Netherlands: The EU General Court’s judgement in Starbucks’ state aid case

In the Starbucks case, the European Commission (EC) concluded an APA issued by the Dutch Tax Authorities in 2008, providing a selective advantage to a Dutch entity (SMBV) of the Starbucks group. According to the EC, the APA resulted in a reduced taxable profit for SMBV and, as such, in unlawful state aid, based on article 107 of the Treaty on the functioning of the European Union (the Treaty).

The EC considered that the TNMM, which was used to determine an arm’s-length remunera­tion for SMBV, did not provide arm’s length pricing and that the CUP method should have been used. The EC argues that, due to the incorrect application of the arm’s length principle as included in the APA, SMBV’s taxable profit is reduced in a way which is not in line with standalone companies, the taxable profits of which are determined by transactions con­cluded on market terms. The Netherlands appealed against the EC’s conclusion and disputed the finding that the APA resulted in a selective advantage for SMBV. In addition to challenging the detailed findings of the EC, they also argued that the EC was not entitled to assess the application of the arm’s-length principle to conclude whether state aid had been provided, based on article 107 of the Treaty.

The Court’s judgement

On 24 September 2019, the General Court (the Court) concluded that the EC was entitled to apply the arm’s-length principle, as incorporated into the Dutch tax system, as a criterion for assessing the existence of state aid. The Court furthermore concluded that although the EC had criticised the transfer pricing methodology used, it had not demonstrated that this had resulted in a selective advantage for Starbucks. Since the EC has not been able to prove the existence of state aid, the Netherlands is not obliged to recover EUR 30 million in unpaid taxes due to the “benefits” as a result of the issued APA.

The Court stated that simply not applying the most appropriate transfer pricing method does not necessarily result in a reduction in taxable profit. Furthermore, the Court argued that the EC would have had to demonstrate that the methodological errors identified in the APA meant that a reliable approximation of an arm’s-length outcome could not be reached and that they resulted in a reduction in the taxable profit.

The EC’s conclusion in the Starbucks case was controversial, as some thought it had crossed the line into interfering with the application by a Member State of its own tax rules. As the OECD stated in the TP Guidelines, “transfer pricing is not an exact science”, so there is no single correct answer.

Removal and next steps

The Court’s judgement is partly good news for corporates as it should give greater certainty to those that have conducted a proper transfer pricing study and obtained an APA. It con­firms that an APA cannot be challenged simply because the EC thinks a different methodolo­gy should have been used.

Both the EC and the Netherlands have the opportunity to appeal against the Court’s judge­ment at the European Court of Justice. Given the response of the Dutch Minister of Finance to the Court’s judgement, it seems unlikely that the Netherlands will take this opportunity. To date, the EC has not announced whether it will submit an appeal against the Court’s judge­ment.