National approaches of EU Member States in concluding bilateral social security agreements with third countries

Introduction

It is impossible to achieve a complete overview of the social security bilateral agreements of all Member States, because there are so many of them and because there is no common database. Even for a particular country it can be difficult to obtain a complete overview. In the words of the European Commission: the process of making bilateral agreements is highly fragmented; EU Member States generally negotiate bilateral agreements without reference to what their partner EU countries are doing. ¹ In order to gain some insight into how the Member States conclude bilateral agreements with third countries, I shall discuss the experience of the bilateral agreements concluded by the Netherlands. This is a country with a long history of migration to and from the territory and is one of the founding States of the EEC, and for this reason the characteristics of the agreements made as well as the main legislative developments can provide a mirror for discussion of the bilateral agreements of other Member States. This contribution, using the example of the Netherlands, discusses the reasons for the Member States concluding bilateral agreements with third countries and the main characteristics of the bilateral agreements that have been made.

Reasons for making bilateral agreements

There is already quite a long history of bilateral agreements which started in Europe in the early twentieth century, when, in this part of the world, the first statutory social security schemes were developed. Bilateral agreements were considered necessary since otherwise workers might have to pay contributions in two countries simultaneously, might not be insured at all, or might not be able to receive the benefits to which they were entitled.

The adoption of the E(E)C/EU coordination Regulations (Regulation 3 ² and 4 ³ and its successors, Regulations 1408/71 ⁴ and 574/72 ⁵ and Regulations 883/2004 ⁶ and 987/2009 ⁷ ) removed this reason within the EEC/EU ⁸ and, as a result of this, bilateral agreements between EU Member States only supplement the Regulations’ coordination rules. This is different for bilateral agreements made with third countries.

As regards the bilateral agreements concluded by the Netherlands, several stages or generations of agreements can be distinguished. In the 1960 and 1970s, there was a shortage of workers, and, for this reason, migrant workers were actively recruited. In order to facilitate their movement, the Netherlands concluded bilateral agreements, inter alia, with Turkey, Morocco and Tunisia. Earlier, immediately after the Second World War, considerable numbers of Dutch nationals had emigrated to the New World, including the United States and Australia and, in the 1980s, it was deemed necessary to coordinate the social security provisions relevant to them. ⁹

Since the turn of the twenty-first century, the policy of the Netherlands towards exporting benefits has become much stricter than before. Export is now possible only if is clear that the conditions for benefit are being fulfilled and that supervision of accurate payment is possible. This also means that Treaties were, and still are, concluded with other countries to regulate the export of benefits.

Apart from these ‘supervision Treaties’, very few new Treaties are concluded nowadays, and, as the government explicitly admits, it happens only if there are particular economic motives to make or revise a Treaty. ¹⁰ Thus, in 2016 a Treaty with China was concluded.

For each of the generations of agreements there may be particular characteristics of the provisions of the agreements. These are discussed in the following section.

Characteristics of bilateral agreements concluded by the Netherlands

International Conventions may serve as an example for bilateral Treaties, may have to be elaborated in bilateral Treaties or may be used to coordinate bilateral Treaties. They may also replace these Treaties and for this reason it is useful to mention them first.

An early global Convention on coordination is ILO Convention 48 which concerns the Establishment of an International Scheme for the Maintenance of Rights under Invalidity, Old-Age and Widows’ and Orphans’ Insurance. ¹¹ This Convention was ratified by 12 countries, of which six are current EU Member States. Three of them have, however, denounced this Convention. ¹²

A later Convention, Convention 157 on the Maintenance of Social Security Rights, ¹³ is even less popular, since it has been ratified by only four countries. ¹⁴ Unlike Convention 48, Convention 157 is not restricted to long-term benefits and is to be elaborated by bilateral and multilateral instruments giving effect to the obligations it imposes. Thus, States that ratify this Convention can define which conditions have to be satisfied. This Convention has, inter alia, a rule on the applicable legislation: Member States have to determine this by mutual agreement. These agreements are to ensure that employees are subject to the legislation of the State of employment. ¹⁵

These Conventions have not been ratified by many countries, and certainly not by many EU Member States. ¹⁶ Still, they could, and can, be used as an example for drafting Conventions. The coordination system of Convention 48 has been a model for several bilateral and multilateral coordination Treaties and for the chapters on long-term benefits of Regulation 3 of the EEC. ¹⁷ This Convention was based on social security schemes which (as they all did at the time) made a link between the duration of insurance and the level of benefit. Currently, however, several countries have systems which do not make such a link (these are Type A schemes in the EU terminology). Therefore, the Convention can no longer serve as an example for countries with such a scheme.

