Patent Boxes: The Rise, the Change or the Fall?

This article discusses the reasons for intellectual property (patent) box regimes, the OECD Base Erosion and Profit Shifting measures intended to counter abuse of such regimes, the consequential changes effected by countries to their regimes and, finally, the effects of these changes and whether these regimes will survive.

1. Introduction

Over the last two decades, national jurisdictions worldwide have come to realize that attracting international business specializing in innovation as well as research and development (R&D) has allowed them to import vitally important knowledge-based technologies and other capital resources, as well as create new domestic markets and expand into promising international markets. Attracting international business that competes on innovation and R&D helps jurisdictions build up their national competitive advantages as countries seek to stabilize their economic development and secure steady employment growth. In order to achieve these ends, countries have systematically provided generous R&D tax incentives to investors. However, this general practice among countries has inevitably led to acute harmful tax competition, giving rise to an international race to the bottom for tax revenues.

This article endeavours to analyse the rise of the various R&D tax incentives introduced by countries to stimulate economic growth and discuss relevant criticisms of such R&D tax incentives for giving rise to international harmful tax regimes. The article focuses on the initiatives taken by the G20 and the OECD, together with the European Union, to address the counterproductive repercussions that the increased leniency of national tax regimes may have on international stability of revenues and profit shifting. It especially concentrates on the Action Plan[1] of the OECD/G20 Base Erosion and Profit Shifting (BEPS) initiative, with the aim of deciding whether these policy measures will have a serious effect on the survival prospects for these potentially harmful tax regimes. Despite any evidence to the contrary, the article maintains that tax incentives should be maintained constructively with regard to the policy discourse on international business and taxation, with enhanced attention being given to issues of real economic substance, thereby establishing a nexus between the development of the valuable intangibles and their exploitation.

2. R&D Tax Incentives and Tax Policy

2.1. Categories of R&D tax incentives

2.1.1. Opening comments

R&D has been clarified as “work undertaken on a systematic basis to increase the stock of knowledge, including knowledge of man, culture and society, and the use of this stock of knowledge to devise new applications”.[2] In practice, R&D can be classified into the following three subcategories: (1) basic research; (2) applied research; and (3) experimental development. Normally, the R&D value chain commences with the early stage of deliberation to create ideas, i.e. basic and applied research, and progresses to the development of valuable intangible assets, such as copyrights, patents and trademarks.[3] Having said that, basic and applied research can generally be considered to be riskier than experimental development if the ensuing financial returns are taken into account.[4] In addition, R&D tax incentives are provided either as “incremental-based”, i.e. the tax benefit is provided to the additional amount of R&D expenses incurred, or “volume-based”, i.e. the tax benefit is provided to the whole amount of the R&D cost of each year, where the latter might be considered to be costlier, in terms of foregone tax revenue, and less effective compared to the former. In contrast to the United States, most international jurisdictions prefer to legislate for volume-based tax incentives, both in terms of simplicity and in avoiding the stop-and-go distortion of the incremental schemes.[5] Finally, R&D tax incentives can be effectively categorized into “input” and “output” tax incentives (see sections 2.1.2. and 2.1.3., respectively).

2.1.2. Input tax incentives

Essentially, input tax incentives refer to the actual costs incurred on both current expenditure and capital expenditure, with capital expenditures being less important than current expenditures, as the former typically represent only 10% to 13% of business expenditure.[6] In particular, current expenditure is incurred largely on labour and supplies relating to R&D activities.[7] On the other hand, output tax incentives broadly refer to the income side and are provided during the exploitation phase of the intellectual property (IP).

From the end of the 20th century onwards, countries have introduced various tax incentives on the cost side using three different methods. First, a tax system can provide an accelerated or even full allowance in respect of a capital R&D asset that would otherwise have resulted from capital asset depreciation based on its useful life. Consequently, the accelerated depreciation reduces the tax base in the early years. Second, a super deduction or extra allowance may be provided on the R&D cost incurred – usually on current expenditures – such as, for example, 60% more than the actual cost, thereby shrinking the tax base. Third and finally, a tax credit may be provided to reduce the tax payable, which is commonly provided by states in respect of current expenditure.[8] In effect, tax credits constitute a more preferable tax policy tool compared to enhanced allowance, as, if there is a change in tax rates, tax credits remain the same, whereas enhanced allowance gives rise to a proportionate change in the tax base.[9] Moreover, tax credits are more attractive to start-ups and small and medium-sized enterprises (SMEs), which have high cash needs, compared to multinational enterprises (MNEs), which have better access to finance and substantial profits within the group to utilize any losses.

2.1.3. Output tax incentives

In contrast, output tax incentives have been introduced by many countries to encourage the full exploitation of the newly created valuable intangibles and to attract foreign mobile capital. In this respect, many European countries have introduced IP box regimes,[10] in respect of which the design varies from country to country. As a result, differentiated IP box regimes can reduce the taxes payable on the licence income received by the owner of the IP, either by diminished tax rates or contracted tax bases, causing the effective tax rates (ETRs) to differ greatly among countries, such as in the case of the ETRs in Malta and France, which are 0% and 15.5% respectively.[11] In addition, these regimes also vary in terms of eligibility of the intangible assets that states can include in their IP box regimes. In the case of the Netherlands and the United Kingdom, for example, IP box regimes are limited to patents, whereas in other countries, like Cyprus and Luxembourg, these regimes extend added benefits to marketing intangibles as well. Furthermore, some countries, such as Luxembourg and the United Kingdom, extend benefits to notional royalties, whereas others limit their benefits to royalty income only. Finally, some countries, like Belgium and the Netherlands, have limited their benefits to self-developed IP assets, while others, such as Cyprus and Luxembourg, have extended benefits to acquired IP assets.

