Splitting of Corporate Taxes in Germany and Formulaic Distribution of a CCCTB – Critical Comparison
Publié en ligne: 30 nov. 2022
Pages: 269 - 290
DOI: https://doi.org/10.2478/ceej-2022-0016
Mots clés
© 2022 Sandra Müller-Thomczik et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
The allocation of cross-border company profits to the states involved according to company domiciles or business activities is a problem that has not yet been solved satisfactorily. At the international level, numerous reform efforts are underway, the majority of which are linked to the reform of the previous profit accrual system via transfer pricing.
See in particular action points 8 to 10 of the BEPS Action Plan (OECD, 2014), Action Plan to Combat Profit Reduction and Profit Shifting. In response, the European Council adopted Directives ATAD I and II (Directive 2016/1164 laying down rules to avoid tax avoidance practices directly affecting the functioning of the single market dated 12 July 2016, OJEU of 19 dated 2016, L 193/1 and Directive 2017/952 amending Directive 2016/1164 as regards hybrid arrangements with third countries dated 29 May 2017, OJEU of 7 June 2017, L 144/1 respectively).
A fundamentally different orientation is given to the formulaic distribution of profits. In 2016, the EU Commission presented a proposal for a directive to implement a common corporate tax base with a subsequent formulaic breakdown of a common consolidated corporate tax base (CCCTB).
In 2015, the EU Commission announced a two-step approach to the introduction of a CCCTB: first a common corporate tax base, then consolidation and sharing (EU-Com, 2015, p. 9).
The commission recently withdrew the specific proposal. However, it still explicitly commits to a formulaic profit distribution in the EU, which is called BEFIT, Framework for Income Taxation (COM (2021) final, p. 11 et seq.). On the one hand, the reform initiatives of the OECD regarding Pillar I and Pillar II are to be continued, on the other hand assets and labour should be taken into account appropriately within the distribution of corporate profits (EU-Com, 2021, p. 12). The implementation of a CCCTB in the EU would render quite a number of benefits, most notably the reduction of tax compliance costs (currently caused by different corporate tax systems and extensive transfer pricing obligations, for example), the avoidance of double taxation, and the reduction of undesirable tax arrangements (Spengel & Stutzenberger, 2018, p. 39 et seq.).
However, formulaic profit sharing has never been introduced at the EU level. It could be possible that the OECD initiatives have changed the situation in such a way that introduction is more likely, but this is not certain.
While the seemingly perpetual global and EU-wide discussion about how to allocate corporate profits goes on, German tax law is familiar with a long-standing, yet scarcely criticised, apportionment of taxable substrate. By implementing the concept of trade tax, the assessment base has been determined since 1936 on the basis of a uniform set of rules and a subsequent apportionment of the assessment base to the participating municipalities by using formula factors. The corporate income tax revenue, to which the federal government and the federal states are jointly entitled, is also divided between the federal state of domicile and the federal state in which a permanent establishment is located pursuant to the provisions of the Trade Tax Act.
Since a formulaic profit distribution would be desirable because of its significant advantages, this study will examine whether the German method of splitting could lead to some considerations of an appropriate and workable formula design for an international profit distribution formula.
Methodologically, after a brief literature overview (Section 2), we use a two-step approach: First we compare the designs (Section 3). For this, we look at the mechanism that is generally accepted (in Germany) and the one that is politically desired, but so far has not been implemented (CCCTB). We will juxtapose both on a factual level and determine that the German mechanism is much simpler and more pragmatic, not only at first impression but also when examining the overlapping factors. Nevertheless, there is hardly any perceptible criticism of the practice of splitting. Based on this recognition, it is questionable what the objectives and requirements of these mechanisms are, and whether they even coincide. If the goals are not comparable, one could not indisputably be used as a model for the other.
In the next step, we analyse the objectives and requirements (Section 4). The goals of the two mechanisms have to be worked out first. To begin with, we analyse different documents, working papers and literature. After we have defined the goals, we will compare them. Here, our approach will be to identify each objective, aimed-for CCCTB apportionment before juxtaposing it with any relevant aspects of the German tax-splitting system.
We determine that, even though there are partial deviations, a closer look reveals significant overlaps; however, the German implementation is far from consistent and prioritises practicability. This leads us to the lessons (Section 5): The German system makes a clear value decision towards practicability, although there is a different set of aims.
For the implementation of formulaic EU profit sharing, the lesson to be learned is that practicability should play a central role in the design of the formula. This lesson is important and helpful to accompany and support the implementation process in the EU.
This study is part of the research on possible formula design for profit allocation in the EU. Much scientific work deals with the formular apportionment of a CCCTB (Jakimovski 2012, p. 29). Furthermore, the formula itself is widely discussed (Kiesewetter, Steigenberger & Stier, 2018, p. 1032). The following articles deal with the formula factors and possible formula designs in a conceptual manner:
Kahle (2007) analyses the apportionment based on factors of production, such as labour, capital and sales, and macroeconomic variables critically. Koch (2010, p. 93 et seq.) and Wellisch (2004) also focus on these factors. Moreover, they deal with a value-added based factor. Hellerstein and McLure (2004), Sørensen (2004), and Agúndez-García (2006) also prove the conception of these factors in detail. Lodin and Gammie (2001, p. 47 et seq) focus especially on the valued-added ratio.
