Introduction and EU law applicable
On 8 March 2017 the Court of Justice of the European Union (hereafter “CJ”) delivered its decision in the Case C-448/15, Wereldhave, dealing with the interpretation of the “subject to tax” requirement provided for the application of the Parent-Subsidiary Directive (both Directive 90/435/EEC and Directive 2011/96/EU). Under article 1(1) of Directive 90/435/EEC, applicable ratione temporis, the latter applies to distributions of profits by subsidiary companies of a Member State to parent companies of other Member States. In turn, article 2 of that Directive defines “company of a Member State” as any company that meets three subjective requirements, one of which being that it must be “subject to one of the [listed] taxes, without the possibility of an option or of being exempt” (article 2(1)(c)).
Facts of the case
The case concerned the distribution of dividends by a Belgian subsidiary to its two Dutch parent companies, holding respectively 35% and 45% of the capital of the former. The relevant dividends, distributed in 1999 and 2000, amounted to approximately Euro 11 millions per year. The parent companies applied to be exempt from the Belgian withholding tax on the dividends, maintaining that they should have been regarded as companies of the Netherlands for the purposes of the application of the Parent-Subsidiary Directive.
In the absence of a decision from the Belgian authorities, the Dutch companies brought an action before the Court of First Instance of Brussels, which held in favor of the applicants. The Belgian State brought an appeal before the Court of Appeal of Brussels, which stayed the proceedings and referred two questions to the CJ for a preliminary ruling.
The first question concerns the possibility to apply Directive 90/435/EEC to dividends received by parent companies, such as those in the main proceedings, which, although being liable to tax under Dutch law, are in fact subject to a zero rate tax in the Netherlands as Fiscal Investment Institutions, provided that all of their profits are paid to their shareholders. In this respect, the referring court doubted whether such companies could be regarded as “subject to one of the [listed] taxes, without the possibility of an option or of being exempt”, as provided for in article 2(1)(c) of the directive.
The second question deals with the issue of whether the freedom of establishment and the free movement of capital should be construed as precluding the taxation of those dividends in Belgium. According to the CJ, however, the absence of any details, in the request submitted by the referring court, regarding the national legal framework applicable to the payment of dividends to similar resident companies, makes it impossible to determine whether the EU fundamental freedoms preclude taxation in Belgium.
The ruling (first question)
According to the CJ, article 2(1)(c) of Directive 90/435/EEC lays down both (i) a positive requirement, i.e. the relevant company must be liable to tax, and (ii) a negative requirement, i.e. it must not be exempt from tax and not have the possibility of an option (p. 31). This double requirement excludes from the scope of application of the directive companies that are not actually liable to pay one of the listed taxes (p. 32). In this respect, the CJ, referring to the opinion of Advocate General Campos Sánchez-Bordona, holds that the application of a zero tax rate is equivalent, in practical terms, to an exemption from tax (p. 33-34).
This conclusion, based on the wording of the directive, is further supported on the basis of a teleological interpretation thereof (reference to the logic and the objective pursued, at p. 35). Indeed, according to the CJ, the directive is intended to cover situations where, in its absence, the exercise by the Member States of their taxing powers might lead to the profits distributed by subsidiaries to their parents being subject to double taxation (p. 39). However, where a parent company is entitled, under the legislation of its Member State, to a zero rate of taxation for all of its profits, the risk of double taxation is ruled out and, therefore, there is no need for the directive to apply (p. 40-41).
The decision of the CJ is worth a comment under, at least, five different perspectives.
First, with regard to the arguments used by the Court to support the ruling, the reference to the telos of the directive is not convincing. Recital 3 of Directive 90/435 provides that “existing tax provisions which govern the relations between parent companies and subsidiaries of different Member States … are generally less advantageous than those applicable to parent companies and subsidiaries of the same Member State; … cooperation between companies of different Member States is thereby disadvantaged in comparison with cooperation between companies of the same Member State; … it is necessary to eliminate this disadvantage by the introduction of a common system in order to facilitate the grouping together of companies”. The goal of facilitating the grouping together of companies of different Member States within the internal market is pursued through the elimination of source taxation (article 5 of the directive) and the application of the exemption/indirect credit methods in the State of the parent company (article 4). Both measures are necessary in order to eliminate the economic double taxation that characterizes cross-border profit distributions as compared to purely domestic profit distributions. Where the dividends distributed by the subsidiary are actually exempted (through the application of a zero rate tax) in the State of the parent, double taxation remains as long as the State of the subsidiary, which has already taxed the latter on its profits when derived, continues to levy a withholding tax on their distribution to the parent. From a policy perspective, there is nothing more unreasonable than withdrawing the benefits of article 5 of the directive – and, thus, allowing the State of the subsidiary to perpetrate the double taxation – based on the argument that the State of the parent company has done its job by exempting the dividends received by the parent. The ruling is not unreasonable on the basis of the wording of article 2(1)(c), but, frankly speaking, it is difficult to justify on the basis of a teleological construction of the directive.
