Which single number is the right one for global corporate tax avoidance?

There isn't one, and anyone who gives you a single figure without a year and a method is overselling it. The OECD's cautious 2014 range is US$100-240 billion. The Tax Justice Network's 2021-data estimate for the corporate slice is US$347.6 billion. The academic estimate of profit shifted into havens is about US$1 trillion, which is a different thing again (shifted profit, not lost tax). Pick the measure that fits your question, and always carry its date.

Is paying an effective rate well below the headline rate illegal?

No. An effective rate below the statutory rate is the normal, lawful result of deductions, credits, allowances and the lawful location of profit and intellectual property. The point of this site is not that anyone is breaking the law. It is that the law lets profit be taxed a long way from where it was earned, and that the tools to do this are only available at a scale most businesses never reach.

Does Pillar Two fix this?

Partly, on paper, and it is early. The OECD's own modelling expects the share of profit taxed below the minimum to fall by roughly 80% and projects US$155-192 billion a year in extra revenue. That is a large dent, not a cure, and implementation across jurisdictions is still unfolding. We will report what the data shows as it arrives, rather than promise an outcome.

Why do you keep citing the Tax Justice Network if it is an advocacy group?

Because its estimates are built on the OECD's own country-by-country reporting data using a peer-reviewed method, and published in full. We label them as estimates, we cite the underlying OECD and academic work alongside, and where a government audit figure exists we give that too. The aim is to show the workings, not to ask for trust.

By the Numbers: Corporate Tax Avoidance, From Primary Sources

Name: By the Numbers: Corporate Tax Avoidance, From Primary Sources
Creator: LoopholeKiln
Published: 2026-06-17T22:46:59.591862+00:00
License: https://creativecommons.org/licenses/by/4.0/

ByLoopholeKiln EditorialPublished17 June 2026

Figures current as of1 January 2026·Corrections

Here is the short version, with the receipts attached. The world loses somewhere between US$100 billion and roughly US$350 billion a year in corporate tax to profit being moved across borders, depending on whose method you use and which year you count. Around 36% of all multinational profit ends up booked in tax havens. None of this is a leak, a scam or a crime. It is the gap between where money is earned and where the rules let it be taxed. Below is every big number we rely on, with its source and its year, so you can check us.

We have pinned each figure once, with its origin and the year of the underlying data. Estimates are labelled as estimates. Where a number has been superseded by newer data or law, we say so. If a figure is missing a primary source, it is not on this page.

The global scale of the loss

Figure	What it measures	Source and year
US$100-240 billion/yr	Revenue lost to base erosion and profit shifting, equal to 4-10% of global corporate income tax receipts at 2014 levels	OECD, BEPS Action 11 Final Report, 2015 (data ~2014)
US$616 billion	Profit shifted to havens, with about 40% of multinational profit in havens, on 2015 data	Torslov, Wier and Zucman, "The Missing Profits of Nations," NBER Working Paper 24701, 2018 (data 2015)
About 40% / about US$1 trillion	Share of multinational profit booked in havens, updated to 2019 on the authors' data portal (the published peer-reviewed paper states a slightly lower 36% of profit on 2015 data)	Torslov, Wier and Zucman, Review of Economic Studies, 2023, and missingprofits.world data portal (2019 update)
US$492 billion/yr	Total tax lost to cross-border abuse: US$347.6 billion corporate plus US$144.8 billion from undeclared offshore individual wealth	Tax Justice Network, State of Tax Justice 2024 (data year 2021)
US$347.6 billion/yr	The corporate slice of that loss, on 2021 country-by-country reporting data	Tax Justice Network, State of Tax Justice 2024 (data year 2021)
About US$1.42 trillion	Multinational profit shifted into havens in 2021, causing about US$348 billion in direct global corporate tax loss	Tax Justice Network, State of Tax Justice 2024 (data year 2021)

A word on why these numbers differ. The OECD's US$100-240 billion is an older, deliberately cautious range from 2014. The academic Torslov-Wier-Zucman work and the Tax Justice Network's estimates use later data and somewhat different methods, which is why they land higher. They are not contradicting each other; they are measuring slightly different things in different years. The honest summary is a range, not a single magic figure.

The reform meant to close the gap

Figure	What it measures	Source and year
15%	The global minimum effective corporate tax rate agreed under Pillar Two	OECD/G20 Inclusive Framework; OECD Economic Impact Assessment, January 2024
€750 million	The annual group revenue threshold above which a multinational falls within Pillar Two	OECD GloBE rules; effective from 2024
US$155-192 billion/yr	Projected additional corporate income tax revenue from the global minimum, equal to 6.5-8.1% of current global corporate income tax receipts	OECD Economic Impact Assessment, January 2024
From 36% down to about 7%	Projected fall in the share of multinational profit taxed below the minimum once Pillar Two applies, a reduction of roughly 80%	OECD Economic Impact Assessment, January 2024
About 55 jurisdictions	Number taking steps to implement the global minimum as of 2024	OECD GloBE implementation summary, 2024

Pillar Two is the first serious attempt to put a floor under the race to the bottom. Whether it works in practice is still being written. What can be said now is that it exists, the threshold and the rate are fixed, and the OECD's own modelling expects it to shrink the problem by most, though not all, of its size.