Nerudova (2008) sees a formula based on labour, capital, and sales as the best solution. In contrast Dahlke (2011, p. 344) sees this formula as only the second best. He provides a detailed discussion of possible factors of formulaic allocation for the EU starting with an overview of the design of formulaic profit allocation applied in the USA, Canada, Switzerland, and Germany (Dahlke, 2011, p. 308 et seq.). Eichner and Runkel (2008, p. 567 et seq.) deal with the question of whether formular apportionment should replace separate entity accounting, and if so, which formular factors should be used. They state that the factor sales may mitigate or even eliminate fiscal externalities caused by the countries’ tax policy. The consideration of demand in a formulaic profit distribution goes back to Musgrave (1972). Roggeman et al. (2013) simulate the splitting results of different formular designs. Lessons from the Canadian example are presented by Mintz (2004, S. 222 f.). Jakimovski (2012) takes a detailed look at the problem of a theoretical justification of profit-sharing factors.
As far as we know, no study has yet looked at possible lessons that can be derived from the German system of tax-splitting.
The splitting of trade tax is based on §§ 28 et seq. GewStG. On the other hand, there is no specific standard for the splitting of corporate income tax. Pursuant to § 2 para. 1 sentence 2 ZerlG, §§ 28–31, 33 GewStG are decisive in this respect. Pursuant to Sec. 2 (1) Sentence 1, an apportionment takes place only when the corporate income tax reaches an absolute amount of €500,000.
According to § 28, para 1, Sentence 1, GewStG, the trade tax assessment amount must be apportioned if permanent establishments for exercising a trade are maintained in different municipalities (Baldauf, 2022, § 28 GewStG, Rn. 13). This is true in the case:
of an establishment extending over several municipalities, of having several establishments in different municipalities, and a relocation of an establishment to another municipality during the year (§ 28 para 1 GewStG; Reichert (2019), p. 699.
The splitting of the trade tax assessment amount is based on the ratio of the sum of the wages paid to the employees working in all the permanent establishments to the wages paid to the employees working in the permanent establishments in the individual municipality (§ 29 para 1 no. 1 GewStG). The concept of an employee is based on the definition of ‘employment’. Pursuant to § 7, para 1, SGB IV, this is defined as non-self-employed work which is carried out in an employment relationship. Instructions and integration into the work organisation of the person/entity giving the instructions are regarded as indications of employment.
Under § 31 (1) of the GewStG, wages are considered as remuneration within the meaning of § 19 (1)(1) of the EStG, provided they are not exempt from income tax under other provisions; supplements for overtime and for work on Sundays, public holidays, and at night are also wages, irrespective of their treatment for income tax purposes.
For the special cases under para. 3 and para. 5 (Baldauf, 2022, § 31 GewStG, Rn. 7–10).
As of the 2015 assessment period, reformed 2021,
Gesetz zur Stärkung des Fondsstandorts Deutschland und zur Umsetzung der Richtlinie (EU) 2019/1160 zur Änderung der Richtlinien 2009/65/EG und 2011/61/EU im Hinblick auf den grenzüberschreitenden Vertrieb von Organismen für gemeinsame Anlagen (Fondsstandortgesetz – FoStoG) vom 03.06.2021, BGBl. I, p. 1498.
This special case of splitting is based on two components:
One-tenth of the standard pursuant to § 29, para 1, no. 1, GewStG, are included the remaining nine tenths are distributed on the basis of a special standard derived from the ratio of the sum of the installed capacity within the meaning of § 3 number 31 of the Renewable Energy Sources Act
Erneuerbare-Energien-Gesetz v. 21.06.2014, BGBl. I, S. 1066, zuletzt geändert durch Art. 1 des Gesetzes vom 23.05.2022, BGBl. I, S. 747.
Operating and office equipment, payments on account, and plant under construction are deducted from tangible fixed assets under point (b) (§ 29 para 1 & 2 (a) of the GewStG), as these assets are not (yet) used for energy production and are therefore not (yet) considered eligible to receive subsidies (Baldauf, 2022, § 29 GewStG, Rn. 28).
In constellations in which an apportionment pursuant to § 29, para 1, no. 1, GewStG is not applicable or leads to inequitable results, a splitting based on the actual circumstances is possible pursuant to § 33, para 1, GewStG (Reichert, 2019, p. 702; see also BFH of 26.08.1987, I R 376/83, BStBl. II 1988, p. 201). Any splitting is manifestly inequitable if it would lead to an inequitable result in an individual case (Baldauf, 2022, § 33 GewStG, Rn. 3). The manifest inequity must be substantial (BFH of 16. 12. 2009 - I R 56/08, DStR 2010, p. 484).
According to the opinion of the supreme court, the legislator has deliberately chosen a rough criterion for the splitting standard in the form of wages that serves the purpose of practicability, so that inconsistencies are accepted in individual cases and do not yet constitute an apparent inequity (BFH of 9.10.1975 – IV R 114/73, BStBl. II 1976, p. 123; Offerhaus & Althof, 2006, p. 626 et seq.). The jurisdiction assumes a manifest inequity, for example, in the case of a substantial, permanent, and exclusive use of temporary workers in a permanent establishment (BFH, 26.2.1992 - I R 16/90, BFH/NV 1992, p. 836; also Bahn, 2016, p. 1370). According to the BFH, however, differences in profitability between different permanent establishments do not justify splitting under § 33 GewStG (BFH of 25.11.2009 - I R 18/08, BFH/NV 2010, p. 941).
