Tax policy colloquium, week 5: Reuven Avi-Yonah’s “Evaluating BEPS”

Yesterday Reuven Avi-Yonah presented the above paper, in front of a packed house.  My thoughts upon reading the paper fell into two main categories:

I.          OECD-BEPS

The paper lauds the aims of the OECD-BEPS process, but concludes that it fell short on both substance and process grounds.  I agree to a significant extent, but (unsurprisingly, given our respective track records) view the issues differently in some respects.

1) What important principles does OECD-BEPS endorse, and might it instead have endorsed?

Reuven has long advocated the primacy in international tax law of what he calls the “single tax principle,” holding that all income should be taxed exactly once, with the key aim being to avoid both double taxation and double non-taxation.  I question this principle’s value, at times a bit harshly, in my international tax book, although perhaps I treat it with a bit more sympathy in a recent short symposium piece.

What’s of interest in the context of the new paper is that it notes a distinct, although not necessarily irreconcilable, principle, that guided the OECD-BEPS process.  It aimed, not just at avoiding double non-taxation (although that was a major aim), but also at “better alignment of taxation with economic activity and value creation.”  For convenience, let’s call this the “alignment principle.”  It holds that each dollar of income not only should be taxed once, but should be taxed by the “right’ country, as defined in terms of where the economic activity and value creation occurred.

Also worth keeping in mind – but in tension with the alignment principle – would be the idea of defining the source of income on a destination basis (i.e., based on where the consumers are), rather than on the origin basis (based on where the production activity occurs).  Reuven has endorsed aspects of this in the past – for example, by suggesting that sales-based formulary apportionment be used in lieu of transfer pricing.  It would tend to source income differently than the OECD-BEPS alignment principle (which is most plausibly interpreted as an origin basis approach), but in both cases, if they were operating properly, source determinations would be less manipulable, and reported profits would be less prone to end up being assigned to tax havens, than under existing practice.

2) Can OECD-BEPS meet its objectives, and, if not, then why not?

Reuven is skeptical, as am I, but our grounds for this, while overlapping, are not identical.  In this regard, we both assign a lot of blame to OECD-BEPS’ retention of the “independent entity” approach, under which commonly owned affiliates are treated as wholly separate (reflecting legal form, rather than economic substance) even if they are engaged in a unitary business.   On the other hand, I am more skeptical than the paper is about the prospects for multilateral cooperation that centers on any consistent approach.

The paper also criticizes OECD-BEPS’ retention of the traditional “benefits principle,” which it notes has dominated international tax doctrine since the 1920s.  Under this principle, when a resident of one country has income in another country, source taxation has priority in the case of active business income, but residence-based taxation plays the lead role in the case of passive income.  More on this in part 2, below.  But I would note that if (as never was even remotely possible), the OECD-BEPS process had ended up endorsing unitary business accounting and sales-based formulary apportionment, then – whether this would have been good or bad on balance – it would have at least made plausible the argument that, assuming sufficiently widespread adoption, multinationals’ tax planning to put income in havens would have been significantly set back. 

Note, however, that since this would be using a destination-based rather than an origin-based approach to determining the source of income, it would have involved not following the “alignment principle.”  Again, however, if the reason for favoring the alignment principle was to reduce manipulability and the assignment of accounting profits to tax havens, this still might accomplish that goal, albeit different means.

3) How well can multilateralism work here?

The paper is far more optimistic about this than I am.  It basically says: given tax competition as a race to the bottom, no country has any real incentive to “defect,” and instead all should cooperate in finding a workable way of taxing multinationals effectively.

I’d offer two main points in response.  First, it’s widely disputed, both in politics and among academics, how one should define both national and global welfare with respect to tax policy towards multinationals.  Imagine trying to resolve a prisoner’s dilemma without clarity about how to measure the end-state payoffs to the players from different sets of choices.

Second, even when cooperation has clear global welfare benefits for all, and even when it’s not a true prisoner’s dilemma, in the sense that everyone can observe what everyone else is doing in real time, and adjust what they are doing accordingly, we don’t always get the cooperation that seemingly should be in everyone’s interest.

