The Wealth Delta Tax: Intellectual Background and Reference Guide
Wealth Delta Tax, wealth taxation, tax theory, Haig-Simons income, accrual taxation, Domar-Musgrave, Harberger taxation, self-assessed taxation, cooperative compliance, public finance, wealth inequality, democratic legitimacy, international tax coordination
Revision History
| Revision | Date | Details |
|---|---|---|
| 0.01 | 2 August 2026 | First Draft |
| 1.00 | 15 August 2026 | Published to website |
| 1.01 | 20 September 2026 | Crosslinks to WFR added in §1 (new papers note), §3, §4, §5, §6 forward notes; §1 note added directing readers to INST, FAL, and LDW as papers published after the original forward-note apparatus |
Abstract
This paper is the reference guide to the intellectual ancestry of the Wealth Delta Tax (WDT). Its nineteen sections cover the bodies of scholarship the WDT project draws on, organised around the problems those bodies address rather than by intellectual tradition. Each section is self-contained: it states what the relevant literature says, where its findings are robust, where they are contested, and where they stop. A forward note at the end of each section directs the reader to the WDT companion papers that develop or respond to the ideas covered there. The paper does not make arguments of its own, does not advocate for the WDT, and does not summarise what the WDT does. It provides the source treatment for external references cited across the companion paper series, in one place, at a level of detail the companion papers themselves do not carry. Readers who arrive from the literature and want to know where the WDT engaged with a particular body of work should use this document as their navigation point. Readers who want to understand what open questions remain in the existing scholarship should consult the companion gap register (LR.A).
Glossary
Accrual-basis taxation: Taxation of economic gains as they arise, regardless of whether a realising transaction has occurred. Contrasted with realisation-basis taxation, which taxes gains only when they are converted to cash through sale or exchange.
Chamley-Judd result: The theoretical finding, established independently by Chamley (1986) and Judd (1985), that the optimal long-run tax rate on capital income in a general equilibrium model is zero. The result has been substantially qualified by subsequent work, particularly Straub & Werning (2020).
Codetermination: A corporate governance arrangement, most developed in German law, under which employees hold seats on a company’s supervisory board alongside shareholder representatives.
Cross-base externality: The fiscal revenue loss to taxes other than the wealth tax itself — principally income tax and VAT — when taxpayers depart a jurisdiction in response to wealth taxation. Documented quantitatively by Agrawal et al. (2025).
Delta base: A tax base defined as the annual change in net worth, rather than the stock of wealth or only realised gains.
Domar-Musgrave mechanism: The property, established by Domar & Musgrave (1944), that symmetric government participation in investment outcomes — taxing gains and refunding losses at the same rate — reduces the after-tax variance of risky investment without a proportional reduction in expected return.
Fiscal rule: A statutory or constitutional constraint on a government’s fiscal choices, typically expressed as a limit on the deficit, debt, or rate of expenditure growth.
Haig-Simons definition: The definition of income as the sum of consumption and the change in net worth over a period, developed independently by Haig (1921) and Simons (1938). The most widely accepted theoretical baseline in tax economics.
Harberger mechanism: The self-assessment enforcement device proposed by Harberger (1965): a taxpayer who declares a value pays a levy on it but must also accept a standing obligation to sell at that declared price, creating an incentive to declare honest values without requiring independent state appraisal.
Lock-in distortion: The incentive created by realisation-basis capital gains taxation to hold appreciated assets indefinitely rather than sell them, because sale triggers the tax liability while continued holding defers it.
Minimum-tax floor: A coordinated international standard setting a minimum effective tax rate on ultra-high-net-worth individual wealth, as proposed in Zucman (2024) and under development in the UN Tax Committee model law process.
Procedural justice: The principle, developed most fully by Tyler (1990), that people’s willingness to comply with rules and authority depends substantially on their perception of the process as fair, regardless of the substantive outcome.
Slippery slope framework: Kirchler’s (2007) model of tax compliance in which the equilibrium compliance rate is jointly determined by the authority’s enforcement power and taxpayers’ trust in the authority. High trust supports voluntary compliance; low trust requires high enforcement power to produce equivalent outcomes.
Sovereign wealth fund: A state-owned investment fund, capitalised from government revenues, managed at institutional arm’s length from the fiscal ministry. Examples include Norway’s Government Pension Fund Global, Singapore’s GIC and Temasek, and Chile’s Pension Reserve Fund.
Use-it-or-lose-it mechanism: The efficiency argument for wealth taxation when returns on capital are persistently heterogeneous, formalised by Guvenen et al. (2023): a wealth tax falls equally on all holders of equivalent wealth regardless of return, shifting the relative burden toward unproductive holders and improving aggregate capital allocation compared to a capital income tax.
1. Introduction and How to Use This Document
Each section covers what a body of literature says on a specific problem. Sections do not depend on each other; the document is organised for selective reading by someone who arrives from the literature wanting to understand where the project engaged with a particular strand of thought.
The structure is problem-based rather than tradition-based. The same scholar may appear in multiple sections because their work bears on multiple problems; the same intellectual tradition may be split across sections because different aspects of it bear on different design questions. The organisation by problem reflects the way the companion papers use the literature: as resources addressing specific analytical challenges, not as intellectual debts to be discharged in a fixed order.
Each section ends with a forward note stating which companion papers develop or respond to the ideas covered. The note is a navigation aid, not a summary.
The document is the publicly readable version of the source material in the project’s master reference database (0.4). It is generated from 0.4 after that database is clean; it is not drafted in parallel with 0.4. All companion papers cite this document for their external references rather than carrying independent literature surveys. The gap register (LR.A) is the companion document: where this paper covers what the literature says, LR.A covers where the literature stops and what the WDT project does in the absence of an answer.
Note on papers published after this document’s forward-note apparatus was completed. Three companion papers published after LR.B v1.00 extend bodies of literature covered here in ways the per-section forward notes do not yet reflect. (INST) extends the political durability literature surveyed in (LR.B §17) by providing an independent derivation of the three failure mechanisms and a structural account of why the WDT’s design properties resist them; it is the natural downstream reading from (LR.B §17) alongside (POL). (FAL) addresses the body of literature on institutional failure and systemic risk that (LR.B §14) and (LR.B §15) touch on in the SWF and precommitment contexts; it develops the WDT-specific failure taxonomy and the Mode A/B/C/D typology. (LDW) extends the labour tax incidence and household welfare literature implicit in (LR.B §2) by tracing the purchasing power consequences for the non-taxed majority of a mature WDT operating under full labour and consumption tax displacement; it is the companion reading to (ENV) for readers interested in distributional welfare effects on ordinary households rather than on the taxed population.
2. The Tax Base Question
Three bases dominate the policy literature on tax design: income, consumption, and wealth.
