A second pillar in these discussions deals with the introduction of an effective global minimum tax.

The ICT sector is measured using industry-level national accounts data and the relevant International Standard Industrial Classification codes (G20 2018, United Nations 2008) for computer programming and information services (Div. 62 63); telecommunications (Div. 61); publishing and broadcasting (Div. 58 60); and manufacturing of computer, electronic, and optical products (Div. 26).

Although fully comparable data are not available, McKinsey Global Institute (2019) estimate that India’s ICT sector accounted for around 7 percent of GDP in 2017–18, mainly reflecting IT and digital communications services.

Some countries consider the DSTs discussed in the following section, merely as an interim solution until international agreement is achieved. Indeed, countries such as Belgium, Czech Republic and Hungary have delayed the implementation of their DSTs, anticipating agreement on a multilateral solution can be reached by mid-2021.

The OECD-IF argues that not all the residual is generated by the market jurisdiction (for example, some reflects risk-taking by the MNE), hence only a portion of the residual profit is reallocated.

For economists, routine returns may resemble a normal return on investment and “residual” profits, resemble rents (earnings in excess of the minimum required by the investor).

For instance, the Made in America Tax Plan would bring the international minimum tax provisions (GILTI and BEAT) closer in line with the OECD-IF’s Pillar 2 proposal. The plan envisages a higher minimum tax rate of 21 percent compared to the OECD-IF.

The definition of consumer-facing business is not yet finalized.

However, as noted above using consolidated data, digital firms account for a sizeable share of residual profits, with the ICT sector alone responsible for 16 percent of total residual profits. Hence the analysis presented here should be viewed as illustrative. Indeed, digital firms are deemed to have greater opportunities for profit shifting, given the importance of intangibles in their production process

Although the share of residual profit is evenly shared across the three continents, affiliates located in Asia tend to be more profitable than their counterparts in other regions, earning a return on tangible assets of 17 percent. The median affiliate in Europe earns a 15 percent return, and those in the Americas and Middle East earn a return of 8.5 percent. The lowest returns are observed in Africa, with a return on assets of only 4 percent.

Results are presented in OECD (2020a) as the percentage change in CIT revenue. IMF staff converted this change in CIT revenue into percentage points of GDP for ease of comparison with the other results presented in this paper.

The top 100 largest MNE groups by revenue have revenue greater than USD 67 billion, with an average revenue of USD 127 billion, and an average residual of USD 1.7 billion. The headquarters of these MNE groups are dispersed equally across Asia, Europe, and the Americas. The average size, in terms of revenue, in each region is similar, but the average MNE group headquartered in the Americas have more than twice the residual profit (USD 3.4 billion) of the average Asian MNE (USD 1.6 billion). The average residual profit of European is the lowest at USD 305 million.

For these estimates, residual profit is defined as profit above 10 percent of unrelated party sales. Only 20 percent of this residual is reallocated based on the share of destination sales in each jurisdiction. The estimates assume that this reallocation is “funded” by countries relinquishing the residual profit to which they currently have taxing rights.

High global revenue thresholds result in common DSTs, effectively capturing mainly important US MNEs, which has been argued to create de facto discrimination (see Hufbauer and Lu 2018, and USTR 2021). The effective targeting of US MNEs by DSTs, however, is a consequence of their market dominance and may be temporary (Avi Yonah 2020).

This type of tax is typified by the 2018 European Commission draft directive for the taxation of digital services, versions of which have subsequently been unilaterally introduced by EU members. The proposed DST, which has since served as a model for a number of EU members states, imposes a 3 percent levy on revenues from certain specified digital activities, which users have “co-created.”

In cases where a double tax treaty does not allow for the creation of a virtual PE, an “electronic transaction tax” is supposed to be applied, though this measure is not yet effective/specified.

Rents—that is, earnings in excess of normal required returns—are an attractive tax base because they can be taxed without distorting a company’s behavior. However, taxing rents is difficult in practice, since some costs are hard to observe and provide deductions for and many sources of rent are not location specific; therefore, taxing them risks driving them elsewhere.

ATAF (2020) proposes a hybrid approach that would deal with abuse risks by determining the DST charge as the higher amount between the direct payments made and a country’s share in apportioned global segment revenue of a company in scope (for instance, the share of a country’s advertising views in global total advertising views).

Potentially contributing to further market concentration in the tech sector.

In 2018 the EU Commission proposed a DST of 3 percent on the gross revenue from activities relying on user participation such as selling online advertising space and intermediary activities that allow users to interact and sell goods and services.

