Dutch retirement migration to Spain and Turkey: Seeking access to healthcare across borders

Abstract This article focuses on Dutch retirement migrants who move to Spain and Turkey after retirement. Retirement migrants move at a stage in their life cycle which can be associated with health deterioration. The need to seek access to healthcare provisions may therefore be important in the migratory experience of retirement migrants. This article provides an analysis at three different and interrelated levels by drawing on an analytical framework of Faist, Bilecen, Barglowski and Sienkiewicz. The article discusses the interrelationship between: (1) European, national, and private rules and regulations on health care which determine retirement migrants’ access to healthcare provisions in the home and host state; (2) retirement migrants’ social networks as a space where collective meaning is given to these rules and regulations and where retirement migrants create preferences for healthcare provisions; and (3) retirement migrants’ strategies to access their preferred set of healthcare provisions in the home and host state. It will be argued that retirement migrants may navigate their way through the healthcare systems and may change their mobility or residence pattern in order to be seen as a resident in the Netherlands or Spain/Turkey in order to access their preferred set of healthcare provisions.


Introduction
Northern European retirement migrants tend to portray themselves and are often portrayed in the literature on retirement migration as active, healthy, and well-off migrants for whom migration to the Mediterranean is a way to pursue the "good life" (see for example Benson & O'Reilly, 2009;King, Warnes, & Williams, 2000). Likewise, retirement migrants are often depicted as "amenity seekers" (Haas & Serow, 2002;Williams, King, Warnes, & Patterson, 2000) or "lifestyle" migrants (Benson & O'Reilly, 2009). Generally, studies on retirement migration focus on the first move after retirement which is often undertaken within the "third age" during which new social and recreational activities can be pursued (Hall & Hardill, 2014). As the aging process progresses however, retirement migrants approach the "fourth age" which can be associated with health deterioration, decrease of choice and autonomy, and less activity (Gilleard & Higgs, 2010). Therefore, retirement migrants may want to secure access to healthcare provisions across borders.
Studies on retirement migration acknowledge the impact of retirement migrants' stage in the life cycle and the need to seek access to (health)care provisions (Bahar, Laciner, Bal, & Özcan, 2009;Benson & O'Reilly, 2009;Gustafson, 2008), yet an in-depth analysis of how the need to seek access to healthcare provisions influences retirement migrants' migratory trajectories is often lacking. The studies of Ackers (2007, 2009) and Dwyer (2002, 2004) are exceptions and provide useful insights by showing the interrelationship between formal (health)care rights and the different ways in which retirement migrants negotiate their European and national rights. Ackers and Dwyer (2004) have shown that legal rules create static and fixed categories which determine retirement migrants' access to welfare provisions. These fixed rules are at odds with the fluid and changeable lives of retirement migrants. By drawing on the experiences of Dutch retirement migrants who move, permanently or temporarily, to Spain and Turkey after retirement, this article builds on these insights and aims to further analyze the links between rules and regulations on healthcare and retirement migrants' choices and strategies to secure access to their preferred healthcare provisions.
By drawing on an analytical framework of Faist, Bilecen, Barglowski, and Sienkiewicz (2015), this article provides an analysis at three different and interrelated levels. It discusses, firstly, the European, national, and private rules and regulations which determine retirement migrants' rights to healthcare provisions in the home and host state. Secondly, the article examines how, within retirement migrants' social networks, collective meaning is given to these rules and regulations and how preferences for certain healthcare provisions are exchanged. Thirdly, the different strategies which retirement migrants may apply in order to access their preferred set of healthcare provisions will be discussed. It will be argued that retirement migrants may navigate their way through the healthcare systems and may change their mobility or residence pattern in order to be seen as a resident in the Netherlands or Spain/Turkey. Before turning to this analysis, the next sections describe the research approach and the analytical framework.

