2.3.1 Tax haven: No or very low effective taxation

The broadest definition of a tax haven is any jurisdiction whose tax laws interact with those of another enabling the avoidance of taxation in that other jurisdiction.^[32] Thus, one could say, a tax haven is relative as every jurisdiction can be a tax haven in relation to another: “a tax haven is like beauty in the eye of the beholder.”^[33] Even countries practising a 35 percent corporate income tax (CIT) rate, such as Malta, may be considered “tax havens” from the Polish perspective, where the CIT rate is 19 percent. It follows from the fact that a Polish taxpayer participating in the profits of the Maltese company is entitled to receive a refund of six-sevenths or even all of the tax paid by this company.^[34] As a result, the effective tax rate of the income generated by the Maltese company and received by the Polish taxpayer will amount to 5 percent or less.^[35]

However, as Brian Arnold put it with respect to tax havens,

Any attempt to measure the relative tax burdens in various countries is extremely difficult and must take into account not only income taxes and corporation taxes, but also commodity, payroll, and other indirect taxes and the value of government services and benefits. Inevitably, judgments about relative tax burdens are very rough.^[36]

Given the difficulties in defining a “tax haven” from a relative perspective, it may be impossible to provide a general and objective definition. Instead, the perspective of every country should be evaluated on a separate basis and in light of the particular circumstances of relevance to taxation. This study does not ignore “high tax” jurisdictions insofar as they may be used for tax avoidance purposes.

Nevertheless, for the sake of simplicity and in order to highlight certain important matters discussed in the study, the term “tax haven” will be considered as a jurisdiction with no or very low effective taxation on all types of income or on such income as may be easily diverted to a CFC established in such a jurisdiction. This is also commensurate with the old nomenclature of the OECD regarding “classical tax haven”.^[37]

As far as I am aware, the OECD does not take the no or very low-taxation criterion into account when defining tax havens insofar as the effective exchange of tax information and transparency are the only things that matter in this connection.^[38] The point is, however, that this study focuses on tax avoidance, not tax evasion, and for tax avoidance purposes the main criterion is no or very low-taxation, while for tax evasion it is the lack of exchange of tax information.

Further, jurisdictions that do not tax income at all or only marginally tend to avoid exchanging tax information and information on the clients of their banks and the participants of entities established on their territories.^[39] So while the lack of an effective means of tax information exchange and transparency matters primarily for tax evasion, not tax avoidance, excluding these definitional factors in this study does not mean that tax havens as defined here would not correspond with the definition of a tax haven under the current international standards.^[40]

2.3.2 Low-tax jurisdiction

The meaning of “low-tax jurisdiction” is concisely determined as a jurisdiction that (i) has low effective taxation on all types of income or on such income as may be easily diverted to a CFC established in such jurisdiction; and (ii) is engaged in the exchange of tax information under a relevant international agreement. The existence of the second factor entails a distinction between “low-tax jurisdiction” and “tax haven” under the current OECD nomenclature and a partial distinction between the two in the present study.^[41]

Low effective taxation is relative and the threshold below which a jurisdiction may be seen as a “low-tax jurisdiction” is usually given as a certain percentage of corresponding effective taxation in a state applying CFC rules. For example, in France^[42] and Italy^[43] the threshold is less than 50 percent; in Portugal less than 60 percent^[44]; in Sweden less than 55 percent^[45]; in Norway less than about 66.6 percent^[46]; less than 75 percent in Mexico,^[47] Spain,^[48] South Africa,^[49] the UK ^[50] and Poland^[51]; and finally less than 90 percent in the US.^[52]

3.1.1 Foreign direct investment (FDI)

According to the OECD, FDI involve the direct or indirect ownership of 10 percent or more of the voting power of an enterprise resident in one economy by an investor resident in another and it is therefore assumed that such a relationship usually leads to the exercise of significant influence by this investor on an enterprise resident in a foreign country.^[66] In my view, however, a person who retains 10 percent or more of a foreign company’s voting rights cannot be automatically considered as having or not having a definite influence over the foreign company – a controlling stake whereby he can take decisions that determine its activities. A person may hold less than 10 percent of the shares of a CFC, but still exert real/factual control over the foreign company, if, for example, the remaining ownership structure is widely spread or control exists due to an agreement between the participant and the foreign company, especially if this company is directly owned by only ten persons holding 10 percent of its shares per capita. The opposite may also be true, i.e. a 10 percent or more ownership may not necessarily confer decisive control over a foreign company if such control is attributed to other persons.^[67] As a matter of fact, the OECD confirms some of the issues I have dealt with above, but it concludes that “the recommended methodology does not allow any qualification of the 10 percent threshold and recommends its strict application to ensure statistical consistency across countries.”^[68]

Moreover, the OECD indicates that FDIs include, among other things, “inward and outward financial transactions/positions between directly and indirectly owned incorporated and unincorporated enterprises.”^[69] This means that FDIs include pure financial flows between parent companies and their subsidiaries, including the transfer of highly mobile incomes to CFCs from their controlling participants. It may therefore be assumed that realizing FDIs through companies established in low-tax EU jurisdictions is likely to be linked to efforts to avoid tax through the use of CFCs if these jurisdictions do not coincide with places where genuine investments are usually made. Distinguishing between tax driven and non-tax driven transactions concerning FDIs is therefore fundamental in order not to confuse strategic uses of FDIs for legitimate economic activities (i.e. non-tax driven activities) with their use for the sole or predominant purpose of avoiding taxes.

In order to ascertain whether taxpayers are avoiding taxation through the use of CFCs, globally and, in the case of Poland, specifically, it is useful to determine whether they make FDIs through companies located in low-tax EU jurisdictions.

