Abstract
Most industrialized countries apply special tax incentives to boost the R&D expenditures of firms. This study considers the design of such R&D tax incentives as applied in the European Union and simulates its effect on the post-tax R&D expenditures of firms in different industries and different profit/loss-situations by means of the simulation model European Tax Analyzer. Any restrictions and progressive tax incentives are explicitly taken into account. Our results indicate that for designing and measuring public support to R&D it is often not sufficient to focus only on tax rate effects of R&D tax incentives and the design of a tax incentive must be in accordance with the framing tax system in order to be effective. As soon as there are any limitations in place, our results suggest a considerably lower impact of R&D tax incentives on the post-tax R&D expenditures than the commonly used B-Index by the OECD. The results clearly illustrate the beneficial impact of immediate cash refunds for unused tax incentives.
This is a preview of subscription content, log in via an institution to check access.
Notes
One important argument pro R&D tax incentives is the belief that it is neutral considering the structure and the choice of R&D projects (e.g. Atkinson 2007). The incentive effect should then be equal for all participants and projects if the policy is not to change the composition of R&D but to raise the level of R&D and to offer the subsidy to small and medium-sized enterprises (SME) as well. Moreover, the financial constraints SMEs are facing due to asymmetric information (Czarnitzki and Hottenrott 2009) are expected to be at least partly mitigated by tax incentives. Ortega-Argilés et al. (2009) give a condensed overview on the role of SME in R&D and the argumentation for tax incentives. However, there are arguments for direct subsidies as well, which are seen to be more likely to influence the composition of R&D, see Tassey (1996, 2007a). Both forms of support for R&D are seen to be complementary. Stoneman (1991) considers the financing of an R&D tax incentive and favors a levy/grant system instead which burdens (less-innovating) firms instead of a country’s budget.
In 2005, 19 out of 27 examined OECD Member States had R&D tax incentives in place (Warda 2006).
Hall and Van Reenen (2000) and Arundel et al. (2008) give an overview. There are several results in the literature that reveal an effect of R&D tax incentives on the relocation of R&D. Paff (2005) found effects that are explained by the relocation of R&D activity to use tax based incentives in California, Wilson (2009) found a strong effect of relocation of R&D among US-states, Billings (2003) found a higher average growth rate of R&D for US-foreign affiliates in countries offering tax incentives compared to countries without incentives and Bloom and Griffith (2001) suggest that UK firms react to a more beneficial treatment of R&D in other countries.
Mohnen and Lokshin (2009) give an overview on empirical studies, including a discussion of indicators used and results.
McKenzie (2008) provides a measure which embeds R&D tax incentives into a marginal effective tax rate and thus not only focuses on R&D expenditures but also on the taxation of resulting profit. This approach—albeit accounting for more tax parameters than the B-Index—does not show the impact of loss situations either.
Tax exhaustion is considered to be one of the reasons that hamper effectiveness of tax based schemes (Stoneman 1991). Yet, a cash refund in case of losses or a reduction in wage taxes can raise liquidity immediately.
Warda (2001) accounts for that by making a distinction between a B-Index for large firms and a B-Index for small and medium-sized firms.
For the sake of comparability of the tax burden across countries and to allow identifying tax drivers, it is necessary to have an identical economic “starting point”. One might receive different results in case one considers country-specific firms. In this study, the focus lies on the incentive’s impact on the firm’s tax burden, seen from a more (tax) technical perspective. We thus abstract from any economic differences between countries.
From an economic point of view, the stock of capital generated by R&D activity like the value of know how is indirectly reflected by the industry specific economic parameters profitability (e.g. flow of income), cost structure and the structure and value of the assets. A higher stock of R&D capital increases the average profitability and vice versa.
As for the financial data, the data from Stifterverband is based on German firms. However, the tax incentives in all countries of the European Union are analyzed in this study, consequently, it would be more straightforward to use European-level data. Since data on both financial data and R&D structures are not available as European averages, we stick to the German data.
Incentives which are available irrespective of R&D like general investment credits are not included. Concerning depreciation of R&D, we consider the respective tax accounting rules of each country for the capitalization and depreciation of self developed intangibles for sale resulting from R&D expenses and for the treatment as expenses if not capitalized.