Outline of the research approach
The research is based on a comparison between Dutch pensioners who move to Spain and Dutch pensioners who move to Turkey after retirement. These two groups are taken into account, because the migratory choices of both groups of retirement migrants are influenced by (health)care considerations. Furthermore, the mobility patterns of both groups of Dutch retirement migrants vary from permanent migration to seasonal or pendulum migration (O'Reilly, 2000). Some Dutch retirement migrants decide to be registered as residents in the country of retirement, whereas others prefer to retain their residence in the country where they spent their working life in order to access a range of familial and welfare resources (Warnes & Williams, 2006).
Comparing Dutch retirement migrants who move to Spain with Dutch retirement migrants who move to Turkey provides insight in the different meaning-making patterns which take place within social networks and the influence of these patterns on retirement migrants' choices and strategies on a micro-individual level. Furthermore, the comparison also sheds light on the differences between movement of retirement migrants within the EU and across the EU borders. Mobility within the EU is facilitated by EU's free movement and social security coordination rules and legislation. Migrants who move to Turkey are not covered by these supra-state rules. This raises the question whether migrants who move to Turkey are facing limitations when in need of healthcare.
The article includes three levels of analysis: the relevant legal reality, retirement migrants' social networks, and the micro-individual strategies of retirement migrants. The data for the last two levels is collected through qualitative fieldwork in Spain and Turkey. The data for the first level -the relevant legal reality -is based on desk research by studying legal documents and national healthcare reports. What is considered to be the relevant legal reality is inspired by the lived reality of retirement migrants. It concerns the rules and regulations on healthcare and registration with which retirement migrants have to deal in their daily lives.
The empirical data for this research is gathered through fieldwork in Spain and Turkey which took place between January and December 2012. A total of 45 in-depth interviews were conducted with Dutch retirement migrants in both countries. All interviewees were retired in the sense that they had given up paid work and spent at least six months per year in Spain or Turkey. Additional data was collected through follow-up interviews between 2013 and 2016 and through interviews with key officials. The interview data was supplemented with participant observation during social gatherings of Dutch organizations and while being present at Dutch-owned bars and restaurants. Conducting participant observation provided further insight into the meaning-making processes within the social networks of Dutch retirement migrants.
The interviews were semi-structured and followed a life-history approach in which the retirement migrants were asked to talk about mobility during their working life, the decision-making process leading to (temporary) movement to Spain or Turkey, the factors which influenced their preference for a certain residence and mobility pattern, and their future expectations. Structuring the interviews in this way enabled me to learn when the law emerges in the daily lives of retirement migrants and how migrants deal with the law. As Sarat and Kearns (1993), (p. 60) argue, studies which focus on law in everyday life should be intensive rather than extensive, as these studies must be open to the multiple ways in which the law is present in the lives of the interviewees. The interviews were analyzed by using Atlas.ti and coded according to the three levels of analysis.
The fieldwork in Spain was conducted on the Costa Blanca, the coastline of the Alicante province. In this region, a total of 24 interviews were conducted with Dutch retirees: 19 interviews with couples, 2 with men and 3 with women. Alicante has the largest share of foreign residents of all Spanish provinces. Dutch retirement migrants are the third largest group of Northern European retirement migrants in the province of Alicante. Currently, 7473 Dutch citizens aged 60 and over are registered as residents in the province. This figure does not include the high number of seasonal migrants and other retirement migrants who have retained their residency in their home country (King et al., 2000;O'Reilly, 2000;Rodríguez, Lardiés, & Rodríguez, 2010).
A total of 21 interviews were conducted with Dutch retirees in the provinces of Aydın (Kayseri) and Antalya (Alanya and Mahmutlar) in Turkey: 14 interviews with couples, 4 with men and 3 with women. Due to a lack of published data, the number of Dutch retirement migrants in Turkey is even more difficult to establish, but it is definitely lower than the number in Spain. Data on residence permits provided by local authorities and the foreigners police shows that in 2012, 576 Dutch citizens were living in Alanya and 412 were living in Mahmutlar. Böcker and Balkır (2012) reported that in 2000, 2607 foreigners, of whom 349 were Dutch, were living in the Aydın province.

Linking the different levels of analysis -a conceptual frame
Studies which acknowledge the risks related to retirement migrants' stage in the life cycle show that legal rules on (health)care influence retirement migrants' decisions for a certain mobility and residence pattern (Coldron & Ackers, 2007), yet what is often lacking is an analytical framework in which the choices and strategies of individuals can be placed. This study moves beyond a one-dimensional analysis of how retirement migrants respond to legal rules by focusing on the interrelationship of legal structures, meaning-making processes, and retirement migrants' agency. This study draws loosely on an analytical cadre developed by Faist, Bilecen, Bargloswki, and Sienkiewicz (2015) to study cross-border social protection. This cadre consists of three main elements which are interconnected and interacting: (1) the relevant reality, (2) activities and strategies, and (3) meanings, interpretations, and definitions.
The "relevant reality" consists of (macro-, meso-, and micro-) aspects which "all function as enabling and restricting opportunities for social agents and for their activities" (Faist et al., 2015;p. 197). It includes the rules and regulations which restrict or enable mobility across borders as well as meso-and micro-level social protection arrangements. The focus in this study is on the rules and regulations which form the context that exists before retirement migrants act (O'Reilly, 2012). Rules and regulations at different (private, national, and supranational) levels may facilitate or impede retirement migrants' mobility and access to different types of healthcare provisions in their "host" and "home" country. How retirement migrants deal with these rules and regulations in practice is determined by their heterogeneities -e.g. migrants' (older) age, health status, legal status, socio-economic status, gender and ethnicity. These different heterogeneities may lead to social inequalties between individuals (Faist et al., 2015).
The "activities and strategies" concern the ways in which migrants cope with risks across borders, according to Faist et al. (2015). Migrants' activities and strategies can be divided into routine and reflexive (thoughtful, purposive, strategic) actions (O'Reilly, 2012). This article focuses on retirement migrants' actions and strategies to cope with the risks associated with deteriorating health and to obtain or retain access to (specific types of) healthcare provisions. These actions and strategies are generally related to migrants' choices for a certain mobility and residence pattern.
The third element consists of the meaning-making patterns which agents create to give meaning to the first two elements (Faist et al., 2015). These meaning-making patterns are created within migrants' transnational social networks and can be seen as social constructions of both activities and structure on the part of the migrants. The retirement migrants move to areas where strong social networks of Dutch retirement migrants already exist. Within these networks, meaning is given to healthcare provisions in the home and host state and possible strategies and actions are exchanged to access a preferred set of healthcare provisions. As will be discussed, migrants' activities and strategies are closely related to these meaning-making patterns within the social networks. Unlike Faist et al. (2015), I have interpreted the meaning-making patterns in this study as mediating factors influencing retirement migrants' activities as will be shown in the next sections.
The following sections will discuss these three different elements in relation to Dutch retirement migrants' move to Spain and Turkey. After discussing the legal framework which determines retirement migrants' rights to access healthcare in Spain and Turkey, the article continues with a description of the meaning-making processes. The meaning-making patterns are important to understand in which context retirement migrants make choices and how retirement migrants' preferences for certain healthcare provisions are created, the choices and strategies will therefore be discussed lastly.

