Keith O’Donnell has been a tax advisor in Luxembourg for more than two decades Photo: Guy Wolff/Maison Moderne

Keith O’Donnell has been a tax advisor in Luxembourg for more than two decades Photo: Guy Wolff/Maison Moderne

Thickening rulebooks may be good for tax consultancies like his, but Atoz’ Keith O’Donnell thinks we’ve had enough of multinational corporate tax reform. He argues that increased complexity is costing companies too much and makes the EU less competitive.

Aaron Grunwald: When you were speaking at a conference organised by the UEL employers’ association earlier this year, you said there’s a “risk that Europe gets stuck on the hamster wheel of constantly trying to redo corporate tax”. What did you mean by that?

Keith O’Donnell: So what I meant by that is, there’s been a huge amount of corporate tax reform over the last few years, mainly driven out of the OECD. Because going back 10-plus years, the G20 agreed that something needed to be done with corporate tax reform, because there was a concern about aggressive tax planning practices, loss of budgets and everything else.

So since then, there was the Beps project launched, which delivered a first set of corporate tax reforms, which in Europe are the directives Atad 1 and Atad 2. And then there was a piece left which has evolved into pillar 1 and pillar 2, and more jargon, I’m afraid. It’s an OECD initiative, making quite profound changes to the corporate tax system, and Europe will, by way of directives, implement this. So there’s been a huge amount of corporate tax change already.

Now my point is, at some point, we have to stop obsessing about corporate tax reform because, to some extent, it’s kind of a law of diminishing returns. You can do more and more, and make things more and more complex, but at a certain point, you’re chasing smaller and smaller gains, first of all.

And secondly, you risk creating huge amounts of complexity for European businesses. And finally, through that complexity, you may ultimately end up countering your original objectives. I’d say, in a strange twist of irony, perhaps what really drove the start of the Beps project, to a large extent, was the US tax code. It was incredibly complex, and it created a huge number of loopholes. And as a result, these loopholes cascaded out indirectly all around the world. And big multinationals are taking advantage of them. So, basically, the whole world tax code has, you could say, gotten reformed largely because of the complexities of the US tax code.

We have to stop obsessing about corporate tax reform
Keith O’Donnell 

Keith O’Donnell Managing partnerAtoz

The point is [that] complexity in itself is bad, because it’s administratively expensive. But it also runs the risk of unexpected consequences. So that’s why I think at a certain point, you have to kind of stop.

The second point is, to focus on corporate tax return and corporate tax reform, at a certain point, I think, is unhealthy. You risk getting into a kind of tax populism. Why do I say that? First of all, because it’s easy as a populist thing to bash big companies and businesses generally, and say they’re all not paying their taxes, they’re not doing the right things. Firstly, because they don’t vote, so they’re kind of an easy target. But secondly, they’re kind of abstract. And the risk is that it distracts from more important reforms.

Corporate tax, OECD-wide, represents typically for any country about 7% to 8% of the tax base. In the budget of a country, this small slice is corporate tax.... If you look at the big problems of the world that can be solved by tax, and clearly everything can’t be solved by tax, but two of the biggest ones that can are the climate and inequality.... If you say you have these two other big problems that need solving and have a tax angle, we’re not spending that much attention on them.

[While the OECD has wrapped up its corporate tax reform efforts,] Europe has just pages and pages of proposals out there to reform corporate tax further. It just looks as if we’re going down the road of saying we’ll continuously try and tweak and change corporate tax in Europe with, I think, very diminishing returns. And secondly, there’s a distraction factor. If we want to do pan-European tax, should we really be focusing on doing more corporate tax--and that was my comment on the hamster wheel--or just running more and more?

For example?

One example of this was Atad 3 or the unshell directive, as it’s called. It was produced as kind of a reaction to things in the press. There wasn’t a lot of work done to think about if this is really necessary, given all the other tax reforms we’ve just talked about. But yet it’s felt this is politically necessary. We’re going to create a lot of complexity, create a huge, huge amount of work, probably deteriorate the competitive position of Europe to some extent compared to other regions, but ultimately create this huge distraction into something which the case hasn’t really been made that it is necessary yet.

In the press articles, there were a series of problems created by shell companies, without saying, given the tax reforms we’re in the process of doing, would these still even exist anymore? Would these tax problems with shell companies, as they call them, still exist? As a firm we analysed a number of the problems that the journalists highlighted; we’re basically saying these are solved. So why are we creating another directive to solve problems that have already been solved?

This has already been solved by the OECD Beps agreement?

Yes, as implemented into European law. So we had the OECD Beps project and then Europe implemented that, basically, through Atad 1 and Atad 2. Now we’re creating an Atad 3 on shell [companies]. My question is: is it necessary? And there’s even further things that the commission wants to do.

