Progressive rates can be part of a reference tax system.

The reference tax system is, in principle, made up of the tax base, the tax rates and the taxable events.


Progressive turnover taxes were thought to be selective because they are levied on gross revenue before costs are deducted and larger volumes of revenue are taxed at higher rates. However, after successive judgments by the General Court and the Court of Justice, it is now clear that progressive turnover taxes should be treated the same as progressive income or profit taxes. Those who make more money, pay more taxes.

On 16 March 2021, the Court of Justice in cases C‑562/19 P, Commission v Poland and C-596/19 P, Commission v Hungary, [1] confirmed two earlier judgments of the General Court in cases T-836/16, Poland v Commission and T-20/17, Hungary v Commission.

The European Commission had appealed against the judgments of the General Court which annulled, Commission decision 2018/160 concerning a tax in the retail sector in Poland and Commission decision 2017/309 concerning a tax on advertising revenue in Hungary. In those decisions, the Commission found that the retail tax and the advertising tax were assessed against the turnover of taxpaying companies and that their progressive character constituted incompatible State aid. The Polish tax had three turnover levels each with a different rate: 0%, 0.8% and 1.4%. The Hungarian tax had six turnover levels with six different rates: 0%, 1%, 10%, 20%, 30% and 40%.

The Commission considered that in both cases the relevant reference tax system was the tax itself. The progressive structure was not part of that reference system. In the Commission’s view turnover taxes must have only one rate so that all companies pay the same proportion of their turnover. Therefore, the retail tax and the advertising tax were considered to be selective because both the marginal and average tax rates varied depending on the amount of turnover of each taxpayer. The Commission was of the view that the progressivity of the tax could not be justified according to the ability of taxed companies to pay on the basis of their available resources, as turnover corresponds only to gross revenue, not to net revenue after costs are deducted.

The General Court, however, held that the progressivity of the tax could be justified on the grounds that companies with larger turnover had more resources at their disposal because they benefited from economies of scale. Their operating costs did not increase at the same rate as their operating (gross) revenue. The General Court also held that the Commission created an artificial reference system and defined an arbitrary rate as the reference rate. Nowhere in the relevant national legislation, the General Court stated, was any definition of such a reference system. Consequently, the General Court annulled the decisions of the Commission.

In the two judgments of 16 March 2021 the Court of Justice in essence told us the following:

  1. State aid rules apply to non-harmonised areas even where there is no explicit reference in the Treaty.
  2. Member States are free to determine their tax systems, as long as they comply with State aid rules.
  3. The reference tax system is, in principle, made up of the tax base, the tax rates and the taxable events.
  4. Member States are free to opt for progressive rates that take into account the ability to pay of taxable natural or legal persons.
  5. Progressive rates may be applied to different types of income.
  6. Turnover is a criterion of differentiation that is a neutral and relevant indicator of the taxable person’s ability to pay.
  7. Progressive taxes can be part of the reference system.
  8. Given the discretion of Member States to define progressive rates, the Commission has to demonstrate that a tax measure is “designed in a manifestly discriminatory manner” in order to prove that it is selective.
  9. The fact that certain conditions must be satisfied by taxpayers who benefit from a tax measure cannot in itself make the measure selective.
  10. Transitional rules cannot be part of the reference system, but are also not in themselves selective.

Of the above statements, the fifth one is not legal but empirical. The Court of Justice should have examined more thoroughly whether it corresponded to economic reality even though it left it to the Commission to prove that in practice such as system could be discriminatory.

Because of the two judgments are to a large extent identical, in this article I examine only the judgment in the Polish case and where necessary I also refer to the Hungarian case.

The test of selectivity

The Commission argued that the General Court erroneously found the retail tax not to be selective.

First, the Court of Justice “(26) noted that, […] action by Member States in areas that are not subject to harmonisation by EU law are not excluded from the scope of the provisions of the FEU Treaty on monitoring State aid […] The Member States must thus refrain from adopting any tax measure liable to constitute State aid that is incompatible with the internal market.”

Then, the Court turned its attention to the issue of selectivity. “(28) So far as concerns the condition relating to the selectivity of the advantage, inherent in the concept of a ‘State aid’ measure, […] that condition requires a determination as to whether, under a particular legal regime, the national measure at issue is such as to favour ‘certain undertakings or the production of certain goods’ over other undertakings which, in the light of the objective pursued by that regime, are in a comparable factual and legal situation and which accordingly suffer different treatment that can, in essence, be classified as discriminatory”.

