Taxation might not top the agenda of the readership of EJIL:Talk!. And yet, we have all heard the chorus of mass media and policymakers decrying the tiny tax liabilities of large corporations operating in the digital economy. Does international law have nothing to contribute to this discussion?
Of course it does. In fact, the international norms governing cross-border taxation have kept states very busy (and quarrelsome) in recent years. The United States and Canada are currently entangled in a diplomatic spat about the legitimacy of Canada’s impending ‘digital services tax’ (‘DST’). When 11 states took similar steps towards DSTs between 2019 and 2021, the USA was equally unhappy – and responded with threats of unilateral retaliatory tariffs under Section 301 of the US Trade Act of 1974.
In parallel, the OECD has been pushing for a multilateral alternative to the taxation of large multi-national enterprises (‘MNEs’) and especially those operating in the digital economy. The OECD’s ‘Pillar One’, to be implemented with a multilateral treaty, seeks to create new taxing rights for so-called ‘market jurisdictions’ – states where particularly large and profitable MNEs generate significant revenue, even if such MNEs are not otherwise connected to these market jurisdictions. Pillar One has faced criticism for only targeting a handful of MNEs, for being excessively complicated and costly to implement and administer, and for not being sufficiently beneficial for developing countries.
The discontent with the current international tax regime and the OECD’s reform efforts does not end here. Most notably, the OECD’s dominance in international tax policymaking has come under fire for failing to include the voices and interests of developing countries in a meaningful way (see EJIL:Talk! of 10 November 2022). This line of criticism has just grown some new teeth. On 22 December 2023, the UNGA adopted resolution A/78/459/Add.8 to develop a multilateral framework convention on international tax cooperation. Although the resolution does not explicitly criticize the OECD, its aim to establish the legal basis for ‘fully inclusive and more effective’ international tax law clearly signals dissatisfaction with the status quo of tax leadership. The voting pattern of the resolution, which was initially tabled by the African Group, also reveals a fault line between countries that are represented in the OECD and those that are not: none of the 111 votes in favour were from an OECD member state. OECD member states have explained their rejection of the resolution with concerns that a UN tax convention might duplicate and undermine ongoing negotiations led by the OECD.
In brief, international tax law is in upheaval. Despite the United States’ efforts to contain the spread of DSTs, and despite the OECD’s efforts to retain its pre-eminence in tax policymaking, the genie is out of the bottle. Future international tax law will reshuffle the exercise of tax jurisdiction. But how? And who will benefit? We are in the midst of a paradigm shift that will determine the answers to these major jurisdictional (and distributional) questions. Given the crucial role that international taxation plays in the international economic order, it is high time for public international lawyers to pay attention and contribute to these debates.
This post concentrates on one of the core causes of the upheaval in international tax law: states’ discontent with the rules that govern taxation in the digital economy.
Trouble in tax paradise
The current turmoil is atypical for international tax law, a field that is otherwise associated with a static legal regime and path dependence. Indeed, the foundational principles of international taxation have barely changed in the past 100 years, when the first model tax conventions were negotiated by the League of Nations. These principles are now enshrined in approximately 3,000 bilateral tax treaties, most of which are highly similar to the OECD or UN Model Tax Convention.
A core principle that underlies these boilerplate treaties is that a state may only tax cross-border income if either the taxpayer (both natural persons and corporations) or the economic activity is physically tied to that state. Under the current regime, business profits can only be taxed by (a) the corporation’s country of residence or (b) by the ‘source country’ of the profits. A state qualifies as a source country under a tax treaty if business profits were generated through a so-called ‘permanent establishment’ in that state’s territory. And that’s the catch: tax treaties define permanent establishments as a certain de minimis measure of physical presence.
Evidently, tethering the definition of a permanent establishment to physical presence causes issues in the digital economy. When the international tax regime was formed 100 years ago, there was no digital economy to consider. Today, physical presence is no longer necessary to make business profits in a state. Corporations offering digital services – through streaming platforms, online marketplaces, social media, and search engines – do not need to ‘set foot’ (or set up a server) in a country to generate profits in that market.
This is where DSTs come in. Since 2019, states across all continents have implemented DSTs. France, for instance, has collected more than a billion Euros in DST-revenue since 2019. DSTs levy low tax rates of 2-5% on revenue from digital services performed by large enterprises. An enterprise is considered ‘large’ if it reaches a certain threshold of global annual revenue (e.g. EUR 750 million) and domestic annual revenue from digital services (somewhere between EUR 3-25 million). Notably, DSTs also apply to taxpayers that have neither their tax residence nor a permanent establishment in the taxing state. Instead, DSTs tie the exercise of tax jurisdiction to the physical presence of the consumers of digital services.
