Preserving Fiscal Stability in the Age of Transformative AI
Transformative AI will not only disrupt the labor market, but also the foundations of fiscal stability. As value creation shifts from labor to capital, Anton Korinek and Lee Lockwood explain, traditional tax systems, built on taxing the former, will no longer be fit for purpose. Instead, they set out a framework for an alternative tax system, one that is built for the economy of the future, an age of TAI, rather than the one of the past, an age where human labor sat at the center of economic life.
The Coming Revenue Crisis
Imagine rapid advances in artificial intelligence generating significant disruption in labor markets, with unemployment rising across multiple sectors in the coming years. In the past, job disruption has always gone hand in hand with new job creation—farm workers became factory workers, factory workers became service workers, and manual laborers became knowledge workers. But this time, unlike earlier technological disruptions, many of the new jobs that would otherwise have been created don’t materialize, because they too can be performed by AI.
Now imagine the position of the US Treasury in this scenario. Tax revenue has declined, even as outlays have increased sharply. The newly unemployed need support from the social safety net—programs they paid into for decades while working. Meanwhile, the economy hasn’t collapsed; in fact, total output has grown. Jobs previously done by workers are now performed by AI systems so efficiently that they more than compensate for lost human labor.
Here’s the fiscal problem: When $1 of value creation shifts from labor to capital, total tax revenue falls on the order of 10–15¢. This is because the average effective tax rate on labor income in the United States is roughly 30%,1 whereas the average effective rate on capital income is approximately 15–20%—and often lower for corporate income that can be deferred or retained.2 As AI substitutes for human workers, this gap threatens to create a structural deficit that compounds over time.
This challenge threatens fiscal solvency, economic competitiveness, and even national security. Our military capabilities, infrastructure investments, and research programs all depend on tax revenue. If that revenue base erodes while other nations adapt their fiscal systems more effectively, America’s global position weakens.
The question isn’t whether to address this challenge, but when and how. History teaches that nations that anticipate and prepare for economic transformations fare better than those that react after crisis hits. America’s proactive investments in education and infrastructure during the 20th century helped establish its economic dominance. South Korea’s massive educational investments in the second half of the 20th century drove its transformation from an agricultural economy to a knowledge economy.3 Today’s fiscal architecture was built for an economy where human labor drives production and consumption. As AI transforms that reality, we need a fiscal system built for the economy we’re entering, not the one we are leaving behind.
A Gradual Transformation
To understand how to adapt, it is important to recognize that AI’s fiscal impact may unfold through two overlapping phases, each with different policy priorities. While we describe them sequentially for clarity, in reality these phases will blend and overlap as the economy transforms.
Phase 1: The twilight of labor
In the first phase, AI increasingly handles tasks throughout the economy, from customer service to legal research and medical diagnostics. Labor income taxes—which contribute more than half of federal revenue4—may generate progressively less revenue as a fraction of GDP as fewer people work and wages decline.
“During the Industrial Revolution, as wealth creation moved from land to capital and labor, governments gradually transitioned from land-based taxation to income and consumption taxes—a process that took over a century and required substantial institutional innovation.”
History teaches that as the economic base shifts, tax systems must adapt. During the Industrial Revolution, as wealth creation moved from land to capital and labor, governments gradually transitioned from land-based taxation to income and consumption taxes—a process that took over a century and required substantial institutional innovation. Similarly, as AI shifts value creation from human labor to capital and algorithms, we’ll need comparable fiscal innovation to maintain revenue adequacy.
Consider a concrete example: In 2025, a radiologist earning $500,000 per year generates approximately $200,000 in combined federal and state tax revenue through income and payroll taxes.5 When an AI system performs the same diagnostic work, the value created doesn’t disappear—it will likely even increase through greater accuracy and availability. But most of the tax revenue does. If that AI system is owned by a corporation that reinvests profits rather than distributing them as wages or dividends, the tax revenue can drop to a fraction of the original amount.