The relevant (legal) reality
In order to understand retirement migrants' preferences for certain healthcare provisions as well as their choices and strategies for a specific mobility and residence pattern, I will firstly discuss the relevant (legal) reality which exists before the retirement migrants act. The relevant legal reality relates firstly to the differences between the healthcare (insurance) systems in the home and the host state and, secondly, to national and EU rules and regulations on registration which determine retirement migrants' entitlement to specific healthcare provisions in the Netherlands and Spain/Turkey. Both aspects of the relevant (legal) reality are interlinked and will be discussed successively.

Differences between healthcare (insurance) systems
The formation and consolidation of current healthcare systems took place within the territorial borders of the nation-state. The healthcare systems of the different states remain nationally organized and closely linked to the territory of these nation-states (Ferrera, 2005;Martinsen, 2005). Within the EU, access to the different healthcare systems is coordinated in Regulation 883/2004 on the coordination of social security systems and Regulation 987/2009 on the implementation of the regulation. However, this does not change the fact that "rights and entitlements vis à vis public healthcare are very much determined by national laws and the systems operating in individual member states" (Dwyer, 2001, p. 315).
The main differences between the Dutch, Spanish, and Turkish healthcare systems are related to the division between public and private healthcare provisions and the quality of insurance packages to which the public healthcare insurance companies provide access. Until 2006, the Dutch healthcare insurance system consisted of a separate public and private system. This distinction was abolished after the introduction of the new Health Insurance Act in 2006. From then on all inhabitants of the Netherlands had to take out compulsory basic private healthcare insurance consisting of a universal healthcare package. This basic healthcare insurance can be supplemented with voluntary additional insurances (Schäfer et al., 2010). The Spanish healthcare system consists of universal healthcare insurance for all residents in Spain and a private healthcare system for which individuals have to obtain additional insurance. The public and private system have their own separate public and private healthcare providers (García-Armesto, Abadía-Taira, Durán, Hernández-Quevedo, & Bernal-Delgado, 2010). Turkey has a centralized universal health insurance scheme which came into force in 2008. It consists of a two tier-healthcare system of both public and private provisions. The healthcare insurance provides access to both types of provisions, but with higher co-payments for private provisions (Tatar et al., 2011;Tello & Baez-Camargo, 2015). How the differences between the healthcare (insurance) systems impact retirement migrants' lives and which (public or private) provisions retirement migrants have access to in the host country depends on the mobility patterns of the pensioners. Retirement migrants who stay on a temporary basis in Spain or Turkey can generally rely on their Dutch healthcare and/or travel insurance and can claim their (unforeseen) healthcare costs in the Netherlands. These pensioners can access both public and private healthcare provisions in the host country (Balkır & Böcker, 2015). Pensioners who travel within the European Union, e.g. Spain, can apply for a European Health Insurance Card (EHIC). This card gives them access to the healthcare system of another Member State during temporary stays and for unforeseen medical treatment. Pensioners who travel to Turkey can get additional insurance and/or travel insurance in order to cover unforeseen healthcare expenses abroad.
Since the introduction of the Dutch Healthcare Insurance Act in 2006, retirement migrants who reside abroad permanently and who receive a Dutch state old-age pension are obliged to register with and pay a contribution to the Dutch Health Insurance Board. 1 The amount of the contribution depends on the cost of the healthcare package provided by the national social security system of the country of residence. The contribution is deducted from the retiree's pension and paid to the social security board in the country of residence. This gives retirement migrants the right to be insured with the social security board in the country of residence -the Seguridad Social in Spain and the Sosyal Güvenlik Kurumu (SGK) in Turkey. This right does not infer a right to the same standard or level of entitlement as previously enjoyed in the country of origin, but entitles the individual concerned to the same services as normally enjoyed by a national of the host state. Retirement migrants who are, based on this system, insured in the national healthcare system in Spain and Turkey have no or limited access to private healthcare provisions in their country of residence. In order to access private provisions, these retirement migrants have to apply for additional private healthcare insurance in Spain and Turkey. 2