“(29) Further, where the measure at issue is conceived as an aid scheme and not as individual aid, it is for the Commission to establish that that measure, although it provides for an advantage that is of general application, confers the benefit of that advantage exclusively on certain undertakings or certain sectors of activity”.

“(30) As regards, in particular, national measures that confer a tax advantage, it must be recalled that a measure of that nature which, although not involving the transfer of State resources, places the recipients in a more favourable position than other taxpayers is capable of procuring a selective advantage for the recipients and, consequently, constitutes ‘State aid’, within the meaning of Article 107(1) TFEU. Accordingly, a measure that mitigates the financial burdens which are normally borne by the budget of an undertaking and which thus, without being a subsidy in the strict sense of the word, is similar in character and has the same effect is also regarded as State aid […] On the other hand, a tax advantage resulting from a general measure applicable without distinction to all economic operators does not constitute such aid”.

Then the Court reiterated the standard test of determining whether a tax measure is selective. “(31) In order to classify a national tax measure as ‘selective’, the Commission must begin by identifying the reference system, or ‘normal’ tax system applicable in the Member State concerned, and thereafter demonstrate that the tax measure at issue is a derogation from that reference system, in so far as it differentiates between operators who, in the light of the objective pursued by that system, are in a comparable factual and legal situation”.

“(32) The concept of ‘State aid’ does not, however, cover measures that differentiate between undertakings which, in the light of the objective pursued by the legal regime concerned, are in a comparable factual and legal situation, and are, therefore, a priori selective, where the Member State concerned is able to demonstrate that that differentiation is justified in that it flows from the nature or general structure of the system of which the measures form part”.

Can progressive rates be part of the reference system?

Next the Court applied the principles above to the case at hand. “(36) The question arises first of all whether, […] the progressivity of rates provided for by the tax measure at issue was to be excluded from the reference system in the light of which it was appropriate to assess whether the existence of a selective advantage could be established, or whether, […] it is, on the contrary, an integral part of that system.”

“(38) Outside the spheres in which EU tax law has been harmonised, the determination of the characteristics constituting each tax falls within the discretion of the Member States, in accordance with their fiscal autonomy, that discretion having, in any event, to be exercised in accordance with EU law. This includes, in particular, the choice of tax rate, which may be proportional or progressive, and also the determination of the basis of assessment and the taxable event.”

Then the Court identified the components that make up the reference system. This is very useful because normally the Court only identifies the reference system in each particular case without defining it in more general terms. “(39) Those characteristics [i.e. the last sentence of para 38] constituting the tax therefore, in principle, define the reference system or the ‘normal’ tax regime, from which it is necessary, […] to analyse the condition relating to selectivity.”

“(40) In that regard, it must be stated that EU law on State aid does not preclude, in principle, Member States from deciding to opt for progressive tax rates intended to take account of the ability to pay of taxable persons. The fact that recourse to progressive taxation is, in practice, more common in the taxation of natural persons does not mean that they are prohibited from using it in order also to take account of the ability to pay of legal persons, in particular undertakings.”

“(41) EU law thus does not preclude progressive taxation from being based on turnover, including where such taxation is not intended to offset the negative effects likely to be caused by the activity being taxed. Contrary to what the Commission maintains, the amount of turnover constitutes, in general, a criterion of differentiation that is neutral and a relevant indicator of the taxable person’s ability to pay […] It does not follow from any rule or principle of EU law, including in the field of State aid, that progressive rates may apply only to taxes on profits. Moreover, like turnover, profit in itself is merely a relative indicator of ability to pay. The fact that it may constitute, as the Commission contends, a more relevant or more precise indicator than turnover is irrelevant in matters of State aid, since EU law on that matter seeks only to remove the selective advantages from which certain undertakings might benefit to the detriment of others which are placed in a comparable situation. The same is true of the possibility of economic double taxation, linked to the combined taxation on turnover and taxation of profits.”

In the paragraph above, the Court downplays the relevance of profit taxes while it makes a pretty strong statement about the neutrality and relevance of turnover taxes. These statements do not emanate from legal principles. They are empirical and the Court should have acknowledged that. But it left a window open for the Commission in future cases to prove the selectivity of turnover taxes by showing how their application discriminates against taxable persons in comparable situations.

“(42) It follows from the foregoing that the characteristics constituting the tax, which include progressive tax rates, form, in principle, the reference system or the ‘normal’ tax regime for the purposes of analysing the condition of selectivity. That said, it cannot be ruled out that those characteristics may, in certain cases, reveal a manifestly discriminatory element, which it is, however, for the Commission to demonstrate.”

