Out-Law Analysis 9 min. read

OECD digital tax proposal will increase tax paid by multinationals


The aim of the latest 'pillar one' digital tax proposal from the Organisation for Economic Cooperation and Development (OECD) is not only to reallocate taxing rights, but also to increase the tax take.

The proposal does this by adding new taxing rights on top of the existing system, targeting all multinational enterprises (MNEs) with consumer facing businesses and not just technology companies.

Much still remains to be decided. Even before the coronavirus epidemic, it looked as if the OECD had perhaps given itself too much to accomplish by the end of 2020, with governments around the world likely to fight for their slice of the new taxing right.

The OECD and national governments have been grappling with the issue of how to tax online economic activity in an age where customers and suppliers are frequently in different countries. In May 2019 the OECD agreed with the G20 a programme of work to come up with proposals to reform the international tax system to address the challenges of the digitalisation of the economy. This programme of work was divided into two 'pillars'.

Pillar one addresses the allocation of taxing rights between jurisdictions and considers proposals for new profit allocation and nexus rules. Pillar two is the 'global anti-base erosion' proposal which seeks to develop rules that would provide jurisdictions with a right to tax where other jurisdictions have not exercised their primary taxing rights or the payment is otherwise subject to low levels of effective taxation.

These plans are not really about digital taxation at all, but about radically changing the system of international taxing rights across all consumer-facing industries.

An OECD statement on 31 January 2020 said that an outline of the architecture of a unified approach to its 'pillar one' proposal has been agreed so it is useful to look at how the pillar one proposal would work.

The two-tier gateau

It now looks certain that the OECD is aiming not only to reallocate taxing rights - splitting the same tax pie into different slices - but also to increase the tax take, making a bigger tax pie. The proposals do this by adding an overlay of new taxing rights on top of the existing system.

Think of it like turning the tax pie into a two-tier gateau. Normal routine profit sits underneath and is divided up between nations on existing principles. However, an extra layer of deemed residual profit sits on top, to be divided up according to a new methodology between all the countries where the MNE operates, so-called 'market jurisdictions'.

So how does it work? As we will see, all the details are yet to be thrashed out, but the outline of the methodology borrows a lot from the world of transfer pricing. Because the OECD refers to three amounts - A, B and C - you may see it called a three-tier system, and that is accurate. But it is also confusing because in terms of new versus old taxing rights, it is really a two-tier system which makes an MNE taxable on amount A (the new top tier of the gateau) and on amounts B and C (existing routine profit underneath).

Amounts B and C

Normal routine profit takes the form of amounts B and C. Remember, we are discussing here an MNE that is, say, outside of its home jurisdiction in X, selling products and services to consumers in countries Y and Z.

Amounts B and C would be the MNE's only exposures to tax in country Y or Z under current rules. Amount B would be fixed remuneration for baseline marketing or distribution functions that take place in country Y or Z. Amount C would be any additional transfer priced profit where in-country functions in country Y or Z exceed the baseline activity compensated under amount B.

Amount B is conceptually easy enough to grasp. In simple terms, you apply the arm's length principle (ALP) from transfer pricing to amounts paid to marketeers and distributors getting products/services out to consumers in country Y or Z; and any permanent establishment that the MNE has in country Y or Z.

This fixed return would consider differences in region and industry, as a price fixed under the ALP should.

Actually, applying transfer pricing rules to calculate that amount is another matter, and one that is already rife with dispute between countries X, Y and Z and the MNE under current rules. The idea here is to lock down an assumed return that reflects the MNE's activity and minimise those disputes. The OECD thinks that no treaty changes are required to implement amount B, but that only holds true if it pins it down enough. To do that properly, it acknowledges that it needs a menu of defined baseline activities, appropriate profit level indicators, returns as fixed percentage(s) at an agreed profit level, benchmarking studies and industry and region differentiation.

That is a lot to lock down, so we might question how the OECD is going to get agreement from 137 countries on the detail.

Amount C is more amorphous. In several ways, it is a realistic acknowledgment by the OECD that the new formula will not be perfect, and leaves wiggle room for country Y or Z to grab a bit more of the lower tier of the tax gateau, where there is more going on in-country under the ALP than amount B recognises. However, because of that, defining amount C is like pinning jelly to a wall, and leaves a wide field of potential disputes open. The OECD recognises this, but that will not help unless it gets the double taxation element right, and by keeping amount C in the mix it is only making its job harder.

Amount A

This is the new taxing right over deemed residual profit. Again, much remains to be decided.

It looks more likely that the £750m revenue threshold from country by country reporting may be set as the trip point for MNEs, although up for consideration is an added layer of aggregate in-scope revenue and/or de minimis thresholds.

If you think the new taxing right only applies to search engines and online marketplaces, think again. It is intended to apply to two broad categories — automated digital services and certain consumer facing businesses — but within those categories are some surprising sub-categories.

Automated digital services include not just the three expected (online search engines, social media platforms and online marketplaces) but also digital content streaming, online gaming and cloud computing.