The race to the bottom, in one line

Figure	What it measures	Source and year
28.2% falling to 21.1%	The average statutory corporate tax rate across all covered jurisdictions, from 2000 to 2023	OECD Corporate Tax Statistics 2023
31.3% falling to 23.1%	The same average across jurisdictions that levy a non-zero rate, 2000 to 2023	OECD Corporate Tax Statistics 2023
Stabilised around 21.1%	Where the average has sat since roughly 2019, after two decades of decline	OECD Corporate Tax Statistics 2024

For twenty years, governments cut headline corporate rates to compete for investment. The average fell by about a third. It is the backdrop to everything else on this page: a lower starting rate, and then a set of tools that take the effective rate lower still for those large enough to use them.

What it looks like for one company

Figure	What it measures	Source and year
About 13.5%	The effective tax rate of one of the world's largest online retailers, on its own filed accounts: a tax provision of US$9,265 million on pre-tax income of US$68,614 million	The retailer's own FY2024 10-K filing, via SEC EDGAR
About 18.8%	The average global effective corporate income tax rate of the six largest technology companies over the decade to 2024, on US$2.5 trillion of profit, against statutory rates averaging closer to 27% to 30%	Fair Tax Foundation, Silicon Six report, 2025

We use these figures because they are exact, drawn from audited filings, and lawful. A profitable company can face a UK main rate of 25%, and a multinational of sufficient scale can show an effective rate of about 13.5%, or a ten-year global average of about 18.8% across six of the largest firms on earth, with nothing illegal anywhere in the chain. These global blended rates are the benchmark the comparison pages set every small business against. The difference is not effort or wrongdoing. It is access to tools that only exist at a scale a small business will never reach. Be cross with the rules, not the filer.

Canada and South Africa, against the same multinational benchmark

These two countries are sometimes shown against a domestic structure, a Canadian-controlled private corporation and a small business corporation, because those are the cleanest like-for-like comparisons in their own countries. But the site's core thesis is the small business against the global giant, so here are both countries set against the same global multinational benchmark as everywhere else.

Figure	What it measures	Source and year
28.3% vs about 18.8% = about 1.5x	A self-employed electrician in Ontario on C$100,000 of profit (2024) against the six largest technology companies' global ten-year average effective rate. The 18.8% is a worldwide blended figure, not a Canadian rate	Canada: CRA, Ontario Ministry of Finance, CRA CPP 2024 maximums; benchmark: Fair Tax Foundation, Silicon Six report, 2025
22.2% vs about 18.8% = about 1.2x	A self-employed professional in South Africa on R600,000 of profit (2026/27) against the same global ten-year average. The 18.8% is a worldwide blended figure, not a South African rate	South Africa: SARS, rates of tax for individuals, 2027 tax year; benchmark: Fair Tax Foundation, Silicon Six report, 2025

Both also have a cleaner same-country structure comparison, set out below for South Africa and on the comparison pages in full: in Canada the sole trader pays about 2.3 times a CCPC's 12.2%, and in South Africa about 1.9 times a small business corporation's 11.8% at R600,000.

South Africa, in one country

South Africa is the newest country on this site, and it is a useful test of the whole thesis because both sides of its comparison are domestic published rates rather than a global blended figure. Here are its headline numbers, each pinned to a primary source.

Figure	What it measures	Source and year
22.2% / 27.5% / 28.8%	The all-in effective income tax rate of a self-employed professional on R600,000 / R900,000 / R1,000,000 of profit, 2026/27 tax year, after the R17,820 primary rebate. There is no social-security wedge, so the all-in and income-tax-only rates are the same	SARS, rates of tax for individuals, 2027 tax year
11.8% to 17.9%	The effective rate on the same R600,000 to R1,000,000 of profit inside a qualifying small business corporation, on the graduated scale (nil up to R99,000, then 7%, 21%, 27%)	SARS, companies, trusts and small business corporations, 2026/27
27%	The standard company rate on taxable profit, held unchanged in the February 2026 budget	SARS / National Treasury, Budget 2026
About R866,000	The profit level at which the self-employed professional's all-in rate crosses the 27% standard company rate: below it the sole trader pays less, above it more	Computed from the SARS individual scale against the 27% rate
About 4% of corporate income tax receipts	The estimated annual revenue loss to multinational profit shifting, roughly R7 billion a year, with the largest multinationals shifting about 78% of their profits offshore	UNU-WIDER, The impact of tax havens on South African revenue
Rank 51 of 70	South Africa's standing on the Corporate Tax Haven Index 2024, with a haven score of 47 out of 100 and a global scale weight of 0.3%	Tax Justice Network, Corporate Tax Haven Index 2024

South Africa splits the usual story. Against the small business corporation scale the gap holds at every level, the sole trader pays between 1.6 and 1.9 times the rate. Against the standard 27% company rate it reverses below about R866,000 of profit, where the sole trader actually pays less. The largest multinationals, meanwhile, are estimated to pay below even 27% through the profit shifting the UNU-WIDER work measures. The same pattern as everywhere else, with the honest nuance that the personal system is genuinely light at the bottom.

Key facts

The global corporate tax loss is estimated at US$347.6 billion a year on 2021 data.
Total cross-border tax loss, corporate plus individual, is about US$492 billion a year.
Close to 40% of multinational profit is booked in tax havens.
The OECD's older, cautious range for the loss is US$100-240 billion a year, from 2014.
The Pillar Two global minimum rate is 15%, applying to groups above €750 million in revenue.
The average statutory corporate rate fell from 28.2% in 2000 to 21.1% by 2023.