Some time ago, these pages predicted that digital services taxes would prove to be sticky things. That is, once governments around the world went through the trouble of implementing DSTs — and once the receipts started to trickle in — legislators would find them difficult to eliminate. DSTs check too many of the boxes associated with a politically expedient tax regime.
The optics suggest that DSTs are a tax on capital income. Never mind that they’re a tax on gross receipts, which is a poor proxy for a firm’s profitability. The optics further suggest that the economic burden falls on nonresident corporations, probably the ones with a track record of profit shifting. Never mind that the burden might easily be passed on to the local residents who consume digital services — or that being a provider of digital services doesn’t tell us anything about whether the firm engages in aggressive tax planning.
Our friends at the OECD are not wrong when they insist that the DST is a flawed creation. There’s a bevy of reasons why DSTs are ill-advised. But it’s debatable whether those arguments will prevail among the folks charged with raising public revenue. Elected officials have sought to export the local tax burden for as long as taxes have existed. Recall what Jean-Baptiste Colbert once said about “the art of taxation” and “the smallest possible amount of hissing”? He might as well have been describing DSTs.
Outside of the United States, DSTs must feel like the perfect tax. There’s little pushback in terms of any detrimental effect on national competitiveness. If enacting a DST made countries uncompetitive, I doubt so many governments would be racing to adopt them. Given the binary choice between enacting a DST or increasing the corporate rate, many legislators would opt for the former.
While this impulse is absent in the United States at the federal level, we’ve witnessed it at the state level. Lawmakers in Maryland were desperate to enact their own version of a DST a few years ago, guided by dreams of forcing out-of-staters to foot the bill for all manner of local spending. In February 2021 they even voted to override a gubernatorial veto. Maryland’s DST was defeated in the courts last year for good reason: It violated the constitutional prohibition against interference with interstate commerce. But you can appreciate the broader point about the tax’s instinctive appeal, which other commentators have described as “the siren song of false prophets.”
Why bother discussing those pesky DSTs? They’re just temporary nuisances . . . correct? Once humanity decides to embrace the OECD-brokered pillar 1 reforms, they’ll vanish into the ether . . . right? I caution readers against placing blind faith in such assurances. The DSTs might go away, as advertised, or they could multiply in number — becoming intractable fixtures of the international tax system.
You already know what the OECD, the tech sector, and academics have to say about DSTs. This week’s column presses you to focus on the political context. As a thought experiment, let’s consider how DSTs are likely to be regarded by the rival political parties in the United Kingdom. Tories have favored the DST because it raises a bit of revenue without being identified as an overt job-killer and without presenting a significant hinderance to growth and investment. The recent succession of U.K. prime ministers, all of them Tories, welcomed the DST as properly serving the national interest. Boris Johnson vigorously favored the DST, despite being otherwise keen to make nice with his U.S. counterparts at every opportunity.
The Labour Party also favors DSTs. Its membership believes the tax promotes economic justice by forcing a few large multinationals (that happen to be foreign) to pay more tax in a country in which they benefit from a massive user base. On the surface, the tax seems to respond to the problem of rampant profit shifting, as outlined by the OECD’s base erosion and profit-shifting project of a few years ago. I say that because of the common perception that the firms that pay DST are the same crew we read about in connection with migrating intangibles and extremely low effective tax rates. Labour members of Parliament support the DST for the same reason they support the range of BEPS-inspired reforms: They translate to modestly increased levels of corporate taxation — a theme that plays well with the base.
That makes DSTs one of the few revenue tools that ideologically opposed lawmakers can support. That’s what I meant by checking all the boxes for political expedience. It’s why I suspect DSTs will be challenging to eliminate.
I won’t deny that DSTs are a crude way to backstop the conventional income tax. Parallels can be drawn to the individual alternative minimum tax, which came about in the United States in the late 1960s when lawmakers saw the need to backstop the individual income tax. It would have been preferable to reform the income tax itself, but that wasn’t in the cards. Once the AMT was up and running, it soon became clear that for a specific pool of affluent taxpayers the AMT was their primary source of federal tax liability. That made it almost impossible to repeal. I wonder if DSTs occupy a similar space.
Backstops need not be elegant; they need only be effective at reaching some of the entities thought to be undertaxed. How effective are DSTs in practice? Because DSTs are a novelty, they don’t have a history. A new report from the U.K. Parliament’s Public Accounts Committee (PAC) goes some way toward filling in the knowledge gap. It takes stock of the country’s initial experience with a DST, introduced in April 2020. As far as I can tell, the PAC report represents the first instance of an informed reflection on how the great DST experiment is playing out. While the report is limited to a sample size of one year (a pandemic year, no less), it’s better than nothing.
U.K. legislators must figure out what to do with the DST by 2025. As enacted, the U.K. DST does not expire. There is, however, a mandatory review after five years. April 2025 is two years away. Based on this new PAC report, it’s possible to make an educated guess as to what that future appraisal might look like. Here’s your executive summary:
- As a gross receipts tax, the DST is harsh on high-volume/low-margin businesses.
