Avoidance and abus de droit: the European Approach in Tax Law

Abstract

Defining tax avoidance has been always a nearly impossible quest for tax lawyers. In Continental Europe, however, it could be said that the notion of "avoidance" is strongly embedded to the concept of "abuse" of a right, being an "abuse", according to the Roman law tradition, the exercise of a right inconsistently with the general principles of correctness, good faith or even with the basic rules of ethics. Therefore the definition of avoidance is not purely legal, but it depends also on other disciplines which influence it. In EU law qualifying avoidance is even more complex as the law of the Union is nearly deprived of any influence by other systems of values, the only general principles to rely on are the fundamental freedoms and the nondiscrimination principle both enshrined in the Treaty. That is why when in recent cases the Court had to rule on the "abuse of law" it did its best to find principles or values to build the concept on. The results were different in direct tax and in VAT cases: the notion of abuse can be wider in some of them and narrower in some others as far as it depends on the background to the specific provisions discussed. In case of the VAT the system is more complex and the court has to strike a balance between the need for neutrality of the tax and the coherence. In direct tax cases only the Treaty is applicable, and together with it the freedoms bestowed upon the taxpayer who can make the most of them in any case, unless the purpose of the scheme implemented is only to save taxes.

1. Introduction, the roots of legal avoidance and abuse of law in the continental legal systems

The concepts of tax avoidance and tax evasion have been analysed in the past by tax academics and practitioners from all round the world. To a certain extent, those[1] who considered them to be “eternal ones” were right.

Avoidance of tax provisions is an outcome of the infinite struggle between the principles of legal certainty on one side and freedom of business activity on the other:[2] between the legal form of the commercial operations and the substance of the aims pursued by the taxpayers.

The purpose of this contribution is therefore less far-reaching. The basic assumptions are that (1) the problems avoidance raises are not limited to tax law and (2) it is necessary to analyse avoidance of law in order to understand better the features and the characteristics of it when dealing with tax law. Another assumption is that (3) the EU tax harmonisation via regulations, directives, soft law and ECJ case law progressively compels the legal systems of the different member States to converge on specific characteristics necessarily found in all the tax planning schemes to be qualified as avoidant, even if these conditions might change from tax to tax. However this final result will be difficult to achieve as the fight against tax avoidance is intrinsically national in its roots, depending as it does on factors that do not necessarily belong to the tax systems; coming back in this way to (1).

Last but not least, the linguistic differences[3] across Europe also play an important role in the distinction between evasion, avoidance and “abuse of law”.

Some of these barriers are easy to overcome, in so far as they depend only on the language and not on the meaning of the concepts used by the legislator or by the judges. To this extent, it is easy to consider the English “tax avoidance” as the French évasion fiscale, the German Steuerumgehung or the Italian Elusione fiscale. On the other hand, “tax evasion” can be compared to the French fraude fiscale, the German Steuerhinterziehung and the Italian Evasione fiscale, for instance.[4]

Some other differences, however, are of a theoretical nature, although strongly correlated to the different languages used: the concept of “abuse of law” is one of these.

In many cases, especially on the continent, the notion of “abuse” is used instead of that of “avoidance of a legal provision”. In this sense the abuse of law could be considered a synonym of tax avoidance in those countries where the former term is accepted.

Despite the differences, the common ground of all these issues relies on interpretation: dealing with tax avoidance or with “abuse of law” still depends nowadays both on the interpretation of statutory law in the light of the Treaty of Rome and on the Constitution of the State (if any).

Some authors noted that a line should be drawn between avoidance considered as the outcome of loopholes in statutes and avoidance that is the result of different interpretations of statutory law or common law.

While the difference is theoretically correct, it must be remembered that this distinction is in any case based on interpretation of the rule of law, and that understanding the rule is an a priori condition to ascertain any loopholes or gaps in the system.

This requirement was well known in the continental (it could be said Roman) experience since the Middle Ages, when the debates on the notion of “abuse” of law and “avoidance” of statutory law arose for the first time, even if they did not immediately involve tax issues.

The historical approach is a good starting point even for current analysis: all in all, the problems faced by the medieval commentators and lawyers in Italy and in other countries were not so different from the current ones.

They had to strike a balance between the statutory laws of the Italian municipalities and the ius commune (what was left of the ancient Roman law) considered to be the legal background that was always applicable if not specifically derogated. Nowadays we almost have to do the same thing while managing the relationship between national laws and EU law. The only fundamental distinction relies on the overall trend of the evolution: in the Middle Ages the ius commune was progressively being substituted by the particularisme of the different municipalities (at least in Italy); in the 21st century it is the national law which has to progressively surrender to communitarian law where necessary.

2. Interpretation of tax law and the origins of avoidance: a lesson from Roman and Medieval law (Non puto delinquere eum, qui in dubiis quaestionibus contra fiscum facile respionderit[5])

The fragment of the Corpus Iuris Civilis compiled and introduced by Emperor Justinian in AD 527[6] and quoted at the beginning of this article must not be emphasised beyond its limit. Tax law was never considered in Roman law as an autonomous discipline, as it is nowadays. The contributions of academics and lawyers on that are scattered amongst hundreds of references that make it impossible to identify a Roman approach to tax law in terms and conditions we are used to[7].

However, there’s enough to understand the Roman reasoning in respect to interpretation, avoidance and abuse of law. That approach constituted the base on which the medieval commentators in Italy (and first of anywhere in the world, in Bologna) built up medieval law in the form of comments (glossa) to the Corpus Iuris in the 12th century.

Basically, all the doubtful cases submitted to the lawyers were answered according to one of two rules: “in dubio contra fiscum” or “in dubio pro fisco”. According to the first one, all the cases where tax avoidance[8] could have taken place had to be resolved in favour of the taxpayer, while under the second one, the solution had to be the opposite. The application of one or the other was a mere consequence not only of the personal position of the interpreter, but also of the source of law which introduced the tax to be applied and the revenue assumed to be avoided. While the ancient custom (consuetudo[9]) allowed an extensive interpretation, statutory law could not.

In any case, the reasons that took (Herennius) Modestinus to answer in that way remained incomprehensible to all the medieval commentators. The first and perhaps most influential of them, Accursius, tried to give an answer that is still all but convincing: “forte est ratio quia (fiscus) dives est”.[10]

For all these reasons, and despite the opinion by Modestinus in the Corpus Iuris, the interpretation of the tax rules across the Middle Ages was far from clear, there being possible two different approaches to tax avoidance: one relying on the form over substance approach and the other on the opposite. The application of the first or of the second basically depended on the source of laws which introduced the tax in the specific circumstances of the case.

Indeed, if the tax were due according to an ancient tradition (custom), both types of interpretation were possible, but if a statute of a municipality introduced the tax, then the strict interpretation (form over substance) had to be preferred. This was not decided according to specific tax rules, but it was done under a general principle of the ius commune that the specific provisions of the municipalities had to be interpreted restrictively,[11] considered as they were a derogation of the law.

The consequence, as is mentioned above, was that the problem of avoidance was solved in different ways, making the substance over form (or vice versa) prevail, depending (also) on the source of the law to be applied.

