SAFE Policy Blog

Digital Tax: The Neglected Value of the Data

Bernd Skiera: Algorithms are of little value without data. Taking these more into account could make the taxation of digital companies fairer

Internet corporations such as Google, Facebook or Amazon pay comparatively low taxes in Germany and other European countries despite higher tax rates in other Member States, even if they make billions of euro in revenues in these countries. This low tax burden of large digital players can be explained by the shifting of “profit-adding” activities” to a location with low tax rates, such as Ireland. Through the internationally applied source principle of taxation this shift translates into a low overall tax load for the respective companies.

Politicians want to tackle this practice: At the end of May, German Chancellor Angela Merkel called for a radical reform of the tax system at the Global Solutions Summit in Berlin in order to better take into account the value of data.

The EU Commission already presented a possible tax reform in March. According to it, companies with a worldwide annual turnover of more than 750 million euros and an online turnover of at least 50 million in Europe should pay 3 percent income tax. In addition, the EU Commission proposed to relax the withholding tax principle. Thus, companies should pay taxes without physical presence in the country where their users are - and not only where the accounting value is created, i.e. at the company headquarters.

Without doubt, a debatable approach. In fact, it is hard for the taxpayer to convey that the craftsman around the corner has to pay significantly higher tax rates than the most valuable companies in the world. In that sense, I think it is good that the EU and the Chancellor Merkel are trying to increase the equality of taxation. The question is, however, whether there are other more efficient solutions.
An obvious solution would be to harmonize tax systems and rates in the European Union. This harmonization would avoid enriching individual countries such as Ireland or Luxembourg at the expense of other Member States.

Another solution could be to take greater account of the value of data in the current system of international taxation. In the current system of transfer pricing, the use of Intellectual Property (IP) plays a decisive role. For example, if certain algorithms are shifted to low tax jurisdictions such as Ireland, this shift will allow the Irish branch to charge a transfer price to other branches for the use of the algorithm, e.g. of a search engine. The actually high profits in countries with high tax rates, such as Germany, is thus greatly reduced by relatively high costs for the use of intellectual property. As a result, the company in Germany hardly has to pay taxes.

What is currently neglected in this respect is the value of data. The algorithms alone are usually not worth much without user data. The same applies vice versa: value is created primarily by the fact that algorithms and data are used together. This further implies that if the data originate in Germany, then – in order to stay in the example – the Irish branch would also have to pay transfer prices to the branch in Germany.

Such data is worth a lot. This can be seen in the prices that external companies are willing to pay for them. An example: In a study with my colleague, Professor Nadia Abou Nabout (today at the Vienna University of Economics and Business) and my former doctoral candidate, Dr. Markus Lilienthal, we have shown that companies paid a price of 8.67 Euro per click for search terms from the financial services industry in search enginge advertising in Germany in November 2012. Therefore, the obvious question is whether the price for the dissemination of this information in the calculation of the tax base has been sufficiently taken into account.

In my view, it would therefore make sense to check whether and to what extent branches of companies pay transfer prices for data. Of course, this cannot be conclusively assessed without insight into the respective tax returns. The low profits of the German branches suggest, however, that such transfer prices are paid not at all or only to a negligible amount. Greater consideration of this intra-group use of data could therefore lead to fairer taxation of digital companies. Another advantage: Such a reform would be relatively easy to implement under the current tax system.

Bernd Skiera is Professor for Electronic Commerce at Goethe University, Frankfurt.

More about this topic: Nadia Abou Nabout, Markus Lilienthal and Bernd Skiera: Empirical Generalizations in Search Engine Advertising, Journal of Retailing, Vol. 90, No. 2, June 2014