2.2. Tax policy considerations on R&D tax incentives

Following Adam Smith’s paradigm regarding simplicity, efficiency, equity and neutrality for a “good tax system”, James Mirrlees recommended that tax policymakers introduce viable tax incentives into their policy considerations so as to realize sustainability in respect of the proposed metamorphosis of tax governance.[12] Remembering that R&D activities are generally not undertaken by all taxpayers, it appears that any related tax incentives would not promote a sustainable tax system by adhering to Adam Smith’s principles, in particular, equality and ability to pay. Moreover, some might argue that R&D tax incentives can result in business decisions that disregard the concept of neutrality, which, in turn, may change negative present value projects to positive ones, giving rise to distortions in a tax system.[13],[14]

Tax policymakers may also be tempted to subsidize R&D activities and use other related tax policy tools to stimulate innovation and foster growth and productivity. The main reason for this practice is that countries may be more willing to support these kinds of projects as they have positive rent or knowledge spillovers and produce other increased social benefits.[15] Unlike governments, companies underinvest in R&D projects whose returns are lower than costs, as they do not seem to care if there are social benefits and positive spillovers in these activities.[16] In addition, SMEs and start-ups that invest in innovation tend to face higher interest rates and more limited access to capital, compared to large MNEs, due to possible asymmetric information in capital markets and market failures. Accordingly, countries may seem more willing to support projects, where SMEs and start-ups would not, that have positive externalities and spillover effects by providing R&D tax incentives.[17] As a result, the private sector might undertake projects with a higher social rate of return, as the tax incentives would make the project profitable. Similarly, as IP is highly mobile, R&D tax incentives can also attract foreign investment in the local economy, increasing foreign direct investment and innovation.[18]

Recently, the R&D tax incentives provided by some governments have been the subject of severe criticism. To begin with, it appears to be very difficult to quantify the marginal spillover effects with regard to the marginal cost needed to subsidize the project due to the limited access to data available by companies for the companies to be willing to undertake a project.[19] Furthermore, R&D tax incentives may be poorly targeted, leading to the subsidizing of R&D activities that would have been undertaken anyway and resulting in “substitution”.[20] In addition, IP assets, being highly mobile, have been used extensively by MNEs for tax planning purposes to shift profits to low-tax jurisdictions and effectively pay less tax.

In contrast, though, it is widely accepted that long-term tax incentives compared to short-term tax incentives tend to produce higher social benefits and positive spillovers, especially in the early stages of SMEs, resulting in countries being generous in providing benefits in respect of R&D tax incentives.[21] Consequently, policymakers need to assess and increase the possibility for providing R&D tax incentives on R&D activities where the social benefit is greater compared to the private benefit.[22]

2.3. Input and output tax policy choices and IP tax incentives

There are many reasons countries use tax incentives to promote R&D. First, input tax incentives can be easily traced, as they can be quantified based on the costs incurred in respect of R&D activities, thereby ensuring that they directly reduce the expected rate of return on a given investment. As a result, input tax incentives are assumed to be simple and effective tax policy tools.[23] On the other hand, in terms of efficiency, input tax incentives are not always easy to quantify, as it is difficult to assess their effect on subsidized projects because these tax incentives may or may not influence investment decisions, depending on business circumstances.[24] Moreover, input tax incentives may be poorly targeted, as they are likely to be provided in respect of research projects, especially in relation to basic research, regardless of their ultimate success. In practice, positive spillovers may not necessarily be realized, which, in turn, may increase the prospect of inefficiency in providing input tax incentives. In addition, input tax incentives are not as effectively constructive in relation to successful IP assets, as output tax incentives are normally granted to successful IP assets (ex post). Furthermore, especially in EU Member States, companies may realize higher returns, as they can successfully claim tax input incentives on R&D expenditure twice, double dipping in respect of R&D costs, due to weak coordination among the EU Member States.[25]

Second, as SMEs are likely to have limited access to capital markets and, therefore, be liable to higher interest rates in respect of loans compared to large MNEs, input tax incentives can be used by SMEs to increase cost efficiencies and strategic effectiveness. In particular, refund tax credits, in cash or to be set off against other taxes, such as social insurance taxes or VAT, can enhance the cash flow position of SMEs in the early stages of R&D activities, where, typically, SMEs do not realize profits.[26] Nevertheless, especially in the European Union, Member States might face State aid investigations, as the provision of input tax incentives can be regarded as selective, both by providing incentives to a particular class of taxpayers or by providing them with attractive refunds on their taxes.[27]

Third and finally, input incentives are generally provided to companies that only internally develop IP assets and not in respect of developing specialized R&D research centres that indirectly pass on the knowledge to companies when the R&D is outsourced.

On the output side and, in particular, in relation to IP box regimes, countries tend to provide tax incentives for the exploitation of newly successful innovative IP assets, as there are also positive external spillover effects outside the company and into the general economy.[28] Having said that, studies have demonstrated that, even if an IP has developed outside a country either by being acquired or contracted out, that IP may be internally exploited, giving rise to increased productivity and growth and resulting in positive spillovers by exploiting further innovative IP assets.[29] Another advantage is that IP box incentives are less risky, as they are only provided to successful IP assets, thereby saving time. In this context, it should be noted that countries have introduced IP box regimes for a number of reasons. First, to attract foreign investment to their territories so as to increase tax revenue. Second, by attracting foreign investments, which increases employment and growth, such countries attract further new innovative and knowledge IP businesses, whose positive externalities spill over into the country. Third and finally, to deter domestic IP businesses from transferring these activities from one country to another.[30]

