DISCUSSION PAPER MTI Global Practice No. 16 October 2019 Colin Clavey Jonathan Leigh Pemberton Jan Loeprick Marijn Verhoeven This series is produced by the Macroeconomics, Trade, and Investment (MTI) Global Practice of the World Bank. The papers in this series aim to provide a vehicle for publishing preliminary results on MTI topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For information regarding the MTI Discussion Paper Series, please contact the Editor, Tito Cordella, at tcordella@worldbank.org. © 2019 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved ii MTI DISCUSSION PAPER NO. 16 Abstract While significant progress has been made in international tax cooperation, the interests of developing economies require greater priority and attention. This paper considers policy options that are currently under consideration in different international fora, drawing on the WBG’s work in supporting tax policy and administration reforms in developing economies. The overall direction of the reform debate is promising and could shift taxing rights towards market jurisdictions, as well as strengthen the tools available to deter profit shifting to low- taxed entities. Some of the proposals under discussion could be adopted in the near-term but need to be better tailored to developing economy needs. This paper identifies five areas for improvement. First, the proposals relating to the allocation of global non-routine profits need to be simplified and incorporate more formulaic approaches. Second, work on the allocation of taxing rights should include in its scope the option to reallocate all non-routine profit, rather than only that part reflecting user value or marketing intangibles. Third, the (income inclusion) idea of targeting lowly taxed profits is sound, but it needs to be a tool for both capital importing and exporting countries. Fourth, detailed guidance is needed on the use of withholding tax as an efficient collection mechanism for source jurisdictions and on the application of mandatory safe harbors with an arm’s length let out. These tools can help ensure efficient administration in developing economies. And, fifth, effective administration requires appropriate access to information. Practical limitations to accessing relevant information for developing economies need to be removed. This should include an amendment to current standards allowing jurisdictions to impose robust domestic filing obligations for CBCR (and/or similar future additional reporting obligations). Corresponding authors: mverhoeven@worldbank.org JEL Classification: F23, H25, H26 Keywords: Domestic Resource Mobilization, International Taxation, Base Erosion and Profit Shifting (BEPS), Transfer Pricing, Profit allocation, Profit diversion, Low-tax Jurisdictions, Minimum Taxation. 3 Executive Summary While significant progress has been made in international tax cooperation, the interests of developing economies require greater priority and attention. The G20 and OECD’s Base Erosion and Profit Shifting (BEPS) Inclusive Framework (IF) published a public consultation document in February 2019, setting out five options for international tax reform. This paper considers these options and five other proposals, from the perspective of developing economies, drawing on the WBG’s work supporting their tax administrations. The BEPS initiative does not fully address the impact of profit shifting by multi-national enterprises (MNEs) on developing economies. The BEPS IF consultation document acknowledges that the 2015 BEPS package did not resolve all challenges to the legitimacy, inclusivity and sustainability of the international tax system. Opportunities for aggressive tax planning remain. It is still possible to endow lowly-taxed entities with sufficient “substance” to justify very high profits. This can be done without making significant real-world changes to an MNE’s operations (and costs are minor compared to tax savings). A globally consistent approach is needed to reduce the risk of tax competition (the “race to the bottom”), and the uncertainty for business that can result from disjointed approaches. Without an internationally accepted approach many policy makers are concerned that taking unilateral action will discourage investment. The growth of digital businesses is placing additional pressure on the existing international tax consensus, resulting in a renewed focus on the value-contribution of markets. While some developing economies represent significant digital markets, their value by comparison to developed markets is smaller. For many countries, activity at the other end of the value chain, predominantly production of raw materials and manufacturing, is a larger part of the economy. The IF reform options seek to address the challenges of both digitalization and ongoing base erosion. They affect traditional business models as well as highly digitalized businesses. Some of the proposed reforms depart from current standards and are potentially beneficial to developing economies. The IF’s willingness to simplify the rules for taxing MNEs and to depart from a strict adherence to the arm’s length principle is a welcome development. The options under consideration are not mutually exclusive. Some build on existing principles, some break new ground, while others are more obviously anti-avoidance rules. Separately, IMF (2019) explores broader options, including greater use of formulary approaches. Reform options must be assessed against lessons from working with resource-constrained tax administrations in developing economies: the need for the rules to be straightforward in their application and the need for transparency. Rules need to be designed to be operable by smaller tax administrations with limited resources and staff with less experience and expertise. More mechanical and formulaic rules, subject to less subjectivity and discretion, are less susceptible to political interference and corruption. Safe harbors, either integrated into the mechanics of the options, or applied in parallel to them, also have an important role in achieving the required simplification. The IF work program should include work to develop mandatory safe harbors for a range of routine activities that countries could apply, subject to 4 an arm’s length let out. Improving access to data is another necessary condition for implementation and of itself improves tax collection and transparency. This will require changes to the current approach to documentation and much readier access to key information about the global activities of MNEs, such as Country-by-Country reports. Options under discussion have varying prospects of achieving consensus and implementation, but a pathway to fundamental reform exists. The ten options considered in this paper are: the user-value proposal, the marketing intangibles proposal, the concept of “significant economic presence”, the extension of withholding taxes and source jurisdiction taxation rights, broad residual profit allocation by formulary apportionment, global formulary apportionment, the destination-based cash flow tax(DBCFT), an income inclusion rule, income inclusion based on a diverted profits rule and a tax on base eroding payments. This paper considers advantages and disadvantages of each option, from the perspective of developing economies, and proposes a way forward, mindful of the obstacles to achieving global consensus. Two of the IF proposals (user-value and marketing intangibles) place more emphasis on the value of the market in the allocation of profits. These options are intended to give jurisdictions the right to tax businesses that have a sustained interaction with their economies, even if they do not have a physical presence there. They do this by deeming part of the non- routine profits of a MNE to be generated by users, or marketing intangibles, and then allocating the right to tax that to the jurisdictions where users or markets are. Achieving consensus on fundamental reform along these lines will be challenging, as the July 2019 meeting of G7 Finance Ministers illustrated, with views differing sharply on the merits of taxes that specifically target digital businesses. The third option, the ‘significant economic presence’ (SEP), is more likely to match the needs of developing economies, provided it has an agreed mechanical formula for the allocation of profits to this ‘presence’. The SEP proposal is focused on highly digitalized businesses and creates a taxing right when a business has a substantial and sustained revenue generation in a country, even if it lacks a physical presence there (possibly by deeming there to be a permanent establishment). The allocation of profits to the SEP could be based on some form of fractional apportionment, and collected, at least in part, by means of withholding taxes applied to payments made for digital goods and services. Withholding taxes could also act as a means of collecting taxes due under the other options that allocate more taxing rights to the source, or user/market jurisdictions. In principle, the SEP proposal has the potential to be the least complex, provided an allocation formula can be agreed. A wider ‘residual profit allocation’ (RPA) approach of the kind considered in IMF (2019) could raise significant additional revenue for developing economies. This is because it would not be restricted to highly digital businesses. It would also be the most effective in reducing administrative complexity and countering profit shifting to low tax jurisdictions , as all of the residual profits of MNEs, irrespective of where they are originally earned, would be allocated to jurisdictions according to a set of factors that are indicative of profit generation, such as assets, labor and sales. The exact metrics and weightings in the formula would need to be agreed and empirical assessments of effects are highly tentative. Because it would radically 5 alter the allocation of taxing rights over MNEs’ profits earned in normal-rate as well as low- tax jurisdictions, it is going to be difficult to achieve consensus on this proposal; the differing reactions to the proposals already tabled for discussion by the IF confirm this. Nonetheless, this option should form part of the discussion on the future of the international tax system, even if it forms part of a longer-term program of reform. The RPA proposal in essence provides for a Global Formulary Apportionment (GFA) for non-routine profits. More common proposals for GFA target all profits for allocation between jurisdictions according to certain factors that are indicative of the contribution made in each location to the generation of those profits. General GFA would address the issues discussed in this paper relating to the complexity and extensive data-requirements of traditional transfer pricing rules based on the arm’s length principle, which create significant implementation challenges for developing countries. However, the difficulties of reaching international consensus make it a long-term solution and a consensus on the slightly narrower RPA approach (albeit suffering from many of the same coordination challenges) may be easier to reach. The DBCFT would replace corporate income with a tax that is similar to VAT but with a deduction for labor costs. It removes many of the distortions in the current system, including the bias towards debt finance, and is proof against profit shifting by MNEs. However, the distributional effects could be large and are subject to significant uncertainty. Consequently, it is difficult to see this securing consensus support in the near-term. While consensus on fundamental reform and more mechanical methods better suited to developing economies is highly desirable, it may be difficult to achieve, in which case the fallback will be greater reliance on anti-abuse rules and/or regional approaches/. The IF proposals include two such rules: the ‘income inclusion rule’ and the ‘base eroding payment rule’ (inspired by elements of the 2017 US tax reform). The approach taken is to be welcomed and both rules are relatively straightforward anti-avoidance measures. They both require tax administrations to have information about the effective tax rates of related parties in other jurisdictions. Under the income inclusion rule, profits of lowly taxed subsidiaries are attributed to the ultimate parent and taxed in the jurisdiction of the parent. This measure is unlikely to directly benefit capital importing developing economies significantly. The ‘base eroding payment rule’ would disallow tax deductions for payments to related parties that give rise to a high risk of base erosion (such as interest or royalties) and that are not subject to a minimum effective rate of taxation. Such a rule could be relatively simple to enforce: taxpayers could be required to self-assess whether such payments meet the minimum effective tax test and information required by tax administrations to enforce the rule should be included in a transfer pricing tax return schedule. This would be a useful addition to the anti-avoidance defenses available to developing economies but could still be undermined by the use of intermediate entities located in normal-rate countries (a form of treaty shopping that may be hard to combat). An additional option is to translate the current income inclusion proposal into a rule designed to meet developing economy requirements based on the concept of ‘diverted profits’. This would counter profit shifting into low tax jurisdictions, offering a relatively simple tool to bring into taxation a share of profit shifted to low-taxed and low-substance entities. It simplifies the vexed issue of what constitutes ‘substance’ and value creation by 6 substituting (in the case of these lowly-taxed entities) a simple rule for the current complex set of technical considerations. If the profits of a lowly taxed entity exceed a certain multiple of costs representative of “substance” (primarily the direct costs of labor) that excess shall be allocated to jurisdictions for taxation in accordance with an agreed metric. There is scope to develop a relatively simple approach that is narrowly targeted, mechanical, with wide application. This and a simplified approach to the definition of substance should be included in the IF work program. All three anti-abuse rules could form a package of fundamental reforms in its own right. Such a package could be ready for immediate implementation, in advance of achieving consensus on wider reforms. The IF could make a recommendation in the near future, similar to the approach taken with the interest deduction ceiling proposed in BEPS Action 4, which could pave the way for the coordinated design and implementation of rules. With prominent precedents for unilateral action by OECD member states, discussions on the reallocation of taxing rules should not slow down efforts to guide the effective use of similar approaches in developing economies. Effective administration of the new rules will require extensive information about MNE operations. The BEPS package has improved access to data by means of a Masterfile and Country-by-Country reports. However, Country-by-Country reports are only prepared by a small number of very large MNEs and the conditions for accessing those reports represent a barrier to access for many developing economy tax administrations. Practical administration of the proposals under discussion requires information contained in Country-by-Country reports, possibly modified to take account of the factors that are agreed as allocation keys under any new system. The current reporting threshold of €750 million will need to be lowered and countries should have direct access to this data by way of local reporting requirements. The international tax rules for MNEs are at a turning point. The solutions must meet the needs of developing economies to maintain the legitimacy, inclusivity and sustainability of the international tax system. New approaches and standards have to reflect the needs of developing economies, both in terms of the operability of the rules, where more mechanical approaches are preferable, greater access to data is supported, and recognizing the position of developing economies in the overall value chains of MNEs. A consensus on the way forward will reduce double taxation and tax competition, provide more certainty to business, help developing economies mobilize much needed resources and achieve a fairer tax revenue distribution. 7 Table of Contents Abstract ...................................................................................................................................... 3 Executive Summary ............................................................................................................................ 4 Table of Contents ................................................................................................................................ 8 Introduction ......................................................................................................................................... 9 International Taxation: Challenges for Developing Economies ....................................................... 12 Tax Administration Capacity ..................................................................................................................... 12 Foreign Direct Investment ......................................................................................................................... 12 Access to Data ............................................................................................................................................ 14 Political Economy ...................................................................................................................................... 15 Complexity and Ongoing BEPS Risks ....................................................................................................... 15 Options for Change ........................................................................................................................... 17 User Value Proposal .................................................................................................................................. 17 Marketing Intangibles Proposal ................................................................................................................ 18 Significant Economic Presence (SEP) ...................................................................................................... 