RESEARCH DOSSIER

Tax Havens

Abstract. Tax havens are a form of jurisdictional economic specialization through which internationally mobile tax bases encounter favorable fiscal treatment. This monograph examines their historical formation, institutional foundations, corporate and household economic functions, measurement, distributional effects and international regulation. The research distinguishes tax haven classifications from offshore financial-center definitions, reported profits from productive activity, financial positions from ultimate ownership, and lawful international activity from established noncompliance. Original empirical research and official statistical records are synthesized through comparative institutional analysis and explicit hypothetical accounting examples. Particular attention is given to the measurement of profit shifting, intermediary investment positions, offshore household wealth, information exchange and the evolving global minimum tax. The analysis evaluates the relationship between local host benefits and cross-border fiscal externalities as well as the identification problems that limit causal and welfare conclusions. Its contribution is an integrated, source-bounded economic synthesis, not a newly named theoretical system or independent global estimate. Sources accessed at the abstract or indexed-excerpt level are identified in the evidence record.

Keywords: Tax havens; offshore economics; effective taxation; mobile tax bases; foreign direct investment; profit shifting; offshore wealth; tax competition; international tax transparency; global minimum tax; fiscal spillovers

Introduction

Tax havens occupy a distinctive position in the international economy because access to a jurisdiction's fiscal institutions can be separated from the place where production, consumption, employment and the ultimate ownership of capital occur. Governments ordinarily tax activities, persons, property or transactions through legally defined connections to territory. Internationally mobile assets and business organizations allow several governments' systems to be relevant to the same economic relationship. A tax haven exploits the demand for these jurisdictional differences by supplying favorable tax treatment to internationally mobile activity or wealth.

The central economic question is how much of the resulting attraction reflects additional productive activity, how much reflects the relocation of legally recognized claims, and how much reflects diminished taxation of income generated elsewhere. Each possibility has a different welfare implication. A jurisdiction that hosts a genuine service industry may create employment and exports while also collecting income that would otherwise be taxable elsewhere. A firm with a low-tax affiliate may expand productive investment in a higher-tax country, or may merely reorganize the recorded allocation of a fixed stock of profit. A household can diversify its portfolio internationally while satisfying every applicable reporting and tax obligation. Its offshore balance is not in itself evidence of evasion.

This monograph examines the economic specialization of tax havens, rather than prescribing an internationally valid legal classification. Its unit of analysis changes explicitly when the research object changes: a tax regime and its date for classification, a firm group for corporate behavior, a household and tax residence for personal wealth, a jurisdictional balance sheet for external positions, and interacting governments for welfare analysis. The study combines institutional history, comparative interpretation of original research, stylized arithmetic and methodological evaluation. It does not estimate a new global tax gap, replicate confidential microdata or assign contemporary criminal or noncooperative status to particular jurisdictions.

Definitions and classification

A tax haven, for the economic purposes of this dossier, is a jurisdiction or identifiable jurisdictional regime that attracts internationally mobile tax bases by supplying substantially favorable effective fiscal treatment relative to relevant alternative locations. The definition refers to a functional role observed within a period and a population of activities. It does not confer a legal status. A preferential regime can exist inside an economy with a substantial ordinary tax system. A jurisdiction can alter the treatment of one type of income while its other fiscal functions remain essentially unchanged.

The historical OECD (1998) report Harmful Tax Competition distinguishes tax havens from harmful preferential regimes and emphasizes the international consequences of selective treatment of mobile activities. The distinction matters because statutory rates alone do not reveal the economic terms of taxation. Exemptions, definitions of taxable income, withholding taxes, incentives, administrative practice and interaction with other countries' taxes all affect the relevant expected liability. The absence of a conventional corporate income tax may coexist with taxes on labor, imports, land, consumption or transactions.

An offshore financial center is a different statistical and economic object. Zoromé's (2007) operational definition of OFCs relates financial services to nonresident demand and domestic economic scale. A center can specialize in intermediation, asset administration or insurance because it possesses sophisticated professional infrastructure. Its external balance-sheet size is informative about its international reach; it is not a measure of its tax differential. Conversely, a selectively favorable corporate regime can attract a large volume of mobile profit without operating a substantial offshore banking center.

Three separate observations are therefore needed in any classification: the extent of fiscal advantage, the scale and character of internationally mobile activity, and the institutional quality of reporting and enforcement. These are research dimensions, not an additive index. A zero value in one dimension does not fix the others. Using a historical published list to define a sample is legitimate if its period and criteria are preserved. Applying the list indefinitely, treating it as official contemporary status, or conflating it with a sanctions list would introduce classification error.

The relevant counterfactual is the effective burden expected in a plausible alternative, evaluated over the life of the investment. Comparing a small jurisdiction's headline rate with a large country's headline rate is insufficient if the income would be exempt in both locations, subject to a residence-country charge, or covered by an international minimum-tax rule. Classification should disclose whether the subject is a country, territory, subnational regime or particular activity; the baseline tax treatment; the period; and the evidence for its role.

Historical formation and institutional credibility

Modern tax havens emerged through the interaction of expanding international commerce with territorial fiscal systems and legal institutions. Palan (2002) interprets the historical phenomenon through the commercialization of state sovereignty: public jurisdiction becomes a source of internationally valuable services. This interpretation helps account for the positive supply of legal infrastructure, fiscal arrangements and registration services rather than portraying offshore activity as geographically accidental.