Nonetheless, IP boxes have been criticized extensively for being an ineffective tool and, therefore, less effective compared to input tax incentives in promoting R&D for a number of reasons. First, IP box regimes increase administration burdens and overall complexity for both taxpayers and tax administrations, as they have to adopt new necessary standards. Second, it is most likely that IP box regimes will result in a loss of revenue because a reduction in either tax rates or the tax base may be significant in respect of a national economy. (A study in the United Kingdom revealed that, as at June 2010, there was a loss of revenue amounting to GBP 1.1 billion a year.)[31] Third, IP box regimes tend to be poorly targeted, as benefits are granted to new successful IP assets that would be exploited anyway, giving rise to inefficiencies in innovation. Fourth and finally, IP box regimes tend to increase tax competition among countries, creating a race to the bottom regarding tax revenues.[32]

That being said, many IP box regimes, which have been introduced, do not require any substance and R&D activities in the residence state. In addition, MNEs tend to centralize their IP assets in an IP holding company for their effective administration.[33] Consequently, MNEs tend to structure their activities, on the one hand, by placing their IP assets in a low or no-tax jurisdiction, usually one with an extensive treaty network so as to reduce the withholding tax paid on royalties in the source jurisdiction and, on the other hand, by locating their R&D activities in a high-tax jurisdiction. As a result, the income in the low or no-tax jurisdiction remains untaxed, while the expenses in the high-tax jurisdiction result in a greater reduction of the tax base compared to a low-tax jurisdiction. MNEs have often been accused of using aggressive tax planning structures such as the “double Irish with a Dutch sandwich” structure or a Luxembourg IP holding company interposed with back-to-back licensing arrangements to realize their aims, such as base erosion and profit shifting.[34]

3. IP Boxes, the OECD and the European Union

3.1. Introductory remarks

Following the negative publicity attracted by MNEs such as Apple, Google and Starbucks for not paying their “fair share” of tax, the G20 authorized the OECD to suggest measures to counter tax planning schemes. In response, the OECD, in collaboration with the European Union, has developed the BEPS Action Plan in respect of the OECD/G20 BEPS initiative and, in October 2015, introduced 15 Actions that “should provide countries with domestic and international instruments that will better align rights to tax with economic activity”.[35] Specifically, the OECD/G20 BEPS initiative does not deal with input tax incentives but, rather, further extends its scope to harmful tax planning schemes, thereby focusing especially on IP structuring. Accordingly, these measures can essentially be either domestic or treaty anti-avoidance rules.

3.2. The “nexus approach” and the new substance requirements under Action 5 of the OECD/G20 BEPS initiative

3.2.1. The proposed new substance requirements

Action 5 of the OECD/G20 BEPS initiative revamped the OECD’s work in the Report on Harmful Tax Competition of 1998,[36] which proposed that the substantial activity requirement is one of the key factors in deciding whether a regime may be considered to be potentially harmful.[37] Effectively, the OECD proposed the “nexus approach”, which, in essence, provides that, for an IP box scheme to apply, there would have to be a nexus and economic substance in the residence state that provides the regime, on the one hand, and the R&D activities giving rise to the relevant IP in the same state, on the other.[38] Moreover, it would take into consideration expenditure that related to any input R&D activities in measuring these activities. In particular, a “nexus ratio” would be used to calculate the income receiving the tax benefit, which would be equal to [(qualifying expenditure + up-lift expenditure)/total expenditure]  overall IP income.[39] Such a situation would ensure that income from the exploitation of the IP, such as royalty income, would have a direct link with R&D expenditures incurred on that particular IP, as the latter would be used to calculate the former. In brief, this income would be considered to be non-harmful and ensures that countries with such regimes would use to stimulate the development of new IP assets and not to attract highly mobile IP resulting in harmful tax competition.

Entitlement would also limited to “qualified assets”, i.e. IP assets and patents, which would be “functionally equivalent to patents”, excluding marketing IP assets, such as trademarks.[40] In this way, expenses would be limited to “qualifying expenditure”, which would be first expenses incurred only by the taxpayer, which would be directly connected to the IP asset, and second, expenses incurred by outsourcing to unrelated parties. Nevertheless, a 30% “up-lift” would be permitted under this formula.[41] Finally, the burden of proof would lie with the taxpayer in calculating the income correctly and the tax administration in verifying such a proof, as the proof would treated as a rebuttable presumption.[42]

3.2.2. Evaluation of the nexus approach

The “nexus approach” has numerous results. First, prior to the OECD/G20 BEPS initiative, IP box regimes had been criticized as not encouraging taxpayers, in particular SMEs, to develop new innovative IP assets and for being available for acquired intangibles or intangibles that existed before the IP regimes were introduced.[43] With the nexus approach, all of these deficiencies would be reduced significantly, as there would have to be a direct nexus between the R&D activities incurred by the taxpayer and the IP income. As a result, it would be difficult for an IP asset to be transferred to a low-tax jurisdiction. In addition, if an IP asset were moved to such jurisdiction so as to be eligible for benefits, it would have to undertake R&D activities in that country in relation to that IP.[44] Similarly, in relation to tax planning techniques used by MNEs, if the location of the IP assets differed from the country in which the R&D activities were carried out, typically contracted out to related parties, there could be a ceiling of 30% on the uplift on the expenses incurred by the related party.[45]

On the contrary, with regard to IP box regimes that were compliant with the OECD/G20 BEPS regimes, simplicity would be undermined for the taxpayers, especially because every year such taxpayers would be required to calculate the income qualifying for the tax benefit and keep track of all years cumulatively, i.e. the nexus approach would be an additional approach.[46] In a similar vein, simplicity would be reduced for tax administrations, which would also have to review and monitor these calculations annually. Accordingly, in order to be efficient, the positive spillovers, together with any increase in tax revenue, would have to be higher than the administration costs and any loss of revenue from the existing regimes combined. Equally, problems would arise in terms of how to ascertain that the income exploited was directly linked to R&D expenditure.[47] A similar deficiency would tend to arise, as the nexus approach focuses on the output side rather than on the input side, thereby following the merits of volume-based incentives as opposed to incremental-based incentives (see section 2.1.). In addition, the nexus approach would remain inefficient in relation to SMEs incurring R&D expenditure, as it would be an output-based rather than an input-based approach.