This basic rule suffered some exceptions, anyway. Perhaps the most relevant one is given by the same Bartolus (from Saxoferrato) in one of his responses (legal advices).[12]

To a certain extent it was a Ramsay approach[13] seven centuries before the decisions by the House of Lords.

3. The long way to a European approach to tax avoidance: from the Roman experience to the statutory anti-avoidance provisions. The continental way and the “abuse of law” (of fraus legis)

As has been noted before, Roman law and medieval law (the ius commune) weren’t directly involved in the phenomenon we call nowadays “tax avoidance”. The legislators and the academics in the first universities across Europe were interested in the interpretation of the law and only in a very broad sense in tax law.

For all these reasons the notion of tax avoidance was never introduced, nor the distinction with tax evasion in the sense mentioned at § 1: while the ius commune progressively lost its efficacy and its binding force in most of the emerging national States across the continent (as had progressively happened in the Italian municipalities before), European legal reasoning lost even the possibility of having a common approach to the same issue.

Tax avoidance was named (and defined) in different ways, according to the national experiences, but at least in theory, the distinction made by A. Hensel and E. Blumenstein in their foundational work on tax law is valid nowadays in almost every country.

According to the first author, “… In this latter case [tax fraud] the taxpayer didn’t fulfil the obligation to pay the tax already due, while in the case of avoidance the taxpayer prevented the tax from being due, avoiding the legal fact, deed or contract to whom tax law linked the payment of a tax …”.[14] From Blumenstein’s point of view,[15] however, tax avoidance occurs when “… According to a carefully planned and intentional procedure the taxpayer enters into a contract or sets up a specific operation which is capable of reducing the amount of the tax otherwise due, or to prevent it; thus far the avoidance is different from the tax evasion. In this latter case the tax is due to all the effects but the determination of its amount by the Tax Administration is prevented by an unlawful behaviour of the taxpayer …”.

Only in the latter case, therefore, (tax evasion) do the authors qualify the behaviour of the taxpayer as unlawful, while in the former it is consistent with law although in conflict with the normality of business activity or the exercise of the same rights.

Arguably it is also here where the legal traditions of the common law countries and the civil law ones began to diverge sensibly from this specific point of view.

The need to levy taxes on the business effectively carried on was achieved by the first ones through the “business purpose test” as developed later on in the case law by English, Australian and American courts[16], and by the second ones through the notion of “abuse of law”.

Even if the effect of the application of the two anti-avoidance provisions should be the same (limiting the taxpayer in the exercise of his business economic freedoms when no other goal seems to be pursued other than an improper tax saving), in common law countries the courts seek to understand what the normal business practice should include, while the continental ones seem to stress the importance of what is not reasonably included in the specific right attributed to the taxpayer.

The outcomes of the two interpretive approaches do not necessarily coincide.

As was mentioned before, Roman law vastly influences the continental interpretations summarised by the German and Swiss authors.

Under the Roman civil law, at first stage, the current notion of “abuse of law” was nonsensical (as it is probably nowadays in common law). In the case of property rights, for example, no abusive enjoyment was conceivable.

The owner of any goods either enjoyed them under law or in an illegal way.

In the second case, civil law had remedies for any person who claimed to be injured by the improper exercise of the right; in the first one, no limit could be imposed on the owner of any goods. The rule of law was clearly expressed in “Nullus videtur dolo facere qui suo iure utitur”[17] or, even better in our perspective, Ulpianus added: “… is qui iure publico utitur non videtur iniuriae faciendae causa hoc facere: iuris enimi executio non habet iniuriam,”[18] which is not completely new to the UK legal tradition, where “… if it was a lawful act, however ill the motive might be, he had a right to do it …”.[19]

The situation was clear, for instance, in the case of slavery. The owner of a slave had any right over him (or better, it, according to the Roman law) including the right to kill him at will (ius vitae necisque): a right that was accorded to the pater familias over all his family in the early stage of Roman law.

After centuries, however, during the later Empire era, this absolute right was somehow clarified without being explicitly limited. In other words, the property right had to be exercised over the slave according to some guidelines, so that the pater familias still had an absolute right “… sed dominorum interest ne auxilium contra saevitiam vel famem vel intolerabilem iniuriam denegetur his qui iuste deprecantur”;[20] soon after the principle was extended by Gaius in the rule “… male enim nostro iure uti non debemus …” or also “… expedit enim rei publicae ne qui re sua male utatur …”.

The turning point of the Roman legal approach towards the abuse is evident.

For the first time the Emperor clearly ruled that the exercise of an absolute right must meet specific conditions. The most important aspect of the revirement relies on the fact that neither Antoninus Pius nor Gaius seem to reduce the content of the property right, which still remains absolute, but rather they appear to introduce external and objective limits to its exercise, because a public interest may require it. This aspect is particularly evident in the passage quoted by the Institutiones, where the author clearly refers to the “common interest” as a limit to the improper exercise of rights. The step from “improper exercise” to “abuse” of right is particularly short.

It could be interesting to question what might have been the reason for such a drastic change (which, although adopted, at the same time respected the traditional content of the right).

Italian authors who faced this issue argued that a reason might have been the advent of the Christian religion[21] or, in the specific circumstances, the personality of the Emperor, who was particularly keen on these issues. In a Christian perspective, slavery was more than inconceivable and any legal rules that allowed it were unacceptable.

So here began a sort of war of attrition between two sets of rules: legal on one side, moral or religious on the other, with the latter influencing the former, or at least their interpretation.

In this early stage of the evolution of the ius commune the abuse of law or any avoidance of the law (including, arguably, tax avoidance) were behaviours consistent with the first set of rules, but not with the latter. The conflict has been a serious issue for all lawyers throughout the ages, and therefore not only for tax practitioners and academics.

The solution the Romans tried to introduce intended to strike a balance between the fundamental provisions of their civil law and the need for a more evolutionary approach. That’s why in the Roman literature abuse is never a factor that limits the right in its exercise, but is rather an external condition that must be observed.

In other words, the Roman legal literature in the aforementioned case never rejected the argument that the owner of a thing has a complete right over it, but rather that it is in the interest of somebody else (the Rei publicae in the Roman literature – we could say the State nowadays) to exercise that right in a specific way and within specific boundaries, and the latter interest prevails.

The common law approach is probably different. Here the “abuse of right” is probably treated in the law of tort, and the “abuse” is associated in most cases with nuisance and the remedies provided by law for the latter.

It’s also worthwhile mentioning that nuisance[22] is relevant and allows compensation to the plaintiff despite the absence of a deliberate intention to offend in the counterpart (i.e., of the Roman animus nocendi) as ruled in Bradford, unless the intention also played a fundamental role in the causation of the event.

The ECJ in its most recent case law seems to adopt a similar approach to the notion of abuse of law in a communitarian context, as is noted below: in other words, the European notion of abuse seems to be founded on objective rather than subjective aspects[23].

4. Avoidance and abuse in the Middle Ages: “lex saeculi merito comprimeretur, justitia Dei conservaretur”[24]

The offspring given birth by the Roman law in its latest stage of evolution flourished across the continent during the Middle Ages.