And the potential Asian user base and revenue is much larger, as Facebook is not (yet) operating in China.

The Statista surveys cover market participants as well as consumers—for India, Indonesia, Thailand, and Vietnam. Reported results for the other economies are extrapolated using indicators of purchasing power of consumers and digital maturity in the economy. E-commerce includes sale physical goods via a digital channel (from all types of devices) to a private end user (B2C) with cross-border purchases attributed to the country of the buyer. E-services capture sales of services and digital goods (event ticket reservation, dating, food delivery, etc.) with an online checkout process. Digital media captures spending on audiovisual media contents and applications distributed online. Digital advertising captures advertisement spending for online channels. (Statista 2020).

See https://www.avalara.com/in/en/blog/2020/09/india-digital-tax-or-equalisation-levy-a-timeline.html.

Assuming that 40 percent of the e-commerce, 70 percent of digital media, 100 percent of digital advertising, 10 percent of e-services, online tourism and online mobility revenues estimated by Statista are in scope and applying the rates of the Indian DST (2 percent on gross revenue in scope, 6 percent on gross advertising revenue).

The European Commission has estimated annual revenue yield from its DST for member states of EUR 5 billion (<0.01 percent of EU GDP). France expects to collect EUR 400 million from the DST in 2020 (0.02 percent of GDP). The United Kingdom estimates that its DST will raise GBP 275 million (0.01 percent of GDP) in 2020–21 rising to GBP 440 million (0.02 percent of GDP) in 2023–24. USTR (2021) finds an equally moderate tax revenue potential when assessing tax liabilities for US firms as a result from the Indian DST: in aggregate, the 86 US firms that are likely in scope of the Indian DST “may face tax payments in excess of USD 30 million per year.”

By January 2021, more than 30 countries had enacted, held public consultations on policy proposals, or announced their intention to introduce unilateral direct tax measures aimed at digital services, see KPMG (2021).

For instance, Malaysia expanded the scope of its tourism tax charged per room per night in 2020 to cover accommodation booked through online platforms and will impose the tax on resident and nonresident digital platform service providers from July 2021.

Similar to tariffs announced following an investigation of France’s DST. These were announced in July 2020, scheduled to go into effect in January 2021 (but have been temporarily suspended).

The African Tax Administration Forum (ATAF) for instance proposed model legislation for Digital Services Taxes to its member states.

Other policy alternatives, such as a destination-based cash-flow tax, are discussed in IMF (2019) and Auer-bach and others (2017).

These alternatives, initially proposed by tax practitioners and academics, have gained traction with policy makers (for example, the G24’s proposal for fractional apportionment), outlined in the Intergovernmental Group of 24’s submission to the OECD on possible solutions to the tax challenges of digitalization. India was the first to propose such a solution, which was then endorsed by other members.

FA is commonly used to distribute CIT revenue at the subnational level, for instance, in countries such as Canada, Germany, Japan and the United States. The European Commission has proposed such an approach for the EU, in the form of the ‘Common Consolidated Tax Base’, wherein the weights are a composite of sales, assets, employment, payroll, and data.

See de Mooij and others (2019) for a detailed discussion of the benefits and drawbacks of formulary apportionment.

FA regimes generally have special treatment only for the extractives and financial sectors.

A discussion of negative residuals can be found in Beer and others (2020).

The following analysis is based on Beer and others (2020).

Routine profits correspond broadly to a normal return on investment. Commonly, a return to tangible assets is chosen at a level which is sufficiently high to encompass a return to all assets (including intangibles).

There could be an incentive to sell to a third-party distributor located in a low-cost jurisdiction. Sourcing rules could be introduced to mitigate this new form of tax avoidance.

As provided for in the OECD (2017) VAT guidelines. A major reform along these lines was introduced in the EU in 2015 to ensure taxation at destination on intra-EU supplies, which previously was based on origin (the location of the supplier). See Dale and Vincent (2017).

Requiring the individual consumers to register and fulfill the necessary steps to remit VAT on a one-of purchase on the internet is challenging (see Box 4).

Without an agent or bill of entry that would help identify VAT and customs duty payments due and the responsible taxpayer.

Levelling the playing field is preferable to pursuing this objective with new tax incentives for digital startups. In Vietnam, for instance, in 2020 incentives aimed at innovative startups were introduced, including a reduction of the corporate tax rate to 10 percent for 30 years (Decree No.94/2020/ND-CP).