Haig (1921) and Simons (1938), working independently, defined income for tax purposes as the sum of consumption and the change in net worth over a period. Any increase in command over economic resources is income, regardless of whether it has been realised. The Haig-Simons definition remains the most widely accepted theoretical baseline in tax economics. The practical divergence between Haig-Simons income and the taxable income assessed in most systems is administrative in origin: unrealised appreciation is excluded from the base on enforcement grounds, not on principled grounds.
The consumption tax tradition, developed most fully by Kaldor (1955) and Bradford (1986), starts from a different position. A person who earns income and saves it is deferring consumption. Taxing the saving imposes a second layer of tax on the same resources. Kaldor’s expenditure tax operationalised the consumption base by taxing only what households actually spend. Bradford’s X Tax reformulated it as a business cash-flow tax plus a wage tax, avoiding the measurement problem that arises when households hold wealth in illiquid forms. Viard & Carroll (2012) defend the X Tax as a practical framework, arguing that a progressive consumption base can meet distributional objectives without the efficiency costs they attribute to capital income taxation.
The two bases reflect different judgments about what the tax system is measuring. The Haig-Simons definition measures the annual accretion of economic power, including unrealised appreciation. The consumption base measures what resources are withdrawn from the productive economy for final use. They produce equivalent tax burdens only when capital returns are certain and uniform across investors; when returns are uncertain or heterogeneous, they diverge substantially in their distributional and efficiency implications.
A third possibility — taxing the stock of accumulated wealth rather than annual income or consumption — sits between the two traditional positions. A stock wealth tax captures the accumulated result of prior income and saving decisions and imposes a burden even on wealth that generates no current income. The empirical literature on stock wealth taxes, surveyed in (LR.B §6), is the most directly relevant body of evidence for the WDT even though the WDT taxes annual changes rather than the stock.
In the WDT project: The WDT uses a delta base — the annual change in net worth — which is Haig-Simons income measured at the individual level but assessed on a net worth basis rather than an income-flow basis. The choice of delta over stock is treated in (WP §3.1) and (MF §3). The challenge to the consumption tax framing at extreme wealth levels is developed in (MF §4).
3. Unrealised Gains and Prior Accrual Proposals
The academic literature on taxing unrealised gains has a long history, but most of it arrives at the valuation problem and then either stops or defers.
Shakow (1986) argued that the failure to tax unrealised appreciation is administrative rather than principled: it reflects the difficulty of valuing illiquid assets annually, not any normative reason to exclude appreciation from the base. He proposed mechanisms for dealing with the valuation problem on a class-by-class basis, but left losses unaddressed. Gains were taxed as they arose; losses generated no equivalent refund. Shakow acknowledged this asymmetry as a limitation but did not design around it, placing the loss treatment problem explicitly in the literature as a named gap rather than an oversight.
Auerbach (1991) approached the same problem from a different angle. His retrospective capital gains tax eliminates the lock-in distortion without requiring annual mark-to-market: gains are taxed at realisation, but an interest charge approximates what annual accrual taxation would have collected during the holding period. The mathematical equivalence to accrual taxation holds under idealised conditions. The limitation is that realisation remains a precondition. Assets held until death, transferred through trust structures, or reorganised to avoid a taxable sale can still escape the charge. The WDT does not require a realisation event.
Vickrey (1939) proposed income averaging to address the progressive rate bunching problem: the tendency of lump-sum income events to attract disproportionately high marginal rates relative to the same income spread evenly over time. The WDT’s assessment window mechanism is related in spirit but distinct in purpose — it addresses single-year valuation volatility for illiquid assets rather than rate bunching on earned income.
Cnossen & Bovenberg (2001) is the closest intellectual ancestor. Developing a capital accretion tax for Dutch tax reform, they tax accrued capital gains annually on a mark-to-market basis, treat full loss offsets as integral, and work through valuation, liquidity, illiquid-asset, corporate, and international problems in detail. They recognise the Domar-Musgrave risk-pooling implications of government loss-sharing. Their conclusion, however, diverges from the WDT’s: annual accrual is theoretically superior, but realisation is necessary for assets that resist periodic valuation. Their preferred system is a hybrid — accrual for easily valued financial assets, realisation with interest for hard-to-value assets, a presumptive net wealth tax for the remainder. The illiquid-asset problem is set aside rather than solved.
Hasen (2017) proposes a progressive accretion wealth tax as a supplement to existing federal taxes, arguing that wealth concentration creates negative externalities from the mere potential power that concentrated wealth confers, independently of whether it is spent. He distinguishes accretion from excise taxation, argues that progressive rates make accretion bases superior on timing grounds, and engages the valuation problem directly, acknowledging periodic assessment of illiquid assets as a known constraint without designing around it. His tax falls on total wealth holdings assessed periodically, not on the annual change in net worth; the delta is not his organising unit. Loss treatment is absent: the paper contains no discussion of wealth declines, no symmetric government participation in downside risk, and no loss offset mechanism. The proposal is framed as one instrument in an optimal tax mix, not a replacement architecture. No valuation system, settlement mechanics, governance structure, or fiscal transition framework is developed.
The OECD’s 2018 study The Role and Design of Net Wealth Taxes uses the phrase “wealth accretion tax” explicitly and describes taxing annual changes in household wealth on an accrual or mark-to-market basis, with full loss offsets permitted. The passage is one paragraph, treating the concept as one alternative among several approaches to taxing capital income. The idea is named; it is not developed.
The WDT’s departure from this prior work is specific. Cnossen and Bovenberg treat valuation difficulty for illiquid assets as a principled limit on accrual taxation. The WDT treats it as a design problem and constructs the four-route architecture to solve it. The reframing in (VAL) — from how the state discovers the correct value to under what conditions understatement ceases to be the rational strategy — is the institutional expression of that fork. A second divergence concerns loss treatment: the prior literature accommodates loss offsets as a practical feature of an income-tax framework; the WDT derives the symmetric refund from reciprocity as a foundational property of the mechanism and capitalises the resulting liability through a dedicated SWF. That derivation and the capitalisation architecture it produces have no identified predecessor in the accrual taxation literature.
The most prominent recent US proposals in this tradition are the Wyden “Treat Wealth Like Wages” white paper (2019) and the 2021 Billionaires Income Tax legislative text, which impose mark-to-market treatment on tradeable assets held by high-income taxpayers, with a lookback charge on non-tradeable assets at realisation. The Biden administration’s 2022 Billionaire Minimum Income Tax proposal included a prepayment mechanism on unrealised gains with, in some circumstances, partial refunds. The structural differences from the WDT are considerable: these proposals operate as minimum tax overlays within income tax frameworks, contain no pre-funded reserve, and carry no governance participation by the taxed population. Miller (2005, subsequent work) proposes a progressive mark-to-market tax for the wealthiest taxpayers on publicly traded securities and specified other assets. Toder & Viard (2016) propose replacing corporate income tax revenues with mark-to-market taxation of shareholder income on accrued gains in publicly traded corporate shares, with averaging provisions for volatility. Both are relevant to the WDT’s corporate architecture but neither takes the individual’s total annual change in net worth as the organising unit of assessment.