Second, IP box regimes before the OECD/G20 BEPS initiative, particularly those that did not require any substantial or economic activity to be undertaken in a resident state, were criticized for not deterring domestic IP assets from relocating to another jurisdiction or attracting foreign R&D activities.[48] On the contrary, IP box regimes that would comply with the OECD/G20 BEPS initiative, would counter, in general, the relocation of domestic R&D activities to a foreign jurisdiction, as the IP exploitation income would be restricted by the nexus ratio that would require a link between the R&D activities and the creation of the IP asset to be exploited. Similarly, the nexus approach would retain the R&D activities and the IP asset exploitation in the same country, permitting countries with generous R&D tax incentive regimes to attract investments to their jurisdiction. As a result, the 30% uplift in the nexus approach would make it more flexible and attractive to outsource some R&D activities, such as basic research, which might not be the case in the resident state.

Third, another reason that pre-BEPS IP box regimes were introduced by policymakers was to attract foreign IP assets that would be exploited domestically, stimulating the direct use of IP assets. As studies have indicated, this would tend to increase knowledge-based capital and spillovers in the residence state.[49] Notably, nexus compliant IP box regimes would not apply in relation to newly created and then acquired IP assets, as the nexus ratio would restrict the income giving rise to the benefit.

Fourth and finally, policymakers introduced IP box regimes before the OECD/G20 BEPS initiative to attract mobile capital, increasing tax revenue and enhancing growth in specific service sectors, such as the provision of services by accountants, bankers and lawyers. This situation is particularly relevant to countries with small internal markets that are ring-fenced. In contrast, in larger markets that are not ring-fenced, the revenue forgone from reducing the ETRs on existing businesses could be significantly higher compared to any additional tax revenue that might derive from attracting new and mobile foreign capital.

3.3. Other protective tax measures for IP box regimes

3.3.1. Controlled foreign company rules and Action 3 of the OECD/G20 BEPS initiative

In principle, Action 3 of the OECD/G20 BEPS initiative, which consists of six building blocks, would effectively strengthen controlled foreign company (CFC) rules.[50],[51] Normally, CFC rules require the resident state to tax undistributed profits derived in the source state where the latter has a significantly lower ETR, such as an IP box regime, sheltering the profits realized there. In this sense, despite the fact that IP box regimes could easily meet the low-tax threshold requirement, certain exceptions could be applied in relation to widely agreed international standards, such as nexus compliant IP box regimes. Similarly, the Anti-Tax Avoidance Directive (2016/1164)[52] has recommended the implementation of CFC rules in all of the Member States, which, in general, are similar to those proposed in Action 3. However, the Anti-Tax Avoidance Directive (2016/1164) includes a carve out provision that makes the CFC rules inapplicable when the CFC in question “carries on a substantive economic activity supported by staff, equipment, assets and premises, as evidenced by relevant facts and circumstances”,[53] which effectively is a similar requirement of a nexus compliant IP box regime, and means that article 7 of the Anti-Tax Avoidance Directive (2016/1164) only restrictedly applies in the European Union.[54] In contrast to the Member States, this rule does not apply to third countries that permit their tax authorities to tax royalty income, even if there is substantial economic activity.

3.3.2. The transfer pricing effect of Actions 8-10 of the OECD/G20 BEPS initiative

Another area that invites serious contention is the controversial issue of IP ownership and IP income entitlement that is addressed in Actions 8-10 of the OECD/G20 BEPS initiative.[55] The aim of Actions 8-10 is to ensure that the transfer and use of IP between related parties are properly allocated in line with value creation. Effectively, Actions 8-10, among other things, provided guidance on the definition of intangibles and, therefore, would try to give rise to certainty for both taxpayers and tax authorities by uniformly using a cross-border definition of what constitutes intangibles. Furthermore, the ownership of the intangibles would not be limited to the legal owner, but could also be assigned to the economic owner. Accordingly, the owner of the intangible, whether the legal or economic owner, could be entitled to IP income depending on a functional analysis for assets used, risk assumed and functions performed. In the context of such a functional analysis, with regard to IP, for an entity to be entitled to IP income, the entity should have the “financial capacity” to assume the risks and at the same time be in a position to bear and control those risks associated with the development, enhancement, maintenance, protection, and exploitation (DEMPE) functions.[56] Consequently, for example, in an extreme case, it could be assumed that an entity that is the legal owner of an IP asset has licence income from its related entities. It could further be assumed that such an IP entity does not perform any functions to control risks or have the financial capacity to assume those risks and, therefore, is only entitled to a risk-free rate of return. As a result, it appears to be likely that taxpayers arrange both their contracts and their substance requirements in line with entrepreneurial risk according to the DEMPE functions.

Nonetheless, in contrast to Action 5 of the OECD/G20 BEPS initiative, in respect of which the substance requirements are effectively in line with the location of the scientists who are basically developing the IP, the substance requirements regarding the recommendations of Actions 8-10 are largely based on risks. In other words, in order for taxpayers to satisfy the new risk approach, they would only need a few high-calibre individuals to bear and control the risks and, at the same time, a cash-rich IP entity that would normally have the financial capacity to assume those risks as well. Consequently, although under Action 5 the OECD has recognized that the people who are developing an IP, such as scientists, are the most important element of the nexus ratio, this contradicts with the OECD’s position with regard to Actions 8-10, in which the key persons would be a few individuals making key decisions who would therefore be in the position of controlling the risks.

