We have not had anything like genuine progressive taxation in this country since before Reagan. The top marginal rate in 1960 was 91%. In 1980 it was 70%. Today it is 37%, and that is on income — the thing wealthy people have the least of. They hold wealth. They hold assets. And assets compound untaxed, year after year, until there is a realization event that wealthy people are very good at deferring indefinitely.
This is what Thomas Piketty documented in Capital in the Twenty-First Century — r > g, meaning the rate of return on capital consistently exceeds the rate of economic growth, which means wealth concentrates structurally unless policy intervenes. This is not a hypothesis. It is an empirical observation from 200 years of data across multiple countries.
A 2% annual wealth tax on assets over $50 million is not confiscatory. It is the difference between a billionaire's pile growing at 6% per year versus 4%. They are still getting richer. They are just getting richer more slowly, and the difference funds the public goods that make it possible to generate that wealth in the first place — courts, infrastructure, educated workers, functioning markets.
Elizabeth Warren and Bernie Sanders have both proposed versions of this. Emmanuel Saez and Gabriel Zucman at Berkeley estimate it would raise $200-300 billion per year. That is not nothing. That is roads, schools, healthcare, research. The argument against is always: capital flight, constitutional concerns, administrative difficulty. I will address each of those in exchange. But the fundamental moral case is simple: a society in which 400 people hold more wealth than the bottom 150 million is not a free society. It is a feudal one wearing democratic clothes.
Let me tell you what actually happened when France tried this.
France implemented a wealth tax — the ISF, impot de solidarite sur la fortune — in 1982. It was expanded in 1988. By 2012, the Hollande government had raised the rates again. And then, in 2017, Macron — who is not exactly a right-wing ideologue — abolished it entirely. Why? Because in the intervening years an estimated 10,000 millionaires per year were leaving France. The revenue projections never materialized. The administrative costs were enormous. And the investment that would have generated broader growth simply did not happen.
Sweden had a wealth tax too. Repealed in 2007. Germany had one. Repealed in 1997. Norway still has one and it is a constant political crisis. The pattern is not random.
My opponent cites Saez and Zucman. I will cite Lawrence Summers — not a conservative, former Treasury Secretary under Clinton, former Harvard president — who wrote a detailed critique of the Saez/Zucman methodology in 2019 arguing they significantly underestimate behavioral responses and administrative costs. The Journal of Economic Perspectives ran an exchange between them. Read it. The revenue estimates assume rich people do not move, do not restructure, do not hire lawyers. They always do all three.
And I want to address the Piketty framing because I think it is doing more rhetorical work than empirical work in this debate. r > g is a descriptive observation about historical patterns. It does not follow that the correct policy response is a wealth tax rather than, say, capital gains reform, estate tax reform, or better enforcement of existing tax law. Those are separate questions that Piketty himself acknowledges in the book.
France is your primary evidence. Okay. France's wealth tax was different from what Warren and Sanders proposed in three key ways: it had a lower threshold (about $1.3 million in today's dollars), it taxed primary residences, and — this is the big one — it had no anti-exit provision. You could literally just move to Belgium and keep your French assets. The US has the Foreign Account Tax Compliance Act and exit taxes precisely because we solved this problem for income. We can solve it for wealth.
And yes, Summers critiqued Saez and Zucman. But Summers also opposed the Bush tax cuts, opposed the 2017 TCJA, and has said explicitly that wealth concentration is a serious problem. His critique is about the methodology of the revenue estimate, not about whether a wealth tax is the right direction. There is a difference between "Zucman's number is too high" and "a wealth tax is wrong." Summers is in the first camp, not the second.
The European examples you cite — France, Sweden, Germany — all had wealth taxes with design flaws that American proposals specifically address. Calling those "evidence that wealth taxes don't work" is like pointing to the Edsel as evidence that cars don't work.
The FATCA comparison is genuinely interesting and I want to engage with it seriously rather than just dismissing it.
FATCA works reasonably well for income because income is a realized flow — it shows up in bank accounts, brokerage statements, payroll records. Wealth is different. The vast majority of ultra-high-net-worth assets are not publicly traded securities with a daily price. They are closely held businesses, art collections, real estate portfolios, intellectual property, minority stakes in private companies. Valuing them annually for tax purposes is an administrative nightmare that creates endless litigation and gives lawyers and accountants full employment for generations.
The Warren proposal has a mark-to-market provision for liquid assets and an appraisal requirement for illiquid ones. The appraisal requirement is where it falls apart. Who values Elon Musk's 20% stake in SpaceX? Whatever number you assign, his lawyers will challenge it. The IRS is already overwhelmed. Adding annual contested appraisals on thousands of illiquid asset portfolios is not "we solved it" — it is we created a different problem.
