Luxembourg has a specific situation in Europe when it comes to commuters: 150,000 commuters are working daily in Luxembourg. The tax situation of those employees is very complex because of the bilateral nature of the double taxation treaties (DTTs) within the European Union, which is a (common) market of now 29 countries.
According to the OECD Model Treaty for the avoidance of double taxation, the taxation right of all (the worldwide) income of a person is in principle allocated to the residence country of that person. There are however important derogations to this principle, one of the most important being the taxation right for salaries. As a commuter is working and producing added value in an enterprise situated in another country, i.e. in the country where he is employed, the taxation right for salaries is allocated to the country where the employer is established. And this rule prevails worldwide for many decades now.
But bilateral DDTs are subject to interpretations by the two partner countries, which might change over time, and especially in times of financial stress for Finance Ministers. Because of their bilateral nature, DTT further do not take into account the EU Common Market and situations where more than two countries are involved. These triangular situations lead to overlapping taxation competencies between Member States of the EU, because the EU has not managed to settle the criteria for the attribution of taxation rights in such situations. A commuter working for an enterprise operating at European level thus might be subject to two or more bilateral double taxation agreements and may be torn between different claims from different Member States may insist more or less on one of the following criteria that give them taxation rights:
- (the residence of) the employer,
- (the residence of) the employee, and
- the presence of the employee during worktime.
Basically, the situation of a commuter with an employment contract in Luxembourg, who is sent to work in a third country should not be treated differently from the – bilateral – situation, where a commuter is sent back, for a limited period of time, in his residence country. This last situation in fact can also be considered as a “hidden” triangular situation.
The allocation of taxation rights in the OECD model agreement is based on the hypothesis that an employer is involved in the relation between the commuter and the tax administration(s), because the employer is obliged (as an agent of the tax administration) to withhold payroll taxes on the salary of his employees, which is however a tax that is finally due by the employee.
A partner of a bilateral treaty wishing to obtain a higher chance to obtain taxation rights on the salary of a commuter than the other partner country, thus should try to have at least two of the three above mentioned criteria (residence of employer, residence of employee, place where the work is exercised) “on his side”. And that is what is happening at present. At least some EU Member States are trying to re-interpret the terms of the existing DTTs in order to obtain more taxation rights on salaries. In fact the whole thing is nothing else that a “battle around the cake”.
Thoroughly thought to the end, the decisive point is effectively “where the employee is”. In fact, when taking into account the fact that the employer and the employee are (during the daytime) normally present in the same country, the country of the employer has two (of the three above mentioned) criteria on its side, that should prevail over the one criterion of the residence of the employee and this for the whole year (article 15, the article in question of the OECD Model Treaty contains a 183 days test), whether the employee is really present in that country at a given day or not.
Nobody contests that it is abusive, when a employment contract with an employer in another country is concluded for the unique reason to benefit from lower tax or social security rates, while the employee is never or very rarely in the employer’s country. Such cases can however be solved by anti-abuse rules and automatic exchange of information. But the real problem is that anti-abuse rules and automatic exchange of information still do not bring more money into empty revenue coffers of the cash-hungry countries. So these solutions do not satisfy them. They thus now are “splitting hairs”, by referring to the fact that some double taxation treaties use the term “resident” to designate the commuter in order to have two of the three criteria (the residence country and the “resident” employee) “on their side” (while other double taxation treaties use the term “employee”).
But this is not all: despite the fact that bilateral treaties, as their name indicates, cannot regulate, or encroach on, the taxation rights of third countries, certain countries now also claim the taxation right for the residence country, when the commuter is present in third country. And this on a daily basis! But besides the fact that a bilateral treaty cannot regulate the taxation rights of third countries, such an approach seems also to be in conflict with the freedoms of the EU Treaty. According to the EU Court of Justice, administrative hindrances can in fact constitute an obstacle to the exercise of the freedom of movement for workers (article 45 of the EU Treaty).
One might wonder why these cash-hungry countries do not refer to the fact that the commuter normally daily returns in his residence country in order to sleep there, and that, without sleep, one cannot efficiently work (in the country of employment).
But at least for the hidden triangular situation of a commuter send back to Germany, Germany has now imposed its point of view and forced Luxembourg (with methods not far away from blackmailing) to accept a modified double taxation treaty: a German resident commuter, working in Luxembourg, who has no employer in Germany, now cannot work (for his Luxembourg employer) for more than 20 days in Germany without being obliged to pay taxes on its salary in Germany (when exceeding the 20 days limit; see documentation below).