Linking tourism, retirement migration and social capital

A general trend in the study of international retirement migration has been the increased attention paid to the social contacts and network connections of the migrants in both the destination and the origin areas. These studies have examined the extent to which migrants build social relationships with their neighbours and the host society while also maintaining social links with their countries of origin, addressing the central role that leisure travel plays in sustaining increasingly dispersed social networks and maintaining the social capital of these networks and of the individuals involved in them. Using a case study approach to examine British retirement migration to Spain, we explore the relevance of transnational social networks in the context of international retirement migration, particularly the intensity of bidirectional visiting friends and relatives (VFR) tourism flows and the migrants’ social contacts with friends and/or family back in their home country. Building on the concept of social capital and Putnam's distinction between bonding and bridging social capital, we propose a framework for the analysis of the migrants’ international social networks. The results of a study conducted based on a sample of 365 British retirees living in the coast of Alicante (Spain) show both the strength of the retirees’ international bonding social capital and the role of ‘VFR's travel and communication technologies in sustaining the migrants’ transnational social practices and, ultimately, their international bonding social capital. It also provides evidence for the reinforcing links between tourism-related mobility and amenity-seeking migration in later life.


Introduction
An emerging trend of research on tourism geographies relates to lifestyle migration and the quest for a better life elsewhere (see Benson & O'Reilly, 2009b, for a review). The concept of lifestyle migration encapsulates the implications and varied manifestations of leisure mobility in contemporary societies. Among lifestyle migration, international retirement migration (hereafter IRM) has been one of the most widely researched forms of leisure/lifestyle mobility. In Europe, the study of IRM has traditionally focused on the analysis of flows from northern to southern regions, the nexus between tourism-related mobility and migration in later life, and the impacts of such population flows at the destination level (see Casado-D ıaz, 2012, for a review). Despite the increasing number of studies centred in the analysis of the experiences of IRM migrants in their new destinations (Casado-D ıaz, 2009;Gustafson, 2008Gustafson, , 2009Huber & O'Reilly, 2004;Oliver, 2008;O'Reilly, 2000), the nature and intensity of the social ties maintained by the migrants in their home countries, for both practical and emotional reasons, remain under-researched.
Therefore, the aim of this article is to contribute to our understanding of the nature and meaning of contemporary mobility in the context of IRM. Using the concept of social capital as a theoretical framework, this study based on a sample of British retirees living in the Costa Blanca, Spain, examines the migrants' interpersonal relationships and transnational network connections with their country of origin. For the purpose of this study, and following Portes (1998), social capital is defined as a social resource that is derived from the formal and informal relationships that exist between the members of a community. It is argued that retired migrants possess, to different degrees, both local and international social capital: they bring with them their 'existing' social capital and then create 'new' social capital in their new place of residence (Casado-D ıaz, 2009).
In this article, we focus on the analysis of the international component of social capital. Specifically, we consider two aspects of the migrants' social lives as indicators of their international social capital: the intensity of bidirectional visiting friends and relatives (VFR) tourism flows (i.e. participating in and hosting friends and relatives visits) and the social contacts with friends and/or family living in the UK. Then, we analyse how migrants' international social capital may differ when taking into account different 'explanatory' variables that include sociodemographic characteristics (age and gender), and what we might consider 'bridging'-related variables (knowledge of the local language and length of residence in the local area) and 'bonding'-related variables (ownership of property in the home country).
The article is divided into four sections. The first discusses recent developments in the field of lifestyle migration, and particularly of IRM. The second addresses the usefulness of the concept of social capital as a theoretical framework for the analysis of the migrants' social ties and networks 'back home'. The third and fourth sections report the data, method and results derived from our case study of British retirement migration to Spain based on questionnaire responses from a sample of 365 British retirees living in the Costa Blanca, Spain. The final section presents the main conclusions and discusses the implications and the relevance of the strong social ties that these lifestyle older migrants have maintained with their country of origin.

