A Closer Look At Apple’s Victory In EU Tax Case

Tax Notes chief correspondent Stephanie Soong Johnston and legal reporter Ryan Finley discuss Ireland and Apple’s state aid win with law professors Ruth Mason of the University of Virginia and Stephen Daly of King’s College London.

David Stewart: Welcome to the podcast. I’m David Stewart, editor in chief of Tax Notes Today International. This week: iWin. In July the European Union’s General Court ruled in favor of Ireland and Apple in a long-running dispute over allegations that Ireland offered illegal state aid to the company’s subsidiaries. The case was one of several state aid enforcement actions brought by the European Commission as it sought to clamp down on tax avoidance by multinationals.

How did the court reach its decision? And what does it mean for the EU and other policymakers?

My guests are Tax Notes chief correspondent Stephanie Johnson and Tax Notes legal reporter Ryan Finley. Stephanie, Ryan, welcome back to the podcast.

Stephanie Soong Johnston: Good to be here. Thanks.

Ryan Finley: Thanks.

David Stewart: Let’s start off with some background here. Apple has been somewhat notorious for achieving a low rate of tax on its overseas earnings. How does it manage to do that?

Ryan Finley: Apple used a variation of the old double Irish structure, which exploits mismatches in U.S. and Irish tax law.

Basically, Apple formed two Irish incorporated subsidiaries, ASI and AOE, which held rights to exploit the Apple group’s intellectual property outside the Americas under a cost-sharing agreement with Apple Inc. The vast majority of the profit reported by these two entities was attributable to these IP rights that they acquired through the cost-sharing arrangement that they co-funded, even though the activities took place exclusively in Cupertino.

The only actual functions that ASI and AOE performed were relatively routine manufacturing procurement service functions, and they took place at their Irish branches. Because the subsidiaries were incorporated in Ireland, but effectively managed elsewhere, they avoided tax residency in any jurisdiction. Their only taxable income was whatever profit could be attributed to the Irish branches, which under two Irish advanced pricing agreements was a very modest amount.

David Stewart: What is state aid and how does it come into play in Apple’s situation?

Ryan Finley: State aid is a part of the European Union’s competition law. In general, state aid is a selective economic benefit given to a business or a group of businesses that distorts competition. There are exceptions, but state aid is usually prohibited by the Treaty for the Functioning of the European Union. Any member found to have given illegal aid has to recover that aid from the beneficiaries.

This all became relevant for Apple when the European Commission opened an investigation into these two APAs — that were granted in 1991 and 2007 — that according to the commission taxed only a small fraction of what ought to have been taxed in Ireland. The commission’s reasoning, which the General Court called an exclusion approach, was basically that the only activities ASI and AOE performed took place at these Irish branches. Therefore, all of their profit must have been attributable to these branches by default. According to the commission, Ireland’s failure to tax this profit was a form of illegal state aid.

David Stewart: What other companies have faced similar scrutiny for under state aid theories?

Stephanie Soong Johnston: As Ryan indicated, state aid rules are about conferring legal advantages to multinationals that operate in the European Union. These state aid rules apply to a lot of different industries and a lot of different rules, not just tax. But in recent years, the commission has been focusing on tax rulings of several multinationals, mostly European but also some American, with mixed results.

The commission in June 2014 looked at a tax ruling that the Netherlands gave Starbucks in 2008. In October 2015 this commission decided that Starbucks did receive state aid and said the Netherlands should have to recover €20 million to €30 million in tax. But the General Court in September 2019 overturned that decision and the commission didn’t appeal it.

You might recall that also in June 2014 the commission announced a state aid investigation into Fiat. In taking a closer look at a 2012 tax ruling that Luxembourg gave to that company, and again in October 2015, the commission said that Fiat did receive state aid and Luxembourg should also recover €20 million to €30 million of tax. That commission decision was upheld by the General Court in September 2019. But Fiat appealed that decision and the case is still ongoing. So, that was a win for the commission.

