A Contrarian Perspective On The Global Digital Services Tax Moratorium

There are two international events that I’ll be following this July. Both take place in France.

The Tour de France has been a midsummer treasure for more than 100 years. It’s a spectacle of human endurance — a bit like running a marathon a day, every day, for three consecutive weeks. The participants have every excuse to jump off their bicycles and quit out of sheer fatigue, yet something inspires them to keep pedaling over the next mountain.

The other struggle will not be televised, but it’s no less compelling in terms of international scope and the drama of overcoming immense obstacles. The Paris-based OECD is racing to finish the multilateral convention (MLC) necessary for implementation of the pillar 1 tax reform project.

If all goes according to schedule, we’re supposed to see the text of the MLC by the end of July — around the same time La Grande Boucle concludes on the Avenue des Champs-Élysées.

How is the perseverance of the OECD staff analogous to that of a weary Tour de France cyclist? Well, you must admire their determination to keep going forward. A less committed bunch would have thrown in the towel long ago.

We don’t know who will win the bicycle race, but we have a good idea what will be contained within the MLC. It will function as the delivery mechanism for unleashing amount A upon humanity, amending our vast tax treaty networks in the process. Amount A is a novel approach for dabbling in the realm of formulary apportionment without ditching the international consensus.

Amount A aims to reassign national taxing rights over a fraction of residual profits for in-scope companies. That’s a subset of a subset of taxable income, levied on a subset of corporate taxpayers. Market countries fume that amount A is too stingy, though it’s preferable to the limited tax base those jurisdictions are currently allowed. Residence countries complain that amount A is too generous. Both sides are correct in their own way.

Will the United States sign the MLC? That’s a sensitive question. It stands out that the United States previously declined to sign the multilateral instrument that accompanied the OECD’s issuance of the final base erosion and profit-shifting reports eight years ago. As falls the MLI, so falls the MLC?

Who knows? But you can understand why Treasury officials might be reluctant to sign the MLC at this particular moment.

A presidential election is just around the corner, in case you hadn’t noticed. The last thing the Biden administration wants is to be seen as forfeiting tax base (read: economic power) to the rest of the world in exchange for a bundle of vague and uncertain benefits.

From the perspective of the Biden campaign, the optimal time for the OECD to unveil the MLC would be sometime in December 2024 — just after the election. Second terms are prime time for complex multilateral projects that lack broad domestic support.

Yes, there’s an argument that amount A advances the U.S. national economic interest. No, it has not been especially persuasive. Amount A promotes global stability by deterring other countries from imposing digital services taxes that would primarily affect a handful of U.S. multinationals.

There’s your sales pitch: Stability is a normative good and Corporate America needs as much of it as we can scrounge. Uncle Sam must go along with pillar 1, or else a few big companies will be aggrieved by DSTs and we could end up with a trade war.

The success of that argument depends on the fear we associate with DSTs. As best I can tell, people in Washington aren’t so intimidated by them. Note that some U.S. states, led by Maryland, are considering (or have already enacted) their own versions of DSTs.

But for the Internet Tax Freedom Act, we might have 30 or 40 U.S. states with DSTs. It should tell us something that state lawmakers look upon foreign DSTs with a degree of envy.

I don’t wish to downplay the threat of global instability, but for many members of Congress the MLC will feel like an exercise in defeatism. If you squint your eyes, those foreign DSTs resemble tariffs. The U.S. Trade Representative knows how to deal with tariffs. The same goes for non-tariffs that cause comparable offense.

Does it matter that DSTs are not discriminatory in their design? Not really. They’re perceived as being discriminatory in effect, and that’s all that matters. Aggressive trade policy is not the exclusive domain of Republican administrations. Many of the tariffs adopted during the Trump administration are still in place under Biden.