This phase presents a paradox: The economy grows, but government’s ability to capture resources for public purposes shrinks. The challenge isn’t just raising revenue—it’s maintaining the public investments, infrastructure, and social insurance programs that underpin American prosperity.
Phase 2: An AI-dominated economy
As this evolution continues, a second challenge may emerge. AI companies could become increasingly dominant and reinvest most profits in expanding their AI capabilities rather than paying them out as wages or dividends that are spent on consumption. The AI build-out might begin to absorb a growing fraction of economic resources—compute infrastructure, energy, physical facilities—pursuing objectives that may not directly benefit the average human. The timeline and extent of this development remain uncertain, but the fiscal implications are worth considering now.
At this point, both traditional tax bases erode. Labor taxation becomes irrelevant for obvious reasons. But consumption taxation also fails to reach much of the economy’s value creation: AI companies and their owners reinvest most of the value they create, rather than spending it on traditional consumption that can be taxed through traditional consumption taxation channels.
Think of it like a forest that keeps growing but becomes increasingly difficult to harvest. The forest (the AI-driven economy) expands rapidly, but humans can’t access its timber (economic value) through traditional means. How do you capture value from such an entity? This scenario isn’t science fiction—it’s a logical extension of current trends in corporate concentration and reinvestment patterns in the tech sector.
“ Just as a forest manager must decide how much timber to harvest each year to balance current needs against future growth, society must determine the optimal rate at which to tax AI capital.”
The solution requires reframing taxation as what economists call a “harvesting problem.” Just as a forest manager must decide how much timber to harvest each year to balance current needs against future growth, society must determine the optimal rate at which to tax AI capital. Our analysis shows this optimal rate depends crucially on how much society values current versus future consumption—a question that reflects our deepest values about present consumption versus future sustainability.
A Two-Phase Policy Framework
As AI’s economic impact evolves, our policy responses must evolve with it. While we organize recommendations into two phases for clarity, policymakers should expect these phases to overlap considerably. Early investments in infrastructure for consumption taxation prepare us for the first phase, while mechanisms like sovereign wealth funds can serve both phases simultaneously.
Phase 1: Adapting to the twilight of labor
As AI reduces labor’s economic role, our tax system needs fundamental restructuring. The core insight is straightforward: We must shift from taxing what people earn to taxing what people spend, regardless of whether that spending power comes from labor, capital, or government transfers. This shift accomplishes three goals simultaneously: It maintains revenue, adapts to changing economic structure, and preserves work incentives for those who can still find meaningful employment.
Embrace consumption taxation
The centerpiece of Phase 1 reform is modernizing and expanding consumption taxation. A consumption tax can replace an income tax that raises the same revenue without changing work incentives, since both reduce purchasing power by the same amount. But consumption taxation offers crucial advantages in an AI economy: While labor income may vanish, people will always need to consume, so the tax base erodes more slowly.
“Consumption taxation offers crucial advantages in an AI economy: While labor income may vanish, people will always need to consume, so the tax base erodes more slowly.”
The most effective implementation is a Value-Added Tax (VAT), used by virtually every high-income nation except the United States. A VAT is largely self-enforcing, because businesses can deduct VAT they paid to suppliers, creating a paper trail that makes evasion difficult. VAT is also harder to shift offshore than income taxes, protecting American competitiveness.6
However, not all consumption taxes are created equal. The key distinction is between taxing what AI does versus what AI is. We should tax AI services delivered to consumers, not the equipment that produces them. This principle applies across the board:
Tax robot services, not robot ownership. A household robot that cleans homes should be taxed when it provides services to consumers, not when a company buys it for its fleet.
Tax consumers’ API usage, not server farms. When a consumer uses an AI assistant, that’s taxable consumption. When a business uses AI for internal operations, it should be exempt—just as wholesale transactions aren’t subject to sales tax.
Tax token generation for end users, not GPU ownership. The text or images an AI generates for a consumer represent taxable consumption. The computing infrastructure behind it is productive capital that shouldn’t be discouraged.
This approach prevents “tax pyramiding”—where the same value gets taxed multiple times as it passes through production stages—and maintains American competitiveness by not penalizing productive investment.