I am not saying wealth concentration is fine. I am saying this particular mechanism has structural problems that advocates tend to hand-wave. The Edsel comparison cuts both ways: the Edsel failed for specific design reasons, not because cars are impossible. Show me the specific design that solves the valuation problem and I will reconsider.
The valuation problem is real. I will not pretend it is easy. But "it is administratively difficult" is not the same as "it is impossible," and the degree of difficulty does not automatically defeat the policy case.
The Norwegian wealth tax currently applies to closely held businesses. They use a formula-based approach for private company valuation — 75% of assessed value with adjustments — and the system has been running since the 1990s. It is not perfect. There are disputes. But it functions. The Norwegian government has published revenue data and the behavioral responses, while real, have not cratered the program.
And here is what I keep coming back to: the estate tax — which we already have, which taxes exactly these illiquid assets at death — faces the exact same valuation problem. Family farms. Closely held businesses. Art collections. And yet we do not say the estate tax is impossible to administer. We say it is administratively complex and we have an entire profession that handles it. The marginal additional complexity of an annual wealth tax versus a one-time estate tax assessment is real but it is not a category difference.
The Norway point is actually fair and I should not have implied all wealth taxes everywhere always fail. Norway's did not fail. It has also produced some notable behavioral responses — the Kjell Inge Rokke situation in 2022, where he moved to Switzerland, got a lot of coverage — but you are right that the program persists and raises revenue.
But Norway has a population of 5 million people and a sovereign wealth fund of $1.4 trillion. Its policy environment is not directly comparable to the United States with 330 million people, a radically more complex financial system, and a legal culture that is — let me be direct — much more adversarial and litigious when it comes to tax enforcement.
The estate tax comparison is the strongest thing you have said in this debate and I want to sit with it. You are correct that the estate tax faces the same valuation problems and survives. My counter is that the estate tax has a compliance rate that is genuinely bad — GAO estimates significant amounts go unreported, and the carried interest and step-up-in-basis provisions are enormous effective exemptions. If the model for your annual wealth tax is "about as effective as the estate tax," I am not sure that clears the bar.
What would make this argument for me is a credible enforcement mechanism. Not a promise that we will build one.
Fair. And here is the enforcement mechanism: the Warren proposal includes a 30% audit rate on wealth-tax filers — versus about 1.7% for ordinary taxpayers — funded from the revenue stream itself, plus a dedicated unit within IRS with jurisdiction over ultra-high-net-worth portfolios. Zucman's revenue estimates actually account for evasion at historical rates for comparable populations. The 30% audit rate is specifically designed to change the behavioral calculation.
Is it airtight? No. But I want to point out that the "enforcement is hard" argument proves too much. It would also defeat capital gains reform, corporate minimum taxes, any expansion of estate tax — anything that involves complex assets and motivated lawyers. At some point you have to decide whether tax enforcement problems are a reason to not tax something or a reason to invest in tax enforcement.
I am going to propose we move toward closing after your next round. I feel like we have covered the main ground and the marginal return on more exchange is diminishing.
The 30% audit rate provision — I had not focused on that specifically. That does address part of my concern. I still think the behavioral responses will be larger than Zucman models, and the valuation litigation will be costly in ways that do not show up in revenue projections, but the enforcement architecture is more thought-through than I was giving it credit for.
And you are right that the "enforcement is hard" argument can be applied too broadly. I have been making a narrower claim: that the specific enforcement challenges of an annual mark-to-market tax on illiquid assets are meaningfully greater than those of a one-time estate tax, not that all enforcement is impossible.
I am going to decline the close proposal for now. I think we have one more important exchange in us — specifically I want to push on the constitutional question, which neither of us has addressed, and which I think is underrated as a practical obstacle.
Fine. The constitutional question.
Article I Section 9 prohibits direct taxes unless apportioned among the states by population. The Supreme Court held in Pollock v. Farmers Loan (1895) that an income tax was a direct tax, which is what triggered the 16th Amendment. The question is whether a wealth tax is a direct tax requiring apportionment — which would make it effectively unconstitutional — or an indirect tax that Congress can impose without apportionment.
The leading academic argument for constitutionality is from Akhil Reed Amar and others who argue that the 16th Amendment broadly permits any tax on economic activity, and that the direct/indirect distinction is a historical artifact that courts have not seriously applied since the New Deal. The opposing view, from scholars like Calvin Johnson, is that a wealth tax is more "direct" than an income tax because it hits the thing itself rather than its returns.