Tourism and lifestyle migration: searching for the good life
Tourism has significantly influenced the mapping of potential destinations for different forms of leisure-oriented mobility, such as second-home ownership, amenity-seeking seasonal migration or IRM, and has also contributed to the intensification and diversification of contemporary mobility in recent decades (Gustafson, 2002;King et al., 2000;Rodriguez, 2001;Williams & Hall, 2000;Williams, King, Warnes, & Patterson, 2000). Research in this field has recently focused on the concept of lifestyle migration, which considers the implications and varied manifestations of leisure-oriented mobility in contemporary societies. As noted by Benson and O'Reilly (2009a, p. 621), lifestyle migration can be conceptualised as the migration of 'relatively affluent individuals, moving either part-time or full-time, permanently or temporarily, to places which, for various reasons, signify for the migrants something loosely defined as quality of life'. While geographers have focused predominantly on the analysis of the spatial distribution and scale of these flows of migrants, the characteristics of the destinations, the links between tourism and seasonal or permanent migration, and the physical impacts at the local level, sociologists and anthropologists have turned their attention to the trajectories, everyday lives, lifestyle choices and experiences of these older migrants, and explored the implications for host and home communities. The importance of issues of gender, age, place, identity, class, nation and community among these amenity-seeking migrants has also been central to researchers in this field (see Benson & O'Reilly, 2009b).
Within lifestyle migration, IRM has been one of the most widely researched forms of lifestyle mobility. In Europe, amenity-seeking migration in later life has typically originated in northern and central countries (mainly the UK, Germany, Scandinavian countries and the Netherlands) in favour of southern Mediterranean regions and Atlantic island and coastal regions with climatic and lower cost of living advantages. These groups of older people have above-average opportunities for international travel. They have the financial resources, are no longer restricted by employment obligations, and the majority have neither dependent children nor parents in need of daily personal care. These amenity-led migrants are primarily motivated by the desire to engage in a more satisfying way of life and in order to do so they search for new spaces that will provide them with more opportunities to enjoy their post-retirement lives. Many studies reported previous tourist visits as the main connection to the area prior to the move (Casado-D ıaz, Kaiser, Warnes, 2004) and then the purchase of second homes as a stepping stone towards IRM (Hall & M€ uller, 2004;M€ uller, 2002). Indeed, the development of mass tourism in southern Europe widened the opportunities for international leisure travel and provided the necessary services and amenities that made these regions attractive for international tourists, second-home owners and future migrants (Gustafson, 2002;Rodriguez, 2001;. Likewise, the different forms of consumption-led mobility generated significant reciprocal flows of VFR, highlighting the important nexus between tourism and migration (Gustafson, 2008;Ill es & Michalk o, 2008;Williams & Hall, 2000). A series of historical developments and material conditions have enabled growing numbers of lifestyle migrants to participate in these forms of contemporary mobility. These include increased life expectancy, higher incomes and greater affluence, the initial relatively lower cost of living and cost of properties in southern Mediterranean countries, and the familiarity of travelling, working and owning a second home abroad . Likewise, the rapid development of communication and transportation technologies has made it possible for greater numbers of migrants to engage in 'transnational lifestyles' while sustaining links with their home countries (Bozic, 2006;Casado-D ıaz, 2009;Gustafson, 2001;King, Warnes, & Williams, 1998). As argued by Warnes (2009, p. 359), 'the ease and low cost of international travel is allowing many older people to exploit, maintain and continue to develop residential opportunities, social networks and welfare entitlements in more than one country'.
A research framework for the study of international retirement migration Pahl (2000) suggests that the term social capital can be seen as an umbrella for explaining the nature of personal communities, social networks, and other forms and styles of social connectedness in contemporary societies. The role of migrant networks in the formation of social capital has been much discussed in recent migration studies (Anthias, 2007;Castles & Miller, 2003;Ryan, Sales, Tilki & Siara, 2008). Social networks might provide different types and levels of social support and, for migrants, these forms of support very often cross national boundaries, allowing for their transnational links to play a supportive role even after migration (Iorio & Corsale, 2012;Ryan et al., 2008). However, as argued by Anthias (2007), not all social networks and social bonds can be referred to as social capital. It is the ability to 'mobilise' these resources for the pursuit of advantage, or the mitigation of disadvantage, which gives them value as 'capital' and makes the concept of social capital meaningful (Anthias, 2007, p. 788). It is both the personal relationships that individuals accumulate when they interact with each other over time, formally and informally, and the amount and quality of the resources derived from those social relationships which constitute the essence of social capital (Bourdieu, 1986).
