Transfer Pricing for Digital Platforms

Aligning transfer pricing outcomes with value creation for digital companies is at the core of the BEPS (“Base Erosion and Profit Shifting”) initiative, led by the OECD. Many of the giant digital companies which are the focus of the BEPS initiative are modeled by economists as platforms. Network effects and user participation have rightly been identified as key features of these markets in the BEPS 2015 reports. However, this article argues that BEPS 2015 reports have not drawn the right conclusions from this and are likely to fall short of their main objective.In digital platform markets, value is primarily created by users either by their work or by their sole presence. The so-called network effects (i.e. the externalities exerted by users on other users on the platform) are generated by the action of users and not directly by the action of the platform itself. The question of the attribution of the value created by network effects thus cannot be solved by functional analyses (as typically performed in transfer pricing projects): functional analyses focus solely on functions, assets and risks at the company level, but, in the context of digital platform, value is created by users, outside the scope of activity of the platform.As a result, transfer pricing rules should be amended to fully acknowledge the idiosyncrasies of digital platforms. Contrary to most papers devoted to this topic, this article does not advocate for a complete rewriting of the international tax rules and rather proposes achievable adjustments to the existing transfer pricing framework. Out of lack of better practical options, we believe that this attribution question should be solved by adjusting the international tax framework and defining a clear rule for the treatment of network effects. More precisely, transfer pricing lawmakers should decide on a rule to attribute these network effects and should clearly establish which legal entity should be entitled the profit generated by network effects. It is our opinion that taxable income generated by network effects should be located where these effects are created. This could be achieved for example by treating network effects in a similar manner as group synergies, i.e. benefits arising from these effects would stay in the country where they are generated