Contemporary legal scholarship on mark-to-market reform (Shaviro, 1989, 2021), Weisbach (2000), Batchelder & Kamin (2019), Avi-Yonah & Mazzoni (2019) engages the administrative and structural problems the Haig-Simons tradition has left open. Batchelder & Kamin (2019) explicitly assess mark-to-market approaches and conclude that the loss treatment question remains unresolved as a design matter. Avi-Yonah & Mazzoni (2019) similarly identify symmetric loss treatment as a necessary feature of any accrual-basis system while offering no design solution. Delmotte (2024), in the current statement of the US legal academy’s anti-annual-wealth-tax position, proposes a deemed-realisation tax at death: a realisation-basis approach without retrospective interest charges and without symmetric loss treatment.
Zelenak (2021) addressed specifically whether mark-to-market losses should generate cash refunds and recommended against them, primarily on political grounds regarding the public reception of large payments to wealthy individuals in down years. His analysis confirms that the cash refund question had been identified in the policy literature and set aside without a design response.
No prior proposal takes the individual’s total annual change in net worth as the organising unit of a complete tax-and-fiscal architecture. The accrual taxation tradition treats annual wealth changes as a method for capturing income more completely within an income-tax framework. The WDT treats the annual net worth delta as the base unit from which valuation architecture, loss treatment, settlement mechanics, corporate attribution, governance, and fiscal transition are all derived. That is a different intellectual project, not a stronger version of the same one.
In the WDT project: The WDT’s response to the loss treatment problem that the prior literature acknowledged but did not resolve — the symmetric refund mechanism, the SWF pre-funding, and the lifetime contribution envelope — is developed in (WP §3.5) and (RATES) The valuation incentive problem that recurs across prior proposals is addressed in (VAL §1) and the mechanism in (VAL §5) and (VAL §7). The most technically serious theoretical objection to accrual taxation — the Arachi & D’Antoni (2022) argument that accrual taxes create intertemporal consumption distortions by billing unrealised gains before the taxpayer has liquidity to pay — is engaged directly in (WFR §4.2.3), which shows that the symmetric refund addresses the loss-year component of the objection for established taxpayers while acknowledging that the objection applies in full at the entry margin where the lifetime contribution envelope floor binds.
4. Symmetric Loss Treatment and Risk-Sharing
Domar & Musgrave (1944) identified a property of tax systems with full loss offsets that has direct consequences for investment incentives. When government participates symmetrically in investment outcomes — taxing gains and refunding losses at the same rate — expected return and variance both fall proportionally. For a risk-averse investor, the effective cost of holding volatile assets decreases, and the distortion that asymmetric taxation imposes on portfolio composition is removed or reduced.
The intuition is clearest in the limiting case. If the government taxes 50% of gains and refunds 50% of losses, the investor’s after-tax position is equivalent to investing half the capital while retaining all the variance reduction from government co-participation. An investor willing to hold a risky asset before the tax may be no less willing after, because the reduction in expected return is matched by an equivalent reduction in variance.
Sandmo (1977) confirmed the same mechanism from a portfolio equilibrium starting point. His analysis of how taxation affects portfolio choice when asset returns are uncertain found that asymmetric tax treatment penalises variance directly: when losses receive no equivalent relief, investors face an environment that makes concentrated, volatile positions relatively more expensive to hold, creating systematic underinvestment in high-risk assets relative to what their risk-adjusted returns would otherwise justify.
Stiglitz (1969) formalised the conditions under which government participation in investment outcomes through taxation can increase risk-taking. His model demonstrates that a proportional income tax with full loss offset can leave the optimal risky portfolio allocation unchanged, because the effective cost of variance falls in proportion to the tax rate.
King (1977) extended the analysis to the cost of capital. When losses generate no relief, the hurdle rate for risky projects rises, and this bias toward lower-variance investment accumulates across the investment universe.
The formal extension of these results to a progressive, delta-based wealth tax has not been accomplished; that extension is identified as a research gap in (LR.A §2.1), and formal numerical modeling can be found in (WFR).
In the WDT project: The symmetric refund mechanism and its grounding in this tradition are developed in (WP §3.5) and (RATES) The full-symmetry design position and the rejection of partial symmetry are stated in (RATES) The moral and foundational grounding for the symmetric refund — as the operational expression of the state’s commitment to downside exposure alongside the taxpayer — is in (MF §7). The formal extension of the Domar-Musgrave result to a progressive delta base — a gap identified in (LR.A §2.1) — is carried out in (WFR §3.2) and (WFR §4.1), which confirm the property holds for the flat symmetric WDT to floating-point precision and establish that the three complications a logistic progressive schedule introduces are all second-order at canonical parameters.
5. Heterogeneous Returns and Efficiency Under a Wealth Tax
The classical analysis of capital income taxation assumes that all investors earn the same rate of return on equivalent capital. Under this assumption, a capital income tax and a wealth tax can always be calibrated to produce equivalent revenue and burden distributions. When returns are persistently heterogeneous across investors, the two tax bases have substantially different efficiency properties.
Fagereng et al. (2020), drawing on twenty years of Norwegian administrative records covering every individual’s full asset portfolio, document that individuals earn persistently different rates of return on net worth, with a cross-sectional standard deviation of approximately 8 percentage points and substantial year-to-year autocorrelation within individuals that survives controls for portfolio composition. The heterogeneity is not a composition effect: individuals earning high returns continue to do so when asset class is held constant.
The wealth-return correlation matters more for the WDT than the heterogeneity finding itself. Moving from the 10th to the 90th percentile of the wealth distribution increases financial asset returns by approximately 3 percentage points, even within asset classes: the upper tail is not only larger in stock terms but, on average, growing faster per unit of wealth.
Guvenen et al. (2023) formalise the efficiency implications in a model calibrated to US data. Under capital income taxation, the tax burden falls disproportionately on productive investors who earn high returns, concentrating the penalty on efficient capital deployment. Under a stock wealth tax, all holders of equivalent wealth pay equivalent tax regardless of their return, shifting the relative burden toward unproductive holders. The efficiency implication is the use-it-or-lose-it mechanism: productive capital gradually concentrates with investors who earn more from it. In their model, the welfare gain from replacing capital income tax with a revenue-neutral stock wealth tax is approximately 8% in consumption-equivalent terms.
The Guvenen et al. result applies to a stock wealth tax. A delta-based tax changes the structure of the efficiency argument: a productive entrepreneur whose wealth position is stable pays nothing under a delta base even if their holdings are large and their return is high. Whether the use-it-or-lose-it mechanism operates more or less strongly under a delta base than a stock base has not been formally modelled. Formal numerical modeling can be found in (WFR).
