The US – EU Dispute Over Digital Services Taxes

I. What Are Digital Services Taxes?

DSTs are taxes on the gross revenue, such as from advertising and data sales, that certain large and highly profitable digital companies earn from consumers through websites that offer the likes of search engines, social media services, and online marketplaces. 

These taxes are imposed without regard to the targeted companies’ physical presence in the taxing country. This distinguishes DSTs from conventional corporate income taxes, which generally require the physical presence of an in-country permanent establishment. 

• The rising use of DSTs reflects EU (and other taxing authorities’) frustration when U.S. digital giants make large profits from interactions with their own consumers, while avoiding the countries’ corporate income taxes due to the physical presence requirement. So far, the EU-wide revenues raised appear to equal several billion Euros per year. DST proponents view physical presence as a badly dated relic of international tax conventions that arose 100 years ago, when business and consumer technologies were far more primitive than they are today. The proponents therefore view DSTs as newly allowing these countries to reach profits that fall within their rightful scope (given these companies’ pervasive interactions with their own consumers), but that have previously been unreachable for a merely technical reason.
• American critics of DSTs view physical presence as still a proper touchstone for taxation – at least with respect to US multinationals that are active abroad. They therefore view DSTs as both wrongfully “extraterritorial,” and as having been, in practice, discriminatorily tailored to fall chiefly on prominent US firms.

DSTs target large digital companies. DSTs typically have revenue thresholds that limit their application to large digital companies that, in practice, are mainly American. France’s 3 percent DST, for example, applies only to companies with worldwide revenues of at least €750 million and French revenues of at least €25 million. This causes it to apply to such foreign companies, from a French perspective, as Alphabet (Google’s parent company), Amazon, Apple, Meta (Facebook’s parent company), and Microsoft, but to few if any French companies other than Criteo.

DSTs must be paid by a corporation even if it offers its services to consumers for free (as do Google and Facebook). In such cases, the tax base typically consists of revenues that the company derives from advertisements that are apparently targeting such consumers.

II. What is the Current State of Play for Digital Service Taxes?

Nineteen countries worldwide have Digital Service Taxes and another 21 have proposed having one. The table shows those European countries that have, or have proposed, a digital service tax. These include Great Britain (2%), France (3% on advertising, digital interfaces, and user data, and 1.2% on streaming), Spain (3%), Portugal (4% for streaming video), Italy (3%), Austria (5%), and Hungary (7.5%). 

Estimates of the market size of digitally delivered services vary from about $2 trillion to $4 trillion annually at present (depending in part on exactly what is being included), with expectations that within a decade it will at least double or triple.

At present, DSTs paid to foreign governments generally cannot be offset by a foreign tax credit. Whether this should change is currently subject to debate around the world.

Many countries have called their DSTs temporary, and expressed willingness to have them replaced by the adoption of OECD Pillar 1, but the latter’s uncertain future and apparently poor prospects for widespread implementation call this scenario into question.

The U.S. state of Maryland is levying a 3% DST that is currently facing federal legal challenges on multiple grounds. Subject to the as yet uncertain legal landscape for state-level DSTs, other states might conceivably choose to follow Maryland’s lead.

III. Distinguishing Features of the Digital Service Tax

One argument that would distinguish DSTs from conventional taxes is that, by nature of the tax and the market power of the large digital firms, these taxes are primarily borne by the firms that make DST payments to the tax-imposing governments. This contrasts with taxes that raise the prices of goods or services and are borne by both companies and consumers. If this is the case, DSTs would raise revenues for the EU with relatively little cost to its consumers and would reduce the profits of US-based corporations and the dividends paid to their shareholders. 

To illustrate this argument through a simplified example: Suppose Facebook’s intellectual property (IP) can be deployed anywhere in the world at no additional cost, and that Meta (its parent company) need not spend anything, either to keep its digital platform up and running, or to tailor its use to any particular country. With regard to (say) the UK, all that it needs to do is (in effect) to answer Yes, rather than No, to the question of whether the platform should be available there. A Yes leads to an immediate positive revenue stream for Facebook at no additional cost to the company since it is essentially costless to offer this platform in the UK (alongside everywhere else). Entering a new market would depend on Facebook getting free money. This simplifying assumption can be justified by the fact that Facebook already has an operating digital platform, and it may also be a reasonable way to think about many digital companies’ revenue-generating activities. In this scenario, use of Facebook in the UK may have no implications for its ability to provide its service anywhere else. Therefore, the company is not in the same position as one that is building a factory, or selling tangible products, and that therefore had to decide (unless it ramps up production) to do so either in Country A or in Country B. Instead, so long as Facebook could clear even a penny of pretax revenue by being active in a given country, the only rational answer it could give to the question of whether it should operate there (leaving aside strategic posturing) would be Yes. Indeed, this would be so even if its revenue from the operations was being taxed at very high rates.

