Globalization has generally been an excellent thing, but one downside is that it facilitates tax avoidance. Multinational firms, facing inconsistent tax rates and rules around the world, structure their operations in ways that channel earnings out of the typically high-tax countries where the value is created. Executives say, quite fairly, that they’re simply responding to the incentives that legislators put in place — as shareholders expect them to do — but the effect has been to siphon trillions of dollars out of public coffers.
How companies accomplish this can be dizzyingly complex. But fundamentally, income shifting centers on the use of foreign subsidiaries. By setting up transactions between related companies and calibrating prices just right, savvy firms can make profit margins vanish in one place and reappear in friendlier jurisdictions.
Stanford accounting professor Lisa De Simone has devoted her research to advancing our understanding of this issue: how income shifting works, what causes it, how prevalent it is — and what, if anything, policymakers can do about it. She has a unique perspective: Before donning an academic gown, De Simone devised tax-mitigation strategies as a consultant for Ernst & Young.
In a pair of new studies, she argues that income shifting is more pervasive than we think — and companies are only getting better at it. We sat down with her to learn more.
So you were once a practitioner of this rarefied art?
Yep, I helped clients, mostly U.S. multinationals, work out the transfer pricing on intercompany transactions — which now, on this side of the desk, I call income shifting. I also helped put together some of these structures you hear about, like the “double Irish/Dutch sandwich.” That’s where, say, royalties for the use of intellectual property are paid by an Irish subsidiary to another subsidiary in the Netherlands, which then pays a second Irish unit, which is tax-resident in the Caribbean. It’s complicated.
That’s a structure the big tech firms favored. It’s been eliminated now, though, hasn’t it?
They’re phasing it out. The EU pressured Ireland to get rid of the quirky residency rule that gave rise to it. Of course, Ireland doesn’t want to lose its advantage, so now it says it’s going to further reduce its already low tax rate. But the latest thing seems to be going through the Netherlands and Luxembourg; apparently some companies have gotten private deals with the governments, although those are also now under fire.
Could you explain what transfer pricing is and how firms use it to reduce their tax liability?
Sure. Say you’re a U.S. multinational and you’re supplying parts to an Irish subsidiary. Ultimately it’s an internal transfer, but you need to stipulate a price. It would be the same between two subsidiaries; transfer prices are just prices paid by related companies. Since they’re not set by competitive negotiation or an open market like normal prices, you need to decide what they should be.
That sounds harmless enough.
Companies have to set transfer prices. It irks me when I see it portrayed as this nefarious activity. But now bring in the fact that tax rates are much lower in Ireland: Now you have an incentive to report earnings there instead of in the U.S., and you can accomplish that by setting a low transfer price. It inflates your profits abroad and deflates them at home. So then transfer pricing becomes part of your tax strategy. Corporations respond to the incentives that governments create.
But firms can’t just set any price they want, can they?
No, there’s an “arm’s length” standard, meaning the price should be the same as if the parties were unrelated. But it’s so hard to enforce. If you’re providing a unique technology or management services to an affiliate, there is no market price. So often what you do is find a “comparable” company in the same business, look at its profit rate, and set a price that yields the same margin. That counts as compliance. Of course you’re going to look at all the possible benchmark firms and cherry-pick the ones that give you the best prices.
So the key is what counts as “comparable,” right? You have a new study on that very issue.
Right. It used to be that if you had, say, a Spanish affiliate, you had to use a Spanish company as a benchmark, because accounting standards varied from country to country. You couldn’t compare book profits across borders; it was all apples and oranges. But there’s been a big push to harmonize financial reporting, especially in Europe — which is great! It certainly facilitates international investment. But it also expands the set of comparables for transfer pricing.
By quite a bit if you’re moving from Spain to the entire EU.
Exactly. Instead of a dozen candidates, maybe you now have 200, and that gives you a wider range of prices to choose from. My hunch, based on my own experience, was that it would lead to more income shifting. So in this new paper, we looked at what happened in European countries that adopted the International Financial Reporting Standards for subsidiaries, and sure enough, it increased by 11%. That translates to a lot of money. It illustrates the kind of subtle interactions you can get between accounting and tax systems. There can really be unintended consequences.
You argue that researchers have underestimated the amount of income shifting that goes on.
These studies typically assume it’s all driven by statutory tax rates, but there’s more to it. For example, I have another new paper that looks at what happens when firms have an unprofitable affiliate. Even if it’s in a high-tax country, it effectively has a zero marginal tax rate, and you can take advantage of that by adjusting your transfer prices — reducing the amount of income shifted out of that company or even shifting income into it. We call that a shift-to-loss strategy.
Even better than a tax haven.
And it doesn’t look as bad! Anyway, that’s one piece of the puzzle that’s been left out. So not only do those studies miss that shift in an unexpected direction, the shift they do measure, from high-tax to low-tax countries, seems smaller because it’s really a net effect. When we included unprofitable affiliates in our analysis, we found that the amount of income shifted in response to a change in tax incentive — what we call the tax elasticity — nearly doubled.
Wow.
Yeah. It means prior studies may have understated income shifting by as much as 50%. In our European sample, instead of $54 billion being moved, we find that it’s more like $99 billion. At the very least, the estimates in prior studies should be considered a lower bound. Also, the elasticity is important for policy, because it tells us that a change in the tax code could have a bigger effect on the tax base than previously thought. Say, for example, a country raises its tax rate, hoping to generate more revenue; you could get enough outward income shifting that tax revenue actually goes down. In a global economy, the corporate tax base is a very leaky vessel!
Is there anything we can do to plug the leaks?
We can’t stop income shifting. Until we have some broad tax reform in the U.S. and change the incentives, it’ll continue. And it’s distorting business decisions. All that cash is sitting idle overseas, and companies can’t bring it home to reinvest or pay dividends unless they pay the taxes they already tried to avoid. Some are clearly banking on the hope that the government will give in and declare a tax repatriation holiday. But then it would start all over again.
How has your perspective changed since you left industry? Is there anything you’ve learned in your research that’s surprised you?
I worked for some of the largest multinationals, so I knew all the tricks, but I wasn’t sure how much of this was going on in the broader economy. At the time there were plenty of midsize firms that didn’t even know what transfer pricing was, and we helped some just to become compliant. But companies were learning by watching, and more and more of them began using their foreign subsidiaries as part of proactive tax planning strategies. So I’ve been trying to find out, is this how the average business operates now? And the answer, for better or worse, appears to be yes.