DSTs are creations of domestic law, without prior international coordination or agreement. As is often the case with unilateral acts that affect foreign corporations, DSTs have faced opposition.
Why the outrage about unilateral DSTs?
The most forceful opposition against DSTs has come from the US. When France adopted its DST in 2019, the US Trade Representative (‘USTR’) initiated investigations under Section 301 of the US Trade Act of 1974. By mid-2020, the USTR had initiated Section 301 investigations in response to DSTs in ten more jurisdictions, namely Austria, Brazil, the Czech Republic, the EU, India, Indonesia, Italy, Spain, Turkey, and the UK.
International trade lawyers know Section 301 as the United States’ tool to impose trade sanctions on states that allegedly violate international trade agreements or engage in ‘unjustifiable’ or ‘unreasonable’ acts burdening US commerce. In the case of DSTs, the US grievance was not the violation of any trade agreements, but that DSTs effectively discriminated against US corporations: most tech giants that reach the revenue thresholds of DSTs are US tax residents.
Interestingly, the USTR did not determine that DSTs violated US tax treaties. Rather, they chose more cautious wording: DSTs were ‘inconsistent’ with prevailing international tax ‘principles’. This is noteworthy because some international tax lawyers have claimed that DSTs violate existing tax treaties.
In brief, some international tax lawyers argue that DSTs are designed to circumvent tax treaties by incurring on revenue instead of business profits. Nominally, taxes on revenue fall outside the scope of tax treaties, which only govern the taxation of cross-border income and business profits (i.e. revenue minus any related business expenses). However, the argument goes, tax treaties must be interpreted and applied in good faith – meaning that a tax with the same effects as an income tax must comply with applicable tax treaties. If states effectively tax cross-border business profits of non-residents without a permanent establishment, they violate their tax treaties. This argument has merit – but only if DSTs are actually taxes on business profits, disguised as revenue taxes. First and foremost, tax treaties must be interpreted and applied in line with objective good faith to reflect a reasonable reading of the treaty. It is secondary that the DST-states’ subjective intention might be to dodge the permanent establishment principle.
The DST-saga under Section 301 ultimately fizzled out. Although the USTR considered DSTs actionable under Section 301 and formally imposed additional tariffs of 25% on approximately USD 3.4 billion worth of US imports from seven DST-states, the USTR repeatedly suspended and finally terminated these actions in 2021. The United States’ sudden change of heart was the result of a political compromise. Under the so-called ‘Unilateral Measures Compromise’ reached in October 2021, the DSTs remain in place but must be repealed once OECD Pillar One takes effect (an issues that we address right below). DSTs accrued prior to that date are creditable against future taxes levied under Pillar One. In exchange, the US withdrew its Section 301 actions.
What about a multilateral solution under ‘Pillar One’?
What might sound like a mutually beneficial compromise between the US and DST-states suffers from a serious flaw. Pillar One aims to create taxing rights for ‘market jurisdictions’. These market jurisdictions may tax part of the profits of very large and profitable MNEs that derive a certain amount of annual revenue from their market. Thus, under Pillar One, neither a permanent establishment nor the corporation’s residence is required to trigger taxing rights; the only physical connection to the taxed profits is the consumers’ presence in the market jurisdiction.
Pillar One was originally intended to create fairer tax rules for the digital economy. However, in its current form, Pillar One has lost this explicit connection to the digital economy: MNEs that do not operate in the digital economy may be subject to taxation under Pillar One just as much as large tech companies. What matters is not what kind of goods and services the MNE offers but whether the MNE reaches the revenue and profitability threshold to be ‘in scope’ of Pillar One.
On 11 October 2023, the OECD published a draft for the multilateral convention (‘MLC’) meant to implement part of Pillar One. This draft convention is not only 212 pages long – and might therefore win the award for the longest convention ever. It also solidifies concerns that Pillar One will never actually become reality: the MLC will only enter into force upon ratification by at least 30 states that simultaneously encompass the residence countries of 60% of the MNEs ‘in scope’ of Pillar One. Almost half the MNEs that are in scope of Pillar One are US tax residents. This means that without US ratification, the MLC cannot enter into force. As the MLC ‘will not pass Congress’, Pillar One is essentially doomed to fail.
