Cooperation, Obligation, Taxation

(1)

Cooperation, Obligation, Taxation

A republican critique of corporate tax avoidance

----Thesis in partial fulfillment of the requirements of the MSc degree in political theory

Samuël van Dijck Student number: 3058409 Supervisor: dr. Bart van Leeuwen Department of Political Science Radboud University Nijmegen Number of ECTS: 18 Number of words: 32879 Date of completion: 24/03/2017

(2)

What is the difference between a taxidermist and a tax collector? The taxidermist takes only your skin. –Mark Twain (1902)

(3)

Chapter 1: Introduction

Benjamin Franklin once wrote a letter to a friend, Jean Baptiste le Roy, in which he expressed his great faith in the newly established American constitution but added that one could never be sure of its endurance, for “nothing can be said to be certain in this world except death and taxes” (Franklin, 1817, p. 266). However, while taxes may indeed have been unavoidable in the 18th_{century, they} certainly aren’t so today. One can hardly open up a newspaper without being confronted with some headline accusing yet another big multinational corporation of paying too little or even no taxes at all.

1.1. The Times They Are a Changin’

The main difference, in this respect, between the world Benjamin Franklin spoke of and the one we live in today is that ours is characterized by a profound international integration of markets in goods, services and capital (Gilpin and Gilpin 2001; McGrew 2014). In other words, we live in a time of substantial economic globalization, driven by changes in transportation technology and spurred on by the removal of trade barriers and capital controls, especially from the late 20th_{century onwards. Under these} circumstances, the international mobility of services, goods and especially capital increased manifold (Ostry, Loungani and Furceri 2016; Piketty 2014; Zucman 2015). The immense growth and increasing complexity of cross-border economic activities that followed these developments has and continues to do so today -dissolved the separation of the world into separate national economic entities in such a way that the it now more than ever operates as a singular system (Hiscox 2014; McGrew 2014).

This process of rapid and far-reaching globalization that is a characteristic of the modern economic system cannot, however, simply be extended to other domains as well. When we take a look

(7)

at the political sphere, matters are still much more firmly situated in a national framework. This is not to say that there are no areas in which we can find quite intensive inter- or even transnational political cooperation. Regarding peace and human rights, for example, we have the United Nations, and there are also important international economic institutions such as the World Trade Organization or the International Monetary Fund. Nevertheless, there are many aspects and topics of political decision making that remain almost entirely situated in national governments. One important example is the issue of (corporate) income taxation. Where capital today crosses borders virtually without any limitations, the power to tax it remains firmly bound to the nation state, grounded in the conviction that the power to tax is one, if not the, central attribute of sovereignty (Rixen 2008, p.5).

So the status quo is characterized, on the one hand, by states’ sovereign right to tax any income – be it capital (i.e. dividends or royalties) or labor (wages) income - reported within their jurisdiction in whatever way they want, and, on the other hand, by far-reaching economic globalization involving high mobility of capital, services and goods. But the former is a framework established at a time far before our current state of global economic integration. It is a tax framework that was designed, quite literally, with a different world in mind. And it is precisely the obsolescence of that framework in our modern globalized world that often leaves governments struggling to catch, or hold on to, the elusive taxes of especially corporations that operate in more than one country (Rixen 2008; Sassen 1996). The extreme international mobility of capital facilitates multinationals in moving their capital income (profits) from the country where it is created (usually high tax jurisdictions), to low tax jurisdictions commonly referred to as tax havens, with the goal of minimizing their tax bill. This is a practice called corporate tax avoidance, and it is the subject of this thesis.

(8)

8 Cooperation, Obligation, Taxation

1.2. Evasion vs. Avoidance

To start things off, we should first of all pay attention to the crucial distinction between tax evasion and tax avoidance. Both are ways of trying to minimize your tax payments, but the central difference between these two concepts is that the former is illegal while the latter is not (Dietsch 2011; Gravelle 2009; Palan, Murphy, and Chavagneux 2013). Tax evasion generally involves wealthy individuals (occasionally also multinationals) moving their assets into so-called secrecy jurisdictions, where they are able to hide it from the tax authorities in their home countries (Zucman 2015). For example, a wealthy Dutch business owner might move part of her financial assets to a Swiss bank where, because of Switzerland’s bank secrecy laws, it can be kept out of sight of the tax authorities in the Netherlands. Now in such a case it is not so that the Dutch business owner is legally exempted from having to pay taxes over that money, but rather that the Dutch government simply doesn’t know how much money she has and how much tax it should levy.

This is different in the case of tax avoidance, which basically entails multinational companies making use of the incoherent and outdated framework briefly introduced above. They use the high mobility of capital to, through all sorts of artificial constructions, move their profits across international borders and report them in low tax jurisdictions (Rixen 2008; Zucman 2014).1_{That means that} in every jurisdiction where they report their profits they are actually paying the taxes they are required to pay by law (Creedy and Gemmell 2011; Dietsch and Rixen 2014; Gravelle 2009). By minimizing their tax bill in this way, multinationals are, therefor, strictly speaking not doing anything illegal.

And this is also generally the defense that multinational companies offer for their tax avoidance strategies. They simply 1 Exactly how this is commonly done will be further explained in chapter 2.

(9)

state that, from their perspective, they are not doing anything wrong since they are meeting all their tax obligations under fiscal law (Atkinson 2015). This is for example the story that Google, Amazon and Starbucks put forth when they were summoned before the British Public Accounts Committee in 2012 to explain why they adopted such aggressive tax avoidance strategies, minimizing their tax payments in the UK. Against the companies’ defense the Chairwoman of the committee replied that, indeed, these strategies were not against the law as such, but that they definitely were cynical and unjust. “We are not accusing you of being illegal” she added, “we are accusing you of being immoral” (Ebrahimi 2012).

1.3. The Project

This thesis can be seen as a normative, philosophical, grounding of that statement made by the Public Accounts Committee. For it seems that indeed the companies’ claim that they were not doing anything wrong because they were not violating any laws is founded upon the assumption that law and morality are necessarily in alignment. In this thesis, however, I will argue that correlation between the legal and the moral domain does not necessarily exist (Shavell 2002; Taylor 1968). More specifically, I will provide a normative assessment of the practice of corporate tax avoidance and show that though it may be legal to avoid paying taxes where you make profits, it certainly isn’t morally acceptable. By doing so, I will be answering the central question of this thesis, which can be stated simply as: is corporate tax avoidance unjust?

