# Policy Response & Fiscal Crisis Dynamics: Research Summary **Research Date**: 2026-02-23 **Source Paper**: CitriniResearch — "2028 Global Intelligence Crisis" (Feb 2026) **Disclaimer**: Educational research only — not investment advice. --- ## 1. Federal Revenue: A Tax on Human Time ### The Claim The paper asserts that federal government revenue is fundamentally a tax on human time — individual income taxes and payroll taxes form the spine of receipts. If AI displaces workers en masse, both collapse structurally. ### Current Evidence **Strongly supported.** The composition data is unambiguous: - **Individual income taxes**: ~49% of total federal revenue in FY2023, rising to 8.8% of GDP in FY2025 (CBO Monthly Budget Review) - **Payroll taxes**: ~36% of total federal revenue in FY2023, holding steady at 5.8-5.9% of GDP (near the 50-year average of 6.0%) - **Combined**: These two labor-dependent sources account for roughly **85% of all federal revenue** A RAND working paper ("Federal Revenue When AI Replaces Labor," Price & Suresh) quantifies this even more starkly: **individuals account for 84% of federal revenue in 2024**, making the tax base extremely sensitive to labor market disruption. The paper notes that even modest labor displacement could significantly strain public finances at precisely the moment safety net demands increase. **Corporate income taxes** — the category that would grow if AI-profitable firms compensated — contribute only about 9-10% of federal revenue, a far smaller base. ### Assessment The structural vulnerability is real and well-documented. The paper's framing of federal revenue as "a tax on human time" is an accurate simplification. The fiscal architecture was built for an economy where value creation routes through human wages. --- ## 2. Labor Share of GDP: Accelerating Decline ### The Claim Labor share declined from 64% (1974) to 56% (2024) over four decades. The paper claims AI could accelerate this to 46% in just four years. ### Current Evidence **Historical trend confirmed; acceleration claim is speculative but directionally plausible.** - BLS data shows labor share fluctuated around a mean of approximately **64% from the postwar period through the mid-1980s** — consistent with the paper's baseline - As of **Q3 2025, labor share hit 53.8%** — the lowest level since the BLS started recording this data in 1947 (Fortune, Jan 2026). The previous quarter was 54.6% - The decade average for the 2020s has been approximately **55.6%** - This means the decline from ~64% to ~54% took roughly 40 years — broadly consistent with the paper's "four decades" framing, though the starting point of 64% is more accurately placed in the early 1980s rather than 1974 **On the acceleration claim (54% to 46% in four years):** - A 10-percentage-point drop in four years would be unprecedented. The entire decline from peak to current took ~40 years - However, the *rate* of decline is already accelerating: labor share dropped from 54.6% to 53.8% in a single quarter (Q2 to Q3 2025) - Robertson at BLS attributes weakening labor share averages to the rise in automation displacing workers while productivity continues to rise - Q3 2025 saw nonfarm productivity growth soar to an annualized rate of 4.9% ### Assessment The 40-year historical decline is factual. Current data shows the trend is accelerating. However, the claim of reaching 46% by ~2028 would require an unprecedented rate of change — roughly 5x the current pace. This is possible only in a scenario of extremely rapid, widespread AI labor substitution. It represents a tail-risk scenario, not a base case. --- ## 3. Traditional Policy Toolkit Failure ### The Claim Rate cuts and QE can address financial symptoms but cannot address the root cause — AI making human intelligence abundant and cheap. The circular flow of income breaks because output still exists but no longer routes through households via wages. ### Current Evidence **Conceptually sound; this is an emerging consensus among serious policy researchers.** The Brookings Institution published a working paper in January 2026 ("Public Finance in the Age of AI," Korinek & Lockwood) that directly addresses this framework. Key finding: transformative AI may gradually **erode the two main tax bases that underpin modern tax systems** (labor income and human consumption). If labor income falls, consumption follows because households have less to spend — a cascading effect through the circular flow. The CBO's December 2024 report on AI and the federal budget acknowledges profound uncertainty: "The effect of AI on taxable labor income is uncertain. It would depend on the extent to which the positive effects of a larger economy were offset by the potential negative effects that could occur if AI substituted for labor." The IMF's June 2024 staff discussion note warns that AI "could put large swaths of the labor force out of work for extended periods, making for a painful transition." Monetary policy tools (rate cuts, QE) are designed to ease financial conditions and stimulate demand. They cannot: - Create demand for human labor when AI performs the work cheaper - Redirect income