Greater cross-border mobility amplifies the value of differences among jurisdictions. A jurisdiction can acquire a comparative advantage in the administration of mobile claims even when its local market is small. Professional expertise, reliable recordkeeping, accessible courts, telecommunications and links to major financial markets may raise the expected quality of that administration. These complementary inputs require investment and institutional continuity. A nominally low rate without reliable administration need not attract significant internationally mobile capital.

Hines and Rice's (1994) Fiscal Paradise is an important early quantitative investigation of the relationship between US multinational businesses and foreign haven affiliates. It establishes the importance of studying the allocation of corporate earnings along with the distribution of investment. Its historical US corporate population cannot be treated as a direct measure of the contemporary global economy. Nevertheless, it illustrates how the phenomenon predates modern automatic information exchange and digital business models.

Dharmapala and Hines (2009) examine which countries become tax havens and report the association of governance quality with haven status in their historical sample. This is evidence that an attractive fiscal rate can be complemented by institutional credibility. Their historical haven list, sample composition and identification assumptions govern the result. Even an instrumental-variables design requires the exclusion restriction: a historical determinant of governance may affect financial specialization through other channels. Economic development, legal heritage, political connections and selection into international markets can affect both institutional quality and tax policy.

History consequently supplies more than a chronology. It shows why changing one tax parameter can fail to reproduce a successful financial center and why reforms that alter transparency need not eliminate demand for the other services supplied by such centers. The existence of an established industry creates continuity, but past success does not guarantee future attractiveness.

Tax bases and jurisdictional economics

Tax systems establish boundaries among source, residence, legal ownership, beneficial ownership and taxable realization. Economic activity generates wages, operating surplus and other income; tax authorities allocate rights to levy charges by reference to statutory connections. The same underlying activity may produce different measurements of taxable income depending on timing, depreciation allowances and adjustments to financial accounts. A financial center occupies one institutional node in this wider allocation.

Tax competition arises because governments anticipate changes in both their own bases and other jurisdictions' responses. For a simple tax with rate t and taxable base B(t), revenue is R(t) = tB(t). Holding other policies fixed, its derivative is B(t) + tB'(t). A lower rate can increase revenue when the induced expansion of the base is sufficiently strong. But the base response combines additional real activity, formal relocation, reporting changes, and activity no longer conducted elsewhere; the derivative identifies none of these components by itself.

An illustrative example clarifies the distinction. Suppose a host collects tax at 5% on a base of 100 units, yielding 5 units. A rate reduction to 4% that expands the measured base to 150 units produces 6 units of revenue. The increase of one unit is a fiscal result for that host. If the additional base consists entirely of profit formerly taxed at 25% in another jurisdiction, the counterpart loses 12.5 units of tax on those 50 units, absent offsetting rules. The consolidated fiscal comparison differs from the host's result. The numbers are hypothetical, hold other behavior constant and do not estimate real-world responses.

Government strategies may also pursue jobs, international integration, fees or professional-services exports. Their relative importance can vary with domestic size, mobility of the base and alternative revenue sources. In an interdependent system, the policy that maximizes one government's local objective need not maximize consolidated welfare. Theoretical analysis by Slemrod and Wilson (2009) in Tax Competition with Parasitic Tax Havens formalizes an environment in which haven activity diverts revenues and encourages costly enforcement. Its conclusions apply to the particular assumptions of the model, rather than establishing that every form of offshore financial service is socially wasteful.

The meaningful comparison is a portfolio of policy instruments and consequences. A tax rate that attracts profitable business may be accompanied by infrastructure expenditure, higher prices for scarce housing, regulatory costs or a concentration of fiscal receipts among a small number of customers. Gross corporate registrations cannot capture these effects.

Multinational enterprises and recorded profitability

Multinational groups combine economic functions distributed across jurisdictions with a common ownership structure. Production, marketing, finance, management and intangible assets can be located in different places. The legal characterization of transactions within the group affects the allocation of recorded revenue, expenses and profit. Economic substance requires attention to people, assets, decisions and risk rather than the formal count of affiliated corporations.

Beer, de Mooij and Liu (2020) review international corporate tax avoidance across several channels, including transfer pricing, financing and tax treaty effects. These channels are useful categories for evaluating aggregate evidence. They are not instructions for structuring transactions. In practice, lawful commercial transactions can generate the same broad accounting categories as disputed arrangements. Whether a particular transaction is compliant requires its facts, governing law and competent authority's decision.

The economics of profit location can be illustrated without reference to any jurisdiction's actual law. Consider a group with pre-tax economic profit of 100 units. If all 100 units are taxable at 25%, its tax expense is 25. If 40 units of the same total are instead taxable at 5% and 60 at 25%, the total becomes 2 + 15 = 17, ignoring residence-country taxes, withholding, minimum taxes, timing differences and enforcement. The difference of eight units is the gross mechanical effect of the assumed allocation. Output, employment and the sum of pre-tax group profit have been held constant. The arithmetic is an accounting counterfactual, not an estimate of attainable savings or a description of a permitted arrangement.