Determining where retirement migrants are supposed to register
The rules which determine retirement migrants' access to the host country's healthcare system are, at first sight, straightforward. Temporary migrants are expected to remain registered as residents in the Netherlands and claim their healthcare costs with their Dutch healthcare and/or travel insurance. Permanent migrants are expected to deregister as residents in the Netherlands, register in the host country and pay a contribution to the Dutch healthcare insurance board in order to be insured in the host country. However, a closer look at the rules and regulations reveals that the European Union, Spain, Turkey, and the Netherlands draw different boundaries between permanent residence and temporary stay. Even within national and EU legislation, different boundaries are drawn. Some of the relevant boundaries are further explained, since the diversity of the rules and regulations may create confusion among retirement migrants. It will appear that the fixed rules to which Ackers and Dwyer (2004) refer are not that fixed because of the diversity in the rules and regulations.
Retirees who move to Spain fall under the EU free movement provisions. The requirements for a stay abroad are laid down in Directive 2004/38 on the right of citizens of the Union and their family members to move and reside freely within the territory of the EU Member States. Pensioners who stay in Spain for less than three months are considered to be tourists or visitors and their stay is consequently not constrained by any formalities. If economically inactive citizens, such as pensioners, stay for more than three months in another Member State, they need to have sufficient resources for themselves and their family members so as not to become a burden on the social security system of the host Member State, and they need to have a comprehensive sickness insurance cover in the host State. 3 As Jorens, Roberts, de Cortazar, Spiegel, and Strban (2010) point out, there is an inconsistency between Directive 2004/38 on free movement and regulation 883/2004 on the general principles regarding the coordination of social security of persons moving within the EU. What in the directive may be considered "residence, " for example living for more than three months in the host Member State, may still be considered a temporary "stay" under regulation 883/2004, because his/her "center of interest" may be located somewhere else. 4 It is thus possible that a person is simultaneously considered to be a resident in Spain under Directive 2004/38, while the same person is referred to as a temporary visitor under Regulation 883/2004. National rules of the host country are also relevant for migrants moving to Spain and Turkey. The Netherlands requires its residents to deregister from the municipal registration register when they intend to stay abroad for more than 8 months within a 12-month period. Dutch private healthcare and travel insurance companies also draw different, and less transparent, boundaries. Healthcare insurances and travel insurance companies can impose restrictions to the amount of healthcare costs temporary visitors can claim abroad and consequently to the number of months one can stay abroad. 5 The Spanish authorities require EU citizens to register with the local authorities, the Municipal padrón, after three months. This is in line with the free movement provisions laid down in Directive 2004/38. Dutch residents in Spain who have signed on to the padrón are however not automatically also tax-paying residents, meaning that this person has officially deregistered as a resident in the Netherlands and that Spain is his or her fiscal domicile. 6 In practice, only tax-paying residents are required to fulfill the residence requirements laid down in Directive 2004/38 concerning comprehensive sickness insurance cover and sufficient resources.
Retirees who move to Turkey do not have free movement rights. Turkey's Law on Foreigners and International Protection No. 6458, Article 11(1), states that foreigners wishing to stay in Turkey for up to 90 days shall obtain a visa that indicates the purpose of their visit from a Turkish consulate. The period of stay in Turkey provided by the visa or visa exemption cannot exceed 90 days within a period of 180 days. Individuals staying longer than 180 days are required to apply for a residence permit (Ikamet). Retirement migrants with an Ikamet can apply for coverage under the Turkish national health insurance scheme, but they are not obliged to do so.
Rules and regulations on healthcare and registration are generally based on the idea that someone's residence is in one state. This fixed assumption does however not fit well with the often flexible migratory patterns of retirement migrants (see also Ackers & Dwyer, 2004). The diversity of rules and regulations on healthcare and registration may on the one hand cause confusion and insecurity among retirement migrants and on the other hand provide room for agency, as will be discussed in the following sections.

Meaning-making processes
Within the social networks of retirement migrants, meaning is given to the rules and regulations and to healthcare provisions in both countries. These meaning-making processes take place within the transnational social space in which retirement migrants move. National states can be: "foils of reference upon which to project ideas, norms, values, etc. The 'other' national system can serve to sharpen comments on, and evaluations and understandings of specific national contexts" (Faist et al., 2015, p. 199).
Within the same transnational social space, people cross borders together with their norms, values, and more generally their frame of reference. These transnational social spaces thus become transnational spaces of comparison (Faist et al., 2015). Dutch retirement migrants actively compare Dutch and Spanish/Turkish legal realities and the healthcare provisions in those countries. Within the transnational networks of Dutch retirement migrants, ideas and perceptions are intensely exchanged. While exchanging perceptions, the relevant legal reality in Spain and Turkey is interpreted and valued.

