Economists Expect Smarter AI by 2030 But Forecast Little GDP Impact
Summary
- • Economists give 61% odds to significant AI progress by 2030, yet GDP and productivity forecasts stay near historical trends
- • Multi-institution survey uncovers a striking paradox between AI capability expectations and economic impact forecasts
- • Labor force participation projected to fall to 58% by 2050 — a level not seen since the 1960s
- • Survey spans 69 economists, 52 AI experts, 38 superforecasters, and 401 general public members
Details
Largest expert forecasting study on AI's economic impact
Forecasting Research Institute, with coauthors from Federal Reserve Bank of Chicago, Yale School of Management, Stanford University, and University of Pennsylvania, surveyed four groups from October 2025 to February 2026: 69 leading economists, 52 AI industry and policy experts, 38 calibrated superforecasters, and 401 general public members.
Core paradox: expect capable AI, not transformative GDP
Economists assign 61% combined probability to moderate or rapid AI progress by 2030, yet their unconditional forecasts for US GDP growth, total factor productivity, and labor force participation cluster near historical trend lines — not acceleration.
47% probability on moderate AI scenario by 2030
Moderate scenario defined as: AI runs semi-autonomous research labs, completes five-day coding tasks, produces high-quality novels, manages complex business roles, and robots navigate home environments. Combined with the 14% rapid-scenario probability, economists see a 61% chance of substantial capability gains within five years.
Labor force participation forecast at 58% by 2050
A rate not seen since the 1960s and well below CBO projections for 2050, suggesting economists anticipate meaningful workforce displacement even without a corresponding GDP surge. This is a key tension: capability gains may not translate to broad economic gains.
Job retraining is the top-rated policy intervention
Retraining programs projected to add approximately 2.76 million workers to the labor force (+1 percentage point participation) and boost GDP growth by 0.2 percentage points in a rapid-progress scenario. Economists broadly support this as a response to AI displacement.
Expert disagreement remains substantial
Some respondents forecast transformative upheaval and large-scale job losses; others expect only modest productivity improvements. Only 14% probability was assigned to the most extreme scenario where AI outperforms all humans and robots handle nearly all physical labor by 2030.
Research: study methodology; Insight: core analytical finding; Stat: quantitative data point; Market Impact: economic/labor effects; Policy: policy response analysis; Context: background and disagreement framing
What This Means
The expert paradox — expecting powerful AI but not powerful GDP growth — suggests AI's productivity gains may diffuse slowly through the economy, or that structural factors like inequality, displacement, and transition costs will offset headline gains. For businesses and policymakers, this is a signal to plan for significant labor market disruption even if aggregate GDP numbers appear stable, and to invest in workforce transition infrastructure now. The wide expert disagreement underscores that AI's economic trajectory remains genuinely uncertain, making scenario-aware policy planning more critical than betting on any single forecast.