The empirical problem is to separate changes of this sort from genuine differences in productivity, asset ownership, business cycles and risk. A financial center's recorded profits may reflect necessary intermediate services, the return to an investment or a statistical convention. Comparisons should distinguish financial from nonfinancial firms and loss-making from profitable affiliates. Ratios that combine profits from positive and negative observations can be unstable; ratios of aggregate sums and averages of firm-level ratios answer different questions.

A group may also face uncertain or contested tax liabilities. Changes in financial-statement provisions can affect reported tax burdens independently of cash payment. This is a reason to compare accrued taxes and cash taxes over appropriate periods, to disclose denominator definitions and to avoid interpreting effective-rate anomalies as individualized proof of misconduct.

Corporate investment and competing interpretations

The welfare consequences of corporate use of tax havens depend in part on whether the activity is complementary to production elsewhere. Desai, Foley and Hines (2006a) investigate the demand for tax haven operations. Their historical firm evidence associates demand for haven operations with substantial foreign business, intensive research and development and intrafirm trade. The result is consistent with a firm-level selection mechanism: the capabilities and incentives of multinational enterprises differ substantially.

Their related study (2006b), Do Tax Havens Divert Economic Activity?, raises the possibility that haven activity and investment in neighboring nonhaven economies may be complementary. Lower overall taxation of a group can increase its after-tax return on real investment. This mechanism does not settle the welfare question: revenue loss, distributional consequences and the financing of public inputs remain relevant. Moreover, stronger firms may undertake both kinds of foreign activity, leading to positive correlations independently of a causal relationship.

Consider two possible counterfactuals after a reform raises the effective tax burden on a group. The first is continuation of production with a different allocation of taxable profit. The second is a reduction or relocation of real capital expenditure. Both can occur simultaneously among different firms. Distinguishing them calls for data on employment, tangible assets, investment and pre-tax profitability alongside jurisdictional tax rates. An increase in real investment in one country may displace investment elsewhere; international welfare cannot be inferred from a single host's employment growth.

Research designs that exploit tax-policy changes need to assess anticipation, differential exposure and contemporaneous reforms. A difference-in-differences estimator is credible only when the treated and comparison groups supply a plausible counterfactual for each other absent treatment. If jurisdictions respond to the same reform, a comparison group may itself be affected, violating the simplest treatment assumption. The appropriate interpretation is often an equilibrium response within a network, not an isolated bilateral change.

Profit shifting measurement and national accounts

The location of booked profit determines some national-account aggregates even when the place of production has not changed. Tørsløv, Wier and Zucman (2023), in The Missing Profits of Nations, exploit unusually high profitability of multinational affiliates in tax havens relative to the comparison population. Their published estimate is that 36% of multinational profits are shifted to tax havens globally. This is a study-specific estimated share of multinational profits, not 36% of all company revenue, wealth, GDP or corporate income-tax receipts. Earlier working-paper figures should not be substituted for the published estimate.

The study's methodology has the advantage of examining economic discrepancies beyond individual tax cases. It also depends on the comparison used to infer unobserved counterfactual profits and the available statistical coverage. Economic activity with unique intangible returns or concentrated financial functions may display higher productivity for reasons other than shifting. Modeling the unobserved baseline is thus essential; the result is an estimate and retains its uncertainty.

National accounts and tax accounts answer related but nonidentical questions. Reallocating a booked profit from a financial center to a production economy changes the statistical interpretation of income, the international distribution of factor payments and potentially measured trade in services, depending on the recorded transaction. Economic residence in statistical accounting is not identical to the nationality of the ultimate owner. Gross domestic product and gross national income measure different concepts: location of production relative to resident income flows. A difference between the two is not a measure of tax abuse.

Revenue consequences must be calculated through the counterfactual liability, not by multiplying an estimated profit amount by a uniform global rate without explanation. Different source and residence tax rules, loss offsets and credits can give substantially different estimates. International fiscal comparisons should preserve the gross amount estimated to have been relocated, the assumed tax rate differential, the relevant tax base and the behavioral response to a change in policy.

Foreign direct investment and intermediary geography

Direct investment statistics capture relationships between resident institutional units. They can record funding through intermediary entities even where the immediate economy is not the ultimate source or destination of the productive investment. In the 2019 IMF working paper What Is Real and What Is Not in the Global FDI Network?, Damgaard, Elkjaer and Johannesen estimate that phantom investment may account for nearly 40% of global FDI. The paper disentangles parts of FDI booked through entities without corresponding domestic economic substance, and reallocates real investment toward ultimate investor economies. This figure concerns investment positions under that study's methodology and historical data, not annual flows or estimates of unpaid tax.

For illustration, suppose an investor resident in A supplies 100 units to an intermediary in B, which invests 100 in a productive enterprise in C. If both transactions meet the statistical requirements for direct investment, the observed outward positions may include 100 from A to B and another 100 from B to C. Two separately recorded positions correspond to a single simplified ultimate financing relationship of 100 from A to C. Consolidation of these economic relationships does not render either individual statistical entry false; they record different claims among resident units.

An immediate-investor table is informative about legal funding routes and counterparty exposure. An ultimate-investor table addresses a different question concerning control or ultimate source of funding, according to the precise statistical convention employed. The distinction matters for studies estimating how trade agreements, proximity, common language, market size or tax rules affect investment. A variable that predicts intermediary location may not predict productive location.