Attempting to answer that question, though rare, is not new in political philosophy. The republican social contract approach to the subject adopted in this thesis, however, is. The issue of tax avoidance in relation to justice has, so far, been mainly addressed from either a global justice view (i.e. Dietsch 2011; Dietsch and Rixen 2014, 2014; James 2013) or from various consequentialist perspectives (i.e. van Dijck (2016) offers a consequentialist

(10)

comparative justice approach, while Stainer et al. (1997) argues for the injustice of tax avoidance on the basis of Mill’s harm principle and Peter Dietsch (2015) criticizes tax avoidance for violating state autonomy). The argument presented in this thesis, however, will be fundamentally deontological in nature, and has a focus on the nation state. In that sense, it is, also because of the contract theoretical foundation of the argument, closer to theories of corporate social responsibility (CSR) since it too argues for a corporation’s responsibility towards the societies in which they operate. However, taxation is a notoriously absent element in CSR theorizing, a lacuna this this thesis will fill for it argues that precisely taxes are a core element of a corporation’s social responsibilities (Dowling 2014). There exist some noteworthy exceptions that do mention tax as a part of CSR (i.e. Christensen and Murphy (2004), Hardeck and Hertl (2014) and Lanis and Richardson 2015) but they fail to ground the responsibility to pay taxes in strong normative principles, which is also a lacuna this thesis fills by grounding the duty to pay taxes in fairness. Moreover, the argument will be grounded in new empirical data that became available only quite recently.

Regarding that last element it is important to note that, traditionally, it has been extremely difficult to show that tax avoidance is actually taking place, due to a lack of information. Multinationals simply were (and mostly are) not required to provide country specific information, but only report on their worldwide operations as a whole (consolidated). Since 2015, however, European banks are required to report their profits, turnover, employees and the amount of taxes they pay on a country-by-country basis as per EU Capital Requirements Directive IV (Aubry et al. 2016). This thus offers a new and unique opportunity for politicians, researchers and the public to, under some assumptions, acquire empirical indications of tax avoidance. An opportunity that this this thesis will thankfully utilize.

(11)

After these factual indications of tax avoidance, then, will follow the central normative argument for the immorality of tax avoidance based on fairness. More specifically, I will argue that corporations’ economic liberty, narrowly defined as the ability to generate, hold or spend (capital) income, depends fundamentally on a social cooperative venture and that the principle of fair play puts upon those who participate in and benefit from that cooperative venture an obligation to the other participants to reciprocate and bear one’s fair share of the (fiscal) burdens. This is, in that sense, a normative foundation for the OECD principle of economic allegiance, which holds that, indeed, corporations should pay taxes were their real economic activity takes place (OECD 2013). Though such a principle for justice in taxation is quite intuitive, it is so far lacking a solid foundation in political theory/ philosophy. That is thus what this thesis seeks to provide.

However, the principle of economic allegiance, by itself, is too weak, I claim, to provide a complete argument for the immorality of corporate tax avoidance: often multinationals are paying taxes, just less than their fair share. I will therefor also argue that a fair share of the fiscal burdens of the state means a burden at least proportional2_{to income. The final normative claim of the thesis is} thus that fairness requires multinationals to pay taxes in the countries in which they generate income, at least in proportion to profits generated. This is not the case for multinationals that avoid taxes in high tax jurisdictions, and the above statement is therefor a normative argument against those tax avoidance practices and a positive answer to the central question.

Finally then, the thesis will look at the implications of this conclusion regarding what fairness requires. For if tax avoidance can indeed be qualified as immoral, then this naturally raises the question of how we should act so as to make sure corporations 2 Proportional taxation, as will be explained alter on, simply means that every taxpayer pays the same percentage, the same portion, of his or her income in taxes.

(12)

engaging in it will stop their aggressive tax planning and start meeting the obligations fairness puts upon them. This thesis will discuss two different ways of approaching this problem. On the one hand there exist the option of simply removing the opportunity to avoid taxes and thereby forcing multinationals to comply. On the other hand there is also a more positive option of trying to change a multinational’s motives, causing it to not avoid taxes even if it were still possible. The first relates to changing the law, the other to promoting corporate social tax responsibility. This thesis will look at both options.

1.4. The Structure: Chapter Summaries

As might already be clear from the section above, this thesis will involve a normative (or critical) and a more constructive (forward looking) part. However, for this critique to gain traction it is important that it is grounded also in a more descriptive or empirical element (Fraser and Honneth 2003). Any defensible normative assessment, that is, “must rely on factual statements about the world” (Caney 2006, p.2). To that extent, the first part of the thesis will be descriptive/ empirical.

Part I, that is, will start in chapter two with a closer look at the context of corporate tax avoidance and the factors that facilitate it. Chapter three will then explain the (logic behind the) main way in which multinationals exploit this context to avoid taxes - namely through profit shifting - and the two most important ways in which profit shifting is commonly carried out: intra-firm lending and transfer pricing. Afterwards, chapter four will introduce some newly available empirical evidence for profit shifting by several of the biggest banks in Europe.

When this descriptive/ empirical foundation of the thesis is established, part II will continue with the normative evaluation of the conclusions of part I. That is, part II will present the main normative argument against tax avoidance. To this extent it will

(13)

start, in chapter five, with the introduction of an important political theory that is commonly employed as a foundation for arguments against the justness of taxes: deontological libertarianism. Basically, deontological libertarianism argues, starting from a fundamental right to self-ownership, that as long as people acquired what they have through free and voluntary exchange, or through just initial acquisition, they have a property right in it. Others can then only justly acquire the property, i.e. income, of another if they receive it through free and voluntary exchange. Taxes do not meet this demand, since they are by definition coercive, and are therefore unjust (Finn 2006). And how can it be unjust to try to avoid an injustice being done to you?