In the event of a serious apparent inequity, the ‘new’ splitting standard must take into account the actual circumstances characterising the individual case and must also be better than splitting pursuant to § 29 (1) GewStG (§ 33 para 1 GewStG). This splitting, which corresponds to the actual circumstances, is carried out, if possible, in agreement among the tax office, the municipality, and the tax debtor and is binding on the tax authorities. It is a case-by-case decision (Reichert, 2019, p. 704. As a substitute, for example, the apportionment according to the proportion of turnover can be applied (Sarrazin, 2021, § 33 para 21 GewStG).
Still open in the case of wages paid under management agreements of sister companies; FG Munich of 27.11.2018, 6 K 2407/15, EFG 2019, p. 376 (pending under BFH - III R 3/19).
Finally, § 33 (2) GewStG provides for the possibility of the municipalities agreeing with the tax debtor on the apportionment if the apportionment standard in § 29 (1) is not applicable (Baldauf, 2022, § 33 GewStG, Rn. 10). Any agreement reached shall be binding.
However, the legal nature of such an agreement is not clear. According to the BFH case law, this is a sui generis agreement similar to an actual understanding (Baldauf, 2022, § 33 GewStG, Rn. 10 f. with further references).
The apportionment of the CCCTB under the proposed CCCTB guidelines provides for apportionment by turnover, assets, and labour as three equally weighted apportionment factors.
The question of choosing the right apportionment factors is considered to be the core problem of profit sharing (Wissenschaftlicher Beirat beim Bundesministerium der Finanzen, 2007, p. 50.)
Using these factors, the proposed CCCTB guidelines identify a formula that is modelled specifically on US law. The Formulary Apportionment has long been valid law in the USA and Canada within the framework of the so-called ‘Unitary Taxation’ (Dahlke, 2011, p. 309, Fn. 2138 for extensive literature on unitary taxation). The basis of the apportionment formula is the so-called Massachusetts formula, which has been recommended since 1933 by the National Tax Association (NTA) as a measure for harmonisation (Weiner, 2005, p. 11; Hellerstein & McLure, 2004, p. 208).
The Massachusetts formula thus found widespread use within the states. This development was reinforced by the Uniform Division of Income for Tax Purposes Act (UDITPA) in 1957, which proposed the Massachusetts formula as a model. UDITPA was incorporated into the Multistate Tax Compact in 1967 as Art. IV in the Multistate Tax Compact (Oestreicher et al., 2008, p. 294; for the further historical course see Hellerstein, 1993; Weiner, 1994, Bökelmann, 1997, p. 115).
Pursuant to Art. 32 (1) CCCTB-Proposal, one half of the labour factor is made up of the total wages of a group member and the other half of the number of employees of this member. This figure is then set in relation to the sum of the wage totals of the group members and the sum of the number of employees.
According to Art. 32 (3) CCCTB-Proposal, the definition of ‘employee’ is based on national legislation, which varies within the Member States. Differences within member states would have an impact on the assessment of the wage bill. Pursuant to Art. 33 (4) CCCTB-Proposal, this also includes the employer's contribution to social security and pensions subject to national provisions.
The decisive factor for the allocation of an employee to a group member is whether the employee receives remuneration from the group member (Article 33(1) CCCTB-Proposal). Under certain conditions, employees may be assigned to another group member if they carry out their activities under the supervision and responsibility of that group member.
The requirements are: The employment relationship has a duration of at least three continuous months and the employees in question account for at least 5% of all employees of the group member from which they receive their remuneration (Art. 33 Para. 2 CCCTB-Proposal).
According to Art. 34 (1) CCCTB-Proposal, the factor consists of the average value of all tangible assets owned by a group member in relation to the average value of all tangible assets. Rented and leased properties are also included.
The allocation is generally made to the economic owner. If this cannot be determined, the assets are allocated to the legal owner (Article 35 (1) CCCTB-Proposal). Allocation to the actual user is also conceivable if the asset represents more than 5% of the taxable value of all tangible assets of the group member actually using the asset.
According to Article 37 (1) CCCTB-Proposal, turnover means the proceeds from all sales of goods or services after deduction of discounts and returns. Tax-exempt income, interest, dividends, royalties, and proceeds from the sale of fixed assets are not included in the turnover factor unless they are income earned in the ordinary course of business.
According to Art. 38 CCCTB-Proposal, the allocation is always based on the place of destination. If This is a so-called ‘spread throw-back rule’ (CCCTB-WG, 2007a, p. 2).
If, on the other hand,
Due to the current developments, the importance of the sales factor, especially at the level of the OECD, has increased considerably: The main characteristic of the Pillar I in terms of the distribution of the taxation rights regarding digitised business models is built on turnover. In particular, an Amout A is provided for depending on the exceeding of certain sales thresholds (OECD, 2021a, p. 1 et seq.) and an Amount B for basic marketing and sales activities based on transfer prices (OECD, 2020b, p. 14 f.).
Article 29 CCCTB-Proposal provides for a safeguard clause according to which the principal taxpayer
For the definition of ‘principal taxpayer’, see Art. 3 para 11 CCCTB-Proposal.
Whereas the apportionment mechanism of the CCCTB uses a three-factor formula, the splitting of German corporate taxes is based on only one factor. The obvious commonality between the two apportionment mechanisms is the labour factor. In terms of content, however, the term ‘labour’ in the formulation of the CCCTB-Proposal and in the statutory definition pursuant to § 31 (1) GewStG only coincide to a limited extent. According to Art. 32 CCCTB-Proposal, the labour factor is composed of half of the wage total of a group member in the wage total of the entire group and half of the number of employees of a group member in the total number of employees of the group. The trade tax apportionment, on the other hand, is based only on the wage total.