Meaning-making processes in Spain
Because of the high number of Northern European retirement migrants on the Costa Blanca, the streets in the towns are dominated by shops, insurance offices, cafes, and restaurants specifically catering to the foreign population. In practically every town one can find one or more Dutch (or Dutch speaking) health specialists. Specifically, the private healthcare sector with provisions for (aging) foreigners is prepared for receiving high numbers of foreigners. At private hospitals, Dutch migrants can receive medical treatment in their mother tongue, either by Dutch staff or with the help of translation services. Translation services are generally also available at public hospitals, but when Dutch migrants need to go to a public General Practitioner, they have to arrange a translator themselves (Gehring, 2013).
Social clubs play an important role in the transnational life of retirement migrants on the Costa Blanca. Social clubs often focus exclusively on one nationality or one language group, and the networks of Northern European retirement migrants are formed along these language lines. At Dutch clubs, but also during informal gatherings, at internet forums, and in local Dutch newspapers, legal information is exchanged and reproduced. Dutch clubs often invite legal specialists and doctors to give informational presentations in order to explain rules and regulations on healthcare. In this way, one can speak of the occurrence of a transnational semi-autonomous social field (SASF) consisting of Dutch pensioners living on the Costa Blanca (Moore, 1973). Within this transnational SASF, Dutch retirement migrants give meaning to and reinterpret rules and regulations related to the relevant legal reality. Due to the intense exchange of legal information, retirement migrants in Spain are generally well-informed about rules and regulations related to healthcare and the different ways to deal with the law.
Social clubs can also be seen as a supporting network when one of its members is in need of (health)care. Club members help each other by providing practical support in the form of translation support, travel support, and aftercare. Since their children generally live in the home country, the care is provided by the other members of the social clubs in Spain.
Clubs are the place where retirement migrants exchange their opinions about healthcare services. As is shown in the following quotation, retirement migrants generally prefer to visit private healthcare providers, since the waiting lines at and communication problems with public healthcare providers are often perceived as problematic.
"I have heard horrible stories from other Club members of [name of the club]. They were not treated in the public hospital even though they were severely ill. One had a stroke, but he didn't speak Spanish and had to wait way too long. You hear these stories a lot here. " (Koos, semi-permanent migrant in Spain). 7 As we will see in the next section, the negative evaluation of Spanish public provisions is directly linked to migrants' strategies to obtain access to private healthcare provisions in Spain or to retain access to Dutch healthcare provisions.
Not only the quality of the healthcare provisions, but also registration rules are discussed among Dutch pensioners in Spain. In their accounts, retirement migrants make a distinction between registration with the padrón and becoming a tax resident: "I think that it is not allowed what I do, because when you stay longer than six months per year [in Spain] you're supposed to be a resident here. [...] I did register with the padrón, that is important for the municipality I heard. They get money for that. These two different rules are very confusing […]. There are people who become a resident here because of financial considerations. There are people who look at that, but I look at my health. If I become ill [I want to go back to the Netherlands] that's why I don't become a resident. " (Sjaan, semi-permanent migrant in Spain).
Like Sjaan, many retirement migrants express the opinion that one should register with the padrón, but that one may become a tax resident. Only the retirement migrants who have become tax residents have to deregister as a resident in the Netherlands. As Sjaan expresses, retirement migrants who may refuse to do this prefer to retain access to Dutch healthcare provisions. Sjaan's statement about registration with the padrón reflects the narrative of the local municipalities: "It is my task to make sure that people register with the local authorities. That is because many people don't do it. That has direct consequences for the policy of the municipalities. Everything is based on the number of inhabitants. Police officers, doctors, government provisions. We try to persuade people to register in a number of ways and we try to show them that it also better for them.

Meaning-making processes in Turkey
Studies on retirement migration to Turkey indicate that there are many similarities between retirement migration to Turkey and migration to more traditional retirement migration destinations such as Spain (Bahar et al., 2009;Balkır & Kırkulak, 2009;Bayır & Shah, 2012;Böcker & Balkır, 2012;Nudrali & O'Reilly, 2009). However, differences can be observed between the density of the networks of Dutch migrants in both countries and the healthcare services which are set up for foreigners.
The fieldwork in Turkey was conducted in the Aydın and Antalya province. In both provinces there were several initiatives to create Dutch clubs. However, tight networks, comparable to Spain, do not exist among Dutch retirees in Turkey. The Dutch retirement migrants in Turkey are organized in looser and more informal ways. Gatherings of Dutch retirement migrants mainly take place in Dutch-owned bars and restaurants, at the Dutch church, and during hikes organized by Dutch retirement migrants. During these gatherings, information about legal rules is exchanged and reformulated, but it happens at a different scale and is less formalized than in Spain. There are no officials invited to explain the rules and regulations, and local newspapers pay less attention to legal aspects. Within the transnational SASF of Dutch retirement migrants in Turkey, rules and regulations about accessing healthcare and registration are mainly exchanged on an informal basis. These loose networks are seen as supportive by the Dutch retirement migrants when it comes to providing informal (after) care and translation services.
Turkey is a relatively new retirement country for Northern European retirees, and assuming that retirement migrants move abroad within their third age, retirement migrants have generally not reached the fourth age yet. 8 This leads to a situation in which Turkey's health and care provisions for foreigners are less developed than in Spain. The system mainly lacks a network of long-term care institutions for foreign elderly people. Healthcare services with translation services are however widely available in the popular retirement areas in Turkey. The general opinion of the quality of the, specifically private, healthcare provisions is positive, as is shown in the following quotation: "I have the impression that the quality in general is really good. The healthcare is very good. The bureaucratically system is however horrible. You have, but you probably know that yourself, two kinds of hospitals here. That is the system. The SGK has their own hospitals and you have the private clinics, of which Anadolu is an example. That is a very good hospital.
[…] Most of the time someone joins from the foreigners department of the Hospital to function as an interpreter. That can either be translation into Dutch or English, that doesn't matter to me. " (Frank, permanent migrant in Turkey).
Practically all interviewed retirement migrants had remained registered as residents in the Netherlands and could thus purchase private healthcare in Turkey. "If you see what problems we face with our house than I don't want to bear the burden of being registered here [as a full resident]. The [private] hospitals here are also corrupt. Extremely good, but they rip you off. " (Gerda, semi-permanent migrant in Turkey).
Although the Dutch retirement migrants acknowledge the quality of the healthcare services in Turkey, they generally decided to retain their residence in the Netherlands because of a combination of legal problems related to housing and a general suspicion of Turkish bureaucracy. The reasons for non-registration are, unlike in Spain, often less directly linked to accessing healthcare, but more to a general distrust in the Turkish bureaucratic system, as Gerda expresses.

