Taxing the digital economy: Good reasons for scepticism

EU Monitor

Digital economy and structural change

May 21, 2019

Taxing the digital economy

Good reasons for scepticism

Author

Dieter Bräuninger

[phone] [e-mail]

Editor

Barbara Böttcher

Deutsche Bank AG Deutsche Bank Research Frankfurt am Main Germany

E-mail: [e-mail]

Fax: [phone]www.dbresearch.com

DB Research Management Stefan Schneider

Original in German: May 2, 2019

Digital taxation is currently a subject of intense debate. While France, the UK, Italy, Austria and Spain have already decided to introduce a digital tax, the EU ministers of finance were unable to agree on an EU-widesystem in March, which the European Commission (EC) had hoped to introduce in order to prevent looming competitive distortions and a fragmentation of the single market as Member States went their own way on this issue.

However, all approaches which are based on the taxation of revenues instead of profits have major flaws. If a company is not allowed to deduct its expenses from its revenues, there is a risk that the tax eats up corporate capital. Start-upcompanies with lower margins would suffer most from such a tax.

Since large digital companies are widely thought to pay inappropriately low taxes, policymakers remain under pressure to act. But the tax gap is often exaggerated and governments themselves have created tax breaks for digital companies.

Tax experts are calling for international collaboration on this issue. The OECD's/ G20's BEPS Project aims at an international solution. This initiative focuses on three major tax policy challenges caused by digital business models: 1. the global reach of business functions and activities, 2. opportunities for profit shifting due to importance of intangible assets such as licences and brands, 3. options for taxing the user contribution to digital value creation locally, i.e. in the country of destination, and not in the country where the company has its seat.

The second issue is being addressed by recent modifications in international tax law; the two others, however, require new approaches. So far, no consensus has been reached on any of the proposals on the table. For example, taxing user contributions to value creation raises tricky definition problems. What activities amount to value creation? Where does value creation really take place? How should the related profits be allocated to the different countries?

As digital services expand into ever new areas of the economy ('smart everything'), the risk of a far-reaching, arbitrary taxation of entrepreneurial activities is increasing. This might become a problem, not least for the German industry, which invests considerable sums into digitalisation.

Disruption, the buzzword of the digitalisation discussions, may become an issue in international tax policy, too. Some of the far-reachingnew rules currently under discussion could undermine basic principles of the current corporate taxation system.

In addition to an (international) digital tax, minimum taxes are one of the concepts under discussion. Their supporters point to elements of minimum taxation included in the recent US tax reform. However, it is still uncertain which way the international community will choose, even though a consensus is to be reached by end-2020.A re-allocationof international tax revenues in favour of foreign markets might lead to Germany losing some of its share.

Taxing the digital economy: Good reasons for scepticism

Fundamental principles for the taxation of
multinational companies*	1

The current, fundamental rules for international taxation have been in place for about a hundred years. Put simply, they are as follows:

-Passive income from entrepreneurial activity, i.e. above all dividends and revenue from licence fees, will be taxed in the recipient's country of residence (residence-based taxation).

-Active income, i.e. profits, will be taxed in the country where the company is doing business and generating revenues (source-based taxation).

-In the case of multinational companies, active income may be subject to taxation in all countries where the company has a permanent establishment. A permanent establishment gives a country the right to levy taxes.

-The second, more complex aspect is the (national) tax base, i.e. the share of a company's overall profits which is allocated to a (specific) country where the multinational runs a permanent establishment.

-As a rule, this share shall reflect the economic activities and the value creation of the relevant permanent establishment. However, this is difficult to determine, as multinational companies tend to engage in large-scale internal transactions, such as the exchange of inputs, services or rights.

-The international allocation of the tax base follows a comprehensive set of rules, which is currently being reviewed at the international level (see p. 5 et seq.).

-In practice, rules for the transfer prices (see p. 6-7) which multinationals have to use for the evaluation of internal transactions play a key role. These rules actually primarily determine the allocation of the (potential) tax revenues to the different countries.

