Data explorer: Unitary taxation and formulary apportionment
- Read this in:
- en
Table of contents
This page brings together the data and visualisations from a new study on unitary taxation and formulary apportionment, conducted by the WIFO Institute and commissioned by the Network of Unions for Tax Justice and the Austrian Chamber of Labour.
For more analysis beyond this data explorer, check out he accompanying blog post summarizing the main findings and implications, and the full research report for detailed analysis and methodology.
The visualisations below explore how the global corporate tax base would be redistributed under different apportionment formulas - rules that determine how a multinational group’s global profits are shared between countries.
Understanding Unitary Taxation
Under the current transfer-pricing system, multinational enterprises (MNEs) are taxed country by country as if each subsidiary were a separate firm. This encourages profit shifting towards low-tax jurisdictions.
Unitary taxation approaches profit allocation differently. It treats a multinational group as a single firm, calculates the group’s global profits, and then distributes those profits between countries using indicators of real economic activity, for example, workers, assets or sales.
Different apportionment formulas give different weight to these indicators, and therefore produce different redistribution patterns. The visuals below illustrate these effects.
Understanding the apportionment formulas
The formulas determine how global profits are allocated across countries. They fall into two groups: single-factor formulas and weighted (multi-factor) formulas.
Single-factor formulasThese allocate 100% of profits using one indicator: | Weighted (multi-factor) formulasThese combine several indicators to allocate profits. Their behaviour depends on the relative weights assigned to each factor. |
Physical assets: buildings, equipment, and other tangible assets | Workforce-weighted (headcount) |
Workforce (headcount): number of employees | Workforce-weighted (wages) |
Workforce (wages): total employee compensation | Multi-factor weighted (EU CCCTB) |
Sales recorded by firms: revenues booked in each jurisdiction | Sales-weighted (double sales) |
Customer location (sales): sales to final consumers, approximated using “output minus exports” Due to data limitations, some scenarios had to be run using sales recorded by firms (notably through turnover data). For such scenarios, the results may be overstated for exporting countries and understated for market countries, such as large Global South economies. These scenarios apply to graph 1-3 and are marked with an asterix (*). |
The exact ratios are provided below.
Ratios for weighted formulas
Workforce-weighted (headcount)(Massachusetts formula – employees) | Workforce-weighted (wages)(Massachusetts formula – payroll) | Multi-factor weighted(EU CCCTB) | Sales-weighted (double sales)(Double-weight sales formula) |
-- 1/3 workforce (headcount) | -- 1/3 workforce (wages) | -- 1/3 sales recorded by firms | -- 1/2 sales recorded by firms |
-- 1/3 physical assets | -- 1/3 physical assets | -- 1/3 physical assets | -- 1/4 physical assets |
-- 1/3 sales recorded by firms | -- 1/3 sales recorded by firms | -- 1/6 workforce (headcount) | -- 1/8 workforce (headcount) |
-- 1/6 workforce (wages) | -- 1/8 workforce (wages) |
Understanding the reallocation effects
Unitary taxation and formulary apportionment would have a significant impact on where corporate profits are taxed.
Tax havens and low-tax countries emerge as the largest losers under any allocation formula. These findings underline the fragility of economic models built on paper profits and heavy tax incentives, and highlight the need to reflect on more sustainable strategies rooted in real economic activity.
Conversely, the different scenarios show a consistent increase in tax revenues for high-income countries. This contradicts the common assumption in international tax debates that high-income, capital-exporting countries would be disadvantaged under formulary apportionment.
Importantly, some of the scenarios outlined below indicate that low-income countries may also lose out, to varying degrees depending on the choice of allocation formula. This reflects both the need for these countries to increase their effective tax rates and the importance of carefully designing sectoral approaches, so that allocation formulas can be properly adjusted to domestic economic contexts. Global South countries are already aware of these challenges. For example, they have successfully advocated for the exclusion of extractive industries from the scope of Pillar One Amount A. A formulary apportionment mostly based on customer presence could indeed have an adverse impact on economies largely reliant on export commodities.
How global tax revenues could change under unitary taxation
How apportionment formulas and tax rates change corporate tax revenues across income groups (graph 1*)
The revenue impact of unitary taxation depends heavily on the tax rate applied to the newly allocated profits. This visual compares three approaches: using statutory tax rates, using effective tax rates and combining statutory rates with a 25% global minimum tax.