Retirement migrants' healthcare strategies
In the remaining part of this article, further attention is paid to retirement migrants' choices and strategies to access their preferred set of healthcare resources when dealing with health-related problems. As already discussed, retirement migrants generally prefer to access private healthcare provisions in Spain or Turkey or retain access to Dutch healthcare provisions.
Temporary migrants generally have the most direct access to both Dutch and Spanish or Turkish (private) healthcare provisions, because they can access both systems with their Dutch healthcare insurance. Retirement migrants who permanently move to Spain or Turkey often try to purchase private healthcare insurance in the host country in order to access the same private provisions as temporary migrants. Retirement migrants who do not have direct access to their preferred healthcare provisions may adjust their mobility pattern in order to access their preferred set of healthcare provision. They may do this in three different ways: (1) they may choose a specific mobility pattern; (2) they may choose a country of residence; and (3) they may hold on to their mobility and registration pattern but try to change the rules and regulations. These three categories will subsequently be discussed.
Rules and regulations do not affect each retirement migrant in the same way. The capacity to achieve their preferences is mediated by retirement migrants' "heterogeneities, " such as age, health status, legal status, socio-economic status, gender, and ethnicity (Faist et al., 2015). Furthermore, wishes, preferences, and needs of retirement migrants change when they approach the fourth age and considerations related to retirement migrants' stage in the life cycle may influence their preferences to access specific healthcare provisions (Ackers & Dwyer, 2004).

Choosing a mobility pattern
Health risks and a more general feeling of vulnerability due to their stage in the life cycle may force migrants to choose a specific mobility pattern. The following example of Henk, a Dutch retirement migrant in Spain, shows how his perception of Spanish healthcare provisions and his health status influences his and his wife's choices in retirement.
Henk and his wife moved to Spain in 2003. After living there for three years, the Dutch healthcare insurance system changed and they were confronted with the fact that they could no longer maintain their private Dutch healthcare insurance. Consequently, Henk's direct access to his Dutch doctor in Spain was restricted. Because of Henk's health problems, it was difficult and expensive for the couple to get additional private healthcare insurance in Spain. Being insecure about his new situation and wishing to retain access to his private doctor, Henk and his wife decided to return to the Netherlands and to become seasonal migrants in Spain. They are now insured in the Netherlands. When in Spain, they continue visiting their private doctor and declare the costs of the hospital visit in the Netherlands by using their EHIC.
The example of Henk shows how rules and regulations interfere in the daily lives of the retirement migrants as well as migrants' agency to deal with this situation. Henk's unstable health situation made him vulnerable to the change of the Dutch healthcare system in 2006 and he therefore felt forced to change his mobility pattern. Henk's illness would have been dealt with in the Spanish public healthcare system, yet the general distrust in these provisions within their Dutch social network made the couple decide to return to the Netherlands and to move to Spain as temporary migrants.
Instead of completely changing their mobility pattern, retirement migrants moving to Spain and Turkey often slightly adjust their mobility pattern, as they believe that they, as temporary migrants, can maintain their Dutch healthcare and travel insurance. The following example of Marie and John shows how they take private rules and regulations of insurance companies into account while deciding on their mobility pattern: Marie: "A couple of years ago I had a problem with my heart. They [the doctors] told me I need heart catheterization and then you get a Dutch girl [from the travel insurance] on the phone asking: 'Did you go with preconceived plans to Spain, because there is a waiting list […] in the Netherlands?' She did not give her permission for the operation. I told her that they could also get a dead body back if they wait too long.
[…] In the end the doctor in Spain did not give permission to travel. I had my treatment here and the travel insurance covered the costs. " John: "There are also things which you don't know at first. We were here and then you hear that people go home after six months, that they cannot stay away more than six months because of the travel insurance or health insurance. " Marie: "It is because the insurance is misleading. If you have a continuous travel insurance, it is often not continuous. You cannot stay away more than six months. If you're back only for one day, they start counting the days again. " (Marie and John, temporary migrants in Spain).
As temporary migrants, Marie and John state that they can access healthcare in both countries -the home and the host state -while either falling back on their health insurance or travel insurance. They adapt their mobility pattern to the requirements of their travel insurance. As Marie and John express, the advantages and disadvantages of different mobility patterns are discussed among the Dutch retirement migrants in Spain. Thinking about mobility patterns is perceived as important, specifically because healthcare and travel insurance companies tend to control declarations of the insured abroad. Insurance companies can, in this way, impact Dutch retirement migrants' mobility patterns, and the companies actively participate in the process of determining where, in which country, retirement migrants are "allowed" to reside. Both examples show the interrelationship between access to healthcare and retirement migrants' mobility patterns. Rules and regulations of private insurance companies are generally perceived by the retirement migrants as more influential on their choices than national or EU legislation on registration. The rules and regulations imposed by states and the EU on registration procedures are often either unknown or perceived as malleable (O'Reilly, 2012) by the Dutch retirement migrants, probably because of a lack of enforcement in practice.