Neither an exceptionally large FDI-to-GDP ratio nor a large special-purpose entity sector is sufficient to quantify fiscal harm. Small denominators can create high ratios, and special-purpose entities can have legitimate commercial uses. Crosswalks should distinguish stock and flow, gross and net, immediate and ultimate counterparts, ownership and risk. Comparing these concepts without harmonization generates misleading rankings.

Conduit and sink functions

Ownership-network analysis identifies roles that conventional country lists obscure. Garcia-Bernardo, Fichtner, Takes and Heemskerk (2017) use an extensive firm-ownership dataset in Uncovering Offshore Financial Centers. Their analysis distinguishes locations where international ownership value accumulates from intermediary jurisdictions that connect ownership chains. They identify five major conduits in their historical network: the Netherlands, the United Kingdom, Switzerland, Ireland and Singapore.

Those findings are conditional on the authors' data, chain definition and measurement method. A named jurisdiction may have substantial domestic production and major cross-border intermediary functions simultaneously. The description of its network role does not assign its present legal or regulatory status. Corporate ownership paths also measure a different object from personal asset holdings, portfolio securities or banks' bilateral external claims.

For economic analysis, the topology indicates that isolated changes in the fiscal regime of one location can lead to reorganization elsewhere. An intermediary whose underlying commercial services remain valuable may retain transactions after certain tax advantages diminish. An entity with little independent economic function may disappear or relocate if the reason for its creation changes. Changes in the number of corporate entities must therefore be interpreted alongside changes in employment, business-service revenue and substantive capital expenditure.

The general lesson is that jurisdictional classification benefits from relational evidence. Demand for favorable taxation creates opportunities for both destination and intermediation services. A count of low-rate jurisdictions cannot capture their respective market roles or their exposure to international reform.

Household wealth and financial intermediation

The economic problem of offshore household wealth differs from corporate profit shifting. Financial wealth is an asset stock, the return on wealth is a flow of income, and a tax liability results from the applicable rules concerning the owner, income, asset and reporting period. A bank deposit, a security held through a custodian and an interest in an investment fund may have different economic and statistical counterpart locations.

Alstadsæter, Johannesen and Zucman (2018) study Who Owns the Wealth in Tax Havens? using macroeconomic information to estimate the international distribution of offshore household financial wealth. A central contribution of this research is the attribution problem: the location of the custodian or intermediary is often observable while the tax residence of the ultimate household is not. Offshore wealth may fall outside directly assembled household balance sheets because the available sources do not identify its owner.

Their later Tax Evasion and Inequality (2019) combines administrative and offshore evidence, illustrating the importance of unequal concealment across wealth groups. The Nordic setting provides unusually rich evidence, but transferring its distributional results to other countries requires assumptions about the representativeness of institutions, enforcement and wealthy households. Selected leaks and voluntary disclosures reveal otherwise difficult-to-observe behavior while introducing selection into the observed sample.

A resident reporting all foreign assets and paying the applicable tax has a different fiscal situation from a resident who conceals taxable income. Economic research must preserve that distinction. Privacy, risk diversification, international business, migration and access to foreign investment products are possible lawful reasons for offshore holdings. Aggregate evidence of undeclared wealth does not establish a particular person's conduct; conversely, legal financial privacy does not exempt someone from applicable tax or disclosure obligations.

To assess a household tax gap, researchers must distinguish the asset's current market value, changes in value, income receipts and taxable events. Multiplying the stock of offshore wealth by a personal income-tax rate treats the entire asset as annual taxable income, an invalid operation in the usual setting. A hypothetical illustration: a stock worth 100 units yielding 4 units of taxable income at a rate of 25% generates 1 unit of annual income-tax liability, before other rules. It does not generate 25 units of liability merely because the underlying asset is worth 100.

Distribution, public finance and development

The distributional consequences of haven activity follow both direct benefits and indirect fiscal responses. Shareholders may receive some of the gain from lower corporate tax. Workers may benefit from additional investment and employment, and consumers may benefit if cost changes pass into prices. The allocation depends on market power, capital mobility, labor mobility, timing and public revenue replacement. Assigning the entire benefit to shareholders or the entire cost to workers without an incidence analysis is unjustified.

The public finance consequences operate through expenditure and alternative tax instruments. A jurisdiction losing a corporate tax base may reduce public services, borrow, raise other revenue or pursue enforcement. Each response affects different people and generations. The OECD (1998) analysis of harmful tax practices identified the potential displacement of tax burdens toward less mobile bases. This proposition is a policy concern with an economic mechanism; the magnitude depends on actual government choices.

Crivelli, de Mooij and Keen (2015) examine base erosion in developing countries through cross-country corporate tax spillovers. International cooperation matters especially where the fiscal system has few readily administered alternatives to taxation of large enterprises. Nevertheless, a fiscal loss as a share of GDP, a share of corporate revenue and a share of total public revenue are not interchangeable. Econometric estimates derived from differences across countries are subject to specification, measurement and omitted-variable risks.

