The Long Reach of European Union Law: Patent Boxes and the Limits of International Cooperation, LV Faulhaber

Tags: supra, European Union, BEPS Project, jurisdiction, EU Member States, EU Law, OECD, Member States, Member State, International Cooperation, main version, boxes, non-EU countries, tax incentives, ECJ, expenditures, income, European Court of Justice, approach, tax avoidance, benefits, direct tax, IP assets, Organisation for Economic Co-operation and Development, international tax avoidance, gross income, domestic subsidiaries, royalty income, Corporate Tax Reform, international tax, EU institutions, jurisprudence, Intellectual Property Box Regimes, Bernard Knight, Intellectual Property Tax, regime, corporate tax avoidance
Content: DRAFT ­ Please do not cite or distribute without permission THE LONG REACH OF European Union LAW: PATENT BOXES AND THE LIMITS OF INTERNATIONAL COOPERATION Lilian V. Faulhaber* From 2013 to 2015, the forty-four member countries of the Organisation for Economic Co-operation and Development (OECD) and the G-20 worked together on the Base Erosion and Profit Shifting (BEPS) Project, which was focused on curbing corporate tax avoidance. This article argues that the outcomes from that project show that European Union law is no longer just a concern of EU Member States. Instead, EU law and the jurisprudence of the European Court of Justice (ECJ) place constraints on non-EU countries by creating downward pressure on International Standards. This article illustrates the effect of EU law on non-EU countries by introducing readers to the BEPS Project's work on patent boxes, which provide reduced rates for income from patents and other IP assets. Well over a dozen countries currently have patent boxes or similar tax incentives, and several U.S. versions have been proposed in the past five years. Countries without patent boxes argue, however, that these tax incentives represent harmful tax competition. In response to this concern, the OECD and G-20 countries developed an approach called the "nexus approach," which was designed to limit the benefits that patent boxes could provide. This article introduces the nexus approach and argues that the two versions of this approach illustrate the degree to which EU law limited the ability of countries to design patent boxes that would eliminate tax avoidance opportunities. The jurisprudence of the ECJ has put downward pressure on international anti-avoidance standards, thereby making it harder for non-EU countries to have robust anti-avoidance rules. This article argues that the ECJ has the ability to create this pressure in the direct tax area ­ and in any other area where the EU does not have independent authority but the ECJ does have jurisdiction. One lesson to be learned from the nexus approach and other recommendations from the BEPS Project is therefore that academics, * Associate Professor of Law, Georgetown University Law Center. From September 2013 through August 2015, I was an Advisor to the Base Erosion and Profit Shifting Project at the Organisation for Economic Co-operation and Development. The views and opinions expressed in this article are mine alone and do not reflect the official policy or position of the OECD. All discussions and arguments in this article are based on publicly available information. I am thankful to Ron Blanc, Jake Brooks, Steve Cohen, Itai Grinberg, Charles Gustafson, Eloise Pasachoff, Katie Pratt, Ted Seto, and participants in the tax colloquium at Loyola Law School, Los Angeles for helpful conversations and comments on earlier drafts. Charles Bjork, Mabel Shaw, and Mikaela Harris provided excellent research assistance.
2
The Long Reach of EU Law
[22-Mar-16
practitioners, and negotiators alike must all be aware of the negative effect that the ECJ's jurisprudence has on regulatory convergence. While discussions of the impact of EU law within the European Union and on EU Member States are important and relevant, the conversation also needs to focus on the impact of EU law outside the European Union and acknowledge that the European Court of Justice is making decisions that limit the ability of the United States and other non-EU countries to set international standards higher than the level permitted by the ECJ. In the direct tax context, this means that countries outside the European Union have lost the ability to police and prevent international tax avoidance to the degree that would have been possible in the absence of the ECJ's direct tax jurisprudence.
22-Mar-16] Patent Boxes and International Cooperation
3
Table of Contents Introduction ................................................................................................ 4 I. Patent Boxes and the Nexus Approach .............................................. 7 A. An Introduction to the Nexus Approach ..................................... 18 B. The Impact of the Nexus Approach ............................................. 25 i. The effects of the nexus approach as a whole: an increase in patent boxes and a decrease in design flexibility ................. 26 ii. The effect of grandfathering provisions: a short-term increase in patent boxes ..................................................................... 29 iii. The effect of tracking and tracing requirements: a long-term decrease in patent boxes ...................................................... 30 iv. The effect of the entity focus: a weakening of the restrictions of the nexus approach .............................................................. 30 v. The overall effect of the nexus approach ............................. 33 II. The Nexus Approach and EU Law ....................................................... 34 III. EU Law and International Cooperation .............................................. 42 A. The Long Reach of EU Law ....................................................... 42 B. What Can Be Done? ................................................................... 46 Conclusion ................................................................................................ 49
4
The Long Reach of EU Law
[22-Mar-16
INTRODUCTION
For decades, countries have answered turned to their tax systems to encourage innovation and entrepreneurship. They have provided tax credits and tax deductions for research and development expenses (R&D), accelerated depreciation for assets used in the R&D process, and reduced wage taxes for the employees engaged in this process. The goal of these tax incentives is to create an incentive for R&D that will eventually create valuable intellectual property (IP) in the form of patents, copyrights, and other assets that will ultimately produce income. In the last several years, more and more countries have also started to provide reduced tax rates on the income arising from IP assets as another way to create incentives for R&D. These reduced rates on IP income take the form of tax incentives referred to as "patent boxes," where income from patents and other IP assets is separated from a taxpayer's overall income and subjected to lower rates. In the last decade, well over a dozen countries have implemented patent boxes and similar tax incentives. Countries that do not have patent boxes, however, see them as examples of harmful tax competition since at least some patent boxes can be used to encourage income shifting by attracting income away from the country where the underlying R&D took place. Countries without patent boxes thus fear that taxpayers who had engaged in R&D in their countries (and benefited from their R&D incentives) could move the income from that R&D to another country, where it would be taxed at a lower rate. In 2013, this criticism of patent boxes overlapped with international criticism of tax avoidance more generally. As newspapers printed stories about large multinationals avoiding taxation1 and domestic legislatures hosted hearings about the same topic,2 the Organisation for Economic Cooperation and Development (OECD) and the G-20 implemented the Base Erosion and Profit Shifting (BEPS) Project, which was a two-year project meant to target international tax avoidance. As part of this project, the OECD was charged with developing an approach to limit the harmful elements of patent boxes. I worked on this project for two years, and I was part of the team that developed the approach that came to be known as the nexus approach. Under the nexus approach, countries are only permitted to provide
1
See, e.g., Vanessa Barford and Gerry Holt, Google, Amazon, Starbucks: The rise of
"tax shaming," BBC New Magazine (May 21, 2013); Charles Duhigg and David
Kocieniewski, How Apple Sidesteps Billions in Taxes, NY Times (April 28, 2012).
2
See, e.g.,Testimony of Apple Inc. before the Permanent Subcommittee on
Investigations, U.S. Senate (May 21, 2013), available at
https://www.apple.com/pr/pdf/Apple_Testimony_to_PSI.pdf; Starbucks, Google and
Amazon grilled over tax avoidance, BBC News (November 12, 2012).
22-Mar-16] Patent Boxes and International Cooperation
5
benefits under patent boxes if those benefits are proportionate to the amount of R&D undertaken by the taxpayer receiving benefits or in the country providing benefits. The nexus approach limits the harmful aspects of patent boxes by establishing a link between R&D and the income that may benefit, therefore constraining the ability of taxpayers to shift income between countries. It does not, however, eliminate all opportunities for income shifting, and it achieves its goal by creating pressures in favor of restructuring and against outsourcing and acquisitions, even when outsourcing and acquisition would not create any income shifting opportunities. And these weaknesses can be explained by one phenomenon: European Union law. Twenty-one of the forty-four countries that took part in the BEPS Project were Member States of the European Union. Therefore, the OECD could not issue any requirements or recommendations under the BEPS Project that would be inconsistent with European Union law.3 Since the European Court of Justice (ECJ) has previously held that domestic R&D credits cannot discriminate based on the location of the R&D, the nexus approach could not take the most logical approach, which would have been to establish a nexus between the location of the R&D and the country providing tax benefits to the income. Instead, it had to create a less intuitive nexus, between the entity incurring R&D expenditures and the entity receiving benefits. This version of the nexus approach will still reduce income shifting, but, because of the constraints imposed by European Union law, it will create more distortions and more possibilities for income shifting than a version that focused directly on the location of the income. These distortions and income shifting opportunities matter for the effectiveness of patent boxes and for international tax competition more generally. But they also illustrate an important and previously unrecognized phenomenon. Previously, discussions of EU law and the ECJ's jurisprudence have focused on the internal inconsistency of the cases or the impact of these cases on Member States.4 This article argues that these discussions need to also focus on the impact of EU law on countries outside the European Union as well. As shown by the nexus approach and other recommendations issued as part of the BEPS Project, EU law is no longer just a constraint on Member States. It is now a constraint on other countries as well, since non-EU
3
The terms "European Union law" and "EU law" are shorthand for the acquis of the
European Union institutions, as well as the Treaty freedoms, cases by the ECJ and the other
courts of the Court of Justice of the European Union interpreting and applying these
freedoms, and relevant directives and regulations.
4
See, e.g., Ruth Mason & Michael S. Knoll, What Is Tax Discrimination?, 121 YALE
L.J. 1014 (2012); Michael J. Graetz & Alvin C. Warren, Jr., Income Tax Discrimination and
the Political and Economic Integration of Europe, 115 YALE L.J. 1186 (2006);
6
The Long Reach of EU Law
[22-Mar-16
countries are now competing in an international tax environment where the Member States cannot police tax avoidance in the most effective manner. Instead, the ECJ's jurisprudence, which has limited the ability of Member States to police tax avoidance, has also put downward pressure on international anti-avoidance standards, thereby making it harder for non-EU countries to have robust anti-avoidance rules. In a nod to Anu Bradford's "Brussels Effect," which describes how the EU has heightened regulatory standards in certain areas due to the combination of EU regulatory authority and market competition, I refer to this downward pressure on worldwide anti-avoidance standards as the Luxembourg effect. In areas such as direct taxation, where the EU has no independent regulatory authority but the ECJ can strike down Member State provisions for violating the freedoms enshrined in the Treaty on the Functioning of the European Union (TFEU), the ECJ's jurisprudence has created a vacuum. Member States cannot pass laws or regulations that violated the Treaty freedoms, but EU institutions also cannot fill this void by passing EU-wide laws or regulations. What the BEPS Project and its outputs reveal is that this vacuum has effects outside of the European Union. Even though the United States and the majority of OECD members were not Member States of the EU, their efforts to combat tax avoidance through the BEPS Project were constrained by EU law, and they now face an international tax environment where it will be difficult to pass anti-avoidance legislation that is more robust than the extremely low standard permitted by the ECJ. This in turn raises significant concerns about the future of international cooperation. If EU law places limits on what non-EU countries can agree to in international negotiations, that is important for those countries to know as they enter into future negotiations or large-scale projects such as the BEPS Project. In order to illustrate the long reach of EU law, this article proceeds in three parts. Part I introduces patent boxes and the nexus approach. This is the first in-depth description of the nexus approach in the literature, and it highlights the fact that the nexus approach in fact has two versions: the "main version," which is subject to the constraints imposed by EU law, and the "footnote version," which is hidden in the footnotes of the report describing the approach and which is not subject to those same constraints. Part I also assesses the likely impact of the nexus approach on the number and design of patent boxes in the future, paying particular attention to elements that have previously not been discussed in the academic literature, including the grandfathering and tracking and tracing provisions. Part II argues that EU law explains why the main version of the nexus approach creates various distortions and income shifting opportunities relative to the footnote version. This Part also argues that EU law imposed limits on the other outputs of the
22-Mar-16] Patent Boxes and International Cooperation
7
BEPS Project. Part III argues that the nexus approach and the other BEPS Project outputs show that the effect of EU law is no longer limited to Member State laws. This Part introduces the Luxembourg effect and argues that EU law is limiting the ability of non-EU countries to police tax avoidance. While this is important in the context of tax avoidance, it also means that the ECJ has the ability to create downward pressure on international standards in the direct tax area ­ and in any other area where the EU does not have independent authority but the ECJ does have jurisdiction. Part III considers several possible solutions to this situation, but, since none of them appear likely in the short term, it concludes that the first step toward addressing this effect of EU law is to acknowledge it. Discussions of the effect of EU law must focus on its effect outside the EU as well as inside the EU, and negotiators, lawyers, and academics outside the EU must realize that the ECJ is changing the legal environment for everyone, not just the Member States of the European Union.
I.
PATENT BOXES AND THE NEXUS APPROACH
Studies have shown that R&D leads to greater economic growth,5 but private parties underfund R&D because they do not necessarily benefit from all the positive spillovers associated with innovation.6 In response to this market failure, countries have stepped in to use their tax systems to create incentives for research and development (R&D). Countries and states have encouraged R&D by granting credits, deductions, and super-deductions for R&D expenditures, accelerated depreciation for machinery and other assets used as part of R&D projects, and reduced wage taxes for employees engaged in R&D, among other incentives.7 These tax incentives all apply to the inputs to innovation, since they provide benefits at the time R&D is undertaken by providing credits or deductions based on the amount of R&D expenditures.8 In recent years, over a dozen jurisdictions have also implemented tax
5
See, e.g., Michael J. Graetz & Rachael Doud, Technological Innovation,
International Competition, and the Challenges of International Income Taxation, 113
Colum. L. Rev. 347, 348 (2013) (stating that the "importance of technological development
to economic growth has been accepted ever since" Robert Solow's 1957 paper).
6
Graetz & Doud, supra note 5, at 349-50; European Commission, A Study on R&D
Tax Incentives: Final Report, TAXUD/2013/DE/315 at 18 (November 28, 2014) (hereinafter
"EU Commission 2014") ("markets left on their own will probably generate less innovation
than would be desirable from society's point of view. The reason is that knowledge is not
completely excludable [and] investments in innovation are more risky").
7
For a sense of the scale of tax incentives for R&D, see EU Commission 2014, supra
note 6 (finding that the 33 countries studied in that article had a total of 80 R&D incentives).
8
For a more detailed description of the variety of R&D incentives that are available,
see, e.g., EU Commission 2014, supra note 6.
8
The Long Reach of EU Law
[22-Mar-16
incentives that provide benefits to the outputs from innovation. In other words, rather than providing benefits at the time that a taxpayer engages in R&D, these tax incentives provide reduced rates to income arising from the IP assets that resulted from that R&D. The primary example of an output-based incentive is a patent box, which taxes income from patents (and sometimes other IP assets) at a reduced rate. These tax incentives are sometimes also called innovation boxes, knowledge development boxes, IP regimes, or the like, but this article uses the term "patent box" to refer to all regimes that provide a reduced rate to income from IP assets. Patent boxes first originated in Europe,9 and their name is generally thought to refer to the box that taxpayers need to check off on a tax return in order to benefit from the reduced rate. Variations of these tax incentives are now in many non-European countries, including Colombia,10 China,11 Israel,12 and Turkey.13 The design of these incentives varies, with some only applying to income from patents14 and others extending benefits to income from copyrights, trademarks, brands, knowhow, and other forms of intellectual property.15 They also vary in terms of
9
France and Ireland first introduced incentives for patent income in the 1970s and
1980s, but the trend of providing benefits for patent income did not take off for several more
decades. In the interim, Ireland eliminated this incentive. See Graetz & Doud, supra note 5,
at 352.
