Making Sense of Profit Shifting: Scott Dyreng
May 21, 2015
Scott Dyreng is Associate Professor of Accounting at Duke University’s Fuqua School of Business.
Professor Dyreng’s scholarship focuses on corporate tax avoidance, international taxation, accounting for income taxes, and profit shifting. His novel research related to various aspects of corporate tax avoidance has been widely cited and published in leading journals in the fields of accounting and finance.
Professor Dyreng received his Ph.D. at the University of North Carolina at Chapel Hill.
In this interview with the Tax Foundation, Professor Dyreng shares his latest insights on the profit shifting phenomenon—specifically highlighting the difficulties associated with measuring and understanding the consequences of profit shifting, the role of data in better analyzing profit shifting, and why a high level of multilateral cooperation might be necessary to materially address profit shifting. This interview is part of our 2015 Tax Foundation Forum series and has been edited for length and clarity.
Tax Foundation: What is known about profit shifting?
Scott Dyreng: The first thing we ought to do is to define what we mean by profit shifting, because I think that means different things to different people. A very broad definition would be that profit shifting is strategic actions taken by firms that result in profits being reported in a tax favorable jurisdiction. And that could really be anything. It could be a transfer price that's set; it could be intra-company debt that’s used; it could be the movement of a plant to a different country. Some people don't like that broad of a definition. They would prefer to have maybe a narrower definition, which would be something like: Strategic actions that are taken that result in profit being reported in a jurisdiction that's different from the one where the economic value is added. So in a situation like that, if you move a plant to a low-tax country, then that's not profit shifting. You've actually moved real operations. So different people think of profit shifting in different ways.
But generally, what's known about profit shifting is that it happens. I think that's pretty well- documented. And most people would agree that, on average, in a large cross-section of firms, if there are incentives to strategically report profits in a way that minimizes tax liabilities, then some firms will do that.
TF: If you would compare what we know about profit shifting with what we don't know, is the realm of the unknown greater than the known at this point?
Dyreng: I think so because we know that profit shifting occurs, but we don't have a very good feel for the magnitude. And we also don't have a very good feel for, at least I don’t think, the consequences of profit shifting or even for the determinants of profit shifting. So in most of our research, we might see that an average firm shifts profits. But what that probably means is that some firms engage in a lot of profit shifting and other firms don't engage in any.
And one question that, at least to me is unanswered, is why two different firms with maybe similar incentives to shift profits don’t behave in the same way.
We suspect that when firms engage in profit shifting they incur some non-tax cost that at some point might offset the benefits of profit shifting. Essentially, when you start tax planning like this, there will be what you could think of as a deadweight loss in the system: Firms aren’t behaving as efficiently as they could. And so the consequences of profit shifting could be under‐investment, overinvestment, or inefficient investments. But these consequences, I don’t believe, are very well understood in the academic literature.
What are the unique factors or drivers that are underlying the profit shifting phenomenon?
I don't think anybody knows. People have speculated. People have speculated that one of the important things is that you need to have some sort of intellectual property because intellectual property enables firms in some sense to strategically set transfer prices or use cost sharing agreements to shift profits. It's a little harder to do that with physical assets or with products that are commodities because the arm’s length transfer price that you might expect to pay is more easily enforced. So I think intangible assets or intellectual property probably matters.
But there are also things that are deeper than that, which might simply be something like the characteristics of the tax director or the characteristics of the CFO, that probably very likely impact the decisions that are made in terms of how much profit shifting to engage in. And there is a lot of research that suggests that personality characteristics of managers have a pretty strong influence on the decisions that they make. And I suspect that holds true in profit shifting, although I don't know of a study that documents that explicitly.
See this 2010 paper by Professor Dyreng for a detailed discussion of how manager characteristics may affect corporate tax avoidance.
How about debt versus intellectual property as a channel for profit shifting?
I definitely think that the debt channel matters. So it's just another channel. There are various channels you can use, and the intellectual property one is really more important for cost sharing agreements and for transfer prices.