Piketty et al. (2023), in a survey paper on capital and wealth taxation, identify three rationales for wealth taxation: the increasingly difficult distinction between capital and labour income at the top of the distribution; the inadequacy of income and consumption as measures of economic wellbeing where capital gains dwarf ordinary flows; and the stronger meritocratic case for taxing inherited wealth than earned wealth. Blanchet & Martínez-Toledano (2023) examine wealth inequality dynamics in Europe and the United States, finding that differing institutional arrangements — particularly differences in pension fund governance and housing market structure — explain much of the cross-country variation in wealth concentration patterns.
In the WDT project: The growth-tier model in (RATES §5.2) is calibrated to the Fagereng et al. (2020) persistence and dispersion estimates. The claim that the tier assignments likely overstate the underperforming share of the WDT population is grounded in the positive wealth-return correlation documented by Fagereng et al. The efficiency case for a delta base relative to the stock wealth tax baseline is noted but not formally extended in (WP §2.5). The distributional concentration of the delta tax base is treated in (RATES §5.1). The formal concentration arithmetic across a four-tier Fagereng calibration — a gap identified in (LR.A §2.3) — is carried out in (WFR §4.3), which traces the Great/Poor wealth ratio under all six candidate tax systems over a 30-year horizon and finds that the dominant split at that horizon is accrual basis versus stock base (286–288× for WDT variants versus 479× for stock-base systems), not flat versus progressive rate.
6. Welfare Comparisons
The welfare case for capital income taxation — or against it — has been the most contested area in public finance theory over the past four decades.
Chamley (1986) and Judd (1985), working independently, showed that in certain general equilibrium models the optimal long-run tax on capital income is zero. The mechanism is accumulation: a capital income tax reduces saving, which reduces the capital stock, which reduces wages in all future periods. These results were influential in establishing the theoretical case against capital income taxation throughout the 1980s and 1990s.
Both results have been substantially qualified. Straub & Werning (2020) showed that the Chamley-Judd zero capital tax result does not hold under more general conditions. In the Judd (1985) model specifically, if the intertemporal elasticity of substitution is below one — which empirical evidence suggests is realistic — the long-run optimal tax on capital is positive and significant. Even in cases where the result holds formally, the economy may take centuries to reach the steady state, and the transition welfare effects may dominate the long-run comparison entirely.
Viard & Carroll (2012) defend the consumption base on distributional grounds, arguing that progressive consumption taxation through the X Tax can meet equity objectives while avoiding what they characterise as the efficiency costs of capital income taxation. Their argument rests on the assumption that saving represents deferred consumption, which the WDT project treats as losing force at extreme wealth levels where holdings may never be meaningfully consumed.
Dias, Iglesias, and Goncalves (2025), using agent-based modelling, find that consumption-based tax systems are structurally regressive in their distributional effects, disproportionately burdening the lowest quintiles of households. Their methodology is not directly comparable to the analytical general equilibrium models above, and the results should be treated as corroborating rather than primary evidence.
The formal welfare comparison between a delta-based wealth tax and the consumption tax alternative has not been done at the level of the WDT’s specific structure. Formal numerical modeling can be found in (WFR).
In the WDT project: The consumption tax challenge is addressed in (WP §8.4). The argument that the consumption base loses force at extreme wealth levels is developed in (MF §4). The formal welfare comparison that includes the delta base alongside all standard candidates at genuine revenue equivalence is in (WFR §3) and (WFR §4). (WFR §5.5) engages the live dispute between Guvenen et al. (2023) and Boadway & Spiritus (2025) on the general-equilibrium capital allocation implications of the use-it-or-lose-it mechanism under a delta base, positioning WFR as a gap-filling rather than side-taking contribution.
7. The Empirical Wealth Tax Record
Several countries have operated annual net wealth taxes, and the empirical literature on their effects is now substantial. The most consistent finding across studies is that the magnitude of behavioural response depends primarily on the quality of third-party reporting and the degree of jurisdictional arbitrage available, not on the headline rate.
Sweden’s wealth tax, studied by Seim (2017) using data from the period before its abolition in 2007, found modest behavioural response elasticities overall, with approximately one third of the measured response reflecting underreporting rather than real changes in saving or investment behaviour. Without the ability to distinguish evasion from real response, apparent elasticities overstate the real economic distortion.
Denmark provides a cleaner picture. Jakobsen et al. (2020) found larger responses at the top of the distribution. Denmark’s strong third-party reporting regime limited evasion substantially, meaning the measured response reflects actual portfolio reallocation rather than underreporting.
Norway provides the clearest evidence on migration responses in a nationally administered system. Jakobsen et al. (2024) studied the period following the 2022 wealth tax increase and found emigration responses that were real but fiscally modest: approximately 22 cents of revenue lost per unit raised, with overall revenues continuing to grow as the remaining base expanded. This figure captures only the direct wealth tax channel; the broader fiscal cost of migration is substantially larger when cross-base effects are included, as documented by Agrawal et al. (2025) and discussed in (LR.B §18).
Switzerland is the structural outlier in the empirical literature. Brülhart et al. (2022) found much larger elasticities than the Scandinavian studies. The explanation is within-country arbitrage: Switzerland’s decentralised cantonal system enables relocation to lower-tax cantons, a form of jurisdictional arbitrage unavailable in nationally administered systems. Swiss findings do not transfer to nationally administered contexts.
Colombia illustrates the reporting infrastructure effect directly. Londoño-Vélez & Avila-Mahecha (2025) found that within the same system and year, assets subject to third-party reporting showed minimal behavioural response while self-reported assets showed high elasticities driven by underreporting.
The wealth concentration data underlying the distributional case for wealth taxation comes primarily from Saez & Zucman (2016) and Piketty (2014), who place the top 1% share of US household wealth at approximately 38%. These figures are contested on methodology grounds by Kopczuk (2015) and Bricker et al. (2016), who argue that the capitalisation method overstates concentration at the top. The direction of the trend — rising concentration since the 1980s — is accepted across methodological approaches. The Saez & Zucman (2019) proposal for a progressive annual stock wealth tax at rates of 1% above $10 million and 2% above $1 billion is the most prominent recent policy comparator for the WDT; its structural differences from the WDT are discussed in (WP §3.1).
In the WDT project: The empirical record informs the RATES revenue model’s pre-behavioural baseline and the BEHAV paper’s identification of reporting infrastructure as a first-order design requirement. (JUR §1.2.1) covers the UK wealth data specifically. The cross-base externality finding from Agrawal et al. (2025) is the dominant qualification on the Norwegian migration estimate; it is treated in (LR.B §18) of this document and identified as an open modelling gap in (LR.A §3.2).
