The agenda on digitalization has evolved swiftly and 2019 has been particularly busy for the Organization for Economic Co-operation and Development (OECD). In January 2019, the OECD published a policy note that finally endorsed the need for examining the allocation of taxing rights. The work under action point 1 was divided into two core elements, one relating to the allocation of taxing rights and the other addressing remaining base erosion and profit shifting (BEPS) issues. A departure from expressed intent of the BEPS program, it was a response to the increasing preference for a turnover based digital tax.
India was among the first to implement an equalization levy at the rate of 6% on business-to-business (B2B) transactions (this was introduced through the Finance Act, which is not an amendment to the Income Tax Act 1961: therefore it would not be affected by the treaty provisions). The tax was heavily criticized for its impact on consumers, with the likely pass forward by digital monopolies. Revenue collected from this tax in 2017–18 was $97.5 million which is 0.15% overall revenues from corporate tax. However, the levy is applicable only to online advertising, explaining in part the small collection.
To further signal India’s stance, the test for significant economic presence (SEP) was introduced in 2018 through an amendment of the definition of business connection in the Income Tax Act 1961 (Explanation 2A was added to Section 9(1)(i)). This had no impact since the law has not been operationalized by setting of thresholds and neither will it take effect unless the treaties are appropriately amended.
Further, the adoption of the new nexus rules or SEP in India is an incomplete reform, since rules of attribution continue to remain a challenge, even if the SEP takes effect. A contemporaneous amendment of these rules is required.
To build on the existing work, the Central Board of Direct Taxes (CBDT) released a document for public consultation on attribution of profits to a PE. The report sought comments on the use of fractional apportionment that would duly recognize the contribution of the demand side or the market jurisdiction. The report made suggestions on potential formulas for apportionment, recognizing that it was a work in progress.
Similarly, other countries are now considering either a withholding tax, a digital services tax or virtual permanent establishment (PE); the virtual PE being a test for economic presence based on the number of users, contracts and/or revenues. In 2019, 42 countries had either implemented or were considering implementation of such measures.
OECD “Unified Approach”
The OECD is concerned that the prevalence of these measures could undermine the existing international tax system. Nearly 200 comments supporting different alternatives were submitted in response to the public consultation following the publishing of the policy note.
Following up on the consultation, in May 2019 the OECD released the programme of work which listed the various proposals. India’s proposals—new nexus and fractional apportionment—were prominent among the modified residual profit split and distribution approach. Later, in October 2019, the OECD followed up the report with the “unified approach.”
The unified approach was an attempt at identifying the common elements of the various proposals and incorporating such elements in an intricately layered process of allocation. This was admittedly to bridge the gaps in the various proposals which, if left unaddressed, would encourage uncoordinated measures (paragraph 7, Secretariat Proposal for a “Unified Approach”, OECD).
Amounts A, B and C were identified as those associated with the new taxing rights, the fixed returns to marketing and additional profits associated with in-country operations above the baseline activity (page 6, Secretariat Proposal for a Unified Approach”, OECD).
The approach is an attempt to restore simplicity, stability and certainty to the taxpayer while moving beyond the arm’s length principle (paragraph 10, Secretariat Proposal For a “Unified Approach”, OECD).
To elaborate, it is being proposed that consolidated financial accounts would be prepared for a multinational group in accordance with the accounting standards of the headquarter country (paragraph 53, Secretariat Proposal For a “Unified Approach”, OECD). This would require estimating the deemed residual profits by separating the routine profits from total profits, an analogy to the residual profit split.
It is unclear at the moment how these routine profits are defined and it is suggested that amount A would seek to approximate the remuneration of routine activities on an agreed level of profitability. The market for digital platforms is commonly known to be monopolistic or oligopolistic (paragraph 16, Interim Report): a feature that has been highlighted in the context of competition law in India, for example in the Google case (Case No. 39 of 2018, Competition Commission of India) in 2018.
Further, the activities related to production or delivery of service are highly integrated. Defining a routine profit in such circumstances will be a challenge.
For example, the profits arising for a dominant tech platform may be considered purely non-routine. A fixed percentage is a simple solution (paragraph 54, Secretariat Proposal For a “Unified Approach”, OECD). Nevertheless, the definitional challenge remains and is perhaps compounded by sectoral differences. An agreement would also be required on this percentage.