Recent research on IRM has explored how retired migrants build social networks in their new communities of residence and the strategies adopted to also retain close ties with their relatives and longstanding friends in their home countries (Casado-D ıaz, 2009;Gustafson, 2008Gustafson, , 2009Huber & O´Reilly, 2004). However, one question that arises in relation to the communities that retired migrants form is the types of social capital generated. As noted by Putnam (2000), different types of networks give rise to different forms of social capital. Particularly relevant to this study is his distinction between bonding social capital and bridging social capital. He argues that bonding types of social capital refer to close social connections between people and are characterised by strong bonds among family members, close friends, neighbours or among members of the same ethnic group. Bonding social capital might also arise within a particular social group bound together by shared identities, interests and place of residence (Healy, 2003). This type of social capital, based on close family and friendship ties, is said to be beneficial to the self-interest of the individual or small groups, and generally good for 'getting by' in life (Putnam, 2000, p. 22). Bridging capital, on the other hand, tends to be inclusive, encompassing people across different social groups and backgrounds, thus encouraging the formation of broader identities and collectivities and, ultimately, social cohesion (Parker & Song, 2006, p. 180). This type of social capital is based on common interest rather than personal closeness or common identification. If bonding capital is good for 'getting by', bridging types of social capital are supposed to be crucial for helping individuals to 'get ahead' in life (Putnam, 2000, p. 19).
It is argued that retired migrants possess, to different degrees, both local and international social capital: they bring with them their 'existing' social capital (international social capital) and then create 'new' social capital (local social capital) in their new place of residence (Casado-D ıaz, 2009). In this article, we focus on the analysis of the international component of social capital, particularly of the bonding type, by examining the migrants' interpersonal relationships and transnational network connections with family and close friends in their home country. So, after relocating to the new home, have migrants' ties with their homeland weakened or have the migrants succeeded in keeping up such relationships? Which factors have influenced this behaviour? In order to investigate these questions, the study analyses migrants' mobility patterns (reciprocal visits by friends and relatives) and their contacts (by phone, Internet, etc.) to ascertain the strength of the migrants' bonds with their home country and, ultimately, the strength of their international bonding social capital.

Keeping the ties: transnational mobility patterns and social contacts
Retired migrants might establish ways of keeping the links with their long-time friends and relatives back in their home countries, such as participating (and reciprocating) in VFR tourism and through long-distance (virtual) communication (regularity of social contact with others, such as speaking, writing, emailing, etc.). According to Urry (cited in Gustafson, 2009, p. 83), this interplay between physical travel and other ways of communicating and maintaining social ties 'at a distance' enables the migrants to sustain the international social ties and networks, and the resources derived from these.
Communication technology has greatly enabled the development of dense social, cultural and economic networks among transnational communities in which mobility has become the norm (Duval, 2004). However, while telecommunication may facilitate ongoing emotional support among migrant communities (Ryan et al., 2008), individuals also rely on physical mobility for the maintenance of geographically expanded and increasingly disperse social networks (Larsen, Urry, & Axhausen, 2006). Leisure scholars have noted the role of leisure in the development of social capital and the strengthening of social ties (Glover & Hemingway, 2005;Glover, Parry, & Shinew, 2005;Rojek, 2005). Research has shown how leisure might provide individuals with a sense of belonging, support and social interactions, hence assisting them in forming community networks and bonds that are important for social cohesion (Bush & Baum, 2001). Similarly, it has been argued that leisure travel, such as VFR tourism, plays an important role in sustaining geographically disperse social networks (Lew & Wong, 2002) and maintaining the social capital of these networks and of the individuals involved in them (Larsen et al., 2006;Larsen, 2008). Leisure (travel) can thus act as a social lubricant for social capital generation . In the context of IRM, leisure has certainly facilitated both informal and formal relationships among the members of communities of older northern Europeans, and the nature of many of these clubs and associations reveals a strong leisure component (Casado-D ıaz, 2006;Gustafson, 2008;O'Reilly, 2000). It has been argued that these newly formed networks of relationships provide the retirees with an informal communal system of self-help and reciprocity and give them a strong sense of community (Casado-D ıaz, 2009).