The commission launched an investigation into McDonald’s in December 2015, taking a closer look at two tax rulings that Luxembourg granted in 2009, and found in September 2018 that there was no state aid. That was sort of a nonissue there.

But, we also are keeping an eye on Ikea. The commissioner launched an investigation into Ikea in 2017, taking a closer look at two tax rulings that the Netherlands gave it in 2006 and 2011. As you might recall, at the end of April, the commission expanded the scope of that investigation. We’re still waiting for the commission’s decision there. 

Apple, however, was the most high profile of these cases. Not only because it was a major household name, but also because of the amount of money that Ireland had to recover — some €13 billion in illegal state aid. That’s what the commission alleged Ireland gave Apple. More to come in this space.

David Stewart: Apple and Ireland won their case. What did the court have to say in its ruling in favor of them?

Ryan Finley: The court held that the commission failed to carry its burden of proof in establishing illegal state aid. To provide more detail on the basis for the decision, I spoke to Steven Daly, a professor at King’s College London.

David Stewart: Let’s go to that interview.

Ryan Finley: Hello, Professor Daly. Thanks for talking to us about the Apple decision today.

Stephen Daly: It’s great to be here.

Ryan Finley: As you know, the General Court’s decision was based on the holding that the commission did not meet its burden of proof in establishing state aid. What about this reasoning makes an appeal more or less difficult for the commission?

Stephen Daly: I think it’s important to actually take a quick step back and point out what the wins for the commission actually were in the case. The commission got a big win in that it is entitled to look at transfer pricing arrangements, and it is entitled to use the arm’s-length principle when it does that.

The commission got a couple of wins in the case, but the problem is basically in the way in which you appeal. The General Court said the commission’s approach was wrong. Because all the evidence that the commission had provided was predicated on that approach, that meant that the evidence that was provided also was no good. That makes appealing much more difficult.

To get into a little bit more of the specifics, what the General Court said was, “You’ve got Irish law on this wrong, and you’ve also got the authorized OECD approach wrong. The way in which you have tried to apply both Irish law and the OECD guidance is wrong.” The reason that it was wrong essentially came down to a failure to look at what was actually happening in Ireland.

With the Irish law, the General Court said, “Well, it used the case of Dataproducts, a High Court decision from 1985 in Ireland, to suggest that the commission ought to have looked at whether the branches in Ireland controlled the intellectual property.” Because the commission did not look at whether the branches actually controlled the IP, that meant that its approach to Irish law was misconceived.

Similar reasoning then on the authorized OECD approach. Because the commission had not actually looked at the risks, the assets, and the functions of what was actually going on in Ireland, that meant that the commission got it wrong there also.

Then finally, for good measure, what the General Court did was it pointed out just how limited the operations in Ireland were and are. By way of brief background, if you can probably tell from my accent, I’m from Ireland. I’m actually from Cork. So I know Apple very well. It kind of hurts a little bit for the General Court to point out just how limited Apple’s operations were and just how insignificant Ireland is in terms of generating value for the Apple company.

Ryan Finley: Do you see any issues with the Court’s reliance on the Dataproducts case? Do you think the reasoning is sound on that?

Stephen Daly: Excellent. I’m so glad that you asked me that. What happened with Dataproducts was you have a Dutch resident company. The Dutch resident company had a trading arm. It had a branch in Ireland. Now, this Dutch resident company also had a bank account in Switzerland, and that was generating interest.

The question in Dataproducts then was whether you could attribute the interest from that bank account back to the Irish branch. Mella Carroll in the High Court in Ireland said, “You need to look at whether the Irish branch actually controlled that asset, that bank account, over in Switzerland.”

The reason that that’s kind of vulnerable here is because it’s certainly not a direct precedent for what was going on in Apple. It’s not even really analogous to what was happening in Apple. It would have been analogous to Apple if the issue was with what do you do with royalties in relation to IP. That will be passive income, just like interest on that Swiss bank account.

