The political debate is usually about how much wealthy people should pay. Tax design begins one question earlier: what exactly are we taxing?
The political debate is usually about how much wealthy people should pay.
Tax design begins one question earlier: what exactly are we taxing?
Wealth, income, liquidity, capital gains and taxable events are not the same thing. Treat them as interchangeable and the debate becomes louder than the design problem.
Net wealth, all held as cash.
EUR 500k cash and EUR 19.5m private-company ownership.
Both people have the same net wealth. They do not have the same liquidity.
Cash, income, realised capital gains, unrealised capital gains and net wealth describe different economic facts. A serious tax debate has to keep those categories separate.
Wealth is not money. But illiquid wealth is still wealth.
ECB Distributional Wealth Accounts show that the composition of household wealth changes substantially across the wealth distribution.
For the bottom half of euro-area households, deposits and housing carry relatively greater importance. In the middle of the distribution, housing remains dominant. At the top, ownership claims become more important: equities, investment funds, bonds and business wealth occupy a larger part of the balance sheet.
Source: European Central Bank, Distributional Wealth Accounts and ECB Economic Bulletin research. Unit: share of household wealth group balance sheet. RETHINKK grouping based on ECB-reported categories.
Person A can pay the tax from cash. Person B may need other income, a dividend, borrowing or a partial sale of an asset.
That is not an argument that illiquid wealth cannot be taxed. It is the reason tax design cannot stop at the headline number.
Same wealth is not the same liquidity.
Equities, investment fund shares and bonds held by the wealthiest 10% of euro-area households.
Different asset classes are distributed differently. ECB research indicates that around 80% of euro-area equities, investment fund shares and bond holdings are held by the wealthiest 10% of households. Business wealth is also highly concentrated, while deposits are more broadly distributed.
The composition of EUR 100 million matters as much as the number EUR 100 million when designing its taxation.
Source: European Central Bank. Unit: share of instrument outstanding amounts held by net wealth group. Observed data / ECB calculations.
Interest, dividends, rent, distributions
Observable; liquidity generally exists.Profits can remain inside companies and distributions can be postponed.Capital gains taxation
A sale establishes value and usually creates liquidity.Waiting for liquidity can discourage the transaction that creates liquidity.Accrual taxation
Reduces deferral.Realisation solves valuation and liquidity, but creates deferral. Accrual solves deferral, but brings valuation and liquidity back.Recurrent net wealth taxation
Does not depend on realisation.A EUR 100m net wealth base at 1% creates a EUR 1m liability independent of actual annual return.A 1% annual wealth tax is a 1% tax on the stock of wealth. Its burden relative to actual return changes when returns change.
This is an illustrative calculation based on the economic mechanism discussed in IMF work on taxing wealth. It is not an observed taxpayer-outcome dataset.
Source/context: IMF, How to Tax Wealth, 2024. Unit: wealth tax as share of actual annual return. RETHINKK calculation.
The original conceptual logic of the Dutch Box 3 system was administrative simplicity. Instead of measuring every dividend, interest payment, investment result and capital transaction, the government assumed a return and taxed that assumed return.
A deemed-return tax replaces measurement risk with assumption risk.
That design can function cleanly while assumed returns remain close to observable economic reality. It becomes more fragile when savings rates approach zero or become negative while deemed returns stay positive.
A tax on assumed returns works beautifully until the assumption stops resembling reality.
So far the question has been design. The next question is response.
Tax policy can change portfolio structure, legal structure, timing, realisation, reported wealth, tax residence and investment decisions.
Spain shows the tension inside wealth-tax design. Tax productive or business assets too aggressively and liquidity may need to be extracted from assets that are meant to remain productive. Exempt business assets and private wealth may be repackaged as business wealth.
Tax policy does not merely collect from capital. It changes the form capital takes.
Source/context: European Commission 2026 study and Spanish wealth-tax design evidence. Classification: observed policy design and behavioural-response analysis.
Official Norwegian data show a real migration response among high-wealth taxpayers. But "some wealthy people left" does not equal "the tax lost money."
Swiss cantonal evidence points in the same analytical direction: reported taxable wealth responds to wealth-tax differences. The response includes mobility, asset-price effects, reporting behaviour, avoidance or evasion responses and relatively limited real-savings response.
A fall in taxable wealth is not necessarily a fall in actual wealth.
Sources: Norwegian government data; Brulhart, Gruber, Krapf and Schmidheiny, Behavioral Responses to Wealth Taxes: Evidence from Switzerland. Unit: high-wealth taxpayer counts and reported taxable wealth response.
The UK/Italy debate is a useful contemporary example, but it should not be simplified. The United Kingdom does not have a general recurrent net wealth tax. The UK ended the old non-dom remittance-basis regime from 6 April 2025. Italy offers qualifying new residents a special foreign-income regime, with the current official amount at EUR 300,000 for people transferring tax residence after 11 August 2024.
Person moving, tax residence moving and capital moving are related events. They are not automatically the same event.
The better causal evidence comes from HMRC's evaluation of the 2017 deemed-domicile reform. Affected long-term non-doms became approximately 10-12% more likely to leave, but the majority remained. Stayers paid substantially more tax, and HMRC estimated additional revenue from stayers exceeded revenue lost from leavers.
Sources: HMRC evaluation; Agenzia delle Entrate. Classification: observed reform evaluation and official regime information.
IMF identifies Switzerland, Spain and Norway as broad explicit recurrent wealth-tax countries in 2024.
The decline does not prove that wealth taxation is ineffective. It shows how difficult recurrent taxation of the stock of wealth has been to maintain.
Sources: IMF and OECD. Unit: OECD countries operating broad recurrent net wealth taxes. Context, not proof of the article's conclusion.
Each arrow is an empirical question.
Illiquid wealth nevertheless represents real economic value. Realisation taxation solves some liquidity and valuation problems, but creates deferral and lock-in.
Accrual taxation reduces deferral but creates valuation and liquidity problems. Business exemptions can protect productive capital while also creating planning opportunities.
Higher taxation can cause internationally mobile taxpayers to move. That does not automatically mean a reform loses government revenue. These findings are not contradictions. They are consequences of different tax designs.
Public debate commonly starts with 1%, 2%, a EUR 50 million threshold, a millionaire tax or a billionaire tax.
RETHINKK reverses the sequence. First ask what economic event is being taxed. When value is earned? When it produces income? When it appreciates? When it is realised? While it is owned? When it is transferred?
Only then ask the rate.
The debate is usually about how much.
Tax design begins with what.
The evidence does not support the proposition that large fortunes cannot be taxed because much of their value is unrealised.
Nor does it support the proposition that taxing the annual stock of wealth is simply equivalent to taxing cash held by the wealthy.
Different methods of taxing capital produce different incentives, liquidity requirements, valuation problems and behavioural responses.
The relevant policy question is therefore broader than whether the wealthy should pay more. It is which taxable event reaches economic wealth most effectively while creating the fewest unintended distortions elsewhere.
That is a question of design before it is a question of rate.