For over a decade, countries have been looking for ways to tax the digital economy. On July 1, 130 countries announced an agreement that would provide a new taxing right to enable a country to tax a portion of digital profits even in the absence of traditional taxable nexus with the country. This new taxing right is known as “Amount A”. The quantum of Amount A remained a mystery until the publication of the OECD’s “Statement on a Two-Pillar Solution to Address the Tax Challenges Arising From the Digitalisation of the Economy” on July 1, 2021 (the “Statement”) which quantified Amount A to be “between 20-30% of residual profit defined as profit in excess of 10% of revenue” for in-scope enterprises. Although this quantum of Amount A represents a political compromise, a solid theoretical basis underlying that compromise is essential to sustaining consensus.
The early proposals to modify profit allocation and nexus rules for the digital economy enterprises, which ultimately produced Amount A, strived to be based on certain subjective criteria, including the concepts of user participation, marketing intangibles and/or the concept of significant economic presence. The contemplated methods for profit allocation were the Modified Residual Profit Split method, Fractional Apportionment method, and Distribution-based approaches, along with the options for business line and regional segmentation. However, the criteria and methods of the early proposals are nowhere to be found to found in the OECD July 2021 Statement, leaving many questions about Amount A still unanswered. While the final compromise on Pillar One eliminates the focus on digital economy and shifts instead to high profitability when defining in-scope MNEs, the “digital essence” still surrounds Amount A. For one thing, the introduction to the Statement continues to refer to the “two-pillar solution to address the tax challenges arising from the digitalisation of the economy.” Moreover, a widely accepted assumption in the final Pillar One negotiations is that high profits are generated by intangibles and those are increasingly concentrated with digital businesses. Therefore, an analysis of Amount A cannot be divorced from the analysis of the factors that contribute to the digital economy.
Tax authorities have long recognized the challenges in taxing the digital economy. Because companies can generate revenue in a country without the need for physical presence, the traditional rules for determining taxable nexus cannot easily be applied to provide a basis for taxation. On February 13, 2019, the OECD released the Public Consultation Document on “Addressing the Tax Challenges of the Digitalisation of the Economy” (February 2019 PCD). The February 2019 PCD outlined several policy proposals that would modify profit allocation and nexus rules based on the concepts of user participation, marketing intangibles and/or the concept of significant economic presence. Each of these proposals would be designed to recognize the value created by a multinational enterprise (MNE) in jurisdictions where the MNE sells its products or services, and bring (at least part of that) value within the new taxing nexus of that jurisdiction.
However, none of the proposals defined the meaning of the “value” created by an MNE in a jurisdiction. The “user participation” proposal focused on the residual profit of an MNE that would remain after routine activities have been allocated an arm’s length return. Only a portion of this residual profit would be attributed to the value created by the activities of users, which could be determined through quantitative/qualitative information, or through a simple pre-agreed percentage. On the other hand, under the “significant economic presence” proposal, the amount that would be apportioned would be the MNE’s “tax base” which could be determined by applying the global profit rate of the MNE group to the revenue (sales) generated in a particular jurisdiction. The tax base would be apportioned by taking into account not only factors such as sales, assets and employees, but also users if they “meaningfully contribute to the value creation process.” Still, under the “marketing intangibles” approach, the current transfer pricing and treaty rules would be modified to require marketing intangibles and risks associated with such intangibles to be allocated to the market jurisdiction under the premise that certain “marketing intangibles, such as customer data, customer relationships and customer lists are derived from activities targeted at customers and users in the market jurisdiction, supporting the treatment of such intangibles as being created in the market jurisdiction.”
A bit more insight into the mechanics of the calculation was provided in the “Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy” which was released by the OECD on May 28-29, 2019. In that document, we got a glimpse of the Modified Residual Profit Split method, Fractional Apportionment method, and Distribution-based approaches, along with the options for business line and regional segmentation. These methods and approaches loosely aligned with the concepts of user participation, significant economic presence, and marketing intangibles from the February 2019 PCD.