3.1.2 Multinational enterprise (MNE)

There is no a precise and legally binding definition of a MNE, but for the purposes of this study, the open-ended definition of the OECD is relevant. It says that MNEs:

usually comprise companies or other entities established in more than one country and so linked that they may coordinate their operations in various ways. While one or more of these entities may be able to exercise a significant influence over the activities of others, their degree of autonomy within the enterprise may vary widely from one multinational enterprise to another. Ownership may be private, State or mixed.^[70]

The literature indicates that MNE can be defined as an enterprise which wholly or partly owns, controls, and manages value-adding activities in more than one country.^[71] MNEs typically consist of groups of companies that are highly interrelated economically and legally and are often under the control of an ultimate parent.^[72] It is therefore likely that the companies used for tax avoidance purposes, for instance CFCs, will be found within the group of MNEs.^[73]

3.1.3 Special Purpose Entity (SPE)

Beyond the data on FDIs and MNEs, I also refer to other relevant data and statistics. In that regard, it should also be explained that these global data and statistics do not refer explicitly to foreign companies used for tax avoidance purposes, usually called CFCs, but to entities that, from a functional point of view, may largely be identified with such companies as Special Purpose Entities (SPEs). Even though there is no legal definition of an SPE, according to the OECD, a company is usually considered as an SPE

if it meets the following criteria: i) the enterprise is a legal entity, a) formally registered with a national authority; and b) subject to fiscal and other legal obligations of the economy in which it is resident; ii) the enterprise is ultimately controlled by a non-resident parent, directly or indirectly; iii) the enterprise has no or few employees, little or no production in the host economy and little or no physical presence; iv) almost all the assets and liabilities of the enterprise represent investments in or from other countries; v) the core business of the enterprise consists of group financing or holding activities, that is – viewed from the perspective of the compiler in a given country – the channelling of funds from non-residents to other non-residents. However, in its daily activities, managing and directing local operations plays only a minor role.^[74]

This definition implies that SPE in its structure and substance vastly corresponds to a CFC. Moreover, the core business of an SPE, which in the OECD’s view is a group financing or holding activities, does not necessary rely on channelling funds from non-residents to other non-residents, as in the case of conduit companies. The group financing or holding activities may either be used for receiving and accumulating an income over long periods and thus serve for tax deferral and/or to repatriate this income later in jurisdictions without taxation or with low-taxation, which corresponds to the core function of CFCs and base companies.

3.1.4 Base companies vs conduit companies

According to the OECD nomenclature, a base company may be considered: (i) a company or an entity treated for tax purposes as body corporate that is (ii) predominantly established in a low-tax jurisdiction by its controlling participants for the purposes of a deferral of taxation on their income (primary sheltering of income) and/or repatriation of this income without taxation or with effectively low-taxation, i.e. tax avoidance (secondary sheltering of income).^[75] In that regard, the function of base companies – accumulating income in low – or no-tax jurisdiction for purposes of tax deferral/tax avoidance – should be distinguished from the function of conduit companies which is “channelling” income from one country to another for purposes of “treaty shopping” and/or “directive shopping”.^[76]

Taking into account the function of CFCs,^[77] it is clear that these companies can be identified with base companies, not with conduit companies. This finding also stems from the research led by the OECD on base companies, which suggests that CFC rules (“Subpart-F type counteracting legislation” in the OECD’s language) constitute specific legislation against the sheltering of income in low-taxed base companies situated abroad.^[78]

Nevertheless, in many cases, tax avoidance schemes, including CFC schemes, do not rely exclusively on base companies, but also on conduit companies, and the combined effects of both will yield tax benefits to the participants.^[79] Likewise, although Polish data and statistics do not specifically refer to companies that are engaged in tax avoidance practices, it appears that they also take into account situations in which such companies are very likely to be engaged.

3.2.1 Global data

The IMF’s Co-ordinated Direct Investment Survey presents interesting data on FDIs realized via entities established in tax havens. In 2010, Barbados, Bermuda and the British Virgin Islands received more FDIs (in sum 5.11 percent of global FDIs) than Germany (4.77 percent) or Japan (3.76 percent) and more outward investments were transacted from those countries (in sum 4.54 percent) than from Germany (4.28 percent). Comparing the number of FDIs by country, in 2010 the British Virgin Islands were the second largest investor in China (14 percent), after Hong Kong (45 percent), and before the United States (4 percent). Also in 2010, Bermuda was the third largest investor in Chile (10 percent); Mauritius the top investor country in India (23 percent); while Cyprus (28 percent), the British Virgin Islands (12 percent), Bermuda (7 percent), and the Bahamas (6 percent) were among the top five investors in Russia.^[80]

Interesting information can be obtained from the OECD Investment Database on FDI stock positions, which are composed of equity and debt (intercompany loans) and represent the value of the stock of direct investments held at the end of the reference period (year, quarter, or month).^[81] Data on FDIs composed of equity and debt are especially interesting from the perspective of tax avoidance because financial instruments related to equity and debt are often used in cross-border intercompany loans for tax avoidance purposes via the hybrid qualification of such instruments whereby the financial instrument is treated by the issuer state as debt and by the holder state as equity, often resulting in a payment of tax deductible interest by the issuer, which is then exempted from taxation in the holder’s state due to the treatment of this payment as dividend (participation exemption).^[82]

These data show that total inward stock investments into the Netherlands (usually used for establishing conduit companies for treaty and EU Directive shopping purposes)^[83] for 2011 were equal to USD 3,207 billion and the amount of USD 2,625 billion was realized via SPEs (note the close similarity in structure and substance of SPEs and CFCs); ^[84] outward stock investments from the Netherlands were equal to USD 4,002 billion, with about USD 3,023 billion being made through SPEs. In the case of Luxembourg (a low-tax jurisdiction which is most likely often used for exploiting CFCs for tax avoidance purposes),^[85] total inward stock investments for 2011 were equal to USD 2,129 billion, with USD 1,987 billion being made through SPEs. Outward stock investments from Luxembourg were equal to USD 2 140 billion, with about USD 1,945 billion being made through SPEs.^[86]

The jurisdictions indicated above have tax systems whereby income (at least various types of passive income) derived by base companies, in the vast majority of cases CFCs, is not taxed at all or only very moderately. The income derived by SPEs functioning as conduit companies (usually established in the Netherlands for treaty and EU Directive shopping purposes), in turn, is usually not taxed because their taxable expenses almost equal their gross income, i.e. their taxable base is close to zero.^[87] In that regard, it is also worth reiterating the close similarity in structure and substance of SPEs and CFCs^[88] and further, that Luxembourg can be seen as a low-tax jurisdiction for exploiting CFCs for tax purposes.^[89]

Furthermore, senior members of the OECD Centre for Tax Policy and Administration (CTPA) have commented on the context of the implementation of OECD/G20 efforts to curb tax avoidance by multinationals, saying that around 80 percent of global gross income is currently (2015) generated by 15 percent of MNEs with large networks of controlled subsidiaries around the world.^[90] These controlled subsidiaries are not necessarily CFCs, but one may assume that at least some of them are, since companies within the group of MNEs are actually controlled by an ultimate parent and their income is dispersed around the world, including low-tax jurisdictions and tax havens.^[91]

These global data imply that a significant percentage of global inbound and outbound investments are realized through entities whose characteristics largely match those of companies established in tax havens and low-tax EU jurisdictions for CFC tax avoidance purposes.^[92] It is therefore highly likely that the taxation of a substantial amount of global income is deferred or avoided via CFCs established within the EU. This assumption may be further verified by a brief discussion of selected international tax avoidance schemes within the EU.