The basic idea was that (1) the correct exercise of any right should not damage others without determining at the same time an advantage to the owner and that (2) any exercise of the right should be consistent with the good faith of the user, who should try to achieve the scope the right was intended to protect or the legislator intended when attributing the right in question.

On the contrary, should the first condition not be met, that would be a case of objective abuse of right, while in the second case, the abuse would be of a subjective nature, according to the medieval commentators.

It is evident that the notion of abuse in tax law is similar to the one depicted above in (2). As was mentioned in § 3, the work of the commentators was aimed at defining the nature of the limits that could have been imposed to the exercise of the right in question, and some Roman commentators discovered them in the moral principles of the Christian religion.

In this sense, the ius naturalis introduced a rule of equity that was capable of derogating the rules of the old Roman law, and even more so the rules introduced in later centuries via statutory law by kings, municipalities, etc. The equity criteria in the Middle Ages soon became the most efficient anti-abuse provision, but at the same time also showed all the limits that instruments like this continue to have nowadays, especially in terms of lack of certainty of the rule of law.

5. The evolution of avoidance in the modern era and the contribution of French legislation to the notion of “abuse”

The boundaries of this article prevent a more detailed analysis of the notion of abuse in law and particularly in tax law in subsequent years, but we must at least mention the rise of the issue again after the Code Napoleon was implemented in France. The legal framework obviously had changed: the sovereign State assumed the monopoly of the sources of law, leaving but a narrow place to the interpreter.

In this context, the notion of abuse was, as it could seem today, nonsensical. The Code implemented straightforwardly a sort of “binary” approach to the law: any taxpayer (or any citizen) could either behave respecting the law, and therefore no abuse was conceivable, or act outside of the law, and then the concept of abuse was insignificant to that respect.

That’s why in the French tradition, and the traditions of those countries which followed the French example, such as Italy, the abuse of law was substantially unknown to academics and practitioners.

Different solutions were not legally possible. As was noted above, the abuse can arise when two different rules, belonging to different formal systems, are in conflict: it could be the case of moral or religious rules against the rule of law (Roman law versus Christian principles) or principles of law against the specific regulations of the case (ius commune versus statutes of the municipalities); but where a public body, such as the State, assumes the monopoly of the sources of law and excludes any other subject, then no sort of abuse is possible, and to a certain extent, the principle decided in Bradford above is confirmed.

More to the point, the concept of abuse as intended in the medieval tradition was no longer acceptable by national legislators across the European countries, while at the same time and to a certain extent, a different notion of abuse began to be implemented and recognised by national courts only in the last century.

The new concept of abuse was grounded on the necessity for the Courts (especially in Italy) to control the use of civil rights, such as those with regards to property, in a way consistent with the social function the Constitution attributed to them.[25] To this extent an abuse could be considered to take place whenever the exercise of any right was not consistent with the social function recognised and protected by the Constitution[26].

It could be argued that the place once occupied by the moral values of the Christian tradition was now taken by a more secular approach, closer to the socialist philosophy[27].

The constitutional approach to tax avoidance, once more, at least in the Italian experience, remains upheld nowadays as far as the duty to pay taxes in Italy is imposed under two basic articles: art. 53 and art. 3[28].

But while the first one is more accurate, introducing the “ability to pay principle” to our legal system, the second one deals with the notion of “economic solidarity” among citizens, leaving significant leeway for the interpreter to bend the notion in different ways.

That means, for example, that in the Italian legal system (and in all those on the continent which have a similar inspiration) the social function of taxes has been used many times as the overarching principle to oppose tax avoidance[29].

Namely, every time a taxpayer avoids the payment of taxes without evading them, he infringes the general principle of solidarity that the Constitution recognises. It is not enough, of course, to qualify art. 3 of our Constitution as an anti-tax avoidance provision, or that it was (and still is) considered a self-executing principle (the most influential authors state that rather, it constitutes a guideline to be followed by the parliament while passing new tax acts[30]), but it is the foundation of the current anti-tax avoidance approach: an interpretive instrument.

The economic solidarity enshrined in the Constitution and, to a certain extent, the social function that must assist the exercise of many rights under our fundamental law, today take the place of what once was the morality of some religion or the general principles of the ius commune.

This historical journey has eventually come to an end.

It was not useless. The most important result of this brief research is that in this perspective, the notion of avoidance is not absolute but might vary through the centuries, and, most important of all, according to the legal framework of tax law (but also of constitutional and civil law), it has to be implemented within.

In other words, it could change from State to State.

The more the system is inspired to social solidarity or accepts a wider notion of welfare, the more it is likely to find an aggressive approach to tax avoidance as an infringement of the duty (and not only of the obligation) to pay taxes.

This is not only a descriptive outcome but is at the same time the starting point for the study of tax avoidance in Europe and, more to the point, in ECJ case law.

The communitarian law, based on freedoms and not on a Constitution[31], had to develop a different approach to the issue: in other words, it had to find elsewhere the reasons and the legal principles to oppose tax avoidance, if it wished to do that.

Eventually, it still has to be seen whether, and how, an EU concept of tax avoidance would be suitable for implementation by the different Member States[32], as the potential conflict could cover fundamental rights and duties of some Member States. The opposite is also true.

It happened in the past that the need to battle against tax avoidance was used by Member States to justify limitations to EU fundamental freedoms, with various outcomes before the ECJ.

Generally, the need for anti-tax avoidance provisions is not a conditio per quam; a derogation to the EU freedoms is allowed: the court always asked for something more or tested the presence of tax avoidance under its standards.

This could potentially constitute another field of conflict between the national experience in combating tax avoidance and the different necessities of the European Union. In other words, some anti-avoidance provisions introduced to enforce the economic solidarity among citizen taxpayers could not work in an EU framework: they could be considered as disproportionate or unreasonable in the light of the communitarian approach.

The Rule of reason, therefore, should succeed where morality or religious or philosophical obligation failed in the past.

6. Rules and boundaries: reasonableness and fundamental freedoms

The fundamental freedoms enshrined in the EU Treaty are not different from the rights analyzed above, starting with Roman law. They are not introduced mainly for tax law purposes but have a relevant impact in this context.

When a national taxpayer adopts an approach to the national tax rule that leans towards tax avoidance, he could try and justify his behaviour under the aegis of a fundamental freedom.

It could be said, therefore, that the abuses of the Treaty rights (freedom of movement of the citizens of the Union, the free movement of goods, the free movement of workers, the right of establishment, the freedom to provide services and the free movement of capital[33]) raise the same problem discussed above while analysing the notion of abuse in different fields.

The solution differs: here it is the rule of reason[34] that distinguishes between proper exercise of the right and abuse (and therefore improper avoidance of tax).[35]

According to this approach, the national anti-avoidance provisions must pass several tests in order to qualify as consistent with EU law.

The first condition is that the specific circumstances of the case might fall into the exercise of one of the fundamental freedoms mentioned above, as far as the conflict (between national anti-avoidance rules and EU law) arises only where the second is theoretically applicable[36].