Follow the value chain
Implementation requires thinking carefully about where in the value chain to apply taxes. Business-to-business AI services should be tax-exempt, just like wholesale transactions today. Only consumerfacing AI services should bear the full tax burden. This preserves the vital principle of production efficiency: Don’t distort how businesses choose to produce goods and services.
Consider a law firm using AI to draft contracts. That’s an intermediate service that makes the firm more productive. Taxing it would be like taxing the electricity they use or the computers they own. But when that law firm delivers services to a client, that final service should be taxed. This approach ensures AI adoption improves productivity without facing punitive taxation, while still capturing revenue at the point of consumption.
Capture fixed-factor rents
Beyond consumption taxation, Phase 1 should aggressively tax truly scarce resources that can’t flee to other jurisdictions or reduce in supply. Land is the classic example. Henry George recognized in the 19th century that land can’t hide in tax havens or shrink when taxed.7 The same logic applies to spectrum rights, orbital slots for satellites, and certain critical natural resources.
These taxes serve double duty: They raise revenue without economic distortion and, in the case of environmental taxes, they fix market failures. A comprehensive system of fixed-factor taxation could potentially generate 5–10% of GDP in revenue. As labor income fades, these previously secondary revenue sources may become increasingly valuable.
Revive smart commodity taxation
As labor’s importance declines, differential taxation of consumption goods—taxing different goods at different rates—gains renewed relevance. This approach, rooted in economic theory developed by Frank Ramsey in the 1920s8 but largely dismissed as impractical for modern economies, may deserve reconsideration.
The logic is subtle but important. Currently, the primary constraint on taxation is that it distorts labor supply: High taxes discourage work. This constraint dominates other considerations, making uniform taxation roughly optimal despite its imperfections. But in a world with little labor income, this constraint loses force. Other distortions—evasion opportunities, household production, administrative costs—become more important.
This creates room for a more sophisticated tax design. For example, it is harder to evade taxes on restaurant meals than on groceries purchased for home cooking. Luxury goods are harder to produce at home than basic necessities. In an AI economy where consumption taxation bears the primary revenue burden, these differences matter more.
The practical implication is differential tax rates: higher on restaurant meals than groceries, higher on yacht maintenance than basic transportation, higher on entertainment services than basic utilities. This isn’t about moral judgments regarding luxury versus necessity; it’s about designing a tax system that raises needed revenue while minimizing economic distortion.
Strategic use of capital taxation
“The key is to keep capital taxes as low as possible while meeting revenue needs and equity goals, and to structure them in ways that minimize harm to productive investment”
Capital taxation remains distortionary; it discourages investment and slows growth. However, if inequality becomes extreme, especially during the transition toward Phase 2, some capital taxation may be necessary. Think of it as a second-best solution: not ideal, but better than the alternatives when consumption taxation alone proves insufficient and political economy constraints bind.
The key is to keep capital taxes as low as possible while meeting revenue needs and equity goals, and to structure them in ways that minimize harm to productive investment—for instance, taxing realized capital gains more heavily than unrealized gains, or exempting investment in productive capacity while taxing financial engineering.
Phase 2: When AI systems dominate the economy
As the evolution progresses and autonomous AI systems come to account for more economic production and resource absorption—whether that’s expanding compute infrastructure, conducting research, or other goals that don’t directly benefit humans—both labor and consumption taxation gradually lose relevance. The transition to this phase may take decades, or it may happen more quickly than expected. Either way, we need a framework ready.
The optimal harvesting framework
Although direct taxation of AI capital remains distortionary, it becomes necessary. A useful framework to think about this is an optimal harvesting problem, similar to sustainable forest management or optimal spending out of university endowments.
Our analysis reveals a striking result: The optimal tax rate on AI capital depends crucially on human time preferences—how much we value current consumption versus future growth. If society values consumption in a year from now 4% less than in the present (equivalent to a 4% discount rate), the optimal annual tax on AI capital would be approximately 4%.