3.2.2 A brief discussion of selected international tax avoidance schemes within the EU

There are several examples of tax avoidance schemes operated by global MNEs whose features are not dissimilar to those of CFC tax avoidance schemes (see Section 2.4). In this section, I will mention some that have attracted the analytical attention of scholars. I will not analyze these international schemes myself, although I do cite the appropriate sources and recapitulate the conclusions that scholars have reached.

Some tax avoidance schemes are particularly interesting since they are known across the world and because of the amount of avoided tax involved. They include schemes run by Google Inc. (“Double Irish Dutch Sandwich”)^[93]; Apple Inc.^[94]; Caterpillar^[95]; and Starbucks.^[96]

After analyzing information from parliamentary hearings in the US and the UK, scholars estimate that Apple, Google, and Caterpillar have managed to reduce effective tax on their foreign-sourced income to about 1, 2, and 4 percent, respectively.^[97] With respect to Apple, the hearing confirms that from 2009 to 2012, “it successfully sheltered U.S. $44 billion from taxation anywhere in the world.”^[98] In relation to Starbucks UK, the hearing indicates that this MNE “has paid £8.6 million of U.K. corporate tax during its 15-year existence on revenue of more than £3.4 billion.”^[99] In the cases of Apple, Google, and Caterpillar, scholars have concluded that one of the most important elements of their schemes is their circumvention of US CFC rules due to certain flaws affecting the latter, most importantly the check-the-box regime and use of EU law.^[100] Similar conclusions can be drawn with respect to other tax avoiders, such as Microsoft, Hewlett-Packard, and Amazon.^[101]

Thus, CFC tax avoidance is possible in relation to income generated by various types of activities, including intangibles (Apple, Google, Microsoft, Hewlett-Packard, Amazon) and tangibles (Caterpillar and Starbucks) and therefore passive as well as active income. Most importantly, the recognized schemes have been effective not only with regard to income from activities that can be located remotely from the physical location of their customers in the EU or elsewhere (Apple, Google, Microsoft, Hewlett-Packard, Amazon, and Caterpillar), but also when such activities must be located in the customer’s location (Starbucks) within the EU.

In it also worth referring to CJEU judgment of 13 March 2007 in Test Claimants in the Thin Cap Group Litigation case^[102] in which US multinationals, Pepsi and Caterpillar, used EU-based companies, in Luxembourg and Ireland (a Swiss company was also involved in case of Caterpillar) among others, in order to provide loans to the UK company.^[103] I am not here to judge whether or not the said EU-based companies were used for CFC tax avoidance purposes, but Antony Ting convincingly argued that the Caterpillar scheme could be seen as a tax avoidance scheme^[104] while the Pepsi scheme, based on the use of the Luxembourg tax jurisdiction, is on the list of the International Consortium of Investigative Journalist revealing secret tax agreements approved by the Luxembourg tax authorities.^[105]

3.4.1 Polish data

Based on the “Report on Polish Direct Investments in 2011” by the Department of Economic Analysis and Forecasting at the Polish Ministry of Economy (Report 2011),^[106]Figure 1 below indicates a rapid increase in the value of Polish FDIs as of 2002,^[107] from PLN 228 ​​million in 2002 to PLN 2 702 million in 2005, even glancing PLN 7 132 ​​million in 2006.^[108] For the period 2007–09, the following figures are given: PLN 3 952 million, PLN 3 011 million, PLN 3 745 million. In 2010 and 2011, FDIs reached 5 484 million and 5 276 million, respectively. The significant increase in the value of FDIs beginning in 2004 may be explained by Poland’s accession to the EU. Polish investors could now enjoy freedom of establishment within the EU and free movement of capital without territorial limitations, both of which are important for FDIs.

Figure 1:

The value of outward FDIs of Polish investors 1994–2011 in millions of PLN.

As the 2011 Report (p. 16) shows, as of 2011, the highest value FDIs were located in low-tax jurisdictions with preferential income tax regimes, Luxembourg and Cyprus; the amounts were EUR 2 851.6 million and EUR 1 257.7 million, respectively. In comparison, considerably smaller amounts were invested in high-tax countries with robust tax systems, such as France and the US: EUR 479.8 million EUR and 236.7 million, respectively.

The “Report on Polish Direct Investments in 2013” by the Department of Strategy and Analysis at the Ministry of Economy (Report 2013)^[109] also gives Cyprus first place as the preferred destination for FDIs, with the value peaking at EUR 443.8 million. Interestingly, this report also says that FDIs in Cyprus were made by companies established there although their transactions were usually conducted and their financial resources usually based outside Cyprus. ^[110] These companies then functioned as typical base and conduit companies that can be qualified as SPEs,^[111] implying that the FDIs were made for tax avoidance purposes.

Another interesting observation follows from the 2013 Report. Over the 2000–13 period, most of the Polish FDIs took place in European low-tax jurisdictions, such as Luxembourg (EUR 6 877 million or 33.3 percent) and Cyprus (EUR 4 514 million, or 21. 9 percent), and in jurisdictions with preferential tax regimes, for example the Netherlands (EUR 1 776 million, or 8.6 percent) and Switzerland (EUR 1 968 million, or 9.5 percent).^[112]

As of 2013, the most attractive among Asian countries was Singapore (EUR 131 million),^[113] taking the place the United Arab Emirates (EUR 19.2 million) held in 2011.^[114] Both are low-tax jurisdictions.^[115]

The 2011 Report (pp. 17 and 18) also underlines the divergence between these Polish FDIs in the above-mentioned countries, particularly Cyprus, Luxembourg, Switzerland, and the Netherlands (and in 2010 Belgium and the Isle of Man), and typical FDIs, such as acquisitions of foreign companies, acquisitions of significant shareholdings in foreign companies or direct investments through the establishment of a completely new branch of a Polish company for purposes of conducting major economic activity. Instead, they relied exclusively on financial flows between parent companies and their subsidiaries, although they remain classified as FDIs.^[116]

