• Home
  • News
  • Reports
  • About Us
    • About TaxWatch
    • Who we are
    • TaxWatch History
    • Privacy Policy
  • Contact Us

tech companies

US set to raise $8.5bn from four tech companies following global tax deal

7th October 2021 by Alex Dunnagan

New analysis from TaxWatch reveals how the US Government is set to be the big winner from increases in taxes on big tech arising from the global tax negotiations currently underway under the auspices of the G20/OECD.

In our latest research, we have analysed the gains that the US Government can expect from imposing a global minimum tax of 21% on Facebook, Google, Apple, and Microsoft. This would result in an extra $5.4bn in taxes from just these four companies, whereas we estimate the total additional tax these companies would pay in all other countries under Pillar One would be $2.5bn under the terms of the G7 agreement set out earlier this year.

The figure of 21% is used because the stated intention of the US Government is to impose a global minimum of 21% on companies headquartered in the US, regardless of the minimum level set through the OECD led process.

Furthermore, we find that the US Government would also benefit from the removal of tax incentives on royalties received by US parents from overseas operating companies (the Foreign Derived Intangible Income incentive, or FDII).

The removal of the FDII incentive is facilitated by the increase in global minimum taxation and therefore should be seen as a benefit of it.

We estimate that just four tech companies will see taxes increase in the US by $3bn per year as a result of the removal of the FDII.

Taken together, this means that the package of reforms will mean a yearly increase in US Tax of $8.4bn from just four companies, as against a benefit of $2.5bn shared between all other countries.

What is significant is that both the global minimum tax and the FDII only impact profits that arise from revenues made overseas, in countries like the United Kingdom where sales are made. The analysis therefore demonstrates that the G7 / OECD deal resolves the question of who gets to tax the offshore billions of tech companies decidedly in favour of the United States, with relatively little being distributed to the countries where these companies operate. This was not necessarily the outcome expected from the OECD led BEPS process, which had a stated goal of making tech companies pay a fair share in the countries where they operate.

The full report, ‘A Fair Distribution’, is available as a web page here, and as a PDF here. This research was featured in The Guardian as an op-ed by our Executive Director George Turner, available here.

A fair distribution?

7th October 2021 by Alex Dunnagan

7th October 2021

US Government set to raise 3x more than rest of the world combined in tax payments from four tech companies after Global Tax Deal

Raising Global Minimum Tax to 21% would see US Government raise an extra $2.5bn in tax from Apple alone

Summary and introduction

In June 2021 finance ministers of the G7 group of nations agreed on proposals to reform the way in which multinational enterprises are taxed. The agreement was the product of over 10 years of negotiations hosted by the OECD under the auspices of the Base Erosion and Profit Shifting (BEPS) programme.

When the deal was announced, the UK Government, which hosted this year’s G7, trumpeted it as a “seismic agreement on global tax reform that will mean the largest multinational tech giants will pay their fair share of tax in the countries in which they operate.”1

This paper presents an analysis of the deal, looking at the accounts of individual tech companies to demonstrate that the United States, where almost all major tech companies are headquartered, stands to gain billions of extra tax dollars from global tech giants as a result of the deal, whilst “the countries in which [tech companies] operate” stand to gain relatively little.

To do this we look at the gains that the US government can expect from imposing a global minimum tax of 21% on Facebook, Google, Apple and Microsoft. This would result in an extra $5.4bn in taxes from just these four companies, whereas we estimate the total additional tax these companies would pay in all other countries in the world under the terms of the G7 deal would be $2.5bn.

The figure of 21% is used because the stated intention of the US Government is to impose a global minimum of 21% on companies headquartered in the US, regardless of the minimum level set through the OECD led process.

Furthermore, we find that the US Government would further benefit from the removal of tax incentives on royalties received by US parents from overseas operating companies (the Foreign Derived Intangible Income incentive, or FDII). The introduction of the FDII in 2017 had already led to tech giants substantially restructuring their tax affairs. US companies have moved vast amounts of profits arising from sales outside of the US out of tax havens and back to the United States. This had expanded the US tax base at the expense of market jurisdictions in advance of the G7 deal. The removal of the FDII incentive is facilitated by the increase in global minimum taxation and therefore should be seen as a benefit of it.

We estimate that just four tech companies will see taxes increase in the US by $3bn per year as a result of the removal of the FDII.

Taken together, this means that the package of reforms will mean a yearly increase in US Tax of $8.4bn, as against a benefit of $2.5bn shared between all other countries arising from just four companies. What is significant is that both the global minimum tax and the FDII only impact profits that arise from revenues made overseas, in countries like the United Kingdom where sales are made. The analysis therefore demonstrates that the G7 / OECD deal resolves the question of who gets to tax the offshore billions of tech companies decidedly in favour of the United States.

This was not necessarily the outcome expected from the OECD led BEPS process.

Tax avoidance and the multinational

The problem that the BEPS programme set out to resolve was clear. Large multinational enterprises had accumulated trillions of dollars offshore by shifting profits out of “market jurisdictions” – the countries in which their customers or users were based, and into tax havens.

These tax haven entities were often an accounting fiction, generating billions in profits every year but with no staff, no offices and no discernable economic activity.

Although companies in all sectors have engaged in profit shifting, perhaps due to the particularly high profile of these companies, and the particularly aggressive stance they took towards the tax system, the behaviour of the tech industry has been a particular concern. In fact, the activities of tech giants warranted their own action point within the BEPS action plan.

The structure of the US tax system, where almost most global tech giants are headquartered, provided huge incentives for US based multinational corporations to move cash out of their international (non-US) markets and hoard it offshore, avoiding taxes on profits made in foreign jurisdictions and a US corporation tax charge that would accrue if they brought the cash home to the US.

A classic tax avoidance structure used by a US company would see intellectual property (i.e. trademarks, patents or software aka IP) developed in the United States sold to an offshore company in Bermuda, which would then licence the use of the IP to companies in Europe and other non-US markets. Companies in these international markets would pay the offshore company high fees for the use of the IP leading to an accumulation of wealth offshore and the elimination of profit elsewhere.

According to the US research institute ITEP, fortune 500 companies alone had accumulated $2.6 trillion in profits offshore by 2017, when the Trump administration enacted far-reaching tax reform.2

The tax proxy wars

A key question facing tax policy makers was therefore whether or not the tax haven profits of US based multinationals and tech companies in particular should have been properly accounted for and taxed in the US, where the IP was developed, or in the “market jurisdictions” where services were sold to (predominantly European and other developed countries). All governments appeared to agree at least that shell companies with no staff and no physical operations should not have been considered to have “earned” any profits themselves.

The OECD’s report which kicked off the Base Erosion and Profit Shifting programme that sought to reform the global tax system raised the question of whether more profit needed to be allocated to countries where sales are made. Under the heading “Jurisdiction to tax” the report contained the following passage:

“In an era where non-resident taxpayers can derive substantial profits from transactions with customers located in another country, questions are being raised as to whether the current rules ensure a fair allocation of taxing rights on business profits, especially where the profits from such transactions go untaxed anywhere.”3

However, perhaps unsurprisingly given the amounts of money at stake, neither the US nor Europe saw eye to eye on how profits accumulated offshore should be divided.

In the US, the IRS has attacked the way in which costs are divided between US entities and offshore entities. It has argued that the subsidiaries of US corporations based offshore, which typically have no staff and no costs, should contribute a greater amount to the costs incurred by their US based parents. In effect, bringing profits onshore to compensate the US entities for their role in developing IP.

In 2020, a US federal appeals court resolved a dispute that had been going on for years between the IRS and Altera, a US based chip maker. The court upheld an IRS regulation from 2003 that offshore companies owned by US multinationals should contribute to the cost of share options granted to employees in the US. As a result of this case, Google and Facebook alone set aside $2bn to comply with the ruling.4

In Europe, the European Commission launched a number of cases against governments, claiming that the favourable tax treatment which allowed multinationals to move money to low-tax jurisdictions breached state aid rules, and compelled European tax havens to levy taxes on profits either accumulated or passing through their jurisdiction.

Although from a European perspective, this has been seen as an albeit imperfect mechanism to claw back some tax which should have been due in Europe, US politicians have viewed EC action against tax avoidance as a raid on US profits.

When Apple was fined €13bn Euros by the European Commission for unpaid taxes arising from a scheme which saw profits transferred out of profitable European markets to a headquarters company that “only existed on paper”,5 Charles Schumer, one of the highest-ranking Democratic senators, said:

“This is a cheap money grab by the European Commission, targeting US businesses and the US tax base.

By forcing their member states to retroactively impose taxes on US companies, the EU is unfairly undermining our ability to compete economically in Europe while grabbing tax revenues that should go toward investment here in the United States”

This fundamental disagreement on who should have the right to tax held up progress on the digital economy workstream of the BEPS programme for years and in the interim, governments sought to enact unilateral reforms.