8. The Valuation Problem
The valuation problem for annual accrual taxation of wealth has been acknowledged across the literature for decades without receiving a systematic design response.
Shakow (1986) addressed the problem on a class-by-class basis, noting that private company equity and other illiquid non-fungible assets resisted the solutions available for marketable securities. The subsequent literature (Auerbach (1991), the Wyden proposals, the contemporary mark-to-market scholarship) has not improved materially on this. Asset-class-specific workarounds are available; a general mechanism that removes the rational preference for understatement is not.
The professional valuation literature provides relevant standards and governance frameworks. The International Valuation Standards Council (IVSC, 2022) maintains methodology standards across asset classes, organised around a preference ordering from directly observable market inputs at the reliable end through to unobservable model-dependent inputs at the least reliable end. IFRS 13 and ASC 820 embed a materially similar hierarchy in financial audit standards, requiring more demanding disclosure as observable evidence becomes scarcer. These frameworks are designed to produce a range of acceptable values consistent with the applicable standards; the WDT requires a single assessed value for the delta calculation, which creates an adaptation requirement the existing standards do not address.
The Pereira Gray (2021), commissioned by RICS following concerns about relationship capture and methodology drift in real estate valuations, recommended mandatory valuer rotation at a maximum of ten years (five for high-risk valuations), an independent quality assurance panel, mandatory separation of valuation from advisory work in the same engagement, and audit trail requirements for valuer-client communications. These recommendations were implemented by RICS in May 2024. They address the professional conduct dimension of the valuation problem — the conditions under which a valuer’s independence from the client relationship is preserved — rather than the incentive design dimension.
Advani et al. (2020), in the Wealth Tax Commission final report, examined the valuation problem from a UK administrative perspective. Their conclusion — that annual valuation of the full wealth distribution is not feasible without substantial new infrastructure — is the baseline from which the WDT’s route-based valuation architecture departs by differentiating asset classes and assigning different mechanisms to each.
In the WDT project: The WDT’s central reframing of the valuation problem — from how the state discovers a correct value to under what conditions honest declaration becomes the rational strategy for most taxpayers across most asset classes — is developed in (VAL §1). The four-route architecture that follows from this reframing is in (VAL §4). The professional valuation governance framework the WDT requires is in (VAL §10). The IVSC hierarchy, RICS standards, and Pereira Gray recommendations are the professional baseline for the Valuation Code described in (VAL §10.5).
9. Self-Assessment and Deterrence
Harberger (1965) proposed, in a paper on Latin American property tax reform, that self-reported values could produce honest declarations without independent state appraisal if the declaration simultaneously constituted an offer to sell at that price. A property owner who declares a low value reduces the annual levy but risks losing the asset to any buyer willing to pay the declared amount. A property owner who declares a high value protects against loss but inflates the levy. Under idealised conditions, self-reported values converge toward true market values through the taxpayer’s own incentives.
Posner & Weyl (2018) elaborated this mechanism in a broader reform context as the Common Ownership Self-Assessed Tax, identifying two properties: the mechanism creates a credible deterrent to underreporting by making the underreporter bear the cost of their own misstatement, and it generates allocative efficiency gains by ensuring assets flow to whoever values them most. Their version is designed as a comprehensive property rights system, not as a valuation layer within a separate tax.
Weyl & Zhang (2022) extend the Harberger framework to depreciating licenses, demonstrating its adaptability beyond static property. Prewitt (2019) survey in the University of Chicago Law Review confirms that no prior application of Harberger-derived deterrence as a valuation enforcement layer within a separate accrual-based tax system had been identified in the legal and economics literature.
The key distinction between the Harberger tradition and the WDT’s application is purpose. The Harberger mechanism in its original and Posner-Weyl form is simultaneously a tax system and an asset allocation mechanism: the forced sale to a higher-valuing buyer is part of the design, not a side effect. The WDT’s Route D auction mechanism uses the deterrent logic — a taxpayer who understates risks losing the asset at their own declared price to competitive bidding — without embracing the allocative purpose. The mechanism is designed to fire rarely, as a deterrent of last resort. The asset need not change hands: if the taxpayer bids successfully at auction, or exercises their right of first refusal at the highest third-party bid price, the basis resets to a market-discovered figure and the deterrent has achieved its purpose.
In the WDT project: The Route D auction mechanism, its relationship to the Harberger tradition, and the distinctions from the Posner-Weyl design are developed in (VAL §11.1) and (VAL §11.2). The three-body Valuation Body architecture that governs the auction trigger is in (GOV §6.1) and (GOV.B §G). The deterrent rather than allocative purpose is the basis for the constitutional property-rights posture of the auction discussed in (WP §8.5).
10. Cooperative Compliance and Procedural Justice
Tyler (1990) found that people’s willingness to comply with rules and authority depends substantially on their perception of the process as fair — specifically, whether rules are applied consistently and whether people feel they have meaningful voice. The threat of sanction matters less than procedural legitimacy. Systems perceived as arbitrary generate resistance even when enforcement is technically adequate.
Kirchler (2007) developed this into the slippery slope framework: tax compliance is jointly determined by institutional trust and enforcement power. High trust supports voluntary compliance at lower enforcement cost; low trust requires high enforcement power to produce equivalent compliance rates, and the resulting compliance is grudging and strategic. A system that begins with cooperation can move toward enforcement-dependence if trust is lost; a system that begins with enforcement faces much higher costs in attempting to shift toward cooperation later.
Gangl et al. (2015b) showed that procedural fairness perceptions specifically mediate the trust-compliance relationship. Taxpayers who perceive the system as treating them fairly comply more voluntarily even where they disagree with the substantive outcome. Gangl et al. (2015a) extended the framework to examine how tax authority interaction styles affect corporate compliance, finding that power-based and trust-based approaches produce different compliance dynamics among corporate taxpayers.
Ayres & Braithwaite (1992) proposed the enforcement pyramid as a framework for regulatory compliance: cooperative approaches should be the starting point, with escalation to more coercive measures reserved for cases where cooperation fails. Beginning with coercion closes off the cooperative option permanently and activates resistance from actors who might otherwise have complied. The framework was developed for corporate regulation but applies structurally to tax compliance design.
Kleven et al. (2011) provide the cleanest field experimental evidence on the deterrence and reporting dimensions of compliance, using Danish administrative data and randomised audit assignment. Their key finding is that third-party reporting nearly eliminates evasion even among households with no audit risk, because misreporting can be detected without an audit. Self-reported income, by contrast, shows substantial sensitivity to audit probability.
In the WDT project: The cooperative architecture’s grounding in the procedural justice literature is noted in (WP §3.5) and (MF §7). (BEHAV §4) and (BEHAV §5) develops the friction taxonomy and design principles through which the cooperative compliance mechanisms are intended to operate. The gap between what this literature establishes and what the WDT claims — that cooperative design effects operate primarily through professional intermediaries in the WDT’s taxpayer population — is identified in (LR.A §3.1).