3.3.3. Treaty entitlement and Action 6 of the OECD/G20 initiative

With regard to treaty entitlement, many countries with an extensive treaty network are generally considered to be a tax-favourable jurisdiction for locating IP assets during the exploitation phase in the sense that article 12(1) of the OECD Model[57] gives an unlimited taxing right to the resident state over any royalty income. On the other hand, it restricts the source state in imposing withholding taxes on royalty payments to non-residents. Action 6 proposes two treaty anti-abuse provisions that are also included in the four BEPS Minimum Standards.[58] First, a limitation on benefits (LOB) clause is proposed, which is an objective test and would require certain specific substance requirements to be met for the relevant tax treaty to apply. Second, the principal purpose test (PPT) has been proposed as a more general anti-abuse provision compared to the LOB test. In effect, a tax treaty would apply only if it was “reasonable to conclude” that the “principal purpose” of providing the benefit accorded with the “object and purpose” of the tax treaty. In contrast to the LOB, which would be an objective test, the PPT would be a subjective test and it would therefore give rise to more uncertainty, as it would be difficult to clarify the meaning of the term “principal purpose”.

The PPT would depend largely on the judgement of tax administrators, which would differ from country to country in terms of what would be regarded as “reasonable to conclude”. In this regard, the Commentaries on the OECD Model provide various examples as to whether and when the PPT should be applied. Accordingly, it appears to be likely that the PPT would not apply “where an arrangement is inextricably linked to a core commercial activity, and its form has not been driven by considerations of obtaining a benefit”.[59] It can be argued that the substance requirements of Actions 8-10 could make the PPT inapplicable without taking into consideration the substance requirements of Action 5, as this was the intention of the OECD.[60] However, it can also be argued that combining the substance requirements of both Action 5 and Actions 8-10 would enhance a taxpayer’s position if a tax official reasonably concluded that granting the treaty benefits would not accord with the object and purpose of the tax treaty.[61] Nevertheless, the PPT could still apply and, as a result, deny treaty benefits, even if the substance requirements of both Action 5 and Actions 8-10 were met, as this would not accord with the object and purpose of the tax treaty. Consequently, in contrast to the other protective measures, the PPT is expected to provide further uncertainty, particularly in granting treaty benefits.

Interestingly, the majority of international jurisdictions opted into the PPT test rather than the LOB test when signing the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (the Multilateral Instrument, or MLI).[62] This was a departure from the prevailing mode of thinking in the United States, which did not adopt the PPT in the US Model (2016).[63] Similarly, article 6 of the Anti-Tax Avoidance Directive (2016/1164) includes a general anti-abuse rule (GAAR), which is similar to the PPT. As a result, treaty shopping and tax planning techniques that favour the shifting of the IP asset to more favourable treaty jurisdictions are limited extensively or may even be ending following the adoption of these rules.

3.3.4. Other treaty anti-abuse provisions

Another treaty anti-abuse provision is included in the US Model (2016) that denies the treaty benefit regarding royalty payments. According to article 12(2)(a) of the US Model (2016), in order to be entitled to the treaty benefit, the recipient and beneficial owner of the royalty payment should not be located in a jurisdiction that is considered to be a “special tax regime”, i.e. a state with a statutory tax rate of less than 15%, on the one hand, or of 60% less than the US statutory tax rate, on the other.[64] Most IP box regimes provide for an ETR below 15%, rendering it non-eligible for treaty benefits. Nonetheless, an exemption applies when R&D activities are undertaken in contracting states, highlighting, in essence, the nexus approach of IP compliant regimes. Equally, a similar provision has been included in the Commentaries on the OECD Model (2017), under which contracting states can deny treaty benefits with regard to either interest or royalties paid to a person resident in the other contracting state that has a special tax regime where a nexus compliant IP box would not be considered to be a special tax regime.[65] Interestingly, no OECD member country nor non-OECD jurisdiction has made an observation in this respect and, therefore, it can be argued that a nexus compliant IP box is widely considered to be non-harmful by various countries. Similarly, the proposed amendment to the Interest and Royalties Directive (2003/49)[66] effectively mandates that the reduction in the withholding tax to zero of the Member State in which the royalty expense is paid may apply if the ETR is at least 10% of the Member State in which the income is received.[67] Nevertheless, as this violates the freedom of establishment, the European Union is proposing to combine the Interest and Royalties Directive (2003/49) with a substance test that is similar to a nexus compliant IP box regime.

3.3.5. Linking rules and switch-over clause

Countries are introducing the “linking rules” into their domestic law. For instance, Germany restricts the deduction of royalty expenses, paid to a non-resident company, up to a specified cap where the royalty income is subject to privileged taxation. However, this rule does not apply where such privileged taxation is considered to be nexus compliant.[68] For this measure to be effective, it must be applied by a large number of countries. Finally, countries tend to include switch-over clauses in the tax treaties that they conclude when providing double tax relief to permit themselves to be able to switch from the exemption method to the credit method. This may be the case where the source state is likely to impose taxes on the income at a low level, presumably an IP box regime, thereby allowing the income to flow untaxed to the resident state.

