Corporate taxation and the quality of research and development

See paragraph 4.40 of HM Treasury, Pre-Budget Report 2009, December 2009. The UK regime will be phased in over four years. In 2013, companies were only entitled to 60 % of the full benefit, which will increase to 70 %, 80 %, 90 %, and 100 % in subsequent years (see Evers et al. 2013).

Empirical evidence confirms that the social returns to R&D substantially outweigh their private returns (see e.g., Griliches 1994; Jones and Williams 1998; Griffith et al. 2004).

Note that recent survey evidence suggests that the average duration of R&D projects until the patent application is rather short, less than two man-years in the large majority of cases (see Sadao and Walsh 2009). As tax reforms are commonly announced months or even years in advance, MNEs that decide on the location of R&D projects and form rational expectations on future tax rates likely obtain a quite precise predictor for the tax burden on the future income stream from the projects. The introduction of the UK patent box regime in 2013 was, e.g., announced in 2009 already. A number of companies declared to boost investment activity in the UK after the government announcement (see e.g., GlaxoSmithKline’s press release on November 29, 2010: http://​www.​gsk.​com/​media/​press-releases/​2010/​government-patent-box-proposals-transform-uk-attractiveness-for-investment.​html). On top, besides the relocation channel spelled out in the main text, the quality of R&D projects may respond to changes in corporate taxation since high corporate taxes reduce the incentive to exert effort and may thus lower the quality of R&D. This applies for owner-managers, as well as for dependent employees if the latter own shares or stock options in the company or owners adjust incentive contracts in response to tax changes.

While most countries tax patent income at the statutory corporate tax rate, some implemented special low tax rates on patent income (see also Sect. 2).

Since tax authorities commonly do not grant tax refunds, this holds true as long as affiliate profits are high enough to ensure the affiliate can exploit the full value of the tax allowance/credit. If the value of the tax allowance/credit exceeds profits, incentives to select high-value R&D projects to countries with attractive R&D tax credit and allowance schemes may prevail (See Ernst et al. 2013).

In a recent paper, Ernst and Spengel (2011) report a faint impact of R&D tax incentives on the number of corporate patent applications. Buettner and Wamser (2009) find positive effects of R&D tax incentives on the volume of foreign direct investment (FDI) of German multinationals.

The EPO is not a body of the European Union and, as a result, the states which form part of the European Patent Convention (the legal basis for the EPO) are distinct from those in the European Union. See: http://​www.​epo.​org/​about-us/​epo/​member-states.​html.

Note, however, that in some countries the principle of economic ownership applies, implying that legal and economic owner may diverge. In such cases, information on the legal owner can be used as a proxy for the party that will eventually be subject to taxation only.

The employer can also transfer ownership of the invented technology to the employee who may then file for patent protection. We consider these cases to be rare events though as firms do not have an incentive to waive ownership of a technology that is expected to earn positive income.

Following previous efforts (see e.g., Abramovsky et al. 2008), the name of the AMADEUS firm was matched to the name of the applicant on the patent application. Note that corporations may take on the role as patent applicant, while patent inventors are necessarily non-corporates.

See Hall et al. (2007). We are grateful to Grid Thoma for providing us with this data.

Forward citations have an important legal function in the sense that they limit the scope of property rights which are awarded to a patent. In the case of EPO patents, inventors are not required to cite prior technology used in the development of their patent, but the references are added by patent examiners. On the one hand, this implies that not necessarily all innovations which draw on an existing patent in fact acknowledge the reference. On the other hand, an external patent examiner has the benefit of following a consistent and objective patent citation practice.

For the purpose of guaranteeing a reasonable level of precision, the construction of the quality measures accounts for an eight-digit IPC classification reported in the patent document.

There were a few exceptions to the credit method. If no double tax treaty was in force for a specific country in a specific year (especially in the 1990s) the unilateral method to avoid double taxation was applied to calculate the effective income tax rate, e.g., deduction of the foreign withholding tax.

