I won't try to summarize what Marty and Itai had to say, although frankly I'd be glad if their comments got wider distribution. (And not just because all publicity is good publicity.) But here is a written version of my remarks at the session, putting the book in context and offering a quick overview:
This book is in some ways a sequel to an earlier book that I wrote, called Decoding the U.S. Corporate Tax. Now, of course you know the iron law of movie sequels. With just one exception, they’re never as good as the original.
The exception isn’t The Godfather, Part 2, which is too slow-moving and self-important. It’s Gremlins 2, which was better than the original for a reason. It was more of a horror-comedy than the original Gremlins.
Fixing U.S. International Taxation isn’t comedy – for that, you’d have to try my novel, Getting It. But it does do one thing that’s different from its precursor.
Decoding mainly just summarizes the existing economics literature on corporate taxation, trying to make it more accessible. I got frustrated when I couldn’t find suitable accounts of the literature for my tax policy classes. Plus, I figured that, if I wrote the guide myself, I’d probably agree with most of it, even a couple of years later.
With Fixing, I initially wanted to do much the same thing for international tax. But there was a problem. I had to adjust for what I considered the sorry state of the literature. As I say in the book, despite more than 50 years of analysis by really good people, at some point it went badly off the rails.
I’m not alone in thinking this. For example, Michael Graetz gave a Tillinghast Lecture here some years back entitled “Taxing International Income: Inadequate Principles, Outdated Concepts, and Unsatisfactory Policies.”
So I concluded that I needed to reinvent the wheel a bit, albeit using familiar public economics ideas. And not to be grandiose, many others have been moving in the same direction as I have. But there was certainly nothing out there putting it all together in a way that I considered suitable.
Let me start with a simple illustration of the underlying problem. Say the U.S. corporate tax rate is 35%, while Germany’s is whatever. So the U.S. taxes all corporate income that’s earned here at 35% , whether it’s earned by a U.S. company, a German company, or anyone else.
If we want, we can also tax U.S. firms’ German income at 35%, or indeed at any other rate we like. But we simply can’t impose any tax on the German income of German firms – or on any other foreign source income of non-U.S. firms.
If only we could tax foreign companies’ foreign source income, that would actually be great for us. Foreigners who owned those companies would in effect be making large tax payments to the U.S. Treasury, which we would get to spend as we liked. And no one would be able to avoid the source-based U.S. tax by investing outside the U.S., rather than here.
But even Dick Cheney, or Vladimir Putin if he were American, would know better than to try taxing foreign firms’ foreign source income. So we face a constraint. There‘s a 2-by-2 grid yielding 4 boxes – the U.S. and foreign source income of U.S. and foreign firms. But the income in one of the boxes we simply can’t tax.
This brings us to one of the key points at which the international tax literature went off the rails. In tax policy thinking, we tend to like neutrality, because, while taxes are a private cost to the taxpayer, they aren’t a social cost. Rather, they’re a transfer from one person to others. If you have two or more choices and the tax system treats them neutrally, it can keep the choices’ relative payoffs in the same order, so it doesn’t induce inefficient behavioral changes that reflect treating taxes as a cost rather than a transfer.
But given the four boxes and the one we can’t reach, there are two different ways we could be neutral here between adjacent boxes: by equalizing U.S. versus foreign investment by U.S. firms, or U.S. versus foreign firms making foreign investments. We can’t be neutral at both margins.
A big mistake has been to assume we should pick one or the other neutrality, and push it all the way while ignoring the other margin. That’s generally not good thinking. When you can avoid making all-or-nothing choices, why wouldn’t you? Chocolate brownies or cheesecake? I think I’ll have a little of each.