But that’s not what’s at issue. What’s at issue is the active trading income. So it’s just not an analogous case. It’s vulnerable in that is not a direct precedent or even an analogous precedent, but it may even be entirely irrelevant. It may have taken the Court in the wrong direction altogether.

Ryan Finley: That seems like that would be a particular option as grounds for an appeal. Does the commission have any other sort of procedural options? The commission are not really restricted by res judicata, like they would be here. Is there anything they could do to just sort of reopen the proceedings on a different basis?

Stephen Daly: It’s an excellent question. Yes. For U.S. listeners, that might seem a little bit odd, but there’s actually nothing barring the commission from starting an entirely new investigation today or tomorrow. We have actually a precedent that I found today looking at this, which was the EDF case.

EDF was a state aid case. It related to a subsidy that was provided in 1997. The original commission decision was from 2003 in that case. The General Court then rejected the commission’s case. It annulled the commission’s decision in 2009. The European Court of Justice also upheld the General Court’s decision in 2012. The commission lost starting in 2003 and lost definitively in 2012.

The commission started a new investigation into EDF in 2013 and used the guidance from the General Court and the European Court of Justice to take the case, the new case, against EDF. It finalized its decision in 2015. The General Court upheld its decision in 2018, and the European Court of Justice then upheld that decision in turn. You had the commission originally losing and then going back, starting a new investigation, raising the exact same facts, and then winning eventually in 2018.

The only thing that you might worry about is limitations. There’s nothing stopping the commission from doing that, but is there a limitation on its recovery of whatever unlawful aid was provided? There is a 10-year rule that comes from EU law. The commission can only look back 10 years. You might think then that means that the commission, if it were to start to new case tomorrow against Apple, could only look back to 2010. That is wrong. The clock stops ticking from whatever happens in 2013. So, the commission can look back as far as 2003, because it originally started its investigation into Apple in 2013, just like EDF.

EDF was started in 2003, and that’s where you can look back to 1997. Even though a new case was taken in 2013, that didn’t mean that the commission had to stop. The commission could still look back at what happened in 1997. It would be the same thing too with Apple. Procedurally, the commission can indeed initiate a new case.

The question then is what should the commission focus on if it were to take a new case? There are aspects of Apple in Ireland that are vulnerable if an entirely different case was taken.

First is residence. We know what happened with Apple was you had companies that were incorporated in Ireland. So from a U.S. perspective, they were Irish residents, but because the companies were not managed and controlled in Ireland, that meant that from an Irish perspective, they were resident outside of Ireland. That meant that you had these stateless companies. Now, was this an innocent outcome on the basis of Irish law?

Well, Irish law actually changed in 1999, but it continued to allow for that exemption for essentially U.S. multinationals who wanted to have a trading arm in Ireland and who wanted to continue to be nonresidents. So, the statelessness is not an accident of Irish legislation. It is purposeful. There’s precedent in EU law in the form of the case of Gibraltar to suggest that even facially neutral rules can give rise to de facto subsidies. If that does happen, then there can be a finding up on local state aid.

The commission could attack the Irish rules themselves. There was actually a whisper of this in 2014, and it seemed to get killed. There’s an article that you can go back and read in the Financial Times in 2014 where the commission’s highlighting that it might attack the Irish residency rules. Anyway, so that might be one option.

The strongest argument, I think, that could be taken is actually a bit what happened procedurally with Apple in Ireland. This it seems to me to be where the commission even came in in the first place. It looked at what happened procedurally in relation to these 1991 and 2007 rulings and said, “Well, there’s things that are a little bit iffy about this.” We have from the original commission decision the suggestion that the Irish ruling in 1991 was driven not by the objective of collecting tax, but actually by of ensuring that Apple would come to Ireland and would employ people.

Now, if it is actually proved that the Irish revenue commissioners went out of their way to grab a favorable ruling to Apple, that is unlawful as a matter of Irish administrative law. That to me seems to be a very good case that could be taken by the commission.