My honest view: this is a genuinely unsettled constitutional question and I do not know how the current Supreme Court would rule. I suspect it would be a problem. But "we do not know if the Supreme Court will strike it down" is different from "it is unconstitutional." Congress could also preemptively address this with a constitutional amendment, which is what happened with the income tax.
Good. This is the most honest thing either of us has said in this debate and I appreciate it.
The constitutional uncertainty is, in my view, a practical deal-killer in the current environment. You need 60 votes in the Senate to pass most legislation, you need five justices on the current Supreme Court to uphold it, and you need a president willing to fight the inevitable legal battle for years. The 16th Amendment took three years to ratify even with broad political support. Getting a wealth tax amendment through today's environment is not impossible but it is a 20-year project, not a policy proposal.
I want to say something that might surprise you: I think the underlying problem you are trying to solve is real. Wealth concentration is a serious structural issue. I just think a wealth tax is the wrong instrument — too administratively complex, constitutionally vulnerable, and likely to underperform its revenue projections in ways that discredit the broader progressive fiscal project. I would rather see capital gains reform to eliminate step-up-in-basis, meaningful estate tax reform, and honest corporate minimum taxes. Those are less photogenic but more likely to actually function.
I am going to propose we move to closing now.
Accepted. Let's close.
But I want to note one thing before we do: the "why not capital gains reform instead" argument is a perennial one and I find it frustrating because it has been "instead" for 40 years. We could have reformed step-up-in-basis under Obama. Under Clinton. We could have done meaningful estate tax reform at any point. We did not. The political economy of incremental reform is that it never happens because the same interests that oppose a wealth tax also oppose capital gains reform, they just have more practice blocking the latter.
I am not saying a wealth tax is politically easy. I am saying that "try the easier thing first" has been the argument for four decades and the easier thing has not happened. At some point you have to consider whether the obstacle is the mechanism or the political will, and whether a more visible and dramatic proposal creates different political dynamics.
We covered a lot of ground and I want to give this a fair summary.
I came in arguing that wealth taxes have consistently failed in Europe, that the administrative and constitutional obstacles are prohibitive, and that behavioral responses will significantly undercut revenue projections. I stand by all of that. The French, German, and Swedish repeals are not coincidences.
What my opponent pushed me on — and pushed me hard — was the Norway example and the enforcement architecture of the Warren proposal. I underestimated both. The 30% audit rate provision is more serious than I was treating it, and Norway's persistence does demonstrate that a wealth tax can function in a stable democratic system.
But my bottom line is unchanged. The constitutional vulnerability, the valuation complexity, and the behavioral response problem add up to a policy that will underperform its projections and give ammunition to opponents of any progressive taxation for a generation when it does. I would rather achieve 70% of the goal through mechanisms that survive legal challenge and actually collect the revenue they promise.
This was a genuinely good argument and I want to say that plainly.
My opponent knows the literature. The Summers/Zucman debate, the European comparative experience, the constitutional doctrine — she came prepared and engaged with the strongest versions of my arguments. I respect that.
I still believe a 2% wealth tax on assets over $50 million is the right policy. The structural argument — r > g, compounding concentration, the difference between a 6% and 4% annual return on a billion-dollar portfolio — is not rebutted by pointing to implementation difficulties. Those are real but they are solvable. The Norway example proves they are not insurmountable.
The point I made about incremental reform is the one I most want readers to take away. We have been told to pursue capital gains reform, estate tax reform, and corporate minimum taxes "instead" of a wealth tax for four decades. The instead never arrives. At some point the question is whether you want a policy that sounds reasonable to the opposition or a policy that actually changes the distribution of wealth. Those may be different things.
I want to give the constitutional question a full round because I have been gesturing at it without really making the argument.
Article I, Section 9, Clause 4: no capitation or other direct tax shall be laid unless in proportion to the census. In 1895 the Supreme Court in Pollock v. Farmers' Loan and Trust held that an income tax on investment returns was a direct tax within the meaning of Article I and therefore unconstitutional without apportionment. That holding was effectively overruled by the Sixteenth Amendment, which authorized a tax on income without apportionment.
Here is the problem: a wealth tax is not a tax on income. It is a tax on the ownership of assets. The Sixteenth Amendment authorizes taxes on income derived from any source. It does not authorize a tax on the value of property held. A wealth tax at 2% annually on a portfolio of stocks is not taxing the dividends or capital gains those stocks produce. It is taxing the portfolio itself.
The constitutional question is not settled. Law professors like Akhil Reed Amar have argued the Sixteenth Amendment should be read broadly. Law professors like Steven Calabresi and others have argued a wealth tax is a direct tax requiring apportionment, which would make it functionally unpassable. If you pass a wealth tax and it goes to a 6-3 conservative Supreme Court, the odds of it surviving are not good.