In this respect, it was noted that the real purpose of the taxpayer (namely the subjective approach to avoidance) is now absolutely irrelevant.[37] The ECJ always judged the application of a fundamental freedom without taking into account the purpose followed by the taxpayer in the specific case, at least in this first stage of the inquiry.

While the subjective aspect is underestimated by the ECJ, the objective aspect plays a more relevant role. The economic activity for which the taxpayer asks protection must be effective and genuine, and it must not constitute a mere device by which to qualify for the protection of the Treaty.

In other words, the business activity must be actual and not artificial: the exercise of the freedom must be consistent with an economic interest of the taxpayer.

It is self-evident that while at first glance the “rule of reason” seems to constitute an ideal concept capable of distinguishing the use from the abuse of a right, and therefore to identify the avoidance, on the other hand, under a more practical approach, it is difficult for everybody, and for the ECJ in particular, to assess in some cases the existence of reasonableness in a business activity[38] carried on by a communitarian taxpayer in the exercise of a fundamental right.

The second feature of the “rule of reason” is so evident that even in recent papers and seminars some influential authors stressed the fact that the ECJ seems to be losing the route followed in more recent years, deciding tax avoidance cases (both in direct taxation and in VAT) without a clear general picture.

The situation may be different from the one represented in the past: the ECJ approach has not changed through the years, but rather the objects of judgement of many recent cases are taking the ECJ into the maelstrom of harmonising its tax approach to an issue that may not be clear also within a national perspective.

The supranational approach can’t but emphasise the divergences, and the decisions reached are, of course, harder to accept for the Member States, as the most recent case law clearly shows.

7. Use, abuse and Cadbury Schweppes: the fingerprints of the EU approach to tax avoidance

Cadbury Schweppes[39] is not the first case, from a chronological perspective, in which the ECJ dealt with abuse of law: it is, however, one amongst those wherein the need to apply a clear concept of abuse has been considered with a particular focus, and where the court had to use once more the rule of reason (and the principle of proportionality to properly apply the Treaty) to provide reliable answers to the involved parties.

It is not important here to describe the facts and the circumstances of the case, which are well known and of lesser relevance to the extent of this research. It is significant, however, to focus attention on the fundamental freedoms allegedly infringed (arts. 43 and 48 of the Treaty: freedom of establishment) and on the anti-avoidance rule challenged before the court: the British CFC regulations.

In this case, the Cadbury group set up an Irish-controlled company to take advantage of the lower tax rates of that country, when compared to the UK rates. The tax saving was then reached through a leasing and financing operation from the latter to the former: the interests were tax deductible in the UK and, of course, taxed at a lower rate on the Irish subsidiary.

In order to avoid this tax deferral, the UK, just like many other European countries, introduced CFC regulations, attributing to the resident parent company the profits (or part of them) realised by the non-resident subsidiary, without any necessity to wait for the distribution of dividends by the foreign subsidiary.

At first glance, a provision like this could impair the freedom of establishment[40] because a resident company could be prevented from establishing a subsidiary abroad if this decision could then immediately and directly increase its taxable base.

The UK provision, considered alone, is reasonable because it is aimed at preventing the foreign allocation of profit where this decision is inspired mainly by tax considerations: the abuse in this case could consist of a business activity developed abroad (and not in the homeland) only for tax purposes.

The answers that Advocate General Léger and the ECJ gave on this issue are quite surprising, but only to a certain extent.

The first statement is that tax planning across Europe is a legitimate way to spare taxes and that the delocalisation of subsidiaries or branches of a company only for tax reasons cannot be challenged by the home State Tax Administration. It could be said, then, that a certain level of tax competition amongst Member States is not only tolerated but reflects the need of a common market. In other words, tax motives are, in themselves, legal motives.[41]

The Advocate General clearly pointed out in his final remarks that when a parent company chooses another Member State in which to establish a company only because in the latter the taxation is significantly lower, it isn’t abusing the freedom of establishment enshrined in the Treaty. This conclusion was, all in all, already upheld in the Centros[42] case, even if the latter was not exactly a tax case.

So far, no iron curtain runs between use and abuse of law, but rather a thin red line that can shape a different border when the abuse is tested under commercial law versus tax law.

For instance, in Centros the freedom of establishment was defended by the ECJ even if in the specific case, the seat of the company set up in the UK was merely artificial, as the main business was developed in Denmark. The same goes for a case in which the company was set up according to a Member State commercial law and then was transferred to the UK or elsewhere. In these circumstances, the company tried to circumvent only commercial law provisions and no abuse was assessed even if the main seat abroad was fictitious or not effectively involved in any business at all.

In tax cases, such as Cadbury, the distinction is not so straightforward, and the artificial nature of the subsidiary could constitute a reasonable limit to the fundamental freedoms of the Treaty.

Of course Centros and Cadbury have a significant common ground. In both cases the court ruled that the nationals of any Member States cannot use EU law to “improperly circumvent”[43] the national legislation or, in another point of the decision, that they are not allowed to “improperly or fraudulently”[44] take advantage of the community law.[45]

When such conditions are not met, then no restriction to the freedom of establishment is allowed in a EU perspective, unless, and this is specifically a case relevant to tax law, other overriding reasons of public interest[46] do not emerge. And even in this case the derogations and the limits to the fundamental freedoms must be appropriate, proportional and consistent with the aim pursued.

It is self-evident that the need of every Member State for specific countermeasures to tax avoidance falls into one of the overriding reasons of public interest, but those provisions must be proportionate to the aims pursued and to the actual dangers. If this condition is not met, then it is the Member State that abuses this possibility.

That’s why the Court plays a fundamental role in defining the notion and the condition of abuse: defining tax avoidance, it also contributes to defining the abuse of right by the Member State.

In EU law this is particularly true: the abuse of the taxpayer is counterweighted by the abuse of the Member State; both subjects are, to a certain extent, in an equal position before the court[47].

Cadbury is paradigmatic to this extent. From the UK point of view, the taxpayer is abusing his right to set up an economic activity abroad, as far as his decision is basically inspired by tax deferral and tax reduction, and from the taxpayer point of view, the UK is abusing its right to override the Treaty freedoms. The concept of tax avoidance (and the closely related “abuse of law”) therefore is at a crossroads the court had to regulate, potentially catching two pigeons with one seed.

8. Resolving a fearful symmetry? The ECJ attempts to tackle tax avoidance

In the previous paragraph it is noted that the EU perspective is a privileged one from which to look at the phenomenon called “tax avoidance” or “abuse of tax law”.

The reasons are quite evident: while sovereign States deal with tax avoidance with a one-way approach, that is, seeking to limit rights and freedoms of the taxpayer in a convenient way, the EU approach compels Member States to do this with respect to some of them, unless other very specific circumstances are not met.

So far, the EU approach to tax avoidance is “two way”: the ECJ in cases like Cadbury judges the limits imposed by the Member States according to limits imposed on those States by the Treaty.

Cadbury is once more helpful to this extent, because the Court set out, albeit not for the first time, the circumstances in which a business operation falls outside the scope of the EU law protection. This situation occurs, just like in the specific case, when together with the “subjective element of obtaining a tax advantage” objective circumstances exist and show that “despite formal observance of the conditions laid down by community law, the objective pursued by freedom of establishment has not been achieved”[48] .

