NBER Study: Five Largest Tech Firms' AI Spending Implies 5-58% Additional GDP Growth by 2030
Key Takeaways
- ▸Big Tech's projected $760B annual AI capex by 2026 implies an expected 2.7x productivity multiplier in the AI sector to avoid bankrupting the firms
- ▸Credible AI boom scenarios would deliver 5–58% additional GDP growth by 2030, with highly uncertain long-term growth centered at ~7% annually
- ▸Markets are pricing in meaningful risk premia (3 percentage point equity premium increase) due to the binary nature of AI productivity realization
Summary
A new National Bureau of Economic Research working paper analyzes the massive capital expenditure commitments announced by the five largest U.S. technology firms—totaling $380 billion in 2025 and potentially doubling to ~$760 billion in 2026—and what these investments imply about expected AI productivity breakthroughs. Using a two-sector open-economy model with rare productivity booms, researchers Jessica and Jonathan Wachter calibrate the spending levels to imply an AI-sector productivity multiplier of approximately 2.7x, with a 50% annual probability window through 2027 for gains of similar magnitude.
The analysis generates a wide range of scenarios consistent with industry forecasts, with estimated cumulative additional GDP growth ranging from 5 to 58 percentage points by 2030, and AI's share of the economy potentially expanding from 8% to 39%. Long-term annual growth expectations average approximately 7%, though with substantial downside risk. The study also quantifies financial market implications: with standard risk preferences (risk aversion of 3 and elasticity of intertemporal substitution equal to 1), the risk-free rate could increase ~0.5 percentage points and the equity risk premium could widen by ~3 percentage points, reflecting investor skepticism about achieving the required productivity gains.
- AI could represent 8–39% of the economy by 2030 depending on whether the productivity boom materializes as anticipated by current investment levels
Editorial Opinion
This research provides crucial macro-economic grounding for the AI investment thesis: it shows that current capex levels are defensible only if AI delivers the extraordinary productivity gains its investors expect. The wide range of outcomes (5–58% additional GDP) reflects genuine uncertainty about whether the technology will deliver. For policymakers and investors, the implication is clear—we're in a bet-the-company moment for Big Tech, and the equity risk premium widening suggests sophisticated investors remain skeptical that AI will fully justify the spending. The next 18–24 months will be decisive.