4. Present and Prospective Influences

The reactions of countries to the release of the Final Reports of the OECD/G20 BEPS initiative vary. To begin with, there are countries that have amended their national laws extensively to comply with the recommendations of Action 5, such as Cyprus and Luxembourg. Cyprus, for example, has achieved the required compliance standards by modifying its existing IP box regime, introduced in 2012, which provided an 80% exemption on gross royalty income. In particular, Cyprus has also moved to limit the eligible assets in respect of trade intangibles by excluding marketing intangibles and by introducing the nexus ratio into the equation, together with the relevant limitations of the qualifying expenditures according to Action 5.[69] Similarly, Luxembourg, which since 2008 has had a similar regime to that of Cyprus, achieved full compliance with Action 5 by excluding marketing intangibles from eligible assets and by incorporating the nexus ratio into their internal law.[70] As a result, both countries can retain their ETRs, i.e. Cyprus at 2.5% and Luxembourg at 5.2%, with regard to their IP box regimes. On the other hand, other countries with regimes that are similar to the Action 5 proposals have made limited changes into their laws by introducing only the nexus formula. Cases in point are the Netherlands[71] and the United Kingdom,[72] as both countries have maintained their ETRs of 7% as of 1 January 2018 (previously 5%) and 10%, respectively.

Furthermore, there are countries that have repealed their existing IP box regimes, such as Liechtenstein,[73] and countries that have alternatively introduced a new IP box regime, like India in 2016. India’s IP box regime has been criticized for not following the Action 5 proposals, despite the fact that India’s IP box regime came into effect after the Action 5 proposal had been published.[74] In particular, as only patents are eligible, India excludes assets that are similar to patents, such as computer software.

India also excludes from eligible persons any non-resident company with a permanent establishment in its jurisdiction, as well as any new owner of the IP asset as the Indian regime only applies to original owners disregarding any subsequent transfer of ownership rights. Moreover, any capital gain resulting from the sale of the IP asset may also not benefit from the reduced rate. Finally, there is no definition of qualifying expenditures as applied in the nexus ratio.

The European Union is committed to requiring countries to adopt the recommendations of Action 5 either directly or indirectly. On the one hand, the European Union, as a supranational organization, requires Member States to adopt the Action 5 proposal directly through the influence of the Commission as reported in the EU Code of Conduct Group (Business Taxation). Judging from the final report issued in October 2017, of all Member States, only France had failed to fulfil any of the required criteria and Italy’s regime appeared to be harmful, as it extended its IP box regime to trademarks.[75] On the other hand, in December 2017, the European Union published a list of blacklisted countries, including those that do not appear to comply with certain criteria such as the BEPS Minimum Standards that include the Action 5 proposal.[76] As a result, 17 countries have been blacklisted for not cooperating with the European Union and another 47 countries have committed to amending their regime following discussions with the European Union. However, the reaction of several countries was immediate. From January 2018, some jurisdictions have been removed from the blacklist, as they have begun to amend their regimes. Similarly, more than 110 countries globally have committed to adopting the findings of the OECD/G20 BEPS initiative, and particularly the four BEPS Minimum Standards that include the Action 5 recommendations.

Briefly, R&D tax incentives, following the nexus approach, are no longer to be distinguished in relation to input and output tax incentives, as income and expenditures must be considered together. This generally applies to both Member States and third countries that follow EU and OECD policies as noted previously in this section. Nonetheless, other countries may introduce R&D tax incentives to the extent that such tax incentives do not conflict with established EU policies on the subject, as they try to trade with the European Union and other third countries that follow EU and OECD policies. According to the latest progress update of the OECD assessment review, which comprises more than 120 member jurisdictions of the inclusive framework, apart from one exception, all of the other IP regimes have now been either amended or abolished so as to comply with the nexus approach.[77] As a result, on the one hand, policymakers must consider both types of incentives together when introducing R&D tax incentives as this might result in either negative ETRs or in providing a double subsidy on the identical R&D activities, as taxpayers may claim both.[78] Furthermore, in the context of the nexus approach, countries that already have an IP box regime most likely will have the following two options: (1) to comply with the nexus approach so as to not be considered a harmful tax regime, thereby avoiding being backlisted; or (2) abolish their regime, as the other contracting state, i.e. the source state, will not be willing to grant treaty benefits in such circumstances (see section 3.2.).

On the other hand, taxpayers will have to consider either the location for IP asset exploitation in advance, as it will no longer be beneficial to move their IP assets to another jurisdiction as under the old regime, or enhance their presence and substantial activities in respect of the existing IP assets situated in a tax-favourable jurisdiction. Undoubtedly, such substance requirements will need to consider: (1) the people that are developing the IP, such as the scientists, to comply with Action 5 and the nexus approach; (2) the people that are taking the risk and the decisions in respect of the royalty income in order to comply with Actions 8-10 of the OECD/G20 BEPS initiative and the DEMPE functions; and (3) the objective LOB test and the subjective PPT. Moreover, MNEs will be interested in a jurisdiction with simple bookkeeping procedures and rebuttable presumptions.

In light of the foregoing, countries with low or no-royalty income tax regimes that are currently nexus compliant will largely be considered as not being harmful and, therefore, comply with both EU and OECD tax policy. Accordingly, bearing in mind that a large number of countries are committed to following the EU and OECD minimum standards, in the author’s opinion, the nexus approach will largely endorse the harmonization of the IP box regimes worldwide, particularly where that involves cross-border royalties. In this regard, the author also believes that nexus compliant IP regimes will continue to exist because additional protective measures, which countries are introducing with regard to royalty income (see section 3.2.), are providing exemptions to nexus compliant regimes. Examples of these protective measures are the implementation of linking rules in Germany, and the denying of treaty benefits where countries have a special tax regime, in the United States. Interestingly, this view appears to be in line, to a large extent, with both the EU Member States under the proposed amendment of the Interest and Royalties Directive (2003/49) and among OECD member countries and non-OECD countries (see section 3.2.).