An economy supplying haven services must be analyzed as a host, not solely as a source of foreign tax losses. Potential gains include local salaries, financial-service fees, fiscal receipts and investment in technology and regulation. Potential vulnerabilities include concentration in an internationally mobile industry, exposure to foreign policy changes, high compliance costs and affordability pressures when demand for scarce local factors increases. A high ratio of booked profits to employment does not translate automatically into local household welfare.

Consider two hosts with the same recorded corporate profit. One has substantial local professional employment and domestically retained fees. The other retains a modest registration fee but little local value added. Real national income and resilience can differ despite identical external profit statistics. A development assessment ought to report locally retained value, public revenues net of administration, employment by occupation, household distribution, external liabilities and alternative productive uses of scarce resources. None of these outcomes follows mechanically from a label.

Secrecy, transparency and information economics

Financial secrecy affects the distribution of information among account owners, service providers and public authorities. The economic value of concealment depends upon the probability of detection, available enforcement and legal consequences. Legitimate confidentiality and information security also have economic value. A well-designed system protects personal data from inappropriate disclosure while permitting authorized tax authorities to obtain the information required for compliance.

Johannesen and Zucman (2014) evaluate the G20 tax haven crackdown, finding evidence consistent with deposits shifting to jurisdictions outside the reach of particular bilateral information agreements rather than necessarily returning home. The mechanism illustrates policy substitution: a rule covering only selected links can redirect mobile activity to uncovered links. Deposit movements are not direct observations of individuals' reasons or compliance, so conclusions remain at the level supported by the research design.

The OECD's (2025) automatic information exchange peer review reports that, for 2024, information on more than 171 million accounts with total reported value close to EUR 13 trillion was exchanged. The number counts financial accounts in the reporting system and should not be read as a number of distinct persons or cases of undeclared assets. The amounts reported are account values, not tax revenue or a measured tax gap. The report also distinguishes legal-framework assessments from effectiveness of implementation, which is essential for evaluating real enforcement capability.

The EU Tax Observatory's Global Tax Evasion Report 2024 (released 2023) estimates a roughly threefold reduction in offshore tax evasion over the preceding decade. This is an estimated change in evasion, rather than elimination of foreign wealth holdings or a count of convictions. Its explanatory framework employs a broad umbrella that includes some legal and contested tax-reduction practices; the household offshore finding should not be used to collapse lawful avoidance, legal residence choices and unlawful evasion into one legal category.

Substantial exchange coverage can reduce the informational advantage of traditional secrecy without erasing differences in tax rates or other institutional services. Effectiveness also depends on the ability of receiving authorities to match records to taxpayers, investigate discrepancies, protect information and apply due process. A country's participation in exchanges is meaningful institutional evidence, while the volume of exchanged records alone cannot prove the extent of compliance improvement.

Data and measurement architecture

Tax haven research needs a measurement architecture built around explicitly separated units and numerators. Different datasets register different economic objects, even when they refer to the same jurisdiction. The following crosswalk is an analytical guide to interpreting evidence; it is not a new quantitative index.

Observation Unit and measure Main interpretation constraint
Statutory tax rate Jurisdiction, tax type and year Does not capture exemptions, credits or effective liability
Corporate profitability Firm group or affiliate, profit relative to stated denominator Composition, loss treatment and productivity affect comparisons
Direct investment position Resident-to-resident cross-border claims at a date Intermediation and pass-through can multiply recorded positions
Offshore household financial wealth Asset stock attributed to owners' tax residence Coverage, ownership attribution and portfolio composition
Information exchange Reportable accounts and values, reporting period Accounts are not unique people, liabilities or measured evasion
Revenue loss Counterfactual difference in taxes for a period Depends on base allocation, effective rates and behavior

The OECD (2026) country-by-country reporting statistics cover fiscal year 2023 in their newest data installment: 60 submitting jurisdictions provided statistics concerning more than 9,400 multinational enterprises. They are anonymized and aggregated and cannot substitute for individual tax audits. Data are particularly useful for examining the relationships among business activity, reported profit and tax, while aggregation may mask firm heterogeneity and limit causal inference.

A credible reconciliation distinguishes reported tax accrued from cash paid, profits before tax from taxable profits, tangible assets from financial claims and employee count from labor compensation. Loss-making entities require separate treatment. Consolidated revenue used to establish reporting scope differs from the sum of affiliate revenues containing intragroup transactions. Exchange-rate conversions, reference periods and confidentiality suppression can change comparisons.

Statistical outliers should prompt inquiry rather than automatic removal or automatic designation of illegality. Investigators should retain the original reporting unit, define rules for exclusions and transformations before evaluating outcomes where feasible, and report how results vary under alternative reasonable specifications. Reproducibility requires the same source vintages, stable identifiers, exact definitions and documented code; it cannot be created merely by an accessible spreadsheet.

The quantitative estimates cited throughout this study are taken from the respective publications, and their data have not been pooled into a synthetic worldwide total. The FDI percentage and the multinational profit percentage, for example, have different dimensions and denominators. Adding them would be equivalent to adding a fraction of a stock to a fraction of an income flow.

Methods, identification and uncertainty

The analysis of tax havens involves at least four distinct empirical tasks. A descriptive study maps the economies or regimes concerned; a statistical accounting study estimates discrepancies in the allocation of stocks or income; a causal study tests the effect of a policy or institutional change; and a welfare study assesses benefits and costs through a stated counterfactual. The evidentiary standards differ.