Previous work has also discussed the strength of bonding types of social capital based on a sample of British retirees living in the Costa Blanca, Spain, and the lack of bridging social capital, specifically in terms of social ties with members of the Spanish local community (Casado-D ıaz, 2009). This study found very limited bridging forms of social capital among this community of retired British migrants, a community bonded by their shared nationality, language, class and interests. Although there were some links with other nationalities and with the host community, the social networks the respondents belonged to were predominantly British. The findings suggested that the lack of contact with the surrounding society was mostly associated with the retirees' lack of knowledge of the Spanish language. The migrants' inability to communicate in Spanish was clearly detrimental to fostering bridging types of social capital at the local level while at the same time reinforced their need for 'bonding' with other members of the British community. It must be noted that the general lack of contact with the Spanish local community (a source for acquiring social capital of the 'bridging' type) was not an isolated feature of British retirement migration to Spain. Similar results were found for the Swiss and Swedish retirees who had relocated on a seasonal or permanent basis to this country (Gustafson, 2009;Huber & O'Reilly, 2004). As argued by Ryan et al., 'migrants' ability to mobilise social capital and successfully engaged in bridging may depend upon the cultural capital (language, skills and educational qualifications) at their disposal ' (2008, p. 677). It is therefore reasonable to assume that migrants' lack of knowledge of the local language might reinforce bonds with the country of origin. Thus, we propose that the migrants' international social capital may differ depending on their proficiency in the local language.
Although the process of migration may weaken family and kinship networks, depriving migrants of a crucial form of social capital (Coleman, 1990;Portes, 1998), it might also encourage migrants to develop strategies to keep their existing international bonding capital (in their home countries) somehow alive. One relevant factor for keeping bonding networks with their homelands and reaffirming social belonging is having another home in their country of origin (Iorio & Corsale, 2012, p. 25). However, this factor must be considered with caution because as noted by Williams et al. (2000, p. 37), although the retention of a home in the UK is an important facilitator of a peripatetic lifestyle, it is not necessarily a sign of attachment to it. Diverse circumstances, such as market conditions for selling or individual financial strategies (i.e. it may be viewed as an investment), or the fact that they did not own homes at the time they retired might also influence this variable. Therefore, we propose that another factor that might influence the ability of the migrants to mobilise the resources attached to their international social ties might be the ownership (or lack) of property in the UK.
The pool of variables considered is completed through the inclusion of age and gender, which have been found to play a relevant role in explaining differences in migrants' social practices in research on IRM (e.g. Casado-D ıaz 2006), and by the examination of the effect of 'length of time resident in the local area' on the migrants' transnational practices in order to further explore the 'time' effect on the variables analysed.

Data and method
In order to explore the proposed framework, a survey was conducted based on a sample of British retirees living in the Costa Blanca, Spain. The Costa Blanca is located in the province of Alicante, in the southeast of Spain, and is one of the most popular destinations for both tourism and retirement migration in Europe. With a mature tourist infrastructure and a pleasurable climate, this stretch of the Mediterranean coast has attracted large numbers of older foreign residents from Northern Europe.
The questionnaire was administered to British retirees aged 50 years or more who lived in Spain for at least 4 months a year. As no sampling frame of immigrants exists, and due to the high levels of participation of retirees in associations reported in previous studies (Casado-D ıaz et al., 2004), access to respondents was sought through the University of the Third Age (U3A). The U3A is a non-profit organisation run by older British volunteers and with local branches in the UK and abroad. Although the U3A branches organise their own social and cultural activities independently, many of the branches located in the Costa Blanca hosted similar leisure-oriented groups based around sports, arts and hobbies or more casual leisure (such as dining at home). The ethos of the U3A is to draw upon the knowledge and skills of their own members for the setting up of interest groups. At the time of the research, there were more than 3000 U3A members in the Costa Blanca area, making them the largest British social group in the region. The survey used self-completion questionnaires distributed and collected by hand through the branches of the U3A in the towns with the largest concentration of older British migrants, that is, Torrevieja, Calpe, Javea, Denia and Vall del Pop. Three hundred and sixty-nine questionnaires were collected via attendance of social meetings and events organised by the different branches of the U3A, where the self-administered questionnaires were distributed and collected by the interviewer. Four questionnaires were unusable due to incomplete responses, leaving a total sample size of 365.
