The Vodafone and Tesco Global decisions: no triumph for EU digital services tax supporters

By Leopoldo Parada, Lecturer in Tax Law, University of Leeds School of Law

The Court of Justice of the European Union (CJEU) on 3 March released two important and long-awaited decisions addressing challenges to the Hungarian telecommunication tax brought by British telecommunications firms Vodafone and Tesco Global.

In both cases, the Court concluded that the special Hungarian tax on telecommunications does not preclude the freedom of establishment just because the tax burden falls mostly on foreign-owned taxable persons, nor does it prevent Member State legislation from imposing taxes using progressive turnovers when the effect of that legislation is that the tax burden falls mostly on foreign-owned taxable persons.

Supporters of digital services taxes in France, Italy, Spain, and other Member States have received these decisions as an absolute legal triumph. However, this triumphalist attitude is far from realistic.

Please do not get me wrong, the decisions are relevant for the discussion on digital services taxes, but they are rather limited and slippery.

Whilst the CJEU confirms in Vodafone and Tesco Global that a pure statistical correlation is not enough to consider that discrimination exists, it avoids referring to other elements that may play a role in determining discrimination, in particular, the role of discriminatory intent, which is present in the majority –if not all– digital services taxes proposed and already in force around Europe.

Indirect discrimination and the use of a pure quantum criterion

The CJEU decisions confirm that the circumstance that high turnovers enterprises are mostly from other Member States, and therefore subject to the higher progressivity of the tax, is not enough to conclude that discrimination existed. This conclusion is nonetheless both predictable and obvious.

Since Hervis Sport, the Court has not been able to establish a consistent formula to determine how much foreign-owned companies are necessary to conclude that indirect nationality discrimination occurs. This is not a fault of the Court, but rather a confirmation of reality.

Indeed, the inherent difficulty of establishing whether a “majority”, “vast majority”, or whatever other quantum criterion used for this purpose appears a priori to be insufficient, and nobody could argue against that.

This is why the CJEU correctly asserts in Vodafone (para 52): “The fact that the greater part of such a special tax is borne by taxable persons owned by natural persons or legal persons of other Member States cannot be such as to merit, by itself, categorisation as discrimination”. 

However, it is precisely the recognition that a quantum criterion is not sufficient to conclude that indirect discrimination exists, that raises a more important question, which is what other criteria should be considered. Unfortunately, the CJEU limited itself to confirm a long-standing obviousness without going beyond.

Discriminatory intent: It matters and should be used

The CJEU has regularly emphasised that what matters in discrimination cases is effects, not intent. This assertion is correct, but perhaps also short-sighted.

If one agrees that the use of a quantitative criterion or a pure statistic correlation is not enough to determine indirect discrimination, the CJEU could use other elements that may help solving cases of covert discrimination, particularly the role of legislative intent, which unlike the Hungarian telecommunication tax cases, is present in the majority –if not all– digital services taxes proposed and already in force around Europe.

If one agrees that the use of a quantitative criterion or a pure statistic correlation is not enough to determine indirect discrimination, the CJEU could use other elements that may help solving cases of covert discrimination, particularly the role of legislative intent, which unlike the Hungarian telecommunication tax cases, is present in the majority –if not all– digital services taxes proposed and already in force around Europe.

As Prof. Ruth Mason (University of Virginia) and I have already suggested somewhere else, the CJEU could use intent in indirect discrimination cases in three different ways:  (I) ignore it; (II) use it to trigger impact analysis; or (iii) use it to lower the quantum of impact required to establish nationality discrimination.

First, the CJEU could simply ignore intent every time a suspicious classification obviously correlates with nationality. For example, when tax residence is used as a classification. In those cases, the CJEU –as it has already ruled before– could avoid any evidence of impact or intent and rule straightforwardly.

Second, the CJEU could also use intent to trigger an impact analysis in discrimination cases. This is what AG Kokott suggested in her opinions both in Vodafone and Tesco Global, and which was clearly not considered by the Court.

In simple words, Member States would not be judged a priori for their tax policy decisions, and the use of a specific distinguishing criterion, simply because it happens to incidentally correlate with nationality. However, where such a distinguishing criterion, which is not intrinsically disadvantageous, appears to be intentionally chosen to disadvantage foreign shareholders, such an intention must be legally relevant. This could be inferred, as AG Kokott suggests, from the principle of prohibition of abuse of rights, under which Member States could not exploit special market situations choosing a particular classification criterion (for example, high revenue thresholds) just to shift the burden onto foreign-owned taxpayers.

Finally, the CJEU could use intent to simply lower the evidentiary burden required to establish nationality discrimination. That is, when a pure quantum correlation shows that a majority of foreign-owned taxable persons bear the tax burden of a new tax, the evidence of discriminatory intent could help the CJEU to conclude that indirect discrimination exists.

As such, if the Court had evidence of discriminatory legislative intent in Vodafone and Tesco Global, it could have used that evidence to conclude that the quantum correlation was neither fortuitous nor a simple matter of chance.

That evidence might have not been available in these two cases, but it appears to exist in all proposals and implemented laws with regard to DTSs in Europe.

Vodafone and Tesco Global conclusions

The slippery attitude of the Court avoiding to discuss the issue of discriminatory intent may be justifiable in the absence of evidence in the case of Vodafone and Tesco Global.

However, this is far from being the case in future litigation regarding digital services taxes, where the CJEU will have to say something more than simply confirming the obviousness that a pure statistical correlation is not enough to conclude that (indirect) discrimination exists.