The OECD's Pillar One Dilemma: To Be or Not B2B

As I’ve already admitted, I’ve been slacking off on studying Pillar One of the Organization for Economic Cooperation and Development’s Two-Pillar tax reform plan. I know I shouldn’t–but when there are only so many hours in a day and so much tax policy to cover (and so many Tweets to write), it’s tempting to de-prioritize a proposal which faces longer odds than a third Independence Day flick.

Tempting, but probably wrong. True, Pillar One must be implemented by a multilateral treaty agreement which will take years to put together, and which the U.S. may never ratify. But while the odds are small, they’re definitely not zero. And tax policy is funny–seemingly academic policy proposals can suddenly become real when no one’s expecting. And when they do, it’s often too late to give them the scrutiny they deserve.

So let’s take a look at it. The OECD recently held an in-person consultation over their most recent progress report on Pillar One, which is a policy meant to capture more of the digital activity that’s not recognized in the traditional, more tangible-based global income tax system. The consultation came after the organization received about 70 comments from taxpayers, organizations and other stakeholders, outlining their areas of concern with the document.

There’s one overwhelming consensus from most of the comments–the complexity of the current plan could be fatal. The proposal lays out a seemingly endless number of definitions, distinctions, categories, formulas, alternative formulas, secondary formulas, fallback formulas, and other blocks of text to explain how the new income allocation system would work. This is especially true for the rules on revenue sourcing, which are likely the most important part of the plan as they will determine which jurisdictions get the biggest cuts of this new revenue pie.

The rules cover a broad array of transactions, which taxpayers must categorize in order to follow the correct revenue sourcing formula. (Unless they fall into a catch-all “Other” silo, which commenters complained was very nebulous.) The distinctions for categories are subjective and could likely lead to conflicts over interpretation, before the formula is even applied.

But it’s not just the complexity, commenters warn. It's also the assumptions behind the procedures. In many cases, the taxpayers are expected to lean on third-party customers–likely other businesses in a supply chain–for new information about the final location of products or consumers.

“The Consultation Document seems to reflect a belief that the full supply chain for a given product, across different companies and geographies, is transparent, when in fact such detail is not typically available,” writes Ernst & Young LLP.

KPMG LLP, another of the “Big Four” accounting firms, warns against assuming “that groups can comply with it's preferred approach simply by devoting additional time or resources to obtaining more information, or that having to solicit customers for more information is not a significant or burdensome undertaking.”

"There are many situations in which transaction-specific data identifying the precise location of a final consumer or the employees of a business customer is simply not accessible," KPMG added.

It’s one thing to ask large corporate taxpayers to divulge information to tax authorities that they’d rather keep to themselves–that’s a constant tug-and-pull of tax enforcement. But it’s another thing to ask them to come up with information they may not even have, and can only obtain by adding new conditions to contracts which could end up altering the economics of the deal.

"It is not commercially realistic to require suppliers to request sensitive business information from their customers," the Silicon Valley Tax Directors Group said.

The commenters asked the OECD to only require information that would be normally available through the course of business--but it's unclear if that could cover all that the project needs. And is it really OK to leave it up to taxpayers to decide what information is reasonable to get? Leaving the requirements as they are could be difficult, or even impossible, in practice. And dropping them entirely would contradict one of the key goals of the project–to tie this new income to the consumer markets which generated it.

How the OECD ultimately solves this question could determine the very nature of the project going forward. This is an area where a long view–having followed this from the very beginning, nearly five years ago–can really help understand the context. Because, originally, Pillar One wasn’t supposed to cover these types of business-to-business transactions at all.

Sometimes, I think I’m the only person who remembers that Pillar One was originally about marketing intangibles. Those are valuable intangible assets, like brand IP or data, which could be derived from a market. The exact nature of these intangibles was a little fuzzy, and the logic of how they were the key to this puzzle took some squinting to understand.

The rationale, touted by OECD and U.S. officials back in 2018 and 2019, went something like this: the arm’s-length standard (the traditional basis of international taxation, using market prices for income allocation) is still king, and should still serve as the backbone of the system. However, there are some limited situations where our current capabilities do not fully account for the valuable intangibles that a company can derive from a market, including through online-only activity. If transfer pricing were perfect, we would recognize these intangibles and allocate more income to the market jurisdictions where they were created. But because we only have our imperfect tools at our disposal, we should use a formula to approximate these assets' value. This should be a portion of “residual” or higher-than-normal income–a proxy for income from intangibles–based on sales, preferably consumer-facing sales.

The problem that this was meant to solve is primarily a political one, not a matter of pure tax policy. European countries wanted to tax digital, tech companies more, claiming the outdated tax rules were letting online income slip through the cracks. The U.S. claimed this was discrimination, and vowed to fight back against any new levies targeting technology or digital companies in particular. The Group of 20 nations asked the OECD to come up with a solution, but the sides seemed too far apart to make any consensus possible.

The marketing intangibles concept could square the circle and cut the Gordian Knot, in theory. It would avoid taxing (mostly American) tech companies by taxing everyone. All sides would be happy–Silicon Valley wasn’t being discriminated against, European countries could tax companies based on an online presence alone, and the rest of corporate America would be happy to pay a bit more in tax to preserve the tax system, (If that last part sounds like very wishful thinking, well, it turned out to be.)