That brings us to this week’s topic. On June 14 the International Chamber of Commerce (ICC) issued a letter to the OECD, urging consideration of an extension of the standstill period, during which no foreign country would impose DSTs. Consider the following proposition, taken from the ICC correspondence:

“Acknowledging the time needed for the ratification and implementation, we would invite the OECD/G20 Inclusive Framework to consider the extension of the so-called DST standstill agreement. This will prevent the enactment of new DSTs while the ratification and implementation processes are ongoing. We are indeed very concerned that the proliferation of DSTs will undermine the possibility to reach an agreement at the OECD/G20 Inclusive Framework, increasing existing tax and trade tensions, while creating further uncertainty and instability for all industries, as businesses become more digitalized.”

The DST moratorium has a built-in expiration date. Under the terms of the inclusive framework’s October 2021 agreement, the episode of coordinated self-restraint runs from October 8, 2021, until the earlier of December 31, 2023, or the coming into force of the MLC.

Because the MLC will not enter into force before the conclusion of the calendar year, we need to think about what’s likely to occur on January 1, 2024. Will dozens of DSTs spring to life — bedeviling the likes of Facebook and Google? That’s what the ICC is trying to prevent, and you can see its point.

The group asks that we should interpret the text of the inclusive framework’s October 2021 statement to omit the reference to December 31, 2023. On the face of it, the request doesn’t seem unreasonable. The tax world is so close to reaching an anticipated moment of climax, why allow the onset of DSTs to spoil the progress?

Here’s the part of the article in which I write unpopular things. Unlike the ICC, I think it could be rather useful if all the countries with DSTs decided to apply them, in full, commencing with the first stroke of New Year’s Day.

Indulge me here. What follows are three reasons why the global DST moratorium should not be extended by a single day.

Reason 1: Loser Spotting

Consider the disconnect between the pool of U.S. businesses deemed to be in-scope for purposes of pillar 1, and the narrower pool of U.S. businesses that would be on the receiving end of DST assessments. Not enough is made of this mismatch.

The CEO of any business that isn’t subject to DST, but expects to be subject to pillar 1, would conclude that the inclusive framework is making these firms pay for somebody else’s problem.

What’s it to the nondigital sector (for example, General Motors or General Electric) if a handful of digital behemoths like Facebook and Google are forced to pay DST in some foreign countries? This scoping mismatch was always there, but we glossed over it by telling ourselves that every business is a digital enterprise.

How many times have you heard somebody declare that the digital economy is the economy? In hindsight, statements like that seem disingenuous and designed to distract attention away from the mismatch between payers and beneficiaries.

What if the shoe were on the other foot? Should Facebook or Google care whether a cost component unique to GM or GE suddenly increased as of January 2024? Would Mark Zuckerberg lose a moment of sleep if the unit cost of automobile airbags went up 3%? I think not.

Someone once remarked that the needs of the many outweigh the needs of the few. A problematic attribute of pillar 1 is how it flips that wisdom on its head, such that the needs of the few (and the highly profitable) seem to outweigh the needs of everyone else.

The swift onset of DSTs from January 2024 would force a concentrated discussion about the misalignment between the losers under pillar 1 and the winners under the accompanying DST standstill and rollback commitments. That’s a healthy discussion to have.

Reason 2: Burden Bearing

The onset of DSTs from January 2024 would allow affected companies to act on anticipated pricing adjustments. We accept that every firm affected by a DST will try to pass the tax burden on to someone else. That’s equally true for other taxes. Nobody absorbs an economic cost if they don’t have to.

I’ve enjoyed reading other commentators who argue that the likes of Facebook and Google will be hard-pressed to recoup their costs by raising the fees they charge to advertisers. Part of that is a reflexive knee-jerk reaction.

Mention any tax, and you’ll find a cluster of economists lamenting that some portion of it will come to rest on the owners of capital — and that is the worst thing ever because the behavioral response is to discourage further capital investment. You need not be dismissive of those claims, generally, to be skeptical of them for DSTs.

One of the themes that comes up in this context is the difficulty that businesses encounter when trying to fully recover the consumption taxes charged on their inputs. We take for granted that the imposition of tax on business inputs is problematic for retail sales taxes, but RSTs are crude instruments by that measure.

VAT, however, is supposed to be different. The entire burden is supposed to be passed on to the final consumer. An established body of research shows that some portion of VAT charged on business inputs isn’t fully recouped. This is a glitch that VAT policymakers are struggling with, although the magnitude of the problem is debatable.