A descriptive comparison needs a coherent sampling frame and measurement period. If a variable is itself used to classify an economy as a haven, subsequent correlations with that variable cannot be treated as independent confirmation of causal significance. Results from narrow tax-regime classifications may differ from results using financial-center or network classifications. Sensitivity analysis should state which jurisdictions and regimes enter each definition.

A statistical accounting study can exploit discrepancies among national accounts, foreign affiliate statistics and external positions. Such methods often involve estimating missing information or assigning a quantity to an ultimate owner using indirect evidence. A discrepancy is a clue to the allocation issue, not a uniquely identified cause. Revisions to the source statistics and assumptions can change inferred magnitudes. Uncertainty should include sensitivity to allocation assumptions, not just sampling standard errors.

Causal analysis confronts endogeneity because governments and firms choose policy and location in response to expected opportunities. A regime may be introduced because investment was already rising or falling. Firms with the greatest international presence may be more exposed to a reform and also follow different trends. For policy studies, researchers should examine pretreatment trajectories, concurrent reforms, time-varying implementation, heterogeneous effects, potential spillovers and attrition. Reporting an average effect across a set of countries can conceal economically important differences.

Instrumental-variable methods require relevance and an exclusion restriction. An instrument that predicts a firm's international expansion may also affect its profitability through business-cycle demand, independently of tax haven use. Studies using a particular instrument must explain the mechanism and defend exclusion in context. A natural experiment is valuable because its assignment mechanism can supply stronger evidence than ordinary cross-sectional associations, but its external validity remains conditional.

Welfare evaluation requires comparing a specified policy alternative against the status quo and valuing real effects separately from transfers. A fiscal transfer among governments is not in itself an equal-sized resource loss for the world; real resource costs arise through distorted choices, administrative expenditure, reduced or misallocated public services and other behavioral or distributional effects. The distribution of tax revenue still matters for welfare because governments and households have different needs and capacities.

The evidentiary limits of this dossier are explicit. It synthesizes retrieved primary institutional records and original research abstracts or accessible excerpts. Some journal full texts are restricted or failed direct retrieval. No original firm-level estimation, independent network reconstruction or comprehensive legal audit has been undertaken. Quantitative claims are restricted to identifiable published estimates and transparent hypothetical illustrations. These limitations constrain the strength of any broad causal conclusion.

International policy and minimum taxation

Institutional reform addresses distinct margins of tax haven activity. Tax information exchange targets hidden ownership and undeclared taxable accounts. Corporate disclosure improves observation of where economic activity and reported income occur. Rules on substantive economic activity address the gap between formal registration and operations. Minimum taxation addresses specified differences in effective tax rates. None of these instruments serves exactly the same purpose.

The OECD Global Minimum Tax overview describes a coordinated system in which in-scope multinational groups calculate jurisdictional effective tax rates under the Global Anti-Base Erosion rules. The 15% minimum applies under the rules to excess profits after the substance-based exclusion, subject to the actual scope and detailed calculations. The broadly stated revenue threshold is EUR 750 million in consolidated revenues in at least two of the four preceding fiscal years, with exclusions and other qualifications. Domestic enactments, transition provisions and accepted safe harbors determine application.

The simplified arithmetic of a minimum tax can illustrate its economic effect. Suppose an in-scope group's GloBE income in a jurisdiction is 100, covered taxes are 5, and the hypothetical substance-based exclusion is 20. The simplified jurisdictional effective rate is 5%, the difference to 15% is 10 percentage points, and applying it to excess profit of 80 gives a stylized top-up of 8. The example omits adjustments, domestic top-up mechanisms, special treatment and safe harbors. It illustrates why the rate floor alone does not imply 15 units of tax on every 100 units of gross income.

A host may respond to the prospective transfer of top-up revenue elsewhere by levying a qualifying domestic tax or by changing other aspects of its investment policy. Substantive operations can retain an attraction even when the fiscal advantage on excess profit is reduced. The final incidence depends on coordinated implementation, the treatment of subsidies, qualifying rules and the group's alternative choices.

The OECD's September 2026 implementation package addresses consistent application, legislative review, the GloBE Information Return and further guidance reflecting the January 2026 Side-by-Side agreement. The existence of this guidance demonstrates that international implementation is evolving. It does not establish that every jurisdiction has enacted the same rules or achieved the same revenue outcome. Country-specific law and actual application need separate dated research.

Individual income and wealth taxation, smaller enterprises outside the GloBE scope, preferential regimes and institutional competition remain distinct questions. Global minimum taxation changes incentives over a specified segment of the corporate economy; it cannot by itself resolve the full economic problems associated with tax havens.

Policy incidence and comparative evaluation

A useful policy assessment starts with a clearly stated target. If the objective is to improve personal tax compliance, an instrument should be evaluated against accurately measured changes in undeclared taxable income, information usability and administrative costs. If the objective is to reduce corporate profit shifting, assess the relation between profit, actual economic activity and tax liabilities. If the objective is local development, assess local value added and opportunity costs. The same policy may perform differently under these objectives.

One evaluation challenge is substitution between jurisdictions, assets and organizational forms. An observed decrease in balances at selected financial institutions can coincide with an increase in other reported or unreported assets. Researchers should not transform incomplete displacement data into an assumption that funds have been repatriated. The historical results summarized in Johannesen and Zucman's (2014) evaluation of information agreements motivate the need for broader information coverage and careful choice of outcome.