*See Devereux, Michael P. and John Vella (2017). Implications of Digitalization for International Corporate Tax Reform. In: International Monetary Fund (ed.) (2017). Digital Revolutions in Public Finance. p. 93

Digital taxes are highly popular in Europe

Digital taxes have been a subject of intense debate in Europe for some time now. Several governments, among them those of France, the UK, Italy, Spain and Austria, have already decided to introduce them. As a rule, they propose to tax certain revenues generated by large (foreign) digital companies doing business with citizens and/or companies from the relevant country at a rate of 2-3% (5% in the case of Austria). The European Commission (EC) cited the national concepts and the related risks for the single market in its call for a uniform digital tax at the EU level. However, the EU ministers of finance were unable to agree on a common concept on 12 March. A tax on (gross) digital revenues is therefore probably off the table. In any case, such a tax was meant as an interim solution until more sweeping rules were established at the European and/or international level (see below).

Failed Commission concept had significant flaws

The digital tax proposal by the EC was to address "the problem that the current corporate tax rules are inadequate for the digital economy". The Commission pointed out that, under international rules, any revenues generated in a given country may only be taxed in that country if the company runs a permanent establishment there (see box). In addition, the Commission wanted to give the Member States the right to tax profits based on national users' contribution to value creation.

The failed proposal for a Directive of March 2018 defines the following as relevant taxable revenues:1

-revenues from placing advertising on a digital interface (i.e. in particular from the sale of online advertising space);

-revenues from making available to users a multi-sided digital interface which allows users to find other users and to interact with them, and which may also facilitate the provision of underlying supplies of goods or services directly between users;

-revenues from the transmission of data collected about users and generated from users' activities on digital interfaces.

According to the proposal, only large companies with worldwide revenues of more than EUR 750 m per year and taxable revenues of more than EUR 50 m p.a. in the EU were to be subject to this tax. The tax rate in the Member States was to amount to 3%.

However, the digital services tax has been criticised right from the beginning.2 Three weaknesses in particular came under fire. First, levying a tax on (gross) revenues instead of profits is problematic. Second, European attempts to introduce a digital tax may weigh on the trade relationship with the US. And third, a digital tax is, in fact, a first step towards a far-reaching revamp of the international tax system.

1European Commission (2018). Proposal for a Council Directive on the common system of a digital services tax on revenues resulting from the provision of certain digital services. Brussels. 21 March 2018.

2Fuest, Clemens et al. (2018). Die Besteuerung der Digitalwirtschaft. Zu den ökonomischen und fiskalischen Auswirkungen der EU-Digitalsteuer. Ifo study commissioned by the Chamber of Industry and Commerce for Munich and Upper Bavaria. Advisory Board to the Federal Ministry of Finance (2018). Response to the EU proposals for taxing the digital economy. For a recent assessment see Olbert, Marcel and Christoph Spengel (2019). Taxation in the Digital Economy - Recent Policy Developments and the Question of Value Creation. ZEW Discussion Paper. No.

19-010.

2 | May 21, 2019

EU Monitor

Taxing the digital economy: Good reasons for scepticism

					Quite rightly, taxing (gross) revenues (minus VAT) instead of profits is regarded
					as a questionable interim solution at best. After all, such a system decouples
					taxes from profits; companies may have to pay taxes in a given country even
					though they do not make any profits there. This problem becomes even more
					serious if companies cannot deduct these digital taxes from the tax debt in their
					home country, for example because they do not make any or only a small profit
					there. In that case, companies would have to draw upon their equity to pay the
					taxes. This is a risk for low-margin companies in particular. A digital tax may
					therefore be harmful to start-up companies during their initial expansion phase.
					The EC has introduced high thresholds for taxation to reduce this risk. However,
					these thresholds lead to another problem: in its initial form, the proposed digital
					services tax appeared to be specifically directed at large US digital companies.
					Thus, the Commission's proposal might have been an additional hurdle for the
					upcoming EU/US negotiations on the cross-Atlantic trade relationship, which will
					be difficult anyway. The US ambassador's criticism of Austria's digital tax
					already points in this direction (FAZ.net, 9 April 2019).
					Third, there are fundamental systematic concerns. The EC plans to tax end
					users' contributions to value creation of digital services at the place where the
					users reside (where their IP address is located). The resultant shift from the
					country-of-establishment principle to the country-of-destination principle is not
					only a major turning point for international tax law, but also harbours significant
					risks. Countries such as Germany, which rely heavily on exports and make
					major digitalisation efforts, might lose in two ways. Their companies might
					become less competitive as they may have to pay new taxes in their export
					markets (higher risk of double taxation), and their domestic corporate tax
					revenues might decline. As the proposed tax is solely targeted at digital
					business models it could create significant definition problems, given the
					digitalisation trend in the economy (see below, p. 11). Moreover, the
					redistribution effects of such a tax are uncertain. Platform or broker services
					providers might pass on (part of) this tax to the end user(s) and/or to sellers.
					Taxation of the digital economy: A controversial issue
Taxes on corporate profits: Relatively				Nevertheless, many stakeholders in Germany are calling for a digital tax, too. Its
steady revenues in the long run		2	supporters point to the gap between the rapid increase in digital revenues and
% of GDP				comparatively low taxes on digital business activities. Of course, this discussion
4					has its reasons. The tax burden on digital companies is often presented in a
				distorted way, and many commentators simply neglect the numerous question
3.5
				marks related to the taxation of revenues instead of profits.
3
2.5					Supporters of a digital services tax are convinced that large, cross-border digital
2					companies are not fairly taxed in numerous countries. The focus is on profits
1.5
				generated in countries where these firms serve numerous customers, but do not
1
				run a permanent establishment. In many countries, this applies above all to
0.5
0					large Chinese and US digital companies. Under the current, international rules
					(see the box on p. 2), such profits are not subject to tax at all in the relevant
					countries. In addition, international digital companies are thought to have
	OECD average	DE	particularly good opportunities to shift profits to low-tax countries, including a
Source: OECD				number of EU Member States.