19 Increasing the Scope of Source Taxation Rights and Withholding Taxes ............................................... 21 Residual profit allocation via Formulary Apportionment ........................................................................ 22 Global Formulary Apportionment (GFA) ................................................................................................. 23 Destination Based Cash Flow Tax (DBCFT) ............................................................................................ 24 Minimum Tax on Outbound Investment: Income inclusion Rule ........................................................... 25 Income Inclusion: A Diverted Profits Rule (DPR) ................................................................................... 26 Minimum Tax on Inbound Investment, Tax on Base Eroding Payments (TBEP) .................................. 27 Interaction Between the Proposals .................................................................................................... 28 Implementation ................................................................................................................................. 29 Conclusions ....................................................................................................................................... 29 References ........................................................................................................................................... 1 Annex 1: Business Lines in the Digital Economy and Taxation Issues .............................................. 3 A Sale and Resale of Goods ................................................................................................................... 3 B Sale and Resale of (Access to) Digital Content or Digital Solutions ................................................ 4 C Sale of Advertising and Data .............................................................................................................. 7 D Multi-Sided Platforms – Ride-Sharing & Short-Term Accommodation Rentals ........................... 10 Annex 2: Indirect Tax and Foreign Suppliers of Digital Goods and Services .................................. 13 Annex 3 Tax Administration’s Capacity in a Developing Economy ................................................ 15 Annex 4 Case Illustrations ................................................................................................................ 17 8 International Tax Reform, Digitalization and Developing Economies1 Colin Clavey, Jonathan Leigh Pemberton, Jan Loeprick, and Marijn Verhoeven Introduction The international rules for allocating taxing rights over MNEs may be at a turning point. The immediate driver for change is the challenge posed by digitalization of the global economy. But the options for change currently under consideration potentially affect all MNEs, not just a small number of high-profile technology companies. Crucially, the current debate provides an opportunity to ensure that change addresses the needs of developing economies. International debate is primarily taking place within the G20/OECD Inclusive Framework on Base Erosion and Profit Shifting (IF). This has resulted in a set of proposals for reform that are set out in public consultation document published by the IF in February 2019 and discussed at a public consultation meeting on 13-14 March 2019.2 Shortly before that meeting the IMF published a paper “Corporate Taxation in the Global Economy”3 that provided a high-level overview of the key economic aspects and implications of alternative options, some of which contained features that are common to the options tabled by the IF. Discussions by the IF are ongoing, so the design of the proposals is likely to develop quite rapidly. The purpose of this paper is to contribute to this debate with the needs of developing economies in mind. The paper draws on the WBG’s operational experience in supporting resource constrained tax administrations with the practical challenges of implementing international taxation rules. It assesses the different options tabled by international organizations and academics in that light, recognizing that developing economies are a diverse group but also that there are some common challenges that most of them face to a greater or lesser extent. The paper devotes more time to the analysis of options that have greater potential of securing consensus support within the timeframe set by the IF. Historically, the development of international tax rules has been a topic of relatively limited specialist interest, with standard setting mainly driven by OECD member states. These standards were then adopted to varying degrees by other countries, sometimes implicitly or explicitly recognizing OECD guidelines,4 and sometimes linked to tax treaty negotiations, which can provide an opportunity to impose OECD standards on partner countries.5 Notably, 1 We gratefully acknowledge comments from colleagues at the IMF’s Fiscal Affairs Department and the African Tax Administration Forum (ATAF), as well as Oleksii Balabushko, Anne Brockmeyer, Jessie Coleman, Michael Durst, Martin Hearson, and Emilia Skrok. This paper’s findings, interpretations, and conclusions are entirely those of the authors and do not necessarily represent the views of the World Bank, its executive directors, or the countries they represent. Please address correspondence to jloeprick@worldbank.org. 2 OECD (2019), http://www.oecd.org/tax/beps/public-consultation-document-addressing-the-tax-challenges-of- the-digitalisation-of-the-economy.pdf 3 IMF (2019), https://www.imf.org/en/Publications/Policy-Papers/Issues/2019/03/08/Corporate-Taxation-in-the- Global-Economy-46650 4 Argentina, Colombia and Kenya are examples of countries where the OECD guidelines were used as a judicial reference by the tax administration or the judiciary See: Cooper et al. (2016): Transfer Pricing and Developing Economies, p.47 5 For a detailed discussion drawing on the UK’s experience see: Hearson. M (2018): Transnational expertise and the expansion of the international tax regime: imposing ‘acceptable’ standards. 9 however, while many developing economies have introduced references to international standards such as the arm’s length principle in their tax rules, many have not, or only very recently, introduced the additional implementing regulations required for administration and/or did not invest in the necessary institutional capacity for implementation. Since the international financial crisis in 2008, the taxation of MNEs has received sustained attention from civil society and policy makers. This refocusing of attention resulted in the G20/OECD BEPS project with a program of 15 Actions designed to update the international direct tax system. The proposed measures included elements that are particularly helpful for resource and capacity constrained tax administrations. One example is the recommendation to introduce a simple limit on deductible interest expense based on a percentage of Earnings Before Interest, Taxes, Depreciation and Amortization (EBITDA). This approach is much simpler to apply and more robust than challenges to excessive interest deductions based on transfer pricing principles, or thin capitalization rules6. Wider acceptance of safe-harbors for low-risk transactions is also a helpful simplification of the transfer pricing regime. Indeed, the concept could usefully be expanded, and the use of safe harbors could be made mandatory.7 On the other hand, other aspects added further complexity to an already demanding system and exacerbate the capacity challenge for tax administrations in developing economies. Moreover, as the IF consultation document acknowledges, the final BEPS package did not fully address the challenges to the legitimacy, inclusivity and sustainability of the international tax system. Aggressive tax planning via low tax jurisdictions and related opportunities for base erosion and profit shifting remain. Annex 4 provides illustrations of persistent aggressive tax planning opportunities drawing on recent operational experience of the WBG in developing economies. As emphasized in the World Development Report 2019,8 developing economies need stronger solutions that are commensurate with capacity, and international and regional coordination to prevent tax competition and a race to the bottom. Although the corporate tax reforms in the United States in 2017 were focused on cutting the headline rate, the Tax Cuts and Jobs Act featured some striking measures designed to prevent erosion of the US base. Two of these measures, (Global Intangible Low Taxed Income (GILTI) and Base Erosion and Anti-Abuse Tax (BEAT)) are highly mechanical in nature and impose minimum levels of taxation on elements of international businesses profits. This suggests that there is a readiness to adopt more radical changes than were envisaged in the BEPS project and to place a floor under the rate of corporate taxation internationally. This could help to counter the adverse effects of excessive tax competition. If similar measures are adopted internationally, they will 6 The adoption of the rule and choice of interest ceilings will however often involve a trade-off between the effective protection of the tax base and investment climate considerations. For instance, in developing economies there is often a public interest case in project finance, sometimes involving highly geared projects infrastructure projects that would not necessarily be covered by the Action 4 carve out for public benefit projects. 7 A safe harbor here means that the tax administration specifies an appropriate transfer pricing method and an associated level or range of financial indicators. Use of safe harbors can be made mandatory while providing for an arm’s length let out and reversing the burden of proof to effectively deal with information asymmetries. See PCT (2017), A toolkit for addressing difficulties in accessing comparables data for transfer pricing analyses. 8 WB (2019), The Changing Nature of Work. 10 disincentivize the routing of investments through entities in low tax jurisdictions.9 These low tax jurisdictions are also used for “round-tripping” of investment by wealthy individuals;10 rather than making direct and taxable investments domestically, funds are routed via opaque structures in low tax jurisdictions, avoiding tax and distorting data on foreign direct investment. Such schemes are hard for tax administrations with limited capacity to tackle, even where they have introduced the legislative tools to do so. The ongoing risk of base erosion and profit shifting arises in both highly digitalized businesses and more traditional business models, affecting both ends of the value chain. This is an important consideration for developing economies. While some represent significant markets in their own right, and markets that are increasingly digital, activity at the other end of the value chain, frequently dominated by production of raw materials and manufacture, is often a proportionately more significant part of their economies. Legitimate policy responses need to reflect the needs of developing economies, both in terms of the operability of the rules, where more mechanical approaches are preferable, and in recognizing the position of developing economies in the value chains of MNEs. The business models emerging in the digitized economy are diverse and there is a general recognition that it does not make much sense to talk about a distinct “digital economy”, as all businesses are increasingly digital in some respect. Annex 1 to this note describes these different business models and some of the tax policy responses that have resulted. Some of these are explicitly characterized as interim measures, pending the emergence of an international consensus. This paper is concerned with direct taxation, but it is worth noting the progress made on indirect taxation. There is an international consensus that VAT should be levied by the jurisdiction in which the consumer is located (the destination principle). In the case of services provided by foreign suppliers, whether that is physical goods purchased via a digital platform, or digital goods and services purchased from a foreign supplier, it is important that foreign suppliers are competing on a level playing field with their domestic counterparts, who will have to charge their customers VAT. Fortunately there is now an effective and widely adopted model for requiring foreign suppliers to account for VAT on their sales to local consumers. This model and examples of adoption/implementation by countries11 is discussed in more detail in Annex 2. The remainder of this paper is structured as follows. Particular challenges facing developing economies are discussed in the next section of this paper. This is followed by a review of the options included in the IF consultation document and other prominent contributions in the light 9 Beer and Loeprick (2019) assess costs for Sub-Saharan economies associated with the conclusion of double tax treaties with low-tax investment hubs. The findings suggest that that treaty-shopping drives nominal investment flows and reduces tax revenue across the region 10 See also Balabushko et al (2017) on round tripping in Ukraine. An important share of recorded foreign investment into Ukraine is likely the result of local funds being channeled and subsequently returned back to Ukraine in the form of direct investment 11 For a country assessment of tax policy options see also Record et al. (2018). Malaysia's Digital Economy: A New Driver of Development, Chapter 5. 11 of practical challenges working on international tax issues. To conclude, some prerequisites for effective implementation in developing economies are assessed. International Taxation: Challenges for Developing Economies It is important to recognize that developing economies are diverse and the challenges they face are not uniform. To take an obvious example, the tax issues that are of most concern to resource rich economies are different from those that are priorities for economies that focus on manufacturing. However, there are some common themes and challenges. Tax Administration Capacity International tax issues are among the most complex that tax administrations have to manage. Transfer pricing in particular involves the analysis of complex fact patterns, the assembly of a great deal of information and an understanding of the complex international standards embodied in the OECD’s and/or UN’s transfer pricing guidelines.12 Identifying and successfully investigating cross-border tax issues generally require a good level of commercial understanding, something that is often in short supply, even in advanced tax administrations. International tax specialists take time to train. Their skills are in considerable demand and experienced staff are often difficult for tax administrations to retain. This is a challenge in most countries but particularly in developing economies which often rely on a handful of key specialist officers, and where the gap between public sector salaries and pay rates in advisory firms can be very large.13 To illustrate the challenges facing many tax administrations a pen picture of an administration that has benefited from operational support by the WBG and other development partners is provided in Annex 3. Tax administrations in developing economies can benefit from rules that are relatively simple and certain in application. In part this is because such rules can be administered by smaller tax administrations, with limited resources and staff with less experience and expertise. However, it also reflects the governance challenges facing some administrations. The exercise of discretion and judgement can create opportunities for rent seeking by officials in administrations where corruption is prevalent.14 Transfer pricing typically involves a range of possible outcomes and so necessarily involves the exercise of considerable discretion and judgement. That necessitates both high levels of skill but also specific governance arrangements and attention from senior management.15 Foreign Direct Investment 12 OECD (2017), Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations. 13 For a longer discussion of this issue and some of the steps tax administrations can take to mitigate the problem see Cooper at al. (2016): “Transfer Pricing and Developing Economies”. 14 For a more detailed discussion of corruption in tax administrations in Africa and how technology may help reduce the incidence of corruption see: https://knowledgehub.transparency.org/helpdesk/approaches-to-curbing- corruption-in-tax-administration-in-africa and https://openknowledge.worldbank.org/handle/10986/29862?locale-attribute=en. 15 For a discussion of the governance of transfer pricing programs see Cooper at al. (2016): Transfer Pricing in Developing Economies, p. 346. 12 Foreign direct investment is a vital source of growth for developing economies. The ability to secure certainty about tax incomes is often as important to foreign investors as the nominal tax rate. This can deter tax administrations from challenging often complex tax arrangements and from adopting effective legislative defenses, especially unilaterally. This may explain why to date only a small number of emerging and developing economies have adopted interest restrictions based on the BEPS Action 4 model.16 Interest Restriction in Vietnam In February 2017 Vietnam published a new Decree 20 governing the taxation of related party transactions. This included a provision that states that: Total tax-deductible interest expenses incurred by the taxpayer in a tax period do not exceed 20% of earnings before interest, taxes, depreciation and amortization (EBITDA). (Unofficial translation) Excess interest expense was identified as a significant tax risk by the Vietnamese government. Excess interest expense was used by many foreign direct investors to reduce the amount of tax paid by their operating entities in Vietnam. In addition, the tendency for businesses to be highly leveraged may well reflect the fact that interest is lightly taxed in the hands of individuals in Vietnam, creating an incentive to capitalize enterprises with loan finance, rather than equity. The measure has had a positive impact on corporate income tax yield. The adoption of the 20% of EBITDA cap resulted in the disallowance of $533m of interest expense in the first six months after it came into effect. The desire to attract foreign direct investment often explains the introduction of poorly targeted and cost-ineffective tax incentives. The redundancy of investment incentives tends to be high in many countries.17 Institutional weaknesses can compound the problem: if government agencies that are tasked with securing inward investment can award tax breaks to firms, they tend to do so with little regard to the cost. Too often there is no effort to track the cost of incentives once they have been given or to carry out a cost and benefit analysis in advance of doing so.18 Further work needs to be carried out to determine the impact of the options discussed in this paper on tax incentives. At a high level, it would be expected that options involving apportionment of profit would deter low-substance preferential regimes. This is because profit allocation factors based on, say, the number of employees and value of assets in the country would allocate little or no profit to low-substance entities within the scope of such regimes. On the other hand, those options would not necessarily impact on typical FDI incentive regimes which require substance. 