Consider carbon taxes and global warming.  Provided that you accept modern science, it’s clear that all countries “should” agree to a suitable carbon tax.  But we don’t exactly see that happening.  This is a classic prisoner’s dilemma – apart from the fact that everyone can see what everyone is doing (or rather not doing) – in that, say the U.S.’s unilateral incentive is to price carbon only at the marginal harm to us from climate change.  Why can’t it easily be solved?  Well, you tell me, but it’s even easier to explain why we don’t see a rush by nations to avoid mutually deleterious international tax competition by agreeing to a particular consistent global regime.

4) Is “constructive unilateralism” the solution?

The paper argues that the U.S. is still enough of a quasi-hegemon to overcome the obstacles to multilateral cooperation by going first (i.e., leading with its chin?).  The basic argument is that, if we reduce our corporate tax rate – say, to somewhere in the range of 20 to 25 percent – and also repeal deferral – so we have worldwide residence-based corporate taxation, with foreign tax credits but also immediate accrual – other countries will follow suit.  I am admittedly skeptical that this would be the upshot of our making this change.

A point of further interest is: What might be going on if, say, the U.S. adopts a particular rule and other countries are then observed adopting a similar rule.  Is this constructive unilateralism in practice?  Even assuming a causal link – i.e., it wasn’t either just coincidence or parallel behavior that perhaps reflected similar, but independently operating, causes – there are at least 3 different things that might be going on.  They have different predictive implications for a case such as the proposed one where we repeal deferral in the hope that others (despite having, in many cases, shifted recently towards the territorial pole) will follow suit.

a) Quasi-hegemony – This is the case where the U.S. gets something to stick out of raw power or influence.  FATCA has had elements of this – the threat of facing a withholding tax or being effectively shut out of market did indeed apparently influence lots of players.  But that was a very particular setting.

b) Stealing a good idea – When Apple makes a hit with the iPhone, its competitors say “Wow, maybe we should do something like that too.”  Indeed, but for patent protection they would likely do it straight out.

Consider the U.S. adoption of CFC rules (under subpart F) in 1962, which was followed by the spread of CFC rules elsewhere in the world.  As I discuss here, there are reasons of self-interest why countries may want to have CFC rules, as a way of protecting the domestic tax base.  I would think that this explanation better fits the actual spread of CFC rules than the hegemon theory.

While countries have reason to protect the tax base by adopting CFC rules, they also have reason for ambivalence about such rules.  Reason 1, they can discourage the use of resident companies to invest either at home or abroad.  Reason 2, they can discourage seeking to avoid foreign taxes, which is contrary to unilateral national self-interest unless there is sufficient tie-in to discouraging domestic base erosion.

The U.S. obviously feels this ambivalence as well.  Even if it was an accident that our check-the-box rules greatly weakened our CFC rules, we’ve deliberately allowed this “mistake” to remain in place for close to twenty years.  So the unilateral national self-interest reasons for having CFC rules, while strong enough to produce widespread adoption of such rules, evidently are not strong enough to consistently persuade either us or others to make the rules particularly tough.  (A worldwide regime without deferral would, of course, be the ultimate set of CFC rules – currently taxing all of one’s CFCs’ income.)

(c) Competing – The U.S. lowered its corporate tax rate in 1986, and many countries subsequently did the same thing.  Suppose we accept that our action in this regard influenced theirs.  The explanation might be, rather than emulation, affirmative tax competition.  Our lowering the rate, and thereby potentially attracting more inbound investment and reported profits (at least, leaving aside the base-broadening aspects of the 1986 change) might not only strike others as a good idea (as under (b) above) but might also make high rates costlier than previously to our competitors.  Under this scenario, our increasing the U.S. tax burden on U.S. multinationals’ foreign source income would not necessarily induce other countries to do the same with respect to their resident multinationals.

II.        REVERSING THE “BENEFITS PRINCIPLE”

Returning to the traditional “benefits principle,” recall the point that, as mentioned by the paper, it was long agreed that active business income should be taxed mainly on a source basis, and passive income mainly on a residence basis. The paper proposes reversing this, and taxing active income mainly on a residence basis, passive income mainly on a source basis. Let's start by looking at the traditional regime, and then at the new arguments here.

