The relevance of depreciation allowances as a fiscal policy instrument: A hybrid approach to CCCTB?


A major goal of the EU Commission in the area of direct taxation is the introduction of a common consolidated corporate tax base in Europe. While hardly discussed in the literature, such a system would limit national discretion over tax depreciation. In a sample of up to 47 countries, we find that the probability of a tax reform that improves the depreciation allowances increases, if the macroeconomic situation is weak. This suggests that changes in depreciation allowances are used as a fiscal instrument for stabilization. A common consolidated tax base deprives national governments from implementing investment incentives via accelerated depreciation. This paper discusses the possible implementation of a hybrid system that combines features of formula apportionment and separate accounting. Such a hybrid system may substantially mitigate transfer pricing problems and other tax planning issues, whilst preserving national discretion over depreciation allowances.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2


  1. 1.

    There is debate about the extent to which FA reduces tax competition in tax rates. See, e.g., Pethig and Wagener (2007).

  2. 2.

    2 * 25% = 50% for machinery and 2 * 7% = 14% for industrial buildings. From 2012 to 2016, Finland reintroduced accelerated depreciation allowances.

  3. 3.

    Egger and Raff (2015) look at tax base broadening in tax competition, but do not discuss changes as potentially stabilizing macro fiscal policies.

  4. 4.

    Available at (Sept. 2017).

  5. 5.

    All information applies to 1st of January of a given year, i.e., if the tax year is different than the calendar year, any reforms introduced in a given tax year that started after 1st of January will not appear in the data until the following year.

  6. 6.

    Countries with tax rate and depreciation information at least since 1983 include: Australia, Austria, Belgium, Canada, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, and United States. Additional countries included at least since 2008: Argentina, Bulgaria, Brazil, Chile, China, Croatia, Czech Republic, Denmark, Estonia, Hungary, Indonesia, India, Iceland, Israel, Luxembourg, Mexico, New Zealand, Poland, Romania, Russia, Saudi Arabia, Serbia, Slovak Republic, Slovenia, South Africa, South Korea, Turkey, and Ukraine.

  7. 7.

    If an investor can choose between different depreciation schemes, we focus on the alternative with the fastest depreciation.

  8. 8.

    For example, the Austrian Ministry of Finance stated support of the business cycle as the reason for the 2009 change (, retrieved January 2018).

  9. 9.

    See Kremer and Ruf (2008) for the German 2008 reform.

  10. 10.

    The output gap is based on yearly IMF World Economic Outlook data and a HP filter (λ = 100). Unemployment rates are also taken from IMF data. While GAP and UE are, as expected, significantly negatively correlated, the size of the correlation coefficient (10%) is modest.

  11. 11.

    Including year fixed effects instead of the trend would drop a substantial number of observations in the logit and conditional logit regressions, because some outcomes would be completely explained by the year dummy. Note, however, that in Tables 2 and 3 a full set of year dummies is used in the LPMs of columns (7) and (8).

  12. 12.

    One may note that macroeconomic literature discusses whether in high-debt countries a consolidation of the budget can provide a positive macroeconomic stimulus. However, several papers seem to imply that such surprising expansionary effects, if at all, derive from expenditure cuts rather than from tax increases. For a recent contribution to this literature, see Alesina et al. (2015).

  13. 13.

    Apart from the difficulty to analyze the interplay between depreciation and this allowance in a static model, a static model does not capture the effects of interest rate changes on the implicit generosity of tax depreciation. This is not the focus of the present paper.

  14. 14.

    For a discussion of weight effects from tax rate changes under traditional formula apportionment, see Mintz and Weiner (2003), Pethig and Wagner (2007), or Kari et al. (2018).

  15. 15.

    A remaining remark refers to possible transfer pricing incentives in the hybrid system. As in a FA system, those are in general eliminated in a hybrid system. An exception applies if an investment good is produced in one part of the multinational, but sold to another part and installed there. In a hybrid system, the sales revenues for this transaction would be taxed at the average tax rate, while the value of the depreciation allowances are higher, the higher the tax rate of the investing part.

  16. 16.

    Indeed, the EU Commission is not required to give any evidence that such a distortion actually applies.

  17. 17.

    See Wissenschaftlicher Beirat beim Bundesministerium der Finanzen (2017).


  1. Alesina A, Favero C, Giavazzi F (2015) The output effect of fiscal consolidation plans. J Int Econ 96(Supplement 1):19–42

    Article  Google Scholar 

  2. Angrist J, Pischke JS (2009) Mostly harmless econometrics an empiricist’s companion. Princeton University Press, Princeton

    Google Scholar 

  3. Bettendorf L, Devereux MP, van der Horst A, Loretz S, de Mooij RA (2010) Corporate tax harmonization in the EU. Econ Policy 63:537–590

    Article  Google Scholar 

  4. Clausing KA (2003) Tax-motivated transfer pricing and US intrafirm trade prices. J Public Econ 87(9):2207–2223

    Article  Google Scholar 

  5. Egger P, Raff H (2015) Tax rate and tax base competition for foreign direct investment. Int Tax Public Finance 22(5):777–810