While the definition of an employee in Germany is based on § 7 (IV) SGB IV, the definition of the CCCTB factor may vary, as it depends on the regulations of the member states. This inevitably leads to divergent assessments in the various EU countries.
The treatment of services contracted out also appears questionable in the CCCTB splitting: Pursuant to Art. 33 para. 3 CCCTB-Proposal, persons who are not directly employed by a group member but who carry out similar activities as directly employed persons are also deemed to be employees (raising this issue, see Oestreicher et al., 2008, p. 357, et seq).
A comparison with the German standard decomposition shows that the number of employees is also considered in addition to the wage total. The aim of this inclusion is to equalise the different wage and salary levels within the EU (CCCTB-WG, 2007a, p. 2). The trade tax apportionment mechanism at least provides for a cap on remuneration that does not exceed €50,000 so that individual high salaries do not distort the apportionment result.
The definition of the wage bill reveals further differences in the recording of social security expenditure. While these are included in the apportionment in accordance with Article 33 (4) of the CCCTB-Proposal, the German legislator takes into account all remuneration received by the employee and thus precisely not the social security expenses via the reference in § 31 (1) sentence 1 of the GewStG to § 19 (1) no. 1 of the EStG. The fact that social security systems within the EU are not uniform is likely to make this particularly important (Dahlke, 2011, p. 325 et seq.).
Further differences are that the CCCTB-Proposal does not explicitly exclude remuneration for vocational training and tax-exempt remuneration,
For the German regulation see § 31 para 2 and 4 GewStG. § 31 para 5 GewStG also provides for the fictitious use of a co-entrepreneur's wage, which results from the German peculiarities of the trade tax liability of partnerships and therefore does not need to be explained further here.
Both apportionment mechanisms also provide, in principle, for the possibility of agreeing on a different apportionment. Table 1 summarises the similarities and differences in the regulatory areas concerned, showing that the German splitting is much simpler.
Comparison of the similarities and differences of the apportionment according to the CCCTB-Proposal and the Trade Tax Act
| Number of factors | 3 factors | 1 factor |
| Composition of the labour factor / content of the wage bill | - Wage bill and number of employees (half) |
- Wage total only |
| - Inclusion of social security expenses | ||
| - Inclusion of services to external parties who perform similar activities to those performed directly by employees | ||
| - Inclusion of temporary workers | ||
| - Presumably inclusion of vocational training | ||
| - Presumably inclusion of tax-free remuneration | ||
| - Capping limit | ||
| Alternative apportionment standards | Possible |
Source: Authors’ own table
Apart from the obvious use of only one factor, the consideration of the number of employees is systematically different. Furthermore, the wage bill in Germany is a significantly smaller factor than in the CCCTB. Nevertheless, this splitting has been taking place in Germany for many decades without any clearly perceptible criticism. Whether lessons can be learned from this for the apportionment of a CCCTB depends largely on the objectives pursued by the apportionment mechanisms. These are elaborated below and then compared with each other.
The trade tax has a long history. Their precursors served as early as the Middle Ages, ‘the protection of the city walls by the guilds’ (Drüen, 2022, § 1 GewStG, Rn. 2). With the real tax reform of 1936, the previously different state legal codifications were unified. Since then, the rules on splitting have remained largely unchanged.
In its basic conception, the trade tax is an object tax (Glanegger & Güroff, 2021, § 1 GewStG, Rz. 14). The tax should not be assessed on the basis of the economic capacity of the trader, but on the basis of the characteristics of the tax object (Drüen, 2022, § 1 GewStG, Rn. 9; Glanegger & Güroff, 2021, § 1 GewStG, Rn. 14). Since the abolition of the trade capital tax with effect from 1998, this circumstance has been the subject of increasing criticism, since from then on only the trade income was taxed and thus the character of an income tax dominated (on the state of opinion Drüen, 2022, § 1 GewStG, Rn. 9).
The justification of the trade tax is based on the principle of equivalence. The trade tax is intended to contribute to compensating a municipality for the burdens directly and indirectly incurred by the trade (RStBl. 1937, p. 696). It has always been problematic that the causal relation expressed in it is not given (Tipke, 2003, p. 1142 et seq).
Tipke further concedes that there is a cost burden from commercial enterprises, but a blanket restriction to commercial enterprises can hardly be justified (Tipke, 2003, pp. 1140, 1142.) There is also an ‘information problem’ (in this respect Hey, 2002, p. 319; Kraft & Kraft, 2018, p. 197, see ‘the trade tax as consideration for municipal services in the form of infrastructure’. For embedding in the legal history and critical appraisal, see Tipke, 2003, p. 1137; BVerfG of 15.1.2008, 1 BvL 2/04, DStRE, 2008, p. 1006.
In the explanatory memorandum to the GewStG 1936, the legislator did not go into more detail on trade tax splitting; in particular, the objective associated with it was not mentioned. Therefore, the overriding objective of the business tax must be applied as an auxiliary measure. In accordance with the principle of equivalence (Baldauf, 2022, § 28 para 3 GewStG; also Wissenschaftliche Dienste des Deutschen Bundestages, 2019, p. 10), the allocation of a financial equivalent to compensate for the subsequent employee costs incurred by a municipality as a result of the permanent establishment
This includes the costs incurred by a municipality as a result of the employees living there, e.g. for the construction of schools, see FG Hessen of 06.04.2005, 8 K 5273/00 The authors believe that the limitation of the reduction to the wage total is not convincing. For example, it is unclear why the wage and not the number of employees should be taken as a basis when the burden imposed on a municipality by a commercial enterprise can be measured in particular on the basis of the number of employees. The apportionment standard is further criticised for not considering the so-called soft burdens associated with the establishment of commercial enterprises, such as reductions in the value of property and environmental pollution, as well as the increase in the importance of the capital factor. (Offerhaus & Althof, 2006, p. 626 et seq).