Several studies appear to support this view. The EC, for example, points out that the effective average tax rate for digital companies averages 9.5% in the EU, i.e. less than half the rate for traditional companies (23.2%; both rates for 2017).3 While it is quite true that digital business models benefit from relatively

3European Commission - Fact Sheet (2018). Questions and Answers on a Fair and Efficient Tax System in the EU for the Digital Single Market.

3 | May 21, 2019

EU Monitor

Taxing the digital economy: Good reasons for scepticism

Effective average tax rates
for digital business models	3
2018, %

30

25

20

15

10

5

0 -5-10-15-20-25-30-35

IT IR LV PT ES FR UK NL PL BE DK AT ES FI SE DE JP US*

* California

Source: ZEW (Centre for European Economic Research), Mannheim and PWC (PricewaterhouseCoopers GmbH)

Calculated lost corporate tax revenue
due to profit shifting	4

Tax losses in ... due to profit shifting to ..., in % of total corp. tax revenues, as of 2015

30

25

20

15

10

5

0

DE FR IT UK ES SE	FI AT PT DK PL US JP OECD
		EU tax havens			Non-EU tax havens

Source: Torslov, T.R. et al. (2018). The Missing Profits of Nations, NBER Working Paper, Online Appendix

low effective taxation in many countries (although not in Germany), this is to some extent due to government measures taken to promote innovation and investment in the digital economy, above all patent boxes. Patent boxes ensure a lower taxation of revenues from R&D activities and intangible assets such as licences and patents. In addition, traditional tangible assets play a relatively small role in the digital sector. Companies from traditional sectors have usually to write off expenses for such assets (machinery and equipment) over several years when they calculate their tax liabilities. In contrast, digital companies can deduct the expenses for the know-how of their employees (i.e. their most important asset) immediately in the form of wage payments.

Other observers point to studies which analyse the absolute amount of tax shifts. The OECD, for example, justified the BEPS Project by saying that tax shifts lead to a loss of 4-10% of total global corporate tax revenues. Based on figures from 2014 (the year in which the study was prepared), this is equivalent to USD 100-240 bn p.a. While a share of one-tenth is quite substantial, the large range of the estimate suggests that the calculations are subject to considerable uncertainties.4 A study by the European Parliamentary Research Service released in autumn 2015 gives even higher figures.5 It puts the equivalent tax losses in the EU at EUR 50-70 bn p.a. or 17-23% of the then corporate tax revenues.

According to a widely noticed, more recent study, multinationals around the world shift c. 40% of their aggregate profits (USD 1.7 tr in 2015) to low-tax countries.6 This phenomenon is particularly evident in countries or regions where tax rates are relatively high in an international comparison. Due to such shifts, the EU loses almost 20% of its (theoretical) corporate tax intake. Within the EU, Germany, France and Italy are suffering most, with Germany losing

28% of its potential tax intake and France and Italy c. 20%, respectively. The US (c. 15%) are also among the major losers. However, in contrast to Germany, the US have recently (2017 tax reform) cut their corporate tax rate considerably and taken additional measures to prevent profit shifts (see p. 15).