Chapter six introduces an argument, introduced by Liam Murphy and Thomas Nagel (2004), against this libertarian perspective on taxation. They argue that, in fact, it doesn’t make any sense to argue against taxes from the perspective of property rights because taxes are prior to them. Property rights only exist, they claim, if there is a state, including a legal framework, which defines and upholds them. And such a state requires taxes for it to be able to exist. Because taxes are a prerequisite for property rights, it seems a logical error to criticize the former on the basis of the latter. I then follow up on this conclusion by arguing that property rights, and in fact several other government provided institutions, are a fundamental requirement for corporations’ economic liberty. Their very ability to generate, hold or spend an income depends on the presence of a state.

In chapter seven it is argued that, in the spirit of social contract theory, the state must be understood as a cooperative venture for mutual benefit, where the benefits in light of this project must be understood as, for example, the ability to acquire, hold and spend income (economic liberty). Following Dagger (1997), then, I will claim that such a cooperative enterprise must fall under the principle of play fairly which involves the duty to

(14)

reciprocate. That is, if one benefits from social cooperation (which corporations do if they make a profit) then one should reciprocate and contribute as well. From that perspective, multinationals have the moral obligation - not towards the state as such but towards the others on whose cooperation they depend - to pay their fair share of taxes.

Part III, then, will answer the question of what a ‘fair share’ of the tax burden means. To that extent, chapter eight will review two traditional principles that are supposed to answer that question and claim that ultimately only one applies, which prescribes that one should at least pay taxes proportional to income. Chapter nine will introduce an argument against these traditional principles and claim that the distribution of the tax base should be understood as an element of the broader issue of distributive justice. Nevertheless, however, it is argued that no theory of justice would allow for a regressive tax system to exist, and that therefore we can conclude from this perspective as well that justice requires one to pay taxes at least proportional to one’s income.

The fourth part of the paper will then look into the implications of this conclusion. In light of the fact that multinationals are not meeting the obligations, towards the communities in which they operate, that fairness imposes upon them, how should we respond? One possibility is, of course, to simply change the international tax framework and national fiscal law. This option will be explored in chapter ten. Chapter eleven will introduce a second approach to the problem. Rather than forcing corporations to comply we can also try to change their motivations and promote the disposition to forgo the opportunity to benefit without contributing even if it comes at some corporate costs. Force is only required if corporations are mere profit maximizers. If, however, we can cultivate a willingness to operate with the good of the community in which they operate in mind, force might not be

(15)

needed in order to make them fulfill their duties.. A final section of chapter eleven then suggests a possible direction one could take in contributing to such cultivation: the promotion of unions.

(16)

Part I

(17)

(18)

Introduction Part I

Part I will consist of three separate, though connected, chapters. Chapter two will specify further the most important characteristics of the current international corporate income tax framework. Chapter three, then, will follow with a sort of ‘tax avoidance 101’ and show how corporations that operate across borders can exploit the international corporate income tax framework and its loopholes to avoid paying taxes (or minimize their tax bills). Chapter four will present some empirical evidence of profit shifting taking place on the basis of newly available country-by-country reports that European banks are required to publish since 2015. Taken together, the three chapters of part I form the factual foundation on which the normative arguments of part II and part III are to be build.

Chapter 2 The International Corporate Income Tax

Framework

3

As was indicated already in the introduction, the current international framework for the taxation of multinationals’ income is characterized by the tension between economic globalization on the one hand, and nation-based fiscal sovereignty on the other. In spite of the rapid globalization of the economic sphere, there exists no comparable international architecture for, or cooperation on, matters of taxation (like there is for trade, for example), nor are national tax systems properly adjusted to a globalized economy and high capital mobility (Feld, Heckemeyer, and Overesch 2013; Hay 2014; IMF 2014; OECD 2014). Rather, current arrangements for the taxation of multinationals’ profits developed over the past

3 Note that (elements of) chapters 2 and 3 are partly based on/ taken from van Dijck (2016) Corporate Tax Avoidance and Justice, a master’s thesis written at the Erasmus University Rotterdam.

(19)

century without any substantial international coordination (IMF 2014, p.9).

In most of the developed world today, the taxation of corporate capital income was introduced around the time of the First World War as sort of a backstop to the personal income tax (Piketty 2014). Quite quickly, however, people found that tough corporate income taxation is relatively straightforward in a closed economy, it becomes much more difficult when corporations start to operate across borders and when the economies in which they operate tend to no longer overlap with fiscal jurisdictions. What should be done, for example, when several different countries seek to tax the profits of the same company? This situation could easily lead to undesirable outcomes like double taxation, which involves a corporation being taxed on its profits twice. In order to deal with such situations the League of Nations decided in 1920 to institute principles for the taxation of corporations with operations in more than one country (OECD 2014). These principles still more or less govern multinational taxation today, even though they have become highly questionable in light of the extreme degree of economic globalization that characterizes our time (Zucman 2015).

2.1. Two Principles

The first principle that is of central importance here is the source

principle, which basically entails that a country is entitled to tax

corporate income that arises within its jurisdiction (Gravelle 2009; OECD 2014). At the time, ‘arises’ was virtually equivalent to ‘reported’. Since capital and goods weren’t as mobile then as they are now, there existed no real difference between reported and generated profits. You simply reported your capital income in the jurisdiction where you acquired it. This has changed fundamentally due to economic globalization, which has caused a separation between profits reported and profits generated in a country (this will be further addressed later on (Dietsch and Rixen 2014).

(20)

A second important principle established by the League of Nations is the so-called arms-length principle, which holds that the entities within a corporate structure that are located in different countries must compute their profits individually and must do so, moreover, as if they were unrelated. This means that each entity must calculate its profits as if it were buying or selling at market price. The arm’s length principle must be seen as an addition to the source principle, because regarding multinational corporations it can sometimes be quite difficult to determine what the source country of it’s profits is (IMF, 2014; Zucman, 2014). Say there is a soda company in the Netherlands, called Dutch Drinks, which has a subsidiary in Canada, Canadian Cans, which produces the cans in which Dutch Drinks sells its sodas. Now, if indeed these two entities are viewed as one for tax purposes it can be quite difficult to see where exactly the profits are created: they are not generated only by the production of the cans nor solely by filling them with soda and selling them. All the profits finally end up with Dutch Drinks, because they sell the sodas. But it does not seem appropriate to tax all profits in the Netherlands: part of the added value was created in Canada. So where should profits be reported? What the arms length principle dictates is that, for tax purposes, Dutch Drinks and Canadian Cans should be treated as separate entities (even though they are parts of the same company) and calculate their profits independently. In other words, Canadian cans must ‘sell’ the cans it produces to Dutch drinks, its parent company, for market price, and accordingly determine its profits. Likewise Dutch Drinks is viewed as ‘buying’ the cans from its subsidiary in Canada, and puts it as a cost in its profit and loss account.