Another challenge is the distribution of newly collected revenues. A successful minimum-tax regime can reduce a differential while reallocating collection rights among host and residence economies. This may leave the global fiscal total larger while distributing the gains differently from a source-country reform. Public-finance evaluation needs to distinguish total liability, who collects the tax, administrative burden, taxpayer incidence and the distribution of public expenditure.

Policy comparisons also need to consider institutions and incentives on both sides of the border. An authority unable to make use of shared information may experience less benefit from exchange than one with sophisticated matching and investigative capacity. A small international financial center can face compliance costs disproportionate to its domestic tax administration. Capacity building can therefore affect both enforcement effectiveness and the economic burden of reform.

An evidence-based evaluation should report effects on real productive investment, recorded income, revenue, compliance resources and the quality of local institutions, then examine whether the same conclusions survive an alternative counterfactual and a longer horizon. Predicted policy outcomes should remain distinguishable from measured outcomes. The relevant question is not whether all international financial specialization disappears; it is whether the targeted economic distortion or information asymmetry is reduced at an acceptable real and distributional cost.

Integrated economic synthesis

The literature supports an interpretation of tax havens as a form of jurisdictional economic specialization that connects fiscal competition with the organization of mobile capital. Three separations are particularly consequential for research.

First, policy parameters and observed economic outcomes must be distinguished. Favorable treatment may attract a formal base, genuinely productive activity, intermediation or a mixture. A statistical concentration of financial claims or reported profits is suggestive of international specialization; its source, scale and fiscal consequences require further evidence.

Second, corporate and household behavior must be analyzed through their respective units. In corporate studies, the central discrepancy is often between recorded profit and the location of business functions, with tax effects measured through counterfactual liabilities. In household studies, the central discrepancy can be between observable financial-account location and the tax residence or disclosure status of the ultimate owner. An information reform can affect one of these margins without directly eliminating the other.

Third, local fiscal success and global welfare are analytically distinct. A host may retain fees, wages and tax even if another economy loses a larger amount of revenue. The appropriate comparison traces transfers and real resource effects separately. Equally, a tax advantage that induces real investment can deliver benefits and costs to different groups. The size and direction of a welfare effect depend on the actual alternative, not exclusively on the statutory rate.

These separations yield a practical interpretation of heterogeneous evidence. When a large offshore ownership network, a large FDI position and high booked profit are observed in one economy, it is tempting to treat them as three independent measurements of the same underlying quantity. In reality, the observations may overlap substantially: one entity can connect all three statistics. Constructing a single aggregate by adding them would double count economic exposure. Conversely, a low measured FDI position does not rule out a substantial household wealth sector.

The evidence also suggests a distinction between the persistence of financial-center functions and the persistence of specific tax advantages. A mature center may continue providing valuable administration, access to markets and professional expertise after changes in minimum-tax rules or reporting. A jurisdiction's economic transition therefore depends on the adaptability of its service ecosystem, the behavior of its international clients and the nature of the institutional reform. The continued existence of an offshore center after a reform is not by itself evidence that the reform failed.

These are analytical conclusions synthesized from existing research and standard economic reasoning. No new named theory, doctrine, empirical estimate or claim of priority follows from them.

Research frontiers

Further scientific progress is likely to depend on the ability to distinguish a jurisdiction's formal financial claims from the location and distribution of real economic outcomes. Publicly available datasets still vary in definitions, periodicity and access. Linking ultimate ownership, productive activity and tax records while protecting personal information could improve both measurement and causal evaluation.

Comparative studies should follow policy changes over multiple years. Tax decisions and asset valuation can respond before a law enters into operation, while employment and infrastructure may respond more slowly than reported profit. A study of an announcement and a study of realized fiscal receipts answer different questions. The transition to global minimum taxation provides motivation for examining several margins together, but institutional changes in 2026 make definitive long-run conclusions premature.

A useful agenda concerns the distributional effects within host economies. International research has often concentrated on the counterpart losses of larger economies or the aggregate scale of offshore financial positions. Local household income, occupational composition, consumer prices and public budgets can reveal the domestic distribution of gains and risks. The same aggregate inflow may distribute benefits quite differently across host communities.

Administrative capacity is another independent research subject. Information may be received in a standardized format while matching, verification and enforcement remain resource constrained. Studies could compare not only the legal existence of exchange relationships but their practical usability and downstream outcomes with appropriate confidentiality and due process.

Finally, the geographic unit of analysis should remain revisable as the system evolves. Countries, dependent territories, special regimes, global cities and financial institutions can occupy different layers of the same network. A historically established list is useful for a bounded study, but a long-run scientific program requires dated definitions and transparent criteria rather than a frozen inventory.

Conclusion

Tax havens arise where internationally mobile economic claims encounter differences in territorial fiscal institutions. Their development reflects interaction among government policy, corporate and household demand, professional services, institutional credibility and the wider network of international finance. An economic explanation must account for both the supply of jurisdictional services and the demand for differential treatment.