The Netherlands did not ratify Convention 157, since it considered its personal scope (which included self-employed and non-economically active persons and persons from countries that had not ratified the Convention) to be too broad. In addition, the absence of a reciprocity rule was seen as problematic. This was at odds with the fact the Convention had to be worked out in further detail by bilateral and multilateral agreements requiring reciprocity. ¹⁸ Furthermore, the Dutch government did not expect a large number of ratifications. ¹⁹ Thus, the relevance of the Convention is limited and, instead of ratifying the Convention, bilateral agreements have been concluded. However, the Convention has a provision for aggregating periods of employment in more than two Member States (Article 7(4)). It thus has a multilateral effect that cannot be achieved by bilateral agreements.

Another framework for making bilateral agreements with third countries is the European Convention of Social Security of the Council of Europe. On the basis of reciprocity, this Convention extends its territorial scope to the countries that have ratified it. Turkey is, however, the only non-EU Member State that has ratified this European Convention, so it has relatively little relevance for the topic of bilateral agreements with third countries.

In addition, the EEC/EU coordination Regulations have been an important example for bilateral agreements. In the following sections this will be discussed.

First generation Treaties: Conventions with countries from which migrant workers came to Europe

Second generation: Agreements with countries to which many Dutch nationals emigrated

In the 1980s in particular, Treaties were made with countries to which Dutch nationals migrated, like the United States, Canada, Australia and New Zeeland. In some cases these persons returned to the Netherlands, and for that purpose coordination rules were also required. However, by 1956, a Treaty was concluded as it was deemed necessary, inter alia, to determine the applicable legislation and to compensate for the long waiting periods for American old-age benefits. These agreements follow the coordination principles of the earlier Treaties, although there are some differences. For instance, in the Treaty with the United States the aggregation of Dutch periods for the purpose of claiming American benefit applies only if a person has completed at least six quarters of coverage under United States law (Article 15). In the reverse situation, 12 months of insurance under Dutch social security law is required. ²⁸

Third generation: Supervision

While the agreements made with countries like Morocco and Turkey in the 1960s and 1970s were meant to facilitate a smooth coordination system, allowing for the export of benefits, the determination of the applicable legislation and the prohibition of non-discrimination, which were inspired by the EU coordination instruments, from the turn of the twenty-first century the bilateral agreements concluded by the Netherlands became more focused on the supervision of a proper application of the benefit rules.

This meant that some Treaties were revised, and Treaties were concluded with some new countries with only supervision rules in the case of the export of benefits to the exclusion of rules determining the applicable legislation and the aggregation of periods (Treaties with South Africa, Surinam, Indonesia, Thailand, and Hong Kong).

These Treaties became necessary because of the adoption of the Act on restricting the export of benefits, which provides that benefits are in principle no longer exportable, unless a Treaty is concluded with that country. In the Treaty concerned, specific enforcement rules had to be included to enable the administrative authorities to verify whether a benefit has been paid according to the applicable rules. These rules have to ensure, inter alia, that the recipient of benefit is indeed the person who is entitled to the benefit, for which purpose his documents have to be verified. For family benefits it is necessary, for instance, to verify that the child concerned is still alive, is following the requirements (e.g. by going to school) and is really the child of the insured person. The country in question must allow the Dutch benefit administration to investigate the situation, if necessary, on its territory. Such enforcement rules are absent in the traditional Conventions and also in EU coordination instruments, but were now required. For this purpose many Conventions had to be revised. This required cooperation with the Contracting State. Since without this cooperation the Dutch Act no longer allowed the export of the full benefit, there were often problems in persuading the other State to cooperate. This was a huge process, requiring elaborate negotiation, and it is sometimes asked whether the costs actually outweighed the benefits from the change.

International Conventions could, however, oblige the Netherlands to continue to pay benefits even without the adjusted agreement. ILO Convention 48, for instance, required the export of invalidity, old-age and survivors’ benefits regardless of the place of residence of the person concerned. Convention 48 therefore obstructed the Act on the restriction of exporting and, for this reason, the Netherlands subsequently denounced it. ²⁹

An additional amendment to the Treaty with Turkey was made in order to be able to verify claims and payments. The competent Turkish authority was obliged to provide proof that the data on the claimant were correct and to specify what information had to be provided, and the competent Dutch bodies were given the right to contact the Turkish authorities or the claimants themselves and their representatives in order to verify the data. The competent Dutch authorities can also refuse, suspend or withdraw a benefit if the applicant or the Turkish administrative authorities provide insufficient information. ³⁰

However, this Treaty was never ratified by the Turkish Parliament because Turkey appeared to be unhappy with the continuous reduction of the rights to social security of Turkish nationals. Decision 3/80 and the European Convention on Social Security required the export of a number of benefits without further conditions on verification, and consequently in practice the export of benefits continued to take place. ³¹

Another development is that the Dutch government wanted to reduce the level of some benefits to the actual costs of living in a country, and this was realised by the 2012 Act on the principle of residence in social security. This reduction is (for Turkish nationals) at odds with Article 5 of the Netherlands-Turkey Treaty and Article 6 of Decision 3/80. This type of provision can also be found in many other bilateral Treaties.