16 Examples include: Argentina, India, Mauritania, Senegal and Vietnam. 17 See examples provide in PCT (2015), Options for low income countries' effective and efficient use of tax incentives for investment. 18 For a longer discussion of tax incentives for foreign direct investors see Chapter 3 of WBG (2017), How Developing economies Can Get the Most Out of Direct Investment. 13 The income inclusion approach (and potentially the DPR), could impact tax incentive regimes that have a substance requirement. This is because it would reallocate low-taxed profit, which would then be subjected to at least a minimum rate of tax.19 Discussions are currently underway, which may result in explicitly carving out regimes compliant with the standards on preferential regimes contained in BEPS Action 5 and those including other indicators of substance.20 Carving out regimes requiring sufficient substance would diminish the effect of the options under consideration on tax incentive regimes that have a substance requirement.21 Access to Data Transfer pricing is based on a comparison between related party transactions and comparable transactions between unrelated parties. In developed economies there is usually a well- established system of financial reporting which enables the compilation of databases of information about the profitability of enterprises and of the prices paid for certain commodities and services.22 Tax administrations in developed economies can afford to access the information contained in these commercially developed databases. For tax administrations in developing economies the cost of access may be an issue but so is the relevance of the data. Information about local enterprises may be scant as financial reporting requirements may be limited, not effectively implemented, or absent altogether. A review of several private databases confirms a scarcity of domestic information that can be used for comparability analysis in many countries (PCT 2017).23 The data reported in developed economies may not meet the standard of comparability required, or may need to be adjusted, adding a further layer of complexity to the transfer pricing process. More reliance on (rebuttable) safe harbors for the return on routine activities could help mitigate this problem. Effective risk assessment of MNEs requires information about the global operations of an MNE. The BEPS package has improved access to data by developing guidance on documentation rules that include a requirement on local MNE entities to provide a master file that includes an overview of the MNE’s operations. It also introduced Country-by-Country reporting obligations. However, Country-by-Country reports are only prepared by a small number of very large MNEs and the conditions for accessing those reports represent a barrier for many developing economy tax administrations.24 19 The DPR does include a rule designed to focus its impact on low-substance entities 20 OECD (2019a). Box 2.1, para 3,1 and 3.2. 21 The IF work program includes an assessment of the economic impact of the options it is considering, including the income inclusion rule. This should include an assessment of effects on wider tax incentive policies. 22 Public availability of information is, however, also heterogenous among developed economies, with a number of countries being heavily underrepresented in the most commonly used databases. 23 When approximating practical requirements for the use of information in transfer pricing comparability analysis, fewer than 1,000 local observations were available for 164 countries. 24 Country-by-Country reports are currently filed in the jurisdiction of the parent entity and provided in a standardized electronic format to other jurisdictions under the terms of bilateral or multilateral conventions including the necessary provisions for the exchange of information. There are specific model agreements between the competent authorities of the parent jurisdiction and jurisdictions requesting the reports. These require the requesting jurisdiction to meet certain conditions, including having the capability to meet international standards on data confidentiality. 14 Political Economy More mechanical rules are less susceptible to political interference. As the pen picture in Annex 3 illustrates, the political context does not just affect support to adopt measures that will protect the tax base. It also affects the management of cases. Auditors can find themselves under pressure to “go easy” on well-connected MNEs that are seen by senior members of the government as important players in the economy. It is harder for auditors to resist such pressure when applying rules that involve a good deal of professional judgement and the exercise of discretion. The OECD’s Inclusive Framework, which includes developing economies, provides a valuable platform for the international standard setting processes. However, challenges exist which can hamper the effectiveness of developing economy input,25 and constrain their ability to ensure that outcomes take into consideration different policy priorities and administrative constraints. Achieving consensus by 2020 on radical reforms of the existing system may prove difficult, with divergences of interest between winners and losers in the allocation of taxing rights.26 Moreover, uncertainty about how some of the options will play out over the longer-term will be challenging for many developing economies and the IF membership. If consensus is not reached, developing economies could consider unilateral or regional adoption of improved anti-avoidance measures that have been designed with developing economies in mind to secure immediate benefits. The TBEP proposals and income inclusion based on a diverted profits rule meet that description. Ideally, the IF should endorse such measures, to coordinate their design and implementation, even if it is not possible to reach consensus on more fundamental reforms. In addition, there are potential benefits to regional approaches to the development and implementation of these types of measures. Perhaps the most important is that regionally aligned approaches deter ‘tax competition’ between member economies. A regional approach also allows all participating states to influence the development of a coordinated policy in a way that is more difficult in developing global solutions, and regional co-ordination can also provide more leverage to influence the shape of global solutions In addition, they reduce the scope for double taxation and double non-taxation and can provide more certainty for, and reduce the compliance costs of, taxpayers. Complexity and Ongoing BEPS Risks 25 See: ATAF (2019), which identifies a need for “a framework in which international tax governance is conducted on an expanded multilateral basis”. 26 The potential differences of view between countries were highlighted in advance of the July 2019 G7 Finance Ministers’ meeting by US concerns about France’s digital services tax (https://www.bloomberg.com/news/articles/2019-07-17/french-defiance-to-trump-on-digital-tax-casts-shadow- over-g-7). However, the G7 remains committed to the IF effort to secure international agreement on a new architecture by early 2020 (https://www.gouvernement.fr/sites/default/files/locale/piece- jointe/2019/07/g7_chairs_summary.pdf) and the IF is working to model the economic impacts of the proposals under Pillars 1 & 2. 15 The G20/OECD BEPS measures include some useful simplifications but also add complexity. The limit on interest deductions already discussed is an effective and simple tool for addressing a form of profit shifting that is prevalent in developing economies. The increased scope for safe harbors discussed earlier is also a helpful simplification but does not go as far as it could have. The BEPS Actions also resulted in significant new complexities. The changes to the definition of Permanent Establishments will help to prevent the artificial avoidance of a taxable presence but the new rules are complex to apply. They are also effective only if treaty partners agree to its adoption and domestic law definitions are updated. Revisions to the transfer pricing rules ushered in by BEPS Actions 8-10 have increased the complexity of transfer pricing. The revised rules aim to track where value is genuinely created more accurately, including in cases where hard to value intangibles are involved. However, the methodology, with its emphasis on accurate delineation of the transaction and a detailed functional analysis, is technically demanding and factually intensive. Consequently, it places an emphasis on the very capabilities that are limited in tax administrations in developing economies. And even after these refinements, opportunities for base erosion and profit shifting remain. This is acknowledged in the IF consultation document and borne out by the experiences of developing economies. Overall, implementation in developing economies of BEPS measures has been lagging, including those related to transfer pricing. This is in part because the BEPS outcomes have generally been grounded in, and an elaboration of, existing guidance, which has added detail and complexity. They require advanced technical skill to apply, together with extensive information and data, particularly regarding entities located in other countries. As a result, it can be difficult for resource constrained administrations to effectively implement many key parts of the BEPS package, including anti-treaty shopping rules and recommendations for valuation of intangibles.27 Annex 4 illustrates these ongoing challenges by way of recent examples derived from the WBG operational support to developing economies28 (summarized in the box below). Recent Examples of Profit-Shifting Identified in West African Economies Annex 4 describes two cases identified during technical assistance provided by the WBG and development partners in West Africa. In both cases significant amounts of profit were transferred to low-taxed entities by means of contractual arrangements with other members of the MNE group, and in both cases, the tax administration sought to test the extent to which the contractual arrangements matched the actual behaviors and arrangements within the MNE group (the ‘substance’). The Annex also describes the factual information, and skills, required to allow the tax administration to comprehensively examine the MNE arrangements. 27 The BEPS guidelines proposed a new approach to the allocation of returns to intangibles based on the principle that in order for the legal owner of an intangible to be entitled to a return from the intangible, it must perform important functions for their development, enhancement, maintenance, protection, and exploitation (the so-called DEMPE functions). However, the methodology, with its emphasis on accurate delineation of the transaction and a detailed functional analysis, is technically demanding and factually intensive. 28 In Africa, a number of country support initiatives have been carried out in partnership with the African Tax Administration Forum (ATAF). 16 Country Case Study 1 involves the shifting of risk (and the associated profit) out of the West African country by means of a contract with a low-taxed associate under which the associate guarantees the West African entity a minimum profit in exchange for an annual fee. The actual fee payable leaves the African taxpayer with no profit in most years. Country Case Study 2 involves inter-company arrangements within an engineering MNE under which key value-adding employees (project managers and engineers) are contractually employed by an associated entity in a low-tax jurisdiction. These employees are then sub- contracted to the West African associate which provides project-engineering services to third party customers. The sub-contract payments are for a fee based on the employee cost and a large mark-up (greater than 50%). The fee shifts profit into the low-taxed entity, leaving the West African entity with losses. Despite the BEPS Actions, it is still possible to divert profits to lowly taxed entities with relatively little substance. As the examples show, significant income can be allocated to income producing factors that are highly mobile and actually involve relatively little decision-making capacity.29 Similarly, although the measures stipulated in BEPS Action 6 concerning treaty abuse are necessary and welcome, they are complex and require extensive factual knowledge of foreign arrangements and operations. For many developing economies a sensible approach is to review their treaty policy much more broadly to minimize revenue losses.30 Options for Change This section provides a brief summary of the main options that have been put forward in recent discussions and assesses them from the perspective of developing economies. These include those proposed in the IF Consultation paper together with options discussed by the IMF (2019) and within the WBG31. The first seven options would make significant changes to the ways in which the international tax system allocates taxing rights over the profits of MNEs. The final three are more focused on tackling tax avoidance involving lowly taxed entities within MNEs. User Value Proposal This proposal would apply to highly digitized businesses that rely on users to create value. For example, users create value when they upload media, information or data that is available to other users via a platform, which can then monetize that data. The proposal involves determining the part of the total business profit attributable to users, which would be allocated to the jurisdictions in which the users are located according to an allocation key. The profit to be allocated would be computed by deducting (from total profit) the profits made from routine activities, calculated using current methods for determining the arm’s length return. The 29 Becker and Englisch (2019) cite several recent studies that evidence profit shifting by MNEs, including through the careful location of IP and risk, addressing the argument made in several responses to the IF consultation document that it is too soon to conclude that the BEPS Actions have failed to fully address profit shifting by MNEs. 30 See 2016 blog post by Jim Brumby and Michael Keen referencing a presentation by Stephen Shay: http://blogs.worldbank.org/governance/tax-treaties-boost-or-bane-development. 31 Leigh Pemberton and Loeprick (2019). 17 portion of that residual profit attributable to user value would be calculated by some agreed method (or an agreed percentage). Jurisdictions would have the right to tax this profit, even if the business has no physical presence there. From a developing economy perspective, the potential to incorporate formulaic approaches is attractive. It would be possible to incorporate formulaic mechanisms (including safe harbors) to determine the portion of residual profit allocable to user-activities, and to the allocation of that profit between countries. But the bulk of MNEs profits are likely to remain subject to the existing transfer pricing regime. The impact of this option on developing economy tax revenues is uncertain and would depend on the definition and measurement of user value, and the allocation formulae adopted. The value-add of consumers in developing economies will depend on the extent to which they interact with these types of digital platforms and may be smaller than in more developed economies because, for example, their consumers may have less access to the internet and relatively low purchasing power. If so, the net result of these proposals could be primarily a redistribution of taxing rights between developed economies. Table 1: User-Value Proposal Primary policy objectives To enhance profit arising in market economies by recognition of the value-added by users of some digital platforms Scope Digital businesses in which users add value to the digital platform. Target Profit in the global supply chain created by users and which is allocable to market economies Potential for mechanistic Medium-Low. The potential to incorporate formulaic or formulaic approaches approaches is attractive. It would be possible to incorporate formulaic mechanisms (including safe harbors) to determine the part of residual profit allocable to user-activities, and to the allocation of that profit between countries. However, other significant profit elements will still have to be addressed using existing transfer pricing methods. Developing economy The impact on developing economy tax revenue is uncertain considerations and depends on the definition and measurement of user value, and the allocation formulae adopted. Developed economy Impact on developed economy tax revenue is uncertain, but considerations likely to redistribute taxable profit toward large market economies. Marketing Intangibles Proposal This proposal could apply to all MNE businesses, rather than only those that are only highly digitalized. It is intended to recognize the ability of businesses to create valuable intangibles that are inherently linked to a market, while having limited, or even no physical presence in that market. It would work by determining the portion of total profit allocable to marketing intangibles