The evidence reviewed here establishes the value of separating profit location from productive activity, financial claims from their ultimate owners, and gross fiscal advantage from final welfare. Published work estimates substantial profit relocation and pass-through FDI; these research results are distinct in scale, time and denominator. Research on offshore wealth reveals significant challenges for measuring personal asset ownership and inequality. Transparency reforms have expanded reporting coverage, while minimum-tax reforms alter some corporate incentives; both demand assessment through actual implementation and outcomes.

For governments, the principal challenge is to observe the relevant economic activity, specify the fiscal base and evaluate alternative rules in an interconnected system. For firms and households, the distinction between legitimate international organization and noncompliance is grounded in applicable law and facts. For economic science, the key requirement is consistent measurement: a tax haven cannot be adequately understood through a headline rate, a single list, a gross balance sheet or a morality-based classification.

This dossier concludes that the most informative comparative assessment employs dated jurisdictional evidence, matched statistical units, explicit counterfactuals, separately evaluated local and international effects, and transparent limitations. Such research makes the economics of tax havens intelligible within the wider Offshore Economics program while preserving the autonomy of legal doctrine, industry-specific finance and behavioral research.

References

Alstadsæter, A., Johannesen, N., and Zucman, G. (2018). Who Owns the Wealth in Tax Havens? Macro Evidence and Implications for Global Inequality. Journal of Public Economics, 162, 89–100. Publication record.

Alstadsæter, A., Johannesen, N., and Zucman, G. (2019). Tax Evasion and Inequality. American Economic Review, 109(6), 2073–2103. DOI: 10.1257/aer.20172043. Publication record.

Beer, S., de Mooij, R., and Liu, L. (2020). International Corporate Tax Avoidance: A Review of the Channels, Magnitudes, and Blind Spots. Journal of Economic Surveys, 34(3), 660–688. DOI: 10.1111/joes.12305. Publication record.

Crivelli, E., de Mooij, R. A., and Keen, M. (2015). Base Erosion, Profit Shifting and Developing Countries. IMF Working Paper WP/15/118. International Monetary Fund, Washington, DC. Publication record.

Damgaard, J., Elkjaer, T., and Johannesen, N. (2019). What Is Real and What Is Not in the Global FDI Network? IMF Working Paper WP/19/274. International Monetary Fund, Washington, DC. Publication record.

Desai, M. A., Foley, C. F., and Hines, J. R., Jr. (2006). Do Tax Havens Divert Economic Activity? Economics Letters, 90(2), 219–224. Publication record.

Desai, M. A., Foley, C. F., and Hines, J. R., Jr. (2006). The Demand for Tax Haven Operations. Journal of Public Economics, 90(3), 513–531. Publication record.

Dharmapala, D., and Hines, J. R., Jr. (2009). Which Countries Become Tax Havens? Journal of Public Economics, 93(9–10), 1058–1068. Publication record.

EU Tax Observatory (2023). Global Tax Evasion Report 2024. Paris: EU Tax Observatory. Publication record.

Garcia-Bernardo, J., Fichtner, J., Takes, F. W., and Heemskerk, E. M. (2017). Uncovering Offshore Financial Centers: Conduits and Sinks in the Global Corporate Ownership Network. Scientific Reports, 7, 6246. DOI: 10.1038/s41598-017-06322-9. Publication record.

Hines, J. R., Jr., and Rice, E. M. (1994). Fiscal Paradise: Foreign Tax Havens and American Business. Quarterly Journal of Economics, 109(1), 149–182. Publication record.

Johannesen, N., and Zucman, G. (2014). The End of Bank Secrecy? An Evaluation of the G20 Tax Haven Crackdown. American Economic Journal: Economic Policy, 6(1), 65–91. DOI: 10.1257/pol.6.1.65. Publication record.

OECD (1998). Harmful Tax Competition: An Emerging Global Issue. OECD Publishing, Paris. DOI: 10.1787/9789264162945-en. Publication record.

OECD (2025). Peer Review of the Automatic Exchange of Financial Account Information 2025 Update. OECD Publishing, Paris. Executive Summary. DOI: 10.1787/bbf150e4-en. Publication record.

OECD (2026). Corporate Tax Statistics 2026. OECD Publishing, Paris. Chapter: Country-by-country reporting statistics. Publication record.

OECD (2026). Global Minimum Tax: Inclusive Framework Releases a Package to Strengthen Consistency and Certainty for MNEs and Jurisdictions. September 2026. OECD announcement. Publication record.

OECD (n.d.). Global Minimum Tax. OECD tax policy thematic overview, accessed November 18, 2026. Publication record.

Palan, R. (2002). Tax Havens and the Commercialization of State Sovereignty. International Organization, 56(1), 151–176. DOI: 10.1162/002081802753485160. Publication record.

Slemrod, J., and Wilson, J. D. (2009). Tax Competition with Parasitic Tax Havens. Journal of Public Economics, 93(11–12), 1261–1270. Publication record.

Tørsløv, T., Wier, L., and Zucman, G. (2023). The Missing Profits of Nations. Review of Economic Studies, 90(3), 1499–1534. DOI: 10.1093/restud/rdac049. Publication record.

Zoromé, A. (2007). Concept of Offshore Financial Centers: In Search of an Operational Definition. IMF Working Paper WP/07/87. International Monetary Fund, Washington, DC. Publication record.