For more of a formal economics example, say there are just 3 drinks – milk, juice, and water – and that we want to tax them in connection with raising revenue as efficiently as possible. Suppose that we simply can’t tax water, and that we’re taxing milk at a high rate, because it’s not very price-elastic – the people who like it will mainly keep on drinking it. Finally, suppose that if we raise the price of juice by taxing it, people will switch into water relatively fast. There’s actually a literature, called optimal commodity taxation or Ramsey pricing, which shows that we shouldn’t choose between milk-juice neutrality, by taxing juice at the milk rate, and water-juice neutrality, by not taxing juice at all. Instead, if we want to raise revenue as efficiently as possible, we should tax juice at an intermediate rate, somewhere between that for the other 2 commodities.
In terms of international tax policy, U.S. source income is milk, foreign firms’ foreign source income is water, and U.S. firms’ foreign source income is juice. We have more market power over investing in the U.S. than over using U.S. firms to invest abroad. Perhaps this helps explain why almost everyone agrees that we shouldn’t actually tax U.S. companies’ foreign source income at the U.S. effective rate. It also helps to show the likely undesirability of setting the U.S. tax rate for U.S. firms’ foreign source income as low as zero.
Let me turn to another big piece of the puzzle which I have ignored so far. When U.S. and German firms earn income in Germany, presumably they will face German taxes. How should we think about that, from a U.S. standpoint? Unfortunately, this is a second place in which the international tax policy literature has gone off the rails.
Suppose the U.S. and Germany both tax U.S. firms’ German source income at 35% (although Germany’s actual corporate rate is lower). And suppose both countries likewise tax the U.S. source income of German firms. Even with foreign taxes being deductible in the residence countries, cross-border investment would face a combined tax rate of almost 60%. Wouldn’t that unduly discourage it, probably to the detriment of people in both countries?
The answer is yes. But the literature made 2 main mistakes in analyzing this concern. First, it defined the problem as double taxation of cross-border investment, rather than as overly high taxation. Focusing on the number of taxes levied, rather than on the overall tax burden imposed, is unhelpful. Personally, I would rather be taxed twenty times at a 1% rate each time, than once at 50%.
Second, the analysis assumed too quickly that, in deciding what to do about the problem, U.S. and German taxes should automatically be viewed as equivalent. From a U.S. standpoint, they are not equivalent. We get the money from our own taxes, but not from theirs. And of course the Germans should think the same way from their perspective.
Now, there may be room for the U.S. and Germany to cooperate by reciprocally deferring to some of each other’s tax claims, to mutual advantage. But you have to actually look at the strategic interactions, not just assume that German and U.S. taxes are the same from a U.S. standpoint. And in fact today we provide a foreign tax credit for German taxes, but they don’t reciprocate. Instead, they generally exempt German companies’ U.S. income. This equally avoids “double taxation,” but it means that German firms in the U.S. will always want to minimize their U.S. taxes, whereas U.S. firms in Germany won’t care about their German taxes if they anticipate claiming immediate U.S. foreign tax credits.
This brings me to the broader problem of how the literature has analyzed the choices that a country has when deciding how to tax resident companies’ foreign source income. Supposedly, there are just 2 respectable options: exempting foreign source income, or having a worldwide system with foreign tax credits.
The problem, from the standpoint of clear thinking, is that these are compound choices. They differ at two margins, not just one.
First, exemption taxes U.S. companies’ foreign source income at a zero rate, while worldwide uses the full domestic statutory rate. Apparently, intermediate statutory rates for foreign source income are verboten.
Second, the two systems differ in their marginal reimbursement rates, or MRRs, for foreign taxes paid. Exemption has a zero MRR – paying foreign taxes has no effect on your U.S. tax liability. However, since this matches the approach’s zero marginal tax rate for foreign source income, exemption is what I call an implicit deductibility system for foreign taxes. Under exemption, U.S. taxpayers want to maximize their after-foreign-tax foreign source income, just as they would with foreign tax deductibility and any positive U.S. tax rate below 100%.
What about a typical worldwide system? In principle, foreign tax credits create a 100 percent MRR for foreign taxes paid on foreign source income. You get every penny back, barring the application of foreign tax credit limits. If this were the entire story, U.S. companies would generally be indifferent to whether the foreign taxes they paid were low or high.