The information on annual bilateral royalty payments used for the construction of the index was obtained from the IMF. Information on countries’ GDP was obtained from the World Development Indicators Database.

Note that, as our main data set comprises patent applications to the EPO, the construction of \(t_\mathrm{e}\) assumes that the firm filed for patent protection in all countries that were members of the EPO in the year of the patent application. For our first sample year 1995, this comprises the countries of Austria, Belgium, Switzerland, Germany, Denmark, Spain, France, United Kingdom, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, and Sweden. Countries that became members during our sample period are: Bulgaria (2002), Cyprus (1998), Czech Republic (2002), Estonia (2002), Finland (1996), Hungary (2003), Island (2004), Lithuania (2004), Latvia (2005), Malta (2007), Poland (2004), Romania (2003), Slovenia (2002), and Slovakia (2002). Furthermore note that royalties paid from parties in the host country of the patentee itself are also accounted for in the construction of \(t_\mathrm{e}\), in the cases where the weighting matrix reflects the country size distribution and the construction is based on the countries in which the firm filed for patent protection. It holds \(tw_{jk}=0\) if \(j=k\).

Note that AMADEUS contains ownership information on a worldwide basis. For most subsidiaries and parents outside Europe, accounting information which allows to proxy for subsidiary size is not available though.

As ownership information is only available in cross-sectional format in the AMADEUS database, affiliates are assumed to not change their host location over time. Consequently, affiliate fixed effects nest host country fixed effects.

As the composite quality index (CQI) may take on negative values, the semi-elasticity is evaluated at the sample average of the variable plus the absolute value of the variable’s minimum: |min(CQI)| \(+\) avg(CQI)\(=2.5289-0.1958=2.3331\), cf. Table 2. It follows that \(0.035/2.3331=1.5\, \%\).

Precisely, a firm is defined to be part of a multinational group if it owns a foreign affiliate with more than 50 % of the ownership shares, or is owned by a foreign parent with more than 50 % of the ownership shares or is owned by a parent firm in the same country which in turn owns at least one foreign affiliate with more than 50 % of the ownership shares.

Note that the sample size is reduced, when we restrict the sample to MNEs and augment the model by affiliate fixed effects in Specification (4) and the total asset control in Specification (5). The sample reduction in Specification (4) on the one hand reflects that not all patent applicants can be matched to firms in the Amadeus data base, and on the other hand that some of the matched patents are filed by national firms. The sample reduction in Specification (5) reflects that the total asset information is missing for a relevant number of firm-year-cells. Note that the restriction to multinational firms in Specification (4) increases the absolute coefficient estimate, while the inclusion of the affiliate fixed effect reduces it. Running the model on the sample in Specification (4) but without affiliate fixed effects yields a coefficient estimate of \(-0.91\) (significant at the 1 % level). The absolute increase in the coefficient estimate in Specification (5) relative to Specification (4) is, moreover, driven by the sample restriction. The coefficient estimate for the sample in Specification (5) without the inclusion of the total asset control is \(-0.96\) (significant at the 1 % level).

Recall that higher tax credits/allowances reduce the B-index.

As a robustness check, we also ran specifications which included the B-index as the only tax measure, which does not change our results.

Man-months of course do not necessarily directly correspond to the actual length of the project until patenting. On the one hand, the project period in months may be shorter since more than one employee may be assigned to the project. On the other hand, the period may be longer as employees may work on several projects simultaneously. In any case, the survey suggests that the time span between the kick-off of the R&D project and the application for patent protection is rather short.

In doing so, we follow Lohse et al. (2012) and code the dummy as 1 if the country’s national tax law requires intra-firm transfer prices to be set according to the arm’s length principle and if the tax law provides further details on the applicability of the principle (e.g., specifies methods which MNEs are allowed to pursue to determine arm’s length prices).

Boehm et al. (2012) analyse patent applications to the EPO and find that geographical patent splits are partly motivated by tax considerations.