Countries are ranked according to the effective tax burden after accounting for incentives (column 5).
R&D expenditures in this simulation are not fixed (see Sect. 2), they increase moderately at rate 1.9% on average, in line with the development of turnover. So the impact of the tax incentive is not very pronounced. A more detailed analysis on the impact of incremental tax incentives is given in the sensitivity analysis.
Additionally, Poland grants incentives for R&D centers which, however, are not considered here.
The latter point is not considered in the simulations, where only expenditures based on the Frascati definition of R&D are included.
Besides the different tax rates in Hungary and the Czech Republic, the large amount of business tax which reduces the corporate taxable base and hinders the deduction of the R&D tax incentive is the second reason why the considered firm faces a substantially higher tax reduction in the Czech Republic than in Hungary, albeit both countries apply the same incentive.
Other countries that do not have an incentive on the increase of expenditures over time show, as expected, slightly lower impacts because the R&D expenditures in period one are lost for their incentives.
The weights given in Warda (2001) are 60% for personnel cost, 30% for other current expenditures, 5% for investments in machinery and equipment and 5% in buildings. Incremental tax incentives are included by introducing an inflation rate since the increment usually is based on nominal amounts.
Similar conclusions can be drawn if one considers the differentials of the two indicators as a percentage of the results based on the European Tax Analyzer. Then, Austria shows a deviation of 10%, the United Kingdom one of 18%. Moreover Portugal, Hungary, and the Czech Republic have the next lowest differences with 35–49%. Also, these three countries do not show high restrictions in the use of R&D tax incentives.
References
Arundel, A., Bordoy, C., Mohnen, P., & Smith, K. (2008). Innovation surveys and policy: Lessons from the CIS. In C. Nauwelaers & R. Wintjes (Eds.), Innovation policy in Europe: Measurement and strategy. London: Elward Elgar.
Atkinson, R. D. (2007). Expanding the R&D tax credit to drive innovation, competitiveness and prosperity. Journal of Technology Transfer, 32(6), 617–628.
Baghana, R., & Mohnen, P. (2009). Effectiveness of R&D tax incentives in small and large enterprises: Analysis of firm data in Québec. Small Business Economics, 33(1), 91–107.
Billings, A. B. (2003). Are US tax incentives for corporate R&D likely to motivate American firms to perform research abroad? Tax Executive, 55(4), 291–315.
Bloom, N., & Griffith, R. (2001). The Internationalisation of UK R&D. Fiscal Studies, 22(3), 337–355.
Bloom, N., Griffith, R., & Van Reenen, J. (2002). Do R&D credits work? Evidence from a panel of countries 1979–97. Journal of Public Economics, 85(1), 1–31.
Czarnitzki, D., Hottenrott, H. (2009). R&D investment and financing constraints of small and medium-sized firms. Small Business Economics. doi:10.1007/s11187-009-9189-3.
Deutsche Bundesbank. (2003). Verhältniszahlen aus Jahresabschlüssen deutscher Unternehmen von 1998 bis 2000. Statistische Sonderveröffentlichung 6. Frankfurt/Main: Deutsche Bundesbank.
Devereux, M. P., & Griffith, R. (1999). The taxation of discrete investment choices. Institute for Fiscal Studies Working Paper 98/16 (Revision 2). London: Institute for Fiscal Studies.
Devereux, M. P., & Griffith, R. (2003). Evaluating tax policy for location decisions. International Tax and Public Finance, 10(2), 107–126.
European Commission. (2001). Company taxation in the internal market. COM(2001) 582 final. Luxembourg: Office for Official Publications of the European Communities.
Hall, R. E., & Jorgenson, D. (1967). Tax policy and investment behavior. American Economic Review, 57(3), 391–414.
Hall, B., & Van Reenen, J. (2000). How effective are fiscal incentives for R&D? A review of the evidence. Research Policy, 29(4), 449–469.
IBFD. (2006/2007). Corporate taxation in Europe. Online-Database. Amsterdam.
Jabobs, O. H., Spengel, C., Stetter, T., & Wendt, C. (2005). EU company taxation in case of a common tax base. Intertax, 33(10), 414–428.