Choosing a residence pattern
Instead of choosing a specific mobility pattern, retirement migrants may also choose a specific residence pattern by maintaining their habitual residence in the Netherlands, while spending most of the year in Spain or Turkey. In this way, retirement migrants retain access to Dutch healthcare provisions while also accessing private healthcare provisions in Spain or Turkey (officially only when it concerns unforeseen treatments). Retirement migrants are generally aware that they should register as a resident in Spain or Turkey, but they argue that non-registration leads to the most favorable healthcare deal. Retirement migrants may consciously choose this option in order to get the best of both worlds, which Ackers and Dwyer (2004) refer to as "cherry picking, " yet retirement migrants may also choose this option because of an unstable health situation. Dutch retirement migrants living in Spain and Turkey who are in need of care, like Michiel and Henny, often refer to their Dutch healthcare insurance as a safety net. Although Michiel and Henny should, officially, deregister as residents in the Netherlands, they have decided to retain their residence in the Netherlands because of Michiel's health status and the fear that the SGK would not cover cancer-related problems. Michiel and Henny are in a more vulnerable position than that of retirement migrants moving within the EU, for example to Spain. The healthcare costs of retirement migrants moving to Spain would also not be covered by private insurances, yet these migrants cannot be excluded from full coverage within the Spanish public healthcare system due to EU coordination regulations and the principle of non-discrimination (see also Dwyer, 2001).
The following example of Nelleke further shows how retirement migrants, specifically in Turkey, may struggle with the decision where to register. Nelleke explains that she went to Turkey together with her second husband. After moving to Turkey, the couple decided to settle there permanently. Not only in practice, but also "on paper": they deregistered in the Netherlands and registered in Turkey. They did so mainly for fiscal reasons. After some years, Nelleke found out that she would not be insured for medical costs when visiting the Netherlands. She could not afford Dutch travel insurance to cover her healthcare costs in the Netherlands but she is unable to obtain Turkish permanent travel insurance to cover medical costs outside Turkey. Nelleke's health is rather unstable and she therefore fears to visit the Netherlands. Nelleke argues that her mobility was impeded by these rules and decided to reregister in the Netherlands and re-obtain Dutch healthcare insurance, because: "I want to see my children and grandchildren and if I can't go to the Netherlands, than that's a problem for me. " Nelleke's husband does not have much contact with his children and he prefers to remain registered as a resident in Turkey. Nelleke continues her story to tell that, at the time of the interview, she is thinking about "moving" back to Turkey again (reregistering in Turkey), because her husband is severely ill and she just broke her hip. Her Dutch healthcare insurance allowed her to get treatment in Turkey, but they told her that she will be repatriated to the Netherlands the next time she needs to be hospitalized, which is against Nelleke's wishes.
This example shows again that healthcare insurance companies may influence retirement migrants' mobility and residence patterns and may interfere in retirement migrants' daily lives, specifically when migrants do not have sufficient financial resources to buy private healthcare insurance or when their health is unstable. Nelleke feels trapped between the different healthcare systems. If Nelleke had moved to an EU Member State, she would have had the possibility to receive treatment in the Netherlands while being a habitual resident in Spain. Retirement migrants who are resident in another EU Member State can apply for an EHIC at the Seguridad Social which provides them direct access to all healthcare provisions in the home country. 9 Migrants who fall outside the scope of Regulation 883/2004 by moving outside the EU do not have this possibility.

