Updated
Updated · The Budget Lab at Yale · Jul 20
Budget Lab Models AI Effects on $6,595 Billion 2030 Revenue Base
Updated
Updated · The Budget Lab at Yale · Jul 20

Budget Lab Models AI Effects on $6,595 Billion 2030 Revenue Base

2 articles · Updated · The Budget Lab at Yale · Jul 20

Summary

  • The Budget Lab released a model estimating how AI could change the tax system and the federal revenue-to-GDP ratio, using CBO’s $6,595 billion FY2030 revenue baseline.
  • The analysis finds the near-term net effect hinges on adoption speed: AI could make IRS audits cheaper and more effective, but taxpayers could also use AI to uncover new avoidance strategies.
  • Its framework applies Karger et al. 2026 AI-driven GDP growth scenarios through the Budget Lab’s small macro model, then layers the implied revenue-to-GDP change onto the baseline; the setup has little to no effect on 2030 GDP growth.
  • The model simplifies several channels—including labor-income dispersion, capital income, retirement distributions and corporate tax assumptions—and the authors say a more comprehensive version is planned.
  • GitHub code and a full methodology were published with the study, which says longer-run simulations would likely show larger effects on inequality than the current 2030-focused exercise.

Insights

Will AI's massive energy demands and hidden costs erase its projected boost to federal tax revenue?
As AI powers both tax collection and tax avoidance, who is ultimately winning the high-stakes revenue race?