Ruminating over Equalisation Levies
Kluwer International Tax Blog
May 23, 2019
Please refer to this post as:, ‘Ruminating over Equalisation Levies’, Kluwer International Tax Blog, May 23 2019, http://kluwertaxblog.com/2019/05/23/ruminating-over-equalisation-levies/
There has been much brouhaha about equalization levies in the context of the digital economy. One of the hotly debated issues is whether such levies are covered by tax treaties at all. In this post, I should like to reflect over this issue as objectively as possible. I shall not, however, delve into issues of whether equalisation levies lack good faith and therefore are tantamount to treaty dodging. Whilst important, these issues are subjective and doubtful in their applicability in countries with a dualist approach to public international law.
Why an equalisation levy?
The digitalisation of the economy is a result of the telecommunications revolution, which allows businesses to attain scale without mass. Whereas a business could not have operated without a physical shop, the use of targeted marketing online enables it to operate through a mere webshop. The requirement of a permanent establishment necessitates a taxpayer to be physically present in a jurisdiction through a fixed place, agents or personnel. That threshold does not appear to be fit for the purpose of securing taxing rights of the state of which the business is not a resident – colloquially referred to as the source state.
It may be argued that not enough has happened to tackle digitalisation since the 1998 Ottawa Conference first recognized that “e-commerce” was ushering fundamental changes to the global economy. It is not surprising that countries would wish to find ways and means to address their problem of losing revenue rather than wait for the perfect solution through consensus, which does not appear to be forthcoming. The equalisation levy is a response to this urgency felt by some jurisdictions.
Conceptually, an equalisation levy should function as a quasi-indirect tax, although the collection and payment mechanisms are unlike a VAT/GST; nor do they feature input credits. Also known as a compensatory or secondary tax, it is levied to equalise “lost” revenue on income paid to persons who may not be taxed, as opposed to who may be taxed. It is therefore levied on and collected from the payer of income, rather than the receiver. This “equalisation” is usually an approximation, and is therefore levied on gross payments. Levying it on net income would be rather cumbersome – nigh impossible – to compute in each case.
Equalisation levies, however, need not always adhere strictly to this description in practice.
India’s Equalisation Levy
India’s Equalisation Levy, for instance, appears to be an adaption of the idea. Worded as a withholding tax, one wonders if it is necessarily one.
With some exceptions, it requires certain taxpayers to “deduct” 6% of the gross consideration for certain online services rendered by non-residents. Thus if INCo (an Indian business) were to pay 1000 for online advertising services to NLCo (a Netherlands resident online advertiser), InCo would be liable to “deduct” 60, and remit 940 to NLCo.
Surely, it is possible that INCo does withhold the 60. In such a case, the tax is an equalisation levy only in nomenclature – much like the European digital services tax proposal. However, it may not always be possible for the recipient of services to withhold the tax – especially from payments made to large service providers. What would the consequences on INCo be, should it remit the entirety of 1000 to NLCo, instead of 940? Must it gross-up the levy and deposit 63.83 as the tax withheld?
According to section 166(3) of the Finance Act 2016, INCo shall, notwithstanding such failure, be liable to pay only 60 to the Indian treasury. Section 171(a) of the Finance Act provides for a 100% penalty for failure to withhold, and section 40(a)(ib) provides for the disallowance of the consideration as a deduction. However, an elaborate conversation with a very high-ranking officer of the Indian Ministry of Finance reveals that it is their policy not to enforce these consequences on taxpayers who remit the levy despite having failed to withhold the tax.
It appears that the Indian Equalisation Levy may be a withholding tax in the black letter of the law, but closer to its nomenclature in spirit. This may well have been the intention underlying its design, given that the law does not so much as refer to the recipient of the consideration as the taxpayer; it refers to the payer of consideration as the “assessee”. The recipient of income, on the other hand, is exempt from tax under Section 10(50) of the Income Tax Act, 1961.
It remains to be seen if a court of law would only follow the letter of the law, or if it would recognise its spirit in practice.
Are equalisation levies covered by tax treaties?
It has been claimed that both modalities explained above – the classical approach and Indian approach – fall outside the scope of tax treaties following the structure of the OECD Model. Before examining the accuracy of those claims, it may be useful to outline Article 2 of the OECD Model, which provides for the general framework of what kinds of taxes are covered by a tax treaty.