When you say ”debt,” you are talking about intra-company debt, and using that to strategically locate the interest deductions in a high-tax country and the interest receipts in a low-tax country. If you have more hard physical assets, then you might be able to strategically use debt, whereas you probably don't have as many opportunities to use your intellectual property.
So I think that debt could matter quite a lot. But we don't know to what extent. As a profession, I don't think we know which of those channels is responsible for what fraction of total profit shifting. There has been research that has documented the channels that are used, whether it's royalty payments or intra-company debt or something like that, but I don't think we understand the relative magnitude of their importance.
Would you say that perhaps the underlying driver of profit shifting is differences in countries’ tax law?
Well, for sure. The thought experiment to run there would be: Give every country exactly the same tax rules and there is no benefit to shifting income for tax purposes.
So broadly speaking, one of the major ways that you can reduce your tax bill is to shift income from one jurisdiction to another. But the benefit in doing that is that the different jurisdictions treat the same income differently. So you might be able to get favorable tax treatment in one jurisdiction or another. But if they all had, in the thought experiment, exactly the same tax rules, or even simpler, if there were no tax, then there would be no tax advantage to profit shifting.
Now, there might be other reasons why people shift profits. So, for example, you might not want your profits to be retained in a country where there is weak corporate governance, because you might be concerned that the dictator in that country will come in and just take over the company. So you want to get your cash out; you want to get your assets out; you want to get everything you can out. That has nothing to do with taxes. It has to do with protecting your interest there. So it's not like taxes are the only reason we would see profits moved around. But from a tax point of view, it certainly has to do with the variation in the tax systems around the world.
How does profit shifting alter the behavior of firms?
I don't think we know very much about that. We think that it might do things like change the location of their investments. So, for example, I am pretty sure, but I can't prove this, that there is more manufacturing of, say, pharmaceutical products in Ireland than there would be had Ireland had the exact same tax rate as the U.S. or the UK. Would we see firms locating manufacturing of pharmaceuticals or whatever it is in Ireland had it not been for the tax rate? There would be some, but probably not as much. But, again, I think the literature is pretty sparse on the consequences of income shifting. It's hard to measure which firm has shifted income, and therefore it's hard to study the consequences of shifted income.
So at this point there is no conclusive evidence on the consequences of profit shifting?
Not that I am aware of. One way to respond to that is: Is there ever conclusive evidence? But the preponderance of evidence, at some point in a literature, becomes so overwhelming that we start to be fairly confident that something is happening.
I think that there is enough evidence to suggest that income shifting happens. But I don't think there is enough evidence to understand the economic consequences, either at a micro-level or a macro-level. So we don't really know how income shifting affects economy-wide investment in the U.S. There are plausible arguments that one could make that would say that it increases U.S. investment, and there are plausible arguments that one can make that would say it decreases U.S. investment. So I don't think there is definitive evidence there.
Does profit shifting matter?
One way to think about it is that almost everybody would agree that introducing any tax in to an economic system will result in a loss of efficiency. And so profit shifting is in my mind a manifestation of that inefficiency that's been introduced in to the system. And so does profit shifting matter? Almost certainly profit shifting is a symptom of an inefficient system. But on the other hand, can we ever get a perfectly efficient system? No, almost by definition, if you want to tax any corporation you're going to have distortions that result because taxation is almost by definition going to introduce inefficiency in to the system. So in that sense, this all matters.
The question to me, though, is not: Does it matter? It's: How does it matter or what does it affect? And that I think is something that is not very well understood. And to be clear, the reason that this is not well understood is because it's very hard to measure income shifting. And the reason it's hard to measure income shifting is because, by definition, you cannot observe what income would have been had there been no shifting. So just to measure income shifting you have to measure something that is unobservable, and the only way to do that is to use some sort of statistical estimation, which turns out to be a hard thing to do. And so if you don't have a good measure of something, then it's hard to study its consequences. So I think that's broadly speaking why we don't have a very solid understanding of what's happening.
Has profit shifting increased over time?