The government argued that the idea is that the level of benefit is first reduced to the costs of living of a country and then that amount is fully exported. However, the Dutch courts did not accept this reasoning and argued that the reduction is incompatible with Article 6 of Decision 3/80. ³² So again, bilateral agreements need to be revised in order to introduce this new rule.

Twenty-two Treaties need to be revised for this purpose, which may take several years. Such negotiations are not always easy, for instance with Morocco. When the Netherlands proposed to revise the existing agreement with Morocco in order to lower benefits, there was a significant diplomatic conflict, during which Morocco threatened, inter alia, to refuse to accept Moroccans who were evicted from the Netherlands (rejected asylum seekers or persons with a criminal record).

Fourth generation: Agreements with countries with intensive trade relations

As of 2008, the Netherlands will only conclude new bilateral Treaties on social security coordination when there is an economic interest involved, in particular where employers no longer wish to pay double contributions, and the need to determine the applicable legislation is relevant. Such Treaties have recently been concluded with China (2016), Japan (2009) and India (2011).

These agreements are focused, in particular, on the posting of workers to and from these countries. For China the Treaty contains posting rules with a maximum period of five years. This provision is meant to limit the obligation for enterprises (so indeed a trade interest) to pay double contributions and to buy voluntary insurance. The reason for concluding this agreement is that in 2011 Dutch workers in China were required to pay Chinese social security contributions. Since the protection by Chinese social security is still quite limited, the Dutch workers were often brought under the Dutch social security system or private insurance was bought for them by the employer. The Treaty was meant to remedy this. The Treaty also states that Chinese workers working in China are no longer covered by Dutch social security.

Since the material scope of the Treaty is limited to old-age, survivors’ and unemployment benefits, it has only a limited material scope (disability benefit, sickness benefit, family benefits and health insurance are not covered). As a result, migrant workers have to buy, inter alia, additional health care in the partner Country.

Since the Chinese social security system is still in an early stage of development, China restricted the agreement to old-age, unemployment and survivors’ benefits (so that Chinese expatriates remained covered by the other schemes). The idea is that after ten years of the agreement coming into force, whether its material scope should be extended will be considered. There may be a problem when a Chinese national comes within the scope of Regulation 1231/2010 (e.g. if a Chinese worker was insured in another Member State before s/he became insured in the Netherlands). In that case, a maximum posting period of 24 months is applicable and after this period s/he will be bound by the legislation of the State of employment. The Treaty does not pay attention to this issue. It permits deviation from the general rules if it is in the interest of the persons concerned (like article 16 of the EU coordination Regulation), but for that purpose the Chinese and Dutch authorities have to agree.

Conclusions

The first generation Treaties were inspired to a large extent by ILO Convention 48 and EEC Regulation No 3 and they adhered to the main principles underlying the Regulation. The export of long-term benefits was regulated in the same way as within the EU and no reduction of benefit was allowed. Whereas EU instruments did not require reciprocity, this was the case for the bilateral agreements. One advantage was that relevant facts about the Contracting Country were taken into account (such as periods spent in Turkey).

The second and fourth generation Conventions were more limited in scope and departed more from the EU Regulations in containing, for instance, longer posting rules. Although longer periods may be explained by the fact that the distance between the countries is considerable, there is no coordination with Regulation 1231/2011. It would be better if bilateral agreements were to take the EU instruments into account and the EU instruments were to allow deviations from them.

The third generation Conventions introduced supervision rules. It is remarkable that EU instruments do not contain such rules. As can be seen in the new EU proposals, such as the proposal to replace Decision 3/80, the plan is to insert these provisions.

Bilateral agreements have not made EU instruments superfluous since, in multi-country situations, bilateral agreements cannot serve as a panacea for all problems. Therefore, the new proposal for Decision 3/80 is to be welcomed. This proposal comes with comparable proposals for Albania, Montenegro, San Marino, Algeria, Morocco, Tunisia, Macedonia, and Israel, ³³ and contain elements of the newer bilateral agreements, such as reciprocity requirements and also supervision rules. ³⁴ However, the proposals do not contain any rule on the determination of the applicable law or on the aggregation of periods of insurance.

There has thus been a fruitful mutual influence between the EU instruments and the bilateral agreement, but there are additional issues where the modernisation and adjustment of the EU instrument can learn from bilateral agreements. Nevertheless, it is important to keep in mind that there may be important political obstacles to fully realising this, since it also means deviating from some more beneficial provisions of the past.