The ruling by the EU’s second-highest court against Brussels over Ireland’s favourable tax arrangements for Apple means the European Commission has lost one battle in its fight against damaging corporate tax practices in Europe. Margrethe Vestager, as competition commissioner, had ordered Apple to return €13bn to Dublin in what she had found to be underpaid corporate taxes. The general court disagreed; an appeal to the highest court is to be expected.

But the commission is determined not to lose the war on member states undermining one another’s tax bases with the result that multinationals are taxed more lightly than their local competitors. Brussels has vowed to double down on its efforts and is looking for additional legal tools. It is right to do so.

It was the paucity of direct legal tools to address the flaws of multinational corporate taxation — tax remains a matter of national power in the EU — that led Brussels to use competition policy tools in the first place. At issue is a bizarre corporate status that Irish tax law used to allow, by which a company could be incorporated in Ireland but not be tax-resident there, even though its Irish corporate status would not make it tax resident under any other countries’ laws either. Apple based its European business in Ireland, but inserted just such a ghostly entity in its corporate structure between the US parent and the Irish branch, and got Irish tax authorities to agree that a good chunk of the profit belonged to the ghostly middle layer. Ms Vestager argued that this violated EU rules on state aid.

It is crucial to understand what the court did and did not say when it struck down her claim. Most importantly, the court did not dispute the legal premise that sweetheart tax deals could constitute illegal state aid. It did, however, rule that the commission’s specific empirical claim — that the Irish tax decisions gave Apple a selective advantage, the test for state aid — had not been proven. In short, it told the commission’s competition lawyers that they did not do a good enough job. 

Specifically, the judges found the commission fell short in two ways. Brussels had argued that because the ghostly non-resident layer had no staff, it could not have done the work from which profits arose, so those profits should be allocated to the flesh-and-blood Irish operations. The court said this “exclusion” argument was invalid and the commission would have to show that the work was actually done by the resident Irish entities. Second, it said that while the commission convincingly proved Dublin was often erroneous and arbitrary in determining where profits should be accounted for — and this finding is important — that in itself did not prove the errors were to Apple’s advantage over other companies.

The court could have gone further: it could have said that the claim was wrong, that Apple did not receive illegal state aid. It did not. It did not even say that the claim could not be proven, only that it had not been. That opens up the possibility for the commission to try to prove its case better, and we must hope it does so.

But it also opens up another possibility: that it can never be proven because the nature of a business such as Apple (and other big multinational companies whose value lies mainly in knowledge and intellectual property) is that much of its value is in excess of what can be easily attributed to any activity done anywhere. As the court points out, the commission says the ghostly middle entities did not do anything that could have created the profits, but the same could be true for the Irish resident entities. And this could be said for any part of the operation, if Apple’s total profits are out of proportion with the cost of the investment and work put in — in other words, if there is a large element of “economic rent”. And this is the case for a lot of the world’s most modern companies.

The standard “arm’s length” test for where profit should be allocated — who would get it if all the corporate entities were engaged in independent free competition against each other — fails when it comes to economic rent, because in the idealised competitive benchmark, such rents would not exist. This economic understanding suggests there is a blind spot in the standard legal reasoning.

The question we really ought to be asking ourselves, of course, is why the introduction of a middle layer in the corporate structure — between Apple’s US parent and the Irish-resident entity — with a ghostly relationship with Ireland but no other country, should in any way affect where profits are properly taxed. The only sensible answer is that it should not (which is no doubt why Ireland has since moved away from the ghost status). Its only function was to find a low-tax home for the large share of Apple’s profits that does not obviously redound to traditional production input.

If the law really allows this, the law is an ass and should indeed be changed. But, as I said above, it does not follow from the general court’s ruling that the law allows it, just that it remains unproven that allowing a company to park its economic rent offshore gives it a selective advantage. But once one focuses on the economic meaning of what is happening, it is hard to see how it could not. It means that unusually profitable companies can avail themselves of lower tax charges than normally profitable ones, because the unusually high part of the profit (the economic rent) will typically be less obviously attached to any production input. This point deserves pride of place when the case goes to appeal.

Coronanomics readables

  • Can the pandemic offer an opportunity to make our societies better? In my long read for FT Weekend, I try to give an answer.

  • One of the less-noticed parts of the EU’s proposed recovery package is equity support for companies. The pandemic has destroyed a lot of equity, so such support is needed to avoid zombie companies or disorderly insolvencies. But as Theresa Küspert and Nils Redeker explain in new policy paper, the European Commission’s proposed design needs to be improved and more rooted in Europe’s industrial policy considerations, and handled by institutions equipped to take them into account.

Other readables

  • Bulgaria and Croatia have simultaneously joined the EU’s banking union and its exchange rate mechanism, the antechamber to the euro. As the European Central Bank’s Fabio Panetta explains in a speech, this means the two countries might be able to adopt the single currency by 2023.

  • The New Yorker reports on the basic income experiment in Stockton, California.

  • In a week where the UK government’s message on Brexit is “we’re moving on, the future is bright”, the Brexit talks reflect the exact opposite.

Numbers news

  • The FT has launched its high-frequency UK economy tracker.

  • The evidence keeps coming in about how unevenly the economic effects of the Covid-19 pandemic fall. In the UK, the Institute for Employment Studies reports that the employment rate among the low-paid has fallen by 4 percentage points but hardly at all among the high-paid. That difference comes in addition to a sharper rise in the share of the low-paid who, while still employed, are away from work.

Get alerts on Corporation tax when a new story is published

Copyright The Financial Times Limited 2021. All rights reserved.
Reuse this content (opens in new window)

Follow the topics in this article