BlockBeats report, June 17: Guolian People's Securities released a research report noting that although the market hopes AI-driven productivity gains will alleviate U.S. debt pressures, historical experience and current realities suggest AI is unlikely to replicate the debt-reduction miracles of the post-WWII or Clinton eras in the short term, making the U.S. debt dilemma difficult to overcome in the near term. By the end of 2025, the U.S. national debt is projected to approach $38 trillion, with net interest payments nearing $1 trillion.
The research report outlines three pathways to reduce the debt-to-GDP ratio: lowering interest rates, boosting economic growth, and reducing the fiscal deficit. Historically, the United States successfully reduced its debt burden in two phases: from 1946 to 1974, it relied on post-war high growth and technological transformation, bringing the debt ratio down from over 100% to approximately 20% over three decades; in the 1990s, it achieved an average primary budget surplus of about 3.2% annually between 1996 and 2001, driven by the internet revolution and fiscal discipline under the Clinton administration.
However, the debt-reduction effectiveness of this AI wave faces two practical constraints. First, the productivity gains from AI are subject to significant lags; according to estimates from the University of Pennsylvania, AI will only boost total factor productivity by 0.05 to 0.1 percentage points between 2026 and 2027, with its contribution rising to approximately 0.2 percentage points only in the early 2030s—far too modest to offset current fiscal pressures. Second, AI accelerates the concentration of factor returns toward capital, systematically eroding the tax base. In the United States, personal income tax and payroll taxes together account for about 85% of federal revenue; AI-driven labor substitution and wage suppression directly threaten this primary revenue source. Corporate income tax, by contrast, contributes only about 10% and is subject to a flat 21% rate, compounded by the cross-border tax avoidance capabilities of major tech firms, making it insufficient to close the gap left by declining personal income tax revenues—creating a paradox where “the more technologically prosperous the economy becomes, the more its tax base withers.”
The report suggests that potential solutions include increasing capital gains and high-income tax rates, imposing a "digital factor tax" on commercial revenues generated by large AI models, and exploring a "robot tax" to subsidize those affected by technological unemployment. However, these measures all face structural challenges, such as difficulties in taxing cross-border flows of AI inputs, the strong political influence of tech giants, and the risk that unilateral tax increases could stifle innovation. The report concludes that fiscal and tax adjustments in the AI era are destined to be a prolonged institutional struggle, and that the U.S. debt problem remains a major, insurmountable obstacle to the U.S. economy in the short term.