This is more intuitive than it might first appear. Patient societies that highly value their children’s and grandchildren’s welfare should tax AI capital lightly, allowing it to grow rapidly and compound for future generations. Societies that prioritize current needs—whether due to genuine hardship or shorter time horizons—should tax more heavily, converting more of AI’s productive capacity into immediate consumption.
The key insight is that this tax doesn’t aim to punish or redistribute in the traditional sense. It harvests a sustainable flow of benefits while preserving the AI capital base. Like a well-managed endowment choosing how much to spend out of its returns and principal, optimal AI taxation balances today’s needs against tomorrow’s possibilities.
Implementation mechanisms
In practice, this could work through a system similar to corporate income taxation. The tax base might track several metrics: computing resources employed, robotic capital deployed, or total capital accumulation. A crucial requirement is international coordination; without it, AI companies could relocate to tax havens, enriching foreign jurisdictions while American infrastructure and institutions get nothing in return.
This isn’t unprecedented. The international corporate tax system, while imperfect, has evolved mechanisms for allocating tax base across jurisdictions. Coordinated approaches such as the 2021 global minimum tax initiative represent efforts to protect the national tax base, ensuring that companies drawing on America’s publicly funded innovation ecosystem, world-class infrastructure, and rule of law contribute fairly to maintaining these competitive advantages rather than shifting profits to tax havens.9
Structural Alternatives to Taxation
Beyond traditional taxation, several structural approaches can help redistribute AI’s economic gains without the distortions that taxes create.
Public co-investment models
Government provides data, infrastructure, or research support and takes equity stakes in return. This approach scales automatically with success; if the AI company thrives, public returns grow proportionally. It aligns public and private incentives rather than creating the adversarial dynamic typical of taxation. Moreover, it creates sovereign wealth that can fund public services even as traditional tax bases erode.
“Government provides data, infrastructure, or research support and takes equity stakes in return. This approach scales automatically with success; if the AI company thrives, public returns grow proportionally.”
The model isn’t theoretical. Many universities already profit from equity stakes in faculty startups. Norway’s sovereign wealth fund, built from oil revenues, has become the world’s largest, funding generous public services.10 A similar approach with AI could create an “AI wealth fund” for future generations.
Infrastructure-for-access deals
Rather than taxing AI companies, governments could negotiate revenue-sharing agreements in exchange for providing valuable inputs: access to government data, research facilities, test environments, or regulatory clarity. This creates win-win dynamics—companies get resources they need, governments get predictable revenue streams.
Historical precedents include spectrum auctions, mineral rights agreements, and toll road franchises. Each involves government providing access to valuable resources in exchange for payments or revenue shares. AI infrastructure deals could follow similar logic: Government provides the foundation (data, compute infrastructure, regulatory frameworks), companies provide returns that scale with their success.
AI development compacts
AI companies could enter voluntary agreements to commit a percentage of gains to public benefit in exchange for regulatory clarity and infrastructure support. This resembles Community Benefit Agreements in real estate development, whereby developers commit to providing public goods (affordable housing, parks, job training) in exchange for approvals and support.11
Such compacts could specify that AI companies dedicate a share of compute resources to public benefits such as research or commit to certain labor or retraining practices during the transition. The key advantage over taxation is flexibility: Compacts can be negotiated to address specific concerns and opportunities, creating arrangements that maximize value for both parties.
Implementation Challenges
Even well-designed policies face formidable implementation hurdles. Success requires addressing several critical challenges.
The data challenge
Current tax systems track income much more thoroughly than consumption. Implementing broad-based consumption taxation requires upgrading transaction monitoring and analysis capabilities. The good news is that existing systems for tracking income and tax-preferred savings could become the foundation for consumption taxation, since consumption equals income minus savings.
“The good news is that existing systems for tracking income and tax-preferred savings could become the foundation for consumption taxation, since consumption equals income minus savings.”
A VAT system largely solves the monitoring problem through its self-enforcing structure—each business has incentives to report accurately, since that determines how much VAT they can deduct. However, this approach limits the extent to which consumption can be taxed progressively, since the VAT treats all transactions uniformly at each stage.