10
See Andrea Prieto, Colombia Decree 121/2014: National Council of Tax Benefits
in Science, Technology and Innovation ­ regulations issued, Tax News Service (Sept. 11,
2014) (describing benefits provided to "new medical products and the software made in
Colombia"). The Colombian regime is noticeably different from most other patent boxes in
that it focuses entirely on income from software, but it is an output incentive that applies to
income from IP assets (i.e., software), and the OECD listed it in its lists of potentially harmful
IP regimes subject to the nexus approach. See Organisation for Economic Co-operation and
Development, Action 5: 2015 Final Report, Countering Harmful Tax Practices More
Effectively, Taking into Account Transparency and Substance, (hereinafter "Action 5 Final
Report"), 63.
11
See Shi Qi Ma, Corporate Taxation ­ China (People's Rep.), IBFD (2015), 30
(describing the benefits provided for "high-new technology enterprise[s]").
12
See Ministry of Finance, Opportunity Israel (Fall 2012), 27-28 (describing the
regime provided to "preferred enterprises").
13
See Republic of Turkey Prime Ministry, Investment Support and Promotion
Agency, Investors' Guide, Special Investment Zones, available at
http://www.invest.gov.tr/en-
US/investmentguide/investorsguide/Pages/SpecialInvestmentZones.aspx (describing the
advantages provided to Technology Development Zones).
14
Belgium and the United Kingdom are countries that limit benefits to income from
patents and extensions of patents. Marc De Mil & Tom Wallyn, 100A IFA Cahiers (Tax
Incentives on R&D) 145,158-59 (2015); UK Corporation Tax Act 2010 Part 8A (last
amended by Finance Act 2015 (Mar. 26, 2015)).
15
See, e.g., Netherlands Corporate Income Tax Act Art. 12b (last amended by Law
No. 35144 (Dec. 30, 2013)) (providing benefits to patented assets as well as intangible assets
22-Mar-16] Patent Boxes and International Cooperation
9
what type of income can qualify. Some only permit royalties from the sale or licensing of IP assets to qualify,16 while others allow so-called "embedded royalties" to qualify for benefits, which means that the reduced rate will apply to a portion of all the sales income from a good or service that was developed using the IP asset.17 Depending on the scope of the IP assets and the income that can qualify, therefore, patent boxes can provide benefits to taxpayers in a wide variety of industries ranging from pharmaceuticals and software to fashion design and car manufacturing. By 2013, patent boxes were in the news as more and more countries adopted them.18 Although these regimes had briefly been introduced in the 1970s,19 it was not until the Netherlands adopted its innovation box in 2007 and the United Kingdom implemented its patent box in 2013 that commentators focused in on the costs and benefits of these tax regimes. Critics of these tax incentives argued that they were poorly targeted,20 that
arising from R&D that received an R&D certificate under the Dutch Income Tax and Social
insurance premium Relief Act); Spain Corporate Income Tax Act ch. 4 sec. 35 (last amended
by Law No. 1/2014 Feb. 28, 2014) (providing benefits to patents, designs or models, plans,
formulas or secret procedures, and rights on information concerning industrial, commercial
or scientific know-how). Only a few, however, go so far as to extend to trademarks and
marketing-related IP assets. Jurisdictions with innovation boxes that extend to marketing-
related intangibles includes Hungary and Luxembourg. Borbбla Kolozs & Annamбria
Koszegi, 100A IFA Cahiers (Tax Incentives on R&D) 365, 375 (2015); Frank van Kuijk, The
Luxembourg IP Tax Regime, 39 Intertax 140, 141 (Mar. 2011).
16
Countries that apply this limit include Hungary. Kolozs & Koszegi, supra note 16,
at 376. France also does not permit certain types of embedded IP income to qualify for
benefits. Georges Cavalier & Jean-Luc Pierre, 100A IFA Cahiers (Tax Incentives for R&D)
303, 312 (2015) (noting that embedded royalties earned by companies that exploit their own
IP may not qualify for benefits).
17
Countries that provide benefits for embedded IP income include Belgium, Israel,
the Netherlands, Turkey, and the United Kingdom. Eric Warson & Ruth Claes, The Belgian
Patent Income Deduction, 50 Eur. Tax'n 319, 322-23 (July 2010); Leon Harris, Government
Upgrades Company Tax Breaks, 61 Tax Notes Int'l 564, 565 (Feb. 21, 2011); Margreet
Nijhof & Michiel Kloes, An Improved Tax Regime for Intangibles in the Netherlands, 58
Tax Notes Int'l 69, 69 (Apr. 5, 2010); Bilur Yalti, Turkey ­ Corporate Taxation (IBFD 2015)
at 16; UK Corporation Tax Act 2010 Part 8A (last amended by Finance Act 2015 (Mar. 26,
2015)).
18
See Annika Breidthardt, Germany calls on EU to ban "patent box" tax breaks,
Reuters UK (July 9, 2013); Vanessa Houlder & Quentin Peel, UK under pressure from Berlin
over tax competition, Financial Times (June 13, 2013).
19
See supra note 9.
20
This argument focuses on the fact that the benefits of patent boxes apply to income
from patents, which in turn means that the benefits are only granted to taxpayers that already
have income-generating patents. They therefore apply after the decision to research was
made, and their application only to successful innovation means that they likely do not
provide benefits to many of the innovators producing positive societal spillovers, particularly
since the type of R&D that is most likely to be underfunded and that is also most likely to
create positive spillovers appears to be basic R&D, much of which may not lead directly to
10
The Long Reach of EU Law
[22-Mar-16
they did not increase R&D sufficiently to offset their significant cost,21 and that they encouraged income shifting and base stripping.22 Advocates argued that they were necessary to maintain a jurisdiction's competitiveness and keep R&D in the jurisdiction in the face of the growing number of patent boxes.23 Despite these arguments, empirical literature on the effectiveness and cost efficiency of patent boxes is quite limited.24 To the extent that there are empirical findings in this area, studies suggest that patent boxes that
income production. See, e.g., EU Commission 2014, supra note 6, at 6 (stating that IP rights
"enable firms to capture a large part of the societal benefits, such that the need for a tax
incentive for protected innovations becomes unclear"); EU Commission 2014, supra note 6,
at 45 (stating that "by subsidizing inventions that do not need a subsidy, patent boxes would
induce inventions that are difficult to patent (and therefore might have high spillovers)
relatively less attractive"); Martin A. Sullivan, A History Lesson for A Future Patent Box,
Tax Notes (Sept. 7, 2015), 1036, 1038 (hereinafter "Sullivan Sept. 2015") ("There is no
readily apparent economic justification for granting patented technologies more favorable
tax treatment than other IP (and in fact, some would argue that there is less of a reason since
this property already enjoys government-favored status)"); Martin A. Sullivan, Economic
Analysis: Patent Boxes, Research Credits, or Lower Rates?, Tax Notes (June 1, 2015)
(hereinafter "Sullivan June 2015") (stating that a patent box "wastes tax benefits on income
from prior research that is now manifesting itself in current income ­ a windfall for prior
work that provides no incentive for new effort").
21
The empirical literature on patent boxes does not conclusively show whether patent
boxes increase R&D. Although one 2015 paper found that, "for each percentage point
reduction in the [corporate income tax] rate thanks to the patent box, the likelihood of
registering a patent in the country concerned will rise" significantly across industries, a
further finding of that same study, was that, unless they require that development take place
in the jurisdiction, patent boxes decrease the probability of inventors moving to the
jurisdiction offering the patent box. Annette Alstadsaeter, Salvador Barrios, Gaetan
Nicodeme, Agnieszka Maria Skonieczna, and Antonio Vezzani, IPTS Working Papers on
Corporate R&D and Innovation ­ No 6/2015: Patent Boxes Design, Patents Location and
Local R&D, 12. Other studies, however, have suggested that patent boxes, even if they do
not increase overall revenue, do attract IP income to the jurisdiction. Rachel Griffith, Helen
Miller, & Martin O'Connell, Ownership of intellectual property and corporate taxation, 112
J. Pub. Econ. 12, 22 (2014).
22
See Breidthardt, supra note 18; Houlder & Peel, supra note 18.
23
See, e.g., Boustany and Neal Release Patent box Discussion Draft, July 29, 2015,
available at http://boustany.house.gov/114th-congress/boustany-neal-release-innovation-
box-discussion-draft/ (hereinafter "Boustany Neal Press Release") (quoting the legislators
that proposed the U.S. patent box as wanting to "begin the conversation on how the United
States can attract and retain the brightest minds and best ideas on Earth" and "attract
innovation and the high-paying, high-quality jobs that come with it"); Government of the
United Kingdom, Corporate Tax Reform, 51 (November 29, 2010) (explaining the proposed
patent box by stating that "[t]he Patent box will aim to reward successful technical
innovation. The Government believes that it is right to introduce this reform now in order to
prevent movement of IP offshore and encourage the development of new patents by UK
businesses, protecting and enhancing the status of the UK as a world leader in this field.").
24
Graetz & Doud, supra note 5, at 375 (concluding in 2013 that "the extant data is
too limited to adequately assess the effectiveness of patent boxes").
22-Mar-16] Patent Boxes and International Cooperation
11
reduce the corporate income tax rate will lead to an increased likelihood that patents (particularly high-quality patents25) will be registered in the country providing the patent box,26 an increased amount of IP income in that jurisdiction,27 and a reduction in that jurisdiction's tax revenue.28 Studies have not yet been able to determine conclusively whether patent boxes lead to an increase in overall R&D, although at least one recent study suggests that in-country R&D increases if a patent box requires that R&D be undertaken in the jurisdiction.29 Even without empirical data showing the actual effect of patent boxes on the overall amount of R&D, countries continue to implement them, and taxpayers have increased their demands for these regimes. By 2013, over a dozen patent boxes and similar tax incentives had been implemented in OECD and G-20 members.30 Several more existed in the European Union and European Economic Area, with Cyprus, Liechtenstein, and Malta all implementing patent boxes with tax rates ranging from 0% to 2.5%.31 Although the provisions that existed in 2013 all provided reduced rates to income from IP, they varied significantly in the benefits that they provided. The tax rates applied to IP income by existing patent boxes ranged from 0% to 19%.32 As mentioned above, the IP assets that could benefit ranged from
25
Alstadsaeter et al., supra note 21, at 15.
26
Alstadsaeter et al., supra note 21, at 12.
27
Griffith et al., supra note 21.
28
Griffith et al., supra note 21.
29
Alstadsaeter et al., supra note 21, at 19.
30
See Organisation for Economic Co-operation and Development, Action 5: 2014
Deliverable, Countering Harmful Tax Practices More Effectively, Taking into Account
Transparency and Substance (hereinafter "Action 5 Progress Report") (listing fifteen IP
regimes that had been identified by 2014).
31
See European Parliament, Directorate-General for Internal Policies, Intellectual
Property Box Regimes (2015), at 7.
32
This range applies to the national-level IP regimes that are listed in Action 5 Final
Report, supra note 10, and that were in effect at the time of publication. The rate that applies
under the Belgian patent Income Deduction is 6.8%. Warson & Claes, supra note 17, at 319;
Peter R. Merrill, James R. Shanahan Jr., Josй Elнas Tomй Gуmez, Guillaume Glon, Paul
Grocott, Auke Lamers, Diarmuid MacDougal, Alina Macovei, Rйmi Montredon, Thierry
Vanwelkenhuyzen, Alexandru Cernat, Stephen Merriman, Rachel Moore, Gregg Muresan,
Pieter Van Den Berghe, and Andrea Linczer, Is It Time for the U.S. to Consider the Patent
Box?, 134 Tax Notes 1665, 1666 (March 26, 2012). The rate under China's Reduced Rate
for New & High Tech Enterprises ranges from 0% to 12.5%. Bernard Knight & Goud
Maragini, It Is Time for the United States to Implement a Patent Box Tax Regime to
Encourage Domestic Manufacturing, 19 Stan. J. L. Bus. & Fin. 39, 50 (2013). The rate under
Colombia's software regime is 0%. Catalina Hoyos Jimenйz, Colombia ­ Corporate
Taxation (IBFD, 2015),19-20. The rate under France's regime is 15%. Cavalier & Pierre,
supra note 16, at 313; Knight & Maragini, supra note 32, at 50. The rate of the Hungarian
regime is 9.5%. Gabor Koka, Changes to Intellectual Property Box Regime Take Effect, 65
Tax Notes Int'l 345, 345 (Jan. 30, 2012). The rate under Israel's regime varies from 9% to
12
The Long Reach of EU Law
[22-Mar-16
only patents and extensions of patents to everything from copyrighted software to know-how and trademarks,33 and the income that could benefit also ranged from only royalties and licensing income to embedded royalties from the sale of goods and services.34 These regimes also varied in terms of whether they limited benefits based on who developed the IP or where the R&D took place. At least one country historically granted benefits only to income from IP that had been entirely developed by the taxpayer,35 but the majority of patent boxes that existed in 2013 also granted benefits to IP that was not self-developed but that was acquired or developed through outsourcing.36 Some patent boxes outside the EU also imposed limitations based on jurisdiction, only providing benefits when the R&D that contributed to the income was done in the jurisdiction providing benefits.37 Patent boxes that had no limits on acquisition, outsourcing, or the location of the R&D provided the greatest benefit to taxpayers,38 while those that restricted
16%. Henriette Fuchs, Israel ­ Corporate Taxation (IBFD, 2015). The rate under the
Luxembourg regime is 5.76%. van Kuijk, supra note 15, at 140. The rate under the Dutch
regime is 5%. Nijhof & Kloes, supra note 17, at 69. The rate under the Portuguese regime
will decrease from 19% to 17% by 2018. Tiago Cassiano Neves, Opening Pandora's Box:
10 International Effects of Portugal's Corporate Tax Reform, Tax Notes Int'l 1223, 1224
(Sep. 23, 2013). The rate under the Spanish regime is 5%. Jason M. Brown, Patent Box
Taxation: A Comparison of Four Recent Europea patent Box Tax Regimes and an Analytical
Consideration of If and How the United States Should Implement Its Own Patent Box, 46
Int'l Law. 913, 927 (2012). The rate under the Turkish regime, which provides a 50%
exemption, is 10%. See Yalti, supra note 17 (stating that the Turkish corporate rate is 20%).
The effective rate under the UK regime is 10%. Tom Scott & James Ross, The New Patent
Box Regime and Corporate Tax Reform in the UK, Int'l Tax J. (Oct. 2012).
33
See supra note 14-15.
34
See supra note 16-17.
35
See Lisa Evers, Helen Miller, & Christoph Spengel, Intellectual Property Box
Regimes: Effective Tax Rates and Tax Policy Considerations, Centre for European Economic
Research Discussion Paper No. 13-070 (Nov. 2013) 10 (describing the previous version of
the Spanish innovation box).
36
Many patent boxes did, however, place limits on either how much R&D can be
outsourced or from whom IP can be acquired. For example, both the Netherlands and the
United Kingdom allow outsourcing and acquisition, but they require that the resident owner
of the IP have taken on the risk associated with that IP. Nijhof & Kloes, supra note 17, at 70;
HM Revenue & Customs, Patent Box: qualifying companies: groups: active ownership
condition (CIRD 210210). Belgium applies a similar requirement that the "overall
responsibility and management of the R&D activities" must rest with the Belgian company.
De Mil & Wallyn, supra note 14, at 160. Luxembourg limits acquisitions between related
parties. Van Kuijk, supra note 15, at 143.