- The DST raises a bit of revenue, with the potential to raise a lot more.
- Some taxpayers pay more in DST than they pay in income tax.
- Some of the economic burden (we don’t know how much) is passed on to third parties.
- The U.S. government detests the DST; everyone else is basically fine with it.
Each of these observations is something that lawmakers would have already known, or could have foreseen, when they enacted the DST. None of these points jumps out as a reason why the DST must be repealed in short order. We should anticipate the 2025 review will conclude that the DST, while clunky and suboptimal, generally works as intended. To abolish it would be to turn away a bundle of revenue that will be difficult to replace.
Absent some external catalyst (for example, a trade war with the United States), MPs may conclude there’s little reason to remove DSTs. We should be worried that self-propagation is now the default position. You’ll need a persuasive justification for eliminating the DST, and you won’t find one in the PAC report. I expect other countries with DST regimes would reach similar conclusions, but let’s not get ahead of ourselves.
Thirteen of Eighteen
By way of reminder, the U.K. DST is imposed at a rate of 2 percent. It reaches corporate groups whose global revenues for in-scope activities exceed £500 million, of which at least £25 million is derived from U.K. users. In-scope activities include online search engines, social media platforms, and online marketplaces.
For the DST’s first year of operation, 2020-2021, HM Revenue & Customs hauled in receipts of £358 million. That’s not a huge sum of money, but it was 30 percent more than forecast. There’s no clear reason why the figure surpassed expectations by that much. A possible explanation involves the COVID-19 pandemic, which forced people to become more reliant on digital services.
According to the report, 18 corporate groups paid DST during that first year, although another 46 are still being assessed for DST liability. The final tally could be higher than the amount referenced above. Of those 18 groups, there’s a subset of five taxpayers that account for 90 percent of all DST receipts. The report doesn’t identify the “big five” DST payers, but you can probably guess as to the usual suspects.
Of the 18 affected firms, 14 paid more DST than expected. The remaining four paid less than expected. Separately, there is another pool of 11 firms that expected to incur DST liability but haven’t, at least to date. They might be among the group whose liability is still being assessed.
Here’s the kicker: Of the 18 firms, 13 paid more in DST than they did in corporate income tax. For anyone who offers DST as a viable creature, there’s your hook. That’s 72 percent. It’s hard to argue that DST flunks as a backstop when it’s the predominant tax being paid by roughly seven in 10 of the affected firms.
How does the 2 percent DST pack more punch than the 19 percent corporate income tax? For starters, the tax bases are entirely different. Companies can report huge revenues but scant profits. Of course, a possible reason for reported profits being so meager is profit shifting. DST is purposefully blind to profitability. This is both a glaring defect and the main reason that many governments are adopting them. An overlooked legacy of BEPS is the desire for a corporate tax base that looks beyond profitability.
It’s tempting to take the above statistic (13 of 18 firms paid more DST than income tax) as evidence that gross receipts taxation can indeed function as a safeguard against profit shifting. Would I prefer to get there by means of a more dignified solution? You bet I would, but that’s not on offer. DST is what’s on offer. At this point, a country is foolish not to have one — if for no other reason than it adds leverage for pillar 1.
There’s more. The PAC report contains a revised projection for cumulative DST receipts over the first five years of operation (running through 2024-2025). That figure comes in at £3 billion. What does that estimate say about the viability of DST? Consider the corresponding cost estimate. According to the report, HMRC spent a mere £6 million to implement DST.
I’m no math major, but a £3 billion return on a £6 million outlay sure sounds like an impressive bang for the buck. I guarantee you that tax administrators in distant corners of the world are paying attention to that detail. They’d be negligent if they weren’t.
DST > BEPS
Let’s remind ourselves what everyone dislikes about DST. It’s acknowledged to be an inferior design; an unsightly gross receipts tax that conflates profits with sales. Fair point, but DSTs do seem to reach taxpayers that excel at profit shifting.
To the extent this new revenue tool augments the fiscal capacity of market jurisdictions, it’s not much of a stretch to say that DSTs are doing the chores that BEPS was supposed to do — at least as to BEPS action 1 on the challenges of the digital economy. The final BEPS reports landed in 2015, which was a long time ago. The ultimate indictment of BEPS is that much of the world perceives the need for a fiscal device as maligned as DSTs.
At the risk of saying wildly unpopular things (and why stop now?), DSTs are a step toward fulfillment of the BEPS policy objectives, albeit through a distinctly unorthodox approach. Recall our proviso about fiscal backstops: Accept that they’ll be crude, judge them by their efficacy.
The PAC report offers a set of recommendations for HMRC and HM Treasury. The last of these recommendations urges HMRC to develop a “contingency plan for what happens if the Digital Services Tax needs to be extended.” You can use your imagination as to why such an extension might be necessary.
The report leaves me more convinced than ever that the U.K. DST will keep rolling along well after 2025 — absent an external restraint. Ditto for the tax’s other early adopters. Legislatures that previously created DSTs don’t suddenly have a reason to repeal them. As said at the outset, these DSTs are sticky stuff.