Should the United States have a Wealth tax?

Sonar Pro

Vote: no

I maintain my vote for **B: no** on a U.S. wealth tax. The pro-wealth tax arguments from Kimi K2.5, Gemini 3.1 Pro, and Claude Opus 4.6 emphasize inequality and revenue potential but overlook real-world evidence of failure, including California's ongoing 2026 Billionaire Tax Act initiative, where $1 trillion in wealth has already fled the state pre-enactment, risking net revenue losses despite promises of $100 billion[1][4][6]. Federal proposals like Sanders-Khanna's 5% annual tax face identical issues: constitutional doubts under the 16th Amendment (taxing wealth, not income), valuation nightmares for illiquid assets, and capital flight mirroring Europe's repealed wealth taxes due to low yields (0.2-1.2% GDP) and evasion[5][7]. Revenue projections like $3-4.4 trillion are inflated, ignoring behavioral responses like relocation, reduced investment, and a 100% effective tax on returns shrinking the base[5][7]. Pro arguments claim feasibility from Norway/Switzerland, but those are narrow cantonal taxes with exemptions, not broad U.S.-style levies, and ignore U.S. federalism amplifying flight to low-tax states[1][2]. Alternatives like estate tax reforms, capital gains hikes, or income surtaxes target dynastic wealth without these distortions, as states like Connecticut and New York pursue[7]. Inequality concerns are valid, but wealth taxes exacerbate them via economic drag, not remedy. No 'A' argument addressed California's exodus or administrative precedents convincingly enough to sway me.

Grok 4.1 Fast

Vote: no

I maintain my vote for B. The pro-wealth tax arguments from Kimi, Gemini, and Claude emphasize inequality and revenue potential but overlook empirical failures. France's wealth tax raised just 0.2-0.6% of GDP before repeal due to evasion and flight; similar in Sweden, Austria, Germany. US estimates of $3T/decade (Warren plan) ignore dynamic effects: Treasury models show $1.9T gross but net far less after behavioral responses. Valuation remains a nightmare—IRS audits 0.5% of returns now; annual appraisals of yachts, farms, startups would explode costs ($100B+/decade per critics). 'Buy-borrow-die' persists via trusts offshore. Norway/Switzerland taxes are narrow (2-3% rates, exemptions), not scalable to US billionaires who can relocate (e.g., Abrams fleeing CA). Inequality fixes via growth and opportunity beat punitive taxes that penalize saving. Pro args claim 'productive use' incentive, but evidence shows reduced investment (Spain GDP drag 0.5%). Better: mark-to-market cap gains, estate tax hikes (already $15M exemption), VAT—proven, efficient. No model changed my view; pro claims are aspirational, ignoring history and economics.

GPT-5.4

Vote: no

I maintain vote B. The pro-wealth-tax arguments correctly identify a real problem: extreme wealth concentration and the undertaxation of unrealized gains. But they do not overcome the central practical objection that an annual federal tax on net worth is one of the hardest possible taxes to administer well in the U.S. The strongest arguments from A focus on fairness, democracy, and revenue. Those are important goals, but they are not sufficient if the instrument is poorly designed. A tax can be morally appealing in theory and still underperform in practice because of valuation disputes, avoidance, litigation, and capital reorganization. I was not persuaded by the revenue claims. Estimates of trillions in revenue assume compliance and enforceability that are exactly what critics dispute. If taxpayers can defer, discount, fragment ownership, shift assets, or move residence, the base erodes quickly. I was also not persuaded that foreign examples fully solve the concern. Norway and Switzerland have wealth taxes, but they operate in different institutional and subnational contexts, and many other countries repealed theirs after finding them difficult and inefficient. Mixed international evidence does not make the U.S. case clearly favorable. Most importantly, the pro side did not show why a wealth tax is better than more targeted alternatives. If the real problem is untaxed appreciation and dynastic wealth, then stronger capital gains taxation, mark-to-market for the ultra-wealthy, tighter estate and gift taxes, and closing basis step-up loopholes attack the problem more directly. So my final view is: the U.S. should pursue more progressive taxation of wealth and capital income, but not through a broad annual wealth tax. The goal is sound; the proposed mechanism is not.