11. Compliance in Professionally Mediated Systems
The cooperative compliance literature in (LR.B §10) was developed primarily around individual taxpayers making decisions in response to their own perceptions of the system. The WDT’s primary taxpayer population does not fit this model. At the wealth levels where material revenue concentrates, the compliance decision is made within an institutional network of specialist advisers, family offices, trustees, and corporate structures.
Klepper & Nagin (1989) established the relevant mechanism: professional advisers reduce accidental non-compliance by ensuring technical requirements are met, but increase strategic optimisation on items where professional judgment creates genuine ambiguity. Advisers are not neutral intermediaries — they are norm transmitters. Clients tend to adopt the compliance posture of their advisers, which means the relevant behavioural unit for professionally advised taxpayers is the advisory relationship rather than the individual. Erard (1993) and Sakurai & Braithwaite (2003) confirmed that the professional mediation layer shapes not only what is reported but how taxpayers conceptualise their obligations.
The OECD’s Cooperative Compliance programme, developed through the Enhanced Relationship framework and subsequent Horizontal Monitoring and ICAP initiatives, addresses large corporate taxpayers and multinationals rather than high-wealth individuals, but the structural analogy is closer to the WDT’s primary taxpayer population than the individual compliance literature. Its consistent finding is that stable, reciprocal engagement — transparency in exchange for reduced uncertainty — improves compliance outcomes among sophisticated actors without depending on the individual fairness responses the Tyler and Kirchler literature identifies.
The political science literature on elite compliance provides a parallel line of support. Levi (1988), Rothstein (2001), and Rothstein & Uslaner (2005) each find that powerful actors cooperate with institutional arrangements when those arrangements are predictable, procedurally consistent, and credibly committed to their stated terms. The distributive outcome matters less than the institutional credibility of the arrangement. For the WDT, the symmetric loss-refund mechanism and constitutional protection of the refund obligation function as credibility signals of this type: not appeals to fairness in any moralised sense, but demonstrations that the state will behave consistently across good and bad years in ways sophisticated actors can plan around.
Kornhauser (2007) and Richardson (2008) find that procedural fairness effects on compliance are present but smaller among higher-income groups than in the general population, consistent with the hypothesis that professional mediation attenuates the direct psychological mechanism without eliminating the institutional credibility effect.
In the WDT project: (BEHAV §7) develops the argument that the cooperative architecture’s primary compliance effect operates through professional intermediaries. The adviser-mediated compliance dynamic is identified as the first-order Phase One evaluation priority in (BEHAV §11). (LR.A §3.1) identifies the gap between what the existing literature establishes and what the WDT claims.
12. Collective Action and Governance of Multi-Constituency Institutions
Olson (1965) on the logic of collective action is the relevant starting point for multi-constituency governance. Small, well-organised groups systematically outperform large, poorly-organised groups in institutional settings where both have formal representation. The mechanism is the free-rider problem applied to collective influence: individuals in a small group have individually significant stakes and individually significant incentives to bear organisational costs; individuals in a large diffuse group have individually small stakes and can free-ride on whatever collective action others undertake.
The implication for WDT governance is direct: the Taxpayer Chamber represents a small, wealthy, well-organised constituency with concentrated stakes. The Dividend Recipient Chamber represents a large, diffuse population with individually small stakes. Formal equal representation is not the same as effective equal participation; the Olson problem predicts that the diffuse constituency will be systematically outcompeted unless the governance design directly addresses the structural asymmetry.
Ostrom (1990) provides a different approach to the same underlying problem. Her empirical work on commons governance identified the conditions under which communities successfully manage shared resources without either state control or privatisation — characterised by clearly defined boundaries, rules matched to local conditions, collective choice arrangements that include affected parties, monitoring, graduated sanctions, and conflict resolution mechanisms. The WDT’s Governing Council is not a commons in Ostrom’s analytical sense, but several of her design principles transfer to the institutional design problem the WDT faces.
Fung & Wright (2003) study participatory governance across a range of institutional contexts, finding that effective participatory institutions require clear decision rights rather than merely voice, institutional support for participation, and accountability mechanisms linking participation to outcomes. Fishkin (2009) examines deliberative polling and citizen assemblies as mechanisms for incorporating diffuse public preferences into institutional decision-making.
Olson’s analysis also applies to the political durability question that (POL) addresses: the taxed population is small enough to organise cheaply, wealthy enough to sustain that organisation indefinitely, and influential enough to access decision-makers and advisory networks. The general population has no equivalent per-capita incentive to organise in defence. This asymmetry is a structural feature of wealth tax politics, not a contingent outcome.
In the WDT project: The Olson problem motivates the lottery-selection model for the Dividend Recipient Chamber and the constituency dissolution mechanism, developed in (GOV §5.1) and (GOV.B §A.3). The WDT’s response to Ostrom’s conditions is implicit throughout [GOV]; the explicit derivation from the mechanism’s own transaction structure is in (GOV §3). The political durability dimension of the Olson problem is treated in (POL §3.4).
13. Mutual and Pension Governance
No existing sovereign wealth fund model has governance participation by the population whose contributions capitalise it. The WDT’s three-constituency Governing Council is without direct institutional precedent. Two adjacent governance traditions provide partial analogues.
Mutual insurance governance is the closest structural analogue. Mutual insurers are owned by policyholders rather than shareholders; their governance typically involves policyholder representation alongside professional management with obligations running to current and future policyholders. Hansmann (2022) provides the academic foundation, finding that ultimate policyholder control may matter more for conveying the institution’s purpose and building trust than for directly affecting management incentives. O’Sullivan (1998) reviews comparative governance of mutual and proprietary insurance companies, including the conditions under which mutuals convert to stock ownership. MacMinn & Ren (2011) provide a comparative analysis of the two organisational forms, covering agency problems and efficiency considerations. The International Cooperative and Mutual Insurance Federation (2020) provides operational evidence that policyholder governance of large pooled institutions is standard practice globally rather than experimental.
The disanalogies with the WDT fund are real. Mutual insurers do not have a public-interest constituency distinct from the policyholder constituency. The WDT fund has a government constituency that mutuals lack. The refund obligation is mechanical and constitutionally guaranteed rather than actuarially discretionary. The mutual insurance analogy is instructive but not directly applicable.
Large defined-benefit pension funds hold assets on behalf of a specific beneficiary population, operate over long investment horizons, and carry pre-committed liability obligations analogous to the WDT’s refund obligations, making the governance literature for this sector relevant. Clark & Urwin (2008) identify four best-practice governance factors: mission clarity, board quality, investment process, and accountability. Well-governed funds maintain clear separation between mission, investment objectives, and accountability structures, with professional investment management under board oversight rather than direct political control. Ambachtsheer (2007) argues that most pension funds operate below best practice because of short-termism driven by political pressure and sponsor influence over investment decisions.