Jacobs, O. H., & Spengel, C. (1996). European tax analyzer. Baden-Baden: Nomos.
King, M. A., & Fullerton, D. (1984). The taxation of income from capital. Chicago: University of Chicago Press.
McFetridge, D. G., & Warda, J. (1983). Canadian R&D incentives: Their adequacy and impact. Canadian Tax Foundation, Doc. No. 70. Toronto: Canadian Tax Foundation.
McKenzie, K. J. (2008). Measuring tax incentives for R&D. International Tax and Public Finance, 15(5), 563–581.
Mohnen, P., & Lokshin, B. (2009). What does it take for an R&D tax incentive policy to be effective? UNU-MERIT Working Paper Series 2009–014. Maastricht: Maastricht University.
OECD. (2002). Frascati manual 2002. Paris: OECD.
OECD. (2007). OECD Science, Technology and Industry Scoreboard 2007. Paris: OECD.
Ortega-Argilés, R., Vivarelli, M., & Voigt, P. (2009). R&D in SMEs: A paradox? Small Business Economics, 33(1), 3–11.
Paff, L. (2005). State-level R&D tax credits: A firm-level analysis. Topics in Economic Analysis and Policy 5(1), 1272, Article 17. Berkeley: The Berkeley Electronic Press.
Spengel, C., & Oestreicher, A. (2009). Common corporate tax base (CCTB) in the EU: Impact on the size of tax bases and effective tax burdens. ZEW Economic Studies. Heidelberg: Springer, forthcoming.
Stetter, T., & Spengel, C. (2006). Taxation of corporations in Canada. A Comparison of Tax Burdens with the United States and Selected Member States of the European Union Using the European Tax Analyzer. European Taxation July/2006, 307–316; 364–374.
Stifterverband. (2007). FuE-Datenreport 2007–Tabellen und Daten. Essen.
Stoneman, P. (1991). The use of a levy/grant system as an alternative to tax based incentives to R&D. Research Policy, 20(3), 195–201.
Tassey, G. (1996). Choosing government R&D policies: Tax incentives vs. direct funding. Review of Industrial Organization, 11(5), 579–600.
Tassey, G. (2007a). The technology imperative. London: Edward Elgar.
Tassey, G. (2007b). Tax incentives for innovation: Time to restructure the R&E tax credit. Journal of Technology Transfer, 32(6), 605–615.
Warda, J. (1996). Measuring the value of R&D tax provisions. In OECD, Fiscal measures to promote R&D and innovation. OECD (96)165. Paris: OECD, 9–22.
Warda, J. (2001). Measuring the value of R&D tax treatment in OECD Countries. Special Issue on New Science and Technology Indicators. OECD STI Review No. 27.
Warda, J. (2006). Tax treatment of business investments in intellectual assets: An international comparison. STI Working Paper DSTI/DOC (2006)4. Paris.
Wilson, D. J. (2009). Beggar thy neighbor? The in-state, out-of-state and aggregate effects of R&D tax credits. Review of Economics and Statistics, 91(2), 431–436.
Acknowledgments
We are grateful to participants of the 2007 CONCORD conference (Sevilla) and to participants of the 2008 congress of the IIPF (Maastricht). All remaining errors are ours. We also acknowledge the help of Shampa Ghosh and Judith Specht.
Author information
Authors and Affiliations
Corresponding author
Annex
Annex
1.1 Description of the European tax analyzer model
The European Tax Analyzer is a simulation model that calculates and compares effective average tax burdens for companies located in different jurisdictions. It currently covers the tax systems of the 27 EU Member States. The effective average tax burden is derived by simulating the development of a corporation’s financial budget over a 10 year period. After the tenth period, the firm is sold at its fair value. The firm’s budget depends on the firm’s annual turnover and expenditures which again depend on the underlying business plan, the economic environment (such as differing interest rates) and the country-specific tax environment. The respective data on the financial structure of the firm and economic data is drawn from Deutsche Bundesbank (2003), the tax data is mainly based on IBFD Tax Databases (2006/2007).
