(Un)Intended Effects of Preferential Tax Regimes: The Case of European Patent Boxes

The development and diffusion of new technologies scores high on the policy agenda of the European Union (EU) as well as its member states. The EU seeks to limit the technological gap towards other major economies. This manifests in the rapid growth of the digital economy that has evoked an intensified interest on the part of the EU and its member countries not only to foster technological innovations, but also to compete for innovative, internationally mobile firms.

In general, governments may want to accomplish this goal by providing a privileged (tax) treatment, such as subsidies or super deduction for firms’ research and development expenditures. Over the last decade, an additional tax instrument - so-called patent boxes - has rapidly diffused in the European Union. [1]  Patent boxes do not provide tax benefits related to inputs required for the creation of innovations, but grant a preferential tax treatment for the income earned from successful innovations and the related intellectual property (IP) rights.

Patent boxes are discussed controversially. Critiques of the patent box regime, foremost the OECD, classify the patent box as a “harmful tax practice”, because it may offer profit-shifting opportunities for multinational enterprises (MNEs). MNEs might use the infrastructure of (high-tax) countries for developing IPs, which they then shift to a country with a patent box, where the income from that innovation is taxed at a preferential rate.[2] The OECD estimates that a substantial amount of tax revenue is lost due to base erosion and profit shifting of MNEs, amounting up to US $240 billion annually (OECD, 2015). Against this background, the question arises to which extent patent boxes facilitate profit shifting of MNEs and whether countries with a patent box indeed benefit from the associated profit shifting behaviour of MNEs.

In a current research project, Koethenbuerger, Liberini, and Stimmelmayr (2019) analyse the effects of the preferential tax treatment of IP income in European countries. They provide a comparison of pre-tax profits of firms in patent box countries – before and after the introduction of patent boxes – and provide the comparison for two types of firms: affiliates that belong to a MNE and those that belong to domestic conglomerates. Only the former type of firm is able to use the patent box for international profit shifting. Changes in pre-tax profits of the latter type of firm are indicative of productivity effects of patent boxes.

The results show that patent-box countries benefited from an increase in income that MNEs report in these countries. Treated affiliates of MNEs report on average 8.5 percent higher profits. However, this estimate cannot solely be attributed to profit shifting into the patent box country. Affiliates of domestic conglomerates report on average 3.5 percent higher profits, after controlling for input choices. To the extent that the magnitude of the productivity effect is representative for the affiliates of MNEs, the estimate uncovers the MNE's actual amount of profit shifting of 5.0 (= 8.5 - 3.5) percent of pre-tax profits.

Still, this estimate of profit shifting might not fully reflect the response of MNEs to shift profits into the affiliate in the patent box country. When looking at the alternative measure where interest payments are not deducted (the EBIT measure)[3], a higher responsiveness of 11 percent is found. The latter estimate shows that the pre-tax profit change of 5.0 percent also captures some shifting of profits out of the treated affiliate through higher interest payments among affiliates of the MNE.  Decomposing the profit shifting response, the plain estimate of profit shifting into the patent box amounts to roughly 15 percent of pre-tax profits.[4]

The findings suggests that the location of the income in the country with a patent box is largely tax-driven and not in line with non-tax reasons for the location of income. MNEs face an incentive to shift income into the patent box country and at the same time use interest payments on internal debt to shift income out of the affiliate in the patent box country.[5] This allows MNEs to save on taxes because of the dual corporate tax rate structure: the inflow of profits is taxed at the patent box tax rate while the outflow via interest payments is subsidized proportional to the higher, regular corporate tax rate. Perhaps surprisingly, the net tax revenue effect of the overall profit shifting response might be negative. Illustrative calculations show that the critical tax differential, which renders the tax revenue change negative, is not implausibly high compared to observed tax rates in patent box countries.

References:

• Financial Times (2017), Ikea's Complicated Tax-Driven Structure faces EU Scrutiny, published 18 December 2017 (https://www.ft.com/content/d332c7a0-e3fb-11e7-97e2-916d4fbac0da. Accessed: 12 June 2018)
• Koethenbuerger, M., Liberini, F., and M. Stimmelmayr (2019), (Un)Intended Effects of Preferential Tax Regimes: The Case of European Patent Boxes, ETH Zurich.
  The paper is a revised version of the working paper: Koethenbuerger, M., Liberini, F., and M. Stimmelmayr (2018), Is it just Luring Reported Profits? The Case of European Patent Boxes, CESifo Working Paper No. 7061.
• OECD (2015), Base Erosion and Profit Shifting Project, 2015 Final Reports, Information Brief.

[1] Mostly after 2007, several European countries including Belgium, Cyprus, France, Hungary, Ireland, Lichtenstein, Luxembourg, Malta, the Netherlands, Portugal, Spain, Switzerland (Nidwalden), and the United Kingdom implemented patent box regimes. The 2017 US corporate tax reform `Tax Cuts and Job Acts' (TCJA) has also introduced a patent box regime for the US tax system.

[2] To limit the use of a patent box for profit shifting purposes, the OECD developed the so-called modified nexus-approach, which aims at allocating the taxable income from successful innovation to the country of IP development.

[3] Earnings before interest and taxes (EBIT) include interest payments, so that differences between pre-tax profit and EBIT changes allow us to detect the impact of other channels of profit shifting, which are unintended consequences of the policy change.

[4] The estimate of 15 percent cannot be compared directly with recent estimates of profit shifting because the treatment effect relates to availability of patent boxes, while existing analyses typically relate the response to tax rate differentials. Once adopting a similar specification by computing the (weighted) tax rate change due to the patent box regimes, which is roughly 20 percentage points, we arrive at a pseudo semi-elasticity of 0.75. The estimate is in line with estimates reported in recent studies on profit shifting, where the consensus estimate is 0.8.

[5] The finding is in line with anecdotal evidence on how IKEA uses the tax benefits offered by the Dutch patent box, where the affiliate in the Netherlands internally acquires IPs to access the patent box benefits. The internal transaction is financed by internal debt with relatively high interest payments to the affiliate in Liechtenstein (Financial Times, 2017).