Consequently, remembering that IP box regimes will continue to exist in a largely harmonized context, this development will lead countries to compete on input tax rather than on output tax incentives. In this sense, the author believes that the most attractive jurisdictions will be those that provide the best combination of a low ETR and, at the same time, generous input tax incentives, minimizing the IP development risk in the early stages of research, such as is the case for tax credits with refunds.[79] As a result, countries introducing such measures will provide further tax incentives to their taxpayers, enabling them to be more attractive. In addition, the implementation of tax incentives on payroll cost, such as in Belgium and the Netherlands, may appear to be attractive.[80] In sum, in order to encourage innovation and attract IP assets, countries will be more willing to provide input tax incentives by focusing on the basic and applied research to minimize IP development risk, resulting in a race to the bottom. Conversely, in order to prevent countries from entering into such a race, R&D direct subsidies could be made more efficient by providing them to universities and similar research institutions that perform basic research.[81] Finally, such direct subsidies are generally considered to be more targeted compared to those for the general commercial and industrial sectors.

5. Conclusions

In practice, it is difficult for anyone to be granted the benefit of IP box regimes without both undertaking the related R&D activities and controlling and assuming the risks in the residence state. Ideally, both policymakers and taxpayers should consider input and output tax incentives with regard to the location of their IP asset prior to making a decision. Furthermore, countries have also strengthened their anti-avoidance provisions when it comes to income arising from IP assets, such as CFC, LOB and PPT rules, washing out any IP box benefits. As a result, it will be hard for countries to attract highly mobile foreign capital. Similarly, it will be difficult for local mobile capital to move abroad. However, all these defensive measures largely exclude nexus compliant IP box regimes and, therefore, substance and R&D activities are of great importance for an IP box regime to apply. Consequently, countries may now be competing in respect of R&D tax incentives that minimize the risks involved in IP development, the most efficient of which are cash refund tax credits and tax incentives on payroll costs.

[1].             See OECD, Action Plan on Base Erosion and Profit Shifting (OECD 2013), International Organizations’ Documentation IBFD, also available at

[2].             OECD, Frascati Manual 2002: Proposed Standard Practice for Surveys on Research and Experimental Development p. 30 (OECD 2002).

[3].             P. Palazzi, Taxation and Innovation, OECD Taxn. Working Paper No. 9, p. 6 (OECD 2011).

[4].             S.E. Shay, J.C. Fleming & R.J. Peroni, R&D Tax Incentives: Growth Panacea or Budget Trojan Horse?, 69 Tax L. Rev. 3, p. 444 (2016).

[5].             I. Guceri & L. Liu, Effectiveness of fiscal incentives for R&D: quasi experiment evidence, Working Paper Series (Oxford U. Ctr. Bus. Taxn., 2015), available at

[6].             Shay, Fleming & Peroni, supra n. 4, at p. 428.

[7].             I. Guceri, Tax incentives for R&D, ETPF Policy Paper p. 11 (Oxford U. Ctr. Bus. Taxn. 2016), available at

[8].             R. Vlasceanu, Chapter 8: Intellectual Property Structuring in the Context of the OECD BEPS Action Plan, in International Tax Structures in the BEPS Era: An Analysis of Anti-Abuse Measures sec. 8.3.3., p. 238 (M. Cotrut et al. eds., IBFD 2015), Online Books IBFD.

[9].             R.J. Danon, General Report, in Tax incentives on Research and Development (R&D), International Fiscal Association (IFA), Cahiers de droit international vol. 100A, sec. 3.1.3. (IFA 2015), Online Books IBFD.

[10].            Which are also referred to as innovation box, patent box, licence box and knowledge box regimes.

[11].            P. Evers, H. Miller & C. Spengel, Intellectual Property Box Regimes: Effective Tax Rates and Tax Policy Considerations, 22 Int. Tax & Pub. Fin. 3, p. 6 (2013).

[12].            Å. Hansson & C. Brokelind, Tax Incentives, Tax Expenditures Theories in R&D: The Case of Sweden, 6 World Tax J. 2, sec. 1. (2014), Journals IBFD.

[13].            Danon, supra n. 9, at sec. 2.1.1.

[14].            Evers, Miller & Spengel, supra n. 11, at p. 19.

[15].            Shay, Fleming & Peroni, supra n. 4, at p. 441.

[16].            Guceri, supra n. 7, at p. 3.

[17].            Palazzi, supra n. 3, at p. 48.

[18].            P. Arginelli, Innovation through R&D Tax Incentives: Some Ideas for a Fair and Transparent Tax Policy, 7 World Tax J. 1, sec. 3.3. (2015), Journals IBFD.

[19].            Hansson & Brokelind, supra n. 12, at sec. 4.1.

[20].            Id., at sec. 4.2.

[21].            I. Guceri, supra n. 7, at p. 22.

[22].            Shay, Fleming & Peroni, supra n. 4, at p. 446.

[23].            Arginelli, supra n. 18, at sec. 4.3.1.

[24].            Hansson & Brokelind, supra n. 12, at sec. 4.2.

[25].            Danon, supra n. 9, at sec. 3.4.

[26].            A. Fairpo, Taxation of Intellectual Property 4th edn., p. 182 (Bloomsbury Prof’l 2016).

[27].            Hansson & Brokelind, supra n. 12, at secs. 2.3.-4.

[28].            Id., at sec. 4.

[29].            Arginelli, supra n. 18, at sec. 3.3.

[30].            Id.

[31].            Evers, Miller & Spengel, supra n. 11, at p. 2.

[32].            R. Griffith, H. Miller & M. O’Connell, Corporate Taxes and Intellectual Property: Simulating the Effect of Patent Boxes, Inst. Fiscal Stud., Briefing Note 112 IFS (2010), available at

[33].            Fairpo, supra n. 26, at p. 420.

