The Netherlands is a world leader in the market for tax planning by multinational enterprises (MNEs). Of all foreign direct investment (FDI, stock) in G-20 countries, 23% is owned by conduit companies located in the Netherlands, which also accounts for 12% of worldwide royalty receipts and 13% of worldwide royalty payments. Much of the money involved in these figures consists of the worldwide earnings of US-based MNEs, which is channelled to tax haven jurisdictions through EU Member States such as the Netherlands, Luxembourg and Ireland. But there are increasing complaints that the Netherlands is also being used for reducing MNE tax burdens in developing countries. For example, more than 40% of FDI in Brazil is owned by Dutch investors; almost 30% in South Africa.
The OECD/G-20 initiatives to curb aggressive tax planning by MNEs (2015), followed up by an EU Anti-Tax Avoidance Directive (2016) implicitly have the Netherlands and Luxembourg as their principal targets. However, neither the OECD nor the EU has proposed to directly tackle the tax-driven flows of money within MNEs. The technique is available: a conditional withholding tax on intragroup payments of interest and royalties. This technique has been contemplated both within the OECD and in EU policy circles. It deserves closer attention.
Dutch international tax policy has traditionally been aimed at low withholding taxes. It has no withholding taxes on interest and royalty payments, and its withholding tax on dividends is usually reduced to (almost) zero in its extensive network of bilateral tax treaties. The idea was that resident MNEs should not be hindered by tollgates in reallocating their worldwide earnings. But the effect of this policy is that the Netherlands is now the leading hub for income flows within MNEs worldwide.
A conditional withholding tax has the potential to considerably reduce tax-driven income flows within MNEs. At a rate of, e.g., 15 percent, it would only be applied by state 1 on firm A1’s payment of interest or royalties to (connected) firm A2 in state 2, when firm A2 enjoys a low or zero tax burden on that income in state 2. Therefore, it directly reduces MNEs’ benefits of tax planning. An alternative form is to apply the tax only when the recipient is a resident of a blacklisted tax haven. This approach, however, ignores the existence of preferential tax regimes in non-tax haven countries.
The idea of a conditional withholding tax has two eminent advantages in comparison to other anti-avoidance measures. One is that it is largely self-enforcing. Once some countries start to make their withholding taxes dependent on other countries’ effective tax, those other countries have an incentive to increase their tax. And indeed, the US has recently adopted the idea as an element of its tax treaties policy. The second advantage is that it considerably reduces the conceptual difficulties in defining “aggressive tax planning” (or, in legal terms, “abuse of the law”). It does so by setting a quantitative minimum standard for MNE tax burdens: the rate of the conditional withholding tax.
A first step for the Netherlands would be to adopt withholding taxes on interest and royalties. These can then be made conditional in its bilateral tax treaties. Political support has been growing over the past few years; the idea now needs to be shaped in more detail. Further background and discussion of EU legal aspects can be found in our paper and another blog aimed at a US readership.