17 October 2019
The Brexit Plague
By John Watson
Let your bells toll! Bring out your dead! A new plague spreads across the world. This time, however, the victims do not come out in lumps. They just lose their minds and assume that a few vague statements of general principle are a substitute for properly worked up proposals. We have seen it in the Brexit debate (remember that incisive analysis “Brexit means Brexit”?). Now it has spread to the OECD and their attempts to ensure that the profits of multinationals are properly taxed, where the consultation paper published by their Secretariat on 9 October leaves plenty to be desired.
Like those wrestling with Brexit, the OECD are under pressure from a number of directions. The first is that the problem is an urgent one. Because the framework set up in the 1920s taxes profits where the work or assets which generate them are situated, a multinational group selling to customers in the UK via the internet can keep its tax bill very low. It pays little or no tax in the UK because the activities and assets generating the profits are abroad. It pays little tax abroad because it has carefully placed those activities and all its intellectual property rights in a jurisdiction which hardly taxes them at all. Bingo! A low tax bill! A big fee to its advisers!
There are lots of variations on this theme. Sometimes no tax is paid in the UK, or the jurisdiction in which the goods or services are sold. Sometimes there is just limited tax because only certain activities are carried on there. Still, in an age where people are expected to pay taxes, it is all unacceptable and something needs to be done.
The next pressure is that it needs to be done quickly. The more physically aggressive jurisdictions, the UK among them, have said that if the OECD does not do something international they will themselves legislate in relation to services and goods sold to their citizens. To keep order, the OECD needs to get in first and that means it has to have workable proposals on the table sometime next year.
Third, the OECD’s proposals have to be accepted by its members who have come up with a number of different ways of solving the problem. The system proposed in the consultation document draws on three of these to put forward a “unified approach”.
As far as generalities are concerned there is a certain amount of low hanging fruit. The obvious approach is to tax groups with highly digitalised businesses in the jurisdictions into which they sell whether or not those groups have a physical presence in those jurisdictions. Then, as the target is large “consumer-facing” businesses, there should be a threshold. It has been suggested that the new regime would only apply where revenue is over €750 million. Thresholds too by reference to the jurisdictions into which sales are made and which would collect the tax. Certain industries, such as those relating to extraction and commodities, would be carved out. Etc, etc, etc…
That is all very nice, of course, but as with Brexit the real question is whether the system will work properly and that is a matter of mechanism. The starting point is that the tax should be an overlay on the present system. Rather than replacing the general principle that profits are taxed where they are generated, the idea is to allocate the group’s “deemed residual profits” between the jurisdictions into which the group sells, presumably on the basis that some figure more or less equivalent to those deemed residual profits would otherwise avoid tax altogether or at least bear very little of it.
Here everything depends upon how the deemed residual profits are determined and that is to be done by taking the profits of the group and deducting so much of them as can be regarded as “routine”. That would be a splendid approach if it was clear how routine profits are to be computed but alas that is not the case. The appendix to the paper indicates that they could include profits attributable to activity but that at least some of the profits attributable to trade intangibles, capital and risk would be excluded. It is all far from obvious how it will work and the reader is left with the uncomfortable feeling that these crucial parts of the system simply haven’t been thought through.
It is interesting to contrast the vague approach here with a concern which is tackled elsewhere in the paper. For those profits which are taxed under the conventional transfer pricing rules it is proposed to introduce fixed returns for particular activities, presumably on the basis that in negotiations as to market pricing the taxpayer always has a huge advantage over the fiscal authority. If that is the case with the ordinary transfer pricing rules, God help the fisc when it tries to negotiate the rules necessary for agreeing non-routine profits with the likes of Facebook and Google.
Quite apart from the practical difficulties, the OECD’s approach of dividing the profits of a digital group between routine profits and other profits, with the latter being the subject of the new tax, needs to be considered from the point of view of double taxation. The broad assumption behind the new tax is that those other profits would otherwise escape tax altogether or at least suffer it at a very low rate. In fact that may not be the case. Suppose that the entire group was set up in the US or the UK with customers in lots of different jurisdictions. Then the “deemed residual profits” would already have borne US or UK tax before any charge was levied under the new rules; to tax them again in the customer jurisdictions would result in double taxation. It follows, and this is acknowledged in the paper, that an elaborate code for tax credit relief will have to be included in the system and that, in turn, suggests an altogether simpler idea.
Why not allocate all the profits of the digital group between the customer jurisdictions according to the relative sizes of the customer base? The profits so allocated to a particular jurisdiction would bear tax at the local corporation tax rate but subject to credit for any tax which the multinational had already borne in the jurisdictions in which those profits were generated. If the multinational was based in the UK where the rate of corporation tax is 19% and sales are to Ruritania where the rate is 25%, the Ruritanian authorities would take the balance of 6%. If the rate in Ruritania was only 15%, there would be no further tax to pay there.
It is not a perfect system and in our example could create pressure on the Ruritanian authorities to push up their tax rate. Some sort of maximum rate might have to be applied to deal with that. Nevertheless it would mean that all digital profits received in a jurisdiction suffered tax at a minimum level. It would also get rid of the difficult concept of “routine profits” with its opportunity for multinationals to exploit the superiority of their negotiating abilities over that of the fiscal authorities and rule setters.
Above everything, however, this would be a much simpler system and a great deal easier to understand. Paragraph 10 of the paper states that the three leading alternatives submitted to the OECD “search for simplicity, stabilisation of the tax system, and increased certainty in implementation”. Quite so. The OECD should do the same.