Just after these principles were agreed upon in the 1920’s whatever globalization existed began to recede and what came after were 60 years characterized by relative protectionism and economic nationalism (Helleiner 2014; McGrew 2014; Ruggie

(21)

1982). Over that period, foreign corporate profits in, for example, the US, accounted for only five present of total corporate profits while that is up to over forty percent today (OECD 2014; Ruggie 1982; Zucman 2014). For that reason, the problems that resulted from the institutionalization of the source and the arms-length principle did not become apparent until some 40 years ago.

With the advent of the neoliberal regime and the Washington Consensus in the late 1970’s came the far-reaching removal of trade barriers and capital controls (Hay 2014; McGrew 2014; Ostry, Loungani, and Furceri 2016; Piketty 2014; Ruggie 2002; Zucman 2015). Together with several profound developments in both transport and communication technology this caused a significant increase in the mobility of capital (Helleiner 2014; Piketty 2014; Zucman 2015). After hibernating over half a century, economic globalization returned at a level more far-reaching and profound than ever before. And it returned with a vengeance, for it hardly seems like a coincidence that the significant rise in the number of tax havens from the 1980’s onwards co-occurs precisely with capital’s increased ability to make the journey to these fiscal paradises (McGrew, 2014; Zucman, 2015). Multinationals could now, with great ease, transfer money and goods to their subsidiaries in foreign countries, which is exactly what they started doing as is evidenced by the fact that over half of all international trade today is made up out of trade within firms (McGrew, 2014). Multinational corporations started moving their capital income around within their corporate structures, enabling them to make use of - among many other things, such as less demanding labor markets - the fiscal regimes in the different countries in which they were incorporated. Where taxation regimes remained bound to national borders, corporations increasingly transcended them (Rixen, 2008; Sassen, 1996). How exactly corporations used this new ability to move profits around internationally to avoid paying taxes will be further explained in the following chapter. Before that,

(22)

however, it is important to address one more fundamental characteristic of the international corporate income tax framework that facilitates multinationals in their tax avoidance activities, namely the requirements for establishment or incorporation (or rather the lack thereof).

2.2. On Mailboxes and the Lack of Substance Requirements

It sounds obvious, but if, as a multinational, you want to move your capital between subsidiaries in different countries, then first of all you need to have subsidiaries in different countries. That is, you need to incorporate, set up or establish a legal business entity, in more than one country first. You need a presence in a certain national jurisdiction in order to be able to make use of, for example, its tax law. The important fact for the purpose of tax avoidance is that there are virtually no requirements to setting up shop, establishing a subsidiary, in a country. That is, there are no demands for any kind of economic substance that incorporation needs to meet: legal presence is sufficient for utilizing national tax regimes (Creedy and Gemmell 2011; OECD 2009; Palan, Murphy, and Chavagneux 2013). The consequence of such lacking establishment requirements has been a tremendous rise, also from the eighties onwards, in the number of “special purpose entities” or so-called “mailbox companies” (Bartelsman and Beetsma 2003; OECD 2009; Sikka and Willmott 2010). Such mailbox companies are business entities that have no employees, and that usually consist of no more than a small room with a computer, serving no other purpose than to provide a company with legal presence in a country, thus enabling them to make use of the tax rules that are on offer there.4_{This lack of substance requirements for companies} that want to incorporate in a certain country, and make use of its tax regime, is an essential element in the system that facilitates 4 This sometimes leads to very absurd situations. In the relatively small office building at 200 Prins Bernhardplein in Amsterdam, for example, over 10.000 businesses are registered.

(23)

corporate tax avoidance (Buettner et al., 2015; Creedy & Gemmell, 2011; Egger & Raff, 2014). Decisions on where a multinational will establish it’s real economic activity (i.e. factories or research and development facilities) depend for the biggest part on factors other than favorable tax regimes, such as a well-educated population, cheap labor, presence of natural resources, or well developed infrastructure (IMF 2014; UNCTAD 2015). But these are not things generally offered in tax havens, and so setting up a subsidiary with real economic activity in such a jurisdiction is not something corporations would want to do. If making use of Bermuda tax law required multinationals to have, say, a factory there, then this would make tax avoidance an expensive activity, since Bermuda lacks virtually all of the factors that are required for making such substantive economic activity worthwhile (Zucman, 2015). Lack of substance requirements enables multinationals to set up mailbox companies in low tax jurisdictions, and make use of tax regimes there, while at the same time having their factories and research and development facilities in locations with, for example, a well-educated population, extensive legal system and good infrastructure, which are by definition high tax jurisdictions since such facilities are paid for precisely through tax revenue. Yes, multinationals will keep on using tax havens as long as this would maximize their profits. But if making use of a country’s tax regime were to require substantial economic activity, it would become much less likely that moving into a tax haven will indeed have a maximizing effect (Dietsch and Rixen 2014; Edwards and Keen 1996; Keen and Konrad 2014; Rixen 2008).

2.3. Summarizing

Summarizing, we have a system in which countries independently decide on their tax regimes, and where companies can incorporate in different countries to make use of those tax regimes without having to attach any actual economic substance to that

(24)

incorporation. Furthermore, multinationals can move capital between their various subsidiaries in different countries with great speed and relative ease. Finally, this situation is governed by two principles: each country subsidiary independently calculates and reports their own profits at an arm’s length price and these profits are then taxed in the country where they are reported. How companies exploit this system to avoid paying taxes we will turn to now in chapter two.

