Thoughts on the US tax system

Sep 30, 2026 | Accounting and bookkeeping, General interest tax news

Suggested MetroTax.net addendum

Below is publication-ready copy that credits the author and source, distinguishes the Economist commentary from U.S.-specific analysis, and frames the issue for accountants without reproducing the original article verbatim.

Editorial context and attribution: This informational addendum responds to themes raised in “What is the right tax system for the 21st century?” by Edward Carr, Deputy Editor of The Economist, published in connection with The Insider. The underlying article considers how fiscal pressures, increasingly complex tax codes, and artificial intelligence may challenge tax systems built primarily around labor income. This addendum is an independent U.S. tax-policy discussion prepared for accounting and tax professionals; it is not affiliated with or endorsed by Edward Carr or The Economist.

What Is the Right U.S. Tax System for the 21st Century?

The United States does not need a single replacement tax to meet the economic and fiscal demands of the 21st century. It needs a more coherent tax mix: one that raises adequate revenue with fewer distortions, treats comparable economic income more consistently, preserves progressivity, and does not depend excessively on wages at a time when artificial intelligence may alter the distribution of income between labor and capital.

That conclusion matters because the present federal system is still heavily reliant on individuals’ earnings. For fiscal year 2026, the Congressional Budget Office projects approximately $2.8 trillion in individual income-tax receipts and $1.8 trillion in payroll-tax receipts, compared with $404 billion in corporate income-tax receipts. In other words, the two principal taxes linked to household earnings and employment represent the dominant share of projected federal tax revenue.

The policy challenge is not merely whether AI will eliminate jobs. It is whether AI-led productivity gains—and the returns generated by data, software, intellectual property, computing infrastructure, and highly scalable businesses—will accrue disproportionately to owners of capital while the tax system continues to rely primarily on wages and payrolls. The answer is uncertain, but the question is no longer theoretical.

The present U.S. system

The current federal system combines several revenue instruments:

Federal revenue source	Principal base	Core function
Individual income tax	Wages, business income, investment income, and other taxable income	General federal revenue; progressive redistribution
Payroll taxes	Employment wages and self-employment earnings, subject to statutory rules	Social Security, Medicare, and unemployment-related programs
Corporate income tax	Taxable corporate profits	Taxation of corporate-level income
Excise taxes	Specified goods, services, and activities	Targeted revenue and, in some cases, behavior-shaping policy
Customs duties	Imported goods	Revenue, trade policy, and industrial-policy effects
Estate and gift taxes	Transfers of wealth	Limits tax-free intergenerational wealth transfer

The Internal Revenue Service identifies income tax, estimated tax, self-employment tax, employment tax, and excise tax among the principal categories businesses must navigate. Employment taxes include Social Security and Medicare taxes, federal income-tax withholding, and federal unemployment tax.

This structure has strengths. Payroll withholding is administratively efficient; progressive income-tax rates can reflect ability to pay; corporate taxation reaches some income before it is distributed to owners; and targeted excises can address costs not reflected in market prices. But the system also contains overlapping tax bases, exclusions, deductions, preferential rates, phaseouts, credits, timing differences, and entity-choice incentives that make equivalent economic activity subject to materially different tax outcomes.

Treasury describes tax expenditures as revenue losses arising from exclusions, exemptions, deductions, credits, preferential rates, or deferrals embedded in federal tax law. Such provisions may advance legitimate social or economic policy goals, but they also complicate the code and can obscure the effective subsidy delivered through the tax system.

AI changes the tax-policy question

AI may increase productivity, profits, taxable economic activity, and ultimately government revenue. It could also improve tax administration if the IRS uses it to identify noncompliance, allocate audit resources more effectively, and assist with complex data analysis. Yet it may simultaneously weaken the wage-based component of the tax base if it reduces labor demand, suppresses wage growth in affected occupations, or shifts income toward capital owners.

The Congressional Budget Office emphasizes that AI’s fiscal effects are uncertain and depend on how technology affects productivity, labor demand, income distribution, and compliance. Workers made more productive by AI may earn more and generate additional income- and payroll-tax receipts. Conversely, workers who are displaced permanently or move into lower-paying work could reduce those receipts. CBO also notes that AI could raise compliance revenue if it strengthens IRS audit capability—or reduce revenue if taxpayers use AI to evade or reduce tax liabilities.

A recent Yale Budget Lab analysis reaches a similar broad conclusion: stronger AI-driven economic growth could raise federal revenue, but revenue gains may be diminished if income moves from labor to capital, particularly because capital income is often taxed at lower rates and some forms of economic gain are excluded or deferred. In one rapid-growth scenario, the analysis estimated 2030 revenue gains of up to $216 billion relative to the CBO 2026 baseline, while cautioning that the distribution of income between capital and labor materially affects the result.