What does this mean for the ‘original’ DST-states that were able to strike a deal with the US in 2021? For now, it looks like they can keep levying their DST – because Pillar One will never take effect.
A second round of US vs. DST: opposition against a Canadian DST
The gamble of France and other countries has paid off, but the Unites States’ crusade against DSTs does not end here. Canada has delayed the adoption of its own DST since 2021, hoping for a multilateral solution with Pillar One. In August 2023, Canada abandoned this strategy when its Department of Finance announced that it would move ahead and enact its own DST, at the earliest on 1 January 2024. What is more, the Canadian DST is set to apply retroactively from 1 January 2022 (not unlike the French DST, which also applied retroactively).
Canada referred to the slow progress of Pillar One to explain this step. Not impressed with this justification, the US and the Canadian business community responded with instant push-back. To date, the USTR has not acted on its threats to retaliate against a Canadian DST with Section 301 tariffs. But another bargaining chip has been brought into play to thwart the DST: the Canadian Business Council suggested that the US might withhold its consent to renew the Canada-United States-Mexico Agreement in 2026 if Canada were to proceed with its DST. Given the dependence of the Canadian economy on trade with its Southern neighbour, it will be interesting to see whether and how Canada’s DST unfolds in 2024.
The future of digital taxation
For years, states have been dissatisfied with the century-old norms of international taxation. Through DSTs, an increasing number of states seek to satisfy their appetite for a bite at the digital economy. At the multilateral level, the OECD’s Pillar One aims – and likely failed – to contain the spread of such unilateral responses to the digital economy. Whether the OECD’s reinvigorated competitor for tax leadership, the UN, will be more successful with a multilateral solution to tax issues in the digital economy remains to be seen. But the UN will certainly try: the recent UNGA resolution also targets ‘taxation of income derived from … cross-border services in an increasingly digitalized and globalized economy’.
As it stands, states that wish to tax tech giants have limited options. They can join the bandwagon of unilateral action (i.e. DSTs) and possibly face off with the US; or hope for a multilateral deal under the auspices of either the OECD or the UN. For now, these states are stuck between Apple and a hard place.
A Bite at the Apple: States’ Struggle to Tax Digital Services
Written by Céline Braumann and Danielle Du MaresqTaxation might not top the agenda of the readership of EJIL:Talk!. And yet, we have all heard the chorus of mass media and policymakers decrying the tiny tax liabilities of large corporations operating in the digital economy. Does international law have nothing to contribute to this discussion?
Of course it does. In fact, the international norms governing cross-border taxation have kept states very busy (and quarrelsome) in recent years. The United States and Canada are currently entangled in a diplomatic spat about the legitimacy of Canada’s impending ‘digital services tax’ (‘DST’). When 11 states took similar steps towards DSTs between 2019 and 2021, the USA was equally unhappy – and responded with threats of unilateral retaliatory tariffs under Section 301 of the US Trade Act of 1974.
In parallel, the OECD has been pushing for a multilateral alternative to the taxation of large multi-national enterprises (‘MNEs’) and especially those operating in the digital economy. The OECD’s ‘Pillar One’, to be implemented with a multilateral treaty, seeks to create new taxing rights for so-called ‘market jurisdictions’ – states where particularly large and profitable MNEs generate significant revenue, even if such MNEs are not otherwise connected to these market jurisdictions. Pillar One has faced criticism for only targeting a handful of MNEs, for being excessively complicated and costly to implement and administer, and for not being sufficiently beneficial for developing countries.
The discontent with the current international tax regime and the OECD’s reform efforts does not end here. Most notably, the OECD’s dominance in international tax policymaking has come under fire for failing to include the voices and interests of developing countries in a meaningful way (see EJIL:Talk! of 10 November 2022). This line of criticism has just grown some new teeth. On 22 December 2023, the UNGA adopted resolution A/78/459/Add.8 to develop a multilateral framework convention on international tax cooperation. Although the resolution does not explicitly criticize the OECD, its aim to establish the legal basis for ‘fully inclusive and more effective’ international tax law clearly signals dissatisfaction with the status quo of tax leadership. The voting pattern of the resolution, which was initially tabled by the African Group, also reveals a fault line between countries that are represented in the OECD and those that are not: none of the 111 votes in favour were from an OECD member state. OECD member states have explained their rejection of the resolution with concerns that a UN tax convention might duplicate and undermine ongoing negotiations led by the OECD.