I only know of one study that looks at that specific question, and that study is by Klassen and Laplante, published in the Journal of Accounting Research. They have evidence that it has indeed increased over time, and that is consistent with what most peoples' intuition. Many profit shifting techniques rely on the ability to use unobservable values of intellectual property, in a transfer pricing strategy, or in a cost sharing agreement or something similar. Because firms have become more and more intangible intense and global, most people believe there is more and more opportunity for income shifting to occur. So my guess is that income shifting has increased over time, but based on one study that I have seen. There could be more studies, but I am not familiar with them all. I believe the evidence would say that it has increased over time. And that would make sense if you believe the story that I told you that it has to do with intellectual property in some sense.
Let’s discuss your and Kevin Markle’s research in relation to the magnitude of profit shifting. What’s the best estimate of profit shifting out of the U.S.?
If you believe our model, which I would be the first to admit is an estimate—it's not perfect—but our estimate would say that somewhere in the neighborhood of 8 to 11 percent of what I would call pre-shifted income ends up being shifted out [of the U.S.]. So pre-shifted income would be income that would have been reported in the U.S., had there been no incentive to shift.
What’s missing for a better understanding of profit shifting?
Profit shifting is hard to study because, by definition, you have to compare what was reported to something that is unobservable, because the unobservable is the amount of income that would have been reported had there not been profit shifting. So we're then forced to use models to estimate what the income would have been in the absence of profit shifting.
And to do that better, as a researcher, what would help me? Better data would help a lot. So more detailed information on the financial state of firms in their foreign subsidiaries or foreign operations, for each country they operate in.
So there's been some movement towards countryby-country reporting. And as a researcher, that data would be very valuable. But I am not sure that the benefits of having that data from a societal point of view would outweigh the cost. I think it's possible that the costs of producing that data may outweigh the benefits.
Can you expand on what you mean by costs?
So to generate a report country-by-country requires a lot of interesting assumptions. For example, how do you assign profits to a country? Is it based on the assets that are in that country? Is it based on the sales that were in that country? Is it based on the market cap of your firm somehow? Or is it based on the number of employees in that country? It's a little hard to decide how a firm should report or aggregate based on country boundaries when the economics of the firm don't necessarily operate under the constraint of a country boundary.
So financial accounting is designed to capture, at least in theory and at a high level, the economics of the firm. But the economics of the firm are not necessarily constrained by country boundaries, so determining how to account for financial results when imposing a constraint that does not necessarily correspond to the economics of the firm becomes a little tricky. And it's possible that if country-by-country reporting were required it would be really hard to standardize across firms because the economics of firms across countries is probably quite idiosyncratic.
So could it be done? Yes, I think it could be done. Do firms have fine enough systems to figure out what's going on in their different divisions? I believe that they do for the most part. But would it be a meaningful measure to investors? I don't know. It probably depends on how it was implemented.
How about models? Where are we in terms of model development of empirical models and how much better can they become?
Well, that's a really good question. In some sense an empirical model is only as good as the data you can put in to it. So even if you have an empirical model that seems in theory like it should work really well, if you put data in to it that's not very reliable, then the estimates from that empirical model will also not be very reliable. With that being said, that's part of the name of the game of being a researcher, to use the tools available to the researcher to improve the model. So do I believe there will be improvement? Yes, I do. Do I believe that it'll be dramatic? I am not sure that it will be dramatic because I think that no matter how good the model gets it's still going to be constrained to some extent by the quality of the data that goes in to the model. And I don't think that the quality of the data that goes in to the models is super high in many cases.
What are possible solutions to address profit shifting?
I think that it depends on what you are trying to accomplish. If you are trying to keep U.S. firms from shifting earnings to foreign countries, then one simple way to do that would just be to lower the U.S. tax rate really low, because then firms would want to shift in to the U.S. But that's not probably a very practical solution.
Theoretically, I think the only way to stop profit shifting would be to have benevolent cooperation among essentially all governments. But that's not going to happen either. So then it becomes a piecemeal approach, where governments try to stop on their own what they can. And as a result, I don't think that there is going to be a lot that takes place.
