Wheeler merkt onder meer het volgende op over het in het OESO-Modelverdrag nagenoeg ontbreken van koppeling van de belastingplicht in het ene land aan de belastingplicht in het andere land ter zake van het toegewezen inkomen:27
“3.3.3. Separation of ownership and tax liability
The allocation rules of the OECD Model are expressed to apply to the person who has various connections with the income, all of which seem to be based on ownership of the income rather than the obligation to pay tax in respect of it. Yet under the domestic law of many countries it is possible
that one person owns income but that a different person is taxed on it. This was the situation in the Russell and Willoughby cases, as in both cases the income was paid to, and legally owned by, the company whereas the legislation in question attributed the income to the individual. In both these cases the tax charge on the individual was imposed under anti-avoidance legislation, but the same phenomenon can also be found outside the avoidance context. The Netherlands, for example, allows companies within a fiscal unity to shift the tax liability of a subsidiary to its parent company. This is achieved by deeming the assets and activities of the subsidiary to belong to the parent for tax purposes,167 so that the tax liability also falls on the parent, although the subsidiary remains the owner of the income.
Nevertheless, it is probably anti-avoidance legislation that most often imposes a tax charge in respect of income on a person who does not own it. One common example is the attribution of the investment income of minor children to their parents, in order to prevent income splitting to take advantage of the lower rates in a progressive rate table. (…)
(…).
In the Netherlands, recent anti-avoidance legislation has been adopted in respect of trusts, which quite consciously attributes income to a person who may never receive any benefit from the income. If an individual resident in the Netherlands settles property on a discretionary trust, the law may now attribute the trust income to the settlor and, after the settlor’s death, to his/her heirs, regardless of whether they have any benefit from, or control over, the income or even whether they know about the trust. Doubtless many other examples can be found in different countries, although they may not be as extreme as some of the legislation discussed here.
In all these situations, the question arises as to the relevance for treaty purposes of the tax liability being imposed on a person who is not the owner of the income. The Netherlands clearly thinks that this is not an issue in the context of the fiscal unity regime; the law was changed in 2003 to use the mechanism described above precisely so that the subsidiary would continue to be able to claim treaty benefits. The previous law deemed the subsidiary to be a branch of the parent company, thereby removing its personal liability to tax altogether and causing it to fail the residence test in Art. 4 OECD Model. The current mechanism, of retaining the potential personal tax liability of the subsidiary but then shifting the attribution of the income to the parent company, in other words relies on the lack of structural importance given by the OECD Model to the liability to tax on
the specific item of income.
Is the converse also true, that a person who does not own income but who is liable to tax in respect of it cannot claim treaty benefits? On the face of the treaty wording it would seem that this person cannot claim treaty benefits, as the income is not “paid to”, “received by” or “derived by” this person. If the income is passive income, there is a further hurdle in that the person claiming treaty benefits on the basis of his tax liability would have to argue that he is the beneficial owner of the income even though he is not the owner. In the Willoughby case the Special Commissioner did indeed draw a distinction between a partnership case, in which both parties have some ownership interest in the income, and the case before him, in which income was deemed to belong to the individual although it was not his “in reality”.
Yet this is an extremely difficult distinction to draw; at which point does the legislation lose touch with reality? Much anti-avoidance legislation is based on the underlying notion that the avoider does have some ownership connection with the income, even though the connection has deliberately been made rather remote in legal terms. The legislator would usually assert that the effect of the legislation is to restore reality, not distort it. Anti-avoidance legislation of this type that applies in a cross-border situation also raises another issue, as the result is often that the income is taxable in the hands of two persons, the legal owner in one state and the person subject to the anti-avoidance regime in another state. Even if the anti-avoidance regime grants a credit for tax paid by the legal owner of the income, the liability of both persons in respect of the full amount of the income remains.”