Then, in October of 2019, the OECD published Public Consultation Document titled “Secretariat Proposal for a “Unified Approach” under Pillar One” (October 2019 PCD), which introduced us to a “Unified Approach.” The Unified Approach proposed a three tier mechanism consisting of Amounts A, B, and C, where Amount A appeared to correspond to the user participation proposal and the Modified Residual Profit Split method; Amount B related to the marketing intangibles proposal and the Distribution-based approaches; and Amount C was defined as the tool to tidy things up if the local entity assumed additional functions above and beyond the baseline activity which is subject to the fixed return in Amount B. It was not immediately apparent whether Amount A corresponded in whole or in part to the “significant economic presence” proposal and the Fractional Apportionment method discussed in the earlier PCDs since both the proposal and the method were contemplated on the basis of the companies’ total pre-tax profit whereas Amount A was expected to correspond to a portion of the residual profit.
One of the major challenges of Pillar One was defining which businesses or segments were in scope. Indeed, while the taxation of the digital economy was Action #1 in the BEPS initiative, the OECD recognized the challenges in ring-fencing the digital economy and postponed its work in this area in favor of the other 14 Action Plans. Pillar One was the latest attempt at resolving these issues and yet the definitional challenges remained. In the meantime, many countries began imposing unilateral digital services taxes and similar levies rather than waiting for an international consensus. This led to global trade disputes and increased tariffs. Recognizing the challenges in defining an in-scope business as well as the criticism that Pillar One discriminated against US technology giants, the Biden administration proposed a change in approach. Rather than focusing on the types of businesses that should give rise to enhanced nexus, it proposed a new approach based on size and profitability. Instead of trying to define an in-scope business, the new approach essentially says that if you are big enough and profitable enough, you should be paying some tax in the countries where you sell your product regardless of whether you have a taxable presence there. That change in approach was reflected in the publication of the OECD’s “Statement on a Two-Pillar Solution to Address the Tax Challenges Arising From the Digitalisation of the Economy” on July 1, 2021 (“the Statement”): “For in-scope MNEs, between 20-30% of residual profit defined as profit in excess of 10% of revenue will be allocated to market jurisdictions with nexus using a revenue-based allocation key.”
Why 20-30%? The October 2019 PCD explained that Amount A would be “determined by simplifying conventions” with regard to both the level of the deemed routine profit and also the proportion of the deemed residual profit that should go to the market. However, none of the OECD pronouncements to-date provided sufficient insight into the potential mechanics of calculating Amount A, and what would likely be the foundation for developing the simplifying conventions. And now, with the “20-30% of residual profit” as an apparent consensus, we still don’t have a good handle on what Amount A is supposed to represent. It is quite possible that the proposed share of residual profit is not really meant to represent anything at all other than a politically obtainable consensus.
The user participation proposal and the significant economic presence proposal were developed around the businesses’ ability to generate income without physical presence. The “significant economic presence [is based on the] factors that evidence a purposeful and sustained interaction with the jurisdiction via digital technology and other automated means” provided certain other factors are met, including, among other, “the existence of a user base and the associated data input [and] the volume of digital content derived from the jurisdiction.” (February 2019 PCD, ¶51) The user participation proposal focused on the value creating activities of an active and engaged user base. (Ibid., ¶21). And yet, the user participation proposal “dismisses the idea that the value created by user activities can somehow be determined through the application of the arm’s length principle, e.g. through hypothesising the user base as a separate enterprise and asking what return it would receive at arm’s length in its dealings with other group entities.” (February 2019 PCD, ¶23) Then why 20-30%?
According to the “Digital Economy Report” published by the United Nations Conference on Trade and Development in 2019 (“2019 UNCTAD Report”), the “first challenge to measuring the digital economy, and therefore its value, is the lack of a universally accepted definition.., which makes international comparisons difficult.” (at 49) This challenge is compounded by the lack of statistical data as well as “the intangible nature of digital data and intelligence”… which complicate accounting for related economic activities in the data-driven economy. (Ibid.)
A typical econometric model – for example the one based on input-output tables – might be of little help. As the G20 “Toolkit for Measuring the Digital Economy” (“Toolkit”) produced in November of 2018 explains, “[t]he basic principle [behind an econometric model] is that the economic output (which can be roughly understood as GDP) could be treated as the result of the input of economic factors, -capital, labor, intermediate product and natural resources-.” (at 79) A return to each of these factors can help approximate the role these factors play in generating the economic output. For example, if the amount of capital input into the digital economy in a jurisdiction could be calculated reliably, then its contribution to the output could also be quantified. However, this contribution would only approximate the size of the digital economy in each jurisdiction but would not provide any insight into the value generated by significant economic presence or by user participation in a digital enterprise that could help determine Amount A.