The WDT fund’s mode B failure risk — the state honouring the fund’s mechanics while abandoning what the mechanics exist to serve — maps onto what Ambachtsheer identifies as the primary governance failure mode in pension funds: the gradual displacement of the long-duration beneficiary interest by the shorter-horizon interests of those controlling investment decisions.
In the WDT project: The governance participation rationale — why it is derived from the mechanism’s transaction structure rather than from an analogy to mutual insurance — is in (GOV §3). The Mode B failure risk and its structural safeguards are in (GOV §4) and (GOV.A §E).
14. Sovereign Wealth Fund Reference Cases
The Government Pension Fund Global (GPFG), Norway’s sovereign wealth fund, is the principal institutional reference case for WDT purposes.
The GPFG is managed at arm’s length from the fiscal ministry by Norges Bank Investment Management, and that institutional independence has held relatively well through periods of fiscal and political pressure. It is explicitly prohibited from investing in Norwegian assets, a rule designed to prevent entanglement between the state’s financial interests and domestic asset prices. It reports publicly on holdings, returns, and investment decisions at a level of transparency that sets a benchmark standard. Its ethical council provides a layer of independent oversight sitting outside professional management.
The WDT fund cannot replicate the domestic investment prohibition directly because it needs calibration against domestic refund obligations, but the Norwegian rule is the institutional precedent for taking state-finance entanglement seriously as a design consideration. Truman (2008) provides the academic treatment of GPFG governance and Frankel (2010) the broader literature on SWF governance quality. Aizenman & Glick (2009) identify recurring governance failure modes in the SWF literature: political interference in investment decisions during fiscal stress, mandate drift toward domestic assets over time, and capture by the finance ministry where appointments, budgets, and information flows remain under ministerial control.
Chile’s Pension Reserve Fund, created in 2006, is the closest structural analogue for a fund pre-funding a specific defined liability. Established to meet future obligations for non-contributory solidarity pensions, funded from copper revenues and budget surpluses, managed by the Central Bank of Chile as fiscal agent with an independent Financial Committee, it maintains strict withdrawal rules and a liability-matching investment mandate. The Alaska Permanent Fund demonstrates that direct public distribution of dividends from a state fund is operationally feasible and politically durable over decades.
In the WDT project: The SWF governance architecture is developed in (GOV §6.3) and (GOV.B §E). The SRR and LRR reserve structure and their sizing are in (RATES §6). The Custodian’s independence requirements draw on the GPFG and OBR institutional models discussed in (JUR §1.5.3).
15. Precommitment and Fiscal Rules
Kydland & Prescott (1977) established the time inconsistency problem for fiscal policy: governments facing severe downturns have strong incentives to deviate from prior commitments, and those incentives are largest precisely when the commitment is most needed. A government committed to maintaining refund payments in all years will face the strongest pressure to suspend those payments in the years of largest refund obligations — the same years of severe market decline when the refund mechanism is most valuable to taxpayers. Without a credible precommitment mechanism, rational actors anticipate the deviation and adjust their behaviour accordingly.
Elster (2000) provides the broader theoretical treatment of precommitment as a constitutional and institutional strategy, examining how Ulysses-style advance binding can produce outcomes that time-consistent decision-making cannot. A refund commitment that survives good times but is suspended under fiscal pressure provides no guarantee.
Kopits & Symansky (1998) define the criteria for well-designed fiscal rules: well-defined, transparent, simple, flexible, adequate relative to the final goal, enforceable, consistent, and supported by sound underlying policies. The checklist applies to designing the WDT’s refund commitment as a statutory rule, particularly in a constitutional context such as the UK where parliamentary sovereignty means formal constitutional entrenchment in the traditional sense is unavailable.
Wyplosz (2005) reviews the comparative track record of fiscal rules versus independent institutions, finding that independent fiscal institutions can outperform strict rules in practice by allowing informed discretion within a mandate-constrained framework. Debrun, Moulin, Turrini, Ayuso-i-Casals, and Debrun et al. (2008) survey the comparative evidence on independent fiscal councils, finding that statutory reporting mandates can raise the political cost of deviation from fiscal commitments without requiring a veto power.
In the WDT project: The precommitment architecture — SWF pre-funding, constitutional entrenchment of the nine enumerated structural clauses, the Administrator’s non-discretionary publication mandate — is developed in (GOV §5.2) and (GOV §5.3) and (WP §8.5). The UK institutional context and the OBR model as precedent for the Custodian’s independence are discussed in (JUR §1.5.3) and (JUR §4.2).
16. Moral Philosophy and the Normative Case
Murphy & Nagel (2002) is the most direct intellectual predecessor on the collective production argument. Their argument is that pre-tax income distributions are not morally prior to tax policy because property rights depend on legal and institutional frameworks the state provides and maintains. The WDT’s argument that extraordinary private wealth is partly a collective product, dependent on the labour of workers, the consumption of ordinary households, and the public infrastructure those households fund, develops this at a more specific mechanistic level.
Hasen (2017) independently derives the argument that wealth as potential power is not reducible to consumption, and that a consumption tax therefore fails to reach the relevant externality. His Part III.A.1 anticipates and refutes the Shaviro-Bankman-Weisbach counter-argument that wealth derives its value entirely from what it can buy. His response is that the power to influence outcomes operates through the mere prospect of conferring benefits, not through actual spending, and that large concentrations of wealth tend not to dissipate into consumption. This is substantively the same position (MF §4) develops, and constitutes independent prior art for the specific normative claim that wealth above a threshold functions as durable power rather than deferred consumption. The WDT’s departure from Hasen is architectural, not normative: he concludes that the externality argument supports adding a progressive accretion wealth tax as a corrective supplement; the WDT concludes that it supports taking the annual net worth delta as the organising unit of a complete fiscal architecture.
Rawls (1971) provides background grounding through the difference principle and the framework of fair terms of social cooperation. The WDT does not claim to be a Rawlsian proposal: it tolerates significant wealth inequality provided that inequality does not undermine democratic institutions, which is a weaker claim than the difference principle makes. The intellectual connection is one of context rather than derivation.
Atkinson (2015) is relevant to the WDT’s specific policy focus. His proposals for reducing inequality include structural changes to capital ownership and returns, not only redistribution through the tax-benefit system. The WDT is consistent with the Atkinson direction without being derived from it.
The treatment of wealth as durable economic and institutional power rather than deferred consumption draws on a tradition that tax theory has not fully absorbed. Galbraith (1952) on countervailing power, Mills (1956) on the power elite, and Hacker & Pierson (2010) on winner-take-all politics describe mechanisms by which concentrated economic resources translate into structural influence over political and institutional processes in ways that are not captured by the consumption-tax framework’s view of wealth as the present value of future consumption.