by deducting from total profit the return from routine activity and the return attributable to “trade” intangibles (for example, those resulting from research and development). This residual profit would be allocated to market jurisdictions (where the 18 consumers are) by way of an allocation key, which potentially includes the volume of marketing expenditure in the country. The right to tax would not require the business to have a physical presence in the jurisdiction. From a developing economy perspective, this option has attractions because it incorporates formulaic approaches to determining the profit deriving from marketing intangibles and the allocation of that profit between jurisdictions. However, the distinction between the two types of intangibles raises some potentially complex definitional issues. The scope of the formulaic approach is broader than under the user value proposal but still leaves tax administrations having to apply the current transfer pricing guidelines to complex situations. The impact on developing economy tax revenue is uncertain. The focus is on value generated in markets the allocation of profits linked to intangibles specific to markets (such as trademarks) only. This means that it does not address the challenge of profit shifting through the location of often more valuable intangibles (such as software) in low-tax jurisdictions, which are the main sources of growth in digitized economies.32 Table 2: Marketing Intangibles Proposal Primary policy objectives To enhance the recognition of profit arising in market economies by a more consistent and enhanced recognition of marketing intangibles Scope Focus is on digital economy, but potentially applicable to other MNE sectors Target Profit in the global supply chain created by marketing intangibles and which is allocable market economies. Potential for mechanistic Medium. It could incorporate formulaic approaches to or formulaic approaches determining non-routine profit (which could employ safe harbors), the split of that profit between trade and marketing intangibles, and the allocation between jurisdictions of the profit deriving from marketing intangibles. Allocation by reference to the size of the market is an option. But, as with the previous proposal, it would still be necessary to follow existing transfer pricing rules in respect of some profits. Developing economy Impact on developing economy tax revenue is uncertain and considerations depends on the definition of marketing intangibles and the mechanism for measuring the profit allocable to them. Developed economy Impact on developed economy tax revenue is uncertain, but considerations likely to redistribute taxable profit toward large market economies Significant Economic Presence (SEP) This proposal is intended to give jurisdictions the right to tax businesses that have a sustained interaction with their economies, even if that does not involve having a physical presence there. Sustained revenue generation will be a key factor in determining whether a business has a significant economic presence, but other factors will also be taken into account (such as user 32 See, for example, Record et al. (2018). Malaysia's Digital Economy: A New Driver of Development. 19 base, the existence of web pages tailored to the local market and locally derived digital content). The allocation of profits to the significant economic presence could be based on some form of fractional apportionment. Unlike the previous two proposals, it is not grounded in the identification of actual or deemed intangibles attributable to a market jurisdiction and represents a more radical approach. The impact on developing countries will depend on the basis of allocation of profits. If this is based on the likely value-add of the user, the impact on developing economies, at least in the near-term, may be modest. Nevertheless, in respect of highly digitalized businesses, the ‘significant economic presence’ proposal, provided that it incorporates a formulaic allocation mechanism, is most likely to match the needs of developing economies, and a withholding mechanism could be used to collect tax from the affected non-resident taxpayer (see below). Allocation formulae have yet to be determined, however. Table 3 : Significant Economic Presence Primary policy objectives To establish a right to tax profit of an MNE on the basis of the MNE’s economic presence in a jurisdiction Scope Digital economy. Target Profit of MNE digital businesses allocable to significant business activities in a jurisdiction Potential for mechanistic High. The approach could incorporate formulaic approaches to or formulaic approaches determining whether a significant presence exists and the profit allocable to a such a presence in a jurisdiction. Developing economy Impact on developing economy tax revenue is uncertain, considerations depending on the definition of ‘significant presence’ and the mechanism for determining the profit allocable to that presence. Developed economy Impact on developed economy tax revenue is uncertain, but considerations likely to redistribute taxable profit toward large market economies. The three profit allocation proposals in the IF consultation document, discussed above, would affect all profits rather than just those that have been shifted to lowly taxed entities, which is arguably the main concern. They will result in an overall redistribution of taxing rights over the profits attributed to either user value or marketing intangibles, regardless of whether that profit is otherwise subject to low taxation. So, while under-taxation is the principal policy driver, the impact is much wider and that may make it harder to achieve consensus. Moreover, the proposals will not catch profit diversion that is not related to user value, or marketing intangibles. Arguably, this limits the utility of the proposals for developing economies. At a technical level, it will be necessary to consider how taxing rights are to be established if either of the first three proposals are pursued, in order to give jurisdictions the necessary taxing rights under tax treaties. This may take the form of an extension of the permanent establishment principles, or the determination of an additional source taxation right. Other changes to the model tax conventions will be necessary too. To ensure speedy implementation of these changes, the existing Multilateral Instrument will need to be amended or supplemented. 20 Increasing the Scope of Source Taxation Rights and Withholding Taxes This option would give jurisdictions taxing rights over non-residents’ income from specific sources and enable the use of withholding taxes as a mechanism for tax collection. Withholding taxes can be especially effective for the taxation of non-residents with no physical presence in a country. Withholding taxes cannot be divorced from the right to tax a non-resident on specified categories of income, which must be established in domestic law and, where relevant, tax treaties. The ‘significant economic presence’ proposal, for example, would create a right to tax non-residents, and the ‘user value’ and ‘marketing intangibles’ options could also be implemented by creating a right to tax the relevant profits at source. In addition, existing source-taxation principles may be extended to incorporate digitally provided services. Imposing withholding taxes on payments for digital services to non-resident taxpayers would apply an established collection model to a new category of payment. Withholding taxes can be an efficient collection mechanism and decrease the scope for avoidance, especially in relation to the taxation of non-residents with no physical presence in a jurisdiction (which is commonplace in highly digitalized businesses). With respect to business to business payments, such a system of withholding could operate alongside some of the options for taxing a share of the residual profit of MNEs discussed earlier. The withholding tax would act as a payment on account of the tax due in the source jurisdiction. The tax withheld would be offset against the final liability once it has been possible to calculate the total due on the share of residual profits attributed to that jurisdiction. This would improve cash flow and compliance but would not reduce the complexity of the profit attribution process. Going further, source taxation could be an option for business to consumer services, for example to include payments made from consumers to foreign suppliers of digital services, such as streaming of music and video. If this approach is applied to the proposals under the IF’s Pillar 1, the amount of withholding would be representative of the tax that should be attributed to the user, or market. To collect the tax, it would be necessary to follow the example of VAT/GST and require the foreign supplier to account for it. From a developing economy perspective, enhanced taxation in source jurisdictions adds to the tax base, and the potential to apply withholding taxes would add to the efficiency and effectiveness of tax collection and enhance cash flow. 21 Table 4: Source Taxation and Withholding Taxes Primary policy objectives To act as a collection mechanism for taxes due as a result of other options that allocate additional taxing rights and/or as a freestanding mechanism, in domestic law and supported by treaties where relevant, that imposes source taxation on payments made in respect of digital goods and services. Scope All MNE businesses receiving the kinds of payments subject to source taxation and the WHT. Target Source (market) jurisdiction profit of highly digitalized MNE businesses Potential for mechanistic High-will be a fixed percentage of the relevant payments or formulaic approaches Developing economy Enhances cash flow and adds to the tax base as it results in considerations increased taxation in source jurisdictions, although this will vary according to the levels of payments being made and received for digital goods and services by each jurisdiction. Developed economy Again, will vary, depending on in and outgoing payment flows considerations for digital goods and services. Residual profit allocation via Formulary Apportionment This approach represents a simplification of the ‘residual profit split method’, an established transfer pricing method. It involves deducting the return from routine activities from total profit, leaving a residual profit, which will then be allocated to jurisdictions according to a set of factors that are indicative of profit generation, such as assets, labor and sales 33. The exact metrics and weightings in the formula would need to be agreed. This approach has a broad scope of application, embracing traditional as well as digital businesses. Coupled with more extensive use of safe harbors (which are available to determine routine returns) and formulaic approaches, this would represent a major simplification of the system, which is a pre-condition to effective implementation in developing economies. However, as it applies to the whole of the residual profit it will result in the reallocation of profits between normal rate jurisdictions. That in turn may complicate agreement of the allocation keys; this is one of the reasons why formulary apportionment has not been adopted in the past. In addition, clarification is needed how this approach would allocate any residual losses. 33 Devereux and others (2019) propose that all residual profit be allocated to market/destination economies. 22 Table 5: Residual Profit Allocation via Formulary Apportionment Proposal Primary policy objectives To allocate all non-routine profits of MNEs to jurisdictions by way of a formula. Scope All MNE businesses will have non-routine profits allocated by formula Target All of an MNE’s profit over and above ‘routine returns’ Potential for mechanistic High. Safe harbour approaches are available to determine or formulaic approaches residual profit (over and above that allocable to routine activities) and formulaic approaches can be used in the allocation of that profit between affected jurisdictions. Developing economy Would represent a radical simplification of the system and so considerations remove a lot of the problems faced by resource constrained tax administrations. The impact on tax revenue would depend on the allocation keys adopted. Developed economy The impact on tax revenue would depend on the allocation considerations keys adopted. Global Formulary Apportionment (GFA) This option would involve, in the first instance, a definition of an MNE group global profit, and then determination of formulae for allocating that profit between affected jurisdictions. It has been discussed amongst academics and NGOs for many years, and would replace the arm’s length principle, which is the current internationally developed approach for the allocation of profit between jurisdictions. GFA can theoretically apply to allocate total MNE global profit or the global profit of specific business lines. Allocation factors could be related to, for example, local sales, local headcount and local sales. The choice of these allocation keys will have a significant impact on how profits are allocated between countries. From a developing economy point of view, GFA would remove the need for traditional transfer pricing rules, which are often complex and expensive to apply and enforce and require extensive information and data. It would also remove a major source of dispute and uncertainty between MNEs and tax administrations. Added attractions are that it is mechanical in nature and largely removes the incentive to divert profits to low taxed entities with little substance. There is currently no international consensus that the merits of GFA outweigh its drawbacks. Its critics argue that in contrast to the arm’s length principle, the results of formulary apportionment would not reflect market dynamics and would therefore be arbitrary. The adoption of this option, which would require global consensus on the allocation keys, is likely to be very difficult to achieve in a short or medium timeframe but should remain under consideration. 23 Table 6 : Global Formulary Apportionment Primary policy objectives To simplify, and potentially make fairer, the apportionment of MNE profit between affected jurisdictions. Scope All MNE businesses. Target Global MNE profits, either in total or per business line Potential for mechanistic High. Apportionment of profit between jurisdictions would be or formulaic approaches according to agreed allocation formulae, such as assets, workforce, and sales. Existing transfer pricing implementation issues would no longer be relevant, as long as adoption is global. Developing economy The impact on developing economies’ tax revenues would considerations depend on the allocation formulae adopted. The highly mechanical approach to profit allocation would address the challenges currently faced with implementation of the arm’s length principle. Developed economy The impact would depend on the allocation formulae adopted. considerations Destination Based Cash Flow Tax (DBCFT) This would replace the existing corporate income tax with a new tax on the receipts of corporations less their expenditure, similar to a VAT. However, it is subject to “border adjustment” under which imported goods consumed locally are subject to the tax but domestically produced and exported goods are not subject to the tax. This option has been proposed by a number of academics,34 most prominently in the lead up to the US tax reform of 2017. As noted in IMF (2019), the DBCFT removes many of the distortions that result from the imposition of corporate income taxes, including the bias towards debt finance. It is also proof against profit shifting by MNEs. However, the distributional effects could be large and are also subject to significant uncertainty. The IMF (2019) finding that low income countries would benefit is counter-intuitive and, may only be sustained over the longer-term if source taxation is retained for natural resources and greater compliance is achieved through a more robust DBCFT system relative to the current system. 34 For a summary, see Auerbach (2017), Demystifying the Destination Based Cash-Flow Tax. 24 Table 7: Destination-Based Cash-Flow Tax Primary policy objectives To address profit shifting through a fundamental reform of company taxation which gives taxing rights to the country where the purchaser is located. Scope All MNE businesses. Target MNE profits Potential for mechanistic Implementation of a DBCFT would be relatively or formulaic approaches straightforward to apply. Transfer pricing issues would no longer be relevant as long as adoption is global Developing economy The impact on developing economies is unknown considerations Developed economy The impact on developed economies is unknown considerations Minimum Tax on Outbound Investment: Income inclusion Rule Under this proposal, the jurisdiction in which the ultimate parent of an MNE is resident would subject to tax the profits of foreign subsidiaries and permanent establishments that are subject to a low effective rate of tax in their jurisdiction of residence or establishment. The proposal has much in common with existing controlled foreign company (CFC) rules adopted by primarily capital-exporting countries. These are anti-avoidance rules aimed at deterring profit shifting to low-taxed jurisdictions. The proposed ‘income inclusion rule’ would seem to primarily benefit capital exporting economies and is therefore unlikely to be beneficial to developing economies, which are also harmed by profit diversion to low-tax jurisdictions. Developing economies will only benefit indirectly if the adoption of the income inclusion rule results in change in behavior by MNEs and a reduction in the profits attributed to lowly taxed entities overall35. However, it is important to note that profit is often directly diverted from developing economies through such mechanisms as base-eroding payments to, and the artificial placement of risk in, low-tax jurisdictions. 35 Any spillover benefits for capital importing economies would be dependent on a very high level of adoption in capital exporting economies. Even then, the benefits may be limited by the principle of territoriality. For example, when the UK reformed its CFC rules to reflect that principle, it made it clear that the target of the provisions was profit diverted from the UK. Profit shifting between foreign subsidiaries would be unaffected. 