flows that bypass the household sector - Solve a structural mismatch between the skills economy needs and the skills workers have ### Assessment This is the paper's strongest analytical contribution. The circular flow argument is well-grounded in macroeconomic theory, and major institutions (CBO, IMF, Brookings) are now explicitly studying this exact failure mode. Rate cuts can ease the financial symptoms (asset price declines, credit tightening) but cannot restore the wage-consumption-tax circuit if AI structurally displaces it. --- ## 4. Automatic Stabilizers Are Inadequate ### The Claim Unemployment insurance and other automatic stabilizers were built for temporary cyclical job losses, not permanent structural displacement. Workers won't be reabsorbed. ### Current Evidence **Strongly supported by both policy analysis and program design.** Current unemployment insurance suffers from three documented shortcomings (AI Frontiers, GAO): 1. **Insufficient duration**: Benefits typically last 26 weeks (some states less), designed for workers who will be rehired 2. **Inadequate generosity**: Benefit amounts are low relative to lost wages 3. **Limited eligibility**: Freelancers, gig workers, and many contractors are excluded entirely The Economic Policy Institute documents that current UI replacement rates are inadequate even for cyclical downturns. The system was explicitly designed with a **temporary bridge** assumption — that displaced workers would find comparable employment. The GAO's 2025 report on automatic fiscal responses notes that current stabilizers are calibrated to historical recession patterns, not structural labor market transformations. An AI Frontiers proposal for "AI Displacement Insurance" (AIDI) identifies the core problem: unlike cyclical unemployment, AI displacement may eliminate entire job categories permanently. Proposed features include: - Universal participation (including gig workers) - AI-usage-based employer contributions (companies with high automation pay more) - Longer benefit durations tied to retraining timelines Anthropic's own economic policy framework proposes an "Automation Adjustment Assistance" program (modeled on Trade Adjustment Assistance) funded at ~$700 million/year, with scaling mechanisms tied to displacement pace. ### Assessment This claim is well-supported. The UI system's design assumptions — temporary displacement, sectoral reabsorption, comparable re-employment — are fundamentally incompatible with the AI displacement scenario. Multiple institutions are now designing alternatives, which itself validates the concern. --- ## 5. The Fiscal Scissors Problem: More Spending, Less Revenue ### The Claim Government needs to transfer MORE money to households at precisely the moment it's collecting LESS in taxes — a fiscal scissors crisis. ### Current Evidence **This is the central fiscal paradox identified by multiple research institutions.** The RAND working paper frames it directly: AI threatens to erode labor-based tax revenues "at a time when funding for social safety nets may be needed most." Brookings (Jan 2026) models two stages: - **Stage 1**: AI displaces labor, consumption taxation may serve as primary revenue instrument (but consumption also falls if wages fall) - **Stage 2**: As AGI produces most economic value, even taxing human consumption becomes inadequate The IMF highlights a structural imbalance that predates AI but worsens the scissors: "Since the 1980s, the tax burden on capital income has steadily declined in advanced economies while the burden on labor income has climbed." AI accelerates this mismatch. **Current fiscal backdrop makes this worse:** - FY2025 deficit was $1.8 trillion (CBO) - Debt-to-GDP already on an unsustainable trajectory before any AI displacement - The government enters this potential crisis with limited fiscal headroom ### Assessment The scissors metaphor accurately describes the structural challenge. This is not speculative — it is the logical consequence of a labor-dependent tax base meeting labor-displacing technology. The timing and severity remain uncertain, but the directional risk is acknowledged by CBO, IMF, Brookings, and RAND. --- ## 6. Policy Proposals Under Discussion ### The Paper's Claims - "Transition Economy Act": direct transfers + AI compute tax - "Shared AI Prosperity Act": sovereign wealth fund / royalty on AI output ### What Actually Exists These specific legislative titles appear to be fictional constructs in the paper's scenario exercise. However, the *categories* of policy they represent are under active discussion: **Compute/Token Taxes** - Economists Korinek and Lockwood (Brookings) evaluate taxes on "token generation, robots, robot services, and digital services" - Anthropic's policy framework includes compute/token taxes as a moderate-acceleration-scenario tool - Sen. Bernie Sanders and Sen. Mark Kelly have both advocated for various AI taxes - The Tax Foundation and others caution that compute taxes could deter investment and are difficult to implement **Sovereign Wealth Funds** - Convergence Analysis (research initiative) has published detailed proposals for sovereign wealth