(25)

Chapter 3. Tax Avoidance 101: How it’s Done

Tax avoidance by multinationals can take on several different forms and often involves highly complex structures (Feld, Heckemeyer, and Overesch 2013; Gravelle 2009; Palan, Murphy, and Chavagneux 2013; Slemrod and Yitzhaki 2002). A complete discussion of the issue would be enough to fill several theses, which is why this chapter will give only a basic introduction to the subject. Even the most complex tax avoidance structures, however, are build around the same basic concepts introduced here (Gravelle 2009). This means that though this section might not provide all details of the issue, it will certainly be sufficiently characteristic of the problem for us to work with it in the rest of the thesis.

The fundamental idea behind tax avoidance by multinationals is simple: report high profits in jurisdictions with low tax rates and report low profits in high tax jurisdictions. Corporate tax avoidance thus generally consists of moving the profits you make between different entities, in different countries, within your corporate structure so as to make it end up in those places where it is taxed the least, thereby reducing the overall tax bill. This strategy for avoiding taxes is commonly called profit shifting (Creedy and Gemmell 2011; Feld, Heckemeyer, and Overesch 2013; OECD 2009, 2013). There are two main strategies that multinational corporations employ to shift their profits between countries from one subsidiary to the other: transfer pricing and intra-firm lending (Creedy and Gemmell 2011; IMF 2014; OECD 2013; Tanzi and Zee 2000). Each will be briefly introduced in turn.

3.1. Intra Firm Lending

In a basic understanding, profit is what is left of turnover when all costs and expenses have been deducted from it. Pre-tax profit, then, is what is left of turnover after costs and expenses but before

(26)

taxes have been deducted. I will, in this thesis, mostly be discussing profits before tax and, unless indicated otherwise, therefore use ‘pre-tax profits’ and ‘profits’ interchangeably.

Now what most multinationals that want to shift their profits from high tax country A to low tax country B do, is artificially increase the costs they incur in A and report them as income in B. For if revenue stays the same and costs are increased then the profit in country A will decrease. Likewise if the costs in country B stay the same and revenue increases then profits will rise. Since profits, or capital income, are in the end what is being taxed (and not revenue), doing so will thus lower the overall tax bill of the company. One tool for increasing costs in one country and reporting them as income in another is intra-firm lending.

In order to explain how intra-firm lending works it might be good to go back to Dutch Drinks, the fictional soda company introduced earlier. This company consisted, so far, of a Dutch ultimate owner (Dutch Drinks), and a Canadian Subsidiary that produces the cans (Canadian Cans). Now, let’s, for simplicity, assume that in both countries the corporate income tax rate is 25% and that the company as a whole makes one billion dollars in pre-tax profits a year. This would mean that over a year, Dutch Drinks would have to pay $250 million in taxes, and that it would be left with $750 million as the actual post-tax result. But, assuming that a company has the goal of maximizing shareholder value, it will continue to look for ways to increase the amount of income it is ultimately left with post-taxes. And it can increase this amount, non-surprisingly, by avoiding taxes. In order to do this the company starts by establishing a new subsidiary in Bermuda, called Bermuda Booze. Bermuda Booze is no more than a broom closet in an office building in Hamilton with no employees, but due to a lack of substance requirements in Bermuda that is enough to establish a legal presence enabling the company to make use of the beneficial Bermuda tax regime (say a 2 percent tax on corporate income).

(27)

What the company does now is to have Bermuda Booze give out a loan to Dutch Drinks, its owner, of, say, $900 million. Now, this is of course quite strange because in fact Bermuda Booze and Dutch Drinks are the same company and it doesn’t make sense to give out a loan to yourself. And indeed, on the company’s overall balance sheet this kind of loan doesn’t change anything. But when one takes out a loan, one has to pay interest over it. And this is where the profit shifting begins. Bermuda Booze gives out a loan to Dutch Drinks and accordingly requires Dutch Drinks to pay interest over that loan. This interest rate it is more or less free to set within certain limits set by the market price. Let’s say that Bermuda Booze gives out the loan at 10% interest. This means that over one year, Dutch Drinks has to pay $90 million in interest to its Bermudian subsidiary. These $90 million are thus an additional expense to the Dutch company (lowering profits there) and extra income for the subsidiary located in Bermuda (increasing profits). Accordingly, overall pre-tax income in the Netherlands and Canada together is decreased by $90 million to a total of $910 million and pre-tax income in Bermuda is now $90 million. The $910 million is taxed at 25% leaving $682,5 million and the $90 million is taxed at 2% leaving $88,2 million. So then the total post-tax profit of Dutch Drinks is $770,7 million. This means that effectively Dutch Drinks has managed to lower its tax bill with $20,7 million, thereby reducing the tax revenue of the Dutch state.

3.2. Transfer Pricing

The second important method that firms apply to shift profits from high-tax to low-tax jurisdictions works through the pricing of goods and services sold between affiliates (Bartelsman and Beetsma 2003; Eden 1998; Grubert and Mutti 1991; OECD 2009). Though it is similar to intra- firm lending in some respects, transfer pricing is in fact the form of corporate tax avoidance that is generally seen as most important and, indeed, also as most costly (Feld, Heckemeyer,

(28)

and Overesch 2013; Sikka and Willmott 2010).

Let’s start from a very basic example. Remember that the arms-length principle says that subsidiaries in different countries must calculate their profits separately, as if they were two unrelated entities dealing with each other on the market (which is also why they can give loans to each other and why a company can pay itself interest). So say that Dutch drinks sells 500 million cans of soda a year at $6 a piece, generating a turnover over of $3 billion. Now let’s furthermore assume that the costs of producing such a can of soda are $1 in Canada (for producing the can) and $1 in the Netherlands (for filling and distributing the can). Canadian cans thus produces 500 million cans and sells them to its Dutch parent for 3 dollars each, causing Canadian Cans to make $1 billion in profits (3-1*500m). Likewise, Dutch Drinks makes $1 billion in profits (6-3-1*500m) and so the company as a whole makes $2 billion in pre-tax profits. Now since in both countries the tax rate is 25%, the company as a whole will end up paying $500m in taxes and will ultimately be left with $1.5 billion in post-tax income. Notice how, since tax rates in both countries are the same, it doesn’t really matter for the company as a whole at what price Canadian Cans sells the cans to the parent company, since, in the end, it will end up paying 25% taxes over its pre-tax profits. This is to show that shifting profits between high-tax jurisdictions is of no, or only very little, benefit to the company as a whole. But there is, of course, a way to reduce the tax bill. To do that, the company will make use of a sort of middleman. That is, instead of selling directly to its parent Canadian Cans will now first sell the cans to Bermuda Booze, which will then in turn sell them to Dutch Drinks. Canadian Cans, however, does not sell the cans for the same $3 a piece to Bermuda Booze, but will rather only ask $1.10 per can. This means that the Canadian subsidiary will now make a pre-tax profit of $50 million (1.1-1*500m). Bermuda Booze, then, will in turn sell the cans to Dutch Drinks for $4.90. Dutch drinks, spending $1 on filling

