Tax Incentives and Corporate Taxation

Testimony on 11 July 2022 before the Subcommittee on Tax Matters of the European Parliament. The recording and the statements of all invited experts are available here.


Tax expenditures are benefits granted to specific sectors, activities or groups of taxpayers through preferential tax treatment; including exemptions, deductions, and lower tax rates.

Governments use them, for example, to promote employment creation, to pursue social welfare objectives, and to incentivise specific patterns of behaviour such as investment decisions or energy consumption.

Let me first give you an overview of why, I think, tax expenditures (in general) should be better scrutinized, and then move into a more detailed discussion of corporate income-related tax incentives (the subset of tax expenditures that today’s session focuses on).

According to the Global Tax Expenditures Database (GTED), over the last 30 years, the global average revenue forgone from tax expenditures was close to 4 per cent of GDP, and slightly below 25 per cent of tax revenue. Based on the latest available report, the average revenue forgone among the 24 EU member states that do report tax expenditure data is 4.7 percent of GDP; ranging from 0.6 percent in Bulgaria and slightly less than 1 percent in Germany and Estonia, to more than 10 percent in Ireland, more than 12 in Finland, and more than 14 in the Netherlands.

Yet, these estimates are conservative and, real numbers are likely to be significantly higher since the state of the art of tax expenditure reporting is worrisome. As shown by the GTED, 116 out of the 218 countries worldwide are non-reporting jurisdictions, i.e. they have never released an official tax expenditure report between 1990 and today. Equally alarming is the quality of tax expenditure reports, when they exist: roughly 26 per cent of the close to 23,000 tax expenditure provision-level data entries in the GTED, does not contain information on revenue forgone. In the EU, three member states (Cyprus, Croatia, and Malta) have never reported the revenue forgone through their tax expenditure systems and, at least 5 others (Czechia, Hungary, Portugal, Romania and Slovenia) only provide limited information, which breaches the COUNCIL DIRECTIVE 2011/85/EU on requirements for budgetary frameworks of the Member States, which requires that “Member States shall publish detailed information on the impact of tax expenditures on revenues.”

The high fiscal cost and opacity in the field are even more worrisome since several tax expenditures can be highly ineffective in reaching their stated goals, and also trigger negative externalities. The impact of environmentally harmful tax subsidies is a case in point of the latter. Even when there is a widespread consensus regarding the urgent need to phase out fossil fuel subsides to tackle climate change, governments worldwide spend millions or even billions of euros on subsidies that incentivize the production and consumption of fossil fuels. According to the latest Inventory of Support Measures for Fossil Fuels published by the OECD, 63% of the more than 1300 total measures documented in the report are tax expenditures, which represents 62% of total support by value.

Let me now move into the case of corporate income-related tax incentives. Governments worldwide implement these tax incentives to attract investment, create employment and boost research and development (R&D) and innovation…all definitely worth supporting policy goals. Yet, sound econometric evaluations providing robust evidence of a causal relationship between a specific tax incentive and the relevant outcome variable are rare and, when they do exist, the evidence is usually mixed.

As often the case, the devil is in the details…and, the design of these tax incentives is key to determine whether they are value-for-money or rather ineffective policy tools. Let me give you two concrete examples for which design features are indeed crucial:

First, the impact of the OECD global agreement of October 2021 on the use of tax Incentives. The introduction of the 15% minimum corporate income tax rate will push governments to repeal tax incentives to bring effective tax rates up to the minimum, if they are not yet there. Hence, as expected, the agreement shall contribute to curb the use of wasteful tax incentives since, according to the World Bank, redundancy ratios can be higher than 80% in some countries, i.e. more than 8 out of 10 investment projects in those jurisdictions would have taken place anyhow, even with no tax incentive in place.

Moreover, reducing the use of tax incentives shall mitigate harmful tax competition, which is indeed one of the goals of the agreement. Yet, whereas rich economies can compete to attract highly mobile tax bases by investing in non-tax-related determinants of firms’ location decisions such as infrastructure and human capital; developing economies often have rather high statutory corporate income tax rates, and rely heavily on tax incentives. Hence, the effect of the new deal might be heterogeneous across rich and poor countries.

In addition, the new tax deal will only affect tax incentives related to the profits of multinational enterprises such as corporate income taxes. Thus, some governments could be tempted to compensate for the reduction in the use of profits-related tax incentives by increasing the use of other tax incentives for investment, including those related to VAT and customs duties. This could add more harm than good; since the latter tax bases are often more reliable, stable and easier to collect than taxes on income. In addition, since these tax incentives would affect the effective tax rates paid by all businesses, and not only those in scope of the new tax agreement, the revenue implications could be significant.

The second point regards the use of tax incentives to boost R&D and innovation. The OECD’s database measuring the tax support for R&D and innovation shows that, on average, tax relief for R&D in the EU jumped from 0.02 percent of GDP in 2000 to 0.10 percent in 2018, reaching almost 0.20 percent in Austria and Italy, and almost 0.30 percent in France.

Again, the effectiveness of R&D tax incentives is mixed and highly dependent on design features. For instance, input-based tax incentives for R&D such as those targeting expenditure and wages have been proven to be more effective than output-based incentives such as intellectual property or patent boxes. The empirical evidence against the use of patent boxes is robust: first, because a patent is an intangible good that is already protected; second, because patent boxes may attract patents, but their impact on actual R&D is negligible; and third, because patents are highly concentrated across a small number of businesses and sectors and, hence, patent boxes discriminate against both un-patentable and unsuccessful research efforts. Yet, in the EU, patent boxes have been increasing both in terms of the number of existing regimes as well as in terms of the generosity of the tax benefit: according to the Tax Foundation, in 2021, 14 of the 27 EU member states had a patent box regime in place, and the tax benefit (i.e. the difference between the statutory corporate income tax rate and the tax rate under the patent box regime) was higher than 33 percent in Malta, roughly 20 percent in Belgium and Luxembourg, and 18 percent in France.

I will stop here and will be happy to answer any questions that you may have, but I hope my main message is clear: tax expenditures are not bad per se, and governments are free to choose their policy goals and how to achieve them. What is unacceptable is that governments keep spending impressive amounts of money through their tax systems with little or no scrutiny. Hence, if tax incentives are used in the context of the corporate income tax reform to enhance competitiveness, their fiscal cost should be estimated and disclosed, and these measures should be evaluated (ideally, ex-ante and ex-post) to ensure that they are fit for purpose, and do not create side effects or distortions, both within the EU market as well as on other jurisdictions, particularly developing economies.

Thank you very much for your attention.