The apportionment is based solely on the ratio of wage totals and, in the BFH's view, can be described as simple and undifferentiated (BFH of 12.7.1960 - I B 47/59 S, BStBl. III 1960, p. 386). In doing so, the legislator accepts that the actual circumstances characterising the individual case are not adequately reflected by splitting on the basis of the wage totals (BFH of 9.10.1975 - IV R 114/73, BStBl. II 1976, p. 123). According to the BFH, a determination of the apportionment standard that does justice to the individual case, would be ‘associated with an unjustifiable amount of administrative work’. (BFH of 12.7.1960 - I B 47/59 S, BStBl. III 1960, p. 386). The considerations regarding the apportionment standard are thus guided by practicality considerations (Baldauf, 2022, § 33 GewStG, Rn. 4),
Generally on the equivalence principle see Kraft & Kraft, 2018, p. 197.
The splitting of corporate income tax is intended to compensate for differences in collection between the local collection
In this respect, trade tax and corporate income tax splitting differ with regard to the time of collection. Whereas in the case of corporate income tax splitting, the tax authority collects the tax at the parent company and the tax is only split if the absolute amount exceeds €500,000, in the case of trade tax, the tax is split after the trade income has been determined. In contrast, the splitting of the trade tax is not intended to pursue a financial equalisation, but to serve as compensation for the burdens arising from permanent establishments. See BFH of 9.10.1975, IV R 114/73, BStBl. II 1976, p. 123; BFH of 12.5.1992, VIII R 45/90, BFH/NV 1993, p. 191. The financial requirements of the federal government and the federal states are therefore not decisive for the purposes of splitting, but only the taxes generated in their own sphere. See BVerfG of 24.6.1986, 2 BvF 1/83 et al, NJW 1986, p. 2630. At the third level, financial equalisation takes place between financially stronger and financially weaker federal states (Article 107 para 2 sentence 1 et seq. of the Basic Law) and at the fourth level, the financial transfer of the Federation to poorly performing federal states takes place within the framework of the Federal Supplementary Allocations (Article 107 para 2 sentence 5 of the Basic Law).
The distribution of tax revenue at the second stage is carried out by means of revenue collection by the tax authorities. In this way, local revenue is used as the standard level and serves as an indicator of a country's tax capacity (Seiler, 2022, Art. 107 GG, Rn. 50). According to the Federal Constitutional Court, this is to be understood as the ‘tax performance of the economy and the citizens of the individual federal state’ (BVerfG of 24.6.1986, 2 BvF 1/83 et al., NJW 1986, p. 2631). This should correspond to the actual economic power, understood as gross domestic product,
Literature, case law, and legislators assume this understanding, see for example Renzsch, 2013, p. 405; BT-Drs. 11/3263 of 7.11.1988, p. 1; in this sense also BVerfG of 12.2.2003, 2 BvL 3/00, p. 1364 et seq.
Due to the integrated economic area of the Federal Republic of Germany, in which regional borders are of no significance for companies (Renzsch, 2013, p. 405), the tax power of a federal state is, however, inadequately represented when looking solely at the collection by the fiscal authorities, and distortions occur.
BVerfG of 24.6.1986, 2 BvF 1/83 et al., NJW 1986, p. 2630. The Federal Constitutional Court uses the terminology ‘real tax power’ there.
The Basic Law, like the Federal Constitutional Court, leaves the apportionment standard to be used open; the only decisive factor is that the distortions are reduced ‘in a relevant manner’ (BVerfG of 24.6.1986, 2 BvF 1/83 et al, NJW 1986, p. 2632; Art. 107 para. 1 sentence 2 GG). However, they do not necessarily have to be reduced completely.
The Splitting Act fills in this margin. Pursuant to § 1 (1) of the ZerlG, the claim arising from the corporation tax is due to the federal state in which the management is located. If a corporation maintains one or more permanent establishments outside this federal state, the tax office responsible for levying the tax (Erhebungsfinanzamt) must apportion the tax due on the income from trade or business among the federal states involved if the income amounts to at least €500,000. Pursuant to § 2 para. 1 sentence 2 ZerlG, the provisions of the Trade Tax Act (§§ 28 et seq. GewStG) must be applied accordingly.
The linking of the corporate tax splitting to an existing law indicates a clear value decision in favour of practicability. On the other hand, there is no discussion of a causation-based recording of tax power. Thus, practicability takes the place of causation justice as the dominant objective.
It is also striking that the fundamental principle of taxation, namely the ability to pay, is not addressed here. Rather, a splitting standard is used that is justified by the equivalence principle.
To analyse the objectives of the CCCTB breakdown, it is necessary to expand the analysis to include objectives and requirements, since requirements for formula design are addressed in parallel with goal formulations. However, the 2011 (EU-Com, 2011) and 2016 (EU-Com, 2016b) proposed guidelines are only partially explicit about the goals and requirements of the apportionment formula.
The introduction of a CCCTB should generally lead to efficient, effective, simple, and transparent company taxation in the EU and prevent tax avoidance and abuse (EU-Com, 2001, p. 18). Justice in the sense of causation justice Efficiency in the sense of resistance to manipulation and Efficiency in the sense of practicability
The apportionment mechanism should also be mutually agreed. This will not be considered further here since the fulfilment of this condition must in any case be established in the EU legislative process.