(29)

and distributing the cans, will then also have pre-tax profit of $50 million (6-(4.90+1)*500m). Bermuda Booze, however, is now making a very significant part of the total profits. It is buying the cans for $1.10 and selling them for $4.90, so it is making a pre-tax profit of $1.9 billion (4.90-1.10*500m). As a whole, the company is still making $2 billion in pre-tax profits. Post-tax profits have changed significantly, however. Pre-tax income in Canada and the Netherlands is taxed at 25%, which means that together the Dutch and Canadian companies pay $25 million in taxes (2*50m*0.25) and are left with $75 million in post-tax profits. Pre-tax income in Bermuda is taxed at only 2% percent, which means that Bermuda Booze is paying $38 million in taxes (1.9b*0.02) leaving it with $1.862 billion in post-tax profits. As a total the company now pays only $63 million (25+38) in taxes and is left with $1.937 billion in after tax profits instead of the $1.5 billion it would have had left if it didn’t route the cans via Bermuda Booze. And this in fact does not even require the cans to actually be shipped to Bermuda. Remember that the Bermuda subsidiary is only a mailbox. The cans would have to be there only on paper, but never have to actually set foot on Bermudian soil. It is truly an artificial construction that allows the company to avoid paying $218 million in taxes in Canada and the same amount in the Netherlands.

Now this is a quite extreme example. The arms-length principle determines not only that subsidiaries in different countries must be seen as separate but also that they must trade more or less at market price. This means that Canadian Cans will have to sell their cans at a price that cannot be too far from the market price of cans, and the same goes for Bermuda Booze if it is to resell them to Dutch Drinks. But the appropriate price of a product can usually only be established by looking at comparables, which means that there can be quite some wiggle room which, at large quantities, can facilitate the avoidance of huge amounts of taxes (Bartelsman & Beetsma, 2003; Sikka & Willmott, 2010;

(30)

Zucman, 2015).

And this is even more the case when we look at so-called intangibles, such as services or intellectual property, which is where the main problem with transfer pricing lies (Bartelsman and Beetsma 2003, 2003; Grubert and Mutti 1991; Zucman 2014). This is so, because it is generally very difficult to find comparables for such intangible assets and thus vey hard to establish the appropriate market price (Eden, 1998; Gravelle, 2009; Zucman, 2015). What is, for example, the market price of the name ‘Amazon’ or of Starbucks’ proprietary coffee blend? How should we value a companies’ specialization in some service? Multinationals are very much aware of these problems of establishing a market price for intangibles and exploit it to the fullest. The way they do it is somewhat of a middle ground between transfer-pricing and intra-firm lending. The basic idea is, like with intra-intra-firm lending, that instead of decreasing pre-tax profits by lowering revenue (i.e. selling cans for less or buying them for more) the same is achieved by increasing costs. But, rather than giving out a loan, our fictional company could now for example have Bermuda Booze offer specialized financial management services to Dutch Drinks. The accountants that offer this supposed advice do not even have to be located in Bermuda (in fact they would not even have to be actual accountants), they would just have to be registered as employees of the subsidiary there. And specialized financial advice is thus particularly something that it is hard to put a prize on. A company can easily argue that there is in fact no other business out there that offers the same kind of expertise. And since Bermuda Booze does not offer its services to a third party a market price is hard to establish. So this facilitates a situation in which Dutch Drinks, the mother company, pays a huge fee to its subsidiary in Bermuda where that fee goes into the books as income, and is thus hardly taxed at all, while that same fee is a cost in the Netherlands where it pushes down taxable income.

(31)

Arguably, the most important source for such constructions is intellectual property (IP) (Bartelsman and Beetsma 2003; Grubert and Mutti 1991; Sikka and Willmott 2010). Say that Dutch Drinks has a soda called Cheap Cola and that it owns the IP rights to that brand name as well as to the brand’s logo. Now, what it can do is transfer these IP rights to a subsidiary it has in a tax haven, say Bermuda Booze. This means that now any party that wants to use the name or logo of the Cheap Cola brand will have buy a license and pay a fee, or ‘royalties’, to Bermuda Booze, and so does the parent company itself. Again, the ‘right’ market price of such goods is extremely hard to determine, if not impossible. For what is the value of the brand name and logo of, for example, Coca Cola? One could easily argue that there is no comparable brand and that there is thus no market price. This means that a subsidiary, Bermuda Booze for example, holding the IP rights to a brand can ask its parent company or another company in the corporate structure virtually any price it wants for the use of that brand name. Again the price the buying party has to pay reduces taxable income in the country where it is located and similarly the receiving party sees its taxable income increase, but is located in a country where royalty income is hardly taxed at all (such as Ireland where Google has most of its patents and trademarks, or the Netherlands where companies like Ikea or bands like U2 and the Rolling Stones have ‘offices’ holding their IP rights).

It is also important to notice that when a multinational wants to avoid taxes its options are not necessarily limited to only one of these strategies. Over the past 40 years or so, companies, or rather the accountancy firms that work for them, have become increasingly inventive in combining many of these different methods involving ever more complex artificial corporate structures enabling them to shift their profits and substantially reduce their tax bills sometimes even to zero (Gravelle 2009; OECD 2013).