There are also failings in terms of the paperwork leading to these rulings. We had in 1990 a meeting between Apple and the Irish revenue commissioners. The Apple tax advisor suggested a particular profit figure that would be taxable for the branches in Ireland. It was €30 to €40 million and the Irish revenue commissioners asked, “Well, where does that figure come from?” The tax adviser said, “Well, there’s no scientific basis for it, but we thought it was so large that you take it as good faith.” The Ireland revenue commissioners obviously turned around and said, “We can’t accept that.”

In 1991 we have another meeting. We have further negotiations. We have a bit of back and forth and there’s a formula that’s agreed. It just so happens that that formula produces a taxable profit figure of the same 1990 figure that had no scientific basis. That’s all a bit suspect.

I think that a commission could predicate a case on the basis of the improper collection of tax methods employed by the revenue commissioners. As I said, the commission won in terms of it is allowed to apply that principle. It is also allowed to apply the authorized OECD approach.

The commission now has that as a stick that it can beat member states with, if it suspects that their ruling practices are out of line. It has that win. That’s practically significant.

It’s also worth noting that in 2014 Ireland actually changed its residency rules so as to prevent stateless companies from actually arising. So, we had that little win as well. That doesn’t come through in a legal decision, but that is very much the commission’s soft power being exercised. The commission has now established itself as a player, whether we like it or not, in the international tax sphere.

Ryan Finley: Thanks very much for speaking with us, Professor Daly.

Stephen Daly: It was absolutely my pleasure. Thank you.

David Stewart: Stephanie, this is all happening in the broader context of international efforts to deal with the taxation of the digital economy. Is this case expected to have some sort of effect on those discussions?

Stephanie Soong Johnston: The decision comes at a particularly sensitive time in global negotiations on this issue. As you might recall, the OECD is leading negotiations, trying to steer nearly 140 jurisdictions toward agreement on a common approach to update the international corporate tax rules by the end of 2020, which that timeline seems a little ambitious at this point.

Just to really quickly review what’s on the table is pillar 1, which deals with the profit allocation and nexus rules, revising those rules, and pillar 2, which ensures a minimum taxation of multinational profits. In some ways this Apple decision highlights these problems in international corporate tax rules. In a lot of ways, it raises the stakes for these countries to come to some agreement on how to fix the rules, so that companies like Apple don’t get away with paying very low rates of tax in the jurisdiction in which they operate.

While there’s no direct link to the two situations, the General Court’s position has political implications for these negotiations. Pascal Saint-Amans, the OECD’s tax chief, told me that the decision “shows that a global solution is even more necessary.” A solution on the table at the inclusive framework addresses some of the underlying concerns about Apple and other multinationals.

What this means for the U.S., I spoke to Ruth Mason, a professor of tax law at the University of Virginia School of Law. She had some interesting insights about what this means for the U.S. against the backdrop of these discussions.

David Stewart: Let’s go to that interview.

Stephanie Soong Johnston: Thanks so much, Professor Mason, for coming onto the podcast. You had mentioned that the judgment was unclear on some critical points. What were those points and why are they important?

Ruth Mason: The decision is really circuitous, and it’s kind of hard to tell what the court means. I have worked out what the reasoning was, and I’ll tell you what I think the court said.

First, if the profit of resident companies is determined at market conditions — this is something that is always true, right? Resident companies, standalone companies, their profits are always determined by actual market conditions. If that’s true, and the purpose of the domestic tax system is to treat group companies and standalone companies on equal terms. For example, by taxing both of their profits.

This is also something that’s almost always or always going to be true. The commission’s entitled to use the arm’s-length standard as a benchmark for verifying that the state properly determined the income of a group member. If the state’s rule, like the Irish rule, is anything like arm’s length, then the commission can use OECD arm’s length to test whether Ireland did its allocation properly. What this all basically adds up to is that OECD guidance is the benchmark for state aid cases involving transfer pricing.

What kind of goes back and forth? Is it a matter of domestic law? Reject this idea that it’s part of EU law, but even though they say it’s part of domestic law, it’s a really capacious interpretation of domestic law. Where is your standard anything like the OECD’s standard? If the answer to that is yes, then the commission can use the OECD standard.