A number of countries sought to impose digital services taxes on the revenues of some tech companies. The UK’s diverted profits tax sought to impose a higher tax rate on profits diverted using artificial structures.

Eventually an agreement was reached at the 2021 G7 Summit in Cornwall to co-ordinate tax policy. As we will explore later in this paper, this essentially resolves the question of who gets to tax big tech profits in favour of the US.

Tax Cuts and Jobs Act 2017

The most significant unilateral policy change enacted before the G7 agreement was introduced by the US government. The Trump administration’s Tax Cuts and Jobs Act of 2017 (TCJA) fundamentally altered the incentives in place for US corporations to keep their cash offshore and paved the way for the G7 agreement.

These reforms came in several parts. A one off tax on accumulated offshore holdings, a new tax on profits allocated to intangible assets in tax havens (global intangible low-taxed income, or GILTI) and a tax incentive on revenues from overseas accruing to intangible assets in the United States (foreign-derived intangible income, or FDII). By targeting foreign derived income, the United States was laying claim to the offshore profits accumulated by its multinationals.

Profits from outside the US accruing to intellectual property based in tax havens would be taxed for the first time under the GILTI regime, whereas, if that IP was brought back to the US, the FDII provided a discount on the tax paid on the profits accruing from revenues earned by that IP. The GILTI and the FDII are both set at a rate of 13.125% (assuming a 21% corporate rate) ensuring that there was no longer an incentive to keep IP offshore. Being set at the same rate means they also work together. If the FDII incentive was removed without also increasing the GILTI rate, then US corporations would simply move their IP offshore, as they had under the previous incentive structure.

As will be explored in more detail later in this paper, in the years following the introduction of the TCJA, US companies responded to the new incentive structure by repatriating intellectual property from offshore jurisdictions to the US.

This clearly shows up in macroeconomic data. A study published on behalf of the Irish Finance Ministry showed royalty payments from Irish companies to the US jumping from €8bn a year on average between 2014-2019 to €52bn in 2020.6 Ireland is a key conduit used by US tech companies to move profits out of Europe.

The G7 deal

The current proposals endorsed by the G7 come in two parts. A new Global Minimum Tax, Pillar Two acts as an expanded GILTI charge, with jurisdictions that host the headquarters of multinational corporations placing a charge on the profits they accrue in tax havens.

The G7 has agreed that this should be set at a minimum 15%, however, as we have seen with the GILTI, countries can go it alone and tax the offshore profits of their corporations without international agreement. The US has said it will adopt a 21% rate and has encouraged others to do the same.

If raising taxes on profits accumulated offshore encourages companies to shift profits onshore, then jurisdictions also have more freedom to tax onshore profits.

As part of the Made in America tax plan President Biden has said that the US will scrap the FDII incentive which had already been declared a “harmful tax practice” by the OECD,7 replacing it with a new, as yet unspecified, R&D subsidy.

The removal of the FDII will mean large increases in US taxation on royalties paid from market economies to the US.

On the other side of the coin, the limited actions by market jurisdictions to claw back some income from digital services companies through digital services taxes are replaced by a limited redistribution of income from large multinational enterprises (MNEs) via the second part of the G7 deal, the so called Pillar I proposals.

Under Pillar One, the world’s largest companies see a portion of their global profits re-allocated to countries where they have a market which are then charged at the local corporate tax rate. Pillar One is limited to roughly the 100 largest corporations in the world.8

Under analysis undertaken by the OECD around the time of the publication of their blueprint for reform, the gains of the new system are weighted massively in favour of Pillar Two, with total gains from both pillars comprising between 2.3%-4% of global corporate income tax revenues and Pillar One accounting for between 0.2-0.5%.9

This is confirmed by our analysis in this paper, which shows that the real effect on four tech companies from the US plan to raise the GILTI rate to 21% and remove FDII is more than three times the additional tax those companies will see from Pillar One.

For companies subject to digital services taxes, the replacement of DSTs by Pillar One represents an effective tax cut.10 11

The impact on US based multinationals of the TJCA the G7 deal and the Made in America tax plan

Now that several years have passed since the Tax Cuts and Jobs Act, the impact of the new incentive structure on US based corporations can be seen very clearly in corporate accounts. In this part of the paper, we look at several companies and see how they have responded to the changing incentive structures they face after TCJA. The analysis shows that although some companies have chosen to repatriate the majority or all of their IP to the US, causing large and dramatic changes in their profitability overseas, some appear to have hedged, keeping some intellectual property offshore.

These companies will be impacted differently by the removal of the FDII and the increase in Global Minimum Taxation.

Shifting profits onshore

Google

In 2019 Google announced that it would no longer be licencing its IP from a Bermuda registered company, Google Ireland Holdings Unlimited Company, and instead would move its IP back to the United States.

The impact of this change on Google accounts was dramatic and immediate. The amount of profit Google declared in the United States (their US tax base) moved from $16.4bn to $37.6bn whereas the amount of profit Google declared outside of the US more than halved, from $23.2bn to $10.5bn.12

Figure 1: Alphabet (Google) 2020 10K, showing US and non-US profits.

Google explains this saying “as of December 31 2019, we have simplified our corporate legal entity structure and now license intellectual property from the U.S. that was previously licensed from Bermuda resulting in an increase in the portion of our income earned in the U.S.”

As a result of this shifting of profit to the United States, the amount of taxes paid to the United States Federal Government doubled, whereas the taxes paid to non-US governments saw a significant decline.

Figure 2: Alphabet (Google) 2020 10K, showing effective tax rates

The tradeoff that Google has made between a discount on its tax bill through the use of offshore companies, with the discount on the FDII shows up in Google’s tax reconciliation. In 2019, Google saw a 5% decrease in its effective tax rate through declaring income outside the US. In 2020, this almost disappeared, and instead the company saw a 3% reduction in its effective tax rate as a result of the FDII. This 3% is worth over $1.4bn.

In comparison, we calculate that under Pillar One, Google is likely to face an additional tax charge of just $307m.

Facebook

Turning our attention to Facebook, we see the same story again, with US profits increasing almost five fold from 2019 to 2020, while non-US profits more than halve.

Figure 3: Facebook 2020 10K, showing US and non-US profits.

In 2020 Facebook saw a 1.9% reduction in its effective tax rate resulting from “foreign-derived intangible income”, a tax break worth some $630m, whilst at the same time the effect of “non-US operations” fell from 5.8% to 2.4%.

Figure 4: Facebook 2020 10K, showing effective tax rates

It is also remarkable that the dramatic changes in where these companies locate their profits, is almost tax neutral, with Facebook’s effective tax rate being 12.8% in 2018, before their IP was repatriated, falling to 12.2% in 2020. The larger payment in 2019 can largely be attributed to a one-off increase in costs associated with “share based compensation” likely to be the impact of the Altera ruling.

Nike

For completeness, we can see that the impacts of the TCJA extend beyond the tech industry by looking at sports brand Nike, where US profits increased by almost 400%, from $593m in 2019 to almost $3bn in 2020. As this is happening, Nike’s previously profitable non-US business goes from a profit of $4.2bn in 2019 to a loss $67m in 2020.13

Figure 5: Nike 2020 10K, showing US and non-US profits

While US profits increase, we see an 8.1% reduction in the effective income tax rate as a result of “Foreign-derived intangible income benefit related to the Tax Act”. Nike state in their accounts that “This benefit became available to the Company as a result of a restructuring of its intellectual property interests”. What this means is that as a result of Nike bringing its IP back to the US, rather than sitting offshore. This 8.1% reduction is worth some $234m to Nike.14

Figure 6: Nike 2020 10K, showing effective tax rates

Remaining offshore

Although it is clear that there has been a very significant shift of profits from offshore back to the US, it appears that some US companies are hedging their bets and either only repatriating a portion of their IP, or continuing to keep all of their IP overseas.

Microsoft

Microsoft’s latest 10-K form contains the following statement:

“In the fourth quarter of fiscal year 2019, in response to the TCJA and recently issued regulations, we transferred certain intangible properties held by our foreign subsidiaries to the U.S. and Ireland. The transfers of intangible properties resulted in a $2.6 billion net income tax benefit recorded in the fourth quarter of fiscal year 2019, as the value of future tax deductions exceeded the current tax liability from foreign jurisdictions and U.S. GILTI tax.”

The reference to the one off tax benefit show that Microsoft expect that moving their IP to Ireland and the US will result in a lower tax bill in the future.

As a result of this, whereas Micrsoft’s non-US profits made up 68% of their total profits in 2018, by 2020, this had fallen to 55%.