37
Regimes with jurisdictional limitations include those in China and Israel. W.
Wesley Hill & J. Sims Rhyne, III, Opening Pandora's Box: Global Intellectual Property Tax
Incentives and Their Implications for the United States, 53. Intell. Prop. L. Rev. 371, 385
(2013); Harris, supra note 17, at 565.
38
See Alstadsaeter et al., supra note 21 (finding that allowing acquired patents to
benefit increases the tax advantage).
22-Mar-16] Patent Boxes and International Cooperation
13
benefits to self-developed IP or IP developed in the jurisdiction were the least generous. As more European countries adopted patent boxes that did not require that the R&D be done in the jurisdiction, countries that were known for having significant amounts of domestic R&D but that did not have patent boxes, such as Germany and the United States, feared that these tax incentives would lead to lower revenues as IP was shifted outside of their jurisdictions and into jurisdictions with patent boxes.39 Several tax reform proposals from Chairman Camp's "Option C" from 2011 to proposals from Senator Feinstein in 2012 and from Representatives Boustany and Schwartz in 2013 included designs for a U.S. patent box,40 and multiple commentators wrote advocacy pieces calling for such an incentive in the U.S.41 At the same time, the countries that did have patent boxes feared that they were losing the ability to tax IP income to other jurisdictions with even more favorable patent boxes. For example, as countries such as Malta and Cyprus implemented patent boxes with rates that were well below 10% and that did not require that the R&D take place in the jurisdiction, Spain modified its patent box to both reduce the rate that applied to IP income and eliminate the requirement that qualifying IP be self-developed, thereby making it easier for more taxpayers to benefit from the lowered rate.42 It was against the backdrop of these debates that the Organisation of Economic Co-operation and Development started work on its two-year project to combat corporate tax avoidance. This project, which was known as the Base Erosion and Profit Shifting (BEPS) Project, was a fifteen-point international tax reform project under which the OECD was granted authority by the G-20.43 The BEPS Project garnered significant political support and attention, partly because international tax avoidance itself was receiving so much attention at the time the project was announced.44 The project was first proposed in the form of the BEPS Report, which laid out the general
39
See Breidthardt, supra note 18; Houlder & Peel, supra note 18.
40
See Jane G. Gravelle, A U.S. Patent Box: Issues, Congressional Research Service
Insight (Oct. 15, 2015).
41
See, e.g., Merrill et al., supra note 32.
42
Compare Evers et al., supra note 35, at 10 (describing the 2008 version of the
Spanish IP box and stating that "only self-developed IP qualifies without exceptions") with
Spain Corporate Income Tax Act ch. 4 sec. 35 (last amended by Law No. 1/2014 Feb. 28,
2014) (allowing for some IP that was not self-developed to qualify).
43
See Itai Grinberg, Breaking BEPS: The New International Tax Diplomacy, ___
Geo. L. J. ____ (forthcoming), for a description of the top-down nature of the BEPS Project.
44
Organisation for Economic Co-operation and Development, Addressing Base
Erosion and Profit Shifting (February 12, 2013) (hereinafter "BEPS Report"). See supra
notes 1-2; Grinberg, supra note 43, at __ (providing examples of how international tax
avoidance has become "front-page news").
14
The Long Reach of EU Law
[22-Mar-16
challenges facing the international tax system,45 and the fifteen specific Action Items were then set out in more detail in the BEPS Action Plan in July 2013.46 Prior to this project, the OECD had already been active in international tax policy.47 The OECD has a Model Tax Convention, the changes to which are the subject of many OECD working party meetings, and the OECD encourages Information Sharing through the Global Forum on Transparency and Tax Administration.48 The OECD also sets out transfer pricing guidelines that are often discussed in working party meetings,49 and it has published numerous reports on topics ranging from aggressive tax planning50 to bribery and corruption51 to the taxation of high net worth individuals.52 Institutionally, the OECD's work on international tax issues is carried out by the Committee on Fiscal Affairs, which is made up of high- level tax officials in each member country; the Centre for Tax Policy and Administration, which is made up of OECD staff; and a variety of working groups and expert groups that meet several times per year to discuss various international tax topics.53 The BEPS Project therefore fit into the existing international tax work that had previously been undertaken by the OECD, but it added all G-20 countries that were not also OECD members to the discussions and it brought with it an accelerated time scale and significantly more publicity and media attention than had attached to previous OECD tax projects. As part of this project, Action 5 of the BEPS Action Plan required the OECD to "[r]evamp the work on harmful tax practices with a priority on ... requiring substantial activity for any preferential regime."54 Although this
45
BEPS Report, supra note 44.
46
Organisation for Economic Co-operation and Development, Action Plan on Base
Erosion and Profit Shifting (July 19, 2013) (hereinafter "BEPS Action Plan").
47
For further discussions of the role of the OECD in tax matters, see, e.g., Grinberg,
supra note 43; Allison Christians, Networks, Norms, and National Tax Policy, 9 Wash. U.
Global Studies L. Rev. 1 (2010); Hugh Ault, Reflections on the Role of the OECD in
Developing International Tax Norms, 34 Brook. J. Int'l. L. 757 (2009).
48
See Grinberg, supra note 43.
49
Organisation for Economic Co-operation and Development, Transfer Pricing
Guidelines for Multinational Enterprises and Tax Administrations (2010).
50
Organisation for Economic Co-operation and Development, Tackling Aggressive
Tax Planning through Improved Transparency and Disclosure (2011); Organisation for
Economic Co-operation and Development, Corporate Loss Utilisation through Aggressive
Tax Planning (2011).
51
Organisation for Economic Co-operation and Development, Bribery and
Corruption Awareness Handbook for Tax Examiners and Tax Auditors (2013).
52
Organisation for Economic Co-operation and Development, Engaging with High
Net Worth Individuals on Tax Compliance (2009).
53
See Ault, supra note 47, at 761-762.
54
BEPS Action Plan, supra note 46, at 18.
22-Mar-16] Patent Boxes and International Cooperation
15
mandate did not mention patent boxes, observers, delegates, and OECD staff read the Action 5 mandate to mean that the BEPS Project had to determine how to align the benefits granted by patent boxes with the substantial activities that led to the income receiving benefits. This work was placed within the context of the OECD's ongoing work on "harmful tax practices," which had begun in 1998, when the OECD published Harmful Tax Competition: An Emerging Global Issue (known as "the 1998 Report") and created the Forum on Harmful Tax Practices (the "FHTP").55 The 1998 Report marked a new phase in international tax cooperation for several reasons. First, the 1998 Report did not just focus on general themes of cooperation and consensus. Instead, it focused on individual regimes implemented by individual countries, and its mere existence suggested that countries were complicit in the race to the bottom. Second, creating the FHTP at the same time as the 1998 Report was published ensured that this focus would continue for many years. Third, other portions of the 1998 Report suggested that it was not merely individual regimes that could be harmful but that entire countries could be named as tax havens. Although this part of the OECD's work fell by the wayside over several years,56 it does highlight the country-specific focus of the FHTP's continued work. Fourth, the framework that was established in the 1998 Report still informs the work of the FHTP, and that framework has a very specific view of what it means for a regime to represent harmful tax competition. Under that framework, if a regime applies a preferential low tax rate to geographically mobile income and is ring-fenced (i.e., the low rate is not available to domestic taxpayers) or lacks transparency or effective information sharing, then that regime will be listed in later reports by the FHTP as "potentially harmful."57 Once a regime is found to be
55
Organisation for Economic Co-operation and Development, Harmful Tax
Competition: An Emerging Global Issue (1998) (hereinafter "1998 Report"). See Ault, supra
note 47, at 767 ("The Report established a new subsidiary body within the OECD, the Forum
on Harmful Tax Practices, which, since 1998, has administered a set of guidelines on tax
practices setting out certain obligations on countries that adopted the Report").
56
Compare 1998 Report, supra note 55, at 21-22 (setting out the factors for
identifying tax havens), with Action 5 Final Report, supra note 10, at 15-16 (listing the work
of the FHTP since the 1998 Report).
57
The 1998 Report describes this requirement by setting out twelve factors for the
FHTP to consider when determining whether a jurisdiction has implemented a harmful
regime. Four of these factors are labeled as "key factors," which means that they are
sufficient for a finding of harmfulness, while the other eight factors can indicate harmfulness.
1998 Report, supra note 55, at 26-34. The interpretation and interaction of these twelve
factors was further elaborated in several later OECD publications. Organisation for
Economic Co-operation and Development, The OECD's Project on Harmful Tax Practices:
2006 Update on Progress in Member Countries (2006), available at
www.oecd.org/ctp/harmful/37446434.pdf; Organisation for Economic Co-operation and
Development, Consolidated Application Note: Guidance in Applying the 1998 Report to
16
The Long Reach of EU Law
[22-Mar-16
potentially harmful, then it must go through an assessment of "actual harmfulness," where the FHTP will consider economic data to determine whether the regime is in fact promoting harmful tax competition. If the regime has is actually harmful, other jurisdictions may impose defensive measures. Since 1998, many jurisdictions have opted to amend or abolish their regimes rather than risk either a label of potential harmfulness or actual harmfulness.58 This earlier work on harmful tax practices provided the context for the OECD's work on patent boxes. Despite the lack of any language in the BEPS Action Plan that focused on patent boxes, the OECD and G-20 member countries involved in this work interpreted the Action 5 mandate to mean that the FHTP should first focus on requiring substantial activities in what they referred to as "preferential IP regimes" (i.e., patent boxes), after which the work would then focus on other preferential regimes. In short, Action Item 5 provided an opportunity for the FHTP to focus on patent boxes, but it did so within the context of a broader project and with the involvement of nonOECD G-20 members that had not been involved in any of the FHTP's
Preferential Tax Regimes (2004, available at www.oecd.org/ctp/harmful/30901132.pdf;
Organisation for Economic Co-operation and Development, Harmful Tax Practices: The
2004 Progress Report (2004), available at www.oecd.org/ctp/harmful/30901115.pdf
(hereinafter "2004 Progress Report"); Organisation for Economic Co-operation and
Development, The OECD's Project on Harmful Tax Practices: The 2001 Progress Report
(2002), available at http://dx.doi.org/10.1787/9789264033993-en; Organisation for
Economic Co-operation and Development, Towards Global Tax Co-operation: Progress in
Identifying and Eliminating Harmful Tax Practices (2001), available at http://dx.doi.org/
10.1787/9789264184541-en. The key factors from the 1998 Report are (i) that the regime
applies no tax rate or a low effective tax rate, (ii) that the regime is ring-fenced (i.e., the
benefits are not fully available to domestic taxpayers), (iii) that the regime lacks
transparency, and (iv) that the regime lacks effective exchange of information. 1998 Report,
supra note 49, at 26-30. In practice, the first factor is necessary but not sufficient for a finding
of harmfulness, so one of the following three factors must also be evident for such a finding.
Along with these four key factors, a regime must also be within the scope of the 1998 Report,
which means that it must apply to geographically mobile income and be preferential (i.e., it
must apply a more favorable rate than the general rate that would apply to corporate income).
1998 Report, supra note 55, at 25-26. Therefore, as envisioned in the 1998 Report, a harmful
regime must apply a preferential low rate to geographically mobile income and either be
ring-fenced or lack transparency or effective exchange of information. The further eight
factors were listed as possible further indicators of harmfulness that must be considered, but
they were not necessary for any such finding. 1998 Report, supra note 55, at 30-34.
58
Ault, supra note 47, at 767-768 (stating that the 1998 Report and the subsequent
FHTP process have "been extremely effective in bringing countries to eliminate regimes
found to be harmful under the criteria of the Report. Of the forty-seven / preferential tax
regimes that had been identified as potentially harmful in 2000, none of the regimes are
deemed harmful at the present time. A number of regimes have been abolished, others have
been amended to remove their potentially harmful features, and still others were found not
to be harmful on further analysis of their actual impact").
22-Mar-16] Patent Boxes and International Cooperation
17
previous work. As mentioned above, the literature on the effectiveness of patent boxes is still quite limited, and, at the time of the BEPS Project, was not conclusive. The literature on tax competition and preferential regimes more generally, however, was more robust. Although the OECD's work in this area was premised on a belief that eliminating preferential regimes would reduce tax competition and the race to the bottom on tax rates, at least some scholars have suggested that preferential regimes are in fact helpful in reducing tax competition. In 2001, Michael Keen suggested that preferential regimes could in fact reduce overall competition by focusing this competition on specific tax bases (i.e., geographically mobile income) rather than allowing jurisdictions to compete across multiple tax bases.59 According to this view of preferential regimes, tax competition that focuses only on a specific tax base will not reduce revenue as much as tax competition that cuts across all income sources.60 Despite this literature, however, the BEPS Action Plan continued the OECD's focus on eliminating or limiting preferential regimes as a way to reduce international tax competition. The OECD therefore had to decide how to address the challenges created by patent boxes in a world where the literature on their effectiveness was not yet conclusive and where the literature on tax competition did not always support the accepted belief that preferential regimes were inherently harmful. Within this context, the OECD effectively split the difference between eliminating patent boxes entirely and allowing them to remain as they were in 2013. Action 5 of the BEPS Report says nothing about eliminating patent boxes entirely, but it also does not permit the OECD and G-20 to leave existing patent boxes standing. Instead, it mandated that the FHTP require substantial activities of these regimes. While Keen might have preferred that patent boxes be left entirely unlimited and countries without patent boxes might have preferred that patent boxes be eliminated entirely, the OECD chose at the start of the BEPS Project to take neither of those routes and instead to require that patent boxes only provide tax benefits to income that arose from substantial activities. This decision was not without empirical support. As mentioned earlier, although the literature on the effectiveness of patent boxes remains fairly inconclusive, at least one empirical study does suggest that requiring that underlying R&D be done in the jurisdiction has the effect of increasing R&D, while patent boxes that do not have this jurisdictional requirement may in fact decrease R&D in the jurisdiction.61 Furthermore, there were clear
59
Michael Keen, Preferential Regimes Can Make Tax Competition Less Harmful,
National Tax Journal Vol. LIV, No. 4, 757 (December 2001).
60
Keen, supra note 59.
61
Alstadsaeter et al., supra note 21, at 12.
18
The Long Reach of EU Law
[22-Mar-16
political reasons for making the decision not to eliminate patent boxes but also not to allow them to stand unchallenged. Within the OECD, a significant minority of jurisdictions had patent boxes. These included Belgium, France, Luxembourg, Spain, and the United Kingdom.62 These countries were unwilling to eliminate their boxes entirely, particularly since many, such as the United Kingdom, had staked significant political capital on the creation of these regimes.63 Jurisdictions without patent boxes, such as Germany, Japan, and the United States, however, were themselves unwilling to allow these regimes to exist without any restrictions, given that these regimes could lead to IP assets that had been created in a jurisdiction without a patent box to be acquired by a taxpayer in a jurisdiction with a patent box just as the IP assets began to produce income. One justification for front-end R&D incentives is that the revenue foregone in subsidizing R&D will be recaptured in the form of tax revenue if and when that R&D is successful and produces income. If, however, other jurisdictions have patent boxes that create incentives to shift that income away from the jurisdiction that funded the underlying R&D, then jurisdictions without patent boxes feared that they would just be funding R&D without ever being able to tax the income that arose out of it. Jurisdictions without patent boxes, particularly those with larger economies and more significant R&D infrastructure, therefore had an interest in requiring that those regimes only be permitted to the extent that R&D was also undertaken in the jurisdiction providing the box. Although this could be seen as encouraging jurisdictions with patent boxes to compete over not just the rate of the box but also the environment for R&D, it suggests that the countries supporting a substance requirement were confident that they could prevail in the latter type of competition, while the same might not have been true in the context of a competition over rates.