Further, the companies located in the above-mentioned jurisdictions were initially controlled by Polish taxpayers and mostly operated as holding companies that were disengaged from any form of genuine economic activity beyond the transfer of dividends between companies in the group. Neither did these companies have subsidiaries in the countries of their incorporation where they were engaged in economic activity. The Report states that, for example, Russia constituted the biggest market for Polish FDIs as of 2011. In that country, Polish FDIs relied mostly on manufacturing and trading (a dynamic development of a network of Polish shops). In comparison, Polish FDIs in Cyprus were mainly involved in establishing holding companies there, without any subsidiaries engaged in active business in that country.^[117]

Taking this information in the round, the majority of the income derived from the FDIs of Polish investors in the last 10 years would appear to originate via CFCs in low-tax jurisdictions. It is therefore highly likely that the sole or predominant purpose of such investments was to create CFCs in low-tax jurisdictions in order to avoid taxation in Poland or elsewhere. It may also be assumed that tax avoidance via CFCs constitutes one of the most widely used ways of avoiding tax by Polish taxpayers and that Poland has suffered from the lack of effective anti-CFC tax avoidance provisions in the period in question.

That being said, it will be interesting to identify and analyze the CFC tax avoidance schemes set up by Polish taxpayers in order to see if there is any correlation between the economic data and statistics on the one hand and the locations of these CFCs as well as the use Polish taxpayers make of them on the other hand.

3.4.2 Selected tax avoidance schemes by Polish taxpayers

There are several instances of CFC tax avoidance schemes put in place by Polish taxpayers that match the features of the schemes described earlier in Section 3.4. This section focuses on three well-known Polish schemes. The first has been named “Double Cypriot Olive with a Polish Potato” in an obvious nod to Google’s CFC tax avoidance scheme known as “Double Irish with a Dutch Sandwich”. The second is called “Luxembourg CFC Scheme” because the Polish taxpayers use the tax jurisdiction of Luxembourg to avoid taxation. The third is known as the “LPP Scheme”, named after the Polish enterprise involved in avoiding taxation via this scheme. After presenting and analyzing these schemes, I will draw the tax-related implications stemming from their exploitation by Polish taxpayers.

I identified each of these schemes by analyzing individual advance tax rulings, publicly available on the website of the Ministry of Finance,^[118] or investors’ relations on the official website of the company involved in the scheme (“LPP Scheme”).^[119] These schemes have not been previously described in the literature by other scholars.^[120] It should also be noted that the analysis is essentially illustrative and may not reflect precisely all relevant tax and investor issues. The accompanying figures were created by myself. For the sake of simplicity and clarity of presentation, I ignore general partners in limited partnerships (LPs, spółka komandytowa in Polish), which are normally Polish limited liability companies (LLCs, spółka z ograniczoną odpowiedzialnością in Polish)^[121] and I do not identify every single investor in the investment funds, such as the Private International Collective Investment Scheme (PICIS).^[122]

Due to the amendment to the Polish tax law that entered into force very recently (1 January 2017) in an effort to prevent tax avoidance, some elements in the schemes are not or may not be currently valid from a taxation point of view. Nevertheless, the presentation and analysis of those schemes remain relevant for understanding the legal, economic and factual contexts existing in Poland at the time of flourishing of blatant CFC tax avoidance schemes.

3.4.2.1 “Double Cypriot Olive with a Polish Potato”: Total tax avoidance of Polish-sourced income

To set up the scheme illustrated below in Figure 2, a Polish taxpayer, being an individual who runs a business in Poland via a Polish LP, first transfers participation in the Polish LP (rights and duties) personally owned by him to a Cypriot Alternative Investment Fund (AIF)^[123] he has recently established. Transfer of participation rights in the Polish LP is done as an in-kind contribution in exchange for shares in the Cypriot AIF, and is a tax-neutral transaction (there is no income to be taxed).^[124] Consequently, the Cypriot AIF participates almost 100 percent in the Polish LP wholly or mainly engaged in active business activity in Poland. The Polish taxpayer, in turn, owns shares in the Cypriot AIF and is entitled to receive dividends from this entity. The Polish taxpayer is an influential investor in the AIF and most likely has the right to select the members of the board of directors of the Cypriot AIF and give them binding instructions. It is therefore very likely that the Polish taxpayer exerts definite control over the Cypriot AIF.

Figure 2:

“Double Cypriot Olive with a Polish Potato”: total tax avoidance of Polish sourced income.

The Polish LP constitutes a foreign PE for the Cypriot AIF under the Poland-Cyprus tax treaty. The income generated by the Polish LP is exempt from taxation in Poland (the source state) as well as in Cyprus (the residence state). The former is based on the subjective exemption from CIT in Poland of collective investment funds with a registered office in a Member State of the EU (Cyprus).^[125] The latter relies on the exemption from taxation under the Cypriot domestic tax law according to which the profits of a foreign PE are exempted from income tax in Cyprus if the PE derives directly or indirectly more than 50 percent of its profits from an active business.^[126] This condition is fulfilled by this scheme since the Polish LP is wholly or mainly engaged in an active business such as construction or pharmaceutics.

Next, the Polish taxpayer establishes a tax transparent LP in Cyprus, to which he transfers shares in the AIF, personally owned by him in advance, to the Cypriot LP. The transfer of shares in the AIF is done as an in-kind contribution in exchange for participation in the Cypriot LP, and is a tax-neutral transaction (there is no income to be taxed).^[127] Consequently, the Cypriot LP owns all the shares that were previously owned by the Polish taxpayer and therefore the Cypriot LP is entitled to receive income from the AIF in the form of dividends.