A second relevant aspect of the ruling deals with the possible distinction between subjective (i.e. concerning the taxable person) and objective (i.e. concerning only certain items of income, in particular the dividend distribution) exemptions for the purpose of article 2(1)(c) of the directive. The CJ has, obviously, taken a position only in respect of subjective types of exemptions, due to the specific facts of the main proceedings. Therefore, drawing an a contrario inference from the ruling and the underlying arguments is a risky and partially unfair exercise. That said, the impression is that both the CJ and the Advocate General attribute relevance exclusively to cases of full exemption, which generally concern instances of subjective, as opposed to objective, types of exemptions (e.g. p. 32-33 of the Court’s decision). In particular, according to the Advocate General, exemption within the meaning of article 2(1)(c) of Directive 90/435 implies that “no payment of the relevant tax is required as the legislature has deemed fit to release a particular class of companies from the obligation to pay the tax” (p. 42 of his opinion), which is true when, “by means of an express legal provision”, a Member State “establishes permanently and in advance [a complete absence of tax] for a certain class of bodies, irrespective of the profits received” (p. 44 of the opinion).
Limiting the concept of exemption relevant for the purpose of article 2(1)(c) of the directive to cases of subjective exemptions is also supported by the systematic interpretation of the directive, being that concept used in order to define the personal scope of application of the directive and not, more generally, to single out cases where the existence of exemptions makes the benefits of the directive inapplicable. Moreover, it would not make much sense, for the same reasons previously highlighted with regard to the teleological interpretation of the directive, to withdraw the benefits of the directive in cases where the Member State of the parent company exempted under its domestic law the dividends received by the latter, since such exemption would clearly pursue the general goal of the directive. This interpretation appears also to play a relevant role in the application of the Interest & Royalties Directive, as it entails that, rebus sic stantibus, the interest and royalties falling within the scope of application of that directive should benefit from the withholding tax exemption, even where they were not taxed in the hands of the recipient due to a specific objective exemption available in its Member State. To ensure that interest and royalties were taxed at least once in a member State, it would thus be necessary to amend the provisions of the directive.
Third, both the CJ (p. 34) and the Advocate General (p. 43-44 of his opinion) appear to limit the cases of exemptions relevant for the purpose of article 2(1)(c) of the directive to full exemptions specifically granted by the law (in advance and in abstracto), while excluding the relevance of instances where companies end up paying no tax in a specific tax year because of merely factual and accidental circumstances, such as being in a loss position, having losses carried forward, or deriving only specific exempted income (e.g. dividends).
Fourth, the decision seems to rule out partial exemptions, as they leave companies actually subject to tax with regard to a portion of their income (p. 32-34).
Finally, the fact that a parent company falls outside the scope of application of the directive, due to any subjective exemption granted by its Member State of residence, does not preclude the possibility for it to have indirect access to similar benefits under the EU fundamental freedoms. Specifically, the freedom of establishment and the free movement of capital provided for in the Treaty on the Functioning of the European Union (hereafter TFEU) oblige the Member State of the subsidiary to grant the non-resident parents with same treatment that it would apply to dividends distributed to comparable resident parents. In this respect, where the law of the Member State of the subsidiary seeks to prevent dividends distributed by resident companies being subject to a series of charges to tax, the situation of a resident parent is generally comparable to that of a non-resident parent (CJ, Case C-303/07, Aberdeen, p. 43-44; Case C-284/09, Commission v Germany, p. 58; Joined Cases C-338/11 to C-347/11, Santander, p. 42). In particular, where a Member State makes the tax treatment of the dividends distributed by its resident subsidiaries (taxation vs. exemption) exclusively dependent on the place of residence of the parent (non-resident vs. resident), the tax treatment of dividends distributed to non-resident parents is discriminatory regardless of whether the parent is subject to tax in the State of residence (see CJ, Case C-303/07, Aberdeen, p. 51) and regardless of the tax situation of the shareholders (CJ, Joined Cases C-338/11 to C-347/11, Santander, p. 28 and 39). The levying of a withholding tax is, therefore, contrary to EU law, unless justifiable on the basis of imperative reasons in the public interest. Only where the domestic dividend exemption is conditional on the tax treatment of the parent company’ shareholders, that treatment becomes relevant also with regard to non-resident parent companies, in order to assess their comparability with resident parents and, therefore, to establish where a discriminatory treatment forbidden by the TFEU is at stake (CJ, Case C‑194/06 Orange European Smallcap Fund, p. 33 and 60).