The derivation of these objectives from the working documents is presented below. The working group's understanding of the term is also discussed. This is done in conjunction with the literature that the working group has drawn on and the Commission's understanding of the proposed guidelines.
The working document
In preparation for the discussion on an appropriate apportionment mechanism, two papers, one by Ana Agúndez-García (Agúndez-García, 2006), a research fellow at the EU Commission's Joint Research Centre (JRC), and one by Joann M. Weiner (see Weiner, 2005), Associate Professor at George Washington University, were provided to the panellists.
Both Weiner and Agúndez-García see justice as a central goal (Agúndez-García, 2006, p. 32 et seq.; CCCTB-WG, 2006, p. 4; Weiner, 2005, p. 52). Agúndez-García also comments at length on the possible content of these objectives. Regarding justice, she states that it is not clear what justice is and that this can only be determined by a value judgement (Agúndez-García, 2006, p. 33). She goes on to present three possible value judgments for selecting equitable apportionment factors (Agúndez-García, 2006, p. 33 et seq.).
Profit generated as a benchmark: If the amount of profit generated is the same, the allocated share is the same. The benchmark used for locating the source of profit is the profit of a part of the enterprise. It is considered problematic that the exact sources of profit generation can hardly be localised. Application of profit factors: If the same profit factors are used, the allocated share is of the same amount. Here, in contrast to the first benchmark, the ratio of the quantity/amount of factors of one part of the company compared to the whole group is used as an estimate for the profit share generated. However, it is unclear exactly which factors are to be decisive. Equivalence principle: A company's presence is considered as the benchmark of the benefits that a company can enjoy locally. Appropriate factors used to quantify the presence could include assets, number of employees, and/or number or value of business transactions.
The working group does not address these benchmarks, nor does it explicitly clarify how it believes justice can be implemented. However, it presents three options for formula design later in the working paper (see in detail CCCTB-WG, 2006, p. 5 et seq.).
Macroeconomic distribution factors, e.g. GDP or national VAT assessment base (CCCTB-WG, 2006, p. 5) A value-added apportionment formula that puts the value added by a part of a company in relation to the company as a whole
The ‘VAT returns’, which would then be adjusted accordingly, should be the starting point for the assessment (CCCTB-WG, 2007b, p. 4). The formula-based splitting on the basis of three factors: labour, capital and turnover.
The macroeconomic apportionment formula is rejected because the connection between the actual entrepreneurial activity and the tax payment is missing (CCCTB-WG, 2007b, p. 3). The value-added approach will also not be pursued further. This is due to the high level of effort involved in determining the data required and the ongoing use of transfer prices to determine the relevant variables (CCCTB-WG, 2007b, p. 4 et seq.). Therefore, the working group decided to pursue the formulaic splitting based on the three factors of labour, capital, and turnover. It is emphasised that all approaches show advantages and disadvantages and, in the manner of a traditionalist way of thinking, reference is made to the good experiences in the USA and Canada with formulaic splitting on the basis of these factors (CCCTB-WG, 2006, p. 6).
The use of the three-factor formula corresponds in content to Agúndez-García's second value judgment, but the justification is accentuated differently: While these factors would ‘reflect as closely as possible the source of income generation’ (CCCTB-WG, 2006, p. 6), they are primarily a benchmark for the The working group reinforces this view by using industry-dependent factors in order to do justice to the special features of profit generation depending on industry peculiarities (CCCTB-WG, 2006, p. 10; CCCTB-WG, 2007, p. 20).
While the working documents understand causation-based apportionment to mean apportionment based on profit generation, the EU Commission's 2016 proposal for a directive introduces the link between apportionment and the generation of value added. The CCCTB is an ‘effective tool for allocating revenue where value is created, using a formula based on three equally weighted factors (assets, labour, and turnover)’ (EU-Com, 2016b, p. 2). It is not clear what relationship the EU Commission sees between the concepts of profit generation and value creation.
At the OECD level, for the portion ‘A’ of the first Pillar (Pillar I), it was first assumed that one part of the profit of certain business activities (in-scope activities) arises through significant participation in market states. (OECD, 2020a, p. 8). Now the decisive factor is solely the achievement of sales and profitability thresholds (OECD, 2021, p. 1 ff.) After deciding on the selection of certain formula factors, their weighting must be determined. In principle, the working group considers this to be a political decision (CCCTB-WG (2006), p. 4; CCCTB-WG, 2007a, p. 7). The model for the CCCTB apportionment formula is the equally weighted Massachusetts formula (CCCTB-WG, 2006, p. 10).
In connection with the weighting of the formula factors, the working group cites another central aspect of justice. The weighting could be used to achieve a balance of interests between production and market countries (CCCTB-WG, 2006, p. 10 et seq.). This would be appropriate since company profits can only be achieved if products are also sold (CCCTB-WG, 2007a, p. 14 et seq).
In the USA, there has been an increased weighting of the turnover factor in recent decades, to which the working group refers (CCCTB-WG, 2006, p. 9 et seq.; in this sense also Weiner, 2005, p. 52). Canada also places greater importance on the turnover factor; the formula consists only of the factors labour and turnover.
This type of justice is called In Musgrave's view, it is not possible to justify whether the source principle should be seen as production-oriented (supply approach) or market-oriented (supply-demand approach); (Musgrave, 1984, p. 234).