In brief, international tax law is in upheaval. Despite the United States’ efforts to contain the spread of DSTs, and despite the OECD’s efforts to retain its pre-eminence in tax policymaking, the genie is out of the bottle. Future international tax law will reshuffle the exercise of tax jurisdiction. But how? And who will benefit? We are in the midst of a paradigm shift that will determine the answers to these major jurisdictional (and distributional) questions. Given the crucial role that international taxation plays in the international economic order, it is high time for public international lawyers to pay attention and contribute to these debates.
This post concentrates on one of the core causes of the upheaval in international tax law: states’ discontent with the rules that govern taxation in the digital economy.
Trouble in tax paradise
The current turmoil is atypical for international tax law, a field that is otherwise associated with a static legal regime and path dependence. Indeed, the foundational principles of international taxation have barely changed in the past 100 years, when the first model tax conventions were negotiated by the League of Nations. These principles are now enshrined in approximately 3,000 bilateral tax treaties, most of which are highly similar to the OECD or UN Model Tax Convention.
A core principle that underlies these boilerplate treaties is that a state may only tax cross-border income if either the taxpayer (both natural persons and corporations) or the economic activity is physically tied to that state. Under the current regime, business profits can only be taxed by (a) the corporation’s country of residence or (b) by the ‘source country’ of the profits. A state qualifies as a source country under a tax treaty if business profits were generated through a so-called ‘permanent establishment’ in that state’s territory. And that’s the catch: tax treaties define permanent establishments as a certain de minimis measure of physical presence.
Evidently, tethering the definition of a permanent establishment to physical presence causes issues in the digital economy. When the international tax regime was formed 100 years ago, there was no digital economy to consider. Today, physical presence is no longer necessary to make business profits in a state. Corporations offering digital services – through streaming platforms, online marketplaces, social media, and search engines – do not need to ‘set foot’ (or set up a server) in a country to generate profits in that market.
This is where DSTs come in. Since 2019, states across all continents have implemented DSTs. France, for instance, has collected more than a billion Euros in DST-revenue since 2019. DSTs levy low tax rates of 2-5% on revenue from digital services performed by large enterprises. An enterprise is considered ‘large’ if it reaches a certain threshold of global annual revenue (e.g. EUR 750 million) and domestic annual revenue from digital services (somewhere between EUR 3-25 million). Notably, DSTs also apply to taxpayers that have neither their tax residence nor a permanent establishment in the taxing state. Instead, DSTs tie the exercise of tax jurisdiction to the physical presence of the consumers of digital services.
DSTs are creations of domestic law, without prior international coordination or agreement. As is often the case with unilateral acts that affect foreign corporations, DSTs have faced opposition.
Why the outrage about unilateral DSTs?
The most forceful opposition against DSTs has come from the US. When France adopted its DST in 2019, the US Trade Representative (‘USTR’) initiated investigations under Section 301 of the US Trade Act of 1974. By mid-2020, the USTR had initiated Section 301 investigations in response to DSTs in ten more jurisdictions, namely Austria, Brazil, the Czech Republic, the EU, India, Indonesia, Italy, Spain, Turkey, and the UK.
International trade lawyers know Section 301 as the United States’ tool to impose trade sanctions on states that allegedly violate international trade agreements or engage in ‘unjustifiable’ or ‘unreasonable’ acts burdening US commerce. In the case of DSTs, the US grievance was not the violation of any trade agreements, but that DSTs effectively discriminated against US corporations: most tech giants that reach the revenue thresholds of DSTs are US tax residents.
Interestingly, the USTR did not determine that DSTs violated US tax treaties. Rather, they chose more cautious wording: DSTs were ‘inconsistent’ with prevailing international tax ‘principles’. This is noteworthy because some international tax lawyers have claimed that DSTs violate existing tax treaties.
In brief, some international tax lawyers argue that DSTs are designed to circumvent tax treaties by incurring on revenue instead of business profits. Nominally, taxes on revenue fall outside the scope of tax treaties, which only govern the taxation of cross-border income and business profits (i.e. revenue minus any related business expenses). However, the argument goes, tax treaties must be interpreted and applied in good faith – meaning that a tax with the same effects as an income tax must comply with applicable tax treaties. If states effectively tax cross-border business profits of non-residents without a permanent establishment, they violate their tax treaties. This argument has merit – but only if DSTs are actually taxes on business profits, disguised as revenue taxes. First and foremost, tax treaties must be interpreted and applied in line with objective good faith to reflect a reasonable reading of the treaty. It is secondary that the DST-states’ subjective intention might be to dodge the permanent establishment principle.