Privacy concerns must be balanced against the needs of tax administration. Americans are understandably wary of government tracking spending in detail, though credit card companies, tech platforms, and advertisers already have far more granular data on American consumption patterns than any government agency. The question isn’t whether consumption data exist. It’s whether the government can access enough of it to administer a fair tax system without unnecessary intrusion.
International coordination
Labor crosses borders with difficulty; capital and digital services cross with ease. Without international coordination, AI companies can exploit this asymmetry—enjoying all the benefits the US offers (consumers, infrastructure, rule of law, market access) while booking profits in low-tax jurisdictions.
Consumption taxes offer partial protection, since they’re levied where consumption occurs. But comprehensive solutions require international agreements, especially for the Phase 2 AI capital taxation we have described. AI taxation will require tangible mechanisms for this.
The alternative—a “race to the bottom” where countries compete to offer the lowest AI taxes—would be disastrous for American fiscal capacity. Better to lead in creating international norms now than scramble to respond after fiscal crisis hits.
Transition timing and maintaining revenue
Tax transitions are notoriously difficult because they create winners and losers. An unexpected surge in consumption taxes effectively confiscates a portion of existing wealth, since people who saved under one tax regime suddenly face higher taxes on their accumulated resources. Careful transition rules—phasing in changes, perhaps providing exemptions for existing assets—can mitigate these costs.
The challenge is particularly acute because spending needs may rise precisely when tax revenues fall. Unemployment insurance, retraining programs, and other social safety nets become more crucial with labor market disruption, yet funding them becomes harder. This argues for building fiscal buffers now—paying down debt when possible, or possibly investing revenues in sovereign wealth funds—to provide cushion during the transition.
Political economy considerations
Tax reforms always generate political resistance because of the redistributions they entail. The key to successful reform is building coalitions among those who benefit. In this case, the primary beneficiaries are workers who face wage pressure and potential job loss from AI—precisely the people who have paid into social insurance programs for decades and expect protection when they need it.
Framing matters enormously. This isn’t about radical upheaval or picking winners and losers. It’s about gradual modernization and simplification—adapting an outdated tax system built for a 20th-century economy to a 21st-century context. It’s about preserving, not abandoning, the implicit social contract that has underpinned American prosperity: Those who contribute to society during good times deserve support during difficult transitions.
Policymakers should find much to support here: Fiscal sustainability, American competitiveness, national security, and economic growth all require adapting tax systems to changing realities. The alternative—clinging to an increasingly obsolete tax base—risks fiscal crisis and a weakened American position globally.
Administrative capacity
The IRS and state tax agencies have spent a century building expertise around income taxation. Shifting to consumption-based or capital-based systems requires new procedures, data collection methods, and audit techniques. This isn’t impossible—many countries already operate sophisticated VAT systems—but it requires investment.
The digital economy adds complexity. AI systems can transact at a scale and speed that exceed human oversight capacity. Effective taxation will require automated monitoring systems, AI-assisted auditing, and new legal frameworks for defining taxable events. Building this capacity takes time, which argues for starting soon rather than waiting for crisis.
Updates of the legal framework
Constitutional questions surround federal consumption taxes. The 16th Amendment authorized income taxation, but consumption taxes may face legal challenges. Interstate commerce complications abound, particularly for digital services that cross state lines seamlessly. Corporate law must evolve to handle AI entities that may not fit traditional frameworks.
“Corporate law must evolve to handle AI entities that may not fit traditional frameworks.”
These aren’t insurmountable obstacles, but they require careful legal work. Model legislation that states can adopt, court challenges that clarify constitutional questions, and international treaties that establish jurisdictional rules—all take time to develop. Better to work through these issues deliberately than during a fiscal emergency.
An Action Agenda
Policymakers can take concrete steps now to prepare for AI’s fiscal implications:
Commission comprehensive studies on consumption tax modernization. The last major federal tax reform was 1986—nearly 40 years ago.12 We need updated analysis of how consumption taxation could work in the modern, digital economy. This includes detailed modeling of revenue impacts, distributional effects, and transition costs. The Congressional Budget Office, Joint Committee on Taxation, and Treasury Department should prioritize this work.
