legal-research tools. Barr cited a Federal Reserve Bank of New York survey for the observation that "many businesses plan to retrain their workers to use AI rather than laying them off." New occupations emerge around AI management and implementation, real wages improve through productivity gains, and employment remains robust, with some worker dislocation that is absorbed.
In the second, transformation, AI handles a broad spectrum of cognitive and physical work, and most existing occupations could face automation. Some roles persist — Barr named doctors, teachers, and judges — because of consumer preference or regulation rather than because AI cannot perform them. In this scenario unemployment could rise significantly as displaced workers struggle with skill obsolescence, and income concentration could worsen if capital owners capture a disproportionate share of the gains. Barr framed expanded economic resources that could address societal challenges as the upside counterweight to these risks.
Four monitoring indicators
Barr outlined four indicators the Federal Reserve would track to assess which scenario was materializing:
- Business AI adoption rates and implementation patterns
- AI capability benchmarks against human performance
- Job openings data and advertised skill requirements
- Occupational employment growth statistics
Monetary-policy implications
Barr identified two considerations for the Federal Reserve. The natural rate of unemployment (u\) may need reassessment as labor demand shifts across sectors. The neutral interest rate (r\) may rise if productivity growth accelerates substantially.
Provenance and positioning
The speech was the first formal Federal Reserve Board speech dedicated to AI and labor markets, and it indicated that the Federal Reserve was monitoring AI's macroeconomic implications, with potential bearing on forward guidance for interest rates. Barr's two-scenario framing was institutionally neutral and did not endorse either the slow-diffusion or the transformation view. The incremental-progress scenario aligns with the slow-diffusion case in The Simple Macroeconomics of AI (Acemoglu) and Why I Think AI Take-Off Is Relatively Slow (Cowen), while the transformation scenario aligns with The 2028 Global Intelligence Crisis (Citrini).
Relationships
- supports: AI Labor Disruption, Labor Disruption Timelines: Who Predicts What and Why, planned AI Bubble Debate.
- related: The Simple Macroeconomics of AI (Scenario 1 anchor), The 2028 Global Intelligence Crisis (Scenario 2 anchor), Why I Think AI Take-Off Is Relatively Slow (Cowen).