Since our sample of British retirees was pragmatic rather than representative, the purpose of our study was to ascertain behavioural trends and retirement patterns among those interviewed, rather to make generalisations from sample to population. Furthermore, the analysis of the transnational practices of those included in our sample might not necessarily mirror those of other British retirees living elsewhere, particularly given the heterogeneity of IRM (Casado-Diaz et al., 2004). In order to investigate the migrants' social capital, this research framework considers two aspects of their social lives: the intensity of bidirectional VFR tourism flows (i.e. participating in and hosting friends and relatives visits) and the social contacts with friends and/or family living in the UK. Mobility patterns were examined by using two main indicators: 'return visits to the UK' and 'visits from friends and relatives'. 'Return visits to the UK' was measured with four items: two metric scales -'number of visits to the UK last year' and 'number of weeks spent in the UK last year' -and two non-metric (nominal) scales -'reasons for visit to the UK' (visit friends, visit relatives, leisure, work/business, to access medical services, etc.) and 'accommodation used in the UK' (own, rented, paid, friends or relatives, etc.). 'Visits from friends and relatives' embraced three non-metric items: 'visitors from the UK' (Are you ever visited in Spain by relatives/friends from your country of origin?), 'people from the UK visiting last year' (children and/or grandchildren, other adult relatives, friends or acquaintances) and the duration of the visit from people belonging to each category ('number of weeks spent by people from the UK during their visit'). The questionnaire also included a variable measuring the reasons for not witnessing visits in the previous year. Transnational social contacts were examined by means of two non-metric scales: 'contacts with the UK through phone, fax and/or mobile' and 'contacts with the UK through the Internet' (in both cases the alternative potential answers were daily, every week, several times a month, once a month, at least once a year, never). Finally, we consider age and gender, the 'length of time resident in Spain' (number of years) and 'knowledge of the local language' (none, some knowledge, quite fluent, fluent). Table 1 highlights the social and educational characteristics of the sample of British retirees interviewed for this study. The average age of the respondents in the sample was 65 years with a slight predominance of women. This might be explained by the fact that many of the social activities organised by members of the U3A in the Costa Blanca were usually led and supported by women, and, since the questionnaires were distributed and collected in these social gatherings, there was a slight overrepresentation of females in the sample. Most of the respondents were either married or lived with their partner (80%) and half of them had completed a university or college degree. Prior to their move to Spain, 60% of the respondents had no experience of living abroad, and, in terms of their previous employment, there was a predominance of professional (41.6%), and managerial and technical (24.9%) occupations. At the time of completion, 97% of those interviewed stated their current occupation as 'retired'. The majority considered themselves full-time residents (77.8%) and owned their homes in Spain (83.6%). In general, the sociodemographic characteristics of the respondents were in line with previous studies examining foreign retirement migration in Europe (e.g. Casado-D ıaz et al., 2004).

Results
The rest of this section presents two sets of data related to the migrants' transnational network connections: VFR tourism (return visits to the UK and visits by friends and family) and strategies for maintaining the social ties with their country of origin. The analysis of the survey results (n ¼ 365) combines descriptive statistics and analyses of variance (ANOVA) in order to identify significant differences between the main variables. Note that in ANOVA analyses, VFR tourism is limited to the 'return visits to the UK' variable, measured by the item 'number of weeks spent in the UK last year'. This item is believed to better measure the degree of contact with the country of origin since it reflects the total amount of time spent in the UK independent of the number of visits undertaken. ANOVA analyses were not conducted for the remaining items due to their qualitative nature. Regarding social contacts we employ both measures: 'contacts with the UK through phone, fax and/or mobile' and 'contacts with the UK through the Internet'. For the purpose of the analysis, age is coded into three categories: the youngest group (YG) includes those respondents aged 59 years or less, the medium group (MG) includes those respondents aged between 60 and 69 years, and the oldest group (OG) includes those respondents aged 70 years or more. 'Length of time resident in Spain' is coded into four categories: 0-3 years (LR1), 3-6 years (LR2), 6-9 years (LR3), and 10 years or more (LR4). Knowledge of the local language is coded into four categories: a few words/none (KNW1), some knowledge (KNW2), quite fluent (KNW3) and fluent (KNW4). Finally, the type of accommodation used when visiting the UK is considered a proxy of ownership of property back in the country of origin (a bonding type of social capital) for further analyses and is coded into three categories: own accommodation (ACC1), with friends/ relatives (ACC2) and paid accommodation (ACC3). The level of satisfaction showed by those interviewed with regard to these social contacts/social ties with the friends and family back in the UK is also discussed.