A. An Introduction to the Nexus Approach
The FHTP therefore had to create a new approach to defining substantial activities. This new approach, referred to as the "nexus approach,"64 was unveiled in the 2014 Progress Report, although not all jurisdictions had yet reached consensus on it.65 On November 11, 2014, after the 2014 Progress Report was published, the United Kingdom and Germany announced that they had reached a compromise that would lead to acceptance
62
See Action 5 Progress Report, supra note 30, at 59.
63
See, e.g., Vanessa Houlder, Treasury sets out patent incentive expansion, Financial
Times (June 10, 2011).
64
Note that some commentators have referred to this as the "modified nexus
approach" due to its many iterations in 2014 and 2015. The nexus approach and the modified
nexus approach are one and the same, and this article uses the former term to avoid confusion.
65
Action 5 Progress Report, supra note 30, at 28-29.
22-Mar-16] Patent Boxes and International Cooperation
19
of the nexus approach by all forty-four countries participating in the BEPS Project.66 The EU Code of Conduct Group then adopted the nexus approach that the FHTP had designed at the end of 2014.67 On February 6, 2015, the OECD itself issued a press release announcing that the nexus approach had been accepted and pledging to finish work on the approach by the end of June 2015.68 In October 2015, the OECD issued the final report on Action 5, which was then accepted by the G-20,69 and this 2015 Report set out the final description of the nexus approach.70 Under this approach, patent boxes and other IP regimes will not be found to be harmful if they require a nexus between the expenditures that contributed to the value of the IP income and the IP income that receives benefits. This nexus is represented by the following equation71:
Qualifying expenditures incurred
to develop IP asset
x Overall income = Income receiving
Overall expenditures incurred
from IP asset
tax benefits
to develop IP asset
Due to a political compromise between the UK and Germany, jurisdictions can permit taxpayers to increase the amount of qualifying expenditures by 30% so long as the resulting ratio does not exceed 100%.72 The nexus approach sets the outer limits within which patent boxes will be found not to be harmful. The nexus approach does not require that countries implement patent boxes, nor does it require that countries implement patent
66
German Federal Ministry of Finance, Germany and UK agree joint proposal for
rules on preferential IP regimes (November 11, 2014), available at
www.budesfinanzministerium.de/Content/EN/Pressemitteilungen/2014/2014-11-11-rules-
on-preferential-ip-regimes.html (hereinafter "German press release"); HM Treasury, HM
Revenue & Customs, and the Rt Hon George Osborne MP, Germany and UK agree joint
proposal for rules on preferential IP regimes (November 11, 2014), available at
www.gov.uk/government/news/germany-and-uk-agree-joint-proposal-for-rules-on-
preferential-ip-regimes (hereinafter "UK press release").
67
See Bob van der Meade, EU Update on patent boxes and the EU Code of Conduct
Group,
International
Tax
Review,
available
at
http://www.internationaltaxreview.com/Article/3430573/EU-Update-on-patent-boxes-and-
the-EU-Code-of-Conduct-Group-Business-Taxation.html; Council of the European Union,
Outcome of the Council Meeting, News Release 16603/14 (Dec. 9, 2014).
68
Organisation for Economic Co-operation and Development, Action 5: Agreement
on Modified Nexus Approach for IP Regimes (Feb. 6, 2015) (hereinafter "Feb. 2015
Agreement").
69
Action 5 Final Report, supra note 10.
70
For the full description of the nexus approach, see Action 5 Final Report, supra note
10, at 24-36.
71
Action 5 Final Report, supra note 10, at 25.
72
Feb. 2015 Agreement, supra note 68. See also German press release, supra note 66;
UK press release, supra note 66.
20
The Long Reach of EU Law
[22-Mar-16
boxes that apply the exact equation above, but it does require that, in order to escape a finding of potential harmfulness, patent boxes cannot provide benefits to any income that would not qualify under the equation above.73 The OECD has thus described it as a "box around the box" because all patent boxes that jurisdictions choose to implement must fall within the confines of the nexus approach.74 One key element of the nexus approach that is not made explicit in the 2015 Report is that there are essentially two versions of the nexus approach: the main version, which must be implemented by any Member State of the European Union that has a patent box, and the footnote version, which could be implemented by non-EU countries. The 2015 Report presents the main version as the only version, since it can be adopted by all jurisdictions, but several footnotes of the Report highlight that jurisdictions outside the EU can choose to design their patent boxes quite differently. In the main version, qualifying and overall expenditures are defined by focusing on which entity undertook them.75 Qualifying expenditures are those incurred by the individual entity benefiting from the patent box and any expenses for outsourcing to unrelated parties,76 while overall expenditures include these expenditures plus all acquisition costs and any expenses for outsourcing to related parties.77 In other words, the nexus ratio can be written as a+b a+b+c+d where a includes R&D expenditures incurred by the taxpayer, b includes expenditures for outsourcing R&D to unrelated parties, c includes expenditures for acquiring IP from related or unrelated parties, and d includes expenditures for outsourcing R&D to related parties.78 Qualifying entities include resident taxpayers as well as both outbound and inbound PEs to the extent that those PEs are subject to taxation in the jurisdiction providing the patent box.79
73
Patent boxes may allow income that would not qualify under the nexus approach to
qualify for benefits in limited circumstances if they treat the nexus ratio as a rebuttable
presumption, but the presumption must be designed such that it can only be rebutted in a
narrow set of circumstances. See Action 5 Final Report, supra note 10, at 35-36.
74
Organisation for Economic Co-operation and Development, Live Webcast: The
BEPS Package (Oct. 5, 2015), slide 28, available at www.slideshare.net/OECDtax/beps-
webcast-8-launch-of-the-2015-final-reports (hereinafter "BEPS Webcast").
75
Action 5 Final Report, supra note 10, at 27-29.
76
Action 5 Final Report, supra note 10, at 27-28.
77
Action 5 Final Report, supra note 10, at 28-30.
78
Action 5 Final Report, supra note 10, at 28.
79
This limitation appears to be based on the premise that it would not be beneficial
for a taxpayer that was not otherwise subject to tax at the full corporate rate on its income in
22-Mar-16] Patent Boxes and International Cooperation
21
In the footnote version, qualifying and overall expenditures are instead defined by where the expenditures were incurred. Qualifying expenditures are all R&D expenditures incurred in the jurisdiction providing benefits, while overall expenditures are all R&D expenditures incurred by the taxpayer, whether domestically or internationally.80 Therefore, in the footnote version, the nexus ratio can be written as a a+b where a includes all R&D expenditures incurred in the jurisdiction providing benefits (whether undertaken by the taxpayer itself or outsourced to or acquired from other parties) and b includes all R&D expenditures incurred outside the jurisdiction (whether undertaken by the taxpayer itself or outsourced to or acquired from other parties). To illustrate the difference between the two versions of the nexus approach, take A Co., which is resident in Country A, a jurisdiction which has a patent box. A Co. has three subsidiaries: Sub A, which is also resident in Country A; Sub B, which is resident in Country B; and Sub C, which is resident in Country C. This is shown below in Figure 1:
Figure 1
Country A
A Co.
Sub A
Country B
Sub B
Country C
Sub C
A Co. earns royalty income from licensing out Patent A, which it owns. The initial R&D for Patent A was done by Sub C, which paid 500 to an unrelated company in Country A to undertake all its R&D. After the rights to the initial R&D were acquired by A Co. for 500, Patent A was further
the jurisdiction to subject itself to taxation just for the sake of receiving a reduced rate.
80
The footnote version is laid out in footnotes 16 and 19 of Chapter 4 of the Action 5
Final Report, supra note 10.
22
The Long Reach of EU Law
[22-Mar-16
developed partly by Sub A and partly by Sub B, each of which was paid 750 for its R&D by A Co. and each of which undertook R&D in its country of residence. These expenditures are listed below in Table 1:
Table 1
A Co. expenditures:
Acquisition of Sub C R&D
500
(undertaken in Country A)
Outsourcing to Sub A
750
(undertaken in Country A)
Outsourcing to Sub B
750
(undertaken in Country B)
TOTAL
2000
Under both versions, overall expenditures would equal 2000.81 Under the main version, there would be 0 in qualifying expenditures because all expenditures for acquisition of the R&D rights from Sub C would be excluded from qualifying expenditures, as would all expenditures for outsourcing to both Sub A and Sub B because they are both separate entities from A Co. Note that this creates an incentive to restructure Sub A so that it either earns the income its own R&D or so that it merges with A Co. Under the footnote version, there would be 1250 in qualifying expenditures, which would include both the 500 in Country A R&D expenditures incurred by Sub C and the 750 in Country A R&D expenditures incurred by Sub A.82 In both versions, the expenditures incurred for R&D undertaken by Sub B would be included in overall expenditures and excluded from qualifying expenditures. In the main version, this is because Sub B is an unrelated entity. In the footnote version, this is because Sub B's R&D was undertaken outside of Country A. As will be discussed in more detail in Section II, this difference is due
81
This is because 500 + 750 + 750 = 2000. Note that this example is simplified
because A Co. paid the same amount to Company C that Company C paid for R&D
expenditures. If Company C had paid 400 in R&D expenditures, all of which were for R&D
in Country A, then overall expenditures in the footnote version would be only 1900. Both
examples assume that Company C was able to show that it had engaged in complete tracking
and tracing to ensure that it did not incur any other expenditures. See footnote 19 of Chapter
4 of the Action 5 Final Report, supra note 10. If Company C was not able to show this, then
overall expenditures would include A Co.'s acquisition costs rather than Company C's R&D
costs in the footnote version as well as the main version.
82
This is again only true if Company C engaged in tracking and tracing that could
show that all of its R&D expenditures were for R&D in A Co. See footnote 19 of Chapter 4
of the Action 5 Final Report, supra note 10.
22-Mar-16] Patent Boxes and International Cooperation
23
to the fact that Member States of the European Union cannot, under the EU treaties, discriminate based on jurisdiction.83 As twenty-one of the forty-four countries involved in the BEPS Project are subject to this constraint, the main version of the nexus approach could not require EU Member States to design patent boxes that violated the EU treaties.84 These two versions, however, represent two very different visions of substantial activities. Under the main version, substantial activities do not include any R&D outsourced to a related party or any R&D done by another party that was then acquired. Under the footnote version, these activities may constitute substantial activities if they were undertaken within the jurisdiction providing benefits, regardless of which entity undertook them, while any outsourcing outside the jurisdiction (whether to a related or unrelated party) does not constitute substantial activities. In other words, substance as defined in the main version depends on who engages in R&D, while substance as defined in the footnote version depends on where the R&D takes place. These two visions of substance may often overlap, but they differ in terms of fundamental principles. Although the two versions differ in terms of their overall focus, they share the same general requirements. Both versions of the nexus approach share the same definition of IP assets and overall income from IP, and they also share the same requirements for tracking and tracing of income, as well as grandfathering. In terms of IP assets, the 2015 Report explicitly states that only "patents and other IP assets that are functionally equivalent to patents" can qualify under a nexus-compliant patent box, and any patent box that provides benefits to other IP assets will therefore be considered potentially harmful.85 The 2015 Report defines functionally equivalent IP assets to include copyrighted software, and it also permits jurisdictions to extend benefits to a third category of IP assets that are "non-obvious, useful, and novel" the taxpayers receiving benefits for such assets fall into a narrowly defined category of small enterprises.86 Both the 2014 Progress Report and the 2015 Report make clear that trademarks and marketing-related IP assets do not fall within the definition of qualifying IP assets.87 Although this distinction is not explained in greater detail, this is likely due to the stated principle of the nexus approach, which is to grant benefits only to taxpayers
83
See infra Section II.
84
See infra Section II.
85
Action 5 Final Report, supra note 10, at 26 and footnote 9 of Chapter 4.
86
Such enterprises can have "no more than EUR 50 million (or a near equivalent
amount in domestic currency) in global group-wide turnover" and may not "themselves earn
more than EUR 7.5 million per year (or a near equivalent amount in domestic currency) in
gross revenues from all IP assets, using a five-year average for both calculations." Action 5
Final Report, supra note 10, at 26.
87
Action 5 Final Report, supra note 10, at 27; Action 5 Progress Report, supra note
30, at 31.
24
The Long Reach of EU Law
[22-Mar-16
that engaged in value-creating R&D activities. If patent boxes were permitted to grant benefits to all income arising from trademarks or other marketing related intangibles, these could arguably grant benefits to all income earned by a company, given that the value of trademarks, brands, know-how, and other non-qualifying IP assets arises from all the activities of a company. This in turn would mean that a patent box was essentially just providing a lower rate for all the income earned by any company with a marketing intangible. In terms of overall income from IP, the 2015 Report leaves a significant amount of flexibility to jurisdictions as to how they can define and calculate qualifying income. Jurisdictions are not required to provide benefits to embedded IP income, nor are they prohibited from doing so. The 2015 Report instead states that jurisdictions should ensure that they do not provide benefits to gross income, although it leaves flexibility to them as to how they define net income, and that they should also ensure that, if they do provide benefits to embedded income, they do so in a way that ensures that routine marketing and manufacturing returns do not receive benefits.88 Because the nexus approach requires a link between expenditures and income, both versions also require that taxpayers that benefit from a patent box must engage in sufficient "tracking and tracing" to ensure that the income receiving benefits did in fact arise from qualifying expenditures.89 In the 2014 Progress Report, the general description of the nexus approach required that taxpayers track and trace expenditures and income either to individual IP assets or to individual products.90 The 2015 Report acknowledges that such narrow tracking and tracing may be impossible for large taxpayers with multiple R&D projects and streams of income, so it also permits tracking and tracing to product families if taxpayers can show that they could not feasibly track and trace to a narrower category and they can show that each product family includes overlapping streams of expenditures and revenues.91 Both versions of the nexus approach also permit the grandfathering of patent boxes that existed prior to a certain date.92 This is consistent with previous work of the FHTP, which allowed regimes that would otherwise be potentially harmful to be grandfathered if "(1) no new entrants are permitted into the regime, (2) a definite date for complete abolition of the regime has
88
Action 5 Final Report, supra note 10, at 29.
89
Action 5 Final Report, supra note 10, at 30-34.
90
Action 5 Progress Report, supra note 30, at 34 and footnote 3 of Chapter 4.
91
Action 5 Final Report, supra note 10, at 31-32. Both versions of the nexus approach
also permit jurisdictions to establish a transitional measure, although the outlines of such a
transitional measure are again left to jurisdictions to decide. Action 5 Final Report, supra
note 10, at 33. The 2015 Report does provide an example of such a transitional measure, but
it makes clear that this is only provided as an illustration of a possible transitional measure
that a jurisdiction could adopt. Action 5 Final Report, supra note 10, at Annex A.
92
Action 5 Final Report, supra note 10, at 34-35.