(32)

It is important to emphasize how profit shifting, particularly through transfer pricing and intra firm lending, is enabled precisely by the context that was described in chapter two. First of all, of course, the boom in capital mobility since the 1980’s facilitates multinationals in moving money from one location (i.e. the Netherlands) to another (i.e. Bermuda) without any restrictions. If there were still capital controls in place this would not be possible or at least not as profitable. Secondly it is clear to see how the lack of substance requirements is crucial to these tax avoidance practices. If in fact setting up a subsidiary in Bermuda required not only a mailbox but, for example, also a factory or a big office with lots of employees actually working there then this would seriously reduce the incentives to use Bermuda for tax avoidance purposes. For establishing a factory on Bermuda would involve great costs and would not be of much value since the country does not offer any of the factors that would make it attractive for real economic activity (i.e. good infrastructure, a well-educated and large population, a strong and large market, etc.). In fact, tax havens are often precisely those countries that are not attractive locations for serious real economic activity. And this in some sense constitutes a vicious circle since many of the factors that make a country attractive for real investment, are precisely those things that are financed by tax revenue. To this point I will return later on. Finally then it should also be clear how the source and the arms-length principle play a facilitating role. If indeed a multinational were simply seen as one entity and taxed as one entity then profit shifting would be pointless (for in the end the profits stay within the same company). The source principle, then, does not allow for taxation of profits reported outside one’s own jurisdiction. That is, the Dutch state cannot tax the profits reported by Bermuda Booze in Bermuda even though they may have been generated within the Netherlands.

(33)

(34)

Chapter 4:

Profit Shifting: Evidence from European Banks

The two chapters above have given a description of the international corporate income tax framework and the ways in which multinational corporations can exploit it to avoid paying taxes. However, though the opportunities are plenty and the incentives seem clear, evidence for profit shifting has traditionally been very hard to find. It is possible, of course, to see that, for instance, Starbucks is reporting a loss in the UK and is consequently not paying any taxes there, but it is much more difficult to show that in fact this is the case because the company is shifting the pre-tax profits they make in the UK to a tax haven. This is the case because companies are only required to report their results on a consolidated basis, which means that they merely have to publish the figures for the company as a whole. So Starbucks only has to report how much profits it makes in total, how much taxes it pays in total, etc. For that reason, it is not possible to see if profits are being shifted within the corporate structure. If profits are shifted from a subsidiary in country A to a subsidiary in country B then the total amount of profits doesn’t change and so if you only get the total amount this shifting of profits is undetectable.

4.1. Sources and methodology

Things are (slowly) changing in this respect, however, which is why this chapter is able to give some empirical indications of profit shifting to substantiate the content of the previous two chapters and to provide an even stronger foundation for the chapters that are to follow. Due to great efforts made by both civil society organizations as well as many political parties all over Europe, the European Union (Capital Requirements Directive IV) has decided in 2013 that, from 2015 onwards, all financial institutions operating within the EU are required to publish, besides their standard annual reports, a country-by-country specification of their turnover,

(35)

profits, employees and tax payments (Aubry et al.2016). In effect, this means that in 2016 all big European banks published such a report for their 2015 results. The reports of eight of the biggest banks in Europe, from three different countries, will be analyzed in this chapter.5_{Though these reports cannot provide direct evidence} of, for example, intra-firm lending or transfer pricing, they can be used to, under some assumptions, provide indications of whether profit shifting is taking place. The remainder of this chapter is directed at providing such indications on the basis of the 2015 country-by-country reports of 6 large European banks (two German banks (Deutsche Bank and Commerzbank), three UK banks (Barclays, HSBC and RBS) and two French banks (BNP and Société Générale).

What country-by-country reporting attempts to do is generate a picture of, on the one hand, where actual economic activity (represented by turnover and employees) is taking place, and, on the other hand, of where profits are located (Murphy 2012). The assumption behind the determination of profit shifting, then, is that the distribution of profits should run more or less parallel to the distribution of real economic activity represented by turnover and employees. Regarding turnover, for example, this means that if a company has more turnover in country A than it does in country B, then more profits should be located in country A as well. This can also be read as the assumption that there are no substantial differences in the profitability (so the ratio of turnover and profit) between countries. Regarding employees the assumption is then that there are no substantial differences in the average productivity (in terms of profit) of an employee. So the average employee in Germany should produce as much profit as an employee in Switzerland. This means that if there are substantial differences in productivity or profitability between countries then this might be 5 The banks under consideration are Deutsche Bank and Commerzbank AG from Germany, BNP Paribas and Société Générale from France and HSBC, Barclays and Royal Bank of Scotland from the UK.

(36)

an indication of profit shifting.

What I have done, based on Richard Murphy’s (2012) methodology for the analysis of country-by-country reports, is first of all divide the group of countries in which the 7 banks are present into two categories: ‘tax havens’6_{and ‘home countries’. The group} of tax havens are constituted by a number of countries with highly beneficial tax regimes, while the home countries are those countries in which the various banks have their headquarters (Germany, France and the UK), which are all high tax countries. Next I determined the following indicators:

Turnover: the total amount of turnover reported by the 7 banks together specified for the total group of countries, for home countries and for tax havens.

Profits: the total amount of profits reported by the 7 banks together specified for the total group of countries, for home countries and for tax havens.

Employees (FTE): the total amount of FTE reported by the 7 banks together specified for the total group of countries, for home countries and for tax havens.

Labor productivity: the average employee productivity of the 7 banks taken together expressed as the amount of profit per FTE (total amount of profit in home countries divided by the total amount of FTE there), specified for home countries and for tax havens.

Home country profitability: the average rate of profitability of the 7 banks taken together expressed as the percentage of turnover that becomes profit, specified for home countries and for tax havens.

I have then, for each indicator, also calculated how it relates to the total of the banks as a whole (expressed as a percentage).