That’s where I get into this sense that we are on this hamster wheel, we’ll constantly keep doing tax reform because we think it’s the right thing to do, never stepping back and saying, does this actually help anything? Is it going to materially raise revenues? Doubtful. Is it going to fix inequalities? Doubtful as well. So to some extent, why are we pouring more and more energy into that? That’s the rabbit wheel analogy, it’s almost like a constant fixation. We’re still concerned there’s somebody out there hiding something. So we just keep doing more and more work on this, instead of maybe looking at more important things.

Are you happy with the OECD agreement as it’s been implemented through the various EU directives? Is it workable and realistic for you?

The first two, Atad 1 and Atad 2, there are still issues in them, but they reflect what was agreed at OECD level. It was a compromise. It wasn’t a bad compromise at all. Credit to the OECD for producing something. It was really like herding cats, but they managed it. So I think these directives make sense. They’re workable, even if they have some flaws in them, but hopefully, over time, they’ll get fixed, or we’ll get used to working with them. So up to then, fine.

Now we have pillar 1 and pillar 2, which, fundamentally, I think are a good idea. They can work. There’s still a lot of complexity in them which I would like to get clarified. But the fundamental idea behind it makes sense.

Pillar 1 is digital taxation, the concern around big digital companies not necessarily having a presence in the country but yet earning huge amounts of money there. That was partly perception, partly reality. Something needed to be done about that.

Pillar 2 is the minimum tax proposals, so saying big multinationals should pay at least 15% effective tax rates in any particular country. That was a huge shift. I mean, an absolute sea change in terms of policy. You know, I think it was the right thing to do. It hasn’t been welcomed universally. But I think in the context, it was the right thing to do. It was a compromise. It was a fair way of dealing with a problem that was a mixture of perception and reality.... ultimately, pillar 2 makes sense. Let’s say big multinationals have to pay a minimum tax of 15%. It stops too much gaming of the system and, provided it’s done in a way that’s reasonably clear, doesn’t create too much uncertainty. I’m fine with that.

There are very valid objectives to Atad 1 and 2, aren’t there? People rightly want companies to pay their fair share of tax. I assume you would agree?


In an ideal world, corporate taxes should be zero
Keith O’Donnell

Keith O’DonnellManaging partnerAtoz

What does a fair share mean for you? What’s a fair tax rate? What should they be paying?

Well, I’ll give you kind of my purest answer on this. I think the traditional taxation of businesses, so corporate income taxation--and I distinguish here, active businesses, so businesses that are providing services, creating goods--is fundamentally a bad idea. Economists will say it’s distortive, it’s inefficient. So at a fundamental level, I’d say, in an ideal world, corporate taxes should be zero.

And why do I say this? Not that I don’t believe those profits shouldn’t be subject to tax. I think taxing them in the company is the wrong way to tax them. In an ideal world, they’d be taxed at the level of the shareholders and investors that benefit from them. Now, that’s an ideal world. It’s very theoretical.

There is a reason why we tax profits at the level of corporates. One reason is fear of avoidance because if you don’t tax them, maybe the investors won’t be taxed on them, which is a very legitimate concern. The second one is that to the extent that a corporate is active in a country, it uses resources. Therefore, it should contribute to paying the costs of the country in the form of taxation. I think that those are both legitimate points. Now, I think there would be at least theoretically better ways to deal with them. But fundamentally, I get that it’s a purist view.

So that’s the kind of purist view which might help to understand why, when you asked me what’s a fair level of taxation for a company, I would personally be in favour of much lower than the average rates. I would prefer if there’s a minimum rate, there’s a maximum range as well. Because then I think companies, which are ultimately creators of richness, of wealth, of employment, of creativity, they would actually invest more back into their businesses. And I think it’s quite possible to say, when they distribute profits out to their investors, you tax their investors. But as long as they’re still reinvesting, then you want to keep their tax bill low.

There are tax systems where countries have chosen not to tax corporate profits until their distribution. Estonia has a system like that, for example. So they say, we’d rather corporates kept investing their profits. And we’ll encourage them to do it by saying we won’t tax them as long as they stay in the company. And once you then pay them out, yeah, then we’re going to tax them. And that gets you that incentive for a company to reinvest the money as opposed to saying, if you’re a successful company, to some extent, we’re going to penalise you for it.

So my theoretical answer would be: zero would be a great number. Now, my practical answer would be 15%. I think all corporates should pay tax at 15%. I preface that by saying that would be for companies that are actively producing goods or services. There are lots of companies that are much more passive, that are basically investment vehicles that just hold assets [such as real estate investment trusts]. I’ve no problem with somebody saying we’ll tax them at a higher rate.

So you would be in favour of higher dividend taxes? Would that be a good trade-off?

Yes, I think taxing dividends at a higher rate, whether it’s by withholding or on receipt, would be a good trade-off.

You talked about the complexity of the tax system. That’s a phrase that I’ve heard a lot from many people, but can you put a number on it or give an example of how much is this costing companies? Big firms will have a big tax and legal team anyway, right? I mean, is it really that much more work?

Well, there’s always going to be some work involved, because tax for big companies is going to be a big line in its expense budget, first of all. So they will look at it closely, and they will hire people to look at it. And secondly, they’ll probably have a lot of legal entities. So there’s a certain amount of inherent complexity in the system.