For tax professionals, the practical implication is clear: AI is not simply an automation issue. It is a tax-base issue. If a larger fraction of national income arises from capital-intensive production, intellectual property, platform economics, data assets, and highly concentrated ownership, policymakers will face increasing pressure to reconsider what income is taxed, when it is taxed, and through which tax base public revenue is collected.

A better direction: rebalance, do not simply raise rates

A durable U.S. approach should not begin with a symbolic “robot tax” or a broad, untested levy on computing power. Those measures are difficult to define, could penalize productivity-enhancing investment, and may be particularly hard to administer across cloud services, embedded AI software, shared computing capacity, and multinational supply chains.

A better framework would rebalance the system across labor, consumption, capital, land, and environmental externalities.

Protect the labor tax base, but reduce its inequities. Payroll taxes provide stable funding for social-insurance programs, but they are concentrated on employment earnings and can impose a comparatively heavier burden on workers than on some forms of capital income. Reform should preserve the financing integrity of Social Security and Medicare while examining wage-base rules, treatment of self-employment income, and the interaction of payroll taxes with credits and benefit formulas.
Broaden the income-tax base before increasing headline rates. A simpler code with fewer narrowly tailored exclusions, deductions, and preferential provisions can raise revenue more transparently than relying solely on higher statutory rates. The relevant question is not whether every tax preference should disappear, but whether each preference has a demonstrable public purpose that justifies its revenue cost and administrative complexity.
Improve the taxation of capital income. If AI increases returns to intellectual property, corporate equity, business ownership, and other capital assets, the system should reduce opportunities for similarly situated economic gains to receive sharply different tax treatment merely because of legal form, realization timing, or access to sophisticated planning. This does not require treating every form of capital income identically, but it does require a more principled relationship between tax burdens on labor income and on high-end capital income.
Consider a broader consumption tax, with household protections. A federal value-added tax or another broad consumption-based tax could diversify federal revenue away from wages. However, consumption taxes tend to be regressive relative to annual income unless paired with refundable credits, rebates, or other protections for lower- and middle-income households. Brookings has argued that, if AI makes labor-income taxation less viable, consumption taxes may need to assume a larger role, while warning against prematurely taxing AI capital assets in ways that distort needed infrastructure investment.
Use targeted taxes where social costs are clear. Carbon-related taxes, congestion charges, pollution fees, and carefully designed excise taxes can both raise revenue and address costs imposed on others. These are not substitutes for broad fiscal reform, but they can improve economic efficiency when designed with transparent use of proceeds and appropriate household offsets.
Reassess land and location-based taxation. Land is immobile, limited in supply, and less susceptible to offshore tax planning than labor income, corporate profits, or intangible assets. Although property taxation is primarily state and local, more serious consideration of land-value principles could help finance local infrastructure and public services while reducing penalties on productive investment in buildings and improvements.
Strengthen administration and compliance. A simpler tax code, improved information reporting, modernized IRS systems, reliable taxpayer service, and rigorous protections for taxpayer rights could increase voluntary compliance without inventing entirely new tax bases. The administrative capacity to enforce a tax system is as important as the rate structure written into law.

What accountants should watch

For accountants, this debate has immediate relevance even if major legislative reform remains uncertain.

The distinction between wage income, self-employment income, business income, capital gains, dividends, and retained corporate earnings will remain central to tax planning and policy analysis.
Entity selection, compensation design, pass-through planning, basis management, and the timing of realization events will remain subject to increased scrutiny where they produce substantially different outcomes for economically similar income.
AI adoption may affect not only staffing and service delivery within accounting firms, but also client classifications of expenditures, capitalization of software and technology costs, research activities, data governance, intellectual-property valuation, and automated compliance processes.
Businesses should expect more attention to whether tax rules adequately capture returns generated by intangible assets, digital business models, cross-border operations, and increasingly automated production.
Any shift toward consumption, carbon, land, or other alternative tax bases would create important state-and-local conformity, nexus, pricing, accounting-system, and taxpayer-communication issues.

Bottom line

The right tax system for the 21st-century United States is not one that taxes “robots” as a slogan or assumes AI will inevitably destroy the revenue base. It is one that can fund public commitments even if income becomes less tied to traditional employment.

The most defensible reform path is a balanced one: simplify and broaden existing income-tax bases; preserve progressive treatment based on ability to pay; reduce unwarranted disparities between labor and capital taxation; use well-designed consumption and externality taxes to diversify revenue; and modernize tax administration without compromising due process or taxpayer privacy.

AI may expand the economy, but economic growth alone does not guarantee a stable or equitable revenue system. The distribution of that growth—between workers, investors, firms, and owners of valuable assets—will determine whether the existing U.S. framework remains adequate.