22-Mar-16] Patent Boxes and International Cooperation
25
been announced, and (3) the regime is transparent and has effective exchange of information."93 Patent boxes and other IP regimes may therefore continue to grant benefits to pre-existing beneficiaries after June 30, 2016, as long as the jurisdiction with the patent box has begun a legislative process to modify the box in 2015.94 Patent boxes may not, however, provide benefits to the extent that the IP assets benefiting from a grandfathered patent box were acquired from a related party after January 1, 2016, if they could not have qualified from a patent box at the time of acquisition.95 In other words, if a taxpayer that benefited from an existing patent box on or before June 30, 2016, acquired an IP asset from an unrelated party in any jurisdiction or from a related party in a jurisdiction with a patent box (including in the same jurisdiction as the taxpayer), it can qualify for grandfathering, including on income from that IP asset. If not, it can qualify for grandfathering on all income other than income from that IP asset. As stated in the February 6, 2015, press release, this grandfathering safeguard was designed to prevent taxpayers from circumventing grandfathering by transferring IP assets into a qualifying patent box at the last minute.96 As of June 30, 2016, therefore, all new patent boxes must comply with the nexus approach, and no taxpayers can benefit from patent boxes that do not comply with this approach after June 30, 2021 at the very latest. Patent boxes that do not comply with the nexus approach by these dates will be found to be potentially harmful and will then be subject to the FHTP's analysis of actual harmfulness. Given the variety of patent boxes that existed prior to this date, as well as their varied objectives and motivations, what does this mean for the future of patent boxes? The following section outlines the likely effect of the nexus approach on both current and future patent boxes.
B. The Impact of the Nexus Approach
Although the effect that the nexus approach will have on the number and design of patent boxes remains to be seen, there are several elements of the approach that suggest the effects it could have. This section first considers the effect of the nexus approach as a whole and then focuses on the grandfathering provisions, the tracking and tracing requirements, and the entity focus of the main version of the nexus approach.
93
Action 5 Final Report, supra note 10, at 34 (citing 2004 Progress Report, supra note
53).
94
Action 5 Final Report, supra note 10, at 34.
95
Action 5 Final Report, supra note 10, at 35.
96
Feb. 2015 Agreement, supra note 68.
26
The Long Reach of EU Law
[22-Mar-16
i. The effects of the nexus approach as a whole: an increase in patent boxes and a decrease in design flexibility
Several jurisdictions that did not have patent boxes before the BEPS Project, including Ireland and Italy, announced in 2014 and 2015 that they were implementing such regimes,97 and the most recent U.S. patent box proposal was presented as being at least in part in response to "the actions of foreign governments pursuant to the OECD Base Erosion and Profit Shifting (BEPS) Project."98 This has led some observers to suggest that one effect of the nexus approach was to make patent boxes more appealing to jurisdictions and possibly even encourage jurisdictions to adopt such regimes.99 It is worth noting, however, that patent boxes had existed for many years prior to the development of the nexus approach, and many countries that did not have them at the time of the BEPS Project had nonetheless considered implementing them.100 The reason that the OECD developed the nexus approach was that, as the number of patent boxes increased, so too did the leniency of these patent boxes. In other words, while many countries with patent boxes at first required some degree of substance for a taxpayer to benefit from the regime, these boxes were eventually amended to require less substance as other countries implemented competing boxes that themselves did not require significant substance. For example, up until 2015, Israel had a patent box that essentially limited its benefits to income arising from R&D undertaken in Israel.101 In 2015, however, Israel announced that it would modify this box to be more in line with the patent boxes in other jurisdictions that did not limit their benefits to income arising from domestic R&D.102 Another example of this trend away from restricted patent boxes can be seen in Spain, which until 2013 had a very restricted patent box that only granted
97
See Ireland Dept. of Finance, Draft Finance Bill (Oct. 22, 2015); Luca Bosco &
Stefano Schiavello, Italy Tax Alert: Implementation rules issued for new patent box regime
(Oct. 24, 2015).
98
Boustany Neal Press Release, supra note 23.
99
See Charles Boustany, Opening Statement: Examining the OECD Base Erosion and
Profit Shifting (BEPS) Project (Dec. 1, 2015) (stating that "the BEPS project ended up
making recommendations that will achieve the opposite result by encouraging countries to
create patent boxes, which will effectively force worldwide companies to shift their business
operations out of the United States").
100
See Tax Reform Act of 2014, H.R. 1 (113th Cong., 2d session), Section 250;
Manufacturing Innovation in America Act of 2013, H.R. 2605 (113th Cong., 1st session).
See also Graetz & Doud, supra note 5 (noting in an article written prior to the BEPS Project
that jurisdictions already felt under pressure to implement patent boxes).
101
Henriette Fuchs, Israel to improve IP tax breaks, Globes (July 8, 2015), available
at www.globes.co.il/en/article-israel-to-improve-ip-tax-breaks-1001051318.
102
Fuchs, supra note 101.
22-Mar-16] Patent Boxes and International Cooperation
27
benefits to self-developed IP.103 In 2013, however, Spain modified this box to allow benefits for income arising from much more IP, including IP that had been developed through outsourcing and acquisition.104 Therefore, while some countries may now feel competitive pressure to implement their own patent boxes, this pressure existed prior to the nexus approach, and the nexus approach ensures that the competition at least takes place on a narrower playing field. That said, there are three ways in which the nexus approach can be seen as having increased the pressure to adopt a patent box. First, the nexus approach does not entirely prohibit jurisdictions from implementing patent boxes. The FHTP could, in theory, have held that a regime that grants benefits to IP income can never require a sufficient level of substantial activities, which would in turn have eliminated patent boxes. It chose not to do so, however, perhaps because many of the countries involved in the OECD's consensus-based discussions of Action 5 themselves had patent boxes.105 But the result was that Action 5 did not eliminate patent boxes, and so it brought greater international attention to these tax incentives without eliminating them. Second, the nexus approach now sets a baseline against which patent boxes will be assessed. While jurisdictions can permit less income to benefit from a patent box than would be permitted by the nexus approach, setting this baseline may make it harder for them to do so since their taxpayers will be able to compare the patent boxes of individual jurisdictions against the nexus approach. This in turn may make it harder for some jurisdictions to maintain no patent box at all or to design a patent box that is significantly less generous than the nexus approach itself because taxpayers can now point to the nexus approach as the permissible outer limit of patent box design. Third, due to the compromise between Germany and the United Kingdom, this baseline is now higher than what countries may otherwise have designed since countries are permitted to increase the amount of qualifying income that can receive benefits by 30%.106 This means that, even if a taxpayer outsourced R&D or acquired IP, it may be able to qualify for more
103
See Evers et al., supra note 35, at 10.
104
Spain Corporate Income Tax Act ch. 4 sec. 35 (last amended by Law No. 1/2014
Feb. 28, 2014).
105
See supra note 62 and accompanying text.
106
The 30% uplift applies to qualifying expenditures, but, since these represent the
numerator of the nexus approach, the effect of this is to increase the amount of income that
can qualify by 30%. The 30% uplift cannot increase the nexus ratio beyond 100%, so
taxpayers with a pre-uplift nexus ratio of greater than 76.9% will only be able to use the
uplift to increase their ratio up to 100% and not beyond. See Action 5 Final Report, supra
note 10, at 27-28.
28
The Long Reach of EU Law
[22-Mar-16
benefits than initially suggested by the nexus approach.107 This could in turn lead to more taxpayer demands for patent boxes in jurisdictions that do not have them or more demands that current patent boxes increase their benefits to be in line with the 30% uplift. One effect of the nexus approach as a whole is therefore that it may have made patent boxes more appealing by not eliminating them entirely and by setting a fairly high baseline. Even if the nexus approach does increase the number of jurisdictions with patent boxes, a further effect of this approach, however, is to restrict the design of those patent boxes. Prior to the nexus approach, there was no limitation on what patent boxes could be designed to achieve. Although most jurisdictions with patent boxes claimed that they were designed to encourage innovation, whatever that may mean,108 these jurisdictions designed these regimes in very different ways. Some of them allowed benefits to go to income from all types of intellectual property, including trademarks, brands, and know-how, while others only provided benefits to income from royalty. Some of them allowed benefits to go to income from the sale of goods created or services provided using IP (this income is referred to as "embedded royalties," since it is thought that a portion of the sale price represents a royalty of sorts for the use of the underlying IP), while others limited their benefits to royalties and licensing income. Some patent boxes required that all the R&D be done by the taxpayer itself or in the jurisdiction providing benefits, while others had no such limitations. Therefore, what these jurisdictions meant by "innovation" varied significantly. Now that the nexus approach applies to patent boxes, all such regimes will have far less flexibility in their design. Jurisdictions may only provide benefits to a designated group of IP assets (except in the case of small enterprises, only patents and their functional equivalents and copyrighted software), so any patent boxes that provide benefits to trademarks, knowhow, and the like will not qualify. Jurisdictions also must show that any income that receives benefits arises from R&D, so they must have a way of separating embedded royalties from the overall sales price for a good or service. And jurisdictions must limit benefits based on the proportion of R&D expenditures incurred by the taxpayer and unrelated parties (or, in the case of the footnote version of the nexus approach, the proportion of R&D expenditures incurred within the jurisdiction). Therefore, even if the number
107
This is only true if the taxpayer incurred some qualifying expenditures. If a taxpayer
has no qualifying expenditures that would be included in the numerator of the nexus ratio,
then the 30% uplift would not increase that taxpayer's benefits from zero. See Action 5 Final
Report, supra note 10, at 27-28.
108
See, e.g. David Gauke, Speech to the Securities Industry Conference (Oct. 3, 2014)
(stating that the UK introduced its patent box "[t]o spur innovation").
22-Mar-16] Patent Boxes and International Cooperation
29
of patent boxes increases, the benefits that these regimes can provide will be constrained by the requirements of the nexus approach, and these requirements are intended to ensure that patent boxes can no longer be used to attract income without also attracting the underlying R&D. Several individual provisions of the nexus approach may offset some of these effects of the nexus approach as a whole. As outlined below, some specific provisions may change both the appeal of patent boxes and the effectiveness of the limits imposed on these regimes. ii. The effect of grandfathering provisions: a short-term increase in patent boxes
Another effect that can already be seen arises out of the grandfathering provisions that allow non-nexus-compliant patent boxes to be grandfathered for five years if they were in effect on or before June 30, 2016. These provisions create an incentive for countries without patent boxes to implement non-nexus-compliant boxes prior to the cut-off date in order to allow taxpayers to benefit from exactly the design features that the nexus approach was designed to eliminate and to receive those benefits for up to five years. Since Germany and the United Kingdom announced the provisional agreement of the nexus approach on November 11, 2014, several jurisdictions have modified their regimes to be inconsistent with the nexus approach or introduced entirely new regimes that are not consistent with the nexus approach, suggesting that, at least in the short term, one impact of the approach could be to increase the proportion of regimes that do not meet the nexus approach. For example, Italy's recently implemented patent box grants benefits to income from trademarks, even though this income is explicitly excluded under the nexus approach, while Israel's recently announced changes to its patent box remove the jurisdictional limit and therefore make this incentive less likely to comply with the nexus approach.109 This effect will be fairly short-lived, however, since grandfathering only applies to taxpayers who were already benefiting from the approach on or before June 30, 2016, and it also requires any jurisdiction benefiting from grandfathering to have started the legislative process for implementing a nexus-compliant regime in 2015. Therefore, in order to benefit from grandfathering, Italy and other countries would have had to have established nexus compliant regimes that would go into effect for new taxpayers on June 30, 2016, and any failure to do so would, by the terms of the 2015 Report, disqualify existing beneficiaries from benefiting from grandfathering.
109
See Bosco & Schiavello, supra note 97; Fuchs, supra note 101.
30 iii.
The Long Reach of EU Law
[22-Mar-16
The effect of tracking and tracing requirements: a long-term decrease in patent boxes
One apparent effect of the nexus approach is to increase the number of patent boxes, but this may only be a short-term effect. This is because the nexus approach has made patent boxes more costly for both tax administrations and taxpayers. In order to ensure that taxpayers benefiting from the patent box in fact engaged in the qualifying expenditures that contributed to its income, the nexus approach requires that taxpayers engage in significant tracking and tracing of both R&D expenditures and the income arising from IP assets.110 Although the 2015 Report permits this tracking and tracing to take place on a product or product family basis, this is still a significant administrative requirement for taxpayers that want to benefit from a patent box and for tax administrations that must determine whether taxpayers have in fact engaged in legitimate tracking and tracing. The nexus approach therefore establishes a significant new requirement, whereby taxpayers that want the benefit of a patent box must increase record-keeping and establish new systems to comply with this box. The nexus approach also adds other complexities to patent boxes, such as requiring taxpayers to have qualifying IP assets or engage in complex calculations to determine what portion of their income from each IP asset, product, or product family represents embedded royalties.111 Although recent patent box proposals suggest that taxpayers have not yet stopped demands for these regimes, these added requirements combined with the limitations on policy objectives outlined above may in fact have the long-term effect of making patent boxes less appealing.112 These requirements could lead to fewer taxpayers lobbying for patent boxes and fewer jurisdictions feeling pressure to implement or maintain patent boxes, which could mean that one possible long-term consequence of the nexus approach could be to make patent boxes less attractive. iv. The effect of the entity focus: a weakening of the restrictions of the nexus approach
As mentioned previously, the main version of the nexus approach that appears in the 2015 Report requires patent boxes to divide expenditures by
110
See supra Section I.A.
111
See supra Section I.A.
112
This argument also suggests that criticisms about the complexity of the nexus
approach or regimes structured after the nexus approach may be unfounded if the effect of
this complexity is to eliminate patent boxes in the long run. See, e.g., Sullivan Sept. 2015,
supra note 20 (criticizing the complexity of the Innovation Promotion Act of 2015, which
shares certain similarities with the nexus approach).
