Prof. dr. Maarten Vink

Johan van der Valk

Sem Duijndam

The qualifying foreign taxpayer obligation (hereafter: QFTO), which entered into force on 1 January 2015, establishes that non-resident taxpayers in the Netherlands may benefit from the same deductions and tax credits as resident taxpayers only if they earn at least 90 per cent of their global income in the Netherlands. Under this new system, these non-resident workers, if they neither earn 90% of their world income in the Netherlands, nor have a sufficient taxable income in their country of residence, risk forfeiting tax benefits (e.g. mortgage-interest deductions for owner-occupied dwellings). Moreover, the rule may especially impact frontier workers and have detrimental economic effects if such non-resident workers decide against employment in the Netherlands and prefer to work in another country. In such a scenario, employers in border regions should be concerned, given that the majority of non-resident workers are employed in areas along the Dutch border. In this inventory of the potential impact of the QFTO, we focus on the group of persons who are employed in the Netherlands, but reside outside of the Netherlands, as they are likely the largest group affected by the rule. The objective of this preliminary ex-post analysis is to examine trends over the years from 2013 to 2016 in the number of non-resident employees in order to see if notable changes occurred in the number and composition of non-resident employees in the Netherlands after the 90%-rule came into force.

Table 1 shows the number of non-resident workers in the Netherlands for the years 2013-2016, as well as the nationalities and countries of residence of the non-resident employees. The number of non-resident employees has increased considerably over this period. Where in 2013 the number of non-resident employees was a little more than 130.000, this number increased to over 185.000 in 2016. This increase, however, is mainly due to the large influx of Polish non-resident workers in this period. The number of non-resident workers living in Belgium or Germany increased just slightly. When we look at Dutch non-residents we see that they mostly live in Belgium or Germany, and that their number increased slightly since 2013.

Looking at employment sector, we see that most non-resident workers work in commercial services. These nonresident workers mainly have the Polish nationality. It is therefore not surprising that the number of non-residents employed in the commercial sector increased sharply since 2013 (from 85.800 in 2013 to 133.300 in 2016), corresponding with the large increase in the number of Polish non-residents over the same period. The number of non-residents working in the industrial sector or public and social services remains fairly constant around 20.000 for the years 2013-2016. Both these sectors mainly employ Dutch nationals, although they also employ a considerable number of Belgians and Germans. Few non-residents work in agriculture, forestry, and fishery and there are also no notable changes visible.

More than half of the non-resident employees work in the cross-border regions. Most of these non-resident workers live in either Belgium or Germany. This is also clearly depicted in Figure 1, which shows the number of nonresident workers residing in Belgium or Germany as a percentage of the total working population for the year 2016 (only this year is shown, because there is not much change over time). Unsurprisingly, most non-resident workers in cross-border regions at the German border are German, whereas those at the Belgian border are Belgian. Some border regions share a border with both Belgium and Germany (Midden-Limburg and Zuid-Limburg). In Midden-Limburg 3.6% of the working population in 2016 lived in either Belgium or Germany, while in Zuid-Limburg this was 5.6%. For most (border) regions the shares remained almost constant over the period 2013-2016, and no common trend is visible. The share of non-resident workers residing in Belgium or Germany over the total working population remains constant at 1% from 2013-2016.

Overall, the preliminary ex-post analysis does not seem to show any compelling effects of the QFTO on the number and composition of non-resident workers in the Netherlands and the Dutch cross-border regions. When we look at the total number of non-resident workers we see an increasing trend, which is persistent over time and does not seem to have been altered since the implementation of the QFTO. For the nationality and work sector of nonresidents we also observe a solid trend over the whole period 2013-2016; the number of Polish and “other” nationals increases, as well as the number of non-resident workers working in commercial services. For the border regions there are also no significant changes visible. However, this analysis does not allow us to focus on those individuals that are most likely affected by the QFTO (those who do not earn 90% of their world income in the Netherlands). Furthermore, the possible delayed effects of the rule cannot yet be assessed, as data is only available until 2016.

For future work, income data from the Dutch Tax Authority will become available, which makes it possible to assess which non-resident workers do not earn 90% of their world income in the Netherlands, and hence which non-resident workers are most likely to be affected by the QFTO. For future inquiries individuals can also be followed over time to research their exact labour and housing mobility. With the use of regression and/or timeseries techniques it can be shown if the QFTO has a significant effect on the housing and labour mobility of nonresident employees.