The Microsoft tax reconciliation shows that in 2021, the company still had a substantial tax benefit arising from earnings taxed offshore of 2.7%, whilst also claiming a 1.3% deduction on their effective tax rate due to the FDII.

Figure 7: Microsoft 2020 10K, showing effective tax rates

This reconciliation table suggests that if the global minimum rate is raised to 21%, then Microsoft will have to pay an additional $1.9bn in taxation. The removal of the FDII would mean an additional $900m to pay.

Apple

Apple is another company that has seen some shift in profits to the US, but appears to be continuing to keep a substantial amount of IP offshore. In 2018, the company earned $48bn in pre-tax profit overseas (66% of total profit), this fell to $38.1bn in 2020 (57% of total profit).

The current tax rate on Apple’s non-US profits was just 8.3% in 2020.

The company stated that as a result of declaring earnings overseas, it saw a reduction in its tax bill of $2.5bn against the standard US corporation tax rate. This fell from $5.6bn in 2018 (which included a period US corporation tax rates were higher). The company does not appear to claim anything under the FDII.

Figure 8: Apple 2020 10K, showing effective tax rates

Where now?

For years multinational companies accumulated trillions of dollars in tax havens at minimal or zero tax rates. Although multinationals from all jurisdictions and in all sectors have been to some extent guilty of these practices, it was always the digital disruptors, almost all of which are headquartered in the United States, that attracted the most attention. This is recognised in the fact that the current proposals put forward by the OECD have been framed as dealing with the challenges of the digital economy, when in reality the proposals will impact a wide range of multinationals regardless of the sector they are in.

In 2017, the US government laid claim to the offshore dollars of US headquartered multinationals, directly taxing the profits they had accumulated overseas and changing their policy framework to encourage them to bring their IP onshore. This has already raised billions from the offshore cash piles of US based multinationals and seen tens of billions of dollars of profit shifted to the US from overseas, substantially increasing the US tax base.

The G7 deal reached on global tax reform largely accepts this position. Countries that host multinationals get to tax the tax haven profits of companies in their jurisdiction, with a small amount of profit from larger companies redistributed to “market” jurisdictions.

For global tech giants, almost all of which are headquartered in the US, this will mean billions of extra tax dollars for the US government – all of which arises from profits on the sales of goods and services overseas.

We calculate that an increase in the global tax rate to 21% would mean that Microsoft would pay an additional $1.9bn in taxation to the US Federal Government, Apple an additional $2.5bn, Google an extra $144m and Facebook an additional $796m based on their latest annual accounts.

The removal of the FDII, which would also see the profits arising from royalties paid from overseas taxed at the current US tax rate of 21%, would mean an additional $1.4bn in tax paid by Google to the US Federal Government, an additional $630m in taxes from Facebook and $924m from Microsoft.

In total, this means that the US will see an increase in tax on profits derived from overseas sales of $8.4bn from just these four companies. By comparison, we would expect all other jurisdictions in the world receive an additional $2.7bn combined from these companies under the Pillar One agreement.

Given the history of the debate around tax avoidance over the last 10 years, which has focused on tech companies in particular removing profits from market jurisdictions, it is remarkable that the response from Finance Ministries in other developed economies to this obvious inequality has been muted. The US appears to have won the argument that the profits arising from from the customers of US companies overseas should largely be taxed in the US.

It is the case that G7 countries will also be able to raise more from multinationals based in their home countries. For example, in the UK, the new proposals would prevent the kinds of controversial profit shifting arrangements that have been used by British companies like Vodafone to keep non-UK profits offshore, and this may well be why developed nations have accepted the Biden proposals.

However, they have done so at the expense of other market jurisdictions. Developing countries, which host very few multinationals, will see relatively little benefit from the deal. As Nigeria’s ambassador to the OECD has said “What I understand, with the . . . rules as currently being developed, is that developing countries may get next to nothing.”15

Data for Facebook, Google, Apple, and Microsoft based on 2020/21 10-K filings

Notes

  1. ‘Income’ refers to what is in US 10-K filings as ‘income before provision for income taxes’. In UK company accounts this often referred to as something along the lines of ‘profit on ordinary activities before taxation’. This is simply taxable profit.

This report is also available as a PDF here. This research was featured in The Guardian as an op-ed by our Executive Director George Turner, available here.

Photo by NASA on Unsplash

1G7 Finance Ministers Agree Historic Global Tax Agreement, G7, 05 June 2021, https://www.g7uk.org/g7-finance-ministers-agree-historic-global-tax-agreement/

2Fortune 500 Companies Hold a Record $2.6 Trillion Offshore, ITEP, 28 March 2017, https://itep.org/fortune-500-companies-hold-a-record-26-trillion-offshore/

3OECD, Key tax principles and opportunities for base erosion and profit shifting, February 2013, p.36

4An obscure court ruling could play havoc with tech companies’ earnings, Marketwatch, 18 July 2020, https://www.marketwatch.com/story/an-obscure-court-ruling-could-play-havoc-with-tech-companies-earnings-2020-07-16

5Comment: ECJ decision should not let Apple off the hook, TaxWatch, 15 July 2020, http://13.40.187.124/apple_ecj_ruling/

6Seamus Coffey, The changing nature of outbound royalties from Ireland and their impact on the taxation of the profits of US multinationals, Irish Department of Finance, https://www.gov.ie/en/publication/fbe28-the-changing-nature-of-outbound-royalties-from-ireland-and-their-impact-on-the-taxation-of-the-profits-of-us-multinationals-may-2021/

7Harmful Tax Practices, OECD, August 2021, https://www.oecd.org/tax/beps/harmful-tax-practices-peer-review-results-on-preferential-regimes.pdf

8The OECD state that “In-scope companies are the multinational enterprises (MNEs) with global turnover above 20 billion euros and profitability above 10%”. A full explanation of which companies Pillar One applies to is available in the Statement on a Two-Pillar Solution to Address the Tax Challenges Arising From the Digitalisation of the Economy, OECD, 01 July 2021, https://www.oecd.org/tax/beps/statement-on-a-two-pillar-solution-to-address-the-tax-challenges-arising-from-the-digitalisation-of-the-economy-july-2021.pdf

9OECD, Tax Challenges Arising from Digitalisation – Economic Impact Assessment, (Paris, 2020) see table 1.1 available from: https://www.oecd-ilibrary.org/docserver/0e3cc2d4-en.pdf?expires=1633339704&id=id&accname=guest&checksum=BB2899422231BC934DA03A1778EE8C89

10This is also likely the case for other countries with their own respective digital services taxes. We have not however carried out an analysis on the effect for countries other than the UK.

11Will the OECD’s global tax deal raise more from tech companies in the UK?, TaxWatch, 27 September 2021, http://13.40.187.124/oecd_pillar1_analysis/

12Alphabet Inc. 10-K 2020, SEC.gov, https://www.sec.gov/ix?doc=/Archives/edgar/data/1652044/000165204421000010/goog-20201231.htm

13Nike 10-K 2020, SEC.gov, https://s1.q4cdn.com/806093406/files/doc_financials/2020/ar/NKE-FY20-10K.pdf

14 The increase of 43.3% to the effective tax rate in 2018 is a result of the one-time mandatory transition tax on deemed repatriation of undistributed foreign earnings, part of the Tax Cuts and Jobs Act 2017.

15Biden’s global tax plan could leave developing nations ‘next to nothing’, Financial Times, 10 May 2021, https://www.ft.com/content/9f8304c5-5aad-4064-9218-54070981fb4d

G7 tax deal represents a tax cut for big tech in the UK – new analysis

27th September 2021 by George Turner

In June 2021 the G7 announced that it had reached an agreement on proposals to change the way that the largest multinational companies are taxed. This included a mechanism to redistribute profits to countries where sales are made (such as the UK), often referred to as “market jurisdictions” . The final agreement is due to be signed off in October 2021 at the OECD.

A new analysis from TaxWatch has demonstrated that tech giants will almost certainly end up paying less tax under the proposals put forward by the G7 than they are currently liable for under the digital services tax.

Scrapping digital services taxes was a key demand of the United States in the negotiations for a new global tax deal, which is expected to be finalised next month at the OECD.

TaxWatch has undertaken an analysis of the formulae used to calculate the DST and the potential future tax liabilities under the G7 proposals, and concludes that with a digital services tax rate of 2%, the current UK DST rate, and a corporation tax rate of 19%, the current UK corporation tax rate, a company would have to have a profit margin of 62% in order to pay more in tax under the G7 proposals than under the DST. Historically tech companies have never achieved these profit margins.

If the UK tax rate rises to 25%, as it is currently scheduled to do, then the profit margin required before a company ends up paying more than under the DST falls to 48%, a margin rarely seen historically.