    Article  Google Scholar 

  6. EU Commission (2011) Proposal for a council directive on a common consolidated corporate tax base, COM 121/4, Brussels

  7. EU Commission (2016) Proposal for a council directive on a common corporate tax, COM 685 final, Brussels

  8. Gammie M, Giannini S, Klemm A, Oestreicher A, Parascandolo P, Spengel C (2005) Achieving a common consolidated corporate tax base in the EU. Centre for European Policy Studies, Brussels

    Google Scholar 

  9. Hines JR (1999) Lessons from behavioral responses to international taxation. Natl Tax J 52(2):305–322

    Google Scholar 

  10. Hines JR (2010) Income misattribution under formula apportionment. Eur Econ Rev 54(1):108–120

    Article  Google Scholar 

  11. Kari S, Laitila J, Ropponen O (2018) Investment incentives and tax competition under allowance for growth and investment (AGI), working paper, VATT Institute for Economic Research

  12. Kiesewetter D, Steigenberger T, Matthias S (2014) Can formula apportionment really prevent multinational enterprises from profit shifting?—The role of asset valuation. Intragroup Debt, and Leases, arqus Discussion Paper No. 175, SSRN 2512174

  13. Kremer J, Ruf M (2008) Belastung der Kapitalgesellschaften nach der Unternehmensteuerreform 2008. Wirtschaftsdienst 88(1):64–72

    Article  Google Scholar 

  14. Mintz J, Weichenrieder AJ (2010) the indirect side of direct investment: multionational company finance and taxation. MIT Press, Cambridge

    Google Scholar 

  15. Mintz J, Weiner J (2003) Exploring formula allocation for the European union. Int Tax Public Finance 10:695–711

    Article  Google Scholar 

  16. Ohrn E (2015) Investment and employment responses to state adoption of federal accelerated depreciation policies. Department of Economics, Grinnell College, Grinnell

    Google Scholar 

  17. Pethig R, Wagener A (2007) Profit tax competition and formula apportionment. Int Tax Public Finance 14:631–655

    Article  Google Scholar 

  18. PwC (2017) Worldwide tax corporate taxes 2016/17.

  19. Sinn HW (1987) Capital income taxation and resource allocation. Eslevier, Amsterdam

    Google Scholar 

  20. Vascega M, van Thiel S (2011) The CCCTB proposal: the next step towards corporate tax harmonization in the European union? Eur Tax 51:374

    Google Scholar 

  21. Wissenschaftlicher Beirat beim Bundesfinanzministerium (2017) Steuervergünstigungen und EU-Beihilfenaufsicht: Problematik und Ansätze zur Lösung des Kompetenzkonflikts mit der Steuerautonomie, Berlin

Download references


Kunka Petkova: Financial support by the Austrian Science Fund (FWF): W1235-G16 is gratefully acknowledged. Alfons J. Weichenrieder: This research is also part of the research program of the LOEWE Center SAFE. Research assistance by Eren Gürer and help with our data inquiry (CBT Tax Base) by Michael Devereux are highly appreciated. We would also like to thank an anonymous referee for very helpful remarks and suggestions, as well as Marko Koethenbuerger, Steeve Mongrain, Leslie Robinson, Martin Zagler, Eva Eberhartinger, the participants of the workshop on Macroeconomic Policy in the Eurozone, the participants of the 74th Annual Congress of the International Institute of Public Finance, the participants of the 2018 ZEW Public Finance Conference, and the participants of the DIBT research seminar for their helpful comments.

Author information



Corresponding author

Correspondence to Alfons J. Weichenrieder.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


Appendix A. Descriptive statistics

See Tables 8, 9, 10, 11, 12, 13.

Table 8 Summary statistics for Table 2 (Machinery)
Table 9 Summary statistics for Table 3 (Buildings)
Table 10 Summary statistics for Table 4 (Machinery)
Table 11 Summary statistics for Table 5 (Buildings)
Table 12 Summary statistics for Table 6
Table 13 Summary statistics for Table 7

Appendix B. Robustness checks

This appendix reproduces Tables 2, 3, 4, and 5, but restricts the sample to the 24 available EU countries in our dataset.

See Tables 14, 15, 16, 17.

Table 14 Macroeconomic situation and the probability of depreciation changes (Machinery)
Table 15 Macroeconomic situation and the probability of depreciation changes (Buildings)
Table 16 Improvements versus deterioration of depreciation (Machinery)
Table 17 Improvements versus deterioration of depreciation (Buildings)