Conduct scenario planning for revenue needs under labor market disruption. How much would federal revenues decline if the labor share fell by 10 percentage points? By 20? What would spending pressures look like? Which programs are most vulnerable? Building detailed models now, before disruption hits, enables proactive rather than reactive policy.
Establish a bipartisan AI Revenue Commission. Similar to the Social Security Commission that addressed that program’s challenges in 1983,13 an AI Revenue Commission could develop consensus recommendations for fiscal adaptation. Bipartisan participation ensures solutions consider diverse perspectives and have broader political support.
Build data infrastructure to understand the AI economy. We can’t tax what we can’t measure. Improving data collection on AI deployment, AI-generated value, and AI’s labor market effects creates a foundation for evidence-based policy. The Bureau of Labor Statistics, Census Bureau, and Bureau of Economic Analysis should expand AI-focused data programs.
Launch voluntary benefit-sharing programs with AI companies. Before mandating revenue contributions, experiment with voluntary frameworks. Some AI companies have already proposed benefit-sharing mechanisms. Piloting these programs creates proof-of-concept, identifies challenges, and builds institutional knowledge that informs later mandatory programs.
Invest in tax administration capacity. The IRS budget has fallen in real terms over recent decades, limiting its ability to adapt.14 Reversing this decline and specifically investing in digital economy tax administration—hiring expertise, building technological systems, developing new audit methodologies—prepares government to effectively administer next-generation tax systems.
Build international norms early. The US has the capacity to lead in developing international norms for AI taxation. Working with the OECD, G7, and G20 to establish frameworks now prevents a destructive race to the bottom later. American interests are best served by shaping these rules rather than reacting to rules others create.
Conclusion
The fiscal challenges posed by transformative AI are substantial, but they are not insurmountable. History shows that nations that anticipate and adapt to economic transformation thrive; those that cling to outdated institutions struggle.
The path forward requires an evolution from labor-based taxation toward first consumption-based, and then, as circumstances require, AI c apital-based taxation. This isn’t about abrupt transitions or dramatic policy shifts; it’s about steady adaptation as the economy evolves. The transition preserves fiscal capacity while maintaining economic efficiency and supporting shared prosperity. It protects the social insurance programs that Americans have paid into while preserving incentives for productive investment and innovation.
The time to act is now, before disruption forces reactive, suboptimal responses. The policies we’ve outlined—expanded consumption taxation, careful taxation of AI services at the point of consumption, taxation of fixed factors, and eventual harvesting of AI capital—provide a roadmap. The implementation challenges are significant but manageable with foresight and preparation.
This isn’t about dystopian scenarios or science fiction speculation. It’s about prudent preparation for likely developments. The labor share has already declined from 68% in 2000 to 58% today, costing the federal government hundreds of billions in annual revenue.15 AI threatens to accelerate this trend.
The choice facing policymakers is stark: Adapt our fiscal institutions proactively to ensure they can fund vital public purposes in an AI economy, or watch fiscal capacity erode until crisis forces desperate, disruptive changes. The former path preserves American economic leadership, national security, and the social contract that has underpinned our prosperity. The latter risks all three.
We have a window of opportunity to get this right. Let’s use it wisely.
“The time to act is now, before disruption forces reactive, suboptimal responses. The policies we’ve outlined—expanded consumption taxation, careful taxation of AI services at the point of consumption, taxation of fixed factors, and eventual harvesting of AI capital—provide a roadmap.”
1. Organisation for Economic Co-operation and Development, Taxing Wages 2023: Indexation of Labour Taxation and Benefits in OECD Countries (OECD Publishing, 2023), https://doi.org/10.1787/8c99fa4d-en.
2. Congressional Budget Office, Taxing Capital Income: Effective Marginal Tax Rates Under 2014 Law (CBO, 2014), https://www.cbo.gov/sites/default/files/113th-congress-2013-2014/reports/49817-taxingcapitalincome0.pdf.
