Nicholas Bloom | The Impact of AI on Productivity

Why It Matters

Uncertain AI‑driven productivity gains could reshape inflation dynamics, making executive‑sourced data essential for central banks’ monetary‑policy strategies.

Key Takeaways

• AI productivity forecasts range from 0.1% to 20% annually
• Survey of 6,000 global executives reveals mixed AI adoption levels
• Firm AI usage low in census, higher in payment data
• Productivity growth has declined steadily since the 1950s, challenging optimism
• Executive insights will inform central banks' monetary policy decisions on AI

Summary

Nick Bloom, Stanford economist, presented a data‑driven assessment of artificial intelligence’s effect on productivity at a San Francisco Fed event. He emphasized that the analysis relies on a large‑scale survey of roughly 6,000 CFOs and CEOs across four countries, rather than theoretical speculation, to gauge how firms are actually deploying AI tools.

Bloom highlighted the stark contrast between historical trends—U.S. labor productivity has been falling from about 3‑4% in the 1950s to near‑zero growth today—and the wildly divergent forecasts for AI, ranging from a modest 0.1% boost to a transformative 10‑20% annual increase. The survey revealed that while individual usage sits at 30‑40%, firm‑level adoption appears low in official census data (around 17%) but much higher in payment‑processor metrics (approaching 50%).

He noted that political affiliation influences AI usage, with Democrats more likely to adopt the technology, and cited anecdotal evidence such as RAMP’s transaction‑based AI adoption figures. Bloom also described the rigorous two‑step recruitment process—phone verification followed by monthly email panels—that yields a roughly 30% response rate from senior executives, ensuring the data’s credibility.

The findings carry significant policy implications: central banks need reliable measures of AI‑driven productivity to calibrate inflation forecasts and interest‑rate decisions. Bloom’s 50/50 assessment—AI could either revive growth or continue the long‑term slowdown—underscores the urgency for better firm‑level data and ongoing monitoring as the technology matures.

Original Description

Please join us on March 24 for a live presentation on the productivity economics of AI by Nicholas Bloom, the William D. Eberle professor of economics at Stanford University. Professor Bloom will explore AI use at the firm-level and its impacts on employment and productivity using new research from surveys of over 5,000 CFOs, CEOs, and executives across the US, UK, Germany, and Australia.

Following his presentation, Professor Bloom will answer live and pre-submitted questions with our host moderator, Huiyu Li, co-head of the EmergingTech Economic Research Network (EERN) and research advisor at the Federal Reserve Bank of San Francisco.

This is a virtual event hosted by the EmergingTech Economic Research Network (EERN). It is open to everyone and will be livestreamed on this page. A recording will be available after the event. We invite you to register and submit a question (https://www.frbsfevents.org/event/EERN-2026March24/register).