25 Table 8: Income Inclusion Proposal Primary policy objectives To deter MNEs from shifting profit from shareholder jurisdictions to low-tax jurisdictions, which would have a right to tax the profit recognised by low-taxed subsidiary entities (and branches). Scope All MNE businesses. Target Profit recognised in low-taxed subsidiary (or branch) entities Potential for mechanistic Low. or formulaic approaches Developing economy Makes no direct contribution to developing economy tax considerations revenues. Impact of ‘spill-over’ benefits for developing countries is uncertain. Developed economy Likely to enhance developed economy tax revenues. considerations Income Inclusion: A Diverted Profits Rule (DPR) This measure would allow every jurisdiction involved in a global value chain to tax a share of the profits diverted to low taxed entities containing insufficient substance. The starting point would be to determine an amount of profit attributable to a low-taxed entity that is representative of that entity’s substantial earning capacity. An y profit in excess of that amount would be allocable to related entities in accordance with an agreed metric. The determination of profit in line with substance would be formulaic; for example, a multiple of costs (primarily the direct costs of labor) incurred in the low-taxed entity’s country of residence36. From a developing economy perspective, a key attraction of a DPR is that it would allocate low-taxed profits to all countries in the same business structure (supply chain), rather than to the parent entity, and therefore directly target profits diverted from developing economies.37 In addition, both the mechanism for measuring excessive profit (that is, the profit over and above that justified by substance) and the allocation of that profit between jurisdictions, can be formulaic. A diverted profits rule’s focus on lowly taxed profits unsupported by substance makes it less likely to give rise to issues of double taxation and disputes between normal rate jurisdictions. 36 For a fuller discussion of this proposal, supported by model legislation see: Leigh Pemberton and Loeprick (2019) 37 If such an approach were adopted, the IF member states should also consider adding a requirement to report the number, and related costs, of employees (or equivalent) located in the respective jurisdiction to the standard transfer pricing documentation requirement. 26 Table 9: Income Inclusion: Diverted Profit Proposal Primary policy objectives To deter MNEs from shifting profit shifting to low-tax jurisdictions from any jurisdiction in which an MNE conducts business. All affected jurisdictions would have a right to tax a proportion of the profit recognised by low-taxed affiliates. Scope All MNE businesses. Target Profit recognised in low-taxed entities over and above that allocable to any substance in those entities. Potential for mechanistic High. Formulaic approaches are available to determine or formulaic approaches ‘excessive profit’ (over and above that allocable to substance) and allocation of that excessive profit between affected jurisdictions. Developing economy Likely to deter profit shifting from developing economies and considerations increase tax revenue. Developed economy Likely to deter profit shifting from developed economies and considerations increase tax revenue. Under both a DPR and a CFC/Income Inclusion rule it will be necessary to decide what tax rate to apply to the included profits.38 Should the income be taxed at the full domestic rate of tax, or at the internationally agreed minimum rate? Either way, if an international consensus can be achieved on what constitutes a “low” rate of taxation, that will help to limit the “race to the bottom” in rates of corporate income tax. Becker and Englisch (2019) discuss a possible rate and the extent to which an international minimum tax will affect the use of tax incentives. Developing economies stand to benefit from both a reduction in tax competition and reduced reliance on tax incentives. However, the definition of what is a “low” rate will have an indirect impact on how high jurisdictions will feel able to set their domestic corporate tax rates. MNEs will still have an incentive to shift profits from jurisdictions with rates significantly above the agreed minimum tax rate. So, it is in the interest of developing economies to ensure that the definition of a minimum rate of taxation is agreed at a level that reflects their need for revenue. Minimum Tax on Inbound Investment, Tax on Base Eroding Payments (TBEP) This proposal targets payment to related parties that give rise to a high risk of base erosion (such as interest, or royalties) and that are not subject to a minimum effective rate of taxation in the recipient entity. Such payments would not be deductible for tax purposes (‘under-taxed payments rule’) or would not be covered by treaty provisions which limit or exempt taxation in the source jurisdiction (‘subject to tax rule’). This approach increases the rights of “source” jurisdictions to levy tax, albeit in the specific context of under-taxation in the residence jurisdiction of the recipient. The ‘under-taxed payments rule’ would be a valuable additional defense for developing economies. It applies to a payment made to a low-taxed related party and would apply whether or not such a payment complies with arm’s length conditions. The rule could be expected to be relatively simple to enforce: taxpayers could be required to self-assess whether such payments meet the minimum effective tax test and information required by tax administrations to enforce the rule would expect to be included in transfer pricing tax return schedule. There are some 38 See paragraph 100 of the IF’s public consultation document (OECD 2019). 27 concerns about the application of this rule in cases where payments flow through entities (‘conduits’) that are subject to tax to entities that are not. As the IF consultation document recognizes, an effective definition of payments subject to the rule will need to be broad. The ‘subject to tax rule’ would be a very useful supplement to the tools developed in BEPS Action 6 to counteract treaty abuse. Many developing economies have entered treaties providing benefits to non-residents that are in practice subject to a very low effective rate of taxation. This gives rise to distorted investment flows, incentives to maximize royalty and interest payments and loss of revenue from source taxation. The scoping of the rules according to ‘effective tax rate’, however, could be expected to create obstacles to enforcement as well as compliance burdens to taxpayers. An alternative could be to use the effective tax rate of a sub-group located in a jurisdiction, which could use data derived from a Country-by-Country report to determine whether payments meet the ‘effective tax rate’ criterion. Otherwise, the rule could be designed to apply only to payments to related parties located in low-tax jurisdictions Table 10: Tax on Base Eroding Payments (TBEP) Primary policy objectives To deter profit-shifting payments from normal-taxed to low- taxed entities Scope All MNE businesses. Target Payments made to low-taxed related-party entities Potential for mechanistic These measures can be relatively straightforward to apply, or formulaic approaches involving a denial of either treaty benefits or tax-deductions. Developing economy Base eroding payments are a continuing concern to developing considerations economies, and TBEP measures can be a useful tool to address them, providing that information on the ownership, location and tax position of recipient parties are available. This requires access to adequate documentation, including CbCRs. Developed economy Base eroding payments are a concern to developed economies, considerations and TBEP measures can be a useful tool to address them. Interaction Between the Proposals Some of the options for change could coexist and some are clearly alternatives. The different proposals for residual profit allocation are alternatives to one another (user value, marketing intangibles and formulary apportionment for all residual profit). The significant economic presence proposal clarifies how a nexus for taxation could be established but will still need to be supported by allocation rules. The income inclusion rule as proposed by the IF and the diverted profits version are alternatives, but either could coexist with a revised approach to profit allocation. Indeed, introducing the diverted profits rule could be an opportunity to explore how best to design profit allocation keys in advance of wider application. The tax on base eroding payments could also coexist as an anti-avoidance measure with revised approaches to profit allocation. Extending source taxation to capture a share of the profit generated in user/market jurisdictions is an alternative to revised rules for profit allocation. The DBCFT and global formulary apportionment would wholly replace the current system for taxing corporate profits 28 Implementation Successful implementation of reform in developing economies is critically dependent on access to information. This is particularly relevant for options that involve reallocation of an MNE’s global profits, or some portion of those profits. Even the measures designed to impose taxes on base eroding payments will need to be supported by access to data to verify taxes paid. The obvious vehicle for meeting this need for information would be MNEs’ Country-by- Country reports, possibly modified to take account of the factors that are agreed as allocation keys under any new system. The current reporting threshold of €750 million would need to be lowered and it would need to be recognized that countries should have direct access to this data by way of local reporting requirements. This is already the case for the transfer pricing master file and there is no reason to retain the cumbersome process of exchange of information under treaties by which Country-by-Country reports are currently distributed. Tax administrations in developing economies may lack the necessary treaty relationships and have difficulty in meeting all the requirements set down in the BEPS Action 13 report and implementation guidance. These are justified by the perceived need for confidentiality, but the trend is towards greater transparency and public Country-by-Country reporting.39 The other condition for success in developing economies is administrability. This paper has highlighted the need for relative simplicity of application if new rules are going to adopted and implemented successfully in developing economies. For the proposals that require an allocation of a slice of all or part of an MNE’s profit countries, a formulaic approach is desirable. And proposals requiring a computation of ‘residual profit’ over and above an arm’s length reward for ‘routine functions’ should include simplified approaches to computing the latter. Given the specific challenges faced by capacity constrained developing economies to ensure the taxation of routine profits, the future work of the IF should encompass a detailed review of baseline profits in relevant industries, to guide the application of mandatory safe harbors with an arm’s length let out as discussed in PCT (2017)), covering for instance contract manufacturing, centralized services and routine distribution. Conclusions The direction of thinking reflected in recent public proposals on the reform of the international tax architecture has the potential to address the challenges developing economies face in seeking to protect their tax bases. Proposals for some simplification and mechanization of regimes are particularly relevant, and the development of a globally consistent approach or common recommendations is helpful to countries reluctant to introduce unilateral measures for fear of the impact on the investment climate. A consensus on the way forward will also help reduce the risks of double taxation and provide more certainty to business. However, the pressing needs for revenue and precedents for unilateral action by prominent OECD countries imply that guidance needs to provided quickly to maintain coordination. Recommendation on appropriate anti-abuse measures may require a shorter turn-around than wider agreement on the re-allocation of taxing rights and could possibly pave the way for wider reform. 39 Public country by country reporting is already required from businesses in some sectors in the EU and is a feature of the Global Reporting Initiative’s recent draft standard for reporting tax and payments to government: https://www.globalreporting.org/information/news-and-press-center/Pages/Draft-standard-tax-and-payments-to- government-public-comment-2018.aspx 29 Achieving consensus by 2020 on more radical reform of the existing system may be difficult. To the extent that proposals result in significant changes to the way in which taxing rights are allocated between jurisdictions, they will create winners and losers. Moreover, uncertainty about how some of the options will play out over the longer-term will be challenging for many developing economies and the IF membership. The IF is currently undertaking work to better understand the economic impact of some of the options for reform. An assessment of likely effects on developing economies will be particularly valuable. Disagreements in the lead up to the July 2019 meeting of G7 Finance Ministers illustrate the challenges that remain to achieving a consensus. However, failure to agree a solution internationally does not leave developing economies with no options. Unilateral or regional adoption of improved anti- avoidance measures that have been designed with developing economies in mind could secure immediate benefits. The TBEP proposals and income inclusion based on a diverted profits rule meet that description. Ideally, the IF should endorse such measures, to coordinate their design and implementation, even if it is not possible to reach consensus on more fundamental reforms, or at least not in the short-term. There could be value in focusing on a package of anti- avoidance measures to address on-going BEPS risks in the short-term, while allowing more time for discussion of fundamental reforms, including some options not currently included in the IF work program. 30 References ATAF (2019). The place of Africa in the shift towards Global Tax Governance: Can the taxation of the digitalised economy be an opportunity for more inclusiveness? https://irpcdn.multiscreensite.com/a521d626/files/uploaded/ATAF%20PAPER%5B1%5D.pdf Auerbach, A. (2017), Demystifying the Destination Based Cash-Flow Tax. NBER Working Paper Series, 23881. Balabushko,O., Beer, S.; Loeprick, J., Vallada,F,. (2017). The direct and indirect costs of tax treaty policy: evidence from Ukraine. Policy Research working paper, No. WPS 7982. Washington, D.C. : World Bank Group. https://hubs.worldbank.org/docs/imagebank/pages/docprofile.aspx?nodeid=27227133 Becker, J. and Englisch, J. (2019) International effective minimum taxation – the GLOBE proposal. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3370532 Beer, S. and Loeprick, J. (2018), The Cost and Benefits of Tax Treaties with Investment Hubs: Findings from Sub-Saharan Africa. Policy Research working paper, No. WPS 8623. Washington, D.C.: World Bank Group. https://hubs.worldbank.org/docs/imagebank/pages/docprofile.aspx?nodeid=30548290 Cooper, J; Fox, R.; Loeprick,J.; Mohindra,K. (2016). Transfer pricing and developing economies: A handbook for policy makers and practitioners. Directions in development. Washington, D.C.: World Bank Group. Michael P. Devereux, Alan J.Auerbach, Michael Keen, Paul Oosterhuis, Wolfgang Schön and John Vella (2019). Residual profit allocation by income. Oxford University Centre for Business Taxation Working Paper WP19/01. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3358291 https://hubs.worldbank.org/docs/ImageBank/Pages/DocProfile.aspx?nodeid=27071555 European Commission (2017), Digital Single Market: Modernizing VAT for Cross Border e- Commerce, https://ec.europa.eu/taxation_customs/business/vat/digital-single-market- modernising-vat-cross-border-ecommerce_en. IMF (2019) Corporate Taxation in the Global Economy. https://www.imf.org/en/Publications/Policy-Papers/Issues/2019/03/08/Corporate-Taxation-in- the-Global-Economy-46650 Leigh Pemberton, J. and Loeprick, J. (2019). Low Tax Jurisdictions and Preferential Regimes: Policy Gaps in Developing Economies. Policy Research working paper, No. WPS 8778. Washington, D.C.: World Bank Group. https://hubs.worldbank.org/docs/ImageBank/Pages/DocProfile.aspx?nodeid=30896182 1 OECD (2017) Transfer pricing Guidelines. http://www.oecd.org/ctp/oecd-transfer-pricing- guidelines-for-multinational-enterprises-and-tax-administrations-20769717.htm OECD (2018), Tax Challenges Arising from Digitalisation, Interim Report. http://www.oecd.org/ctp/tax-challenges-arising-from-digitalisation-interim-report- 9789264293083-en.htm. OECD (2019), Consultation Document – Addressing the Tax Challenges of the Digitalization of the Economy. http://www.oecd.org/tax/beps/public-consultation-document-addressing-the-tax- challenges-of-the-digitalisation-of-the-economy.pdf OECD (2019)a. Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy. http://www.oecd.org/tax/beps/programme-of- work-to-develop-a-consensus-solution-to-the-tax-challenges-arising-from-the-digitalisation-of- the-economy.pdf. Platform for Collaboration on Tax (2015) Options for low income countries' effective and efficient use of tax incentives for investment: Tools for the assessment of tax incentives. Platform for Collaboration on Tax. Washington, D.C.: World Bank Group. https://hubs.worldbank.org/docs/imagebank/pages/docprofile.aspx?nodeid=25251583 Platform for Collaboration on Tax (2017) A toolkit for addressing difficulties in accessing comparables data for transfer pricing analyses. Washington, D.C.: World Bank Group. https://hubs.worldbank.org/docs/imagebank/pages/docprofile.aspx?nodeid=27121843 Martin Hearson (2018), Transnational expertise and the expansion of the international tax regime: imposing ‘acceptable’ standards, Review of International Political Economy, 25:5 Record,R., James L.; Larson,B. Robert; T. Sharifuddin,S.; Chong,Y.. (2018). Malaysia's Digital Economy: A New Driver of Development . Washington, D.C.: World Bank Group. https://hubs.worldbank.org/docs/imagebank/pages/docprofile.aspx?nodeid=30402610 WBG (2017) How Developing economies Can Get the Most Out of Direct Investment http://www.worldbank.org/en/topic/competitiveness/publication/global-investment- competitiveness-report WBG (2019), World Development Report on “The Changing Nature of Work” http://www.worldbank.org/en/publication/wdr2019 2 Annex 1: Business Lines in the Digital Economy and Taxation Issues 1. The digital economy is a broad term, comprising a range of business lines, some with commonalities in terms of the main tax compliance issues arising, but several with important idiosyncrasies. Just as there is no commonly-agreed definition of what comprises the digital economy, and how it should be measured, neither is there a clear consensus on how to classify or categorize its various segments or economic actors. Complicating matters, some leading digital firms operate not a single coherent business model as such, but across a number of distinct business lines. Sometimes, but not always, the digital firm may operate across its various business line using the same or similar branding. The remainder of this section discusses the principal business lines observed in the digital economy. This taxonomy is based on the nature of the underlying economic transactions and what this implies for tax policy and compliance. The description of each business line is followed by some examples of how different countries have sought to tax the business line described. This sets the scene for the discussion of policy options in the next section. A Sale and Resale of Goods 2. Definition: Businesses that acquire products, including control rights, from suppliers and resell them to buyers; resellers control prices and assume liability towards customers; they do not allow for the interaction of end-users and they do not necessarily require customers to affiliate to the online interface (OECD 2018). 3. Mapping transactions: The final consumer makes an online purchase of a good which is either acquired by the reseller from a local supplier or imported by the reseller from a foreign supplier. This is the least complex of the digital business lines and, due to the tangible nature of the goods delivered, comprises the set of transactions that, in principle, it is easiest for the tax administrations to track. 4. Taxation issues: a. The main risk posed to direct taxation by this business model had been the way in which some resellers have structured their businesses to avoid having a taxable presence in countries where they have significant physical operations. This was possible because the definition of a permanent establishment in model tax conventions had not kept pace with the way functions that had traditionally been regarded as simply preparatory or auxiliary in nature had become a core part of the business models of online retailers. As was mentioned in the introduction, this issue was a specific focus of BEPS Action 7, which upgraded the definition of permanent establishments. Countries can incorporate that revised definition in their treaties by means of the Multilateral Instrument (MLI) that resulted from BEPS Action15. b. The tangible nature of the goods being imported means that the tax administrations can rely on existing systems of border control to secure payment of indirect taxes and duties due on the imports. However, the growth in cross-border online purchases has called into question existing exemptions for low-value consignments. These allow packages with a value below a set threshold to be imported without 3 payment of indirect taxes or duties and have been adopted quite widely to keep down the costs of administration and compliance. 5. Recent international experience: Increasing numbers of countries are signatories to the MLI and are reviewing their domestic charging provisions to ensure they fully exercise their taxing rights over permanent establishments.40 Advances in customs screening technology have brought the administrative cost of applying customs duties and indirect taxes even to small transactions to close to zero. There remains, however, significant variance worldwide in the maximum declared value of such imports allowed to pass customs without attracting taxes or tariffs. The current de minimis threshold in Mexico is $50, $800 in the U.S.41, C$20 in Canada and €22 in the EU. Moreover, there have been divergent trends, with the U.S. having recently increased its threshold from $200 to $800 and with policy discussions in Canada having centered around a significant increase in its threshold, to around $200. The recently concluded United States-Mexico-Canada Trade Agreement heralds changes to de minimis arrangements governing trade between the three countries. Canada will raise its de minimis level, from C$20 to C$40 for taxes and will also provide for duty free shipments up to C$150. Mexico will continue to provide USD $50 tax free de minimis and also provide duty free shipments up to the equivalent level of USD $117.42 By contrast, Australia eliminated its de minimis threshold (previously approximately $730) as of 1 July 2018, and now requires B2C exporters in other countries with annual sales above A$75,000 to register with the Australian tax authorities, and to charge and remit VAT. The EU has also legislated to eliminate its de minimis threshold from 2021 as part of its new VAT regime for e-commerce. The changes are expected to generate approximately $8bn in extra VAT revenues annually for EU governments. In line with the destination principle, VAT will be charged in the jurisdiction of final consumption. To ease the administrative burden on sellers, couriers and customs officials, the EU will in parallel establish a digital online portal (One Stop Shop) for the online cross-border sale of goods, similar to that for digital services that is already in existence. B Sale and Resale of (Access to) Digital Content or Digital Solutions 6. Definition: These firms essentially operate as re-sellers of digital content, much in the same manner as goods re-sellers buy and sell goods, or of access to that digital content for the period of the subscription. There are some important distinctions, however, in respect of the delivery mechanism. Since the ‘product’ is delivered in purely digital form, there is no need for the provider to employ physical infrastructure for in-country delivery, and therefore even less need for the provider to have a PE in the country. In addition, payment is typically in the form of a periodic subscription, engaged in through a web-interface, giving the subscriber access to or use of – but not ownership of– the digital content for the duration of the subscription. Due to the similar nature 40 For details of the current signatories of the MLI see: http://www.oecd.org/tax/treaties/beps-mli-signatories-and- parties.pdf 41 Unlike other jurisdictions cited here, the U.S. does not levy a Federal GST or VAT, so its collection at the border does not arise. 42 https://ustr.gov/about-us/policy-offices/press-office/fact-sheets/2018/october/united- states%E2%80%93mexico%E2%80%93canada-trade-fa-1# 4 of virtual provision, and therefore similar implications for tax authorities, this category can also include digital solutions such as cloud computing etc. 7. Mapping transactions: Typically, the final consumer makes an online purchase or subscription, often transacting with a non-resident supplier of digital content which is unlikely to have a permanent establishment in the territory where consumption takes place. The transaction is processed electronically – typically through the domestic banking system or by means of card payment. This category also includes bespoke digital solutions, such as web design and cloud computing, which are typically, if not necessarily, provided on a B2B basis, either as a subscription or one-off contracted payment. 8. Taxation issues: a. Where the foreign supplier genuinely has no taxable presence in a country, even under the revised definition of a PE, this business line does not raise any novel direct tax issues. In principle, it is no different from the case of a foreign manufacturer who send goods to a domestic customer. In practice, suppliers of digital services may also engage in more interactive transactions with their customers, which may raise issues that are discussed in relation to the sale of advertising and data. b. Imports of digital content generate a liability for VAT, with the onus on the purchaser to declare the purchase to the tax administrations and to pay the relevant tax. Where the final consumer is a business registered for tax purposes in the home country, it can declare the purchase and VAT to the tax administrations in the regular course of its business and claim the input VAT against VAT collected from its own customers. For the business, VAT payable on the imported digital content is therefore purely notional and does not represent an additional cost (other than the administrative cost of registering the transaction with the tax administrations). However, for example, the vast bulk of subscribers to streaming services are likely to be individual final consumers. Here, the VAT payable is not notional in the sense that it cannot be off-set against another tax liability and should be remitted to the tax administrations by the subscriber or purchaser. Compliance with this requirement, however, is complicated, burdensome, and rarely either respected or controlled. Malaysia and Indonesia are examples of countries where the consumer is in principle liable to account for VAT on digital services supplied from abroad but the tax is not collected in practice. 9. Recent international experience: More than 50 countries have adopted the Guidelines’ recommendations for imposing VAT on the direct supply to consumers of services and intangibles by foreign suppliers, including most OECD and G20 countries.43 The European Union, for example, has levied VAT on nonresident suppliers of telecommunications, broadcasting, and electronic services, regardless of scale, since January 1, 2015. Those businesses which do not have 43 OECD, Tax Challenges Arising from Digitalisation, Interim Report , 2018, p. 103, http://www.oecd.org/ctp/tax- challenges-arising-from-digitalisation-interim-report-9789264293083-en.htm. 5 a permanent establishment in any EU country choose a member state with which to be identified, and are then assigned a VAT number pertaining to that country. The business must then apply the VAT accounting and remittance rules of the country of consumption. An online portal, known as the mini one-top-shop (MOSS), allows suppliers to register, submit quarterly returns, and pay the tax due.44 According to the OECD, more than €3 billion in taxes have been raised in the EU in this way.45 Australia adopted a similar approach in July 2017, requiring foreign suppliers that exceed a turnover threshold of A$75,000 to account for GST on digital and professional services.46 Within ASEAN, Singapore announced in its February 2018 budget that GST will be imposed on imported digital services, including the streaming of music and movies and downloaded applications and Malaysia will apply its sales and services tax to imported digital services from 1 January 2020. 47 A number of Latin American countries have been experimenting with i) requiring foreign digital suppliers to register with the tax administrations, and to collect and remit VAT directly, and ii) split payments, whereby financial intermediaries are tasked with imposing a charge for VAT on the consumer when processing payments for digital services and remitting that portion of the transaction directly to the tax administrations. The UK is also exploring the possibility of introducing such split payment mechanisms.48 10. Since 2003, France has levied a retail excise tax on audiovisual content. First conceived to apply to off-line consumption of DVDs and VHS, for example, it was extended in 2004 also to cover online paid video-on-demand services. The tax is levied at a flat rate of 2% (with a higher 10% rate on content of a violent or pornographic nature) on the sale or rental of such content in France or, in the case of online content, to a consumer habitually domiciled in France. The residence of the supplier is relevant, but they are nonetheless required to report, collect and remit the tax (OECD, 2018). 11. In August 2018, Chile became the latest Latin American country to signal its intent to impose a charge on the import of digital content and services with the announcement that it would introduce a freestanding (i.e. independent of its 19% VAT rate) 10% charge on firms such as Netflix and Spotify. It is envisaged that the 10% rate will extend beyond audiovisual content, to include subscriptions to internet platforms, advertising and promotional services, the purchase of software as a service and online storage. The tax will be collected as payment is made by the 44 European Commission, “Digital Single Market: Modernizing VAT for Cross Border e -Commerce,” December 2017, https://ec.europa.eu/taxation_customs/business/vat/digital-single-market-modernising-vat-cross-border- ecommerce_en. 45 OECD, “Tax challenges arising from digitalisation: More than 110 countries agree to work towards a consensus - based solution,” March 16, 2018, http://www.oecd.org/ctp/beps/tax-challenges-arising-from-digitalisation-more- than-110-countries-agree-to-work-towards-a-consensus-basedsolution.htm. 46 Australian Taxation Office, “GST on Imported Services and Digital Products,” September 2017, https://www.ato.gov.au/Business/International-tax-for-business/GST-on-imported-services-and-digital-products/. Digital services include streaming or downloading music, movies, applications, games, and e-books. Professional services include architecture or legal services. 47 Inland Revenue Authority of Singapore, “GST on Imported Services,” February 2018, https://www.iras.gov.sg/irashome/GST/Consumers/GST-on-Imported-Services/. Malaysian Budget 2019, page 12: https://www.treasury.gov.my/pdf/budget/speech/bs19.pdf 48 The UK Government issued a consultation document on the split payment method of VAT collection in March 2018: https://www.gov.uk/government/consultations/alternative-method-of-vat-collection-split-payment 6 administrators of electronic payments (generally this means that credit and debit card providers will act as withholding agents). The tax administration will need to keep a list of foreign digital services providers and of the local withholding agents. C Sale of Advertising and Data 12. Definition: This business line typically relies on offering users ‘free’ access to a service, such as a search engine or a social network, who through the use of the service generate data or content, which is in turn monetized, either through the sale of online advertising space or the direct sale of data for the purpose of, for example, targeted advertising through other platforms. 13. Mapping transactions: Monetary transactions typically consist of the payment by a business, sometimes through an intermediary, to the digital provider for either online advertising or for user data. There is a school of thought, underpinning for example the EU’s proposal to tax such digital services, that the receipt by users of the ‘free’ service in exchange for the provision – whether consciously or inadvertently – of data should be considered not only as an economic transaction, but as a key contributor to value creation, and therefore a legitimate justification for taxing the revenues of the digital provider in respect of value created within the borders of the relevant tax jurisdiction. According to the EU (2018): “The extent to which platform users have to provide personal data in exchange of an access to a service varies significantly from one activity or company to another as well as the degree of users' awareness and level of involvement (passive versus active participation) in giving away personal data. More and more personalized user data enables better targeting of marketing messages and increases the value of the advertisement medium. This value is difficult to measure for tax purposes. The specificity of such revenue model is that it disconnects users from the revenue sources”. 14. Taxation issues: The direct payments for advertising and data do not raise any particularly novel taxation issues. The open policy issue concerns direct taxation of the profits generated by platforms that have little or no physical presence in jursidictions where many of their users are located. Even where the ultimate target audience of the online advertising is demonstrably in a given jurisdiction, it does not necessarily follow that either the entity purchasing advertising space or the digital provider would have a permanent establishment in that jurisdiction (under existing PE rules). Even if the digital provider did have a permanent establishment in the jurisdiction, the question of how much profit to attribute to that PE, and therefore how much corporate income tax to levy, is not easily answered. These issues are central to the ongoing work of the Task Force on the Digital Economy and its efforts to achieve a consensus on what changes, if any, are needed to the international rules of taxation. In the meantime, as the original BEPS Action 1 report acknowledged, some countries are introducing, or contemplating, “interim measures” designed to secure some tax revenues from this type of business line, even when there is no taxable presence in the country under existing international direct tax rules. 15. Recent international experience: In June 2016, India introduced a 6% ‘equalization levy’, on specified digital services, essentially online advertising, but with the legislation drafted in such a manner as to allow for an extension in scope through secondary legislation. In essence, the 7 equalization levy was introduced as a free-standing tax, independent of either VAT or income tax, but as a proxy for corporate income tax on foreign suppliers who did not have a permanent establishment in the country. The charge is only applicable on B2B transactions, and when the consideration exceeds approximately $1,500 in a given tax year. Responsibility for declaring and remitting the levy remains with the domestic purchaser rather than with the foreign supplier. The introduction of such a tax or levy was envisaged – albeit not recommended, as such – under BEPS Action 1, thus it is not considered to be contrary to its guiding principles. In each of its first two fiscal years of operation, the levy has reportedly raised approximately $90m. An industry-wide consultation is underway as to the application of the levy and exploring whether it needs to be modified. Government sources have been quoted as suggesting that the rate could be increased to 8%, for example, while an expansion in scope to bring B2C transactions above a certain threshold and other digital services and content, such as audiovisual streaming, into the tax net are also understood to be under consideration. 