Under this scenario, the price of advertising on Facebook might be little if at all affected by whether or not there was a DST. Suppose that, before the imposition of a DST, Facebook sets a price per advertisement that earns it $3 billion in revenues from British advertisers, and that this price maximizes its profits. If the UK then imposes a 2 percent DST, Facebook’s after-tax profits are reduced by $60 million. But Facebook may have no incentive to change its price from the one that maximized its profits before the imposition of the DST if the tax has no effect on the cost of providing an advertisement on its site. Essentially, the tax is directed to revenue, not to units sold, so (with the zero marginal cost assumption) there is no incentive to change prices to affect the number of units sold.

There is also the issue of whether there are close substitutes for advertisements on Facebook. If Facebook loses much revenue if it raise prices because of competition from EU-based companies then it also has less incentive to raise prices.

This scenario is clearly not entirely true. Surely Facebook continues to spend money both operating and optimizing its digital platform for global use, and specifically on managing and maximizing its UK net revenues. How true is it that digital giants such as Facebook can deploy their IP anywhere in the world at an extremely low marginal cost? Does the zero-marginal-cost model, while not completely true, come close enough for its conclusions to be largely applicable? After all, surely a firm such as Facebook must expend some resources on tailoring its digital platform to (say) the UK market, and then on taking further steps to optimize its derivation of UK-related advertising revenues.

There is an alternative view whereby prices do rise. The economist Jane Gravelle noted in a recent Congressional Research Service study: “DSTs are excise taxes (taxes on sales), not taxes on profits. Generally, excise taxes are passed on to customers. In [the case of a DST], the tax will be passed on to advertisers (who in turn are likely to pass it on to their customers) and online consumers.” To illustrate this point, consider a more conventional scenario, like when it is costly for an automobile manufacturer to produce another car. When a tax is imposed on each car sold, the auto company will raise its price to reflect the higher cost of producing each car (the marginal cost). The increase in the price paid by consumers depends upon their sensitivity to the price increase – if they are relatively insensitive, the profit-maximizing price will rise by more than if they are relatively sensitive. In this case, the tax is borne, at least partially, by consumers. In this case, too, it depends upon whether consumers see other opportunities as close substitutes – if so, the taxed company is more limited in its ability, and desire, to raise prices. Moreover, we should also keep in mind the point that being active in a given country is not just an all-or-nothing, Yes or No, proposition. One can expand or contract one’s degree of local involvement (and expected gross revenue) at the margin. In addition, depending on the availability and cost of funds for a given digital company that is subject to the DST, along with the marginal cost-efficiency of its operations as they are scaled up, the company might indeed find itself choosing between marginal investments in different jurisdictions. This as well might push towards one’s accepting the excise tax view.

As yet, there is little conclusive evidence on which of these two views is correct. One problem with efforts to analyze the incidence of DSTs is that, in the short run, empirical evidence could be affected by companies’ initial strategic efforts to affect political decisions; they might pursue local political goodwill by making a show of not raising their prices or they could attempt to discourage the DSTs by explicitly raising prices to dramatize the proposition that DSTs do indeed get passed on to locals.

IV. Policy Considerations

DSTs are a current point of international contention. As recently as December 2025 (the time of this writing), the United States has threatened to retaliate against EU firms if the EU does not remove its DSTs. This is a continuation of threats from August 2025, when president Trump vowed he would impose substantial new tariffs and restrict U.S. chip exports for all countries that did not remove DSTs and related regulations as well as earlier threats. The concerns about DSTs predated the second Trump Administration – the Biden Administration had also voiced its opposition to these taxes.

Both American and EU officials seem to view DSTs as primarily borne by firms rather than consumers. But given the prevailing uncertainties with regard to the digital giants’ operations and economic incentives, we cannot as yet tell whether the U.S. and the EU countries – which join each other in assuming that the latter’s DSTs are principally burdening Americans, rather than Europeans – are right or wrong. Perhaps the correct answer to who bears the incidence of these taxes will become clearer in the near future. But for now, the most defensible conclusion is that either view – DST incidence mainly on American shareholders, or on European consumers – might soon prove to be the more accurate one.

Surprisingly, the underlying economic incidence questions appear to make no observable political difference. The U.S. and EU policymakers’ shared assumptions could ultimately prove to be right or wrong, but there is great uncertainty. Thus, the DST dispute truly is an instance of the old saying, “Ready, fire. aim!” Neither side really knows how the disputed DSTs are affecting their economic welfare. It is curious how willing they both are to carry on anyway.