Transnational mobility patterns: VFR tourism to and from the country of origin In order to analyse the strength of the migrants' international social networks, the proposed research framework considers in the first place the intensity of VFR tourism, both to and from their country of origin. Table 2 summarises the retirees' mobility patterns regarding return visits to the UK. The frequency of return visits to the UK is a powerful indicator of the strength of the social ties kept by the retirees with their communities of origin. The questionnaire results showed that the vast majority of the retirees visited the UK in the previous year in order to see their family and friends. Thus, only 10% of those interviewed stated that they had not travelled to the UK in the previous year, whilst 28% of the retirees had travelled to the UK at least once in the previous year and 62% had taken two or more trips suggesting very strong links with friends and relatives in their country of origin. When asked about the reasons for those visits, the respondents stated that they were primarily for the purpose of visiting family (78%) and friends (44%), a fact that was also reflected in their preferred choice of accommodation (54% stayed with friends or family), and the length of the visit was between 1 and 6 weeks in 61.6% of the cases. Additionally, Table 3 shows that, as well as being participants in these tourism flows, most respondents became recipients of VFR travel. Almost all of them (98.1%) hosted visitors from the UK during the past year with most visits lasting between 1 and 2 weeks. The most frequent visitors were children and/or grandchildren (76.4%), although the number of other adult relatives who travelled to Spain to see them was also important (50%). The strong kinship networks maintained in the UK were also evident by the high proportion of respondents that had also hosted visits by friends (66.8%). The scarce number of respondents (1.6%) who did not witness visits from the UK during the previous year believed that the mains reasons for this could be the increasing cost of international travel and the fact that they might not have sufficient room to lodge them, not because they might have lost contact with them. These results exemplify the significance of VFR tourism to international retirement destinations and suggest that the respondents have managed to retain close ties with their relatives and longstanding friends in their home countries through both participating and hosting visits from friends and relatives, as stated in previous studies on IRM (Casado-D ıaz, 2009;Gustafson, 2008;Huber & O'Reilly, 2004;. The amount of visits (frequency of contacts) provides evidence of the fact that these retirees still rely heavily on existing social networks in their home countries. These results are relatively surprising and reflect that this set of migrants are not drifting away however long they stay in Spain. One possible explanation might be the personal circumstances of this set of migrants, where many of them are well educated and resourceful, and they are aware of the importance of maintaining their 'international bonding social capital' for emotional as well as instrumental reasons (e.g. possible return, onset of illness).
To shed some light on these results, we ran ANOVA tests in order to explore the existence of significant differences between the main variables. Regarding sociodemographic characteristics, our findings show that there are no gender differences for the variable 'number of weeks spent in the UK' (t ¼ 0.06, p > 0.05). However, there are some significant differences regarding age. The ANOVA tests provide evidence of the fact that the older the respondents, the least number of visits they took to the UK (F ¼ 4,418, p < 0.05), with the oldest group showing a significant different (lower) mean (i.e. few weeks spent in the UK) in comparison to the medium group (mean difference: MG -OG ¼ 0.463, p < 0.05). Results also indicate that there are no differences regarding the 'length of time resident in the local area' in terms of the 'number of weeks spent in UK' (F ¼ 0.687, p > 0.05), but there are regarding 'knowledge of the local language' (F ¼ 4.14, p < 0.01). Specifically, we observe that the least knowledge of the local language declared, the more number of weeks spent in the UK. The two groups that state they have poor knowledge of Spanish spend more number of weeks in the UK than the other two groups (mean differences: KNW1 -KNW4 ¼ 1.25, p < 0.01; KNW1 -KNW3 ¼ 0.81, p < 0.01; KNW2 -KNW4 ¼ 1.02, p < 0.01; KNW2 -KNW3 ¼ 0.58, p < 0.01). Also, as expected, the 'ownership of property in the UK' has an effect on the number of weeks spent in the UK. Those who still own properties in the UK spend more number of weeks on average than those who do not (mean differences: ACC1 -ACC2 ¼ 1.16, p < 0.01; ACC1 -ACC3 ¼ 1.38, p < 0.01).