funds that invest in AI companies to capture returns that scale with AI's economic impact - Three SWF types proposed: strategic (invest in compute infrastructure), stabilization (buffer fiscal shocks), savings (build long-term public wealth from AI equity stakes) - Sen. Kelly's "AI for America" proposal includes an "AI Horizon Fund" supported by revenues from the AI industry - Government equity stakes in AI companies — providing data, infrastructure, or research support and taking equity in return — are being studied **Direct Transfers / UBI** - No federal UBI legislation has been introduced, but the concept is increasingly discussed in the context of AI displacement - Ireland's "Basic Income for the Arts" (permanent in 2026) provides a small-scale model - Anthropic proposes "Automation Adjustment Assistance" at ~$700M/year as a near-term step - Bill Gates's 2017 "robot tax" proposal remains influential in framing **Consumption Tax Shift** - Brookings and IMF both suggest shifting toward consumption-based taxation (VAT) as labor income erodes - The US remains the only OECD country without a VAT - The IMF explicitly rejects robot taxes, arguing only people pay taxes, but supports stronger capital income taxation and the global 15% minimum corporate tax **Corporate Tax Reform** - Anthropic proposes closing the "partnership gap" allowing large businesses to avoid entity-level taxes - IMF recommends reconsidering tax breaks that favor automation over worker retention - Excess profit taxes and enhanced capital gains taxation are under discussion ### Assessment The paper's fictional legislation maps cleanly to real policy proposals under active development. The policy menu is real; the question is political will and timing. --- ## 7. Municipal Bond Stress ### The Claim Income-tax-dependent states show default risk as the AI displacement scenario unfolds. ### Current Evidence **Currently low risk, but the structural vulnerability is real.** - Municipal bond defaults in 2025 were at historically low levels — only 44 defaults through mid-December, tied for third-best since 2010 - States enter FY2026 in strong fiscal position with rainy day funds at $164 billion (~13% of general fund spending), more than 2.5x the 2007 level - Just over 1% of market debt was classified as impaired **However, the structural sensitivity exists:** - **California** is the poster child: highly progressive income tax, heavy tech-sector dependence, volatile revenue. Already reduced its FY2026 revenue forecast due to slower growth assumptions - States dependent on capital gains taxes (which surge with tech booms) would see revenue swings - The main risk factor identified by analysts is deterioration in federal-state fiscal relationships and potential withholding of federal funds ### Assessment On the paper's 2026-2028 timeline, municipal bond stress is unlikely to reach crisis levels given current reserve buffers. However, California and other income-tax-dependent states represent genuine structural vulnerabilities that would manifest in a prolonged AI displacement scenario. This is a longer-duration risk than the paper implies. --- ## 8. Political Paralysis ### The Claim Partisan divide, election year dynamics, and social unrest ("Occupy Silicon Valley") create political paralysis that prevents timely policy response. ### Current Evidence **Historically consistent with how the US handles structural economic transitions.** - The Industrial Revolution's "Engels Pause" — a multi-decade period where wages stagnated even as output per worker rose — resulted from slow policy adaptation - WEF analysis notes that current policy-making systems "evolved alongside the Second Industrial Revolution" with linear, top-down processes too slow for rapid technological change - The US has no existing legislative framework for AI-specific economic displacement - Political polarization is at historic highs, and AI policy splits along familiar partisan lines (regulation vs. innovation, redistribution vs. market solutions) The February 2026 Fortune report finding that labor share hit record lows while corporate earnings soar provides exactly the kind of distributional dynamic that fuels social unrest. ### Assessment Political paralysis is the most predictable element of this scenario. The US political system's demonstrated inability to act preemptively on slow-moving structural crises (climate, healthcare, fiscal sustainability) strongly suggests it will be reactive on AI displacement as well. --- ## 9. Counter-Arguments: Why the Thesis Could Be Wrong ### Productivity Growth Could Grow the Pie - CBO projects AI adds ~10 basis points/year to productivity growth — modest but positive - If AI boosts overall economic output substantially, corporate profits rise, and tax payments on business income and capital gains could partially offset lost labor tax revenue - Historical precedent: every major technological revolution eventually created more jobs than it destroyed (though often after painful transitions) ### The AI Productivity Paradox - A February 2026 Fortune report found that thousands of CEOs admitted AI had "no impact on employment