The customer’s location is commonly determined by a combination of the information on the customer’s payment profile (credit card information, bank account details), residence (billing address or home address), and their internet access (the internet protocol of the device used or the country code of their SIM card [if transaction made through a mobile device], location of the consumer’s fixed landline). Most countries require two pieces of nonconficting information to make this determination.

The EU broadly defines the nature of in-scope digital or electronically supplied services to be covered under new VAT legislation. The EU VAT Directive, specifies electronically supplied services to mean “services delivered over the internet or an electronic network and the nature of which renders their supply essentially automated and involving minimum human intervention, and impossible to ensure in the absence of information technology.” A list of examples is provided in the accompanying VAT regulation, along with selected exclusions.

In some cases, exemptions are explicitly carved out for example, online gaming (Norway), professional services provided over email (EU), and the provision of education and health services through digital means, areas that are typically zero-rated in domestic legislation in these countries.

Other examples outside the region include Argentina and Costa Rica.

The Indonesian Directorate General of Taxation issued the first list in July 2020. This included companies in the spotlight of the digital tax debate, namely Amazon, Google, Netfix, and Spotify. Four subsequent lists were published between July and November, bringing the total number of in scope businesses to 46.

There are some exceptions. In Argentina, Azerbaijan, and Bangladesh, for example, the liability falls not on the non-resident supplier but on the local payment provider. Where that is the case, specification of transactions for which payment providers need to withhold VAT is required, likely narrowing the scope of covered supplies in practice. While this approach can thereby help address compliance management challenges, it is unlikely to be a desirable long-term solution in most countries. A hybrid approach is pursued in Costa Rica, where the government does allow voluntary compliance by companies on the in-scope list, but if they do not comply, the payment is withheld by card issuers (for example, Visa, MasterCard).

Tresholds for cross-border digital services for VAT registration are typically at the same level or below the domestic requirement for mandatory registration. Some countries do not include a threshold, even if one exists in domestic legislation (for example, Moldova), or a modified lower threshold may apply. As part of its general rules, Indonesia provides for a second threshold related to customer traffic levels, which is used in conjunction with the monetary threshold—the threshold is reached if the amount of traffic or access in Indonesia exceeds 12,000 users annually or 1,000 users monthly. As discussed above, currently, however, registration requirements are in practice limited to companies directly referenced in regularly published official lists.

No input credits can be claimed by foreign registrants in the simplified registration process. Consequently, abuse risks are less of a concern than with regular registrants. If a foreign service provider wants to claim input tax credits for supplies made in another country, the usual requirement is to establish a place of business in the country and to go through the regular registration process.

From July (January) 2021, the EU (and United Kingdom), require platforms to withhold and remit VAT on low-value parcels on behalf of sellers.

Australia introduced 10 percent GST on cross-border sales of services and digital products imported by Australian consumers on July 1, 2017, following enactment of Tax and Superannuation Laws Amendment Bill 2016. The budget initially estimated revenue collection of A$150 million during the first year (FY2017–18), followed by A$200 million in FY2018–19.

https://www.avalara.com/vatlive/en/vat-news/thailand-vat-on-digital-services-near.html.

Applying the standard VAT rate and assuming that 100 percent of transactions of digital media content, 10 percent of all e-commerce transactions, 5 percent of digital advertising, and 15 percent of e-services, mobility and travel services captured by Statista are provided by unregistered remote suppliers to final consumers and/or unregistered registered entities. For details on the Statista survey see also discussion in section 2 on Digital Services Taxes.

In Croatia, for instance, a compliance management campaign launched in 2018 drew on a comparison of domestic tax returns with digital platform data regarding hotel and lodging accommodation sold on behalf of Croatian suppliers. About 40 percent of Croatian vendors using the platforms covered in the campaign either did not register or declared significantly less income for tax purposes than they received from platform-facilitated sales (World Bank 2021).

Recent guidance provided by OECD (2020b) on model reporting rules for platform operators provide a useful reference framework. See also OECD (2019) on the different approaches to leveraging the prominent role of digital platforms for the collection of VAT.

Fall Economic Statement, 2020, Annex 4 https://www.budget.gc.ca/fes-eea/2020/report-rapport/anx4-en.html.

Trough the use of P2P platforms, increasingly efficient small businesses may better compete with and displace larger incumbents. If the rise of the sharing economy means that incumbents are being displaced and replacement P2P activity remains below tax thresholds, governments risk losing revenue as income and profits are dispersed across many smaller businesses instead of concentrated in large profitable companies (Aslam and Shah 2017).