22-Mar-16] Patent Boxes and International Cooperation
31
entity. If the expenditures were incurred by the entity or for outsourcing to an unrelated party, then they will be qualifying expenditures that increase the amount of income that qualifies for benefits. If they were expenditures for acquiring IP from another entity or for outsourcing to a related party, then they will reduce the amount of income that qualifies for benefits. Taken together, these requirements mean that the main version of the nexus approach is designed to discourage both the shifting of income from entity to entity and the shifting of innovation from entity to entity. Whereas existing patent boxes focused more on the shifting of income, the nexus approach now requires that they also focus on the shifting of innovation and only grant benefits to income when the innovation itself was undertaken by the entity receiving benefits. This in turn suggests one vision of substance that underlies the main version of the nexus approach. Because this version maintains a strict focus on entities rather than on jurisdictions, it suggests that the fundamental issue underlying patent boxes was not that income was being taxed in one jurisdiction while the activities were taking place in another jurisdiction. Instead, the concern about substantial activities was that they were taking place in one entity (regardless of its location) while the income was being allocated to another entity (regardless of its location). According to the logic of the main version of the nexus approach, this means that a taxpayer that structures itself such that one domestic subsidiary engages in certain R&D activities while another domestic subsidiary earns the income from those activities is engaged in impermissible tax planning. The footnote version of the nexus approach, in contrast, focuses on jurisdictions rather than entities. This version grants benefits only to the extent that the R&D expenditures were incurred for R&D undertaken in the jurisdiction granting benefits. If a taxpayer pays for R&D undertaken in another jurisdiction, those expenditures will reduce the amount of income that can benefit. This version again discourages both the shifting of income and the shifting of innovation, but the focus is not on shifting from entity to entity but rather from jurisdiction to jurisdiction. Under the vision of substance supported by this version, base erosion and profit shifting takes place not through shifting between entities but rather through shifting between jurisdictions. This version is not concerned with whether the R&D was undertaken by one entity or another, but it is instead concerned with where the R&D took place, and subjecting the income to tax in one jurisdiction when the R&D activities took place in another jurisdiction is seen as impermissible tax planning under this view of substance. One effect of this entity focus could be a distortion of taxpayer investment and innovation decisions. Since the nexus approach now only grants benefits to income that arose from R&D undertaken by the entity or
32
The Long Reach of EU Law
[22-Mar-16
within the jurisdiction, companies that would otherwise have divided their R&D across entities or jurisdictions are now discouraged from doing so purely for tax reasons. Of course, patent boxes are themselves distorting taxpayer decisions by granting benefits based on the location of income or the type of activity contributing to that income. The nexus approach is therefore not itself creating a distortion that would not exist in a land of unrestricted patent boxes. Instead, it is placing a restriction on patent boxes that are already creating distortions, and it may in fact offset some of the normatively undesirable distortions created by patent boxes, such as the incentive in favor of shifting income away from the underlying substance in order to benefit under a patent box.113 Furthermore, governments have been intentionally creating distortions in this sector for years, whether through direct subsidies, front-end deductions and credits, manufacturing incentives, or other non-tax provisions. Therefore, any criticism of the distortions created by the nexus approach must acknowledge that incentives for R&D have historically created an entire backdrop of distortions against which any changes must be measured. Yet, under the main version of the nexus approach, the focus on shifting between entities means that domestic subsidiaries are treated as separate entities. Therefore, taxpayers with multiple subsidiaries within a Member State that want to benefit fully from a patent box must restructure to ensure that the entity that earns income from an IP asset is the same entity that incurs the R&D expenditures for that IP asset, which may require a different corporate structure than would otherwise be selected for business reasons. Moreover, since the main version does not distinguish between outsourcing to domestic or foreign entities or between acquiring IP developed through domestic or foreign R&D, this version of the nexus approach creates a disincentive against both outsourcing to all related parties and all forms of acquisition. This could therefore lead companies to view outsourcing and acquisition as more costly than before, even if these would otherwise be the most efficient ways to develop an IP asset or increase the amount or quality of innovation. A second effect of this entity focus could be to allow some shifting of income across jurisdictions since branches are not separate taxable entities. Consider the earlier example illustrated in Figure 1,114 and imagine that Sub B is instead Branch B, a branch of A Co. The 750 paid to Branch B and then used for the R&D done in Country B could now be treated as 750 paid by A
113
See Graetz & Doud, supra note 5, at 407 (noting in an article written prior to the
BEPS Project that "there is little reason today to believe that the decision regarding where to
locate such income turns on where the related R&D is performed or where the spillovers
occur).
114
See supra Section I.A.
22-Mar-16] Patent Boxes and International Cooperation
33
Co. for R&D rather than for related-party outsourcing. It would therefore become a qualifying expenditure, even though the R&D was undertaken in Country B, and the nexus ratio as calculated under the main version of the approach would now be 37.5% ­ or 48.75% if Country A applied the 30% uplift. Therefore, in this example, no income would qualify for the box if Sub B remained a separate entity, but almost half of IP income could qualify if Sub B were instead treated as a branch. The entity focus could thus encourage companies to establish branches instead of subsidiaries, and it could also permit income to qualify even when the underlying R&D took place elsewhere It is unlikely that this would lead to significant income-shifting because work in other areas of the BEPS Project should limit the ability of taxpayers to have such an arrangement respected for tax purposes. Under the Action 7 output, if Branch B had the significant R&D facilities necessary to develop Patent A, Branch B would likely be treated as a taxable PE in Country B.115 This would therefore mean that Country B would have taxing jurisdiction over the 750 paid to Sub B, whereas previously this payment from head office A Co. to Branch B would have been a wash for Country A tax purposes. This in turn means that it is unlikely that taxpayers will shift a significant amount of R&D to foreign branches, but the main version of the nexus approach does allow for some slippage at the margins, since taxpayers now have an incentive to shift just enough R&D to a branch so that it will increase the nexus ratio while not triggering PE status. The incentive to try to strike this balance would not exist in the absence of the entity focus of the main version of the nexus approach. This entity focus therefore may weaken the restrictions of the nexus approach and allow some degree of jurisdictional income-shifting compared to the footnote version, which focuses directly on where the R&D took place rather than which entity undertook it v. The overall effect of the nexus approach
The effect of the nexus approach therefore remains to be seen, but it appears that the nexus approach could increase the pressure for jurisdictions to adopt patent boxes, at least in the short term, to take advantage of grandfathering provisions and to respond to taxpayer demands that may have increased since the publication of the Action 5 Report. In the long run, however, the tracking and tracing requirements of the nexus approach may decrease the demand for patent boxes, and the overall requirements of the
115
See Organisation for Economic Co-operation and Development, Action 7: 2015
Final Report, Preventing the Artificial Avoidance of Permanent Establishment Status (2015)
(hereinafter "Action 7 Final Report") (making it more difficult for the branch in the above
example not to be treated as a permanent establishment).
34
The Long Reach of EU Law
[22-Mar-16
nexus approach will change the design of patent boxes since these regimes must now require a nexus between R&D expenditures and qualifying income. The nexus approach's focus on entities, however, suggests that taxpayers may still be able to use patent boxes to shift some income, and patent boxes applying this nexus will themselves create distortions and incentives that may be unrelated to the goal of keeping R&D in the same jurisdiction as the income receiving benefits.
II. THE NEXUS APPROACH AND EU LAW
If the nexus approach had only included the footnote version, several of the weaknesses identified above would not exist. First, there would be fewer opportunities for taxpayers to receive benefits for income even when R&D had not taken place in the jurisdiction providing benefits. Second, taxpayers would be free to outsource to any parties they saw fit or acquire IP assets at any stage of their development with no effect on the amount of income that could benefit, so long as the underlying R&D had itself been undertaken in the jurisdiction. A third benefit of using only the footnote version is that, to the extent that existing research can provide guidance on the design of patent boxes, the literature suggests that having a jurisdictional requirement is necessary to increase R&D within the jurisdiction. Finally, the footnote version is more consistent with the goals of the BEPS Project, given that the description of substantial activities in the BEPS Report focused on shifting income between jurisdictions, not entities.116 The nexus approach, however, does not only include the footnote version, and this version is in fact hidden away in the footnotes of the 2015 Report. Why, then, did the OECD agree on a version of the nexus approach that discourages outsourcing and acquisition, regardless of where the underlying R&D takes place? Instead of choosing a version that focuses on jurisdiction, why did the OECD settle on a more complicated and less intuitive version that focuses on the entity that undertook the R&D rather than the jurisdiction where the R&D took place? The answer to those questions is that the treaty freedoms of the European Union would not permit the Member States of the EU to adopt a patent box that focused on jurisdiction. In previous case law, the European Court of Justice (the ECJ) has concluded that R&D incentives that only apply to R&D undertaken in the Member State providing the incentive discriminate in violation of the EU's treaty freedoms.117 R&D credits have historically been provided only to R&D undertaken in the jurisdiction for several reasons.
116
See BEPS Report, supra note 44.
117
See Laboratoires Fournier, Case C-39/04 (10 March 2005); Sociйtй Baxter, Case C-
254/97 (8 July 1999).
22-Mar-16] Patent Boxes and International Cooperation
35
First, it's generally understood that the positive spillovers resulting from R&D are local, so R&D credits should focus on creating an incentive for local R&D and thereby increase the local spillovers. Second, if the R&D supported by R&D credits is successful, a jurisdiction can recapture some of the revenue losses due to the credits by taxing the income from the income arising from the successful R&D. This recapture is only possible, however, if the income is earned within the jurisdiction and therefore subject to taxation. Third, jurisdictions generally want to be able to ensure that credits are being used to support actual R&D, and oversight of this is easier when the R&D is in the jurisdiction providing the credits. Despite these reasons for limiting R&D credits, however, the ECJ struck down the jurisdictional limits on Member State R&D credits and required that they either be provided for R&D undertaken throughout they EU or not be provided at all. A substantial activities requirement produced by the OECD, which includes twenty-one EU Member States,118 could therefore not mandate any differential treatment based on the location of R&D. Instead, it had to find a different lens through which to view substantial activities, and the FHTP chose entities. Rather than treating R&D differently depending on where it took place, the main version of the nexus approach therefore treats R&D different depending on who undertook it. This in turn led to the main version distinguishing between related and unrelated parties rather than between domestic and foreign R&D. Outsourcing to all related parties is excluded from qualifying expenditures, while outsourcing to all unrelated parties is included in qualifying expenditures. As stated in both the 2014 Progress Report and the 2015 Report, the reason for this was that it was assumed that taxpayers would not choose to outsource the actual value-generating portion of R&D to an unrelated party, so any such outsourcing would likely be at the margins.119 In terms of acquisitions, since there is no way for the EU nexus approach to distinguish between acquisitions where the underlying R&D took place in the jurisdiction and those that are just shifting IP out of another jurisdiction, the cost for all acquisitions must be excluded from qualifying expenditures. The focus on entities is thus intended to achieve a similar outcome to a focus on the location of the R&D, but it must allow for some slippage, where some external R&D will be permitted to increase the nexus
118
The countries that are both Member States of the European Union and OECD
members include Austria, Belgium, the Czech Republic, Denmark, Estonia, Finland, France,
Germany, Greece, Hungary, Ireland, Italy, Luxembourg, the Netherlands, Poland, Portugal,
the Slovak Republic, Slovenia, Spain, Sweden, and the United Kingdom. Norway and
Iceland, members of the European Economic Area, and Switzerland, a member of the
European Free Trade Association, are also OECD members and therefore participated in the
BEPS Project.
119
Action 5 Final Report, supra note 10, at 29-30. See also Action 5 Progress Report,
supra note 30, at 32-33.
36
The Long Reach of EU Law
[22-Mar-16
ratio and some internal R&D will end up decreasing the nexus ratio, because EU law prohibits any Member State tax incentive from focusing directly on the location of the R&D. The main version is therefore more complicated and likely more costly to taxpayers than appears necessary. Any other version of the nexus approach that would comply with the EU treaty freedoms would, however, not prevent base erosion and profit shifting. Since the prohibition on discriminating based on the location of R&D means that a patent box or other back-end tax regime would be consistent with EU treaty freedoms if it were to provide benefits to all R&D undertaken throughout the EU, one option for the OECD would have been to design the nexus approach so that R&D undertaken anywhere in the EU would qualify the resulting income for benefits. This option, however, would itself create an incentive for base erosion and profit shifting within the EU, since taxpayers could benefit from a patent box in a low-tax Member State with limited support for innovation even if they had undertaken all the R&D in a different EU Member State that had the infrastructure, educational system, trained employees, and technical support necessary for that R&D. In other words, this would permit taxpayers to shift income anywhere within the EU, even if the underlying activities had taken place in another EU Member State. So the nexus approach is not as weak as it could have been, and the OECD produced the most robust version of the approach that was possible given the constraints of EU law.120 The fact remains, however, that it was because of EU law that the nexus approach does not take the logical approach of focusing on jurisdiction. The logic underlying this outcome is the logic of the single market: in order to be one supranational market with no internal barriers, Member States cannot prevent the shifting of R&D or income within the EU. Yet, given that the EU institutions still have no affirmative authority over direct taxation, this logic means that the effect of EU law is in fact to
120
Since the main version of the nexus approach was designed to comply with EU law
and adopted by the Code of Conduct Group, see infra notes 135-138 and accompanying text,
this article assumes that, if a nexus-compliant patent box is challenged in front of the ECJ,
the Court will find it not to violate the EU treaty freedoms or the prohibition on state aid.
Given the ECJ's constantly changing view of the scope of Treaty protections, however, it is
not impossible that, despite the OECD's efforts to design an approach that is consistent with
EU law, a Member State taxpayer could challenge it as a violation of Treaty freedoms or the
state aid prohibition. If this were to happen, and if the ECJ were to find that a nexus-
compliant patent box was inconsistent with the TFEU's protections, this would strengthen
the arguments in this article about the detrimental effect of EU law on international
cooperation. It would also raise fundamental questions about the interactions between the
institutions of the EU, given the Commission's participation in the working party meetings
that led to the creation of the nexus approach, the Code of Conduct Group's adoption of the
approach, and the Commission's decision not to initiate state aid investigations of patent
boxes due to the ongoing debates over the nexus approach.
22-Mar-16] Patent Boxes and International Cooperation
37
allow and encourage base erosion and profit shifting. In many ways, EU law underlies the entire BEPS Project. In the years leading up to this project, the ECJ struck down or limited a wide range of Member State anti-avoidance rules. In general, many anti-avoidance rules treat a payment, transaction, or person differently when there is a crossborder element. Therefore, a domestic taxpayer who engages in purely domestic transactions will be treated differently for tax purposes than a nonresident taxpayer or a domestic taxpayer who engages in cross-border transactions. The reason for this is that, when a transaction or event is purely domestic, the country imposing taxes knows what happens on both ends of the transaction: if a payment is deducted in the hands of the payor, then it will be included in the hands of the payee, and the taxing jurisdiction will be assured that it will be taxed at least once. When a transaction or event is not purely domestic, however, the opportunities for non-payment of taxes increase because one jurisdiction cannot be sure of what is happening in the other jurisdiction. For example, if the deduction on the payor side was premised on the assumption that the payment would be taxed on the payee side, that assumption can no longer be supported if the payee is in another jurisdiction. It is therefore common to distinguish between domestic and cross-border transactions in tax legislation.121 In the European Union, however, the ECJ struck down rules that treated domestic and cross-border transactions differently. With these decisions, Member States lost many tools in the fight against tax avoidance. They could not impose withholding taxes on dividends or similar payments to other Member States, even if those payments were excluded from income or otherwise subject to benefits in the recipient Member State.122 They could not impose robust controlled foreign company ("CFC") rules on subsidiaries in low-tax Member States to prevent income from being shifted into those subsidiaries.123 They could not limit the deductibility of interest payments to other Member States, even if those interest payments were excluded from income or otherwise subject to benefits in the recipient Member State.124
121
See, e.g., 26 U.S.C. §§ 951-965 (providing rules targeting foreign controlled
corporations that do not apply to domestic controlled corporations).
122
See Denkavit Internationaal BV, Denkavit France SARL v. Ministre de l'Йconomie,
des Finances et de l'Industrie, C-170/05 (Dec. 14, 2006) (holding that a tax that is imposed
on dividends paid to non-resident parents and that is not imposed on dividends paid to
resident parents is a violation of the Treaty freedoms).
123
See Cadbury Schweppes pic, Cadbury Schweppes Overseas Ltd v. Commissioners
of Inland Revenue, C-196/04 (Sept. 12, 2006) (holding that controlled foreign company rules
that apply to non-resident subsidiaries and not to resident subsidiaries are a violation of the
Treaty freedoms to the extent that the subsidiaries to which they apply are not wholly
artificial arrangements).