The General Court is pretty harsh on the commission’s application of the OECD arm’s-length standard. The commission didn’t apply the OECD arm’s-length standard as far as the General Court was concerned. If the OECD standard is the right standard, it’s very hard to see how the Commission could snatch a victory on appeal.

What the commission would have to argue is that the General Court was wrong in the law, so that the commission is not bound by the OECD authorized approach. Instead, the commission is entitled. Its argument would run to use its own arm’s-length standard as a matter of EU law. If the commission gets to use its own standard, it can make up the content for that standard, including that it can use this allocation by exclusion approach.

In the final decision in Apple, the commission came with guns blazing. It said, “It’s our standard. We get to define it. We interpret EU law. We’re the enforcer of the state aid rules, the prohibitions of state aid. We apply benchmarks all over the place in all different areas. There’s a whole bunch of them. They all have separate names. And this is the one that we use in transfer pricing cases.” It was treating tax like lots of other areas of law.

Like when member states make loans to member states, the commission asks, “Well, would a private party make the same loan on the same terms?” This is called the market economic operator tests. So, the commission said with some justification, looking analogously at other areas of law, this is just the benchmark for tax cases. The problem is, in my view, that the commission sort of confused the idea. The requirement that a member state has to hold companies at arm’s length with the arm’s-length standard. The two are different.

The idea that that state has to hold private enterprises at arm’s length is just to say you can’t subsidize. That’s a state to company arm’s-length standard. But the commission turned that into a mandate to apply what I would call pocket to pocket arm’s length. The idea that parts of a company have to keep other parts of the company at arm’s length. That’s not an EU law idea. That’s an idea that’s grown up out of the OECD. There’s a standard. Companies add to it. It accretes over time. But that’s not the same as the principle that underlies the prohibition of state aid, which is that states have to keep companies at arm’s length.

Stephanie Soong Johnston: What does the General Court’s decision in Apple mean for the ongoing talks on the OECD on pillars 1 and 2?

Ruth Mason: I don’t think that there’s any really clear implication that comes out other than their implications from Apple, for the power of the commission. The commission was flexing is muscle; it came out with, as I said, guns blazing with this arm’s-length standard, and then it contracts.

The commission comes out, POW! Then it gets the reaction. It gets the reaction from the U.S. It gets the reaction from the member states and multinationals. I think it sort of thought, “Wow, we didn’t really quite maybe know what we were getting ourselves into.” Then the commission really narrows its legal arguments. Once it gets to the General Court, it’s saying, “Oh, we were just using a benchmark.”

But notwithstanding that it lost in Apple, I think the commission has still established itself. It is an actor in international tax. It has a view and everybody’s paying attention.

Stephanie Soong Johnston: You mentioned the U.S. needs to really get itself to the table. It really has a lot to lose if it doesn’t. Can you elaborate on that point? What should the U.S. be doing after Apple, as it engages with the OECD on the solution?

Ruth Mason: One lesson that might be drawn from Apple that I think would be wrong would be for the U.S. to say, “See, there’s no problem. Right? We don’t have to cooperate. We don’t have to compromise because Apple won. The commission has basically been told that this means of combating tax avoidance is not going to be that effective over time.”

But the problem isn’t just Apple. The problem isn’t just commission state aid investigations. It’s bigger than that. There’s digital taxes. There’s an EU proposal for digital tax. All these unilateral proposals for digital tax. Progress is being made on those unilateral actions with or without the U.S. The problem with those unilateral actions is they lead to double tax and double tax reduces growth and reduces cross border investment for everyone. It’s in everybody’s interest to avoid that. The way to do that is to arrive at a compromise.

Stephanie Soong Johnston: Excellent. Thank you.

David Stewart: Stephanie and Ryan, thank you very much for your insights and for the interviews you did.

Stephanie Soong Johnston: You’re welcome. Thank you.

Ryan Finley: Thank you.