The DST-saga under Section 301 ultimately fizzled out. Although the USTR considered DSTs actionable under Section 301 and formally imposed additional tariffs of 25% on approximately USD 3.4 billion worth of US imports from seven DST-states, the USTR repeatedly suspended and finally terminated these actions in 2021. The United States’ sudden change of heart was the result of a political compromise. Under the so-called ‘Unilateral Measures Compromise’ reached in October 2021, the DSTs remain in place but must be repealed once OECD Pillar One takes effect (an issues that we address right below). DSTs accrued prior to that date are creditable against future taxes levied under Pillar One. In exchange, the US withdrew its Section 301 actions.
What about a multilateral solution under ‘Pillar One’?
What might sound like a mutually beneficial compromise between the US and DST-states suffers from a serious flaw. Pillar One aims to create taxing rights for ‘market jurisdictions’. These market jurisdictions may tax part of the profits of very large and profitable MNEs that derive a certain amount of annual revenue from their market. Thus, under Pillar One, neither a permanent establishment nor the corporation’s residence is required to trigger taxing rights; the only physical connection to the taxed profits is the consumers’ presence in the market jurisdiction.
Pillar One was originally intended to create fairer tax rules for the digital economy. However, in its current form, Pillar One has lost this explicit connection to the digital economy: MNEs that do not operate in the digital economy may be subject to taxation under Pillar One just as much as large tech companies. What matters is not what kind of goods and services the MNE offers but whether the MNE reaches the revenue and profitability threshold to be ‘in scope’ of Pillar One.
On 11 October 2023, the OECD published a draft for the multilateral convention (‘MLC’) meant to implement part of Pillar One. This draft convention is not only 212 pages long – and might therefore win the award for the longest convention ever. It also solidifies concerns that Pillar One will never actually become reality: the MLC will only enter into force upon ratification by at least 30 states that simultaneously encompass the residence countries of 60% of the MNEs ‘in scope’ of Pillar One. Almost half the MNEs that are in scope of Pillar One are US tax residents. This means that without US ratification, the MLC cannot enter into force. As the MLC ‘will not pass Congress’, Pillar One is essentially doomed to fail.
What does this mean for the ‘original’ DST-states that were able to strike a deal with the US in 2021? For now, it looks like they can keep levying their DST – because Pillar One will never take effect.
A second round of US vs. DST: opposition against a Canadian DST
The gamble of France and other countries has paid off, but the Unites States’ crusade against DSTs does not end here. Canada has delayed the adoption of its own DST since 2021, hoping for a multilateral solution with Pillar One. In August 2023, Canada abandoned this strategy when its Department of Finance announced that it would move ahead and enact its own DST, at the earliest on 1 January 2024. What is more, the Canadian DST is set to apply retroactively from 1 January 2022 (not unlike the French DST, which also applied retroactively).
Canada referred to the slow progress of Pillar One to explain this step. Not impressed with this justification, the US and the Canadian business community responded with instant push-back. To date, the USTR has not acted on its threats to retaliate against a Canadian DST with Section 301 tariffs. But another bargaining chip has been brought into play to thwart the DST: the Canadian Business Council suggested that the US might withhold its consent to renew the Canada-United States-Mexico Agreement in 2026 if Canada were to proceed with its DST. Given the dependence of the Canadian economy on trade with its Southern neighbour, it will be interesting to see whether and how Canada’s DST unfolds in 2024.
The future of digital taxation
For years, states have been dissatisfied with the century-old norms of international taxation. Through DSTs, an increasing number of states seek to satisfy their appetite for a bite at the digital economy. At the multilateral level, the OECD’s Pillar One aims – and likely failed – to contain the spread of such unilateral responses to the digital economy. Whether the OECD’s reinvigorated competitor for tax leadership, the UN, will be more successful with a multilateral solution to tax issues in the digital economy remains to be seen. But the UN will certainly try: the recent UNGA resolution also targets ‘taxation of income derived from … cross-border services in an increasingly digitalized and globalized economy’.
As it stands, states that wish to tax tech giants have limited options. They can join the bandwagon of unilateral action (i.e. DSTs) and possibly face off with the US; or hope for a multilateral deal under the auspices of either the OECD or the UN. For now, these states are stuck between Apple and a hard place.
Share this:
Related
Categories
Tags
Leave a Comment
Comments for this post are closed
Comments