The paper also analyses the proposals that have recently been put forward by the G24 group of countries, which advocate for more profit to be reallocated to countries where revenues are made (such as the UK) under the OECD formula. If these proposals were adopted, then tech companies would be likely to end up paying more in tax in market jurisdictions, with the result that profit margin required before a company pays more dropping to 44% assuming a 19% corporate tax rate or 36% assuming a 25% corporate tax rate.

The full report can be found as a web page here and as a PDF here.

 

Will the OECD’s global tax deal raise more from tech companies in the UK?

27th September 2021 by George Turner

27th September 2021

Analysis shows that tech giants would pay less in taxation in market jurisdictions under G7 global tax deal than under current digital services taxes.

Proposals put forward by the G24 group of nations ahead of next month’s crucial meeting of the OECD would lead to tech companies paying more.

Making tech companies pay their fair share?

In June 2021 the G7 reached an agreement on a new system of taxation for the world’s largest companies.

The deal unlocked ongoing negotiations at the OECD and was billed as a way of confronting the “tax challenges that arise from the global digital economy” and a way of making sure multinationals pay “their fair share in the countries they do business”.1 Final sign off on a deal at the OECD is expected in October.

The agreement was the product of years of negotiations at the OECD following public outcry at the low levels of taxation being paid by large tech companies – some of which had seen their effective tax rates fall to single digits.

Although tech companies have insisted that the low tax rates they have achieved have been a result of government incentives, the reality is that the industry was a heavy user of tax avoidance schemes.

Talks aimed at addressing the problem had been taking place at the OECD for many years. In the interim, many countries around the world grew increasingly frustrated with a lack of progress and started to implement their own unilateral measures to tax digital companies and “make them pay their fair share”.

These so called “digital services taxes” (DSTs) were a controversial measure that sparked trade disputes between the United States and European countries that adopted them.

One particular aspect that aggravated tech companies was that DSTs operate as a tax on the turnover in jurisdictions where services were sold. This meant that they did not come under tax treaties and therefore could not be credited against taxes paid in a company’s home jurisdiction in the say way that corporate taxes can be offset against liabilities at home. They were a tax that came on top of any corporate tax liability.

From the point of view of tech companies, a new global system of allocating taxable profit to jurisdictions to replace a patchwork of DSTs would be easier for tech companies to administer, allow them to manage their tax liabilities on a global level, and have the added benefit of wiping the slate clean following years of tax avoidance.

For this reason, tech companies have long argued for tax reform and were supportive of the G7 deal when it was announced. A spokesperson for Google told Reuters: “We strongly support the work being done to update international tax rules. We hope countries continue to work together to ensure a balanced and durable agreement will be finalised soon.”2

Nick Clegg, Global Vice President of Public Affairs for Facebook Tweeted: “Facebook has long called for reform of the global tax rules and we welcome the important progress made at the G7. Today’s agreement is a significant first step towards certainty for businesses and strengthening public confidence in the global tax system. We want the international tax reform process to succeed and recognize this could mean Facebook paying more tax, and in different places.”3

But will companies like Facebook end up paying more tax in market jurisdictions like the UK?

In this paper, we demonstrate that for companies subject to the DST rate of 2%, the DST rate in the UK, it is a near certainty that they will end up paying less in tax in “market jurisdictions” such as the UK under the G7 proposals than under existing digital services taxes. If the UK supported the position of the G24 group of nations, more profit would be allocated to the UK, with the result that profit margin required before a company pays out more tax under the OECD formula than the current DST being closer to the profit margins observed at large tech companies.

We come to this conclusion based on a mathematical analysis of the formulas used to calculate the DST and the new tax base under the OECD proposals.

The effective tax cut that the G7 deal represents is even larger for countries with higher DSTs like France.

The G7 deal

The agreement reached at the G7 was based on a blueprint put together by the OECD.

The new proposals come in two parts, Pillar 1 and Pillar 2. Pillar 1 is a mechanism to re-distribute profits made by some of the world’s largest multinational companies to countries where they make revenues (such as the UK). Pillar 2 is a mechanism to stop companies from moving their profits into low tax jurisdictions. In this paper we focus on the Pillar 1 proposals. Under Pillar 2 all of the benefit goes to the country where the multinational is headquartered, and as such it is not relevant to the distribution of profit to market jurisdictions.

The OECD blueprint outlines a three step process to find the “allocation amount” under Pillar 1, which is the amount of profit that will be available to market jurisdictions to tax.

The G7 deal largely adopted the blueprint, and agreed the various percentages required to determine the allocation amount, which had been left as an open question.

Following the G7 agreement, we can now see what the impact will be on the tax base of large corporations using the blueprint process and the allocation percentages set by the G7.4

Step 1 – Allowance for “routine” profit

The first step in the OECD process is to isolate what the OECD call “routine profits”. This applies a simple percentage to the profit before tax of a multinational as an allowance that will excluded from any reallocation. The G7 agreement set this as 10%.

Step 2 – Global Reallocation

The next step in the process is to create a pool of profits that will be available for reallocation to market jurisdictions. This is a simple percentage of the remainder of a company’s profit before tax after step one. The G7 set this reallocation percentage at 20%.

Step 3 – Redistribution

The reallocation pool arrived at after step two is then carved up amongst market jurisdictions who can then apply tax the profits allocated to them.

The G7 agreement did not give any details on how this would work, however, the OECD blueprint states that the key will be based on the revenues a Multinational Enterprise(MNE) makes in any in-scope country.

For the purposes of this paper, we have assumed that the allocation is based on a simple proportion of total revenue made in any country.

The three steps can be expressed as a formula as follows

Local tax base = Global revenues x (profit margin – 10%) x 20% x local market share

This local tax base can then be multiplied by the local corporation tax rate to find the amount of additional taxation raised by the pillar 1 process.

Digital Services Taxes

As part of the G7 agreement reached on the new global tax deal, countries should dismantle domestic digital services taxes. The rational for this is that the DST was a stop gap to ensure that some tax was paid in market jurisdictions until a global agreement of the kind proposed by the OECD was agreed.

Digital services taxes are crude and simple taxes that place a flat rate of taxation of the revenues of a company derived from the provision of certain digital services in the country.

In the UK, the DST is a tax of 2% on the revenues derived from search engines, social networks and online marketplaces with a tax free allowance of £25,000,000

The UK DST can also therefore be expressed as follows:

DST revenues = (Global revenues x local market share) -£25m x 2%

Comparing DST revenues with Pillar 1 revenues

For companies subject to the DST, i.e. those that derive their income from in-scope activities, we can easily compare the amount of tax raised by the DST against the new pillar 1 formula. We have already done this exercise using the real corporate accounts of several global tech giants and our estimates of revenues in the UK here.

However, we can also determine under what conditions the DST will equal tax revenues raised under pillar 1 because by putting the two equations together we get:

(Global revenues x local market share) -£25m x 2% = Global revenues x (profit margin – 10%) x 20% x local market share x local tax rate

We can see here that for any given company, their global revenues and local market share will be the same on each side of the equation, meaning that the only two variables that will change whether or not the DST raises more than Pillar 1 for any company subject to the DST will be determined by the profit margin of the company and the local tax rate.

This equation can therefore be rearranged to isolate profit margin, allowing us to answer the question, what profit margin would a company need to achieve in order to end up paying more under the pillar 1 than the DST at a given local corporate tax rate?

Profit margin = ( (Global revenues x local market share) -£25m) x 2%/ (Global revenues x 20% x local market share x local tax rate) + 0.1

If we take a company that has global revenues of £10bn, and a market share in the UK of 10% of global sales, and a UK corporate tax rate of 19% we get the following:

Profit margin = ( (£10bn x 0.1) -£25m) x 0.02 / (£10bn x 0.2 x 0.1 x 0.19) +0.1

Which gives us

Profit margin = 62%

So, under the current UK corporation tax rate of 19%, a corporation would have to achieve a pre-tax margin of 62% before it started paying more under the pillar one agreement than under the digital services tax.

The UK corporation tax is currently scheduled to increase to 25%, so we can re-run the exercise at that tax rate with the same notional £10bn company.

Profit margin = [1]£10bn x 0.1) -£25m x 0.02 / (£10bn x 0.2 x 0.1 x 0.25) +0.1

This gives us

Profit margin = 49%

The equation is highly sensitive to increases in the DST rate. If we look at the same company, and apply the French DST rate of 3% and the French Corporate tax rate of 28%, then the profit margin required is 68%.

Real profit margins of tech giants

Of course in theory, a company could end up paying more under the pillar one process than under the digital services tax if it achieved very high profit levels.

However, a quick glance at the real historical profit margins seen by tech giants demonstrates that the kinds of profit levels required before a company ends up paying more under pillar one are rare. As such, it is a near mathematical certainty that companies will end up benefiting from a deal that removes digital services taxes and replaces them with a Pillar 1 allocation under the deal announced at the G7. This is a highly significant finding, given that the purpose of the G7 deal was to come to a more equitable taxation of the digital economy.