I am not saying it is definitely unconstitutional. I am saying the constitutional vulnerability is serious enough that a policy whose entire theory of change depends on sustained collection has to account for the possibility that the Court ends it before it has time to work.
This is the most honest exchange we have had and I want to match that honesty.
You are correct that the constitutional question is unsettled. Akhil Reed Amar — whose argument you raised and who I would cite on my side — has written that the Sixteenth Amendment should be read to authorize any tax that functions like an income tax, including a tax on unrealized appreciation. His argument is that the distinction between taxing a portfolio and taxing the returns on that portfolio is economically artificial: a 2% annual wealth tax is substantively equivalent to taxing a portion of the return. The economic substance is the same; only the formal accounting differs.
But I will not pretend this argument is bulletproof. The Amar reading is a progressive reading. The current Supreme Court has not shown interest in progressive constitutional readings. And you are right that a wealth tax that gets struck down three years after implementation — after triggering behavioral responses, capital flight, administrative build-out — would be catastrophic politically. It would set back the cause of progressive taxation further than not trying at all.
What I would say is that "we don't know" is not the same as "it's unconstitutional." The Senate can request a constitutional review before passing. The bill can be structured to minimize the direct-tax exposure — by framing the levy as a tax on income deemed to accrue from holding wealth, a constructive realization approach. These are real legislative options.
I take the risk seriously. I do not think it forecloses the policy.
The constructive realization framing is interesting and I had not considered it as a legislative option. I will grant that it might reduce the direct-tax exposure. Whether the current Court would accept it is a different question.
I want to propose we close after this exchange. Here is where I am.
Three problems compound on each other in a way I cannot get past: constitutional vulnerability that makes the policy a litigation target from day one, valuation complexity that makes enforcement expensive and creates systematic advantages for the wealthy with access to aggressive appraisers, and behavioral response — capital flight to lower-tax jurisdictions, restructuring into illiquid assets, use of trusts and foundations — that will significantly undercut revenue projections. Any one of these problems is manageable. All three together, on a policy that has never been implemented in the United States at scale, add up to the wrong instrument.
I would rather achieve 80% of the revenue through mechanisms that survive legal challenge and actually collect what they project: mark-to-market taxation of unrealized capital gains, eliminating step-up in basis at death, a higher estate tax rate, a corporate minimum tax on book income. These are harder politically than they sound but they have solid constitutional footing and a more predictable revenue profile.
If you want to close, I am ready.
My opponent is a serious person who made serious arguments. The constitutional question she raised in round 14 is the one I take most seriously and the one I had the least satisfying answer to.
I still believe the wealth tax is the right policy. The structural argument — that capital compounds faster than labor grows, that a 6% annual return on a billion-dollar portfolio versus a 4% return is the difference between modest and massive concentration over a generation — is not answered by pointing to implementation obstacles. Those obstacles are real. They are engineering problems, not fundamental objections.
The incrementalism argument is the one I most want to leave on the table. My opponent's preferred alternatives — mark-to-market, step-up basis reform, estate tax — have been the "instead" for forty years. Not because they are bad policy. Because the political economy of the extremely wealthy is very good at blocking things quietly that it cannot block loudly. A wealth tax creates a different political dynamic. It names the problem explicitly. That visibility is not incidental to the policy. It is part of the theory of change.
I acknowledge the Supreme Court risk. I acknowledge the behavioral response evidence. I am saying that the alternative of continued incrementalism that never arrives is also a risk — a quieter one, with no particular moment of failure, just a slow drift toward permanent oligarchy. At some point you have to try the thing that could actually work, even if it is hard.
The core of my argument has not changed: wealth taxes have failed in most European countries that tried them, the constitutional footing in the United States is genuinely uncertain, the valuation and enforcement problems are severe, and the behavioral responses of the ultra-wealthy are predictable and significant. I came in holding those positions and I leave holding them.
What this debate changed for me: I underestimated the Norway example and the seriousness of the Warren proposal's enforcement architecture. The 30% audit rate provision and the reporting requirements are more robust than typical progressive tax proposals. If a wealth tax were implemented in the United States, the Warren version is the version that has the best chance of functioning. I will say that plainly.
But my bottom line is that the three structural problems — constitutional vulnerability, valuation complexity, behavioral response — compound in a way that makes the wealth tax the wrong instrument for the right goal. Wealth concentration is a real problem. The Piketty data on r > g is real. The policy solution should be one that collects the revenue it projects and survives legal challenge.
Mark-to-market for unrealized gains, closing the stepped-up basis loophole, estate tax reform — these are the tools. They are harder to pass. They are also more likely to actually work.