It is a helpful benchmark for future rulings even if the conditions are deeply influenced by the circumstances of the case, by the rules challenged (CFC legislation) and by the specific ways and means to set up an improper tax-saving scheme.

In the text of the sentence it is possible to underline three specific aspects which played an important role in the final decision: three conditions, two positive and one negative, that help to define better the notion of abuse, in both the mentioned ways.

The first one, the negative, is for the Member States.

From the Court’s point of view, “anti-tax avoidance” provisions are matters for the different Member States: a need for harmonising them arises neither from the Treaty nor from any specific directive, excluding some incidental steps in directives whose primary target was different.[49]

As a matter of principle, every Member State is free to implement any of them as far as it deems necessary to safeguard the general principles of the system and the general duty to pay taxes.

However, once those provisions limit or, broadly speaking, interfere with the exercise of any fundamental freedom, their application can be allowed only if targeted at contrasting “wholly artificial arrangements”[50] concluded by the taxpayer aimed at “circumventing the application of the legislation of the Member State concerned”.[51]

This is not the first case[52] in which the court has ruled on this, but it could be considered one of the most clear, taking into account the impact on typical anti-avoidance provisions (CFC regulations) whose use is quite widespread across Europe.

The second ones, the positives, are for the taxpayer.

Freedom of establishment means that (1) the Treaty protects not every kind of business, but only those that are aimed at actually pursuing an economic goal through a fixed establishment for an indefinite period. Moreover, (2) the business must underpin a genuine economic activity: this means that it must reflect, as the Court says, a “real economic activity”.

Even if closely related, the two sets of conditions do not necessarily coincide all the time.

First of all, time matters for the positive conditions and not for the negative one. “Establishment” in the meaning of the Treaty means a place of business qualified by stability in time and an intrinsic attitude to be developed (according to the specific case) for an indeterminate period (depending on the nature of the business developed). The Treaty of Rome seems to guarantee protection where a business is carried on in another EU country only through a set of investments that will develop during the time.

The first conditions don’t seem to require such a qualified duration. So, for instance, an anti-avoidance provision aimed at contrasting a business operation in another Member State that is not going to create in that other State a fixed establishment under the meaning of the word in Cadbury could not find an adequate protection under art. 43.

This will not mean that the Treaty doesn’t cover the mentioned operation, as it could fall into the protection of another article or another principle of communitarian law (for instance, the free movement of capital[53]).

Timing is not the only difference between the two sets of conditions, then.

It could be argued that a second and a third difference rely on the subjective and objective elements, because later on in Cadbury the court realises that even if, theoretically speaking, the distinction between genuine activity and artificial arrangement could suffice to introduce an acceptable difference, from a more practical point of view, the aim is still far from being reached.

That’s why the Court stresses the importance of the subjective element[54] in the abuse of law: underlying this requirement is the intention to obtain a tax advantage. Only if this condition is met, and the motive is the exclusive reason which pushed the taxpayer to set up the entire operation, is the anti-avoidance provision consistent with EU law.

But this is not enough: the Tax Administration of the home State has to demonstrate[55] that the taxpayer didn’t achieve, in the specific circumstances of the case, the objective pursued by the freedom of establishment.[56]

In other words, that there has been a detournément, that is, a deviation, of the ratio of the rule in question. The taxpayer invoked the freedom of establishment, but the operation he set out is not intended to establish anything in the other Member State.[57]

In this way the Court sought to strike a balance between the two abuses underlined before: on one hand, the abuse of the taxpayer exercising his freedoms and rights in a improper way (inconsistent with the duty to pay taxes and the solidarity principle that are introduced in almost every European Constitution, if any, or fundamental law of the State) and on the other hand of the State introducing an anti-tax avoidance that could constitute an overkill, in some cases, of the business operation.

A clear example of this were the British CFC regulations: unable to distinguish true tax avoidance from the exercise of the freedom of establishment.

The Court tried to reach the best solution available, considering that the playing field was all but level, in the circumstances of the case.

In other words, tax avoidance is perceived differently (and answered differently) in light of the general principle of any tax system[58]. If, like in many Member States, a duty of economic solidarity is present, then a certain answer is possible; if, as in EU law, this principle is not present (the Treaty is not a Constitution and arguably it will never become one), the Court can’t but use overarching rules of reason in a desperate attempt to find an acceptable solution for all the parties involved.

This answer is, however, valid not for all EU law, but only for those aspects that are not harmonised and where the tax avoidance and the abuse of law can be tested only in the light of the fundamental freedoms enshrined in the Treaty.

Where the EU legislation is more detailed and contains principles and rules applicable to all the Member States it is possible for the Court to implement a notion of abuse and of tax avoidance that is more compatible with the system it has to work within, without any need to refer to anything but the rule of reason.

In the case of the VAT, the Court has a complete and detailed legislation built upon several directives, later consolidated into one text[59], and a principle, the neutrality of the tax, that must necessarily inspire the different legislations of the Member States without entering into any conflict with other issues at a constitutional level (such as solidarity or the ability to pay taxes…).

The notion of abuse (and of avoidance) shall, therefore, be flavoured differently.

9. Abusing neutrality? Playing with the VAT deduction in Halifax

The most striking difference between Cadbury and Halifax[60] is clearly summarised by the court in §§ 57 and 62 of the latter judgement.

According to the first paragraph of the sentence, the notion of abuse (if any) in the VAT mechanism is not qualified by any subjective element.[61] In other words, the VAT application must be necessarily inspired by two general principles, facility in the application and certainty of the legal relationship between businesses.

From this point of view, any possibility of tax reassessment which is grounded on subjective factors, such as the deliberate will to avoid the payment of taxes by setting up a complex commercial operation, could constitute, per se, an unacceptable impairment of that goal, introducing an element of uncertainty and unpredictability to the reassessment which could interfere with the application of the VAT mechanism.

At first glance, it seems therefore that the taxpayer could enjoy a greater leeway in the abuse of the VAT provision. The impression is arguably wrong, and all in all it would be surprising that in a harmonised field of tax law the possibilities for tax avoidance could be greater than those in non-harmonised sectors[62].

The court clarifies this point later on in the judgement, stating that clearly the sixth directive (and not the 2006/112/EC directive) is aimed also at preventing tax evasion, avoidance and abuse of law,[63] but that this objective must be achieved with respect to the legal certainty of the business relation between entrepreneurs localised in different Member States of the Union.

In other words, priority must be given to the need for trust in business relations or, more to the point, to tax consequences of the business relations. Every entrepreneur across Europe must know in advance the amount of the VAT (if any) applicable to the sale of goods or purchase of services in a cross-border dimension.[64]

It could be argued at this point that this second necessity is peculiar to the VAT system and no issues of this kind were raised by the Court (or at least were raised without such a priority) in cases involving direct taxation and fundamental freedoms. The first and foremost effect, as mentioned above, is that the anti-avoidance provisions have to pass a tougher control of their compatibility with EU law.