The following chart gives us the historic pre-tax margin of Google, which has never risen above 37%.

The following chart is the pre-tax profit margin of Facebook. Facebook’s profit margins have been as high as 53% in the past, but more recently have hovered around the 40% mark.

Source: macrotrends.net

The G24 proposals

The fact that the G7 proposals involve very little distribution of profits away from the home jurisdiction of multinational companies (most of which are headquartered in G7 countries) has not gone unnoticed.

The G-24 group of developing nations has put forward alternative proposals to those adopted by the G7.5

The most signifiant change to pillar 1 is to the allocation percentage, the percentage of a multinational’s profits (above the 10% hurdle) that would be available for reallocation to market jurisdictions. Under the G24 proposals this should be at least 30%.

If we rerun our formula above, substituting an allocation percentage of 20% to 30%, we find that assuming a 2% DST and a 19% corporate tax rate, a company would only have to have a profit margin of 44% before Pillar 1 started to raise more revenues than the DST. With a corporate tax rate of 25% that falls to 36%.

It is therefore in the best interests of the UK to support the proposals put forward by the G24 group and seek a higher allocation percentage.

This report is also available as a PDF here.

This article was amended on 27th September to correct a mistake in the calculation of the impact of the G24 proposals.

1G7 Finance Ministers Agree Historic Global Tax Agreement, HM Treasury, 05 June 2021, https://www.gov.uk/government/news/g7-finance-ministers-agree-historic-global-tax-agreement

2Reaction to the G7 minimum tax agreement, Reuters, 05 June 2021, https://www.reuters.com/business/reaction-g7-minimum-tax-agreement-2021-06-05/

3Nick Clegg, Twitter, 1437hrs 05 June 2021, – https://twitter.com/nickclegg/status/1401171338358714374?lang=en

4HM Treasury, G7 Finance Ministers and Central Bank Governors Communique, 05 June 2021, https://www.gov.uk/government/publications/g7-finance-ministers-meeting-june-2021-communique/g7-finance-ministers-and-central-bank-governors-communique

5Comments of the G-24 on the Statement on a Two-Pillar Solution, G24, 19 September 2021, https://www.g24.org/wp-content/uploads/2021/09/Comments-of-the-G24-on-the-IF-July-Statement.pdf

References[+]

References
↑1 £10bn x 0.1) -£25m x 0.02

Tech giants undermine African countries by failing to collect VAT

12th July 2021 by Alex Dunnagan

African countries are missing out on significant tax revenues because multinational tech giants are failing to collect and pay VAT on services, according to new research.

A new report published by TaxWatch finds that Microsoft, Google and Facebook are not collecting VAT on sales made to customers in most African countries, even in some countries where they have a local office.

The companies say that they will only collect VAT and sales taxes in countries that have levied specific taxes on digital products, claiming that it is up to the customer to pay any taxes due.

However, analysis from TaxWatch shows that these companies should be required to register for VAT under existing VAT rules in many African countries if they make significant sales.

VAT, a tax on the final consumption of goods and services, is an important source of revenue for tax authorities. An analysis from the OECD shows that on average VAT accounts for 30% of tax revenues in Africa, as opposed to taxes on individuals, which make up just 15.4%. Taxes on corporations account for 18.6% of tax revenues on average.

Since 2013 a number of African countries have changed their laws so VAT is collected on products, such as advertising, sold to customers.

The report does not make any estimate of the amount of money that could be raised if digital companies registered for VAT, as this would require detailed knowledge of the amount of sales made in each country. However, evidence from countries that have enacted a digital sales tax shows that the gains could be significant.

In 2014 South Africa became one of the first countries in the world to introduce specific legislation on VAT and digital services. The South African Revenue Service has said that it has collected an additional R600m a year since the introduction of the rules.

TaxWatch Executive Director, George Turner, said: “Big tech has for years traded on the myth that because they invoice their clients from offshore, they somehow have no obligation to pay any taxes. Our research demonstrates that this simply is not the case. If these companies have made significant sales in any country that levies VAT, then they should register to pay VAT locally.

“Google, Facebook and Microsoft wouldn’t get away with such an approach in Europe or North America anymore.

“I would encourage tax authorities across the continent to cast a close eye on the activities of these companies in their jurisdiction, and pursue them for any VAT payments that should have been made. This is likely to be a far more significant issue than any corporate tax avoidance engaged in by tech companies, as VAT makes up a far higher proportion of tax revenues than corporate profits.”

When questioned on the findings of this report, a Microsoft spokesperson stated: “Microsoft is fully compliant with all local laws and regulations in every country in which we operate. We serve customers in countries all over the world and our tax structure reflects that global footprint.”

Google did not respond to our requests for comment.

The report is available on our website here and in PDF here.

This research was featured in Law360 and PQ Magazine.

Eight tech companies in the UK avoided an estimated £1.5bn in 2019 – New Research

2nd June 2021 by Alex Dunnagan
  • £45.4bn in revenues
  • £9.6bn in profits
  • £296m in tax paid
  • £1.5bn in tax avoided

Eight large tech companies in the UK made an estimated £9.6bn in profit from sales to UK customers in 2019, a new analysis by TaxWatch shows.

​But by moving money out of the UK, these companies ended up declaring a fraction of these profits in the accounts of their UK subsidiaries, radically reducing their tax liabilities.

​Amazon, eBay, Adobe, Google, Cisco, Facebook, Microsoft, and Apple faced UK corporation tax liabilities of £297 million in 2019.

That puts the total amount of tax avoided by the companies in the UK at an estimated £1.5bn in 2019, the latest year where figures exist.

Large US-based technology companies have tens of millions of UK users and make billions in sales to UK customers. The UK is unarguably a significant source of corporate profits for these companies. But a glance at the accounts of their UK-based subsidiaries shows that little of this profit ends up in the UK.

As Finance Ministers seek to negotiate a new international settlement on how large multinational companies should be taxed, understanding the scale of tax avoidance by global digital giants is key to evaluating the outcomes of any deal.

​The latest figures are an update on TaxWatch’s 2018 study, Still Crazy After All These Years, which looked at the activities of Google, Cisco, Facebook, Microsoft, and Apple, five of the largest technology companies in the world, over the period 2013-2017. Last year we updated these figures taking 2018 into account. This year three companies were added to our analysis, eBay, Adobe, and Amazon.

The full report is available here, and as a PDF here.

This report featured on ITV News, Channel 4 News, and in several newspapers.

Eight tech companies in the UK avoided an estimated £1.5bn in 2019

2nd June 2021 by Alex Dunnagan

2nd June 2021

Introduction and Summary

The UK market is an extremely important market for US based technology companies. These well known companies realise substantial sales in the UK and as such are some of the most significant economic operators in our country.

For example, In 2020 Amazon made £19.4bn in sales to UK customers, a 50% rise on the previous year and more than 20 times the largest physical bookseller in the UK, Waterstones.1

It sold more goods to UK customers than Morrisons, the fourth largest supermarket chain in the country. The advertising revenues of Google and Facebook now surpass those made by ITV, or major publishers such as News UK, the owners of the Times Newspapers.

Given how profitable the digital economy has proved to be, the vast sales made by the global digital giants should result in these companies being some of the largest corporate tax payers in the UK. However, the accounts of the UK businesses of these major international companies show that very little profit is ever declared in the UK and as such, little corporation tax is paid on the profits these companies make from UK customers. These profits have instead re-appeared in tax havens where the company pays little or no tax at all.

TaxWatch has been tracking the tax payments, revenues and profits of the tech giants in the UK since we began in 2018. Each year we have sought to estimate the amount of revenues generated by UK customers of major tech companies, how much profit should be attributed to those sales, and how much tax has been avoided by these companies by moving their profits offshore.

This year our analysis takes on a special significance, as Finance Ministers seek to negotiate a new international settlement on how large multinational companies should be taxed. Understanding the scale of tax avoidance by global digital giants is key to evaluating the outcomes of any deal.

In 2018, we started by looking at the top five tech companies, Apple, Alphabet, Facebook, Microsoft and Cisco Systems. This year we have added Amazon, eBay and Adobe. The numbers we produce below are for the 2019 financial year, which is before the impact of the Covid-19 pandemic. The reason for this is that the UK accounts for the companies in our study are only available to 2019.

Our analysis has a simple methodology. We look at the global profit margin of a company, the percentage pre-tax profit the company make on every dollar of sales. This is the average profit the company makes across all operations in all jurisdictions. We then apply that average profit margin to the amount of sales made in the UK to calculate the amount of profit generated in the UK. The UK sales are either taken directly from the global accounts of the company where they are reported, or estimated using a variety of different methodologies. We then apply the UK headline tax rate to the amount of profit estimated for the UK, which gives us an estimate of the tax liability that each company should have in the UK. We then compare this to the actual taxes paid in the UK by their local subsidiaries.