To some extent, these high figures are due to the study's methodology. The authors use the same production function for the activities of local subsidiaries of foreign multinational companies and for those of domestic companies. This means that they assume productivity to be identical, provided that the ratio of labour and capital input is the same. However, economies of scale and (a likely) advantage in terms of know-how suggest that the respective foreign subsidiaries are more productive and tend to generate higher profits in low-tax countries. And above all, the authors assume that all movements of intangible assets to countries with below-average tax rates which do not go hand in hand with actual investments and/or the establishment of financing companies in these countries are tax-shift strategies.

To sum up, the differences between the taxation of traditional and digital business models are to some extent due to political support for the latter. Nevertheless, empirical studies confirm that large (foreign) companies (particularly from the digital sector) pay relatively low taxes in high-tax countries in the EU. The importance and urgency of the issue are, however, controversial.

4See also Hentze, Tobias (2019). The challenge of moving to a Common Consolidated Corporate Tax Base in the EU. German Economic Institute, Cologne, IW Report 2/19.

5European Parliamentary Research Service (2015). Bringing transparency, coordination and convergence to corporate tax policies in the European Union. I - Assessment of the magnitude of aggressive corporate tax planning. Study for the European Added Value Unit.

6Torslov, Thomas R. et al. (2018). The Missing Profits of Nations. NBER Working Paper 24701.

4 | May 21, 2019

EU Monitor

Taxing the digital economy: Good reasons for scepticism

Strong growth in cross-border
royalties & licence fees	5

% yoy

10.0

5.0

0.0 -5.0

-10.0

-15.0

2013	2014	2015	2016	2017
		FDI				royalties & licence fees

Source: Unctad

New opportunities for profit erosion and profit shifting?

Institutions and groups which support a digital tax obviously see a need for action. However, any actual steps should be justified by more than just a hunch that tax revenues from the digital sector are inappropriately low. The OECD explains three relevant characteristics of digital business models:7

-Wide reach or importance without physical presence ("scale without mass"). Companies can use the internet and online platforms to create long- distance cross-border relationships with numerous customers and do business abroad without running a permanent establishment in other countries. Under current law, there is no nexus for a taxation of the profits they generate abroad.

-Considerable reliance on intangible assets, including intellectual property (IP). Brand names, the development and use of software (including algorithms which analyse large amounts of data generated by business on internet platforms) and creative (digital) content play a major role for the production or provision of services via the internet. As a result, modern, digital services can be provided at low variable costs. Expenses for royalties or licence fees have risen by an average of almost 5% p.a. during the past five years (2013-2017), whereas global goods trade and direct investment have increased by only just about 1% p.a. (Unctad World Investment Report 2018, p. 22).

-Participation of (end) customers or users in value creation and high value of data. Companies which use digital platforms to interact with their customers can analyse customer behaviour with the help of high-performance algorithms and use these data to increase their revenues, for example by re-selling the data and/or putting customer-specific advertisements on the platform (against a fee).

While these features characterise digital business models, from the OECD's vantage point, they are obviously not limited to such businesses. Therefore, efforts to ring-fence such models or activities to establish a separate base for taxation would be problematic. Still, the supporters of a digital tax draw two conclusions from these ideas. First, they argue that digital business models make it particularly easy for multinationals to shift profits to countries with low tax rates. This behaviour has wider implications: If goods and services which were produced or provided domestically so far are now offered from abroad, a country will not only lose tax revenues, but remaining domestic providers will come under (additional) competitive pressure if the foreign country sets comparatively low tax rates. Second, many digital business models are supposed to be based in fundamentally new value creation processes, which are not taxed adequately (and, above all, not in the proper countries) under the current rules.

The international BEPS initiative for more fairness, coherence and transparency aims to secure national tax revenues

The international measures taken against profit shifting and base erosion within the framework of the BEPS (Base Erosion and Profit Shifting) initiative of the OECD and the G20 are based on these arguments, too. The initiative's main goals are to create a coherent international tax system, to improve transparency, to provide (more) certainty for businesses that do not take aggressive positions, and to secure (national) tax revenues. In general, it aims to tax profits in the country where they are generated. The international

7See OECD (2018). Tax Challenges Arising from Digitalisation - Interim Report 2018, p. 24 et seq.

5 | May 21, 2019

EU Monitor