It is not a matter of abuse here, considered as a balance between fundamental rights and the need of every Member State to raise money for welfare, but also a matter of predictability of the tax consequences[65]. The dimensions of the overall judgement are therefore different, and we could argue that the concept of avoidance and abuse is no longer the same.

Looking to this issue from the point of view of the taxpayer, then, the need for neutrality of the tax and the priority of legal certainty become the two fundamental benchmarks with which to assess the compatibility of the anti-avoidance provision to EU law.

That’s why it is not entirely accurate to state that the step-transaction approach followed by the ECJ in all the cases involving avoidance of VAT allows more loopholes to be abusively exploited by the fraudulent taxpayer. Rather, it is the mechanism of the tax that imposes this approach.

The Court implicitly clarifies this point later on in the Halifax judgement, when it notes that even if the step-transaction approach has to be favoured in VAT cases, and even if it’s up to the national Courts to assess the eventual abuse of the neutrality (in most of the cases, of the VAT deduction input by the taxpayer), taking into account the specific characteristics of the operation.

From the ECJ approach, the “purely artificial nature”[66] of the operation from an objective point of view and the “legal, economic or personal relations”[67] between the parties (that is, the seller and the purchaser of the goods, for instance) are significant. If these conditions are met,[68] the operation can be disregarded for VAT purposes and the deduction of the tax denied by the Tax Administration.

The Court is walking once more on a minefield[69] and seems quite aware of that. It is contradictory to stress the importance of legal certainty and soon after to introduce elements of judgement recalling a sort of “substance over form” approach that is quite unfit to give ex ante a clear solution, depending as it does on the specific circumstances of the case.

That is probably why soon after, the Court clarifies that, anyway, the abuse must be assessed under “a number of objective factors” and that only after that positive assessment of the operation it is possible to say that the essential (not only the main) reason for the existence of the specific operation is to obtain a tax advantage.

All in all, the approach seems to recall the one analysed before, where the Court stressed the importance of the effective exercise of a fundamental right or freedom enshrined in the Treaty (freedom of establishment, of movement of capital, etc., …), but here the soundness of the business is the condition to be met and the neutrality of the tax the feature to be preserved as far as possible.

10. Concluding remarks: where do we go from here?

The first part of this brief research began with a quotation from an author[70] indicating that the issue relative to tax evasion and, first of all, avoidance, is an eternal one. After ten years the same author implicitly confirmed his opinion some months ago at another academic meeting[71]. Moreover, other academics recently presented a different and original approach[72] to the notion of tax avoidance, which seems to focus not only on legal issues but also on different factors derived from the business experience, such as risk management and the need for an overall different approach to the taxpayer by the HMRC, at least in the UK. In a relationship based on trust, there is no space for avoidance[73].

Both the approaches are absolutely correct, and where the first underlines a problem, the second tries to offer a new solution with different means.

However this is intended to be an investigation made from a legal point of view and therefore in a quite traditional way. The purpose has not been to provide another approach to tax avoidance, tax evasion and abuse of law. Such a purpose would be too far-reaching and impossible to achieve. The goal has been, to a certain extent, to try to understand the notion of avoidance and abuse of law from a different perspective.

The basic assumption of this research was that the notions of tax avoidance and abuse of tax law must be understood and interpreted according to what happens in the other fields of the national legal system. The historical evolution of each of them, and of the Roman law, is determinant in clarifying when abuse takes place.

This perspective has supported the idea that avoidance is not an “eternal problem” in tax law, but an issue that arises when legal provision collides with other sets of rules of a non-legal nature: of course this conflict depends on the national background and on the strength (real or perceived) of the non-legal rules[74].

When the ECJ tries to harmonise the different tax systems as far as this is allowed by the Treaty, it progressively has to solve the same problems faced by the different national States across the centuries. However, while in every Member State the abuse is always a judgement of comparison between an individual right (not necessarily of the taxpayer) and other principles enshrined in the legal system (ability to pay, solidarity, etc., …), in Europe these principles are missing, in so far as something more of a Treaty between States will not be implemented.

This is helpful towards explaining why the Court, in some cases, seems to wander in obscurity[75], recalling concepts such as “genuine business” or even “non-artificial business” that although sufficient to solve the problem from a theoretical point of view are of little use from a practical point of view. The practitioners need something more, as the same ECJ implicitly admits in all cases involving VAT.

The answers now provided, even if workable in direct taxation, are not sufficient. Assessing avoidance only in cases of letterbox companies (a case that in the Italian point of view could be qualified, to a certain extent, as tax evasion and not as avoidance) is of little help for all those States that are striving to defend constitutional principles of economic solidarity, while it could be considered enough by others which made a different choice, providing limited welfare and at the same time requesting a lower level of taxes.

All in all, tax avoidance and abuse of tax law are not “eternal problems” in tax law, but issues that are only partially related to tax law and depend first and foremost on the culture and the historical evolution of every State, issues about which the European tax lawyer has little to say, especially without a constitutional background.

[∗] Ph. D., Senior lecturer, University of Ferrara, Faculty of Law, Italy and Research Fellow, Taxation Law and Policy Research Institute, Monash University, Australia. Email: marco.greggi@unife.it.

[1] F. Vanistendael, Judicial interpretation and the role of anti-abuse provisions in tax law, in G. S. Cooper (edited by), Tax Avoidance and the Rule of Law, Amsterdam, 1997, p. 153; P. Pistone, L’abuso del diritto nel diritto tributario internazionale, in (edited by V. Uckmar) Diritto tributario internazionale, Padova, 2005, p. 813 ss.; G. Maisto, Multilingual texts and interpretation of tax Treaties and EC Tax law, Amsterdam, 2007, passim.

[2] In Europe this feature is evident particularly in VAT cases. The ECJ, for instance, has underlined this conflict in Halifax plc, Leeds Permanent Development Services ltd, Countrywide Property Investments ltd v. Commissioners of Customs & Excise, C-255/02 at §§ 71 and 72. For a more detailed analysis, see below § 9.

[3] M. Mossner, Association of European tax professors: a common language in taxation within Europe, in EC Tax Review, 1999, p. 158. The author noted that “A first step in developing coordinated tax systems within the Europe Union is therefore to develop a common language within Europe”.

[4] Some misunderstanding on the continent is, however, always possible: note how the same word “evasion / evasione” has two completely different meanings in France and Italy. C. Sacchetto, Multilingual issues under internal laws with a particular emphasis on tax law: the Italian experience, in Dir. Prat. trib. int., 2007, p. 881. Another clear example of the difficulties encountered while translating legal concepts is given by P. Pistone, Centro di attività stabile e stabile organizzazione: l’IVA richiede una evoluzione per il XXI secolo ?, in Riv. Dir. Trib., 1999, III, p. 31.

[5] Corpus Iuris Civilis 49.14.10.

[6] The Corpus was ended in 533, but it is well known that further amendments were introduced almost until the death of the emperor.