Paragraph 1 states that the treaty applies to taxes on income and on capital imposed by a contracting state, its political subdivisions etc. Paragraph 2 explains what are considered to be taxes on income and on capital. Importantly, it states that taxes imposed on elements of income are covered by tax treaties. Paragraph 3 allows contracting states to list the existing taxes to which the convention apply particularly, although the list is not exhaustive (although some countries may prefer to exclude the first two paragraphs and provide for an exhaustive list of current taxes covered by the treaty). Most importantly, paragraph 4 states that the treaty applies, inter alia, also to any substantially similar taxes that are imposed after the date of signature in addition to the existing taxes.
Both approaches to equalisation levies should be examined separately against this backdrop.
Examining the Indian approach
It has been India’s stated position that its Equalisation Levy does not fall within the scope of tax treaties. According the letter of the law, the levy is but a withholding tax. The mere fact that the levy is imposed on the consideration, and not the net income should make no difference on two counts. First, the gross receipts are but an “element of income”, the other elements being deductible expenses. Secondly, gross basis taxation is explicitly allowed for in tax treaties – for example, on income from immovable property, dividends, interest, royalties and fees for technical services. The Equalisation Levy should qualify as a substantially similar tax to these.
My discussions with the official also reveal that the Ministry adopts another, rather convoluted, argument to contend that the Equalisation Levy is not substantially similar to income taxes. This argument emanates from the differing scopes of the Income Tax Act, 1961 and the Equalisation Levy within the Finance Act, 2016. Whilst the former applies to the entirety of India, the latter does not apply to the state of Jammu & Kashmir. This is claimed to be the case because the Income Tax Act was enacted under powers granted by Entry 82 of the 7th Schedule of the Constitution of India, whist the latter was enacted under Entry 97 thereof. The argument made by the ministry is: the two taxes are substantially different because they have been enacted under two different entries of the 7th Schedule of the Constitution of India.
It would be useful to recall that the Indian judicial tradition is to try to follow the “international tax language” [see: Vishakhapatnam Port Trust v. CIT, 1984 (38) CTR 1 (AP)] to the extent possible. It seems unlikely that the argument that the two taxes are dissimilar merely on account of the constitutional clauses under which they were enacted would succeed. Reference must be made to the OECD Commentaries in this regard, which serve as a good guide in construing the international tax meaning of what is covered by Article 2.
Without delving into issues of Constitutional law in India, the differing geographical scopes of the two taxes should also not matter since Article 2 applies to taxes levied by the State, or a political subdivision or local authority thereof. It would be particularly odd if a tax levied only in the political subdivision of Jammu & Kashmir should be included within the scope of the tax treaty, but a tax levied in all others but Jammu & Kashmir should be excluded.
If the letter of the law is followed, it appears to be a hard sell to treat the Equalisation Levy not to be regarded as a tax covered by tax treaties modelled upon the OECD Model.
Examining the classical approach
Whilst not a typical income tax, an equalisation levy is not unknown to tax treaties. Article 10(5) explicitly restricts the imposition of a tax on undistributed profits triggered by the payment of dividends paid by a non-resident company of the state seeking to impose such a tax. This tax on undistributed profits is nothing but a classical case of an equalisation levy on dividends paid by a non-resident company.
Even though such an equalisation levy may not be explicitly mentioned in Article 2, it is clear that it is covered by the tax treaty. It would seem that equalisation levies on non-dividend payments should qualify as substantially similar taxes to the one covered under Article 10(5). However, it is one thing for a tax to be “covered”, but quite another for a distributive rule to “deal with” the item of income or capital on which such a tax is imposed. An equalisation tax on dividends is dealt with in tax treaties (and that too in a very limited scenario) because Article 10(5) specifically addresses it. It is unlikely that the equalisation levies aimed at addressing the challenges of the digitalisation of the economy could be said to have been dealt with by the distributive rules of a tax treaty, despite being technically “covered”.
It has been near-universally acknowledged that equalisation levies are not desirable in the long run, and have been proposed or adopted only as interim measures. Interim measures, however, run the risk of metamorphosing into permanent features of the law should a long-term solution not be found soon enough. Perhaps it may be useful for the OECD to devise a provision along the lines of Article 10(5) to deal with equalisation levies more generally.
All views are personal.