16. In 2016, France extended its excise tax on audiovisual content (see. Paragraph 19, above) to cover such content provided to final consumers for free but monetized through the display of online advertisements to viewers. This would capture the revenues of services like Youtube, and those with similar business models. The tax functions in a similar manner to that previously in effect on videograms (e.g. VHS, DVD) and paid online video-on-demand content, except that the tax base consists of the consideration paid for the display of advertisements or sponsorships linked to a particular online audio-visual content. Taxpayers are granted an allowance, on which the tax is exempt. This allowance is 4% but increases to 66% where the content is generated by private users for the purpose of sharing between members of a community of like interests. Of the remaining consideration, the first €100,000 is also exempt from the tax (OECD, 2018). 17. In March 2018, the EU Commission published proposals that, if implemented, would see a similar 3% tax introduced on online advertising, as well as some other digital service imports. the tax would be levied on the gross revenues from digital activities in which users play a major role in value creation. Specifically, the tax would apply to revenues from selling online advertising space, intermediary activities that allow users to interact and sell goods and services, and the sale of data.49 The Commission has estimated that a 3 percent tax could raise €5 billion a year. The measure was intended to be an interim tax, pending longer-term reform but did not command support from all member states. In the absence of an EU consensus, member states, including the UK, France and Spain, have put forward their own digital services taxes. 18. In October 2018, the UK Government announced that it would introduce a Digital Services tax in April 2020. This decision reflects uncertainty about the outcome of the discussions about possible reforms to the global tax system and a determination to tax digital businesses on the value they derive from their UK users. Like the EU measure it is narrowly targeted at larger MNEs in the digital economy. It will charge a 2% tax on the revenues of digital businesses whose revenues 49 European Commission, “Digital taxation: Commission proposes new measures to ensure that all companies pay fair tax in the EU,” March 21, 2018, http://europa.eu/rapid/press-release_IP-18-2041_en.htm. 8 are linked to UK users. The “in scope” business models will be the provision of a search engine, a social media platform, or an online marketplace. Tax will be payable if the in-scope business model generates at least £500 million globally and more than £25 million in the UK. The tax is expected to raise around £400 million each year.50 France passed legislation introducing its version of the Digital Services Tax in July 2019. It is structured in a very similar way to the UK measure, imposing a 3% levy on revenues generated in France by digital businesses with global revenues exceeding €750 million, of which €25 million or more is generated in France. 51. Spain’s digital services tax also has similar features, targeting revenues from online advertising, multi-sided digital interfaces and users’ data. It also imposes a 3% levy on revenues generated in Spain by digital businesses with global revenues exceeding €750 million, of which €3 million or more is generated in Spain.52 19. In 2016, the tax administrations of Israel issued a circular clarifying their position in relation to permanent establishment (PE) and VAT issues insofar as they apply to foreign suppliers of digital goods and services to Israeli residents. The circular was drafted so as to be consistent with the overarching principles then emerging from the OECD’s BEPS process. In respect of PE rules, the Israelis treat differently digital suppliers resident for tax purposes in jurisdictions with which Israel has a double taxation agreement to those with which it does not. In respect of the former, the new interpretation as set out in the circular essentially limits the extent to which exemptions from PE status for certain ‘preparatory or auxiliary services’ can be used. In particular, the circular notes that business operations which constitute a significant digital presence in Israel may not necessarily be considered to be exempt from constituting a PE. It further notes that a business may be deemed to have such a significant digital presence where, for instance, it has a significant number of contracts signed with Israeli residents via the internet, where services offered are used by many Israeli residents online, or where internet services are customized for Israeli users. The attribution of profits to the Israeli PE, deemed such by virtue of that significant digital presence, is done using the arm’s-length principle.53 In respect of firms tax resident in jurisdictions with which Israel does not have a double taxation agreement, however, the authorities’ existing ‘domestic source rules’ are used to establish whether that firm produces income originating in Israel. This means that a PE is deemed where the foreign firm has physical premises in Israel, where its operations are supported by an Israeli agent or representative, or where the firm has a significant economic presence in the country. Indications of a significant economic presence could include the provision of online services relating to Israeli customers, a high volume of online transactions with Israeli customers, the provision of online services customized to Israeli customers, etc.54 50 The UK published draft legislation in July 2019: https://www.gov.uk/government/publications/introduction-of-the- new-digital-services-tax 51 https://www.senat.fr/espace_presse/actualites/201904/creation_dune_taxe_sur_les_services_numeriques.html 52 For more information about recent developments in the EU see: https://taxfoundation.org/digital-services-tax- revenue-estimates/ 53 Deloitte, World Tax Advisor, 24 June 2016: http://newsletters.usdbriefs.com/2016/Tax/WTA/160624_2.pdf 54 Idem. 9 20. Italy adopted a Levy on Digital Transactions in 2017, effective since 1 January 2019. The levy will be applied at a rate of 3% on monetary transactions with both resident and non-resident suppliers of digital services to Italian resident B2B clients for which value is created by users or via user-generated content, and which is not captured by existing corporate tax rules. The tax liability is to remain with the supplier, although suppliers will be exempt where they do not carry out at least 3,000 taxable transactions in a calendar year. As in the case of India, however, the duty to collect and remit the tax lies with the purchaser. The levy will not be creditable against any other Italian taxes, except that Italian resident suppliers will be able to deduct it from their corporate tax liability. This transaction tax is designed to fall outside the scope of double tax treaties, and is expected to yield $235m per year (OECD, 2018). 21. In 2014, Hungary introduced an advertising tax on the net sales revenue (exclusive of VAT) for the sale of advertising time or space, by either resident or non-resident firms. The tax was broad in scope, covering traditional media, outdoor display and online advertising. In the case of online advertising, the display of advertisements in the Hungarian language was deemed sufficient to create a nexus for the payment of the tax in Hungary. While the supplier or publisher is liability to register with the tax administrations and to pay the tax, a secondary tax obligation is also imposed on the local advertiser where the latter cannot provide the tax administrations with a formal declaration by the primary taxpayer that it recognizes its tax liability. This is essentially an enforcement measure, and this secondary tax liability is not subject to the reverse-charge mechanism. In 2017, the authorities raised the rate of the advertising tax from 5.3% to 7.5% while maintaining the 0% rate on the first HUF 100m. The authorities have reported low compliance rates among non-resident suppliers, and therefore only limited tax revenues (OECD, 2018). D Multi-Sided Platforms – Ride-Sharing & Short-Term Accommodation Rentals 22. Definition: A digital supplier provides an online platform, which acts as an intermediary between drivers and passengers, typically in urban areas, or between accommodation hosts and guests. In return, the platform provider charges a fee or commission to the ultimate service provider. The tax treatment of the platforms themselves do not raise any tax issues that are not addressed in connection with the other business lines. However, the existence of these platforms raises some tax compliance issues, or opportunities in relation to the service providers operating via these platforms. The actions some countries have taken in response to the emergence of these platforms is described here, as it may be instructive for other countries. The subsequent policy discussion does not include any options that are specific to multi-sided platforms, although they may be relevant to them, depending on their business model. 23. Mapping transactions: In multi-sided ride-sharing platforms, there are multiple transactions in play. The most immediate transaction is that between the service provider (driver) and final consumer (passenger), which typically involves electronic payment through the platform itself. The digital firm takes a commission, typically up to 20% of the total transaction amount, with the remainder constituting income for the service provider. Thus, there is a clear income tax liability on the part of the service provider. Depending on the prevailing tax laws in the relevant state or 10 municipal jurisdiction, the initial transaction between service provider and final consumer may also generate indirect tax liabilities, whether in the form of a value-added or other forms of tax. 24. In the case of accommodation, the final consumer, or guest, makes on online payment to the service provider, or host, through the digital platform. This incorporates a percentage ranging up to 15% which is charged by the digital platform to the service provider. Transactions relating to furnished accommodation may be subject to VAT. There may be local taxes too, if temporary accommodation attracts a hotel tax or similar levy. Generally, it is the service provider who is responsible for making a declaration to the tax administrations in respect of income earned through such digital platforms and indirect taxes due. 25. Taxation issues: In common with other digital businesses, the platform provider may have a taxable presence in the country in which the services are provided and may be affected by other measures that target digital businesses. However, the novel issue raised by multi-sided platforms is the involvement of service providers in sectors where tax compliance is traditionally low. This has prompted some countries to try and exploit the emergence of these platforms to achieve improved tax compliance on the part of the service providers. In some cases, this has included measures to exclude a certain level of income from taxation. 26. Recent international experience: In France, digital platforms are required to provide the ultimate service provider with a breakdown of their tax liability and reporting requirements with respect to each transaction, as well as an annual statement of gross income. From 2019, it is expected that the digital platforms will be required to report this information directly to the tax administrations (HMRC, 2018). Estonia has allowed for service providers on some digital platforms to opt-in to having their data sent to the tax administrations so as to allow for pre- population of the individuals’ tax declaration forms (HMRC, 2018). Many States in Mexico charge ‘hospitality taxes’ on the provision of accommodation. Where this is the case, Airbnb, for example, collects and remits this portion of the payment directly to the relevant tax authority. In May 2018, Airbnb agreed to automatically report to the tax administrations in Denmark the income earned by its hosts through the platform. In parallel, the tax administrations increased the threshold under which such income would not be subject to income taxation. 27. As of March 2017, ride-sharing platforms in Uruguay have been subject to a dedicated regime which requires them to submit a monthly report to the tax office identifying all drivers and their income earned through ride-sharing activities. The ride-sharing platform is obliged, moreover, to withhold a proportion of each drivers’ monthly transactions, rising over time from $27 to $108. These withheld amounts are remitted to the tax office the following month and can be deducted by the driver from their eventual liabilities to the tax office and / or social security office. Where drivers do not have the requisite permits for the provision of passenger services, the online platform is liable for taxes, debts, fines and surcharges accrued by the driver. 55 At sub- national level, ride-sharing platforms operating in Montevideo, the capital, are required to register with City Hall, to report the names of all drivers with City Hall, and to withhold and remit a levy 55 https://www.bna.com/insight-uruguaytaxation-ecommerce-n73014481908/ 11 amounting to $0.06 for every kilometer driven. In addition, ride-sharing platforms also fall under the remit of new taxation rules governing the digital economy which became effective as of 1 January 2018. Such platforms were made subject to the non-resident income tax at an effective rate of 6% (or 12% levied on 50% of the transaction amount) and to VAT at a rate of 22% where both final consumer and ultimate service provider are Uruguayan residents (and half, or 11% when one or other is a non-resident. Responsibility for withholding the tax lies with the online platform, except where the client is a registered resident taxpayer for corporate income tax purposes.56 28. Belgium is another country to significantly increase the tax-free allowance for income earned through the sharing economy. Since mid-2016, income up to €5,000 earned through the sharing economy is taxed at the rate of 20%. With the application of a 50% allowance, however, this equates to an effective tax rate of 10% on such income so long as the total does not exceed €5,000 in any one tax year. The tax is withheld and remitted by the digital platform, thereby reducing the scope for evasion and minimizing the administrative burden on service providers. All payments must be made electronically so that they can be traced. When income exceeds the €5,000 threshold, it is deemed to be a professional activity, meaning that normal business taxes apply. However, since the letting of real property and movables are excluded from the regime, ride- sharing (such as Uber) falls outside the remit, as does the pure accommodation element of hosting guests (as opposed to the provision of ancillary services such as breakfast). 29. Since 2015, Airbnb has been supplying tax administrations in Ireland with information relating to the income of its hosts operating in that country. In 2018, as part of its enforcement efforts the Irish tax administrations wrote letters to property owners informing them that their tax affairs were under investigation, including in relation to income from the letting of short-term accommodation during the 2014-2016 period. 30. As of 2018, Spain has imposed an information exchange obligation on intermediaries, explicitly including multi-sided digital platforms, involved in short-term and tourist accommodation rentals. The intermediaries are obliged to inform the tax authority of each transaction relating to a property on Spanish territory. The new ‘model 179 form’ which the intermediary must submit requires the following information: name of property, name of person with the right to use the property if different from owner, identification of persons or entities renting, details of the property being rented, number of days the property is to be used for touristic purposes, the amount of the transaction, the contract or transaction number assigned by the intermediary, date when rental will commence, date of transaction, and payment details.57 31. In August 2018, Chile signaled that it would be introducing a freestanding 10% tax on imported digital services, which would also include multi-sided platforms such as Airbnb, although not ride-sharing platforms which are to be subject to separate, comprehensive legislation addressing also a range of issues beyond taxation. 