Applying statutory tax rates results in broad revenue gains, with high-income countries capturing most of the increase. These results, however, do not mean much because in practice profits are taxed at reduced rates or exempt due to various tax incentives.
Using effective tax rates, which better reflect what multinational groups currently pay, leads to smaller and sometimes slightly negative global outcomes. A major driver of these outcomes is the low rate effectively applied in many countries. Under a worldwide switch to formulary apportionment, tax competition between countries is expected to be curbed. This is because corporate profits would be allocated according to objective factors, rather than being influenced by accounting manipulation. In such a system, effective tax rates could rise without countries fearing the loss of real investment. Combining statutory rates with a 25% minimum tax generates more stable results. Short-term losses are largely reduced or reversed, and even low-tax investment hubs tend to raise more revenue under this scenario. This highlights why a 25% minimum tax is a key priority for the labour movement: it helps ensure multinationals pay a fairer share and prevents low-tax jurisdictions from undermining the benefits of unitary taxation.
Switch between tax rate scenarios and click on the bars to view detailed information for each income group under different apportionment formulas.
How countries’ corporate tax revenues shift under unitary taxation
(Table 1)
Use this table to check the estimated change in corporate tax revenues for each country under all apportionment formulas, using statutory tax-rate assumptions.
How the global corporate tax base would shift under unitary taxation
How different apportionment formulas change the tax base across income groups (graph 2*)
This chart compares how much the tax base changes for each income group under different unitary taxation formulas.
Investment hubs consistently lose tax base, while high- and upper-middle-income groups gain the most in absolute terms, reflecting their larger scale of real economic activity.
Workforce-weighted formulas shift relatively more tax base towards countries with large labour footprints, while formulas that give more weight to tangible assets (such as the EU CCCTB) produce smaller redistributions. Sales-weighted formulas channel more gains towards big consumer markets.
Use the menu at the top of the chart to switch between apportionment formulas. Click on each bar to view the detailed apportionment factors behind the results.
Data considerations
These estimates rely mainly on aggregate country-by-country reporting (CbCR) data. While this provides broad global coverage, the dataset has two key constraints. In some smaller jurisdictions, the baseline tax base appears unusually small because the aggregated figures are dominated by large loss-making multinationals, which distorts percentage changes. In addition, although activities span roughly 200 jurisdictions, the data come from multinationals headquartered in only 56 countries, which shapes the overall distribution.
These limitations do not change the broad regional patterns but do affect the precision of individual country results. More methodological detail is provided in the section below.
How the tax base shifts across world regions under different formulas (graph 3*)
After exploring the differences within each region, the regional averages offer a clearer view of the broad global shifts produced by unitary taxation. Across almost all formulas, the largest reductions in corporate tax base occur in regions dominated by financial and investment hubs, where significant profits are currently booked despite relatively limited real economic activity.
Once these reductions are accounted for, the reallocated tax base is distributed mainly towards regions hosting large markets, significant production activity, or substantial workforces. North America, Western and Northern Europe, and parts of Western Asia and Eastern Europe generally show positive movements under most formulas.
A further theme running through the results is that high-income and upper-middle-income regions retain the overwhelming majority of the global corporate tax base under all formulas. This is not a consequence of the reform itself, but a reflection of today’s global economic structure: the main apportionment factors, workers, assets and sales, remain concentrated in these regions, and the formulas largely track where economic activity actually takes place.
How country-level tax-base changes under different formulas (table 2)
Use this table to check the estimated percentage change in the corporate tax base for countries across all apportionment formulas.
Once these reductions are accounted for, the reallocated tax base is distributed mainly towards regions hosting large markets, significant production activity, or substantial workforces. North America, Western and Northern Europe, and parts of Western Asia and Eastern Europe generally show positive movements under most formulas.
A further theme running through the results is that high-income and upper-middle-income regions retain the overwhelming majority of the global corporate tax base under all formulas. This is not a consequence of the reform itself, but a reflection of today’s global economic structure: the main apportionment factors, workers, assets and sales, remain concentrated in these regions, and the formulas largely track where economic activity actually takes place.