The initial capital stock includes the firm’s total assets and liabilities. The assets consist of real estate, office and factory buildings, plant and machinery, office equipment, intangibles (patents and royalties), financial assets, shares in other corporations (both domestic and foreign), inventories, trade debtors, cash funds, and deposits. The liabilities include new equity capital, long-term and short-term debt, and trade creditors.
The business plan which drives the development of the capital stock includes estimates on production and sales, acquisition of goods, personnel cost, R&D expenditures, investment, distribution, and costs of financing. The model accounts for different sorts of goods which are either stocked or sold and different approaches to determine production costs. Moreover, different depreciable assets with differing useful lives are implemented; reinvestments are undertaken at the end of an asset’s useful life, on the basis of the asset’s replacement cost. Economic data includes rates of price increases and credit and debit interest rates.
For the sake of comparability, the underlying firm and economic data is constant irrespective of where it is located. Due to this necessary assumption any differences between pre- and post-tax data in the model can be solely attributed to the applied country-specific taxation rules.
The effective tax burden is the difference between the pre-tax and the post-tax value of the firm (both in €) at the end of the simulation period (i.e. period 10). The value of the firm represents the equity, which includes the initial capital stock plus the cumulative net income of each of the ten periods. At the end of period ten, the tax value of assets and liabilities may differ from their fair value, depending on the tax rules which are to be applied. These hidden reserves and liabilities are added to the taxable income in period ten and are taxed accordingly. As a consequence, only the effects of different tax accounting rules on the liquidity are taken into account. Any remaining loss carry-forwards at the end of the simulation are dissolved liquidity-related whereas a devaluation of 50% is made if there are no restrictions for the use of loss carry-forwards and a devaluation of 75% if there are any restrictions. The computation of the absolute effective average tax burden requires two steps.
In the first step, the pre-tax value of the firm at the end of the simulation period is determined. The pre-tax value of the firm is derived from the estimated cash flows and the value of the net assets at the end of the simulation period. Thereby it is assumed that any amount of surplus cash flow at the end of a single period can be invested at a given interest rate and any deficit can be covered by borrowing money at a given debit rate (balancing investment or credit). The interest receipts or expenses plus the amount of the underlying balancing investments or credits are considered for the calculation of the cash flow in the following periods. The value of the net assets at the end of the simulation period is computed by deducting the liabilities of the corporation from the assets.
Pre-tax cash flow at the end of the simulation period |
+ Value of the net assets at the end of the simulation period |
(= Assets in the capital stock at replacement prices |
− Liabilities in the capital stock at nominal values) |
= Pre-tax value of the firm at the end of the simulation period |
In the second step, the post-tax value of the firm at the end of the simulation period is determined. This equals the pre-tax value of the firm minus tax liabilities for each period minus the taxation of hidden reserves and liabilities at the end of the simulation period. Thereby, the model makes use of the country-specific tax-accounting rules and all relevant features of corporate income tax as well as all other taxes payable at the corporate level (such as business taxes, real estate taxes or payroll taxes). The detailed simulation of cash flows within the firm allows for the implication of complex tax rules such as progressive tax rates, limited allowances, tax credits, and—as it is a multi-period approach—loss carry-forwards and loss carry-backwards.
The differential between the pre-tax and the post-tax value then is the effective average tax burden, measured over a time-period of 10 years.
Pre-tax cash flow at the end of the simulation period |
− Tax liabilities in each period |
= Post-tax cash flow at the end of the simulation period |
+ Value of the net assets at the end of the simulation period |
(= Assets in the capital stock at replacement prices |
− Liabilities in the capital stock at nominal values) |
−/+ Tax liabilities on hidden reserves/tax refunds on hidden liabilities |
= Post-tax value of the firm at the end of the simulation period |
Pre-tax value of the firm at the end of the simulation period |
− Post-tax value of the firm at the end of the simulation period |
= Effective average tax burden |
Rights and permissions
About this article
Cite this article
Elschner, C., Ernst, C., Licht, G. et al. What the design of an R&D tax incentive tells about its effectiveness: a simulation of R&D tax incentives in the European Union. J Technol Transf 36, 233–256 (2011). https://doi.org/10.1007/s10961-009-9146-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10961-009-9146-y