What these sub-categories have in common is that they are all provided on a standardised basis to a large population pool. Note, though, that customers and users are conflated together into one amorphous bag, even though in commercial reality they are not the same thing.

Consumer facing businesses will have to have a significant and sustained engagement (directly or, more worryingly, indirectly) with their markets to be caught. This does not only include personal computing products (e.g. mobile phones) but also extends widely to include clothes, toiletries and cosmetics; branded food and refreshments;  franchises, e.g. in the restaurant and hotel sector; and automobiles.

These lists are slightly mixed bags, probably reflecting a combination of the 'hit lists' of the governments involved in the process. Whilst the OECD is at pains to say that not just any consumer facing business will be caught, it is not at all clear yet where the lines will be drawn.

Nor is it just consumer facing business affected. Whilst business to business sales of intermediate components/products are to be largely excluded, the OECD cannot resist flagging up exceptions that are to be pulled back in (e.g. branded components commonly acquired by consumers). It is keen to keep such exceptions to a minimum. Whilst the oil and gas extraction industry, raw commodity sellers, and services involving human judgment (such as lawyers and accountants), plus shipping and airlines, all fall outside the scope of the new taxing right, however, the position is uncertain for the financial services sector.

Why bother with two categories? This approach allows the OECD to construct two different sets of conditions.

For automated digitalised businesses, only the revenue threshold will apply. Cross that threshold, and you are hit.

For consumer facing businesses, additional factors are to be set to determine whether there is 'significant and sustained engagement' with markets. These may involve passing a local in-scope revenue threshold over a set period (taking into account the size of a local market, subject to a minimum level), the existence of a physical presence in-country, or targeted advertising. Such 'plus factors' currently look amorphous and ambiguous in scope, though. How do you pin down that 'extra something' that an established brand has in its markets? It will be a question easier asked than answered.

A big elephant in the room remains the US 'safe harbour' proposal; rhetoric aside, this looks a lot like an 'opt-out' election. An opt-out is unlikely, and if the US is insistent this may prove an insurmountable obstacle. There may, though, be flexibility around phasing in and some scoping elements to bring them on board.

Calculating amount A

The OECD remains wedded to a formulaic approach to calculating the quantum of the new tax, intended to identify the portion of residual profit allocated to market jurisdictions (and using consolidated group financial accounts to do so).

The idea is to catch only the profit portion exceeding a certain level, and profit before tax in the accounts is the preferred accounting measure in assessing this.

A number of other issues remain to be determined. These include whether business line or regional segmentation is worth pursuing and whether different business models should have specific sourcing rules for revenue. Another issue to be resolved is whether weighting for different degrees of digitalisation should be taken into account (so-called 'digital differentiation', such as the digital value creation in a gaming platform versus selling perfume).

A trickier area yet to be determined is the allocation key which will divvy up the MNE's sales for the allocation of amount A to market jurisdictions Y and Z. Unless it is absolutely locked down, this is likely to be an area rife with dispute.

Double taxation

So, how is the OECD going to stop double (or triple or more) taxation arising? That is the trickiest bit of the whole endeavour. The short answer is it does not know yet, although it does acknowledge the problem. The working principle is clear enough: the entity liable for the tax on amount A should be the one earning the residual profit. When that profit is a deemed profit, however, the principle is not so easy, and the all important allocation key that gives country Y and Z their share is not the end of the problem.

The OECD sees no significant overlap between amounts A and B, and considers that disputes may be limited on amount B once it has set the fixed rates of return correctly. Given the level of existing disputes under the ALP in the world of transfer pricing, this may be naïve, unless it can hammer down every aspect of amount B. Short of developing a gigantic grid which sets the ALP for every consumer facing business in the world, this may prove difficult.

Likewise, the overlap between amounts A and C can be fixed by a suitable mechanism to eliminate double taxation, but that is going to be a tall order. Agreement between the 137 countries that a new binding dispute prevention and resolution mechanism is required for amount A is a start. Formulating such a mechanism will involve innovative approaches, however, giving it enough bandwidth to do the job and enhancing the mutual agreement procedure recently introduced. A new multilateral instrument to make this happen is almost inevitable. Otherwise, countries X, Y and Z will all be fighting over amounts A, B and C and refuse to give relief. The OECD's idea of having a review panel made up of participating countries is interesting but perhaps naïve: like the Eurovision song contest, votes will almost certainly be driven more by politics than the merits of the matter.

Next steps

The OECD and the 137 countries involved are committed to moving forward. Key policy features on the 11 areas in the revised programme of work are to be agreed by July, and a consensus reached by the end of 2020. They need to find a solution to combat mushrooming unilateral measures, as well as the US's threats of trade war, but actually getting to that solution will be easier said than done.

Outside public statements of unity, serious questions must be asked about what happens next given the US's stance on safe harbours and the sheer mountain of other issues to work through. The OECD has its work cut out.

This is based on an article that appeared in Tax Journal on 20 March 2020.

We are processing your request. \n Thank you for your patience. An error occurred. This could be due to inactivity on the page - please try again.