The Polish taxpayer, in turn, is entitled to receive income from the Cypriot LP on an ongoing basis, since under the Poland-Cyprus tax treaty, it constitutes a foreign PE for him. Since the Cypriot LP is tax-transparent from the point of view of both Cypriot and Polish tax laws, the look-through approach applies to dividends paid by PICIS to the Cypriot LP – the recipient of the dividends is considered to be a foreign tax resident, i.e. a Polish taxpayer, but not a Cypriot taxpayer. Therefore no taxes are withheld on dividends paid to non-resident individuals, because all dividends paid out of Cyprus to non-residents are exempt from tax under Cypriot domestic law.^[128]

Dividends received by the Polish taxpayers via the Cypriot LP qualify as profits from a foreign PE, i.e. business income taxable under Article 7(7) in conjunction with Article 10(4) of the Poland-Cyprus tax treaty. The credit method applies to such income in Poland pursuant to Article 24(1)(b) of the Poland-Cyprus tax treaty. The foreign business income of Polish taxpayers, however, qualifies under Polish domestic law as income from economic activity and falls within the scope of the so-called tax abolition relief.^[129] The tax effects of this relief are analogous to the exemption method.^[130] Consequently, this income is not taxed in Cyprus nor is it in Poland.

As a result of this tax avoidance scheme, business income originally generated in Poland avoids taxation altogether on a current basis. Since Polish domestic law allows for the final non-taxation of the income repatriated from Cyprus, this scheme cannot be deprived of its tax effectiveness by amending the treaty with Cyprus. It could be though by amending the Polish domestic law.^[131] The tax effects of this scheme were confirmed by several advance tax rulings.^[132]

The directors of some tax chambers, however, have challenged the opportunity to exempt from taxation an AIF’s income derived in Poland because the conditions for exemptions from CIT under Article 6(1) point 10a of CITA are not completely satisfied. ^[133] This position, however, was very recently rejected by the Supreme Administrative Court (SAC) in its judgment of 24 July 2015.^[134] The SAC decided that a Cypriot AIF (PICIS) was similar to any Polish investment fund subjectively exempted from taxation and that its income, derived via Polish tax-transparent (before 1 January 2014) partnerships limited by shares, was not subject to CIT in Poland.

Other directors refused to provide tax abolition relief arguing that the income received by Polish individuals via a Cypriot PE did not derive from a non-agricultural economic activity under Polish tax law but from financial capital (e.g. divided, capital gains etc.)^[135] There is also the argument that the qualification of income as business profits under Article 7 of the Poland-Cyprus tax treaty cannot be binding for the purpose of applying the Polish domestic tax law and therefore providing tax abolition relief. The tax authorities have not been challenged on this count by taxpayers before the courts. On the one hand, it seems that the considerable risk of losing the case convinced taxpayers to drop the scheme rather than try to defend its probity before the court. On the other hand, the most recent individual tax rulings confirm the positions of Polish taxpayers that income derived by them via a Cypriot LP, which merely participates passively in profits of other companies, qualifies for the tax abolition relief.^[136] This implies that the scheme may be in force and brings desired tax effects for the Polish taxpayers.

In my view, the Polish tax authorities could have achieved the same result by simply concluding that the Cypriot LP did not constitute a PE for the Polish individual since it was not carrying on a business activity as required by the autonomous tax treaty’s meaning.^[137] On this basis, its income cannot be considered as business income under Article 7(7) in conjunction with Article 10(4) of the Poland-Cyprus tax treaty but remains dividend income under Article 10(1) of the treaty. Neither therefore does its income qualify for tax abolition relief. However, due to the lack of appropriate anti-avoidance provisions, such as CFC rules, the Polish tax authorities attempted to use various, and in my view inappropriate, interpretative means to prevent Polish taxpayers from successfully exploiting the blatant CFC tax avoidance scheme they had earlier accepted. Had the Polish tax authorities appropriately interpreted the notion of PE under the Poland-Cypriot tax treaty, there would be no issue with the scheme at all. It is not necessary to invoke anti-avoidance provisions to achieve this.

Due to the amendments in Polish tax law, in particular in relation to taxation of investment funds,^[138] the scheme analysed above and illustrated below does not allow for avoidance of taxation on income derived via a tax transparent partnership from 1 January 2017. It means that taxation on business income originally generated in Poland cannot be currently avoided. However, this scheme could have been used for such purpose until 31 December 2016. The scheme still allows for the indefinite deferral of tax on income derived via a Cypriot CFC.

3.4.2.2 The “Luxembourg CFC Scheme”: Total tax avoidance of income generated in Poland on an ongoing basis by an individual

The procedure and the tax-related consequences of setting up this CFC scheme, which involved the establishment of an investment fund in Luxembourg and exchange of its shares for the rights and duties of the Polish LP, is analogous to that described immediately above in Subsection 3.4.2.1 concerning the Cypriot AIF.

Figure 3:

“Luxembourg CFC Scheme”: total tax avoidance of income generated in Poland on an ongoing basis.

The Luxembourg Investment Fund (Lux. Inv. Fund) participates in almost 100 percent of the rights and duties of the Polish LP involved in an active business activity in Poland. The Polish LP constitutes a foreign PE for the Lux. Inv. Fund. The income generated by the Polish LP is exempt from taxation in Poland (state of source) as well as in Luxembourg (state of residence). The former is based on the subjective exemption from CIT in Poland of collective investment funds with a registered office in an EU member state, i.e., Luxembourg.^[139] As a result, all types of income generated by the Polish LP are exempted from taxation in Poland. Neither is this income taxed in Luxembourg, since the Lux. Inv. Fund is subjectively exempted from income tax in that state so that its profits, derived via a PE in Poland (the Polish LP), are not taxed in Luxembourg.

There is no income tax on dividends paid by the Lux. Inv. Fund to the Polish taxpayer (individual) because dividends paid by the Lux. Inv. Fund are exempt from withholding tax.^[140] Such dividends were exempted from taxation in Poland under Article 24(1)(a) of the Polish-Luxembourg tax treaty until the end of 2013. Double non-taxation of dividends paid by the Lux. Inv. Fund to its Polish shareholders has not been possible since 1 January 2014 due to the amendment in the Polish-Luxembourg tax treaty; the exemption method with regard to dividends was replaced by the credit method.

The result of this CFC scheme is that neither the income from current business activity generated in Poland nor the one from its repatriation from Luxembourg to Poland were taxed until 31 December 2013. These consequences of the CFC tax avoidance schemes with the Luxembourg CFC were sanctioned by many advance tax rulings,^[141] which explains why the majority of the FDIs by Polish investors was realized in Luxembourg in the years 2000–2013.^[142] This scheme still allows for the indefinite deferral of tax on income derived via a Luxembourgish CFC but not on income sourced in Poland and derived by the CFC via a Polish tax transparent partnership (Figure 3).^[143]

3.4.2.3 “LPP Scheme”: Tax avoidance on income from royalties and director fees

The following scheme was used by Poland’s biggest clothing company, LPP (LPP).^[144] The scheme serves primarily to avoid taxation on royalties received from the licensing of trademarks and director fees.