4.2. Data and Interpretation: Indications of Profit Shifting

6 List of countries considered to be a tax haven can be found in Palan et al. (2013, p.41-44)

(37)

The results of the country-by-country data analysis can be found in the following table:

Indicator Value % of total

Total turnover (in $m) 220637

-Home country turnover (in $m) 94848 43%

Tax haven turnover (in $m) 39960 18%

Total profits (in $m) 32315

-Home country profits (in $m) 1988 6%

Tax haven profits (in $m) 19207 59%

Total full time employees 940125

-Home country full time employees

346023 37%

Tax haven full time employees 101379 11%

Home country labor

productivity (in $m)

0,01 17%

Tax haven labor productivity (in $m)

0,19 551%

Home country profitability 2% 14%

Tax haven profitability 48% 328%

The results in the table are quite convincing. When we look, first, at turnover we see that the 7 banks combined generate revenue of over $220 billion. Of this turnover, 43% percent is reported in the home countries and 18% in countries that can be characterized as a tax haven. Of the total amount of 940000 workers that the banks employ together, 37% is working in one of the home countries while only 11% of total employees is employed in a tax haven. This shows that the real economic activity of these 7 banks, represented by turnover and employees, is located only for a very small part in tax havens and for the greater part in non-tax havens of which the home countries are the most important segment. How different are things for profits! As we can see from the table, the banks together generated over 32 billion dollars in pre-tax profit in 2015. 59% of those profits are reported in tax havens while only 6% are reported in the home countries. This is a staggering difference and already a Part I: Avoiding Taxes 37

(38)

clear indication that profit shifting is taking place. If indeed the assumption holds that on average the profitability of business and the productivity of an employee does not vary significantly between countries then we should see a picture in which profits and economic activities are distributed among the different jurisdictions in more or less the same proportions. But that certainly is not the case. We can see the same if we look at the two other indicators. On average an employee in one of the home countries earned about $10.000 in profits for the banks in 2015. Their colleagues in tax havens however were about 19 times as productive, generating about $190.000 in profits for their employers. Similarly, it is clear from the table that business in the home countries had a profitability rate of only 2 percent. This means that, on average, about two cents remain as profits from every dollar turnover. In tax havens this is 24 times as high. There almost half of every dollar turnover is preserved as pre-tax profit.

It is thus evident that on the basis of the country-by-country reports the banks are required to produce and the calculations made, business seems to be much more profitable in tax havens and employees there much more productive. But there are no reasons, or so we assume, that profitability and productivity would be higher in tax havens than in non-tax havens. In any case we could quite safely say that there is no reason that productivity (19x) and profitability (24x) would be this much higher in fiscal paradises.

A final step allows us to estimate the profits that are detracted from the tax bases, the total amount of taxable income, of the home countries. Since we assumed that there are no reasons to think that productivity and profitability should substantially vary between countries, we could say that in fact in every country these values should be (more or less) the same as they are for the banks overall. Now average labor productivity for the banks as a whole is $34.000 per FTE and average total profitability is 15%. So, based

(39)

on the assumptions we made we can argue that these are the values that would apply in each country if in fact no profit shifting were taking place. This means that we can calculate the profits we would expect these banks to report in the home countries, if they were not shifting profits, in two ways: one based on labor productivity and one based on profitability (though the principle is the same). Regarding FTE productivity we simply multiply the amount of employees in the home countries with the average employee productivity of the banks as a whole. On the basis of that calculation we would expect the banks to report a profit of about $12 billion in the home countries, nearly $10 billion higher than what they actually report there. The same calculation for tax havens shows that we would expect the banks to report only about $3.5 billion there instead of $19 billion, a difference of over $15 billion. From this calculation it also becomes evident that the home countries are the biggest losers: two thirds of the total amount of profits overreported in tax havens seems to come from them. Doing the same calculations with profitability gives us a similar picture with the home countries losing out on about $11 billion. These are profits that the home countries would have been able to tax if they weren’t shifted out of their jurisdictions. Instead, these high tax countries get to tax only a fraction of the profits generated there. And yes, over that fraction the banks are usually paying the taxes they are required to pay. In 2015, BNP Paribas for example paid a tax rate of 40% in France over the profits actually reported there. But, again, this is only over reported profits while the data suggests that the expected profits of BNP Paribas in France were about $1.5 billion higher. This means that the tax rate it pays over profits generated in France is only about 5%.

It is important to keep in mind that these are only rough estimates. The assumptions on which they are build are not rock solid: it is very likely that there does exist some variation between profitability and productivity between countries. However, such Part I: Avoiding Taxes 39

(40)

variation would undermine only very small results. The differences between tax havens and home countries that we have found can certainly not be characterized as such and should therefor be interpreted as serious and significant indications of profit shifting taking place. And, consequently, the results are also an indication of tax avoidance for the shifting of profits happens clearly in the direction of low tax jurisdictions. Profits are shifted to fiscal paradises where much lower tax rates apply than in the countries in which they are mostly generated and from where they are shifted.

Conclusion Part I

Part I was meant to function as a factual and empirical basis for the normative project of the thesis that follows in part II to IV. To this end, it started, in chapter two, with the description of the economic and fiscal framework in which multinationals operate today. Most importantly it emphasized the mismatch between economic globalization and the increased mobility of capital that came with it on the one hand, and the outdated source and arm-length principle combined with a lack of substance requirements on the other. Companies can relatively easily set up a subsidiary in another country that then for tax purposes is seen as a separate entity of which the profits that it reports are to be taxed by the jurisdiction in which it is located. These companies can then, utilizing the free movement of capital, employ several methods, such as transfer pricing and intra-firm lending, to shift profits from one subsidiary to another within their corporate structure, so as to increase pre-tax income in low pre-tax and decrease it in high pre-tax jurisdictions. Exactly how these methods work was described in chapter three. Chapter four then provided some empirical evidence, which became available only recently, in order to show that indeed we have very strong reasons to believe that profit shifting is taking place on a large scale.

(41)

confirm another point made earlier: economic activity requires certain facilities that are financed by tax revenue. States that facilitate the development of real economic activity will almost necessarily be high tax states, and indeed we saw that the most substantial economic activity, expressed as turnover and employees, is taking place in high tax countries. But because of the high taxes in such countries, companies also have an incentive, so to speak, to get their pre-tax profits out of there. There thus exists a tension: multinationals have an incentive to go to high tax countries for generating profits, but at the same time they are also motivated to get their profits out of there. This is thus by definition a framework in which profits are made in one country but the taxes that are supposed to pay for the facilitation of making profits are not paid in the same location. There seems to be something inherently unjust about this situation. It seems only fair that corporations pay the taxes they are required to pay in those countries where they have their economic activity, something that the OECD calls the principle of economic allegiance. Part II will provide a normative argument for that principle.

Cooperation, Obligation, Taxation