So can we plug some holes? Sure we can. We can change the way we enforce transfer pricing or we can change the rules on intra-company debt. But it's a little bit like that game Whac-A-Mole. Whack one down and another one pops up. And as long as there is an incentive for firms to shift profits to some other jurisdiction that has a lower tax rate, then it'll probably continue to happen. So I don't think that piecemeal effect is going to be highly effective. I just think that firms will find a way to work around whatever rules are put in place. And the only way to make it so that they can't do that would be to get a high level of cooperation among essentially all countries, or at least major countries, otherwise I think it's going to be just more of the same.
An interesting example is the U.S., where all the U.S. states have their own tax system, and nationwide there’s a formulary apportionment system. But as long as you have a state like Nevada or Delaware, that is in some sense a "defector" where they allow no corporate tax or no corporate tax on certain types of income, then you'll find companies that can exploit that variation. And one way to solve all of that problem among U.S. states would be to get all of the U.S. states to agree to have basically complete coordination and give up their autonomy. But even within the U.S., states are not willing to do that.
The point is, if you have sovereign countries, it just seems unlikely to me that you are ever going to get all of the countries to sit down and agree in a benevolent way to all act in harmony and unison. Because once you start to do that, then it really increases the incentive for some other country to be a defector and to break away from the game.
But a more important question to me is: When we tax corporations, who are we really taxing? And is that who we really want to tax? Because somebody is bearing the burden of that tax, whether it's the customers, employees, or the shareholders, and I don't even think we understand that very well. Just understanding that would probably help guide our policy decisions a little bit more.
How would you compare a worldwide versus a territorial system in addressing profit shifting?
There is a little bit of evidence that would suggest that profit shifting might increase a little bit if you are in a territorial system versus a worldwide system. And the evidence that I've seen would suggest that's probably true.
But I think more important than that is to understand what other consequences there might be, and there is at least some argument to suggest that domestic investment might increase if we were in a territorial system because firms wouldn’t end up with their cash trapped abroad. They may shift more income out but then they might invest more here, which might offset the shifted income that was moving out. It's not really clear. Some people would argue, and I even argued this in a paper, that firms that have no financial constraints and can borrow very cheaply are essentially behaving right now as if we are in a territorial system, because they are shifting out every penny that they can possibly shift out. If they need cash at home, they can just borrow, and I think some firms have done things similar to that.
There are other firms who can't borrow so cheaply because they have too much debt or something like that, and for those firms a territorial system would make income shifting more profitable because they could shift the income out and then bring it right back and not worry about incurring an incremental U.S. tax. But if they were to do that under the worldwide system, they basically don't save any tax at all. And they just incur the burden of shifting out in the first place. So a territorial system has benefits and I think it has costs. I think those benefits and costs are somewhat understood, but I don't think they are fully understood.
And another thing to think about is that there is probably no pure territorial system, nor is there really a pure worldwide system anywhere in the world. It really depends on how it's implemented. And the devil is always in the details.
What major developments on profit shifting do you expect in the next 1-2 years?
Not much. I think the BEPS project is interesting. And it's fun to think about from an academic point of view. But I don't think that it's going to have a massive influence on policy because I don't think that the sovereign governments are going to have much of an incentive to adhere to whatever suggestions that the BEPS project comes out with.
However, I think that the models are going to improve over time—that usually seems to happen. And I think that people will be creative and figure out ways to test the consequences of income shifting and the causes possibly. So I believe there will be progress made there. But it's going to be a little tricky with the data that we have. But people get creative and that's the name of the game.
Which resources would you point to for someone who would like to learn more about profit shifting?
There are a lot of very thoughtful commentaries on this topic, and I do think it probably would be very valuable to read beyond what is written in the popular press because the popular press is not always the most objective in terms of how the academic research is presented. Regardless of what your political leanings might be, sometimes I think going to the academic literature, where you can see more carefully thought out arguments highlighted, can be valuable.