In the aforementioned working document from 2006 (CCCTB-WG, 2006), the aim is not only a fair mechanism but also an efficient one (CCCTB-WG, 2006, p. 2). However, there is no concrete definition of efficiency. The working group merely refers to the Agúndez-García essay already referred to. According to this approach, efficiency is achieved when decisions are neutral, enforceable, simple, and at the lowest possible cost (CCCTB-WG, 2006, p. 4). These characteristics show an understanding of efficiency that can be divided into two types:
Efficiency in the sense of resistance to manipulation (decision neutrality, enforceability) Efficiency in the sense of practicability (simplicity, cost reduction)
In Agúndez-García's view, the allocation mechanism would be efficient if it were decision-neutral, i.e. if investment decisions were not made dependent on the formula factors. Related to this, enforceability would exist if no tax avoidance by means of profit shifting were possible (Agúndez-García, 2006, p. 38).
Whether there could be distortions in the allocation of the taxable substrate as a result of the formulaic allocation would depend on the extent of the mobility of the factors. The assets factor is seen as highly mobile, while the labour and turnover factors (by destination) are seen as comparatively robust. Whereas in the case of the labour factor, the tax burden can only be influenced by a real relocation of employees,
Intangible assets are not included in the asset factor, not only because of the measurement problem but also because of the existing possibilities of manipulation (EU-Com, 2007, p. 19).
The working document from 2007 also explicitly places the prevention of tax avoidance at the centre of its objectives (CCCTB-WG, 2007a, p. 5). According to this, the apportionment mechanism should be as difficult as possible to manipulate and should not be based on factors with an easily changeable local allocation. This should prevent tax rate differentials within the EU from being easily exploited.
This would also avoid undesirable tax competition in the EU (EU-Com, 2006, p. 12; see also in detail CCCTB-WG, 2007a, p. 19 f). To avoid this, the formula would also have to be designed in a uniform manner, as otherwise distortions would arise (CCCTB-WG, 2006, p. 7; see also CCCTB-WG, 2007a, p. 7).
Remaining manipulation incentives after the implementation of the second OECD Pillar (Pillar II) (OECD, 2020c) in national law are no longer as extensive because of the harmonisation of tax rates (CCCTB-WG, 2006, p. 12; Wissenschaftlicher Beirat des Bundesfinanzministeriums, 2007, p. 53.
Proposed minimum tax policies were the components ‘income inclusion rule’, ‘switch-over rule’, ‘undertaxed payments rule’, and ‘subject to tax rule’. See ‘Tax Challenges Arising from Digitalisation – Report on Pillar Two Blueprint’, Tz. 410 et seq. An agreement was achieved on July 1st by OECD-Countries (OECD, 2021).
In order to ensure efficiency in the sense of practicability, the Agúndez-García vision would have to satisfy the postulate of simplicity (Agúndez-García, 2006, p. 38).
Weiner also sees simplicity and practicability as important criteria when assessing the introduction of a CCCTB (Weiner, 2005, p. 55). Specifically on the definition of assets. Intangible assets are not included, in addition to existing possibilities of manipulation, also due to the difficulties of measurement (CCCTB-WG, 2007, p. 19).
The working group continues to strive for the simplest possible application for taxpayers and tax administration (CCCTB-WG, 2007, p. 5). Such practicability would fulfil a second essential postulate in the form of cost-effectiveness. According to Agúndez-García, this would first be the case if the costs of administration were sustainable for the entire tax system. This would be the case, for example, if already known data can be used, which the company needs for other purposes, or if the data can be easily verified by the administration (Agúndez-García, 2006, p. 38).
Regarding costs, the working group stresses that the extensive costs of a transfer pricing system would be eliminated, at least within the EU (CCCTB-WG, 2006, p. 11). Overall, the splitting of the CCCTB was a ‘simple and cost-effective method’ in terms of obtaining the required data (CCCTB-WG, 2006, p. 12).
The apportionment mechanism of the CCCTB aims at justice and efficiency and the central manifestation of justice is causation justice. For this purpose, a presumed relationship between the factors labour, capital, turnover, and the generation of profit or value added should suffice. Furthermore, the formula is intended to achieve a balance between production and market states.
The goal of efficiency is expressed in a formula design that is as resistant to manipulation as possible, as this should ensure the enforceability of the tax claim. Efficiency should be sought by designing the system in such a practicable way that taxpayers and the administration can act cost-effectively.
In the following, the objectives of the CCCTB apportionment will be compared with those of the splitting of corporate taxes in Germany in order to examine whether lessons can thus be drawn from the splitting of trade tax for international profit apportionment. Therefore, we juxtapose each previously elaborated objectives of the CTTB apportionment with any relevant aspects of the German tax splitting.
The CCCTB strives for a causation-based profit distribution, but this terminology is not used in the German corporate tax code. The
In German taxation, the ability-to-pay principle must always be observed as the fundamental principle of taxation (Hey, 2020, § 3 para. 40).
For further considerations on the ability-to-pay principle in an international context, e.g. with regard to equal treatment of nationals and the resulting need to avoid double taxation (Jakimovski, 2012, p. 146 et seq.).
The apportionment for corporate income tax purposes is based on tax capacity, which must be measurable in terms of performance and thus be determined according to causation. Paradoxically, the corporate income tax apportionment does not use tax capacity on an individual's taxable capacity as a basis for its assessment, as would be consistent with the ability-to-pay principle, but instead uses a macroeconomic variable. Even though the objective of fairness of causation is found in both the corporate income tax apportionment and the CCCTB apportionment, their concretisation differs and is not consistent for the German corporate income tax.