Furthermore, there is significant cross-pollination between domestic and foreign digital and non-digital industries. As the Toolkit explains, “value added of domestic [information and communication technologies (‘ICT’)] industries is embodied in a wide range of final goods and services meeting final demand both at home and abroad. Similarly, domestic value added (DVA) from other industries (‘non-ICT‘) can be embodied in final ICT goods and services consumed globally.” (Toolkit, p.58)
This sentiment is echoed by the OECD’s Inter-Country Input-Output (ICIO) Database, http://oe.cd/icio, and Trade in Value Added (TiVA) Database, http://oe.cd/tiva: Specifically, that the added value that is embodied in final goods and services in the ICT industry in any given country contains components of value added from various countries and industries. It is no surprise then that the DVA has been, and will continue to be, difficult to measure. It is also obvious that parsing out the value which could form the basis of Amount A will remain a challenge.
If the “top down” measurement using input-output tables is too crude, perhaps the “bottom up” approach will be more promising? What if we started at the user/customer level and tried to determine the value of the user, user data, or user generated content (notwithstanding the February 2019 PCD, ¶23, that this approach has been dismissed)?
According to the 2019 UNCTAD Report, an attempt to value individual data may not be worthwhile. “…[I]ndividual data are of little or no value. Value emerges once data are compiled in large volumes and processed to provide insights and enable data-driven decisions by individuals, businesses, governments and other organizations. Thus, it is the capacity of digital platforms to aggregate, process, transmit, store, analyse and make sense of data that allows them to generate value.” (at 30) The 2019 UNCTAD Report further argues that “[w]hile these data can have a significant potential for value creation, it is not possible ‘ex ante’ to assess this value. The use of the data is not known at the collection stage, especially not by the data producer… It is only after their use that their value becomes certain… Since the economic value emerges with the processed information and knowledge, it is only then that market-like features can be observed. At this point, the data are controlled by the platform owners, who also receive the proceeds of this value.” (Ibid.) In other words, data has value only in the “processed” form which it attains by being subjected to the analysis and packaging by the tools that frequently reside outside of the new nexus jurisdictions.
One can argue, however, that the data would not have attained its “processed” form if not for the raw inputs – which is why the World Economic Forum (WEF), in a report published in January of 2011, labeled personal data a new economic “asset class.”[1] Since then, there have been numerous attempts to put a price tag on individual data, including a personal data calculator launched in June of 2013 by the Financial Times,[2] and a personal data platform called Handshake.[3] Whether one can reliably estimate the value of this new “asset class” relative to the value of other assets that contribute to value creation by a digital enterprise is unclear. Even using the personal data calculator as simple as the one developed by the Financial Times, the range of values can be quite wide. Furthermore, the value of user data will be different across industries and products, will depend on whether the data is volunteered, observed, or inferred, and will change over time. It is quite likely that the consensus of the Inclusive Framework members on the “20-30% of residual profit” share is a shortcut, or a patch, given the dearth of objective market parameters and the lack of consensus between esteemed international bodies (such as WEF and UNCTAD) on the value of building blocks of digital economy.
Yet personal data is only one component of the value of user participation in the digital space. Other components include the value generated by consumption of goods and services, user contribution to the content, and the network effects, but our attempts to understand the complexity of the underlying data are still in their infancy. We should not be surprised to discover that profit drivers in multi-sided platforms will be different from one-sided platforms; that the user value will depend on whether s/he generates content; that critical mass matters; and that traditional approaches to determining customer value will not work in the digital economy. However, we are not yet ready to quantify any of these factors without rigorous academic inquiry. We may not even know the correct questions to ask. How then do we know whether the 20-30% residual profit share is the correct answer?
Amount A represents a political compromise designed to stave-off proliferation of digital services taxes and at the same time provide some basis for allowing countries to tax a portion of residual profit of the largest most profitable multinationals. Only time will tell whether that compromise will hold and whether countries will continue to accept the trade-off in tax revenue. As H.L. Mencken once said, for every complex problem there is an answer that is clear, simple, and wrong.
As a reminder, Mayer Brown will be hosting a free webinar on the OECD agreement and President Biden’s tax proposals on July 15. Sign up here.
[1] http://www3.weforum.org/docs/WEF_ITTC_PersonalDataNewAsset_Report_2011.pdf
[2] https://ig.ft.com/how-much-is-your-personal-data-worth/#axzz2ePHWXB3Khttps://ig.ft.com/how-much-is-your-personal-data-worth/#axzz2ePHWXB3K