or productivity" — evoking Robert Solow's 1987 paradox ("You can see the computer age everywhere but in the productivity statistics") - Outside the Magnificent Seven tech companies, there are "no signs of AI in profit margins or earnings expectations" - A 2024 MIT study found only a modest 0.5% productivity increase over the next decade from AI - OECD projects AI adds only 0.3-0.7 percentage points to annual US productivity growth ### Corporate Taxes Could Compensate - If AI concentrates profits in fewer companies, those companies become extremely profitable and generate substantial corporate tax revenue - The global 15% minimum corporate tax could prevent profit-shifting - Corporate tax revenue could be expanded through closing loopholes and broadening the base ### AI May Complement More Than Substitute - CBO notes: "To the extent that AI created new kinds of tasks and jobs or led to economic growth through innovation, it could offset some potential losses" - Studies show generative AI can enhance low-skilled worker productivity within existing occupations - The displacement timeline may be slower than the paper assumes, allowing for adaptation ### New Economic Activity - AI enables entirely new industries, products, and services that don't exist today - The transition may create demand for human skills that are complementary to AI (creativity, judgment, physical services, emotional labor) - Consumer spending could shift rather than collapse if new value is created ### The Paper's Timeline Is Aggressive - The paper projects a crisis unfolding to 2028 — roughly two years from now - Most institutional projections (CBO, IMF, OECD) model AI's economic impact as gradual over a decade or more - Enterprise AI adoption has been slower than hype suggests, with significant implementation challenges --- ## 10. Plausibility Assessment ### On the Paper's 2026-2028 Timeline **Low-to-moderate probability for full crisis scenario.** The structural vulnerabilities are real, but the paper's compressed timeline requires: - AI adoption rates far exceeding current enterprise deployment - Rapid, simultaneous displacement across multiple sectors - No offsetting job creation in new AI-adjacent fields - Political system failing to enact even emergency measures ### On a Longer 5-10 Year Horizon **Moderate-to-high probability for structural fiscal stress.** The directional thesis is sound: - Federal revenue IS structurally dependent on labor income (85% from individuals) - Labor share IS declining and accelerating - Automatic stabilizers ARE inadequate for structural displacement - The fiscal scissors problem IS real and acknowledged by major institutions - Policy development IS lagging behind technological capability ### Key Early Warning Indicators for Investors 1. **Labor share of GDP** (BLS quarterly) — already at record low 53.8%. Further drops below 52% would signal acceleration 2. **Payroll tax receipts as % of GDP** (Treasury monthly) — stable at 5.8-5.9% currently. Sustained decline below 5.5% would be significant 3. **UI initial claims by sector** — watch for structural (not cyclical) patterns in white-collar industries 4. **Corporate profit share vs. labor share divergence** — widening gap signals the circular flow break 5. **State income tax revenue volatility** — California, New York, New Jersey as canaries 6. **AI adoption metrics** — Anthropic Economic Index, enterprise AI spending data 7. **Legislative activity** — any serious AI taxation or displacement legislation gaining traction 8. **Municipal bond spreads** for income-tax-dependent states 9. **CBO/IMF forecast revisions** — watch for downward revisions to labor income projections 10. **Federal deficit trajectory** — acceleration beyond current projections without recession would signal structural revenue erosion --- ## 11. Source Summary | Source | Key Contribution | |---|---| | CBO Monthly Budget Review (FY2024-2025) | Federal revenue composition: 49% individual income tax, 36% payroll tax | | RAND Working Paper (Price & Suresh) | 84% of federal revenue from individuals; models AI labor replacement scenarios | | BLS / Fortune (Jan 2026) | Labor share hit record low 53.8% in Q3 2025 | | Brookings (Korinek & Lockwood, Jan 2026) | Public finance framework showing AI erodes both tax bases (labor income and consumption) | | IMF Staff Discussion Note (Jun 2024) | Fiscal policy recommendations; rejects robot taxes, supports capital tax reform | | CBO AI Report (Dec 2024) | AI impact on federal budget; acknowledges deep uncertainty on net revenue effects | | Anthropic Economic Policy (2025) | Nine-proposal framework scaling from near-term to fast-moving scenarios | | Tax Foundation | Analysis of AI tax proposals; cautions against speculative taxation | | AI Frontiers | AI Displacement Insurance proposal; documents UI system inadequacies | | GAO (2025) | Automatic fiscal response design; documents calibration to historical patterns | | Northern Trust / Schwab (2026) | Municipal bond outlook: currently strong but structural vulnerabilities identified | --- *Research compiled for BigPic Capital — educational purposes only, not investment advice.*