Technical Notes
Data sources
The report works with three main types of data:
Firm-level accounts (unconsolidated ORBIS data)
These are company-level financial statements for individual legal entities. They provide detailed information on assets, employees and profits, but coverage is uneven: data are relatively good for many European groups and weaker for some large economies and low-tax jurisdictions.
Aggregate country-by-country reporting (CbCR) data
These are aggregated disclosures by multinationals on where they book profits, pay tax and employ staff, published at a country–country level. They offer broad global coverage but in highly aggregated form, and they do not allow loss consolidation to be modelled directly.
Hybrid approaches combining firm-level and aggregate data
The hybrid method links firm-level ORBIS information with aggregated CbCR and other macro statistics (including foreign affiliate data) to approximate apportionment factors such as sales by destination. This approach offers the widest country coverage, but it inherits limitations from both underlying sources.
For estimating potential changes in tax revenues, the graphs on this page use the hybrid approach that combines firm-level financial accounts (from ORBIS) with aggregated CbCR data.
For the graphs on this page, the allocation of the corporate tax base across countries relies mainly on aggregate country-by-country reporting (CbCR) data.
Treatment of small baselines
Some countries report a very small corporate tax base in the underlying data. In these cases, even modest absolute changes can result in extremely large percentage movements, which are not meaningful when data are incomplete or unstable. To avoid presenting misleading relative effects, the tax-base visualisations on this page exclude countries whose baseline corporate tax base is below 1 billion USD/EUR. These jurisdictions remain included in the full report - particularly in the detailed tables and discussion - but are omitted here to ensure that the graphical patterns remain interpretable.
International loss consolidation has different short-term and long-term impacts
Switching to unitary taxation would bring together all the past losses of multinational groups at the global level. Because losses can be offset against profits, this produces a temporary dip in the worldwide tax base, strongest in the first years of a new system, then fading as older losses expire.
Most simulations therefore show a smaller tax base in the early years, with effects becoming more modest over time.
The largest-coverage parts of the analysis use aggregate data (from country-by-country reporting and hybrid methods).
These data do not allow the loss-consolidation effect to be modelled.
As a result:
The global tax base is held constant in these simulations, even though in reality it would fall temporarily.
Some gains may appear slightly overstated, especially where results rely on these aggregate methods.
This does not change the direction of redistribution, but it affects the scale.
Results are highly sensitive to which tax rate is assumed
The graphs on this page focus on how the tax base is redistributed under different formulas. What countries actually collect in revenue will ultimately depend on:
• the tax rates they apply nationally, and
• any global minimum tax or other common rules that are agreed.
The report shows that overall global revenue effects are sensitive to these choices. Using higher or lower effective tax rates in the simulations can shift the aggregate result from a small global loss to a small global gain. The pattern of who gains and who loses in terms of tax base, however, is much more stable across assumptions.
Sector-specific tax regimes are not included
The simulations treat all profits as if they were taxed under the same statutory rate.
This means the model does not take account of:
higher tax rates applying to certain sectors (e.g. oil and gas), or
special regimes such as patent boxes or incentives.
Countries with large extractive sectors or significant special regimes may therefore see actual revenues differ from the modelled averages.
Behavioural changes are not modelled
The simulations assume that firms do not immediately change where they place workers, physical assets or sales in response to unitary taxation.
In reality, some adjustment is likely over time. However, relocating real activity, shifting factories, offices or jobs, is much more costly than shifting profits on paper, so very large movements are unlikely in the short term.
The exact size of any behavioural response is uncertain, and the graphs should be read as medium-term static estimates, not forecasts of every future adjustment.
Measuring sales and customer location
Some of the indicators used to allocate profits in the formulas are difficult to measure directly, especially at the global level.
Two issues are particularly important:
a. Sales by destination must be approximated
Firm accounts usually record where sales are booked, not where customers are located.
To estimate customer-location sales, the study must rely on a proxy: output minus exports, drawn from sector-level data.