Further, it is suggested that the existing transfer pricing approach would continue to apply to the routine activity.
The need for the above suggested change to allocation is presumed since at the moment no profit will be attributable to qualifying entities based on functions, assets and risks (FAR). However, the Indian tax treaties do not contain a revised version of Article 7 which prescribes attribution by way of FAR (paragraph 38, CBDT consultation draft). This seemingly conflicts with the Indian position.
Further, even if such approach is applied to digital corporations, the disputes related to routine profits will remain and the fractional apportionment to the PE would only be to the extent of residual profit.
Once the split of the profits is complete between the routine and non-routine, a portion of the latter associated with market jurisdiction (paragraph 57, Secretariat Proposal For a “Unified Approach”, OECD) would be distributed among qualifying entities based on sales (paragraph 60, Secretariat Proposal For a “Unified Approach”, OECD).
The report presents as an example social media’s non-routine profits from algorithms, software, data and valuable brands (paragraph 57, Secretariat Proposal For a “Unified Approach”, OECD). The inseparability of these elements often compounds the problem. It could be relatively difficult, if not impossible, to separate these elements in the case of machine learning algorithms that do not require human intervention and rely on good data.
Further, a clawback mechanism is being proposed for loss-making enterprises. Deep discounts are frequently offered to customers by platforms to gain traction; this has been observed in the case of India. If the losses are sustained in market jurisdictions from profits in other jurisdictions, unless there is specific legislation to prohibit these as anti-competitive, there might be a sustained clawback from the profits of the market jurisdictions.
This limitation must be considered, and for this purpose, the application of rules relating to amount B must be strengthened.
To allocate such profits, firstly a new nexus rule will be necessary to establish that the jurisdiction has the requisite taxing rights. It is being suggested that a standalone rule, a top up to the PE rule, based on revenue threshold, will be introduced. This threshold will be defined taking into consideration certain activities such as online advertising. The SEP or the virtual PE recommended by the EU have multiple threshold criteria including those based on user base and contracts.
A purely revenue-based criterion would vitiate the proposed amendment since the remote execution of sales on digital platforms could perpetuate inadequate assignment of profits based on sales.
The association between sales and revenues may have to be articulated through thorough economic analysis; however, it may well be worth looking at contracts concluded as one of the criteria. The thresholds may have to be designed well. Therefore, sales alone cannot establish nexus.
This is particularly relevant for countries such as India where there is misalignment of share in global revenues reported from digital services and that of digital buyers. India’s share in global digital buyers is close to 10%, whereas the share in global revenue is approximately 1%.
In so far as the distribution functions are concerned, it is proposed that these be treated as currently under the existing transfer pricing rules. Yet marketing and distribution remain widely contested.
To assure that dispute is minimized a fixed remuneration will be explored, since negotiated returns are expected to provide certainty. This is another element that would require extensive agreement.
It is possible that there may be disagreement related to compensation of the distribution function, and foreseeing such dispute an amount C is suggested. That is, any dispute between taxpayer and market jurisdiction over any element of the proposal should be subject to legally binding and effective dispute prevention and resolution (paragraph 30, Secretariat Proposal For a “Unified Approach”, OECD).
As for dispute prevention, advance pricing agreements are an alternative.
A legally binding dispute resolution, though not explicitly mentioned, is arbitration. However, the experience with the selection of mandatory binding arbitration (MBA) in the MLI is reflective of the potential uptake. Only 26 jurisdictions (as per country positions submitted as in February 2019) applied MBA to their tax treaties and these predominantly comprised OECD countries.
India is particularly opposed to arbitration in tax disputes and it is not likely that there will be a change in stance for the unified approach.
Lastly, carve outs are being considered in the application of approach since it is being considered that it should apply to consumer-facing businesses. However, the challenges arising from digitalization may not be restricted to only consumer-facing operations. It is being proposed that other than the extractive industries, financial services may be considered. Financial services are rapidly evolving and present significant potential for automation. Carve outs, if at all, must therefore be thought through.
The unified approach consists of many elements that require consensus as well clarity. There are general concerns that have been observed, however there are identifiable elements to which India may not easily agree.
Suranjali Tandon is Assistant Professor at the National Institute of Public Finance and Policy, New Delhi.
The author may be contacted at: suranjali. firstname.lastname@example.org
This column does not necessarily reflect the opinion of The Bureau of National Affairs, Inc. or its owners.