It should be stressed that the figures we produce are estimates of the profits made by these companies in the UK. They are not definitive, however, in the absence of any accurate reporting on profits made from UK customers by major multinational companies, we believe our methodology gives a good understanding of the amount of profit made by each company in the UK.

We gave each company the opportunity to respond to our figures. Most declined. Amazon strongly disputed our figures. Facebook told us that they had long supported efforts to reform the international tax system and that currently the tax system does not allocate profits to where customers are located.

There are good economic reasons for taking this approach we have taken to attributing profit to the UK. As companies move into new markets, sales volumes increase and the marginal cost of production falls. As a result, the profit on each unit of sales increase. As most companies start in their home market and then move to new markets at a later stage of development then their profits will be higher in their foreign operations.

The accounts of the major tech giants appear to support this theory, with profit margins higher on non-US sales than on US sales. For example, in 2019, 79% of Facebook’s profits were made outside the US, even though non-US sales only accounted for 55% of revenues. Adobe reported that it made 86% of its profits outside the United States in 2019, despite the majority of its sales being to US customers.

By using the average profit margin across the group of companies, it means that the sales made in foreign jurisdictions share the costs of product development and other costs with the home jurisdiction. Our methodology therefore arguably underestimates the true scale of tax avoidance in some of the companies in our study.

Results

We find based on our estimates of profit arising from the UK market, that the eight global tech giants underpay their taxes by £1.5bn a year. The majority of this comes from just two companies. The largest tax avoider based on our methodology is Apple, which sees an underpayment of £518m, followed by Alphabet (Google), with £452m.

This is a significant finding. The current solution put forward by the OECD to redistribute profits made by large multinationals to countries like the UK (known as the Pillar One proposals) foresees a total tax benefit of between $5-$12bn across all jurisdictions (i.e. not just the UK) on an analysis which was based on a dataset of 27,000 multinational groups.2

In fact, the OECD proposal which would only seek to redistribute profits that exceeded “routine profits” set at a 10% margin may mean that a company like Amazon would not be impacted at all.

The impact of global minimum levels of taxation on the tech giants

Although US based global tech giants have traditionally paid very little in corporate taxation in markets such as the UK, companies do pay taxes in the US. Indeed, in defence of their position, Facebook told that they pay the majority of their taxes in the US.

This was clearly demonstrated in our research published in April 2020, which found that the effective tax rate of several large technology companies was four times higher in US than the rest of the world. In simple terms, this means that US corporations pay four times more tax on each dollar profit they make in the US than they do on each dollar of profit their make outside of the US.

It is important to be clear that the significant tax bills that these companies face in the US have historically not been the product of profits from Europe being moved back to the United States to be taxed there. Large US corporations with a significant presence both in the US and the rest of the world pay their taxes on profits arising from sales to US customers, but by moving profits out of foreign jurisdictions into tax havens, effectively eliminate taxes paid on taxes on profits declared in the rest of the world – which make up a significant amount of the total profits of these companies. The result of this is that the effective tax rates of these corporations have sat between the US headline rate and zero.

To some extent this changed in 2017, when US tax reform introduced a form global minimum taxation, the GILTI, a charge placed on the profits of US corporations declared in tax havens. In addition, the 2017 reforms included a form of tax subsidy on profits moved back to the United States, the FDII. This provided a powerful incentive for companies to move their profits back to the United States where they would be taxed at a lower rate.

More research is needed to look at how firms responded to this change. However it is clear that Google responded by moving more profit back to the US. This is clearly shown in the latest Google 10-K which shows that between 2019 and 2020 the amount of profit the company allocated to its international operations (which previously almost all arose in Bermuda) declined by more than 50% from $23bn to $10.5bn. All of that profit re-appeared in the United States which saw its share of Google’s profits jump from $16bn to $37bn over the same period. The net result of this is that US federal tax payments more than doubled from $2.4bn in 2019 to $4.8bn in 2020, whilst taxes paid to non-US governments declined from $2.7bn to $1.7bn.

All of this highlights the fundamental importance of distribution in any agreement on global tax reform that ensures that more profit is allocated to the countries where real economic activity, users and customers are located. Countries should also reserve the right to take unilateral action to ensure that profits are properly taxed in their jurisdiction regardless of any deal on global minimum rates.

Could a global minimum tax rate benefit the UK?

There has been much confusion over the idea of global minimum tax rates, with even some Finance Ministers appearing to misunderstand what a global minimum means in practice. A global minimum tax rate is not an agreement between all countries of the world to increase their tax rates above a minimum level.

Instead a global minimum can be imposed by countries on companies headquartered in their jurisdiction on profits that arise overseas (the US already has a form of global minimum taxation).

When a company moves profits into tax havens, the tax administration in its home jurisdiction places an additional tax charge to top up the tax payments of the company. This severely limits the potential benefits of a tax haven.

This means that in practice, it simply does not matter if any jurisdiction refuses to sign up to a minimum tax rate as the finance ministers of Ireland, or Hungary have threatened. If a jurisdiction adopts a tax rate below the global minimum imposed on a foreign multinational corporation operating in country, all that will happen is that the country with the lower rate will be giving up tax revenues to the multinational’s home country.

The negotiations at an OECD level are for OECD members and other jurisdictions to co-ordinate economic policy so that each jurisdiction will impose a minimum tax on its corporations, resulting in a patchwork of rules that will create a system of global minimum taxation.

Although global minimum tax rates therefore primarily benefit the home jurisdictions of any multinational companies they apply to, there are benefits to the UK of adopting the policy and there are broader benefits that arise to all jurisdictions. Firstly, the UK is itself home to a great number of multinational corporations that themselves avoid taxes on their foreign profits. UK participation in a system of global minimum taxation would therefore ensure that substantial amounts of revenues could be raised from UK multinationals.

Secondly, the policy mitigates against the tendency for some jurisdictions to engage in a race to the bottom through cutting taxes to encourage foreign multinationals to establish subsidiaries in their countries. It places a floor on the race to the bottom where previously there was none.

Thirdly, the disincentive to use tax havens means that profit shifting is less likely. There are multiple costs to profit shifting through the potential for enforcement action, reputational damage, and the cost of paying tax advisors to create schemes. Reducing the incentive on companies to engage in profit shifting may therefore mean that fewer companies engage in the practice. Less profit shifitng will benefit the UK.

Finally, there are also some measures currently being negotiated as part of the OECD global minimum tax rate package that seek to protect the tax bases of non-home country jurisdictions. But it is not clear how effective these will be.

However, as long as there is a difference between the global minimum rate applied to any particular company, and the headline corporation tax rate of the country where they operate, there will be an incentive for companies to shift profits could remain. For example, if the global minimum rate applied to US corporations by the US Government was 21%, and the UK headline rate rises to 25% as planned, then there will still be an incentive to move profits out of the UK and into tax havens, or back to the US.

As such, it is in the UK’s interests to seek to negotiate a global minimum rate of taxation at least as high as 25%, the rate which the UK Government is currently seeking to legislate for.

Figures

eBay

The online marketplace eBay launched in the UK back in 1999. eBay’s 10-K shows a global pre-tax margin of 20%, and UK revenues of £997,172,883 for 2019, which gives us an estimated profit on UK revenues of £201,951,765.

With the UK’s 19% corporate tax rate, this means eBay had an estimated UK tax liability of £38.4m.

eBay UK had a UK current tax charge of £5,260,000 in 2019.

Based on the above, we estimate eBay avoided £33m in UK taxes in 2019.

eBay did not respond to our request for comment.

Adobe

Adobe’s 10-K shows a global margin of 33%, and UK revenues of £616,487,008 for 2019, which gives us an estimated profit on UK revenues of £205m. With the UK’s 19% corporate tax rate, this means Adobe had an estimated UK tax liability of £39m.

Adobe Systems Europe Limited, a UK company, had a £2,040,000 tax charge in 2019.

Based on the above, we estimate Adobe avoided £37m in UK taxes in 2019.

Adobe did not respond to our request for comment.

Amazon

Amazon’s 10-K shows a global pre-tax margin of 5%, and Amazon have confirmed to us that their UK revenues for 2019 were £13,730,000,000,3 which gives us an estimated profit on UK revenues of £684m.

With the UK’s 19% corporate tax rate, this means Amazon had an estimated UK tax liability of £130m

Amazon Services UK Limited and Amazon Web Services UK Limited had a £14,582,000 current tax charge in 2019 between them.