“(44) In the present case, […] the Polish legislature, by the tax measure at issue, introduced a tax on the retail sector based on the monthly turnover generated by that activity, which, contrary to what the Commission maintains, is a direct tax. […] The Commission has not established that that progressivity of the rates, adopted by the Polish legislature in the exercise of its discretion in the context of its fiscal autonomy, was designed in a manifestly discriminatory manner, with the aim of circumventing the requirements of EU law on State aid. In those circumstances, the progressivity of the rates of the tax measure at issue had to be regarded as inherent in the reference system or the ‘normal’ tax regime in the light of which the existence, in the present case, of a selective advantage had to be assessed.”

While the Court earlier directed the Commission to prove empirically the selectivity of a turnover tax, in paragraph 44 above it raised the standard of proof because the Commission will have to demonstrate that a tax measure is “designed in a manifestly discriminatory manner”. But, when, as in the Polish case, the legal text of a tax measure simply states that different rates apply to turnover above certain thresholds, how can anyone show that that is manifestly designed to be discriminatory? How can the Commission prove the intention of a measure that is made up of a few rates and a few thresholds for the purpose of raising revenue, as was the formal aim of the Polish measure?

The Court concluded that “(45) the General Court did not therefore err in law in holding, […] that, by considering that the progressive scale of the tax measure at issue was not part of the reference system in the light of which the selective nature of that measure had to be assessed, the Commission had incorrectly relied on an incomplete and fictitious reference system.”

Because “(46) an error in determining the reference system necessarily vitiates the entire analysis of the condition relating to selectivity”, the Court upheld the judgment of the General Court.

Deductibility of losses carried forward

In the Hungarian case, the Commission also argued that the General Court erred in finding that the partial deductibility of losses carried forward, allowing undertakings to deduct 50% of their losses, was not a selective advantage.

The Court of Justice reiterated the test of selectivity and went on to point out that “(58) the fact that only taxpayers satisfying the conditions for the application of a measure can benefit from a measure cannot, in itself, make it into a selective measure […] Nor can the selective nature of a measure be inferred from the mere fact that it is of a transitional nature, since the decision to limit its application ratione temporis, with a view to ensuring a gradual transition between old and new tax rules, falls within the discretion of the Member States”.

“(59) In the present case, […] the Hungarian legislature sought to moderate the tax burden borne by the most economically vulnerable undertakings for the first year of their liability in respect of the tax measure at issue […] Since, from the outset, it was intended to be transitional, that mechanism cannot be regarded as part of the reference system or the ‘normal’ tax regime, in the light of which the analysis of its selective nature must be made, even if it were to resemble a basis-of-assessment rule.”

“(60) It is therefore necessary to examine whether the mechanism for the partial deductibility of losses carried forward introduces a difference in treatment as between operators which are, in the light of the objective pursued by the Law on advertisement tax, in a comparable factual and legal situation.”

Then the Court observed that “(61) that mechanism introduces a distinction between, on the one hand, undertakings with losses carried forward in respect of previous years, provided that they did not make a profit in 2013, and, on the other hand, undertakings that made a profit in that year, as only the former are entitled to deduct those losses carried forward when calculating the basis of assessment for the tax measure at issue in respect of 2014.”

“(62) In the light of the objective of redistribution pursued by the Hungarian legislature in adopting the Law on advertisement tax, as evidenced by the progressivity of the tax measure at issue, those two categories of undertakings are not in a comparable factual and legal situation. The choice of a basis of assessment expressed according to turnover does not render inconsistent, in relation to that objective, the adoption of a transitional measure taking profit into account, since the latter also constitutes, […] an indicator which is both neutral and relevant, even though it is relative, of undertakings’ ability to pay.”

“(63) The criterion relating to the lack of profits in respect of the 2013 financial year is in that regard objective, since the undertakings concerned, from that point of view, have a lesser ability to pay than others on the date of entry into force of the Law on advertisement tax, during the course of 2014.”

“(64) Consequently, the Hungarian legislature was entitled, without infringing EU law on State aid, to combine, in respect of the first year of application of that law, the measurement of ability to pay resulting from the amount of turnover with a measure enabling losses carried forward by undertakings that did not make a profit in 2013 to be taken into account.”

“(65) The fact that the undertakings liable to benefit from the mechanism for the partial deductibility of losses carried forward were already identifiable on the date on which the tax measure at issue was introduced is not, in itself, capable of calling that conclusion into question.”

Therefore, the Court rejected all the pleas of the Commission and dismissed the appeal.

[1] The text of the two judgments can be accessed at:

Commission v Poland:

Commission v Hungary: