A Reagan-era promise returns: Trump sells AI growth instead of self-funding tax cuts

Forty-five years ago, Ronald Reagan's administration cut the top federal income tax rate from 70% to 50%, promising that faster growth would make up the lost revenue. It did not: over Reagan's two terms, US gross national debt roughly tripled. Every subsequent wave of Republican tax cuts ended the same way, with a wider deficit. Now, according to a Guardian column dated 11 October 2026, the promise has a new engine: artificial intelligence. The author predicts it will fail again.

The portion of the text available to us states the main thesis: even if AI sharply accelerates the economy, its impact on public finances will be muted. We do not see the full argument, so what follows is editorial analysis, not a summary of the author's case.

How the old argument works and where it breaks down

The "taxes pay for themselves" logic rests on the Laffer curve: at a sufficiently high rate, cutting it expands the tax base enough that revenue rises. In theory, that is possible. The problem is that US rates have long been below the point where the effect would work. That is why the cuts of 1981, of 2001–2003, and of the 2017 law, which lowered the corporate rate from 35% to 21%, produced growth but did not cover the lost revenue.

AI adds a new variable to the scheme. If the technology raises productivity, GDP will grow, and in theory so will the tax base. Proponents of this scenario expect the boom to lift private investment, corporate profits and employment in adjacent industries.

Why an AI boom is not a tax boom

Several mechanisms can keep the gains from reaching the Treasury:

• Income structure. If the benefits of AI go mainly to owners of capital, which is taxed at lower rates than wages, revenue grows more slowly than GDP.
• Labor market. Income tax and social insurance contributions remain the main source of federal revenue. Automation that shrinks the wage bill hits exactly those.
• Tax breaks on capital investment. Data centers and chips are bought with accelerated expensing, so an investment boom reduces taxable profit for years.
• Deflation. If AI makes services cheaper, the nominal base that taxes are levied on grows more slowly than real output.
• Scale of the deficit. To pay for a large tax cut, AI would have to deliver not a modest boost to growth but a sustained, multi-year acceleration, which no one has yet recorded.

That is the substance of the "muted effect" formula: even a successful AI scenario does not guarantee a successful budget outcome.

Why this matters for the tech industry

AI is becoming an argument in fiscal policy. Growth expectations tied to generative models, agents and infrastructure are now priced not only into company valuations but also into the justifications for tax decisions. If the bet on AI-driven growth fails, or pays off without a tax return, the budget will not be the only one to pay. The political risk will fall on the sector itself: disappointment with the "AI dividend" usually turns into tighter regulation and a search for new revenue sources, including among companies that benefited from the boom.

What readers should do with this

Companies investing in AI infrastructure should build tax scenarios into their models alongside technological ones: changes to capital investment breaks, corporate income taxes and possible levies on automation. Investors should separate two claims: "AI will accelerate the economy" and "the government will get money from it." The first can come true without the second. And when reading political forecasts of AI-driven growth, check the numbers against history: since Reagan, tax cuts have never fully paid for themselves, and the new version of the promise has no data yet that would set it apart from the earlier ones.