It is the case that Amazon’s Luxembourg operation, which is the entity that charges UK customers, reports its profits and revenues directly to HMRC (although it does not provide figures for the UK publicly). We assume that this entity did not have a significant UK tax liability because its accounts show that it is loss making.

Based on the above, we estimate Amazon avoided £115m in UK taxes in 2019.

Amazon responded to our request for comment, with an Amazon spokesperson saying:

“These calculations are wildly inaccurate, and do not include the majority of taxes we pay in the UK. Our total tax contribution in the UK was £1.1 billion during 2019 – £293m in direct taxes and £854m in indirect taxes. During 2019, our international consumer business was loss-making as we continued to invest heavily.”

Alphabet (Google)

Until 2016 Google reported the revenues it made from the UK in its US 10-K filing. On average, around 9% of Google’s global revenues came from the UK between 2014 and 2016. We applied this average figure to Google’s 2019 global revenues to estimate the revenues generated from the UK in 2019.

Alphabet’s US 10-K shows a profit margin of 24%. Using this margin, we estimate the profits on UK revenues to be £2,689bn.

At a 19% UK corporation tax rate, that would amount to a tax liability of £511m.

Google UK Limited’s accounts show a current UK tax charge for 2019 of £58,987,000.

Based on the above, we estimate Alphabet avoided £452m in UK taxes in 2019.

Alphabet did not respond to our request for comment.

Cisco

The accounts of Cisco Systems International report separately on the revenues the company makes from UK sales, which were £1,699,691,993 in 2019.

Cisco’s US 10-K shows a pre-tax margin of 28%. Using this margin, we estimate the profits on UK revenues to be £616m.

At a 19% UK corporation tax rate, that would amount to a tax liability of £117m.

In 2019, Cisco International Limited and another UK subsidiary, Cisco Systems Limited, were charged £48,919,131 in tax between them.

Based on the above, we estimate Cisco avoided £68m in UK taxes in 2019.

Cisco did not provide us with a comment on our findings.

Facebook

In order to estimate Facebook’s real revenues in the UK, we looked at Facebook’s average revenue per user (APRU), which is published in a chart, broken down by region, appended to the company’s US stock market filings. We then took the mid-point between the US APRU and European APRU basing this calculation on the assumption that the UK would be at the top end of the European APRU range, but less than the US. Using the online tool Napoleon Cat, we calculated an average of 42,000,000 Facebook users in the UK for 2019.

On the basis of these numbers we estimate Facebook’s revenues from the UK to be £2.9bn in 2019.

Facebook’s US 10-K shows a pre-tax margin of 35%. Using this margin, we estimate the profits on UK revenues to be just over £1bn. At a 19% UK corporation tax rate, that would amount to a tax liability of £194m.

Facebook’s UK accounts show a current UK tax charge for 2019 of £40,049,000.

Based on the above, we estimate Facebook avoided £154m in UK taxes in 2019.

We spoke with Facebook, who informed us that they have changed their structures to be more transparent about where they generate revenue, including in the UK. Facebook have stated that they have long supported the OECD process which is looking at new international tax rules for the digital company.4 They also said that currently the tax system does not allocate profits to jurisdictions where customers are located and pay the majority of their taxes in the United States.

Microsoft

Microsoft’s UK accounts do not disclose any information about how much the company earns from UK customers. All of the revenue earned by Microsoft’s main UK subsidiary, Microsoft Limited, is income earned from other Microsoft subsidiaries.

Microsoft’s UK sales are booked by an Irish subsidiary, Microsoft Ireland Operations Limited. Its accounts provided a figure for revenues earned from the UK until 2015. Between 2013 and 2015 the proportion of the company’s global revenues that came from the UK ranged between 3.12% and 3.6%.

To estimate revenues that Microsoft gained from the UK in 2019 we applied an average of 3.44% to the global revenue of Microsoft.

We estimate that Microsoft generated revenues of £3.4bn in 2019 from UK customers. This would yield an estimated profit of £1.2bn and a tax bill of £2225m in 2019. Microsoft Limited had a tax bill of £34,194,000 in 2019.

Based on the above, we estimate Microsoft avoided £191m in UK taxes in 2019.

Microsoft did not respond to our request for comment.

Apple

Our figures for how much revenue Apple makes from the UK market is derived from an estimate of the amount of money spent by UK customers on iPhones in the UK, which was constructed from data on smart-phone penetration and market research.

This shows that £6.3bn was spent on iPhones in the UK in 2017. This accounted for 6% of Apple’s global iPhone sales. If we assume that other Apple products have a similar market share in the UK, and that the amount UK is worth as a per cent of the global total remained the same, then in 2019 Apple would have made revenues of £12.7bn in the UK.

Applying Apple’s global pre-tax profit margin of 25% implies an estimated profit of £3,2bn and a tax bill (at the UK’s 19% corporation tax rate) of £610m.

Apple (UK) Ltd, Apple Europe Ltd, and Apple Retail UK had a £92,839,000 UK tax charge between them in 2019.

Based on the above, we estimate Apple avoided £517m in UK taxes in 2019.

Apple did not respond to our request for comment.

eBay

Adobe

Amazon

Google

Cisco

Facebook

Microsoft

Apple

Total

Current UK Tax Charge

£5,260,000

£2,040,000

£14,582,000

£58,987,000

£48,919,132

£40,049,000

£34,194,000

£92,839,000

£ 296,870,132

UK Estimated Revenues

£997,172,883

£616,487,008

£13,730,000,000

£10,979,551,791

£1,699,691,993

£2,904,305,592

£3,408,827,130

£12,699,873,462

£ 47,035,909,859

Estimated Profits on UK Revenues

£201,951,757

£205,090,395

£684,047,882

£2,687,957,516

£616,194,000

£1,019,302,521

£ 1,183,417,748

£3,208,820,181

£ 9,806,782,000

UK Estimated Tax Liability

£38,370,834

£38,967,175

£129,969,098

£510,711,928

£117,076,860

£193,667,479

£ 224,849,372

£609,675,834

£ 1,863,288,580

Estimated UK Tax Avoided

£33,110,834

£36,927,175

£115,387,098

£451,724,928

£68,157,728

£153,618,479

£ 190,655,372

£516,836,834

£1,566,418,448

This report is available as a PDF here.

 

1 Amazon reports UK sales rose by 51% in 2020, The Guardian, 03 February 2021, https://www.theguardian.com/technology/2021/feb/03/amazon-reports-uk-sales-rose-by-51-in-2020

2 Tax revenue effects of Pillar One, Tax Challenges Arising from Digitalisation – Economic Impact Assessment, OECD, 12 October 2020, https://www.oecd-ilibrary.org/sites/0e3cc2d4-en/1/3/2/index.html?itemId=/content/publication/0e3cc2d4-en&_csp_=60cbdb3f912de71310706737fc50a27f&itemIGO=oecd&itemContentType=book

3 Email from Amazon to TaxWatch, 31 May 2021

4 Facebook boss ‘happy to pay more tax in Europe’, BBC News, 14 February 2020, https://www.bbc.co.uk/news/business-51497961

Amazon’s ballooning revenues demonstrate the need for transparency in tax affairs

4th February 2021 by Alex Dunnagan

The news that Amazon’s sales in the UK have ballooned over the last year have sparked the traditional headlines about Amazon’s tax bill (or lack of it).

This article looks to tease out some of the complex issues that arise from Amazon’s tax structure in the UK.

There is no doubt that Amazon has engaged in tax avoidance in Europe and the UK. Its structure, with a European hub in Luxembourg and satellite companies in the UK and other markets, was originally designed to exploit Luxembourg’s low VAT rate on books and other tax benefits.1 The way that the company divided profits between the UK and Luxembourg (with very little profit being declared in the UK) was based on the fiction that Amazon’s European companies were separate entities operating independently of each other.2

Much of this changed in 2015. Ahead of the UK’s introduction of a “diverted profits tax” which was designed to attack structures such this, Amazon established a UK “branch” of its main European HQ in Luxembourg.3

Its UK company, Amazon UK Services, which delivers goods on behalf of its Luxembourg parent, continues to operate, making a small amount of profit and paying a little corporation tax each year. However, the branch structure means that the company would also start reporting the sales made to UK customers from Luxembourg to HMRC and pay corporation tax on those profits.

So was that the end of the story with regards to Amazon’s corporation tax in the UK? No.

The branch structure means that Amazon does not publish any public accounts that detail the sales and profits it makes from UK customers or the taxes paid on them. All that is there is the company’s Luxembourg accounts, which show that as a whole, the company had no net tax liability and instead receives tax credits.4 It is possible that there was some corporation tax paid in the UK, which was offset elsewhere, but we just don’t know.

Amazon’s global accounts state that the Luxembourg entity did make a profit in 2020 as a whole, but that this was was offset by deferred tax assets.