124
Lankhorst-Hohorst GmbH v. Finanzamt Steinfurt, C-324/00 (Dec. 12, 2002)
38
The Long Reach of EU Law
[22-Mar-16
They also could not have any rules that required residents of other Member States to have engaged in substantial activities in order to receive tax benefits, since this was seen as disproportionate to the goal of preventing tax avoidance.125 Essentially, the ECJ barred Member States from policing tax avoidance within the European Union because it disallowed rules that discriminated against non-resident taxpayers and recognized only a very limited exception for the prevention of tax avoidance. This in turn meant that EU law could at least in part explain many of the patent boxes that most concerned the jurisdictions that wanted the BEPS Project to target these provisions. Countries that opposed patent boxes were most concerned by those that had no limitation on the location of the R&D or the identity of the party who undertook the R&D, but EU law was what prohibited Member States from limiting their patent boxes to income from domestic R&D. This meant that, by the time of the BEPS Project, the European Union had become a safe space for tax avoidance. Taxpayers could structure their transactions such that they took advantage of the lack of withholding taxes between jurisdictions and the lack of anti-avoidance rules within EU Member States. While the jurisprudence of the ECJ was not the only cause of the base erosion and profit shifting issues facing countries at the time of the BEPS Project, it was a key factor in creating these issues. This can be seen in the fact that several of the BEPS Action Items focused on the very areas where the ECJ had struck down or not permitted Member State rules as violations of the treaty freedoms. For example, Action 2 focused on hybrid arrangements, where the same transaction or entity is treated differently in two different jurisdictions.126 One way to address these arrangements is for a jurisdiction to change its treatment based on the other jurisdiction's treatment, but this is the type of differential treatment that the ECJ has previously not permitted. Action 3 focused on CFC rules, which the ECJ had severely limited.127 Action 4 focused on interest deductibility rules, which the ECJ had also limited.128 In effect, these Action Items highlight that one goal of the BEPS Project was to overcome the limits that had been placed on anti-avoidance efforts by the ECJ. And yet, as shown by the nexus approach, efforts to overcome the
(holding that thin capitalization rules that apply differently to non-resident shareholders and
resident shareholders are a violation of the Treaty freedoms.
125
See Lilian V. Faulhaber, Sovereignty, Integration and Tax Avoidance in the
European Union: Striking the Proper Balance, 48 Colum. J. Transnat'l L 177 (2010)
(defining the "wholly artificial arrangements doctrine," according to which the ECJ only
permits Member States to justify a discriminatory measure as a means to prevent tax
avoidance if that measure targets only wholly artificial arrangements).
126
BEPS Action Plan, supra note 46, at 15.
127
BEPS Action Plan, supra note 46, at 16.
128
BEPS Action Plan, supra note 46, at 16-17.
22-Mar-16] Patent Boxes and International Cooperation
39
ECJ's jurisprudence were themselves thwarted by the Court's interpretation of the freedoms protected in the Treaty on the Functioning of the European Union (TFEU). The nexus approach is therefore striking in that it did work within the confines of EU law to address a problem to which EU law contributed, but it also highlights that the BEPS Project was hampered from the start given that many Action Items were limited in what they could achieve based on the legal constraints that applied to EU Member States. This was true for several other Action Items as well. The Reports on Action 3 and Action 4, for example, included several pages that explained the restrictions imposed by the EU treaty freedoms and made clear that no recommendations in those Reports should be interpreted as requiring Member States to violate those freedoms.129 Given that those freedoms were the very ones that led to weak CFC and interest deductibility rules in the first place, the fact that the OECD had to work within those constraints limited the degree to which the OECD could advocate rules that would eliminate tax avoidance opportunities. EU law therefore limited the ability of Member States to police tax avoidance both before and after the BEPS Project. The OECD did propose recommendations and requirements that went as far as possible at combating tax avoidance within the constraints of EU law, but these constraints prevented it from adopting the most robust possible anti-avoidance rules. This effect of EU law is particularly clear within the context of the work on patent boxes, because this is the one area where the OECD's outputs include both the desired outcome and the more limited outcome that is necessary to comply with EU constraints. By including the footnote version but explicitly limiting this version to non-EU countries, the OECD made clear that the main version was a deviation from the version that could have been proposed in the absence of EU law. Although the ECJ's interpretations of the treaty freedoms limited the outcomes under the BEPS Project, this was not the only interaction between EU institutions and the BEPS Project. Other EU institutions and groupings, including the Council, the Commission and the Code of Conduct Group, in fact contributed to strengthening the outputs under the BEPS Project. For example, in the context of Action 2, which dealt with hybrid mismatch arrangements, some Member States had interpreted a piece of EU legislation as preventing the types of rules that were proposed in the Action 2 report.130
129
See, e.g., Organisation for Economic Co-operation and Development, Action 3:
2015 Final Report, Designing Effective Controlled Foreign Company Rules 17-18 (2015)
(hereinafter "Action 3 Final Report") (discussing the constraints placed on CFC rules by EU
law).
130
See Organisation for Economic Co-operation and Development, Action 2: 2015
Final Report, Neutralising the Effect of Hybrid Mismatch Arrangements (2015) (hereinafter
40
The Long Reach of EU Law
[22-Mar-16
This legislation, known as the Parent-Subsidiary Directive, prevents Member States from imposing taxes on certain intragroup payments.131 The Action 2 report, however, included several recommendations, one of which was the imposition of a tax on intragroup payments. This recommendation said that, if a subsidiary had been permitted to deduct a payment to a parent, the parent country could then impose taxes on that payment rather than allowing it to escape taxation. Some observers saw this and other Action 2 recommendations as violating the Directive. In response to this concern, the Council of the European Union approved two amendments to the Directive in 2014 and 2015. The first of these amendments132 directly supported Action 2 by requiring that a parent impose a tax on a payment to the extent that it had been deducted in a subsidiary resident in another Member State. The second amendment then made it easier for Member States to adopt other Action 2 recommendations that may also have run afoul of the ParentSubsidiary Directive, thereby eliminating an EU legal constraint that would have significantly weakened the effectiveness of the BEPS Project's recommendations.133 Another example of cooperation between the European Union and the OECD can be seen in the Code of Conduct Group's adoption of the nexus approach. In 1997, the EU established the Code of Conduct for Business Taxation, which targeted harmful tax practices within the EU.134 Since then, the Code of Conduct Group has been charged with eliminating these practices, and it does so by applying five criteria when assessing regimes in Member States: (i) "whether advantages are accorded only to non-residents or in respect of transactions carried out with non-residents," (ii) "whether advantages are ring-fenced from the domestic market," (iii) "whether advantages are granted even without any real economic activity and substantial economic presence within the Member State offering such tax advantages," (iv) "whether the rules for profit determination...depart from internationally accepted principles," and (v) "whether the tax measures lack
"Action 2 Final Report").
131
Directive 2011/96/EU.
132
Council Directive 2014/86/EU amending Directive 2011/96/EU on the common
system of taxation applicable in the case of parent companies and subsidiaries of different
Member States (8 July 2014).
133
See Council Press Release, Parent-subsidiary directive: Council agrees to add anti-
abuse clause against corporate tax avoidance, ST 15103/14 PRESSE (December 9, 2014);
Directive 2015/121/EU (January 27, 2015). This directive added an anti-abuse provision to
the Parent-Subsidiary Directive that would instruct governments not to grant benefits under
the directive to arrangements designed to obtain a tax advantage.
134
Conclusions of the ECOFIN Council Meeting on 1 December 1997 concerning tax
policy (98/C 2/01).
22-Mar-16] Patent Boxes and International Cooperation
41
transparency."135 More recently, the Code of Conduct Group opened discussions on Member State patent boxes, many of which were later assessed by the FHTP. These discussions took place in the context of the Group's discussions of the third criterion, which requires that advantages only be granted where there is real economic activity and a substantial economic presence within the Member States offering the tax advantage.136 On November 20, 2014, just after Germany and the UK released their compromise proposal on the nexus approach, the Code of Conduct Group announced that it was adopting the main version of the nexus approach as developed by the FHTP in the context of the BEPS Project.137 This adoption increased pressure on Member States who were participating in the BEPS Project to adopt the nexus approach themselves, since they would now be subject to it at the EU level regardless of whether or not the OECD adopted it. Therefore, opposing the nexus approach no longer provided any benefit to these Member States. This effect can be seen by the fact that, before the FHTP assessed any patent boxes, the Code of Conduct Group assessed all Member State patent boxes and concluded that none of them were compatible with the nexus approach.138 Furthermore, this expanded the scope of the nexus approach by subjecting all twenty-eight EU Member States, including the seven who were not members of the OECD or G-20, to assessment under the nexus approach. In practical terms, this meant that Cyprus and Malta patent boxes, both of which applied rates of 2.5% and neither of which would have been considered under the OECD's nexus approach since neither Cyprus nor Malta is an OECD or G-20 member, were now subject to the nexus approach. The institutions of the European Union therefore have contributed to the success of the BEPS Project by reforming EU-level legislation and adopting the nexus approach in the Code of Conduct Group. The Commission also took part as an observer in many of the working party and FHTP
135
Conclusions of the ECOFIN Council Meeting on 1 December 1997 concerning tax
policy (98/C 2/01), Sections B(1)-(5).
136
Conclusions of the ECOFIN Council Meeting on 1 December 1997 concerning tax
policy (98/C 2/01), Sections B(3). On its face, the third factor requires that substantial
activities occur within the jurisdiction granting tax benefits, but the Code of Conduct Group
does not appear to have expressed concerns that this requirement is inconsistent with the
ECJ's tax avoidance jurisprudence.
137
Van der Meade, supra note 66. In the media, there was some confusion about what
organization had developed the nexus approach. See, e.g., Liz Loxton, Closing the loophole:
Tax breaks on IP and patents, Economia (Feb. 5, 2015), available at
http://economia.icaew.com/finance/february-2015/closing-the-double-irish-loophole
(quoting a lawyer as referring to "the EU's Forum on Harmful Tax Practices," which does
not exist). This was developed by the FHTP and adopted by the Code of Conduct Group.
138
Van der Meade, supra note 66.
42
The Long Reach of EU Law
[22-Mar-16
meetings that led to the OECD's recommendations, and it also stepped back from its 2013 review of patent boxes as state aid.139 What the two versions of the nexus approach highlight, however, is that, despite these positive contributions on the part of other EU institutions, EU law remains a significant constraint on the ability of Member States to police tax avoidance. The ECJ's jurisprudence created opportunities for the base erosion and profit shifting that led to the BEPS Project, and it also limited the efforts by Member States and other OECD and G-20 countries to curtail these opportunities during the BEPS Project. While this is not necessarily the intention of the Court, which achieves this result merely by interpreting the Treaty freedoms to prevent discrimination against non-residents, the result of this interpretation is to create a welcoming environment for tax avoidance within the European Union. That effect is worrisome enough when it just constrains the ability of Member States to police tax avoidance. As will be discussed in the next Section, however, the BEPS Project expanded the impact of the ECJ's proavoidance jurisprudence so that it is now not only Member States who are constrained in what they are able to do to prevent tax avoidance. Instead, the ECJ's jurisprudence has now made it harder for even non-EU countries to police tax avoidance and curtail tax competition.
III. EU LAW AND INTERNATIONAL COOPERATION
EU Member States were not the only countries that were party to the BEPS Project. Although they made up twenty-one of the forty-four participants, the majority of the countries taxing part in the BEPS Project were not EU Member States. But, as shown above, the results of the BEPS Project were constrained by legal requirements that applied only to those twenty-one Member States. This in turn illustrates an important new development in international relations: it is no longer just the Member States of the European Union who are subject to EU law constraints. These constraints now effect countries outside the European Union as well. A. The Long Reach of EU Law
Although the two different versions of the nexus approach mean that EU Member States are the only countries that must implement the main version, they do not mean that other countries cannot implement the main version. Instead, non-EU countries are free to adopt either the main version or the footnote version (or a mix of the two). And even though the footnote
139
See, e.g., Ajay Gupta, Is the EU Preparing for Failure on BEPS?, Tax Notes Int'l
815 (Sept. 7, 2015).
22-Mar-16] Patent Boxes and International Cooperation
43
version is more consistent with the goals of Action 5, it is likely that many jurisdictions even outside the European Union will end up implementing at least a version of the main version of the nexus approach. One of the main reasons for implementing a patent box is to compete with other countries for both revenue and R&D activity. Therefore, when countries have a choice between a tax incentive that will be more attractive to taxpayers or one that will be less attractive because it provides fewer benefits, many countries will choose the former. The main version of the nexus approach allows for patent boxes that are generally more attractive to taxpayers because they are likely to allow some extraterritorial R&D to qualify for benefits, while the footnote version does not.140 That said, the main version also creates more complexity by requiring greater tracking and tracing and creating pressures against outsourcing and acquisition, so some non-EU countries may adopt the main version, while others may incorporate some combination of the main version and the footnote version, such as a patent box that focuses on which entity undertook the R&D but then adopts jurisdiction-focused rules on acquisitions. Even if some non-EU countries adopt a purely jurisdiction-focused patent box, however, the mere existence of the main version of the nexus approach changes the competitive environment in which countries are deciding to adopt patent boxes. The fact that the EU Member States were able to shape the nexus approach so that it focused on entities rather than location means that taxpayers now know that, at least in the European Union, patent boxes will allow for some degree of income-shifting, where non-local R&D can qualify income for benefits. This in turn means that non-EU jurisdictions that see themselves competing with EU jurisdictions for taxpayers and income will feel pressure to allow similar slippage in their own patent boxes. This in turn shows that EU law no longer has an effect just on the Member States of the European Union. Instead, when combined with the competitive pressures facing countries as they design tax incentives, it changes the legal landscape for other countries. In other words, when EU Member States are part of an international or transnational negotiation, EU law shapes those negotiations and can have an effect on the law that applies in non-EU countries as well as in Member States. This lesson can be seen in some of the other outputs from the BEPS Project as well. In the context of CFC rules, for example, the apparent goal of including Action 3 was to push for more robust CFC rules worldwide.141 The outcome from Action 3,
140
This is because of the treatment of unrelated party outsourcing. The treatment of
acquisitions, however, is less generous under the main version, so this may overall end up
being less generous. Jurisdictions could, however, choose to adopt the main version as it
applies to outsourcing and the footnote version as it applies to acquisition.
141
BEPS Action Plan, supra note 46, at 16 (outlining the positive effects of CFC rules).
44
The Long Reach of EU Law
[22-Mar-16
however, only set out non-binding recommendations for countries that wished to adopt CFC rules, and it expressly acknowledged that these rules could not be designed in a way that was inconsistent with the EU treaty freedoms.142 Given that EU treaty freedoms were what had weakened Member State CFC rules to begin with, this acknowledgment did not offer much comfort to observers who had hoped that the BEPS Project would lead to stronger rules than those permitted under the ECJ's jurisprudence.143 EU law therefore has a much broader reach than has previously been acknowledged. Although many academics have discussed the internal inconsistencies of the ECJ's direct tax jurisprudence144 and the effect of this case law on Member State tax provisions,145 the nexus approach shows that the ECJ's direct tax jurisprudence has an effect on tax provisions outside of the European Union. This means that the European Union and its court system are not only important for EU legal experts, but instead for all countries that are negotiating with the EU or competing with the EU for taxpayers, income, or anything else. In an earlier article,146 Anu Bradford identified what she referred to as the "Brussels effect," whereby the European Union has raised the floor for regulatory standards through market harmonization.147 In the process identified by Bradford, standards in areas where the EU has regulatory authority (such as antitrust law, privacy, human health, and the environment)148 have been heightened even outside the EU because the EU's higher regulatory standards have pushed up all regulatory standards due to market pressures.149 What I identify here is essentially the Luxembourg
142
Action 3 Final Report, supra note 129, at 17 (stating that "EU Member States will
need to ensure that they make choices that are consistent with EU law").
143
The Action 3 Final Report did suggest that Member States could design their CFC
rules to be more robust than what they currently had. This recommendation suggested that,
while most Member State CFC rules were designed to target only wholly artificial
arrangements, they could also be designed to target a wider variety of subsidiaries as long as
that included both resident and non-resident subsidiaries, or they could be designed to target
"partly wholly artificial arrangements," or they could be designed to apply more broadly so
long as they could be justified by a need to maintain the balanced allocation of taxing power.