Now, having said that, if we take, for example, the Atad 3 proposal. There was an estimate out from the IMF. They said, for example, in Luxembourg, there might be about 40,000 entities which could be regarded as a special purpose vehicle. Let’s say they have to file a tax return each year, and we need to deal with this. And then our [Atoz] estimate is year one, it’s probably in the region of €5,000 a company. So you take 40,000 companies and multiply by €5,000, you see already we’re into the hundreds of millions.... you’ve really got to ask, could they be spending that money on something better?

Our firm and our businesses as tax advisors have been growing faster than the world economy for a number of years. To me, there’s something wrong there. We shouldn’t be growing faster in the world economy, we should be growing more slowly. Because the system should be getting more efficient at collecting taxes.

You shouldn’t need to hire an army of lawyers and advisors
Keith O’Donnell

Keith O’DonnellManaging partnerAtoz

People should be spending less on people like me, not more. If they’re spending more, that means the complexity is growing faster than the problem. And that, to me, is a terrible problem. If you take it to a ridiculous extent, in X years’ time, businesses like ours would be way oversized and will be a drag on the economy.

There’s a fine balance. We all understand that we need to pay a fair amount of tax. But at some point, you shouldn’t need to hire an army of lawyers and advisors to be able to figure it out. And that’s kind of where we’re heading at the moment.

Isn’t that good for your business and a lot of firms in Luxembourg?

As they say in French, I shouldn’t ‘cracher dans la soupe’ [bite the hand that feeds you]. You know, I could be saying, or I should be very happy, that the OECD and the European Commission are creating lots of work for me, but fundamentally, we’d rather advise clients that are being successful than benefit from an effective drag on the economy in the form of fees paid to us.

We’d much rather deal with successful clients that can be more dynamic, more entrepreneurial, because they deal with much simpler systems, as opposed to saying, let’s create very complex systems, and [then] have to pay people like us lots of money. I’d much prefer that Luxembourg and Europe had a very dynamic, successful economy.

We’ll still do fine. No need to worry about unemployed tax lawyers, you know, we’ll do just fine. [I’d prefer] if we had a successful economy and a smaller piece of it than a bigger piece of an economy that’s been slowed down.

During your UEL talk, you argued that the hamster wheel of tax reform could hit competitiveness in Europe, particularly with Asia and the US. The US tax system is known to be incredibly complex. Is there really a risk that the European tax system will be more complicated than the US? I suppose in some Asian countries the system might be a bit simpler, but I’m a little sceptical of the competitive risk.

You’re right to be sceptical. It’s really tough to model that because one of the difficulties with some of these proposals is that when people try to say which companies lose or which companies gain, it’s basically impossible to model... if you increase the tax in Country A, maybe Country B is going to increase its tax so that Country A doesn’t get as much. In the same way, when you’re trying to work out the knock-on effects of changing taxes in Europe, it’s quite difficult to be definitive about it.

However, having said that, fundamentally, as you make your country and your system more complex, more expensive, you either lose competitive advantage, or somebody else gains competitive advantage over you. So that’s why I’m always very determined to say, ‘before you add complexity to something, are you really sure it’s necessary?’

When we start to layer on too much complexity continuously, we will ultimately lose that competitiveness, and whether it’s versus the US, versus the UK, whether it’s versus Asia, I can’t quantify that, it’s too difficult.

Does Luxembourg need to compete on tax rates?

It’s, again, really hard to know, because lots of people have tried to work this out and I think there have been studies that have said Luxembourg is a net loser, some that Luxembourg is a net gainer. That’s just in budgetary terms. Do we need to be competitive? Yeah, I believe we do.

You know, if you look at Luxembourg’s competitive positioning, what it has is [an] economy that’s very open, very dynamic. I think it needs to continuously encourage businesses to be here. We’ve got to be realistic, we’re not a big country, we don’t have the critical mass of having [a large] internal market.

So we do need, in my view, to be competitive. Maybe we need to reduce corporate tax rates. It’d be necessary, first of all, to see how all this kind of stabilises. But ultimately, I wouldn’t rule that out. I think it would be something legitimate for Luxembourg to say we want to have a competitive corporate tax rate, because at the moment, we’re mid-tier at best. Maybe at some point, if we want to continue being a business-friendly country, one of the things we may need to do is cut the nominal corporate tax rate.

Quick glossary


The Organisation for Economic Cooperation and Development is a policy forum for 38 mostly wealthier countries.


The OECD’s base erosion and profit shifting project started in 2015, with the aim of closing loopholes that allowed multinational corporations to “exploit gaps and mismatches in tax rules to avoid paying tax.” Global rules were outlined last autumn. A total of 141 countries and territories, including Luxembourg, have signed on to the framework.


The EU’s anti tax avoidance directives aim to harmonise member states’ tax rules, including but not limited to provisions of the OECD’s Beps framework.

This article first appeared in the April 2022 edition of .