56 https://home.kpmg.com/xx/en/home/insights/2018/05/tnf-uruguay-taxation-of-the-digital-economy-implementing- decree.html 57 https://www.agenciatributaria.es/AEAT.internet/alquiler_vivienda/apartamentos_turisticos.shtml 12 Annex 2: Indirect Tax and Foreign Suppliers of Digital Goods and Services There are two principal ways to collect VAT/GST on services provided by a foreign supplier: require the domestic customer to account for the tax by way of a recharge or require the foreign supplier to account for the tax. The second option is the approach recommended by the International VAT/GST Guidelines.58 Where the transaction is business to business, it is relatively simple to identify transactions involving foreign suppliers of digital services and for the domestic business to account for the recharge as part of its regular reporting for VAT/GST purposes. The supply of digital services to consumers typically involves a large number of low value transactions and a recharge payable by these retail consumers is hard to enforce and costly to administer, for taxpayers and the tax administration alike. Where a charge on the consumer exists in principle, it is usually not collected in practice.59 Ensuring a level playing field between digital and traditional businesses is an important consideration for VAT policy makers. If purchases made online are not subject to VAT but traditional purchases in a bricks and mortar store are, this will drive consumers to substitute online purchases for traditional ones. Furthermore, bricks and mortar retailers are likely to react by offering consumers the opportunity to buy “online” in their stores; they can then “collect” the goods immediately afterwards. The overall result will be a major erosion of the VAT base. If it makes sense to ensure that domestic suppliers in the digital economy account for VAT on their sales, what are the implications for foreign suppliers? Traditional models of collecting from the supplier are not effective, because they cannot be directly regulated by the domestic tax authority. As discussed, foreign suppliers of good and services that are wholly digital (games, movies, etc.) may have little or no physical presence in the country in which consumption takes place. But failure to subject consumption of goods and services purchased from foreign suppliers will disadvantage their domestic competitors and threatens to erode the VAT base over time. Foreign suppliers will have a price advantage and domestic digital entrepreneurs will have an incentive to relocate. The digital economy is characterized by high levels of international mobility. If foreign suppliers are required to account for VAT/GST, it will be important to have an effective compliance strategy to ensure the charge is effective. This will involve making use of third-party data, including data from domestic internet service providers (ISPs), to identify the supply of digital services by foreign providers to domestic users. The International VAT/GST Guidelines encourage tax administrations to make more use of instruments enabling exchange of information and mutual assistance in collection to support indirect tax compliance.60 Local legislation may also require all online traders who carry on businesses in, or with customers in, a country to register for non-tax reasons. This can be an additional aid to securing compliance with their tax obligations. 58 See Chapter 3 of the International VAT/GST Guidelines, OECD 2017. 59 Even if consumers are aware that there is a theoretical liability, the amounts involved for each purchase are trivial and there is often no obvious way to pay. Similarly, enforcement of the liability is not practical for the tax administration. 60 Ibid, paragraph 2.22. 13 The widespread adoption of the approach advocated in the International VAT/GST Guidelines reflects the fact that foreign suppliers are generally compliant. In terms of value, the market is dominated by a small number of highly visible e-commerce platforms and suppliers of digital content to consumers. Even where they have no physical presence in a jurisdiction, they will comply with the requirement to account for consumption taxes. However, countries need to put in place suitable mechanisms to make the compliance process simple for these suppliers. This is usually achieved by way of a simplified online process for registration, declaration and payment of the tax due. The OECD guidelines have been supplemented with practical guidance on these implementation issues.61 Countries need to consider eliminating any de minimis exemption of low value consignments from VAT. Allowing tax and tariff-free import of small value items by consumers tilts the playing field in favor of foreign e-commerce retailers, since the domestic retailer must charge the prevailing goods or sales tax and would have had to pay any relevant tariffs on merchandise imported for resale. A country’s ability to eliminate the exemption may be constrained by the terms of free trade agreements it has entered into. 61 http://www.oecd.org/tax/consumption/mechanisms-for-the-effective-collection-of-vat-gst.htm 14 Annex 3 Tax Administration’s Capacity in a Developing Economy ADMINISTRATIVE CAPACITY African Country A is making significant efforts to rebuild its state following periods of serious disruption. This rebuilding includes the enhancement of its capacity to carry out effective audit of MNEs operating there and collect income tax from those MNEs. While the total number of MNEs operating in the country is small, they account for a large proportion of the country’s income tax base. The income tax law affecting these MNEs is broadly in accordance international standards and includes comprehensive transfer pricing rules and most of the recommendations and standard developed in the OECD/G20 BEPS process. The tax administration is focused on capacity-building, for which there is top-level commitment and management support. Although the tax administration is relatively small, it has built up a team of 6-7 auditors to examine the returns of the largest MNEs operating in the mining sector, which presents the most significant tax risk. The team has benefitted over many years from assistance and training from international and regional organizations and from specialists provided by African and non-African countries. Nevertheless, Country A continues to struggle to build a sustainable capacity to effectively audit, and collect tax from, MNEs. In part, this is due to the nature of the international tax rules and prerequisites for their effective implementation. In addition, there are a number of institutional factors (many of which Country A has in common with other developing economies). International tax rules, including transfer pricing rules, typically have a number of features: - The ability to effectively apply transfer pricing (and related) rules in large MNE audits, requires auditor skills and experience that is generally recognized to require 2-3 years of practical experience to master. - The application of international tax rules frequently requires a large amount of data and other information, including information and documentation held outside the local country. Local fact-finding can be time-consuming and requires co-operation from the taxpayer and effective sanctions against non- cooperation. Accessing information held abroad requires exchange of information agreements with other tax jurisdiction, and the infrastructure to implement them. - The application of transfer pricing rules frequently relies on financial data and other information on transactions between unrelated non-related entities. (ie. data on ‘comparables’) - The application of transfer pricing rules rarely results in a definitive answer, with the result that negotiation to achieve resolution is typically seen in more experienced and developed tax administrations. In addition, most fact-patterns leave room for judgement and discretion, which bring risks of malpractice and corruption. 15 - Taxpayers need an incentive to engage with auditors; this may be lacking if the appeals and judicial processes are inexperienced in the issues, and MNEs perceive that judicial action is unlikely. - Most international tax rules are within the scope of tax treaties, which means that the results of an audit of an MNE may be subject to MAP. Inexperienced tax administrations often find it challenging to effectively negotiate with an experienced and well-resourced treaty partner. Institutional factors in Country A include: - A relatively small pool of officers from which to draw and train specialist auditors. The tax administration struggles to maintain a team of specialists with the critical mass required for long-term skills-transfer necessary for sustainability. - The cost and benefit of assigning the most skilled and senior staff to transfer pricing audits is not always clear as the main revenue sources are indirect taxes and investment of resources and the best minds to enhance administration and compliance management on these may well be more cost-effective. - Very little experience auditing and applying complex international tax rules in MNE audits. Most auditors are used to focusing on checking whether documentation exists rather than the appropriateness of the documentation provided by taxpayers. - Inexperience within the tax administration of negotiation and settlement with large taxpayers and their experienced professional advisors. - There is no experience in the appeals system or judiciary of considering technical law relating to MNE taxation - The tax authority has no access to comparable data required for the application of transfer pricing rules. - The country has limited international information-exchange instruments - There is a culture of poor taxpayer compliance, engagement and cooperation with the tax administration, perhaps because there is expected to be no consequences of non-compliance and non-cooperation - Poor IT and records systems - Management systems focused on financial (non-technical) audits and short-term revenue raising - Unease at rigorously auditing large MNEs, and applying penalties, for fear of a detrimental effect on FDI and country aid. Country A has experienced high- level political lobbying, and consequent interference, on specific audits. This has included representation by the jurisdiction of the ultimate parent company, which is a large aid donor. - Country A continues to face corruption issues. The risk of corruption is exacerbated in audit cases where a large degree of judgment and discretion applies. 16 Annex 4 Case Illustrations Country Case study 1 XYZ Group is a large multinational group whose ultimate parent entity and corporate headquarters are located in an OECD member country. The group carries out engineering services in locations throughout the world. XYZ has formed a subsidiary in Country B (XYZ-B) which carries out engineering services for third-party customers in Country B, including in the construction and mining sectors. XYZ-B has substantial operations in Country B, where it maintains an office in the capital city as well as capital assets such as trucks and machinery. It directly invoices its customers and meets all the costs of its business. It pays a royalty under a license agreement for the use of the XYZ trademarks and recompenses the head office for the costs of providing a number of centrally-provided services. Both of these payments have transfer pricing implications, but the audit focused on a third payment to a related entity in a low-tax jurisdiction (XYZ – LTJ). This payment is intended to recompense the XYZ-LTJ for indemnifying XYZ-B against commercial risks. The payment was designed to be set each year at an amount that left XYZ-B with a low profit margin, determined according to a benchmarking study of independent providers of engineering services. The annual payment was relatively large, and in practice stripped XYZ of a large proportion of its taxable profits. XYZ provided extensive transfer pricing documentation, prepared by a Big-4 advisor, to support the arrangements. The Country B tax administration took the view that the payment would not have been made at arm’s length between independent parties. An important argument to support this position was that local fact-finding had established that, contrary to the contention in the transfer pricing documentation, XYZ-B’s business was substantially managed by XYZ-B in County B and that all key business decisions regarding the management of the key commercial risks were made locally62. XYZ contented that personnel in XYZ-LTJ did in fact manage a number of risks at a high level, by setting, monitoring and managing for example, pricing policy, customer credit, foreign exchange risk criteria etc. In order to assess and analyze the fact-pattern, the Country B tax administration, and ultimately the courts, need to: 62 It should be noted that internationally agreed transfer pricing principles, including post-BEPS. allows risk (and the profit associated with that risk) to be reallocated to the entity in which it is managed and controlled, and which has the financial capacity to bear the risk, in cases in which these depart from the contractual terms. In this case, one of the arguments of the tax authority was that the risks were in fact controlled in Country B and it is thus inappropriate to shift the profit arising from those risks to the XYZ-LTJ. 17 - understand the business and sector in sufficient depth to identify the significant sources of entrepreneurial profit, and the business risk management and control activities associated with each - understand the group’s business processes, and, in particular a detailed understanding of XYZ-B’s operations and the actual management and decision- making procedures - understand the activities carried on in XYZ-LTJ, including the presence of any personnel with the ability and authority to take business decisions affecting the risks borne by XYZ-B, and who actual does so - Assess the entrepreneurial profit earned by XYZ-B arising from the management and control undertaken by personnel of XYZ-LTJ, and which is therefore allocable to XYZ-LTJ. In this case, it was established that personnel were located in XYZ-LTJ, but whether they were able to, and actually did, take entrepreneurial decisions related to the risks borne by XYZ-B was a matter of dispute between the taxpayer and the tax authority. Issues arising from this case study are: - the profit shifting created by the indemnity arrangements significantly reduced the profit recognized in XYZ-B - the information required to carry out a sufficiently robust audit on the group’s business operations, including those carried on in other countries, was very extensive and, in respect of information concerning foreign operations, problematic to obtain. The transfer pricing documentation provided no relevant detail. - The technical issues were very complex, requiring a knowledge of domestic law and the approaches contained in the OECD Transfer Pricing Guidelines - The taxpayer made available professionally-produced inter-company agreements and documentation to support their position and were supported by highly qualified advisors and lawyers. A number of observations may be made: - The volume of resources, and technical expertise needed, to carry out even this aspect of the audit is significant, and likely to be beyond that available in developing economies, and some developed countries. - If all information and analysis is fully carried out, there remains much scope for debate and disagreement with the taxpayer, and possibly the tax authority of ZXY-LTJ, on the extent, if any, of the profit allocable to XYZ-LTJ. Inevitably, agreement will involve negotiation and discretion. 18 - It is conceivable that business operations have been (or could be) structured in such a way as to place real decision-making functions in the hands of very few personnel located in XYZ-LTJ, thus potentially justifying significant shifting of profit earned in Country B. Country Case Study 2 ABC Group is a large multinational group whose ultimate parent entity and corporate headquarters are located in an OECD member country. The group carries out telecom infrastructure consultancy services in locations throughout the world. ABC-C is a subsidiary of ABC Group, located in Country C. It provides consultancy services to third party customers in the telecom sector in Country C. It has two offices in Country C which carry out sales and marketing activities as well as the provision of consultancy services. It employs eight permanent technical consultants. When required for the delivery of large projects, ABC-C engages highly qualified consultants, engineers and project managers (numbering over 100 in some years). These personnel are employees of ABC -LTJ, a group subsidiary located in a low tax jurisdiction. These personnel are not physically located in the low tax jurisdiction. Country C tax administration conducted an audit of ABC-C and established that ABC-LTJ contracts these employees to ABC-C at a rate of cost-plus 65%. It was also established that the same arrangements are used by the group throughout its global business. During the audit it was established that three senior employees and three support staff are based in ABC-LTJ. These employees carry out high-level management of the company’s contracting employee, who globally number about 400 at any one time. Day-to-day administration of the recruitment, contracts and remuneration of these employees is contracted by ABC-LTJ to offices in the head-office jurisdiction. It is established that ABC- LTJ is very profitable. The ABC transfer pricing documentation states that the three senior managers located in ABC-LTJ direct the internal consultancy side of the business. This includes determining the number and job-specifications of employees, approving salary, other remuneration and training, and managing and approving their deployment throughout the ABC group. The transfer pricing documentation contends that the activities of these employees in ABC-LTJ constitutes the key-profit-generating functions in the group and justifies the high level of profit recognized in the low tax jurisdiction. A number of observations may be made: - The taxation principles involved in cases such as these are complex, as is their practical application, which requires substantial amounts of fact-finding. In this case, for example. the tax administration would need a deep understanding of the businesses of both ABC-C and ABC-LTJ, and their business decision- making processes. - depending on the facts, there may be some plausibility in the taxpayer’s contention that the activities carried out by the relatively small number of employees in the low tax jurisdiction constitutes the management and control of key business risks, although there remains significant scope for differing interpretation of facts and the taxation principles that apply, as well as for 19 determining how much profit is appropriately allocated to ABC-LTJ. Inevitably, agreement will involve negotiation and discretion. - Arrangements such as these are commonplace. They can give rise to significant profit-shifting, backed up by extensive and plausible agreements and documentation. The resources needed to counter such arrangements are large and advanced skills are needed do so. Many developing economies, and some developed countries, lack the capacity to spot and effectively challenge such issues. - It is important to note, also, that schemes such as this one may plausibly achieve profit-shifting to low tax jurisdiction on the basis on placing key risk-taking functions in the hands of a relatively small number of employees which can then be physically relocated into a low tax jurisdiction. 20