See Action 3 Final Report, supra note 129, at 17-18. All these suggestions were placed within
the constraint of EU law, however, and the fact that the Action 3 Final Report did not require
any Member States to implement CFC rules means that these design options will not change
the environment for CFC rules unless Member States affirmatively want to test the limits of
EU law.
144
See, e.g., Mason & Knoll, supra note 4; Graetz & Warren, supra note 4.
145
See, e.g. Faulhaber, supra note 125.
146
Anu Bradford, The Brussels Effect, 107 Northwestern U. L. Rev. 1 (2012).
147
Bradford, supra note 146, at 3.
148
Bradford, supra note 146, at 19.
149
Bradford, supra note 146, at 3.
22-Mar-16] Patent Boxes and International Cooperation
45
effect.150 In an area where the EU does not have any regulatory authority such as direct taxation,151 the EU cannot raise international standards through market convergence, so the Brussels effect does not apply. What the nexus approach shows us, however, is that the opposite outcome arises. Instead of just leaving individual country standards as is, the ECJ reduces these standards when it finds that individual Member State provisions are inconsistent with the Treaty freedoms. These reduced standards are then exported through international competition. The reduced ability of Member States to police tax avoidance therefore does not just affect Member States, but it changes the entire competitive landscape and creates pressures for other countries to also reduce their anti-avoidance standards. This effect was made particularly clear in the context of the BEPS Project, since the goal of that project was to raise international standards limiting tax avoidance. In many of the Action Items, the OECD and G-20 achieved this goal: they introduced rules to combat hybrid mismatch arrangements, which very few countries had even attempted to prevent;152 they made it harder to avoid PE status;153 they established new requirements for exchanging and collecting taxpayer information;154 and they modified the transfer pricing guidance to allocate more income to jurisdictions where value creation took place.155 These outcomes could be achieved because these were areas where the ECJ's jurisprudence did not apply. In the context of hybrid mismatch arrangements, for example, the Council's amendment to the Parent-Subsidiary Directive was sufficient to remove at least some limits on Member State action. PE status was a change made to bilateral treaties, and, although the ECJ has previously challenged treaty provisions, the establishment of PE status is further from the ECJ's jurisprudence than an individual domestic law affecting cross-border taxation. Exchange of information was outside the scope of the ECJ's jurisprudence because this information was required of all taxpayers, and the transfer pricing changes were just made in the guidance, which the ECJ sees as generally consistent
150
While the Commission and other EU institutions are located in Brussels, the
primary seat of the European Court of Justice is Luxembourg.
151
Professor Bradford acknowledged that the Brussels effect only applies in areas
where the "missing regulatory propensity...reflects a preference for heterogeneity in the EU"
and identified direct taxation as one such area. Bradford, supra note 146, at 59.
152
Action 2 Final Report, supra note 130.
153
Action 7 Final Report, supra note 115.
154
Action 5 Final Report, supra note 10, at 45-60 (setting out rules for the spontaneous
exchange of rulings); Organisation for Economic Co-operation and Development, Action 13:
2015 Final Report, Transfer Pricing Documentation and Country-by-Country Reporting
(2015).
155
Organisation for Economic Co-operation and Development, Actions 8-10: 2015
Final Reports, Aligning Transfer Pricing Outcomes with Value Creation (2015).
46
The Long Reach of EU Law
[22-Mar-16
with EU law. But in areas where the ECJ had already ruled on the inconsistency of Member State rules with EU law, the OECD was not able to raise international standards as high as it could have in the absence of EU law. EU law therefore has an effect that goes well beyond the borders of the EU ­ and Bradford's arguments about regulatory convergence go well beyond the borders of the EU's regulatory authority. Regulatory convergence in fact happens in reverse in areas where two requirements are met: (i) the EU has no regulatory authority and (ii) the ECJ has struck down the ability of Member States to pass legislation. When both of these requirements are met, the Member States face a legislative and regulatory vacuum, where they cannot implement their chosen provisions but no EU institution has the authority to replace those provisions with EU-wide harmonization. This vacuum in turn affects countries outside the EU, creating a lower bar against which to compete and in turn reducing international standards. This means that, while the debates over the internal consistency of the ECJ's direct tax jurisprudence are important and relevant, they are not the only debates that should be taking place around this case law. Instead, there needs to also be a focus on how this jurisprudence affects countries beyond the EU as well as Member States and a discussion of what can be done. The following Section considers some possible answers, but, as shown by the nexus approach and other outcomes of the BEPS Project, most of these answers require the ECJ and the other EU institutions to limit these constraints, which they have thus far been unwilling to do.
B. What Can Be Done?
The long reach of the ECJ's direct tax jurisprudence exists because of the combination of the ECJ's jurisdiction over cases arising under the freedoms protected by the TFEU and the other EU institutions' lack of authority to pass direct tax legislation without the unanimous support of Member States. To address the effect of this combination, the European Union institutions and Member States either need to expand the EU-level direct tax legislation that fills the gaps left by the ECJ's direct tax case law or curtail the ECJ's jurisdiction over direct tax cases. In recent months, the Commission has made moves toward expanding the small number of regulations and directives (the two forms of EU legislation) in the tax area. It has done so by proposing what it refers to as the Anti-Tax Avoidance (ATA) package, which includes a proposed ATA Directive.156 This Directive includes six possible anti-avoidance rules that
156
European Commission, Proposal for a Council Directive laying down rules against
tax avoidance practices that directly affect the functioning of the internal market (January
22-Mar-16] Patent Boxes and International Cooperation
47
Member States would be required to adopt: (i) an interest deductibility rule, (ii) an exit tax, (iii) a switch-over clause (under which a Member State can switch from the exemption method to the credit method for income received from another Member State with a low tax rate), (iv) a general anti-abuse rule, (v) a CFC rules, and (vi) a hybrid mismatch rule.157 In order for this Directive to go into effect, all twenty-eight Member States would need to vote in favor of it. Once it was implemented, each Member State would then also need to pass domestic implementing legislation. It therefore still requires significant Member State action, but the Commission is holding this up as a major step forward in the EU-wide fight against tax avoidance. Despite the Commission's claims, this effort to integrate Member State anti-avoidance efforts through EU-wide harmonization in fact illustrates just how difficult it will be for Member States to offset the Luxembourg effect through political agreement. First, any agreement would require the unanimous support of all Member States. Given that twenty-one Member States were involved in Action 3 of the BEPS Project, which was intended to lead to countries adopting stronger CFC rules but instead set out general best practices which were not required, it is unlikely that those same Member States would agree to an EU-wide Directive requiring all Member States to adopt CFC rules.158 Second, several of the Directive's proposals are not entirely consistent with the BEPS Project's recommendations that cover the same issues. Member States are therefore unlikely to agree to legislation that would result in subjecting them to two inconsistent sets of rules. The hybrid mismatch rule proposed by the Commission, for example, requires the residence state to adopt the characterization of the source state for both hybrid instruments and hybrid entities.159 This is different from the recommendations set out by the OECD, which explicitly opted not to recharacterize income160 but instead set out rules denying deductions or forcing inclusions.161 Although the two approaches, if adopted, should both eliminate hybrid mismatches, they are fundamentally different in that
28, 2016) (hereinafter "Proposed ATA Directive").
157
Proposed ATA Directive, supra note 156, at 16-22.
158
See supra 141-143 and accompanying text.
159
Proposed ATA Directive, supra note 156, at 21.
160 See Action 2 Final Report, supra note 130, at 26 (stating that "[t]he primary and
defensive rules are limited to adjusting the tax consequences that flow from the difference in
the tax treatment of the instrument and should not generally affect the underlying character
of the payment (e.g. whether it is treated as interest or a dividend) or the quantification or
tax treatment of a taxpayer's overall gain or loss on the acquisition or disposal of an asset
acquired under a financial instrument").
161
See Action 2 Final Report, supra note 130, at 11-12 (describing the general
approach applied to hybrid instruments and entities, under which the primary rule would
allow the source country to deny the deduction and the secondary rule would allow the
residence country to impose taxation).
48
The Long Reach of EU Law
[22-Mar-16
recharacterization of a payment can potentially have greater consequences than just denying a deduction or imposing taxation. The Commission's proposal is therefore fundamentally inconsistent with the OECD's recommendations. Third, and most important, even if Member States were to give their unanimous support to the Directive, the Directive itself is still subject to the Treaty freedoms as interpreted by the ECJ. Therefore, all the proposed rules in the Directive are limited by the ECJ's existing jurisprudence. This means that the Commission has only proposed interest deductibility rules, exit taxes, and CFC rules, among others, that are consistent with EU law, which in turn means that they do not go any further toward preventing tax avoidance than many of the rules that already exist in Member States.162 EU-wide legislation thus does not seem to be the answer to the current stalemate since even that would be subject to the Treaty freedoms. What, then, could the ECJ itself do to address this situation? One option would be for the Court to take a more deferential approach when deciding cases involving Member State implementation of internationally agreed recommendations. Although the ECJ is not a political body, many commentators have noted throughout the years that its deference to Member State rules and its interpretation of the Treaty freedoms fluctuate in line with the political pressures facing the European Union.163 Given that the Commission participated in OECD meetings about the BEPS outputs and that twenty-one Member States of the EU participated in the BEPS Project, the ECJ could interpret the BEPS outputs as the result of international cooperation to which it should defer, and it could therefore see any restrictions arising from these as outside the scope of the Treaty freedoms.164 This could provide Member States and non-EU countries with more certainty that, if they were to agree to heightened standards in any future international agreements, those standards would be more protected than individual Member State rules. This would, however, contrast with the ECJ's selfpresentation as an apolitical body that is merely interpreting the Treaty freedoms, so it may be unlikely that the Court would follow this approach. If harmonization at the EU level is unlikely and the ECJ is also unlikely to constrain its own jurisdiction, then the only remaining solution is for Member State action. Member States could amend the Treaty, either to deny the ECJ jurisdiction over direct tax cases or to require less than unanimous support over direct tax legislation. Either of these could address
162
See Proposed ATA Directive, supra note 156, at 16-21.
163
See, e.g., Servaas van Thiel, The Direct Income Tax Case Law of the European
Court of Justice: Past Trends and Future Developments, 62 Tax L. Rev. 143 (2008).
164
An opportunity to exercise such deference could arise, for example, if a nexus-
compliant patent box were to be challenged by a Member State taxpayer. See supra note 120.
22-Mar-16] Patent Boxes and International Cooperation
49
the impasse, either by allowing more Member State tax provisions to stand or by replacing them with more EU-wide tax provisions. In the context of anti-avoidance legislation, choosing only the first option of denying the ECJ jurisdiction would permit more Member State anti-avoidance rules to stand but would do nothing about harmonizing these rules. Choosing only the second option would harmonize anti-avoidance rules across the EU, but their effectiveness would be limited because they would still be subject to the existing EU law constraints. Therefore, the most effective option for Member States concerned about losing control over their ability to police tax avoidance would be to choose both options, which would allow greater harmonization of direct tax provisions while also ensuring that those provisions would not be subject to challenge in violation of the Treaty freedoms. In the current political environment, however, choosing either of these options seems just as unlikely as Member States agreeing to direct tax legislation or the ECJ voluntarily constraining its jurisdiction. Member States have previously considered and rejected the idea of rescinding the ECJ's jurisdiction over direct tax cases,165 and Member States have also shown themselves unwilling to eliminate the unanimity requirement for direct tax legislation, given that direct taxation still represents a fundamental exercise of sovereignty. In the near term, therefore, the situation is likely to remain as it currently is: Member States cannot pass anti-avoidance legislation that discriminates based on jurisdiction, and the EU institutions cannot pass any anti-avoidance legislation at the EU level. And this in turn means that, at least in areas where the ECJ has struck down Member State rules, international tax standards are themselves going to be pushed down by the effect that EU law has on the competitive landscape. This effect is illustrated in the nexus approach, where the definition of substantial activities in the context of patent boxes was constrained by EU law and ended up focusing on entities rather than location, but it is also illustrated in several other of the BEPS outputs as well.
CONCLUSION At the start of the BEPS Project, several countries and commentators expressed concern about the effect of patent boxes on the international tax environment. They claimed that patent boxes were leading to reduced tax rates and greater opportunities for base erosion and profit shifting.166 In response, the OECD and G-20 designed the nexus approach to require these
165
See Ruth Mason, Made in America for European Tax: The Internal Consistency
Test, 49 B.C. L. REV. 1277, 1280 (2008).
166
See Breidthardt, supra note 18; Houlder & Peel, supra note 18.
50
The Long Reach of EU Law
[22-Mar-16
tax provisions to require substantial activities. While the nexus approach did impose limitations on patent boxes and may even lead to a reduction in the number of patent boxes in the long-term, it also included two separate versions: the main version, which focused on entities, and a version hidden in the footnotes, which focused on the location of the underlying R&D. As suggested by the BEPS Report that the OECD issued in 2013, as well as empirical literature on the effectiveness of patent boxes, a limitation based on the location of the underlying R&D would have been the most logical way to curtail patent boxes and ensure that they achieved their stated goal of increasing the amount of R&D in a jurisdiction. And yet the main version of the nexus approach focuses on entities, which creates incentives to restructure and disincentives for outsourcing and acquisition, as well as more opportunities for avoidance than a location-based approach. The reason for this deviation away from the more logical approach was EU law. Even though EU law created many of the base erosion and profit shifting opportunities that led to calls for the BEPS Project, the involvement of Member States of the European Union in the Project meant that the outputs of the project were themselves limited by EU law. The pro-avoidance jurisprudence of the European Court of Justice thus did not just have an impact on Member States of the EU. Instead, it had an impact on other countries as well, since these countries were now competing for revenue and taxpayers in an environment where patent boxes could be designed to comply with the more lenient main version of the nexus approach. The nexus approach therefore illustrates the long reach of EU law. No longer do decisions of the ECJ just reduce the ability of Member States to pass direct tax legislation. They now also push down international standards on taxation through the Luxembourg effect: where the lack of EU-wide regulatory authority and the ECJ's jurisprudence combine to create a vacuum, that vacuum will lead to lower standards internationally since other jurisdictions will be competing with the lack of law in the European Union. In the context of tax avoidance, this means that it is now harder for all countries ­ including those not subject to the EU treaties ­ to police tax avoidance because the ECJ has interpreted anti-avoidance rules to violate the Treaty freedoms. Although the ECJ and Member States could theoretically address this problem by reducing the ECJ's jurisdiction over direct tax cases and agreeing to EU-wide tax legislation, these options seem unlikely, at least in the near future. The lesson to be learned from the nexus approach is therefore that academics, practicing lawyers, and negotiators alike must all be aware of the effect that the ECJ's jurisprudence has on regulatory convergence in the direct tax area. While discussions of the impact of this jurisprudence within the EU are important and relevant, the conversation also needs to focus on
22-Mar-16] Patent Boxes and International Cooperation
51
the impact of this jurisprudence outside the EU as well and acknowledge that the European Court of Justice is making decisions that limit the ability of the United States and other non-EU countries to police and prevent international tax avoidance.

LV Faulhaber

File: the-long-reach-of-european-union-law-patent-boxes-and-the-limits.pdf
Author: LV Faulhaber
Author: Anne
Published: Tue Mar 22 11:34:21 2016
Pages: 51
File size: 0.31 Mb


, pages, 0 Mb

Has modernism failed, 10 pages, 0.12 Mb

, pages, 0 Mb

, pages, 0 Mb
Copyright © 2018 doc.uments.com