The following explains the steps that were taken in order to establish the scheme and the consequences to the taxpayer of its exploitation.

1. In 2011, LPP, a joint-stock company with tax residence in Poland, the Cypriot Gothals, contributed in kind to the ownership of two trademarks with a value of PLN 556 million (EUR 133.5 million). As a result of this transaction, LPP received nearly 100 percent of the shares in Gothals and Gothals received ownership of the trademarks. This transaction did not trigger taxation in Poland.^[145]

2. The day after this transaction, Gothals transferred ownership of these trademarks to a Dubai company, Jaradi, in the form of an in-kind contribution. This transaction did not bring about taxation either.^[146] The result of these transactions was that LPP received a nearly 100 percent share of Jaradi, while Jaradi Limited received a nearly 100 percent share of Gothals. LPP, in other words, controlled both companies fully. It should be noted here that this transaction and the transaction indicated in point 1 above were repeated in 2013 with respect to LPP’s three trademarks with a value of PLN 510 million (EUR 122.5 million).

3. Next, Gothals entered into a licensing agreement with Jaradi, which authorized Gothals to sublicense the trademarks to third parties. Royalties paid by Gothals to Jaradi are not subject to withholding tax in Cyprus because they are exempted from taxation under Article 13(3) of the Cyprus-UAE tax treaty. Moreover, the royalties paid by Gothals to Jaradi constituted deductible expenses for Gothals, and were consequently deducted from the tax base of that company. Owing to the low effective tax rate in Cyprus on income obtained from the exploitation of trademarks (about 2 percent)^[147] together with the low taxable base due to the level of deductible expenses, income received by Gothals is not taxed in Cyprus.

4. Further, LPP entered into sublicensing agreement with Gothals, entitling LPP to use the trademarks in order to conduct its business, mainly the promotion and sale of clothing in Poland, the Czech Republic, Slovakia, Hungary, Lithuania, Latvia, Estonia, Ukraine, Russia, Romania, and Bulgaria. Consonant with the sublicensing agreement, the royalties paid by LPP to Gothals are tax exempt in both countries due to the provisions of the EU Directive on Interest & Royalties (2003/49).^[148] The amount of royalties paid by LPP to Gothals constitutes tax deductible costs for LPP.^[149] Hence, these royalties decrease the tax base of LPP in Poland in relation to its worldwide income that is taxable in that country.

5. Income derived by Jaradi from royalties received from Cyprus is not taxed in Dubai, since ZEA adopted the territorial tax system whereby foreign-sourced business income, capital gains, dividends, interest and royalties are not taxed in Dubai.^[150] In this context, it is important to note that royalties are taxed neither at source in Poland nor in Cyprus, because tax paid at source could not be deducted from the tax payable in Dubai (no income tax in Dubai = no deduction of tax paid abroad).

6. Because the CEO of LPP, Marek Piechocki, is the director of Jaradi Limited, the significant proportion of the income derived by Jaradi Limited is repatriated to Poland in the form director fees, none of which is taxed in Dubai in relation to individuals (either residents or non-residents) under UEA domestic tax law,^[151] nor in Poland under Article 19 of the Poland-UEA tax treaty.^[152]

7. Since LPP fully controls Jaradi and Gothals, it decides when and how much of the income accumulated by these companies will be distributed. LPP may decide that the income accumulated by Jaradi Limited will not be distributed in Poland for many years, which amounts in reality to an unlimited tax deferral. If dividends were paid from Jaradi to LPP, they would be taxed in Poland. This is the reason why the company entitled to receive dividends from Jaradi in the first place is Gothals, and not LPP. In this way, the profits accumulated by Jaradi are distributed to Polish participants without taxation: the dividends paid by Jaradi to Gothals are neither taxed in Dubai nor in Cyprus under their respective domestic laws.^[153] The outcome is that dividends paid by Gothals to LPP are exempt from taxation in both Poland and Cyprus in accordance with the provisions of the Parent-Subsidiary Directive (2011/96).^[154]

As far as I know, the “LPP Scheme” has been, at least implicitly, accepted by tax authorities^[155] and I was unable to find any publicly available material indicating otherwise. Interestingly, LPP spokesman stated in May 2015 that from 1 January 2015, LPP would be attributing income from Gothals under the Polish CFC rules.^[156] Whether LPP did pay tax on Gothals’ income because of the application of Polish CFC rules, I was unable to determine (Figure 4).

Figure 4:

“LPP Scheme”: tax avoidance on income from royalties and director fees.

4.1.1 Company law in the case law of the CJEU and its effect on tax avoidance under EU law

According to Art. 49 TFEU that governs freedom of establishment,

Within the framework of the provisions set out below, restrictions on the freedom of establishment of nationals of a Member State in the territory of another Member State shall be prohibited. Such prohibition shall also apply to restrictions on the setting-up of agencies, branches or subsidiaries by nationals of any Member State established in the territory of any Member State.

In light of Art. 49 TFEU, CJEU already stated in its judgment in the Centros case that

The right to form a company in accordance with the law of a Member State and to set up branches in other Member States is inherent in the exercise, in a single market, of the freedom of establishment guaranteed by the Treaty.^[167] [Emphasis added]

This means that the purpose of the freedom of establishment does not limit the choice of establishing a company in a Member State, since it is inherent in the enjoyment of this freedom. Taxpayers may do so for whatever reason, including tax avoidance, and it will be in line with the purpose of the freedom of establishment insofar as it does not contradict this purpose.^[168] This stems from the CJEU case law and can be found in addition to the Centros^[169] in its judgments in cases such as Segers,^[170]Inspire Act,^[171] and Überseering.^[172]

This case law of the CJEU regarding company law strongly influenced the reasoning of the CJEU in tax avoidance cases which implies that “normal” tax avoidance^[173] is permissible under EU law. For instance, in the landmark Cadbury Schweppes case, the CJEU stated, in paragraphs 36–38 and 50^[174] and in reference to a previous case, that:

However, the fact that a Community national, whether a natural or a legal person, sought to profit from tax advantages in force in a Member State other than his State of residence cannot in itself deprive him of the right to rely on the provisions of the Treaty (see, to that effect, Case C-364/01 Barbier [2003] ECR I-15013, paragraph 71).