Deferred tax assets are often the result of losses a company has made in previous years that it can offset against profits made today or in the future. Interestingly, the company also said that it still believes that its earnings in Luxembourg are uncertain, and as such, has written off some of their deferred tax assets on the basis that it will be unlikely to generate enough profit to offset previous losses against future profits.5

So whilst business across the world is booming (and has been booming for years), and the company is recording its highest ever profits, the accountants appear to believe that its European operations may continue to be loss making (or at least not make much profit) for a significant time to come.

For a company that was built partly on tax avoidance all of this is hardly reassuring, and Amazon’s attempts to obfuscate the issue raises even more concern.

Specifically, Amazon publish what they call their annual economic impact in the UK.6 This attempt at assuaging criticism of the company’s historic tax dodging seeks to set out the company’s total contribution to the UK economy, including the total amount of taxes it pays. This includes direct taxes the company pays, like corporation tax, as well as taxes it pays on behalf of consumers, such as VAT. The company states that in 2019 it paid £293m in direct taxes.

The problem with this is that the figure for direct taxes includes Employer’s National Insurance, which is a large bill for a significant employer like Amazon and will dwarf any corporation tax liability they have. The comparisons of this figure that some media outlets have made with other UK retailers’ corporation tax bills are grossly misleading.

Looking at Amazon’s global accounts, it must be said that the company does state that it pays taxes overseas. The latest annual report for 2020 published on February 3rd shows that the company had a current tax liability of close to $1bn outside the United States on profits of $4bn.7 However, although this tax rate of around 25% is more than the UK’s statutory tax rate, there is no indication whatsoever as to where this is paid, it might all be paid on Amazon’s operations in Asia or anywhere else. We just don’t know.

The UK government could easily rectify this lacuna. Every major multinational needs to report their economic activity, including their profits and taxes on a country by country basis. This report is then given to tax authorities on a confidential basis.

In 2016 MPs voted to give the power to ministers to compel the publication of these reports (for all major companies, not to single out any individual one). However, the government has yet to take up the offer.8

If the government were to do this, or Amazon were to publish their country-by-country report themselves, this would go a long way to establishing the reality of Amazon’s tax position in the UK and other jurisdictions around the world.

Two final things worth noting. Amazon’s published figures for how much profit it makes outside of the United States may well be significantly deflated by royalty payments paid by its overseas operations to the US.

Amazon’s latest corporate accounts show an adjustment in their tax liability against the US headline corporation tax rate of $372 for Foreign income deduction.9 This is the Foreign-Derived Intangible Income (FDII) provisions of the US tax code which gives US based corporations a discount of 7.75 per cent on royalty income on intellectual property they bring in from overseas. This suggests that Amazon’s international operations sent back almost $5bn in royalties from overseas, vastly more than the total profit they made outside the US.

Secondly, Amazon’s global accounts also show that the company received $639m in tax credits.10 These appear to be primarily related to US government subsidies on research and development. The use of tax credits to help subsidise things like research and development has become increasingly popular with governments around the world in recent years. As we have previously pointed out in our work on the creative industry tax credits in the UK, by design, much of these credits end up subsidising very large profitable corporations. It seems that the US is no different in this regard.

Our Director discussed some of the issues raised in this blog on Times Radio on 04 February 2021.

Photo by Christian Wiediger on Unsplash

1 David Pegg, From Seattle to Luxembourg: how tax schemes shaped Amazon, The Guardian, 25 April 2018, https://www.theguardian.com/technology/2018/apr/25/from-seattle-to-luxembourg-how-tax-schemes-shaped-amazon

2 Clair Quentin, Risk-Mining the Public Exchequer, Journal of Tax Administration Vo. 3, No 2 (2017), http://jota.website/article/view/142

3 Simon Bowers, Amazon to begin paying corporation tax on UK retail sales, The Guardian, 23 May 2015 https://www.theguardian.com/technology/2015/may/23/amazon-to-begin-paying-corporation-tax-on-uk-retail-sales

4 Mark Sweney, Amazon given €294m in tax credits as European revenues jump to €32bn, The Guardian, 21 April 2020, https://www.theguardian.com/technology/2020/apr/21/amazon-given-294m-in-tax-credits-as-european-revenues-jump-to-32bn

5 Amazon 10-K 2020, p.63 available from: https://d18rn0p25nwr6d.cloudfront.net/CIK-0001018724/336d8745-ea82-40a5-9acc-1a89df23d0f3.pdf

6 Dayone Blog, 2019:Amazon’s Economic Impact in the UK, https://blog.aboutamazon.co.uk/jobs-and-investment/2019-amazons-economic-impact-in-the-uk

7 Amazon Inc 2020 10-K p.62

8 Out-Law News, UK MPs back power to make multinationals’ country-by-country tax reporting public, https://www.pinsentmasons.com/out-law/news/uk-mps-back-power-to-make-multinationals-country-by-country-tax-reporting-public

9 Amazon Inc 2020 10-K p.62

10 Amazon Inc 2020 10-K p.63

US effective tax rate over four times higher for tech companies

8th April 2020 by George Turner

A new study has shown that large technology companies have historically paid more than four times in tax on their US profits than on profits made in the rest of the world.

In our latest study, we looked at pre-tax profits reported by major multinational companies in the technology sector. Our study looked at Microsoft, Apple, Alphabet, Facebook, Cisco Systems, Adobe, Intel and Nvidia.

Under US stock market rules companies have to report the amount of their pre-tax profits that are made overseas and the taxes paid to foreign governments.

The study found that over the last five years technology companies have faced a tax rate of just 9.6% on profits generated outside of the US. By contrast, the same companies have seen a tax liability of 45% on profits generated in the United States.

Part of this significant gap is explained by the large, one-off tax bills faced by companies in the US to deal with historic tax abuse following tax reform in 2017. For example, in 2017 Apple faced a tax bill of 71% on its US profits. Google had an effective tax rate of 120% on its US profits. For some companies, these large charges also appear in 2018 and 2019 as new rules were issued by the IRS on how to account for the tax reforms brought in in 2017.

Average ETR 2015-2019 Microsoft Apple Alphabet Facebook Cisco Adobe Intel Nvidia Total
Foreign total tax rate 14.57% 8.21% 8.54% 5.21% 13.29% 6.94% 10.17% 3.12% 9.60%
Foreign income as % total 72.22% 67.31% 56.59% 66.06% 68.37% 67.14% 41.63% 57.74% 63.51%
US current tax rate 76.27% 68.44% 37.89% 52.42% 76.10% 32.22% 30.03% 10.06% 54.88%
US total tax rate 52.13% 48.08% 39.71% 48.36% 80.94% 20.81% 31.27% 3.75% 45.32%

However, pre-tax reform there were still very significant gaps between the rates these companies paid on US profits and on non-US profits.

In 2016, Google had a tax bill of just 7.6% outside the US and a rate of 28.7% on its US profits. In the same year, Facebook paid just 2.6% of non-US profits in tax, whereas in the US it faced a tax bill of 30.9% on US profits.

Recently, the gap appears to have closed, following significant tax cuts in the United States, which saw the headline rate of federal corporation tax fall from 35% to 21% in 2017. At the same time action by tax officials around the world has increased the focus on tax avoidance by multinational companies.

In 2019, the US based technology companies in our study had a total foreign tax bill of 13.6% on profits generated outside of the United States. The total US tax bill was 15.6% of US profits, or 25.4% on a current tax basis.

The worldwide average tax rate was 26% in 2019 when weighted by GDP.

The figures call into question the claims made by companies on why they pay so little tax outside of the United States. Frequently when challenged companies claim that the reason that non-US governments see relatively small tax payments in their jurisdiction is due to the fact that profits should be allocated to the United States – where the value of the product is created.

However, TaxWatch’s study shows that the majority of US tech companies state in their annual accounts that most of their profits are made outside of the United States. On average the companies in our study reported that 63.5% of their profits were generated outside of the United States.

In 2019, Facebook states that 79% of its pre-tax profit was made outside of the US. Adobe claimed that it made 86% of its profits outside of the United States, on which it paid a tax rate of just 7.2%. Nvidia, the maker of high end graphics cards made 50% of its profits outside of the US, on which it paid just 3.7% tax.

The study also showed significant differences between the corporation tax liabilities of different companies on their non-US profits.

Between 2015 and 2019 Microsoft paid 14.6% of its non-US profits in tax, whereas over the same period Nvidia paid just 3.12% of its foreign earnings in tax. Most companies achieved figures in the single digits.

To download a copy of this briefing in PDF – click here.

Photo by Allie Smith on Unsplash


TaxWatch

c/o M & A Solicitors

38 Coney Street

York

YO1 9ND

 

info@taxwatchuk.org

Copyright © 2023 TaxWatch

 

 

Follow us on X (Twitter)