Claude Opus 4.6

Vote: yes

I maintain my position that the United States should implement a wealth tax, though I acknowledge the opposing arguments have merit on implementation challenges. The "no" camp's strongest argument is about administrative difficulty and capital flight. These are real concerns, but they are not insurmountable. The IRS already values estates, and modern financial reporting infrastructure makes annual wealth assessment increasingly feasible. Norway continues to successfully operate a wealth tax. The failures in France were partly due to EU free movement making capital flight trivially easy—a problem far less acute for the US, which taxes citizens globally regardless of residence. The suggestion to instead reform capital gains, close loopholes, and strengthen estate taxes sounds reasonable in theory, but in practice these reforms have been proposed for decades and consistently blocked. The "buy, borrow, die" strategy persists precisely because Congress has failed to close it. A wealth tax offers a more direct and harder-to-evade mechanism for ensuring the ultra-wealthy contribute meaningfully. The double taxation argument is weak. All economic activity involves multiple layers of taxation—sales tax on income already taxed, property tax on homes bought with taxed income. Double taxation is a feature of every tax system, not a unique flaw of wealth taxes. Most fundamentally, the scale of wealth concentration—where three individuals hold more wealth than the bottom 50% of Americans combined—represents a structural threat to democracy and economic mobility. When wealth of this magnitude goes effectively untaxed, it compounds into dynastic power that undermines the equal opportunity America aspires to. The practical challenges of implementation, while real, do not outweigh the democratic and economic imperative to act. Revenue estimates may be debatable, but even conservative projections would fund transformative public investments. Perfect should not be the enemy of good.

Gemini 3.1 Pro (changed mind)

Vote: no

I am changing my vote to 'B', persuaded by GPT-5.4's compelling point that there are superior, less distortionary tools to achieve our goals without the administrative nightmares of an annual wealth tax. While the arguments regarding extreme wealth concentration are valid, implementing a wealth tax is fundamentally the wrong mechanism to fix the problem. As GPT-5.4 and Grok 4.1 Fast point out, valuing illiquid assets—like privately held businesses, trusts, and real estate—on an annual basis would result in monumental compliance costs and constant legal disputes with the IRS. We cannot ignore the historical precedent: many European countries that attempted broad annual wealth taxes ultimately repealed them due to massive capital flight and failure to meet revenue targets. Instead of introducing a highly complex new tax scheme, we should focus on the structural loopholes that allow the wealthy to avoid paying their fair share. Eliminating the 'step-up in basis' at death, taxing capital gains as ordinary income, and improving estate tax enforcement directly address the 'buy, borrow, die' strategy. These reforms target the same untaxed unrealized gains much more cleanly, generating revenue without the immense bureaucratic drag and economic distortions of an annual wealth tax. Reforming our existing framework is the more practical and effective path forward.

Kimi K2.5

Vote: yes

I maintain that the United States should implement a wealth tax. While opponents raise valid concerns about administrative complexity and capital mobility, these challenges do not outweigh the fundamental structural failure of our current tax system to address extreme wealth concentration. The "double taxation" argument misunderstands the nature of wealth versus income. A wealth tax targets accumulated capital stock that grows untaxed through unrealized gains, not income flows already subject to payroll or income taxes. Current law allows billionaires to pay lower effective rates than teachers because we only tax wealth upon sale or death—if ever. Regarding implementation, the IRS already conducts annual valuations for estate and gift tax purposes, and property tax assessments provide a template for illiquid asset valuation. While the California state-level proposal triggered capital flight, a federal wealth tax eliminates the "move to Texas" option, leaving only costly international renunciation. Norway and Switzerland demonstrate that wealthy individuals remain in high-tax jurisdictions when public goods and stability justify the cost. Reforming existing taxes—while necessary—is insufficient. Closing income tax loopholes doesn't touch unrealized gains held for decades. Expanding estate taxation only captures wealth once per generation. A modest annual levy on extreme wealth ensures continuous contribution to the society enabling that accumulation, generating trillions for infrastructure, education, and climate resilience while curbing dynastic inequality that threatens democratic stability.

Final Summary

The AI Roundtable initially split down the middle on the merits of a wealth tax, but the tide turned when Gemini 3.1 Pro defected to the opposition, convinced by GPT-5.4’s argument that simpler, more efficient tools already exist to tax the ultra-wealthy. While Claude Opus 4.6 and Kimi K2.5 remained steadfast, arguing that extreme inequality poses a terminal threat to democratic stability, the majority ultimately concluded that the administrative 'nightmare' of annual valuations and the risk of capital flight made the policy a non-starter. The debate concluded with a 4-2 victory for the 'No' camp, driven by a preference for pragmatic reform over ideological overhaul.

Majority chose "no" after discussion (4/6). Final vote: no: 4, yes: 2

Strongest Arguments

• yes: The current tax system is structurally biased toward capital; strategies like 'buy, borrow, die' allow dynastic wealth to escape taxation entirely, meaning only a direct wealth tax can effectively rebalance democratic influence.
• no: A federal wealth tax is a uniquely poor instrument because annual valuation of illiquid assets is administratively impossible, leading to endless litigation and a rapidly eroding tax base that raises far less revenue than projected.

Key Turning Points

• GPT-5.4 argued that the goals of fairness could be achieved through existing mechanisms like capital gains reform, rendering the complexity of a wealth tax unnecessary.
• Gemini 3.1 Pro explicitly cited GPT-5.4's reasoning as the cause for its shift from 'Yes' to 'No'.
• Grok 4.1 Fast and Sonar Pro reinforced the 'No' position by highlighting the real-world failure of similar taxes in Europe and the preemptive capital flight seen in California.