It’s not necessarily bad that the policy followed the politics here–that’s how things work in a democracy. But the politics has led the project through so many obstacles and U-turns, the policy itself has morphed into something most don’t recognize.

Defining “consumer-facing” products turned out to be more complicated than it sounded, and it became more complicated still after the OECD added cloud computing. The policymakers created a second category, “automated digital services,” to account for that and other digital businesses which didn’t clearly fit into the consumer-facing role. (Online advertising and data collection were also included, with the online viewer/user standing in as a proxy for the consumer.)

Eventually they included a dizzying number of subcategories and exceptions, which participants found to be unworkable. In an example that former Treasury official Chip Harter loved to highlight, in one draft diamonds (considered a finished product) were covered by Pillar One while pearls (falling under a fisheries exemption) were not. The project stalled, with the U.S. threatening to bolt.

The new Biden administration in early 2021 came in with a new proposal to clear up the mess and simplify all of these headaches. The consumer-facing/automated digital services categories were gone–instead, the plan would target only the world’s most profitable companies, under the logic that they were likely the ones benefiting the most from valuable intangibles.

By adding business-to-business (B2B) transactions, the policy started to look more and more like a value-added tax–but without the established invoice rebate system that makes a VAT run relatively smoothly. It makes you wonder why the OECD is trying to do this under the rubric of an income tax system at all. (The answer, again, is politics: the consumption-based European digital services taxes that this is hoping to replace are allegedly a temporary patch to the income tax system.)

In trying to avoid the morass of complexity, the OECD may have run into it on the opposite side. The B2B model couldn’t be grafted onto Pillar One without creating a whole new host of problems, as the latest consultation and commentary shows.

We don’t normally think of it this way, but most of the dominant tech companies that rule so much of our lives are B2B. Facebook and Google are primarily advertising companies, providing a useful service for free to capture people’s eyeballs. In the meantime they collect and sell data about us, an even more valuable commodity. As the saying goes, we aren’t their consumer, we’re their product. Even Amazon profits more from its web hosting and other online business services than its indispensable retail machine.

Excluding these companies, for a project that purports to capture new digital income, is not an option. But how to include them in a consumer, market-based model may require an answer that they haven’t found yet–even after solving the relatively simple question of how to use online viewers as a proxy for the destination of advertising or data collection transactions. (Which I’ve always suspected isn’t as simple as it seems.)

Moving backwards is always difficult for projects like these, but some commenters suggested ditching B2B transactions altogether.

The National Foreign Trade Council noted that "non-customer revenues," by definition, "are not derived from any market jurisdiction."

Of course, then we're back to deciding what to do with diamonds and pearls.

What this demonstrates is how difficult it is to answer these questions under the pretense of giving the existing system a simple patch--something that no one ever really believed in the first place. These are fundamental questions about the nature of markets and value creation, and it's looking more and more like they can only be answered by re-examining some of the basic foundations of the global tax system. But will there ever be political will for that kind of overhaul?

DISCLAIMER: These views are the author's own, and do not reflect those of his current employer or any of its clients. Alex Parker is not an attorney or accountant, and none of this should be construed as tax advice.

LITTLE CAESARS: NEWS BITES FROM THE PAST WEEK

Lawmakers aren't wasting any time before leaning on Treasury about how to implement the tax provisions of the Inflation Reduction Act. Several Democrats wrote a letter to the department urging it to stand "firm against requests to dilute the regulations in such a way as to undermine the clear intent of the law." It's a little unclear what they're talking about, but the treatment of private equity and "split-off" corporate reorganizations are some of the issues that have already arisen. Meanwhile, Senate Finance Chairman Ron Wyden (R-Ore.) in his own letter also gave Treasury a list of priorities he expects to see it pursue with its newly beefed-up budget, including to "crack down on offshore tax evasion."

While policymakers remain at a standstill on the other side of the OECD project, Pillar Two, tax administrations say they're still moving forward with implementation. Following a recent meeting, the OECD's Forum on Tax Administration released a statement outlining several priorities, including to "intensify joint consideration of the detailed administrative and capability aspects" for Pillar Two and to "leverage our experience from applying existing multilateral tax certainty tools" for the new challenge. (Also of note: the OECD released a progress report from the prior international tax overhaul, on the country-by-country tax transparency initiative. Can never have too many irons in the fire.)

Not exactly a news item, but tax wonks all seem especially interested in the acronym wars arising from the Inflation Reduction Act. The crafters of the 2017 Tax Cuts and Jobs Act spent at least some time coming up with catchy-sounding acronyms (BEAT, GILTI) while others (FDII, pronounced "Fiddy") caught on despite many objections. The IRA has acronym problems from the get-go, and folks haven't even decided what to call its tax centerpiece, the new corporate minimum tax on book income. But there's momentum to use its official legislative title, the Corporate Alternative Minimum Tax or CAMT--and worse, to pronounce it "Cam-Tee." (Doesn't it kind of rhyme with "panty," one of the most loathed words in the English language?) Or we could go with the vaguely edible-sounding BMT (book minimum tax). Could always be worse--when Congress enacted a similar tax in the 1986 tax reform, it was called the Business Untaxed Reported Profits or BURP tax. Excuse me.

PUBLIC DOMAIN SUPERHERO OF THE WEEK

Phara, the Living Goddess, appearing in Zegra, Jungle Empress #2 in 1948. Ruler of Kait, she can control lions.