A multinational corporation operates in multiple tax jurisdictions and at the end of the day earns profits- at times enormous profits. The real issue is how the profits so earned will be taxed. Who will tax them and how much? These are very basic questions and one wonders that an international tax system that is operating for over a century should have fixed the rules by now. The intense discussion and debate of the past decade or so in the context of the BEPS project indicate that that is not the case at all and there are still gaps and holes, often deliberate and it is particularly important for developing countries to properly understand the nuances and press for fair solutions.

In the good old days when a company operated trans-border, due to distance and time involved, the options available were quite limited. It could do so essentially in two /three forms. It could form a subsidiary in the local jurisdiction and pay taxes on its domestic profits at the normal tax rates. But then nothing prevents the subsidiary to deal with other parts of the MNC and in such circumstances, it is easy to divert local profits to jurisdictions where taxes were low or non-existent. To tackle such issues, the Arm's length principle was invented. All the transactions of the local subsidiary with the other parts of the MNC must be at arm's length as if the transacting entities were separate and distinct from each other, while essentially the MNC is a single entity interested in multiplying its overall profits.

The other ways the MNC could operate was by establishing a fixed place through which the business would be carried on and a third way was to appoint an agent to do all the functions for the enterprise. A fixed place and agency thus became the principal pillars of the concept of permanent establishment. With the expansion of global trade, multinationals soon figured out that they could mix and match, form subsidiaries in some places, open liaison offices in others, while appointing agents who did not have the authority to bind the principal in other places. In the case of manufacturing, goods could be made through contract manufacturers.

The rules of the international tax game designed by the OECD and its predecessor organisations that were the main providers of capital were and still remain such that foreign business can be taxed in a market jurisdiction only if there is a Permanent Establishment (PE) and the rules for finding a PE were made even more difficult than finding a needle in a haystack. The problem gets more complicated with the coming of age of the digital way of doing business and the traditional pillars of international taxation naturally came under severe strain.

Realising the difficulties, over the years the OECD, very reluctantly made some concessions, the most important of which is the exception for construction business where a PE was deemed to exist if the activities went on for more than a certain number of days in a year. Some concessions were also made in the preparatory and auxiliary exemptions. Besides, many of the changes were made not in the rules but in the Commentary (including an optional provision of a service PE) that were to be followed by the OECD Member countries. This raised the issue of the feasibility of such a concept to be followed by non-OECD member countries who by and large can be equated with the developing countries. The UN Model, initially meant to be a model suitable for developing countries, added a few more categories to the PE but mostly followed the OECD Model and the OECD officials always seemed to control the UN through purse strings and through the intellectual inputs on the ground that OECD had years of experience in administering the international tax architecture and their accumulated wisdom should be made use of.

Till the last decade, this system somehow held together and developing countries continued to get ripped off their genuine share of revenue as there was no coherence in their approach in negotiating tax treaties and they were scared of foreign investment. Signing very one sided tax agreements were touted to them as the passport to the much needed capital. Things began to change following the financial crisis and even western states looking for additional resources found out that multinationals particularly from the USA and also particularly those operating in the digital sphere were gaming the system, often times not paying taxes anywhere. Thus began the OECD BEPS project which then got sanctified by other G-20 member countries. And some of the developing countries got a say in the standard setting, limited though it might have been. Most of the developing countries and even some of the newly inducted OECD member countries are too respectful of the existing system and do not try to question the rationale of the existing rules. However, debates became necessary; particularly in the context of action plan 1 relating to digital economy. The delegate(s) from India actually did a commendable job in putting forth the irrationality of the existing system and in pressing for change. The division within the OECD also helped and we had the equalisation levy and subsequently the significant economic presence in the domestic law, primarily as a result of the deliberations in the task force on digital economy.

The Equalisation levy was mercilessly criticised by most of the Indian tax practitioners projecting various doomsday scenarios. The bogey of foreign investment is often bandied about. It was the same story for the significant economic presence test. The fact however is that many countries in the developed world actually followed India and adopted similar measures. It is true that these are still very hot topics and consensus eludes. Nevertheless, the home grown nay-sayers, whether acting as proxy for other interests or not became somewhat muted. The significant economic presence test, particularly in the non-treaty situations is also not operational since the rules of profit attribution to the business connection have yet to come in force.

These voices have become active again when in April 2019, the CBDT put out a public discussion draft on the rules for attribution of profits to permanent establishments generally and also the rules for profit attribution to significant economic presence. In the discussion draft the working of the arm's length pricing method as means of profit allocation to a PE has been questioned. Once again, one comes across fervent appeals from transfer pricing practitioners beseeching the CBDT not to jettison the Arm's length standard.

In this context, it may be worthwhile to note the reasoning given in the draft report. It is also important to note that the report is only at the draft stage and suggestions have been invited from the stakeholders. It is also not the case that there are no problems in practical implementation of the solutions as proposed in the draft. For the present though, it is important to note the shortcomings of the existing system and how it works against the interests of the developing countries for an intellectual justification for proposing an alternative.

To recapitulate, under the OECD devised rules of international business taxation, the first obstacle to be crossed is the concept of a permanent establishment. Ever since, India became an observer at the OECD, India has been consistently voicing its differences over the OECD interpretation and more than 70 reservations have been expressed and most of them relate to PE provisions and commentaries thereon, particularly in relation to e-commerce and service PE.