As to freedom of establishment, the Court has already held that the fact that the company was established in a Member State for the purpose of benefiting from more favourable legislation does not in itself suffice to constitute abuse of that freedom (see, to that effect, Centros, paragraph 27, and Case C-167/01 Inspire Art [2003] ECR I-10155, paragraph 96).

As noted by the applicants in the main proceedings and the Belgian Government, and by the Cypriot Government at the hearing, it follows that the fact that in this case CS decided to establish CSTS in the IFSC^[175] for the avowed purpose of benefiting from the favourable tax regime which that establishment enjoys does not in itself constitute abuse. That fact does not therefore preclude reliance by CS on Articles 43 EC and 48 EC (see, to that effect, Centros, paragraph 18, and Inspire Art, paragraph 98).

It is also apparent from case-law that the mere fact that a resident company establishes a secondary establishment, such as a subsidiary, in another Member State cannot set up a general presumption of tax evasion and justify a measure which compromises the exercise of a fundamental freedom guaranteed by the Treaty (see, to that effect, ICI, paragraph 26; Case C-478/98 Commission v Belgium [2000] ECR I-7587, paragraph 45; X and Y, paragraph 62; and Case C-334/02 Commission v France [2004] ECR I-2229, paragraph 27).

All this suggests that seeking for the most favoured tax route by taxpayers is permissible under EU law even if it may be perceived as tax avoidance from the perspective of the domestic law of Member States. This explains the existence of tax avoidance schemes within the EU law, as indicated and discussed in Section 3.2.2 and Section 3.4.2 above. However, the question is: does not EU law impose any limits on tax avoidance?

4.1.2 Limited effect of company law case law of CJEU on tax avoidance cases: “normal” tax avoidance is permissible under EU law, while “qualified” (=wholly artificial) tax avoidance is not

Even if what has been said above about the CJEU case law in company law cases, implies that tax avoidance is tolerated under EU law, this tolerance has its limit. This limit, in my view, follows from the limited influence of CJEU case law/company law on tax cases and therefore the case law cannot be automatically transposed to tax case law.

By referring to its previous case law,^[176] in paragraph 55 of its judgment in the Cadbury Schweppes case, the CJEU concluded that the restriction of the freedom of establishment entailed by applying the UK CFC rules may be justified if they apply only “to wholly artificial arrangements which do not reflect economic reality, with a view to escaping the tax normally due on the profits generated by activities carried out on national territory”.^[177]

In the light of the above, it seems that the CJEU distinguishes between two types of tax avoidance: (i) in a wide sense, the taxpayer tries to escape the tax normally due on the profits generated by activities carried out on national territory (“normal” tax avoidance) and (ii) in a narrow sense, the same taxpayer resorts to wholly artificial arrangements (“qualified” tax avoidance).

The current discussion allows to say that the concept of wholly artificial arrangements is used by the CJEU for creating a ban on tax avoidance at the national level that is unacceptable at the EU level. That is to say, it seems that the CJEU is concerned about whether there is a real establishment in a Member State hosting a CFC from an abuse of EU law perspective. It is noteworthy that the CJEU in its quite recent judgment of 12 June 2014 in joined cases C-39/13, C-40/13 and C-41/13 SCA stated that:

it is settled case-law that that ground [the risk of tax avoidance] does not constitute, by itself, an autonomous justification for a tax restriction on freedom of establishment if it is not relied on in conjunction with a specific objective of combatting wholly artificial arrangements which do not reflect economic reality and the purpose of which is to escape the tax normally due.^[178]

AG J. Kokott in her opinion delivered on 12 September 2006 in the case Oy AA (C-231/05) stated in the same vein that:

prevention of this form of “tax avoidance” is not a separate ground of justification which can justify a restriction on freedom of establishment. The fact that undertakings seek to profit from the differences between national tax systems is a legitimate form of economic conduct, and is indeed inevitable in an internal market in which taxation of corporations is not harmonised. Accordingly, an undertaking cannot simply be prevented from moving its seat to another Member State which offers more favourable conditions of taxation.^[179]

This means that combating tax avoidance only with the use of wholly artificial arrangements constitutes an autonomous justification for a tax restriction on fundamental freedoms triggered by anti-tax avoidance rules by Member States. Whereas counteracting a tax avoidance is not such a justification. To put it differently, tax avoidance is permissible under EU law, while tax avoidance via wholly artificial arrangement is not.

Furthermore, in its judgment in the Cadbury Schweppes case, the CJEU stated that a CFC reflects economic reality if its incorporation corresponds to “an actual establishment intended to carry on genuine economic activities in the host member state” (paragraphs 65–66). However, it is argued in the doctrine and is evident from settled case law, that this statement should not be understood narrowly.^[180] In that respect, and in reference to the CJEU’s case law in company law cases such as Segers,^[181]Centros,^[182] and Inspire Act,^[183] Luc de Broe advocates an extremely narrow understanding of the phrase in the Cadbury Schweppes case quoted above. In his view, a company established in one member state as a purely passive investment company, a letter box company or the like, is to be protected under the freedom of establishment insofar as this company has a controlling shareholding in an active subsidiary established in another member state.^[184] Interestingly, this author openly admits that, according to his understanding of the said CJEU case law, residents of non-member states have an opportunity to engage in Directive and treaty shopping.^[185]

I do not agree with his opinion due to the following reasons.

CJEU indeed first confirmed in paragraph 38 of the Cadbury Schweppes case the doctrine stemming from the Segers, Centros and Inspire Act cases, according to which establishing a company in a member state for “the avowed purpose of benefiting from the favourable tax regime which that establishment enjoys does not in itself constitute abuse”. It follows from the fact that the purpose pursued by the freedom of establishment does not limit the choice of establishing a company in a member state, since it is inherent in the enjoyment of this freedom.^[186] Its purpose is to assist EU residents in establishing subsidiaries/branches/agencies in other EU member states. This is the definite effect of freedom of establishment on company law case.