This approximation can:
overstate gains for very large markets (e.g. the US and China), and
understate complexity in global value chains.
b. Sales recorded by firms may include intra-group transactions
When revenues include related-party sales, results can be distorted by transfer pricing arrangements. Excluding these data is not always possible across all countries
Access the full report
Corporate taxation and employment: dispelling the race-to-the-bottom myth
Technical notes
Data sources
The report works with three main types of data:
Firm-level accounts (unconsolidated ORBIS data)
These are company-level financial statements for individual legal entities. They provide detailed information on assets, employees and profits, but coverage is uneven: data are relatively good for many European groups and weaker for some large economies and low-tax jurisdictions.
Aggregate country-by-country reporting (CbCR) data
These are aggregated disclosures by multinationals on where they book profits, pay tax and employ staff, published at a country–country level. They offer broad global coverage but in highly aggregated form, and they do not allow loss consolidation to be modelled directly.
Hybrid approaches combining firm-level and aggregate data
The hybrid method links firm-level ORBIS information with aggregated CbCR and other macro statistics (including foreign affiliate data) to approximate apportionment factors such as sales by destination. This approach offers the widest country coverage, but it inherits limitations from both underlying sources.
For the graphs on this page, the allocation of the corporate tax base across countries relies mainly on aggregate country-by-country reporting (CbCR) data.
For estimating potential changes in tax revenues, the graphs on this page use the hybrid approach that combines firm-level financial accounts (from ORBIS) with aggregated CbCR data.
Treatment of small baselines
Some countries report a very small corporate tax base in the underlying data. In these cases, even modest absolute changes can result in extremely large percentage movements, which are not meaningful when data are incomplete or unstable. To avoid presenting misleading relative effects, the tax-base visualisations on this page exclude countries whose baseline corporate tax base is below 1 billion USD/EUR. These jurisdictions remain included in the full report—particularly in the detailed tables and discussion—but are omitted here to ensure that the graphical patterns remain interpretable.
International loss consolidation has different short-term and long-term impacts
Switching to unitary taxation would bring together all the past losses of multinational groups at the global level. Because losses can be offset against profits, this produces a temporary dip in the worldwide tax base, strongest in the first years of a new system, then fading as older losses expire.
Most simulations therefore show a smaller tax base in the early years, with effects becoming more modest over time.
The largest-coverage parts of the analysis use aggregate data (from country-by-country reporting and hybrid methods).
These data do not allow the loss-consolidation effect to be modelled.
As a result:
The global tax base is held constant in these simulations, even though in reality it would fall temporarily.
Some gains may appear slightly overstated, especially where results rely on these aggregate methods.
This does not change the direction of redistribution, but it affects the scale.
Results are highly sensitive to which tax rate is assumed
The graphs on this page focus on how the tax base is redistributed under different formulas. What countries actually collect in revenue will ultimately depend on:
• the tax rates they apply nationally, and
• any global minimum tax or other common rules that are agreed.
The report shows that overall global revenue effects are sensitive to these choices. Using higher or lower effective tax rates in the simulations can shift the aggregate result from a small global loss to a small global gain. The pattern of who gains and who loses in terms of tax base, however, is much more stable across assumptions.
Sector-specific tax regimes are not included
The simulations treat all profits as if they were taxed under the same statutory rate.
This means the model does not take account of:
higher tax rates applying to certain sectors (e.g. oil and gas), or
special regimes such as patent boxes or incentives.
Countries with large extractive sectors or significant special regimes may therefore see actual revenues differ from the modelled averages.
Behavioural changes are not modelled
The simulations assume that firms do not immediately change where they place workers, physical assets or sales in response to unitary taxation.
In reality, some adjustment is likely over time. However, relocating real activity, shifting factories, offices or jobs, is much more costly than shifting profits on paper, so very large movements are unlikely in the short term.
The exact size of any behavioural response is uncertain, and the graphs should be read as medium-term static estimates, not forecasts of every future adjustment.
Measuring sales and customer location
Some of the indicators used to allocate profits in the formulas are difficult to measure directly, especially at the global level.
Two issues are particularly important:
a. Sales by destination must be approximated
Firm accounts usually record where sales are booked, not where customers are located.
To estimate customer-location sales, the study must rely on a proxy: output minus exports, drawn from sector-level data.
This approximation can:
overstate gains for very large markets (e.g. the US and China), and
understate complexity in global value chains.
b. Sales recorded by firms may include intra-group transactions
When revenues include related-party sales, results can be distorted by transfer pricing arrangements. Excluding these data is not always possible across all countries.