In the WDT project: [MF] provides the full moral foundations treatment: the foundational axiom (MF §2), the collective production argument (MF §3), wealth as power (MF §4), the cognitive moral hazard of the current system (MF §5), and the terminal goal of democratic flourishing (MF §6). The named compromises where the theoretical framework meets administrative necessity are in (MF §9).
17. Democratic Legitimacy and Wealth Concentration
Dahl (1985) argued that economic inequality of sufficient magnitude is incompatible with political democracy because concentrated economic power translates directly into concentrated political power through mechanisms formal democratic institutions cannot neutralise. The durability of democratic institutions is therefore not independent of their underlying distribution of economic resources. Institutions may formally distribute political rights equally while operating within social conditions in which highly unequal resources produce highly unequal capacities to organise, communicate, lobby, finance, and sustain political influence.
North (1990) provides the complementary institutional perspective: political and economic institutions create the incentive structures within which power is exercised and reproduced. Where concentrated wealth can systematically shape those institutions or the rules they administer, economic concentration can become self-reinforcing rather than merely reflecting prior market outcomes. The democratic problem is consequently not exhausted by any single act of corruption or formal capture. It concerns the longer-term capacity of unequal economic power to alter the institutional environment through which political choices are made.
Wilson (1980) similarly emphasises the importance of organisation and concentrated interests in political processes. Benefits concentrated among relatively small groups may generate stronger incentives for sustained political organisation than diffuse interests spread across much larger populations. This creates a structural asymmetry: citizens may possess formally equal votes while possessing very unequal capacities to monitor, organise around, and influence the policies affecting them. Wealth concentration can intensify that asymmetry by supplying a durable private resource base for political action.
Pierson (1994) shows how political institutions and policy choices can generate path dependence, making established distributions and policy settlements difficult to reverse even after their original political conditions have changed. Pierson (2000) extends this analysis to advanced industrial democracies, emphasising the political obstacles to reversing entrenched arrangements. Applied to wealth concentration, the implication is not that every increase in inequality inevitably produces democratic capture, but that sufficiently durable concentrations of economic power may create feedback mechanisms through which political influence helps preserve the economic conditions from which that influence derives.
Bartels (2008) provides empirical evidence for the US context: elected representatives are systematically more responsive to the preferences of wealthy constituents than to those of ordinary voters, a pattern that holds across party lines and over time. Gilens (2012) extends this finding, showing that when the preferences of high-income Americans diverge from those of the median, policy outcomes align with high-income preferences at rates substantially above what chance would produce.
These findings do not establish a simple causal story between wealth concentration and democratic capture; the methodology of this literature is contested. The direction of the evidence is nevertheless consistent across distinct approaches: concentrated wealth creates the conditions for concentrated political influence, while formal democratic institutions may attenuate those effects imperfectly. The institutional, organisational, and path-dependent mechanisms identified by North, Wilson, and Pierson provide a broader account of how such influence can become durable and self-reinforcing rather than appearing only as a series of isolated interventions.
Piketty et al. (2023), drawing on this evidence, identify the breakdown of standard income and consumption measures as adequate proxies for economic wellbeing — where capital gains dwarf ordinary income flows — as the second distributional rationale for wealth taxation, alongside the difficulty of distinguishing capital from labour income at the top of the distribution.
In the WDT project: The democratic legitimacy argument and the terminal goal — preserving the conditions under which democratic capitalism remains legitimate and sustainable, not equality of outcome — are developed in (MF §6) and (WP §1). (POL §3) examines the political durability dimension of this argument, including the failure modes that have ended previous wealth taxes.
18. Exit Taxation and International Coordination
The existing international landscape for exit taxation is documented in the OECD (2025) working paper, which confirms that 14 of 38 OECD members levy some form of exit tax on individual unrealised capital gains at the point of departure from tax residence, treating departure as a deemed disposal at market value. Design varies considerably: some cancel the charge after a defined period abroad, some allow instalment payment, and the interaction with bilateral tax treaties creates complexity where treaties predate exit tax provisions. Split taxation — under which the departure state taxes accrued gain up to departure and the arrival state taxes subsequent appreciation — is the most technically sophisticated approach and the direction in which treaty practice is moving.
The OECD’s 2025 study and the European Commission’s April 2026 study on wealth taxation identify limited evidence on the international mobility of ultra-high-net-worth individuals specifically. The Jakobsen et al. (2024) Norwegian study partially fills this gap; the conditions that make Norwegian estimates tractable (strong national administration, limited within-country arbitrage, CRS information exchange) are also the conditions the WDT’s reference design is built to create or approximate. The Kleven et al. (2024) survey of the broader taxation-and-migration literature confirms that migration responses are concentrated among high-earning and high-wealth individuals and that magnitude depends heavily on the availability of realistic alternative jurisdictions and the quality of information exchange between them.
Agrawal et al. (2025) find that wealth-tax-driven migration generates income tax and VAT revenue losses approximately six times larger than the direct wealth tax revenue loss — the cross-base externality. This finding has not been incorporated into any existing wealth tax behavioural model. The Norwegian 22-cent estimate captures only the direct wealth tax channel; total fiscal cost is substantially larger once cross-base effects are included. The modelling gap this creates is identified in (LR.A §3.2).
Zucman (2024) published a blueprint for a coordinated minimum effective taxation standard for ultra-high-net-worth individuals under Brazil’s G20 presidency. The proposal is structured around Pillar Two’s logic: a 2% annual minimum tax on approximately 3,000 billionaires globally, with participating countries able to top up undertaxed wealth in non-participating jurisdictions. The UN Tax Committee’s Subcommittee on Wealth and Solidarity Taxes is developing a model wealth tax law in the same direction. Both proposals take a stock wealth tax as their reference instrument and contain no loss-refund mechanism; the interaction of a refund-based system with a minimum-tax floor creates a design question the existing coordination literature does not address. That interaction is identified as a research gap in (LR.A §4.3).
The lessons from Pillar Two on the political economy of coordination are relevant to the WDT’s Phase Two international agenda: agreement across 140-plus jurisdictions took approximately a decade from the launch of BEPS in 2013 to implementation in 2024, and the US exemption of its multinationals from the January 2026 rules introduced significant instability into the framework. Domestic WDT implementation does not require international agreement as a precondition; the Zucman blueprint follows the same sequencing logic.
In the WDT project: The exit and closure design is developed in [CLOSE] throughout: the no-punitive-exit-taxation position (CLOSE §4.2), the bridging facility (CLOSE §5), and the re-entry rule (CLOSE §6) are the three settled structural positions. Jurisdiction-specific legal implementation remains open (0.0 #8). The cross-base externality as a Phase One transitional exposure is treated in (CLOSE §9.2) and (BEHAV §9.2).