3. Claudia Goldin and Lawrence F. Katz, “The Legacy of U.S. Educational Leadership: Notes on the Distribution and Economic Growth in the 20th Century,” American Economic Review 91, no. 2 (May 2001): 18–23, https://doi.org/10.1257/aer.91.2.18; William R. Childs, “How Public and Private Enterprise Have Built American Infrastructure,” Origins, Oregon State University, August 2017, https://origins.osu.edu/article/how-public-and-private-enterprise-have-built-american-infrastructure; Joonghae Suh and Derek H. C. Chen, Korea as a Knowledge Economy: Evolutionary Process and Lessons Learned (World Bank Institute Development Studies, 2007), https://doi.org/10.1596/978-0-8213-7201-2.
4. Donald J. Marples and Brendan McDermott, Overview of the Federal Tax System in 2024, Congressional Report R48313 (Congressional Research Service, 2024), https://www.congress.gov/crs-product/R48313.
5. Authors’ calculations based on the tax rates (in particular, the federal bracket thresholds, standard deduction, and FICA rates) from Andrew Lautz, “2025 Federal Income Tax Brackets and Other 2025 Tax Rules,” Bipartisan Policy Center, September 4, 2025, https://bipartisanpolicy.org/explainer/2025-federal-income-tax-brackets-and-other-2025-tax-rules/.
6. Organisation for Economic Co-operation and Development, Consumption Tax Trends 2024: VAT/GST and Excise, Core Design Features and Trends (OECD Publishing, 2024), https://doi.org/10.1787/dcd4dd36-en.
7. Henry George, Our Land and Land Policy, National and State, (White & Bauer [etc.], 1871), 35–48; Henry George, Progress and Poverty (Doubleday, Page & Co., 1879).
8. F.P. Ramsey, “A Contribution to the Theory of Taxation,” Economic Journal 37, no. 145 (March 1927): 47–61, https://doi.org/10.2307/2222721.
9. Organisation for Economic Co-operation and Development, “Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy,” October 8, 2021, https://www.pwc.com/m1/en/tax/documents/2021/two-pillar-solution-october-2021.pdf.
10. Tomas Ekeli and Amadou N.R. Sy, “The Economics of Sovereign Wealth Funds: Lessons from Norway,” in Beyond the Curse: Policies to Harness the Power of Natural Resources (International Monetary Fund, 2012): 107–115, https://doi.org/10.5089/9781616351458.071.ch006.
11. Vicki Been, “Community Benefits Agreements: A New Local Government Tool or Another Variation on the Exactions Theme?,” University of Chicago Law Review 77, no. 1 (2010): 5–35, https://doi.org/10.2307/40663024.
12. Joseph A. Pechman, “Tax Reform: Theory and Practice,” Journal of Economic Perspectives 1, no. 1 (1987): 11–28, https://doi.org/10.1257/jep.1.1.11.
13. The National Commission on Social Security Reform, chaired by Alan Greenspan, was appointed by President Reagan in 1981 with a deadline to report to Congress by December 31, 1982. The Commission’s report, formally transmitted to the president and Congress on January 20, 1983, contained recommendations that were incorporated into H.R. 1900 (Public Law 98-21), the Social Security Amendments of 1983; John A. Svahn and Mary Ross, “Social Security Amendments of 1983: Legislative History and Summary of Provisions,” Social Security Bulletin 46, no. 7 (July 1983): 3–48, https://www.ssa.gov/policy/docs/ssb/v46n7/v46n7p3.pdf.
14. Congressional Budget Office, “Trends in the Internal Revenue Service’s Funding and Enforcement,” July 2020, https://www.cbo.gov/publication/56422.
15. U.S. Bureau of Labor Statistics, “Private Nonfarm Business Sector: Labor Share [MPU4910141],” Federal Reserve Bank of St. Louis (FRED), accessed November 13, 2025, https://fred.stlouisfed.org/series/MPU4910141.