Going to court -trying to change the rules and regulations
A third strategy applied by Dutch retirement migrants living in Spain was to retain their mobility and residence patterns, but to try to change the rules and regulations instead. After the Dutch Healthcare Insurance Act came into force in 2006, retirement migrants living permanently in Spain could no longer maintain their Dutch healthcare insurance. These migrants suddenly fell under the special regime contained in Regulation 1408/71 (now regulation 883/2004) for sickness benefits for pensioners. More concretely, the pensioners were required to pay contributions in the Netherlands and were entitled to healthcare in the state of their residence in accordance with the legislation applicable there. Their previous private health insurance contracts were terminated in as far as coverage was taken over by the legislation in the state of residence. In order to fight this change, retirement migrants in Spain set up a foundation and joined forces with pensioners living in Portugal, France, Belgium, and Germany. They started a lobbying campaign and initiated a series of legal proceedings. 10 These retirement migrants made use of the tight network of Dutch retirement migrants living permanently in Spain to mobilize money and people to litigate against this change and its negative consequences for the Dutch pensioners living in Spain.
The pensioners achieved the introduction of a country of residence factor (instead of unitary contributions) after a ruling of a local court in summary proceedings. 11 The pensioners continued their lawsuit because their main argument, which stated that the pensioners should have a right to choose whether or not to make use of the healthcare offered under the legislation of their country of residence, was rejected. In subsequent cases this argument was also rejected by other Dutch courts and by the CJEU in the van Delft ruling 12 as well as by the ECtHR 13 . For an analysis of the court rulings, see the work of van der Mei (2013).
Although the courts dismissed the claims of the retirement migrants, the example does show that the networks of retirement migrants in Spain are strong and that the migrants have the capacity to fight state rules, which are generally seen as non-malleable. In Turkey, a comparable network was not formed, and similar collective forms of protest were not mentioned by the Turkish respondents. This can probably partly be explained by the loose network of Dutch retirement migrants in Turkey.
Retirement migrants generally decide more often to change their mobility or residence patterns than to fight rules and regulations in court, since going to court may take many years and retirement migrants are aging during this process. One of the migrants involved in the foundation stated: "We are old and it will take years before we hear something. We are soon dead and then the problem is solved as well. "

Conclusion
Retirement migrants move at a stage in their life cycle which can be associated with health deterioration. The need to seek access to healthcare provisions is therefore important in the migratory experience of retirement migrants. In this article, an interrelationship has been assumed between rules and regulations at a macro-level, meaning-making processes at a meso-level, and migrants' strategies at a micro-individual level. This conclusion recapitulates how these levels are interlinked.
When seeking access to healthcare, retirement migrants have to deal with rules and regulations which determine in which country they can access healthcare provisions. Different sets of rules and regulations on a state, bilateral, and supra-state level, but also on a private level, determine that retirement migrants can only have one habitual residence in either the home or the host state. This conflicts with the fluid lives of the retirement migrants and their dual place attachment. Retirement migrants may actively deal with this categorization in three different ways. They may choose a specific mobility pattern, choose a specific residence pattern, or unite as a group and fight the rules and regulations in court. Their actions are generally focused on seeking access to private healthcare provisions in the host country and retaining access to Dutch healthcare provisions.
Retirement migrants' actions are mediated in two different ways. Firstly, their actions are informed by the exchange of information within social networks. Within retirement migrants' transnational social networks, meaning is given to healthcare provisions and information is exchanged about how one could or should deal with these rules and regulations. Furthermore, the social network in Spain is also a place where migrants unite in order to fight rules and regulations. Secondly, retirement migrants' actions are mediated by their heterogeneities. Specifically, age, health status, residence status, citizenship, and socio-economic status appear to be important factors influencing retirement migrants' access to healthcare provisions and their capacity to deal with the law. For example, retirement migrants with more financial resources are more footloose (Ackers & Dwyer, 2004) because they can buy private healthcare insurance in the host country, and retirement migrants who are healthy are not forced to deal with the law.
Private rules and regulations, imposed by Dutch healthcare and travel insurance companies in particular, influence retirement migrants' decisions in practice. The formal rules and regulations on residency imposed by states and the EU appear to be malleable (O'Reilly, 2012) because of a lack of coherence in the laws and/or a lack of enforcement in practice. Rules and regulations of healthcare and travel insurance companies are less malleable, because these companies have the power to determine who can access which provisions and when. These companies appear to act as unexpected gatekeepers who control access to healthcare provisions.This observation should be further researched in order to get a better understanding of the role of private actors in migration decisions of individuals.
When comparing Dutch retirement migrants moving to Spain with those moving to Turkey, two observation can be made. Firstly, it becomes clear that the composition of retirement migrants' social networks in the host country influences the ways retirement migrants assess healthcare provisions and consequently influences retirement migrants' strategies. A strong network of Dutch pensioners in Spain creates a form of agency which enables migrants to fight state decisions and to be less (passively) subjected to state decisions. Secondly, it also shows that when there are no supranational rules to facilitate retirement migrants' access to the host state, migrants may end up in a vacuum when they face serious health challenges, since countries, in this case Turkey, may exclude migrants from full coverage. As a consequence, retirement migrants often choose to maintain their residence in the Netherlands as a safety net.