Even after crossing the first obstacle, if a PE is found, there is the question of how much of the profits can be attributed to a permanent establishment. Here, it is interesting to note that the OECD member countries have been debating amongst themselves for years before the report relating to profit allocation to a PE was published in 2008. This was then made part of the OECD 2010 Model. The 2010 Model has now adopted what is known as the Authorised OECD Approach or AOA. Some Indian practitioners swear by the AOA. OECD Member countries may be bound by the "authorised" approach but why should other countries unquestioningly adopt the AOA particularly when the consequences of the same for the revenue of the developing countries are definitely negative.

Here, the discussion paper very lucidly explains the flaws of the AOA and how it further reduces the taxable pie of the market economies.

The basic rule of attribution of profits to a PE is contained in Article 7(2). In this context, it will be instructive to compare some aspects of the 1978 OECD Model Article 7, which with minor changes continued till 2008, with the current Model.

Now when the rules of international taxation were being drawn up it was expected that PEs would maintain books of accounts in the jurisdiction they were operating and it would be possible to find out the profits therefrom. In fact, the 1978 Commentary on Article 7(2) in paragraph 11 stated: "In the great majority of cases, trading accounts of the permanent establishment - which are commonly available if only because a well-run business organisation is normally concerned to know what the profitability of its various branches- will be used by the taxation authorities concerned to ascertain the profit attributable to that establishment… Thus when books of accounts are available, the profits disclosed in the accounts subject to adjustment should be adopted."

However, contrary to these assumptions, in majority of cases, the basic problem in profit attribution to PEs arises because companies in the first instance always claim that there is no PE in the source country and hence there is no obligation to maintain any separate books of accounts. Subsequently, if it is found that a PE actually exists, then there is the further question of what profits are to be attributed to a PE.

A preliminary look at Article 7(2) will suggest that even though a multinational enterprise is a single enterprise, a separate entity approach has to be applied for the purpose of profit allocation. This has been interpreted to mean applying the arm's length principle. It is in this connection, that the discussion draft released by the CBDT exposes the common fallacy of conflating arm's length principle with arm's length price. It has been pointed out that the Arm's length principle requires the existence of entities that are completely independent of each other, and thereby able to fulfill their primary objective of maximizing their own profits, without being affected by the impact of their actions on the profits of another enterprise. Since this is a condition that seldom gets satisfied in the real world, it would generally require that a PE that is not acting for maximizing its own profits is compensated for all unfavourable obligations that adversely affect its profits. In most of the cases, this is not done and hence it may not be possible to apply Article 7(2) and profits attributable to a PE may have to be determined by some other method, most notably by taking recourse to Article 7(4).

Now, as can be seen from the table, in the revised version of the OECD Model, the determination of profits of the PE is specifically subject to Article 7(2) and Article 7(2), in turn, specifically refers to the functions, assets and risk analysis (FAR). OECD has successfully explained to the world that this is a logical extension of the doctrine of separate entity as was always found in Article 7(2) and therefore the transfer pricing guidelines relating to arm's length price determination in the case of related entities should be fully applied in the PE situation also. Consequently, the apportionment method that was found in OECD's own Model till 2008 was done away with.

CBDT's discussion draft then lays bare the invidious side effects of adopting such an approach particularly from a market country's perspective.

As explained in the report and as common sense dictates, when one is concerned with taxation of business profits, it should be ascertained as to what the elements of such profit are. Profit is quantum of sales less cost of sales. Sales represent the demand side and as a corollary, the market economies. Cost of sales represents the supply side. In a purely domestic situation, roughly speaking sales-cost of sales will be profit on which tax is levied. In a cross border situation also, that should be the main principle. In fact, in the pre-2010 OECD Model, in so far as it related to profit allocation, both the demand side and supply side factors were taken into account for apportionment of profit.

However, this was changed in the post 2010 model without any discussion with or concurrence of the developing countries whose markets are up for exploitation. India, of course, has voiced its opposition in strong terms but that is of no significance for the OECD. As pointed out in the discussion draft, the current OECD prescription of profit allocation to a PE solely on the basis of FAR, is to ensure that taxation of profits take place solely on the basis of the contributions made by supply side factors while completely ignoring sales and thereby the contributions made by the maintenance of the markets and the demand side factors to profitability of a foreign enterprise. By attributing profits only on the basis of FAR, representing supply side factors, and excluding sales from the equation, the contribution of market jurisdictions to the profits derived by an enterprise from that jurisdiction stand completely ignored.

The CBDT paper gives economic and expert analysis to support its proposition and is a valuable addition to the thinking on the subject. It is in this context, that the report suggests an alternative of fractional apportionment of profits based on practice in India as also based on the current discussions in the EU and also in the USA.

The Committee did not support the full formulary apportionment approach, but considered that fractional apportionment with equal weightage being given to sales (representing demand) and manpower and assets (represent supply including marketing activities). This is permissible under article 7 of the treaties that India has signed and is also sanctioned by Rule 10. The Committee also observed that where some profit of the enterprise is already taxed in India as in the case of a dependent agent, then that part should be excluded from the profit to be allocated to the PE.