By contrast, the said effect (allowing for establishing a company in a member state freely) is not definite in tax law cases. Something more is required to comply with the purpose of the freedom of establishment in such cases in order to avoid the abuse of that freedom. This illustrates paragraph 55 of the Cadbury Schweppes in which it was indicated that a restriction on the freedom of establishment may be “justified on the ground of prevention of abusive practices”, i.e. prevention of “the creation of wholly artificial arrangements which do not reflect economic reality, with a view to escaping the tax normally due on the profits generated by activities carried out on national territory”. That is to say, the purpose pursued by the freedom of establishment is not to assist taxpayers in avoiding taxation by means of the creation of wholly artificial arrangements in member states which do not reflect economic reality. This constitutes the definite effect of the objective pursued by freedom of establishment on tax law cases.

The effect of the objective pursued by freedom of establishment is thus different in company law cases, on the one hand, and on tax law cases, on the other. As Alexandre Saydé indicates, assessing the purpose of the EU law concerned, for example the purpose of the freedom of establishment, “permits [one] to distinguish cases such as Cadbury Schweppes and Centros, in which artificial cross-border movements are respectively proscribed or ratified.”^[187]

Furthermore, in paragraphs 65–66 of the Cadbury Schweppes case, the CJEU made it clear that “the incorporation of a CFC reflects economic reality” – that is to say it should not constitute a wholly artificial arrangement – if that incorporation corresponds “with an actual establishment intended to carry on genuine economic activities in the host Member State”.

In my opinion, it is useful to try and determine the degree at which economic substance/activity is sufficient for being fully protected by the freedom of establishment and with more exactness than under the CJEU in the Cadbury Schweppes or even Fred Olsen cases. By achieving that degree, I mean the situation in which a CFC scheme enters the legitimate scope of the freedom of establishment and therefore a taxpayer benefiting from that scheme by means of tax avoidance cannot be taxed under CFC rules since such taxation could not be effectively justified by the prevention of wholly artificial arrangements, i.e. would be disproportionate to achieve that aim. In order to determine that degree more precisely, it seems useful to refer to the CJEU case law outside the field of company taxation, in particular to VAT case-law.

Before turning to VAT case-law, I will provide some arguments to justifying my having recourse to it as its use in the analysis of the concept of “economic substance/economic activity” with respect direct taxation provisions (CFC rules) may be questioned.^[188] First, I am not relying on VAT cases solely in order to determine what genuine economic substance or activity mean under the freedom of establishment. Rather, I use such case-law in the context of and in addition to the case-law on direct taxation. Second, as pointed out by AG Jacobs in his opinion on Cassa di Risparmio di Firenze (note 26),

the same concepts in different areas of Community law [now: EU law] should, as a general rule, be given identical meaning, unless otherwise justified by the nature or specific features of the area in which that concept is being inserted and which may warrant an ad hoc reading.

Third, in direct tax cases, the CJEU frequently refers to its previous VAT model of reasoning while examining the compatibility of domestic direct tax law provision with fundamental freedoms as in the Cadbury Schweppes case (paragraph 64). Thirdly, as stated by D. Weber,

there is no reason not to apply the VAT case law when exercising the treaty freedoms. In such a situation, there could also be a transaction that ultimately has, on its own, a non-fiscal goal, but the route to achieve this might have been taken for purely fiscal and artificial reasons. The “essential aim” case law then offers scope to combat this abuse.^[189]

With these arguments in mind, I observe that VAT case law allows to distinguish two variants of holding or asset management entities when assessing whether or not they carry on genuine economic activities or have economic substance sufficient for being fully protected under the freedom of establishment.

Thus, the mere shareholding of a CFC in other entities, without an active engagement in other entities in order to gain profits from them, such as managing their resources, does not constitute a genuine business activity and is therefore not protected under the freedom of establishment.^[190] Interestingly, this finding largely converges with the finding stemming from the definition of economic activity under the Polish tax law^[191] and autonomous meaning of the term “business activity” under Article 7 OECD MTC.^[192]

In other words, a company (e.g. CFC) established in a member state must carry out a genuine economic activity itself in order to be protected under the freedom of establishment, as mere passive engagement in companies established in other member states that conduct genuine economic activities is not sufficient. The evaluation of activities should therefore be done at the level of the company established in a member state, not of its subsidiary or branch.^[193]A fortiori, a company’s activities are not considered to be genuine economic activities if they are outsourced to companies in other member states, because, from the perspective of economic activity, it generally makes no sense for a company “to outsource its key entrepreneurial management activities to another party without maintaining control over the risks [and over resources] connected with these activities”.^[194]

In the Segers, Centros and Inspire Act cases, it was evident that the companies were not merely involved in passive investments of other entities in order to avoid taxation, but carried out genuine economic activities as defined by the freedom of establishment, if not in their member states of establishments, then in other member states through their branches or agencies or subsidiaries.^[195] By contrast, in the Cadbury Schweppes case, the companies in question (two CFCs, CSTS and CSTI) were established in Ireland (Dublin) by their parent company resident in the UK (CS), only to take advantage of the Irish tax regime (the International Financial Services Center in Dublin) without any intent to pursue genuine economic activities in Ireland or elsewhere.^[196] Consequently, it was logical and justified, not only on account of various effects of the objective pursued by freedom of establishment mentioned above, but also of facts relating to the Cadbury Schweppes case, that the CJEU did not stop its reasoning with paragraph 38, which reflected the outcome of Segers, Centros and Inspire Act cases (=definite effect of the freedom of establishment on company law cases), but developed it in paragraphs 55–68, which are specifically relevant to tax avoidance cases. Such development did not occur in Segers, Centros and Inspire Act cases, since it was not needed due to various legal questions, effect of the purpose of freedom of establishment and facts than in the Cadbury Schweppes case. In particular, the former case law did not deal with tax avoidance structures, while the latter case did and therefore the outcome of the former case law (company law) cannot have a definite legal effect on cases similar to this in the latter case law (tax law).

The observation above leads us to the concept of abuse of EU law in tax cases which delineates the limit for tax avoidance under EU law. In that regard, the interesting question is whether the concept of abuse of EU law is sufficient to effectively prevent tax avoidance, which may distort the establishment and development of an internal market, by Member States or at the EU level under the EU. The former case involves the application of domestic anti-avoidance measures, while the latter case involves prospectively introduced EU secondary law, e.g. anti-avoidance EU Directive (unlikely to happen but not impossible, especially in the context of the Anti-Tax Avoidance Package released by the European Commissions on 28 January 2016).^[197]