[34].            Vlasceanu, supra n. 8, at sec. 8.2., pp. 225-229.

[35].            OECD, supra n. 1, at p. 11.

[36].            OECD, Report on Harmful Tax Competition (OECD 1998).

[37].            OECD, Action 5 Final Report 2015 – Countering Harmful Tax Practices More Effectively, Taking into Account Transparency and Substance p. 23 (OECD 2015), International Organizations’ Documentation IBFD [hereinafter the Action 5 Final Report (2015)].

[38].            Id., at p. 24.

[39].            Id., at p. 25.

[40].            Id., at p. 26.

[41].            Id., at p. 27.

[42].            Id., at p. 35.

[43].            See sec. 2.

[44].            P. Arginelli, Chapter 5: The Interaction between IP Box Regimes and Compensatory Tax Measures: A Plea for a Coherent and Balanced Approach, in EU Law and the Building of Global Supranational Tax Law: EU BEPS and State Aid sec. 5.4.1., p. 109 (D. Weber ed., IBFD 2017), Online Books IBFD.

[45].            OECD, Action 5 Final Report (2015), supra n. 37, at p. 27.

[46].            Id., at p. 29.

[47].            Arginelli, supra n. 44, at sec. 5.3.2., p. 106.

[48].            See sec. 2.

[49].            Id.

[50].            OECD, Action 3 Final Report 2015 – Designing Effective Controlled Foreign Company Rules (OECD 2015), International Organizations’ Documentation IBFD.

[51].            Arginelli, supra n. 44, at sec. 5.6., p. 126.

[52].            Council Directive (EU) 2016/1164 of 12 July 2016 Laying Down Rules against Tax Avoidance Practices that Directly Affect the Functioning of the Internal Market, arts. 7 and 8, OJ L 193 (2016), EU Law IBFD, [hereinafter the Anti-Tax Avoidance Directive (2016/1164)].

[53].            Id., at art. 7.

[54].            Arginelli, supra n. 44, at sec. 5.6., p. 124.

[55].            OECD, Actions 8-10 Final Report 2015 – Aligning Transfer Pricing Outcomes with Value Creation (OECD 2015), International Organizations’ Documentation IBFD.

[56].            R. Collier & J. Andrus, Transfer Pricing and the Arm’s Length Principle After BEPS p. 209 (Oxford U. Press 2017).

[57].            Most recently, OECD Model Tax Convention on Income and on Capital (21 Nov. 2017), Models IBFD.

[58].            Vlasceanu, supra n. 8, at sec. 8.3.2., p. 235.

[59].            OECD Model Tax Convention on Income and on Capital: Commentary on Article 29, para. 182, Example M (21 Nov. 2017), Models IBFD.

[60].            R.J. Danon, Chapter 2: Intellectual Property (IP) Income and Tax Treaty Abuse: Relevance of BEPS Actions 5 and 8-10 for the Principal Purpose Test, in Taxation of Intellectual Property under Domestic Law, EU Law and Tax Treaties sec. 2.4, p. 26 (G. Maisto ed., IBFD 2018), Online Books IBFD.

[61].            Id., at sec. 2.4, p. 27.

[62].            Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (7 June 2017), Treaties IBFD [hereinafter the Multilateral Convention or MLI].

[63].            US Model Tax Convention on Income (17 Feb. 2016), Models IBFD.

[64].            Arginelli, supra n. 44, at sec. 5.6.4., p. 129.

[65].            Para. 85 OECD Model: Commentary on Article 1 (2017).

[66].            Council Directive 2003/49/EC of 3 June 2003 on a Common System of Taxation Applicable to Interest and Royalty Payments made between Associated Companies of Different Member States, OJ L 141 (2013) (as amended through 2013), EU Law IBFD.

[67].            See also Arginelli, supra n. 44, at sec. 5.6.3., p. 127.

[68].            M. Valta, Chapter 1: Taxation of Intellectual Property (IP) in Domestic Tax Law, in Maisto ed., supra n. 60, at sec. 1.3, p. 13.

[69].            N. Neofytou, The IP Box Regime in Cyprus Has Now Been Amended pp. 1-3 (Redimus Publications 2016).

[70].            R.O. Hoor, K. O’Donnell & M. Bentley, The New Luxembourg IP Tax Regime: Opportunities for the Post-BEPS Era pp. 3-5 (Bloomberg BNA TPIR 2017).

[71].            D. Oosterhoff & B. de Nies, Evaluation of the Innovation Box, 23 Intl. Transfer Pricing J. 6 (2016), Journals IBFD.

[72].            Fairpo, supra n. 26, at p. 236.

[73].            European Commission, EU Code of Conduct Group (Business Taxation) (Report to the Council of the European Union 2017), available at

[74].            P. Jain, India’s New Patent Box Regime, 82 Tax Notes Intl. 13, pp. 1281-1283 (2016).

[75].            European Commission, supra n. 73.

[76].            Council of the European Union, The EU list of Non-Cooperative Jurisdictions for Tax Purposes (2017), available at

[77].            OECD, Harmful Tax Practices – 2018 Progress Report on Preferential Regimes Inclusive Framework On BEPS: Action 5 (OECD 2019).

[78].            Danon, supra n. 9, at sec. 3.2.

[79].            C. Shi, IP Boxes in Light of the BEPS Project and EU Law – Part II, 56 Eur. Taxn. 9, sec. 4. (2016), Journals IBFD.

[80].            Vlasceanu supra n. 8, at sec., p. 242.

[81].            Shay, Fleming & Peroni, supra n. 4, at p. 455.

The article was published in the IBFD Bulletin for International Taxation, 2019 (Volume 73), No 5.

Christos Theophilou
T: +357 22 875 720