Appendix C. A basic framework for thinking about a hybrid system

The total cost of an affiliate a located in country A may be decomposed into \( \gamma = 1 \ldots G \) different components; each of these components may be subject to different definitions. One possible definition could be the one used by country A. Alternatively, there could be a uniform EU definition for the various cost components. Hence, the effective size of a cost component \( \gamma \) for subsidiary a may either be \( C_{a,A}^{\gamma } \) or \( C_{a,EU}^{\gamma } \). A vector of G indicator variables \( v^{\gamma } \) may be used to denote whether the cost category is left in the national domain (\( v^{\gamma } = 0 \)) or in the harmonized FA domain (\( v^{\gamma } = 1 \)). A similar decomposition can be done for revenue types \( \rho = 1 \ldots Q \). Revenues may either follow a national definition \( ( {R_{a,A}^{\rho } } ) \) or a common European Union definition \( ( {R_{a,EU}^{\rho } } ) \). Again, a vector of Q indicators variables \( i^{\rho } \) may denote whether the revenue category is left in the national domain or in the harmonized FA domain. Hence, in a simple example of a multinational operating in two EU countries, A and B, the tax base \( P_{A} \) allocated to country A is then denoted by

$$ \begin{aligned} & P_{A} = \mathop \sum \limits_{\rho = 1}^{Q} \left( {1 - i^{\rho } } \right)R_{a,A}^{\rho } - \mathop \sum \limits_{\gamma = 1}^{G} \left( {1 - v^{\gamma } } \right)C_{a,A}^{\gamma } \\ & \quad + {\kern 1pt} \varphi_{A} \left( {\mathop \sum \limits_{\rho = 1}^{Q} i^{\rho } R_{a,EU}^{\rho } + \mathop \sum \limits_{\rho = 1}^{Q} i^{\rho } R_{b,EU}^{\rho } - \mathop \sum \limits_{\gamma = 1}^{G} v^{\gamma } C_{a,EU}^{\gamma } - \mathop \sum \limits_{\gamma = 1}^{G} v^{\gamma } C_{b,EU}^{\gamma } } \right), \\ \end{aligned} $$

where \( \varphi_{A} \) is the apportionment factor that will depend on the fraction of real economic activity occurring in affiliate a relative to the complete company group. Besides the exact formula for this factor, the political process needs to decide about the various indicator functions. A situation in which all \( i \)’s and \( v \)’s are zero reflects SA, while FA is characterized by all \( i \)’s and \( v \)’s equaling one. Conversely, having heterogeneous values for the indicator variables implies a hybrid system that allows for national discretion over defining particular cost or revenue categories.

Besides the decision on the r’s and c’s, there is the decision about the factors that determine \( \varphi_{A} \) and \( \varphi_{B} \). In principle, there could be differing \( \varphi \)’s depending on which cost or revenue category is split up. Allocating interest expenses may make it more natural to use the distribution of capital across countries, whereas the allocation of management remunerations may suggest a stronger role for overall payroll as an apportionment factor. However, for simplicity, Eq. (C1) introduces a common set of apportionment factors for all revenue and cost types.

Note that a common set of apportionment factors is also reflected in the FA proposal by the EU. Yet, starting from the fact that company profits are derived from the difference between various revenue and cost items may facilitate thinking out of the box. A generalized characterization of apportionment of global profits may be written as

$$ \begin{aligned} & P_{A} = \mathop \sum \limits_{\rho = 1}^{Q} \left( {1 - i^{\rho } } \right)R_{a,A}^{\rho } - \mathop \sum \limits_{\gamma = 1}^{G} \left( {1 - v^{\gamma } } \right)C_{a,A}^{\gamma } \\ & \quad + \mathop \sum \limits_{\rho = 1}^{Q} \varphi_{A}^{\rho } i^{\rho } R_{a,EU}^{\rho } + \mathop \sum \limits_{\rho = 1}^{Q} \varphi_{A}^{\rho } i^{\rho } R_{b,EU}^{\rho } - \mathop \sum \limits_{\gamma = 1}^{G} \varphi_{A}^{\gamma } v^{\gamma } C_{a,EU}^{\gamma } - \mathop \sum \limits_{\gamma = 1}^{G} \varphi_{A}^{\gamma } v^{\gamma } C_{b,EU}^{\gamma } , \\ \end{aligned} $$

where the indexation of \( \varphi_{A}^{\rho } \) and \( \varphi_{A}^{\gamma } \) signals that the formula to apportion different cost and revenue categories may differ.

Why could it be an advantage to let the apportionment factors vary for different cost and revenue categories? From Hines (2010), we know that incentives for inefficient mergers and investments are more likely to occur if the factors do not appropriately explain profits across different affiliates that are subject to formula apportionment. For instance, imagine a profitable Swedish company whose income is taxed in Sweden at a very high rate. Providing that the European companies are required to allocate their profits among affiliates relying to a large extent on the location of employment, the profitable Swedish company will then have an incentive to acquire another company in a low-tax country with a large labor force (in spite of being unprofitable). In this case, the Swedish profit might be attributed to the low-tax country where it will be subject to less taxes. The above-mentioned example illustrates how the formula apportionment could create incentives for changing the ownership structure of companies and their operations, in order to decrease their tax burden. Therefore, allowing the weights to differ across various cost and revenue categories can lead to a better fit of the factors in explaining profits.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Petkova, K., Weichenrieder, A.J. The relevance of depreciation allowances as a fiscal policy instrument: A hybrid approach to CCCTB?. Empirica 47, 579–610 (2020).

Download citation


  • Corporate taxation
  • Investment incentives
  • Macro fiscal policy

JEL Classification

  • H2