What did source country priority for active  business income actually mean, in an era when tax systems were mainly (at least nominally) worldwide rather than territorial?  More particularly, multinationals’ foreign business operations, even if ultimately subject to home country taxation, would immediately face source country taxation.  The home country would offer not only deferral (assuming the use of foreign subsidiaries), but also foreign tax credits.

Passive income, meanwhile, while potentially subject to source country withholding taxes that were creditable in the residence country, would often face withholding tax rates significantly below the taxpayers’ home country income tax rates.  Plus, treaties would typically arrange a reciprocal reduction or even elimination of the withholding tax for each others’ residents.

For active business income in particular, this approach was often rationalized on “benefits” grounds that I find unpersuasive.  First of all, benefit is not high on my list of normative tax principles to begin with.  Also, it seems to treat inbound active investment as if it were pollution, imposing costs on the source jurisdiction that therefore, naturally enough, demands recovery of those costs even if they have a public goods character and thus are not marginally linked.  But countries tend to want inbound investment, which they may view as having positive externalities associated with it, so demanding reimbursement under a benefits theory is a bit of an odd frame here.

A better historical argument for source priority – leaving aside that this simply happens to be what countries agreed to – was that the source country was often in a better position to observe and measure locally earned active business income.

There may also have been an elasticity issue at work here.  Even if companies’ residence was not, as such, highly elastic in the past, the question of which company (from which country) would make a given investment may have been susceptible to tax influence.  This factor is washed out, however, if all would face the same source-based tax.  Plus, arguably source was less elastic in the past, e.g., if high transportation and communications costs meant that one needed to produce locally.

Nowadays, the source of active business income is clearly mobile.  There’s tax competition, lower transportation and communications costs, greater importance of highly mobile intangible factors of production, highly advanced tax planning and profit-shifting technologies, etc.

Thus, there’s much to gain from taxing active business income on a residence basis, rather than a source basis, IF there aren’t similarly bad elasticity problems on that side of the ledger.  This is basically what the paper asserts.  The idea is as follows: all G-20 and EU countries agree to impose worldwide taxation on a residence basis, with a tax rate of at least 20 percent.  This would eliminate the relevance of corporate residence mobility except as to (a) tax rate differences within the permissible range (which therefore might push towards the bottom and (b) tax residence outside this set of countries.  The paper argues, however, that (b) is not a big problem if all of these countries agree to base corporate residence on a meaningful “headquarters” rule.

I am skeptical, however, about the prospects for such agreement, and I’m agnostic about the issue of residence migration outside the EU plus the G-20 under a meaningful headquarters rule.

Just as an aside, if the paper’s suggested approach were indeed adopted and succeeded, arguably there would still be effective source country priority on the taxation of active business income since, with foreign tax credits still being part of the regime, source countries would have every reason to impose taxes up to the typical residence company rate.

Okay, onto passive income.  There’s an ongoing debate regarding whether the main approach to addressing avoidance and evasion in this area should emphasize information reporting (a la FATCA), or the greater use of withholding taxes.  The paper urges the latter, on the ground that cooperation by tax havens can more readily be disposed of under a withholding tax approach than under an information reporting approach.  This is related to an argument in the paper that, because investors are risk-averse, even those in the BRIC countries want to invest predominantly in the US, the EU, and/or Japan.  This ostensibly means that only those 3 sets of actors need agree, in order for a withholding tax regime to be workable and acceptably comprehensive.  I did not entirely understand the argument, for two main reasons.  First, even the risk-averse can certainly earn, say, bank interest or dividends that comes from a Caymans entity and thus is treated as Caymans source (and hard to treat otherwise), even if in substance the underlying $$ are actually invested in the US, the EU, or Japan.  Second, I didn’t fully understand the paper’s distinction between the relative capacity of tax haven intermediaries to undermine withholding taxes, on the one hand, and information reporting, on the other.