In order to implement the splitting according to tax capacity, the splitting of corporation tax uses the splitting mechanism of trade tax for reasons of practicability. This serves to implement the principle of equivalence. This is unsystematic and can only be justified by conceding a hierarchy of values in favour of practicability.
In the
Figure 1 shows the results of the comparison of the CCCTB's objective causation justice with the relevant aspects of the German corporate income tax apportionment and the trade tax splitting.
Figure 1
Relevant aspects of the German system regarding causation justice
Source: Authors’ own figure
A partial objective of the CCCTB is to ensure that the tax rules do not provide tax avoidance opportunities. The formula must therefore be designed to be tamper-resistant. This objective is not formulated for either trade tax or corporation tax. However, it is implicit in the ability-to-pay principle, according to which equal treatment of taxpayers should also be sought (BVerfG of 16.3.2005 2 BvL 7/00, NJW 2005, p. 2448; for the OECD context see OECD, 2015, p. 34).
Also known as the horizontal dimension of performance. According to the Federal Constitutional Court, horizontal tax justice means that taxpayers with the same ability to pay are also taxed at the same rate, while in vertical terms higher incomes must also be taxed fairly in comparison to lower incomes (BVerfG of 16.3.2005, 2 BvL 7/00, NJW 2005, p. 2448).
Thus, resistance to manipulation is also a partial objective of the decomposition of corporate taxes in Germany. However, it is of lesser importance. This can be seen particularly in the use of only one factor. This is much more susceptible to manipulation compared to a formula consisting of multiple factors. In this respect, there is a dominance of resistance to manipulation by practicality. Figure 2 summarises how the CCCTB's requirement of resistance to manipulation appears in the German system.
Figure 2
Appearance of the objective resistance to manipulation in the German system
Source: Authors’ own figure
Practicability is an explicitly stated objective of CCCTB apportionment and is an essential (Hidien, § 2 para 26 ZerlG) pillar of an efficient apportionment mechanism. The rules should be simple and secure, and therefore cost-effective. It also ensures the predictability of the distribution outcome.
In Germany, causation is clearly dominated by practicability. In contrast, the discussion on the distribution mechanism of the CCCTB reveals equally weighted efforts to also achieve causation justice. In addition, the interests of the market countries should be taken into account, and the formula should be designed to be as resistant to manipulation as possible. These similarities and differences are illustrated by Figure 3.
Figure 3
Similarities and differences regarding resistance to practicability
Source: Authors’ own figure
When it comes to content, the objectives of the splitting of corporate taxes in Germany and the CCCTB apportionment are substantially similar. When comparing the objectives, however, a clear difference in their relationship to each other becomes evident: While the German apportionment of corporate taxes gives higher priority to practicability, the CCCTB's apportionment mechanism attempts to achieve the objectives with equal priority. Figure 4 provides an overview of the similarities in content and the relationship between the objectives.
Figure 4
Comparison of the objectives of splitting according to German corporate taxes and the CCCTB
Source: Authors’ own figure
The comparison shows that objectives are comparable in essential aspects, but are not consistently implemented in Germany for reasons of practicability. The main funding of this study is the clear value decision towards practicability, although there is a different set of aims:
Justice should be sought in CCCTB sharing that coincides with the principle of ability to pay immanent to German taxation. However, there are difficulties associated with the implementation of a causation-based and thus performance-based profit distribution. In particular, the discussion on the exact relationship between the apportionment factors and the generation of profits is controversial (Section 4.2.2.2).
This objective is not even pursued in the national apportionment of corporation tax; rather, it is more important to find a practicable apportionment standard. Even if this cannot be measured completely independently of performance, Germany accords practicability a greater value. This value decision is also found in the splitting of trade tax.
In regard to the national splitting of corporation tax, it should be noted that, despite the specific objective of approximating tax capacity, a formula with only one factor is considered sufficiently representative. This could be a lesson for international profit sharing to gear the requirement for justifying formula factors not so much to performance and the associated causation fairness, but even more clearly to practicability. Such an objective could be particularly relevant when searching for additional or different factors to reflect digital business models.
At the national level, it should be noted that levels three and four of the fiscal equalisation system provide for aid to financially weak countries (Section 4.1.2). This could be one reason why practicability is highly accepted, but there are also redistribution provisions at the EU level within the EU budget as well as further financial aid (Kullas et al., 2016). It should still be examined to what extent this could have an impact on the acceptance of a practicability design.
A direct comparison of the objectives shows that distributional issues on the sales side and resistance to manipulation are not discussed in connection with the national splitting. While the former addresses a typical international distributional conflict between production and market countries, manipulation resistance certainly addresses a problem that is not unknown at the national level (Eichfelder & Zander, 2018, p. 1313 et seq.).
In the national splitting approach, labour is used despite the known potential for design. This also endorses its usability for international profit sharing. By basing its allocation on three factors, however, the CCCTB goes much further here.
The splitting of corporate taxes in Germany is recognisably committed to practicality by being based solely on the labour factor. This makes them easy to implement, and the effects are transparent and predictable. This would also address key objectives of the CCCTB apportionment formula. However, the latter strives more for justice, especially in the sense of causation justice, and for efficiency in the sense of resistance to manipulation.
In view of the mass procedure to be expected when the CCCTB is applied, the strong orientation of trade tax splitting towards practicability could serve as a model for a simpler design of profit apportionment in the CCCTB, especially since it has not yet been clarified what a causation-based formula design should look like. For the implementation of a formulaic EU profit sharing, the lesson to be learned is that practicability should play a central role in the design of the formula. This lesson is helpful in accompanying and supporting the implementation process in the EU.