Glaring gaps in the new digital services tax law

There has been a lot of commentary in the media in respect of the newly introduced digital services tax. While the rationale for the introduction of the digital services taxes is understood generally, there appear to be some concerns on whether the law as passed achieves the objectives desired by the Kenya Revenue Authority (KRA).

The first issue is the scope of the digital services tax (DST) and what type of entities it applies to. The Income Tax Act, the primary legislation, states that the digital services taxes will apply to a digital marketplace which in turn is defined in the same Act as a platform that enables the direct interaction between buyers and sellers of goods and services through electronic means.

Accordingly, the digital services tax is therefore applicable to a narrow range of entities – namely, those that provide a platform for third party buyers or sellers to interact. It therefore would hold that a business that is selling its own goods on a digital platform is not providing a platform for sellers and buyers to interact but is selling its own products. Where a company does not provide its own goods but provides a platform for others to provide its services or goods to buyers, then it would be caught within the definition and liable to DST. To bring it close to home, ride hailing apps that connect drivers and riders would be considered a digital marketplace while an ecommerce venture that sells its own product online would not be a digital marketplace. This interpretation is consistent with the definition of a digital marketplace as proposed in various other countries.

However, the current draft of the DST regulations, which are secondary legislation, appears to extend the application of the digital services tax to a host of activities that do not qualify as a digital marketplace or platform providers as per the definition in the main part of the Income Tax Act. The DST regulations appear to suggest that streaming companies that are providing their own content or any company that sells goods online would be caught up under the digital services tax.

It is well established under the rule of law that the regulations of an Act cannot supersede the ambit of the primary legislation. In the case of the DST regulations, the scope of digital service tax has been expanded through secondary legislation. Given that every business has a digitised component to it – if a wide application of a digital services tax is adopted, it will inevitably end up incorrectly and illegally taxing a whole range of businesses from 2021 under the digital services tax.

The legislation also states that the digital services tax only applies to entities that have income “accrued or derived” from Kenya. The requirement that Kenyan income tax is only applicable on income that is ‘accrued or derived from Kenya’ is not new. The principle that underpins the Income Tax Act requires a company to carry out its dominant activities physically in Kenya before such income is taxed. That is why for example a Japanese car exporter sitting in Kyoto, Japan and exporting cars to Kenya is not subject to income tax in Kenya while a company selling Japanese cars in a bazaar in Jamhuri park is subject to Kenyan income tax.

The requirement that the income must have “accrued or derived” from Kenya is the very reason that a global e-commerce entity with no operations was not subject to tax in Kenya until the law was amended – because its dominant activities were outside of Kenya. If the “accrued or derived from” did not act as a limit to taxation, there would have been no need to introduce a separate “digital services tax”. The digital services tax as drafted has therefore been limited in its application by the explicit requirement in the provisions introducing the digital services tax that such a tax is only applicable to income accrued or derived from Kenya. One can argue that we are therefore no better off now than we were prior to the introduction of the digital services tax legislation.

Given the drafting of the legislation, it is therefore likely that taxpayers may mount legal challenges as to the scope of the tax as there are gaps between the law as drafted and KRA’s understanding of the law.

One final point to note. Whenever transaction or turnover based taxes are introduced in the Income Tax Act, it is likely that those taxes will not be paid by the person earning the income and will simply be pushed to the consumer. They will no longer be a tax on ‘income’ but a tax on ‘consumption’. Just like withholding taxes on nonresidents, the digital services tax may be pushed to Kenyans through an equivalent price increase and this will only increase the tax burden on Kenyans.

This phenomenon does away with the argument that the taxes are introduced for the purposes of fairness and equity by taxing nonresident digital companies – far from it, they only make goods and services more expensive for an already overburdened Kenyan taxpayer.

A possible solution for Kenya is the treaty article proposed by the UN on automated digital services. This proposal is straightforward to understand and has a clear definition of what constitutes automated digital services without necessarily expanding the scope of taxation to unrealistic limits. This proposal may be worth consideration as a stop gap measure.