[7] Some Authors (S. Douma and F. Engelen, Halifax plc and Others v. Commissioners of Customs and Excises: the ECJ applies the abuse of rights Doctrine in Vat cases, [2006] BTR 430; see also E. Van der Stok, General anti avoidance provisions: a Dutch treat [1998] BTR 151) stress the importance of the Roman background, quoting from the Corpus iuris that “In fraudem legis agere” is the behaviour of those “qui salvis verbis legis sententiam ejus circumvenit” [who without infringing the words of the law, deceive the purport thereof] (Dig. 1.3.29). Fraus legis is the concept still used in Dutch Tax law to fight back tax avoidance.

The reference is still nowadays of extreme interest to tax lawyers, together with another one in the same Corpus, but in this case by Ulpianus (1.3.30) “Fraus legis enim fit, ubi quod fieri noluit, fieri autem non vetuit” [Fraus legis occurs when a certain behaviour is not wanted without being vetoed – by the legislator –]. However. While the definition of fraus (circumvention) or abuse is clearly settled, the reader must be aware that Roman lawyers developed their conclusions in Civil law, and not in Tax law. The abuse mentioned in the Corpus Iuris dealt with inheritance law (Lex Cinicia de donationibus, 204 b.C.), and it was the task of the subsequent Academics to extend this reasoning to tax law, using interpretive arguments (such as the analogy or the extensive interpretation).

[8] Of course the two general rules mentioned above were applicable not only to tax avoidance but in all cases where the wording of the tax provision was not clear enough to produce a straightforward application. It is obvious, however, that the tax avoidance cases were the most relevant ones, as well as those where the rules mentioned played the most relevant role. Of course tax evasion was not involved in the distinction.

[9] “Consuetudo servatur in vectigalibus” [Use must rule in the application of taxes], Ad. Dig. N. I, dig. XXXIX, 4, 4.

[10] “Maybe because the Tax Administration is richer than the taxpayer” in Digestum novum Lugduni, 1550, col. 1478.

[11] See the comments on this topic to the L. Caesar § expeditis ff. n de publicanis (ad dig. N. II, Dig. L. 4, 18).

[12] Consilium 135 ad L. I. It happened that a municipality introduced a tax to be paid by all those who entered the walls to sell skins and leathers in the market of the town and that the tax was due because of the occupation of the market soil by the goods to be sold. Then the administration of the municipality discovered that many merchants had begun to sell their skins while keeping them on their shoulders and arms, without occupying any portion of the market soil. They also refused to pay the tax, claiming that statutory law was to be interpreted restrictively and that the payment was due only for the occupation of the soil of the market. Obviously that was not the case.

It could be argued that this was one of the first cases of avoidance in the history of tax law. Quite surprisingly Bartolus answered that in cases like these it was necessary to distinguish leathers and skins used by the merchants for their personal purposes from the others. The tax was due for the latter; despite their being sold without occupying any portion of the market. Arguably, Bartolus in this case applied the principle of substance over form to overrule the strict interpretation of the statute introducing the tax (named Gabella).

[13] W. T. Ramsay Ltd. v. Internal Revenue Commissioner [1981] UKHL 1; [1981] 1 All E.R. 865.

[14] A. Hensel, Steuerrecht, Berlin, 1924. Translation by the Author from the Italian version by D. Jarach, Diritto tributario, Milan, 1956.

[15] E. Blumenstein, System des Steuerrechts, Zurich, 1951. Translation by the author from the Italian version by F. Forte, Sistema di diritto delle imposte, Milan, 1954.

[16] See for instance Y. Keinan, The many faces of the economic substance’s two-prong test: time for reconciliation?, N.Y.U. J. L. & Bus., 2005, 371 ss.; J. Rector, A review of the economic substance doctrine, 10 Stan. J.L. Bus. & Fin., 2004, p. 173.

[17] Dig. 50, 17, 55. “It is never in bad faith for one to exercise his right”.

[18] Dig. 47, 10, 13, 1. “The one who exercises his own rights never damages others: indeed the exercise of a right is never an offence”. To a certain extent, the step-transaction doctrine could be seen as grounded on this basic assumption: where every step of a more complex transaction is taken according to the law of contracts, no abuse is conceivable and therefore no avoidance occurs.

[19] Lord Halsbury in Bradford v. Pickles [1895 A.C. 587].

[20] Inst. 1, 8, 2; Dig. 1, 6, 2 “It is in the same interest of the owner not to refuse help in case a slave is maintained in a condition of starvation, famine or subject to disproportionate work”. It is worthwhile remembering that this excerpt is taken from a ruling of the Emperor Antoninus Pius.

[21] G. Grosso, Abuso del diritto, Enc. dir., Milan, 1958, I, p. 161.

[22] A. Gambaro, Abuso del diritto II, Diritto comparato e straniero, Enc. giur. Treccani, Roma, 1988, I, p. 6 and 7; the author also quotes from Grandel v Mason, in 3 D.L.R., 1953, 65 and Hollywood Silver Fox Farm v Emmett, in [1936] 2 K.B. See also U. Magnus, Tort law in general, in (edited by J. M. Smits) Elgar encyclopedia of comparative law, Cheltenham, 2006, 725.

[23] See also Poiares Maduro’s opinion in Halifax at §§ 70 and 87

[24] “[Sometimes] it is necessary to modify the application of the law to let God’s justice prevail”.

[25] In the Italian Constitution, the right of property is recognised and protected only in so far as it is exercised also in a social interest. Even if the wording of the article didn’t have a significant impact on the subsequent legislation that is still in force nowadays, it was the heritage of the Communist Party in the Constituent Assembly after the Second World War. So far, an abuse of right has been considered every time the right has been exercised without taking into due account the social function attached to the right.

[26] The constitutional background of the “abus de droit” theory is clearly explained by P. Harris, ‘Abus de droit’ in the field of value added taxation [2003] BTR, p. 136.

[27] C. Salvi, Abuso del diritto, in Enc. Giur. Treccani, Roma, I, 1988, p.3. See also C. Sacchetto, Ethics and Taxation, Dir. e prat. trib. int., 2007, I, p.9 ss.

[28] A. Di Pietro, The principle of equality in European taxation – Italy, G. Meussen (ed. by) The principle of Equality in European taxation, The Hague, 1999, p. 118; W. B. Barker, The three faces of equality: constitutional requirements in taxation, Dir. Prat. trib. int., 2007, p.405.

[29] The situation is therefore significantly different from the American one, see for instance Y. Edrey, Constitutional review and tax law: an analytical framework, in American University law review, 2007, 118; B. Ackerman, Taxation and the constitution, Colum. L. Rev., 1999, p. 39.

[30] See A. Di Pietro, cit., p. 119.

[31] It is worth while mentioning that the draft scheme of an European constitution, refused by French and Dutch citizens after national referenda contains no specific provisions in this respect. Nonetheless the existence of a European constitution, enshrining social and political rights could have grounded better any extensive approach to the “Abuse of law” theory.

[32] This aspect is specifically clear where EU law directly legitimates an anti abuse approach, such as for instance in the Mergers Directive (1990/434/EEC), or in the Parent – Subsidiary one (1990/435/EEC): see B. J. M. Terra and P. J. Wattel, European Tax Law, The Hague, 2005, p. 563.

