Digital Services Taxes: Is There an End in Sight?

Since 2018, the tax world has seen multiple rounds of proposals to change how digital companies are taxed. The most common, the Digital Services Tax (DST), popped up as a European Commission directive and then made its way into the national laws of multiple countries across the globe.

Ongoing international tax negotiations were partially motivated to provide an alternative to DSTs. This week the Organization for Economic Co-operation and Development (OECD) once again demonstrated that the elimination of DSTs and similar policies is still part of the plan.

Let’s remember how we got to this point. In 2018, the European Commission and several national governments argued they needed special rules to tax digital companies. They believed that users in their countries (and their data) were creating value for the companies that would not be taxed there.

The user-created value theory harkens back to older debates about when value is created. Are California oranges valuable when they are picked, when their juice is squeezed, or when their juice is consumed? (The same questions could be asked of French grapes—or even Californian grapes, for that matter.)

The answer to this question impacts where a company should be taxed on its profits. If the users of a digital platform create value (or profit) for a company, then taxing that profit where the users are would make sense. If, however, the user data is worthless until it is run through algorithms developed and deployed in another country, then the profits from processing that data could justifiably be taxed in the other country.

Long-standing international tax principles generally allocate taxable profits where a company’s workers, production, and owners reside.

But that creates a problem for countries that want to tax based on where a digital company’s users are located. If digital companies’ profits are already being taxed according to current principles, an additional tax in users’ locations would be double taxation—taxing the same base more than once.

That problem is easily solved if you ignore existing international tax rules and simply give in to the political drive to tax large digital companies. Enter DSTs.

The European Commission proposal was for a 3 percent tax on the revenues (not profits) of companies with worldwide revenues of €750 million ($797 million). There were multiple other design elements, but U.S. analysts commonly viewed it as a tariff on U.S.-based digital companies.

Revenue-based taxes are regressive. Even if the policy targets “large” companies, a revenue-based tax will result in a higher effective tax rate for a less profitable company relative to one with a higher profit margin. Therefore, it’s good policy to tax profits rather than revenues.

The Commission proposal failed to achieve unanimous support at the EU council, but individual countries including France, Spain, Italy, and the United Kingdom (albeit no longer in the EU) plowed ahead with their own versions.

To address the growing chaos and trade threats from the U.S., the OECD set out to design a proposal that would change where companies pay taxes to partially assuage the concerns of the governments that pursued DSTs.

That policy is now known as Amount A of Pillar One. Put simply, Amount A takes a slice of corporate profits from the largest and most profitable companies (not just digital companies) and moves it away from the jurisdiction that currently taxes it to a jurisdiction where a company’s products are sold to consumers. In a way, this splits the proverbial baby. Most of that company’s profits are still taxed according to the long-standing principles of profit allocation, and a portion is allocated using Amount A.

The problem: Amount A is complex and requires a multilateral tax treaty to even be enforced.

Pillar One would, however, eliminate DSTs. In a new document released this week, the OECD presented draft language that suggests countries might only benefit from Amount A if they remove their DSTs or similar policies (these could include policies like the Indonesian and Indian equalization levies). What the documents did not include is a list of what constitutes a DST or “relevant similar measures.”

The OECD did lay out some principles that could help identify problematic policies (see here for the specific language):

1. The tax is driven by the location of customers or users.
2. It is primarily a tax on foreign businesses.
3. It is something other than a tax on income and is beyond agreements to avoid double taxation.

These principles are a good start, but the incentive to remove a DST will depend on whether a country sees a better tax revenue outcome from Amount A. In turn, those revenue numbers will depend on how the rest of Amount A gets negotiated.

Also, it seems unlikely that these principles will result in “all” DSTs being removed as agreed in October 2021. There is room for governments to work around the principles above. A DST could get past the second principle by applying it to both domestic and foreign businesses in a somewhat balanced way.

Many questions remain unanswered because the full Pillar One agreement hasn’t been reached yet. But as it stands, Pillar One would usher in the end of many DSTs (though perhaps not all) at the cost of increased complexity (in an already complex and uncertain system).

The question that taxpayers and governments across the world will have to answer: is it worth it?