The suggested parallel is that a similar fate awaits U.S. businesses as to unrecoverable foreign DST charges. That is, Facebook and Google will never be able to fully recover their DST assessments by proportionally increasing the fees they charge advertisers in the relevant foreign markets. The argument goes that proportional pricing adjustment won’t do the trick for DSTs, because they don’t work perfectly for VAT.

Putting aside that many proponents of DSTs wish for that to be the case (for example, some advocates of the French DST actively want the burden to fall on nonresident shareholders, because only then does it effectively export the tax burden), there’s little reason to think things will pan out that way. There are fundamental differences between VAT and DSTs in terms of how the taxes are charged and internally credited.

There’s a multiplicity to VAT in that it comes at a business from every angle. A large business might have many thousands of inputs from hundreds of different vendors. It can be difficult to track how much accrued VAT a business has paid over a relevant accounting period.

In contrast, there’s a singularity to a DST in that it comes from a single source. The task of recouping an isolated tax charge through pricing adjustments is quite different from the situation you have with VAT.

Naturally, if the cost of digital advertising becomes too extreme, advertisers will be tempted to seek less costly ways to reach their audience. That’s true with any price increase.

The real concern, then, isn’t that outfits like Google and Facebook can’t effectively pass the DST burden on to advertisers, but that doing so might influence demand for their platforms. Then again, we’re talking about a sector of the economy that routinely experiences excess profits and that has greatly benefited from lax antitrust enforcement.

How competitive is the marketplace for online search engines, or for social media platforms on which you can upload your kid’s graduation photographs? What’s the price elasticity for digital advertisements placed on those websites? I don’t know, but it would be nice to find out.

Going live with DSTs as of January 2024 would allow us to field-test the efficacy of pricing adjustments. We might learn that the burden of foreign DSTs settles mostly on foreign consumers, negating the desirability of the tax to those who assumed it would fall on nonresident 1-percenters.

Reason 3: Disrespect

Finally, there’s a practical concern with ignoring the expiration date contained in the inclusive framework’s October 2021 statement. What do you replace it with?

Should the OECD announce that the standstill period must remain in place for one additional year, extending through the end of 2024? That would get the project past the next U.S. presidential election.

What if another 12 months isn’t enough time for every participating government to ratify the MLC? These matters take time, you know. How do we feel about a two-year extension, running through the end of 2025? Could I possibly sell you on a five-year extension?

The temptation, here, is to do away with a date-specific expiration. That would have the OECD retain the standstill requirement until the MLC is widely ratified by a critical mass of nations, defined to include the United States. Isn’t that essentially what the ICC is asking for? Let’s hold off on DSTs until the Americans have made up their minds.

I shouldn’t need to explain the problem with that approach. You’re waiting for Godot, but Godot never arrives. Keeping DSTs switched off until Congress ratifies the MLC — however long that takes — would see other countries’ tax laws assigned to indefinite limbo. It disrespects their political sovereignty.

What nation would enact a tax law, only to delegate the effective date to the unchecked discretion of foreign lawmakers? I can’t see how an open-ended extension of the DST moratorium period could possibly fly with the other members of the inclusive framework.

Après Moi, Le Déluge . . . or Not

I welcome the ICC letter because it gives voice to what everybody is thinking: What the heck is going to happen with all these DSTs after December 31 — which isn’t so far off? Taxpayers deserve an answer. The next move belongs to the OECD.

In contrarian fashion, I’ve spoken my piece that the global DST moratorium should not be extended. That said, I suspect a one-year extension running through December 31, 2024, is the path of least resistance — and the choice most conducive to salvaging the entirety of pillar 1.

OECD officials can rationalize the delay without much fuss. It acknowledges the practical difficulties of ratifying an instrument as complicated as the MLC. The display of patience affords further opportunities for constructive input from diverse stakeholders. But really it’s just to get us past the election.

Think of it this way. If Biden wins the election, a Treasury official can sign the MLC in early 2025 and our thoughts will turn to the ratification process. If Biden loses . . . well, we might have other things on our mind.