[33] Respectively articles 23, 39, 43, 46 and 49 of the Treaty.

[34] See Cassis de Dijon, C-120/79.

[35] D. Weber, Tax avoidance and the EC Treaty freedoms, The Hague, 2005, p. 253.

[36] In Leur Bloem (C-28/95) however communitarian tax law was considered applicable even to a purely National case.

[37] D. Weber, cit., p. 254.

[38] Or also in a specific commercial operation, such as in Part Service (C- 425/06) still pending in front of the ECJ. In this case the taxpayer (unreasonably) unbundled a single financial leasing contract into a loan and a leasing. The overall commercial operation didn’t change significantly, but the tax saving was considerable.

[39] Cadbury Schweppes plc, Cadbury Schweppes Overseas ltd v. Commissioners of Inland Revenue, C-196/04.

[40] In the specific case art. 53 was invoked by the taxpayer as well. The Court refused to go into details as far as art. 43 was considered sufficient to assess the infringement of the Treaty. According to a settled case law, when an infringement of art. 43 is recorded, it is not necessary to examine the same issue under another point of view.

This conclusion was possible in the specific circumstances of the case because the host State of the subsidiary (Ireland) is an EU member country. At the moment, therefore, it is not possible to confirm beyond any reasonable doubt that the decision in Cadbury would have been confirmed where that country would have been a non EU one.

It could be argued, in this latter case, that art. 56 (applicable also to non EU taxpayers) would have taken the place of art. 43 in the reasoning of the Court.

[41] N. Vinther and E. Werlauff, Tax motives are legal motives – The borderline between the use and abuse of the freedom of establishment with reference to the Cadbury Schweppes case, European taxation, 2006, p. 384.

[42] Centros ltd. v. Erhvervs-og Selskabsstyrelsen, C-212/97. In this case the UK was chosen because of the lesser conditions to be met in order to set up a limited company, in comparison to those in Denmark.

[43] See Cadbury, cit., at § 35.

[44] Once more Cadbury, cit., at § 35. It is important to note how the court uses words and attributions such as “improper”, “fraudulent”, “genuine” (at § 54) that recall the Roman experience, where the abuse of law was discovered by using moral or ethical concepts.

[45] Before Centros see also Knoors, C-115/78 and Bouchoucha, C-61/89.

[46] Cadbury, § 47.

[47] This situation resembles to a certain extent the relation between “Abus de droit” and “Abus de pouvoir” in the French doctrine, as Harris mentioned in his article quoted above.

[48] See Cadbury at § 64.

[49] There are references to tax avoidance and abuse of law in virtually any directive involving direct taxation in the EU. The same could be said also for VAT, even if the situation in that case is much more complex and will be analysed in a subsequent paragraph of this research. In the case of the Sixth Directive, for example, the notion of avoidance and abuse were mentioned in art.13A, 13B, 14, 15, 22.8, 27.1 (see again P. Harris, cit., p. 147)

[50] See Cadbury, § 51.

[51] See Cadbury, § 51.

[52] See also ICI, § 26; Lankhorst – Hohorst, C-324/00, § 37 where the court challenged the “thin capitalisation rules” in the light of the same principle (freedom of establishment); De Lasteyrie du Saillant, C-9/02, § 50; Marks and Spencer, C-446/03, § 57.

[53] And in this case even non EU companies could qualify for the protection of the Treaty.

[54] It is not the first time we meet such a condition in the definition of “abuse of (tax) law”; see § 2 above. Even if the subjective element is not decisive per se, it still shows a certain importance in the reasoning of the Court.

[55] The burden of proof is neither clearly established in this case nor in the others preceding it (see, for instance, Emsland – Stärke, C-110/99 at §§ 52 and 53; Halifax and others, C-255/02, at §§ 74 and 75). However from § 64 of the decision in Cadbury emerges the need for the Member State to demonstrate something more of the subjective element; that’s why it could be argued that it is the burden of the Tax Administration to demonstrate that the business operation is not sound at all.

[56] See Cadbury, § 64.

[57] See Cadbury, § 64.

[58] P. Harris, cit., p. 136.

[59] Directive 2006/112/EC.

[60] It is not the intent to summarise here the facts of the judgement. Anyway, Halifax was submitted to the ECJ after the British Revenue discovered a VAT deduction mechanism set up by the Halifax company and other subsidiaries aimed at allowing them to deduct the VAT even in cases and circumstances where the core business activity of the parent would not allow it. Basically, running a VAT-exempt business (see Halifax, at § 12) it would have been impossible for Halifax to deduct the VAT paid for repairs on real estates. That’s why the parent company involved in the operation of a considerable number of companies of the same group (see Halifax, § 19), allowing them to recover the VAT. The Revenue considered the entire operation a sort of abuse of the step-transaction approach that still inspires the VAT mechanism, and therefore the commissioners refused any claims of deductions. See also H. L. McCarthy, Abuse of rights: the effect of the Doctrine on VAT planning, in BTR, 2007, p. 160.

[61] A similar conclusion was upheld also in BLP Group, C-4/94.

[62] The context of the two cases is different. In Cadbury the taxpayer challenged the compatibility of a national anti abuse provision with EU law; in Halifax the Commissioners sought to implement an anti-abuse rule relying only on EU law despite the fact that the national provisions didn’t say anything about that.

[63] In this sense, the case Gemente Leusden and Holin Groep, C-487/01 and C-7/02, at § 76.

[64] See Netherlands v. Commission, C-326/85, § 24 and Sudholz, C-17/01, § 34.

[65] This aspect was not debated in Cadbury as far as it involved the existence of an anti avoidance provision. Even if its compatibility with EU law was discussed, every taxpayer was aware of its existence and applicability to all qualifying cases.

[66] Halifax, at § 81.

[67] Halifax, at § 81.

[68] Even if the Court is not clear on this point, it could be argued that the subjective conditions (special relationships between the involved parties) could be considered a factor emphasising the artificial nature of the operations.

[69] This could constitute a partial answer to the provocative thesis by H. Rasmussen in On law and Policy in the European Court of Justice. A Comparative study in judicial policymaking, Dordrecht, 1986, passim.

[70] F. Vanistendael, at footnote 1.

[71] F. Vanistendael, Halifax and Cadbury Schweppes: one single European theory of abuse in tax law ?, in EC Tax Review, 2006, p. 192; see also the report by J. Freedman, G. Loomer and J. Vella, Moving beyond avoidance? Tax risk and the relationship between large business and HMRC, pp.18 and 42.

[72] J. Freedman, G. Loomer, J. Vella, Moving beyond tax avoidance? Tax risk and relationship between large business and HMRC, Oxford, 2007.

[73] Commenting on the Varney Delivery Plan and the relationship between specific taxpayers and the HMRC, J. Freedman, G. Loomer and J. Vella noted that a vital element “… is the desire from both sides for a relationship based on mutual trust” (see J. Freedman, G. Loomer and J. Vella, cit., p. 5).

[74] Avoidance varies according to the legal tradition of each State and to the specific tax it refers to.

[75] M. Cappelletti, Is the European Court of Justice “running wild”?, in European law Review, 1987, p. 3.