As it happens, in practice U.S. companies are usually anything but indifferent to their foreign taxes. They want to minimize those taxes, just like our taxes, and they are very good at doing it. This reflects the real world consequences of another important U.S. international tax rule, called deferral, under which you don’t have to pay U.S. taxes on your foreign subsidiaries’ earnings until you’ve officially brought the money home, which you may never actually need to do.
In Fixing, I analyze how deferral and foreign tax credits interact to affect U.S. taxpayers’ incentives. In short, each rule has bad incentive effects when considered in isolation, but somewhat reduces the other’s bad effects.
Deferral encourages costly tax planning maneuvers, in lieu of just using internal funds efficiently. But if you have enough foreign tax credits, you can bring your foreign earnings home tax-free. So foreign tax credits can reduce deferral’s undesirable incentive effects.
On the other side of the ledger, consider again the effect of foreign tax credits in reducing cost-consciousness with respect to foreign taxes. If you anticipate either permanent deferral or a future reduction in the tax rate on repatriations, you may be highly foreign tax cost-conscious after all.
Still, a system with both rules is guaranteed to have high tax planning costs and other deadweight loss relative to the revenue actually raised. So having both deferral and foreign tax credits, as we do today, is definitely not a good answer.
The interaction between these two bad rules puts us in what I call an “iron box.” Scale back deferral and you worsen the problems caused by the foreign tax credit. Scale back foreign tax credits and you worsen deferral. Change either or both of them and you alter the overall effective U.S. tax rate on foreign source income – the merits of which have nothing to do with deferral and foreign tax credits in particular.
Let me just mention two more themes from the book. First, on a more theoretical note, I spend a lot of time devising what I’d like to hope were final burial rites for a set of concepts that have played much too large a role in international tax policy debate. These are single-bullet global or national welfare norms such as capital export neutrality, capital import neutrality, capital ownership neutrality, national neutrality, and national ownership neutrality. Each of these concepts has an acronym, such as CEN or CIN, so I deride using any of them as the “battle of the acronyms” or “alphabet soup.” Lawyers have tended to over-use these terms more than economists, but plenty in both groups have come to realize that they don’t help much. I just try to summarize and spell out a case against them that was already well-known in the best circles.
Second, the international tax issue that has rightly gotten the largest headlines in recent years is rampant base erosion and profit-shifting by numerous well-known companies. This reflects the tax planning opportunities that the companies have in the present environment, what with high global capital mobility and the huge economic role of intangibles. I don’t have as much that’s new to say about these problems as I do about the basic structural elements of our international tax system, which is why I don’t emphasize them more. But two points about them that I do make in Fixing are the following.
First, it’s well-known that we can’t actually get the source rules right, because source is not a well-defined economic idea. But we can try to make it costlier for multinationals, in a business and planning sense, to treat so much of their income as arising in tax havens. This is exactly how economic substance rules limit aggressive tax sheltering – for example, by requiring undesired downside economic risk as the price for successfully claiming huge artificial losses. In the international tax realm, this approach is likely to require paying far less heed than we currently do to the formal legal lines between commonly owned entities.
Second, all else equal, it’s actually a good thing, from the U.S. standpoint, if a U.S. company pays less foreign taxes, rather than more. Again, the foreign taxes are just a cost from our standpoint, because we don’t get the money. But the complicating problem, which helps to explain why exemption countries have anti-tax haven rules, is that, when we see significant income being attributed to a haven, and we know that it couldn’t actually have been earned there, we have reason to suspect that we, and not just foreign governments, may be among the losers. In effect, income’s being reported in a tax haven is a tag that indicates its potentially suspect status from the standpoint of protecting our domestic tax base.
A reasonable response to this concern may involve treating the foreign taxes that our companies pay as effectively better than just deductible, but worse than fully creditable. The Baucus international tax reform plan in Washington actually takes a stab at doing this, in an apparently treaty-compliant way.
There’s a lot more I could say – the book is about 200 pages long, after all – but instead why don’t I stop here, and listen with interest to what Marty and Itai have to say.