Transnational social contacts: keeping in touch 'at a distance' As well as visiting (and being visited by) long-term friends and family in their country of origin, most respondents kept in touch with them by regularly calling and/or emailing them. Table 4 shows the intensity of the respondents' contacts with friends and family back in the UK. A large proportion of the respondents (73%) have contact with the UK on a daily/weekly basis through phone, fax and/or mobile, and around 55% of the sample used the Internet to communicate with friends and/or family back in the UK on a daily or weekly basis. The majority of the respondents also 'felt' they had kept close relations with their friends and relatives in the UK (see Table 5). Most of them 'strongly agreed' or 'agreed somewhat' to the following statements: 'I regularly keep in touch with friends/ relatives in the UK' (93%); 'Back in the UK I have close friends who I could always turn to for help' (80%); 'I keep close personal relations with friends back in the UK' (70%). These results provide evidence of the intensity of the contacts and also the importance of communication technologies, particularly the Internet, for the maintenance of the migrants' kinship and friendship networks and the resources derived from them, which is their international bonding social capital.
The survey also included a number of questions designed to ascertain the level of satisfaction of the respondents with a series of statements related to the frequency and intensity of those contacts with friends and family back in the UK ( Table 6). Most of the respondents were either 'very satisfied' or 'somewhat satisfied' with the contacts maintained with family and friends in the UK (82%), the frequency of visits by friends and family (72%), and the frequency of their visits to the UK (65%). Again, these results confirm the presence of high levels of international bonding capital for the sample analysed (in terms of VFR tourism and social contacts).  ANOVA tests were used to explore the existence of significant differences for the explanatory variables. No significant differences were found between men and women in terms of the 'frequency of contacts with the UK through the Internet' (t ¼ À0.41, p > 0.05) and 'frequency of contacts with the UK through phone/fax/mobiles' (t ¼ À0.39, p > 0.05). Regarding age, no significant differences were found for the variable 'frequency of contacts with the UK through phone/fax/mobiles' (F ¼ 0.51, p > 0.05), but they were found for the variable 'frequency of contacts with the UK through the Internet' (F ¼ 11.10, p < 0.01). Specifically, we find that the older the respondents, the least frequent were Internet-based contacts with people back in the UK (mean differences: YG -OG ¼ 1.26, p < 0.01; MG -OG ¼ 0.71, p < 0.01; YG -MG ¼ 0.55, p < 0.05). Results also show differences regarding the 'length of time resident in the local area' in terms of the 'frequency of contacts with the UK through phone/fax/mobiles' (F ¼ 2.81, p < 0.01) and also for the 'frequency of contacts with the UK through the Internet' (F ¼ 9.15, p < 0.01). Those migrants who had lived in Spain for longer had fewer contacts than the 'youngest' group (those living for less than 3 years in Spain). This has been observed for the 'frequency of contacts with the UK through phone/fax/mobiles (mean differences: LR1 -LR4 ¼ 0.41, p < 0.05) and for the 'frequency of contacts with the UK through the Internet' (mean differences: LR1 -LR4 ¼ 1.29, p < 0.01; LR2 -LR4 ¼ 1.09, p < 0.01). No significant differences were found for the factor 'knowledge of the local language', nor for the 'frequency of contacts with UK through phone/fax/mobiles' (F ¼ 0.71, p > 0.05) or the 'frequency of contacts with UK through Internet' (F ¼ 0.77, p > 0.05). Finally, the 'ownership of property in the UK' has no effect on the frequency of contacts with the UK for both the 'frequency of contacts with the UK through phone/fax/mobiles' (F ¼ 1.76, p > 0.05) and the 'frequency of contacts with the UK through the Internet' (F ¼ 2.40, p > 0.05).

Conclusions
This article has contributed to recent developments in the study of IRM by exploring the relevance of transnational social networks and the role that leisure travel, particularly VFR tourism, plays in maintaining these social ties and, ultimately, in sustaining the migrants' international bonding social capital. We have argued the potential of the concept of social capital and Putnam's (2000) typology of bridging and bonding forms of social capital for the analysis of the transnational social practices of international retired migrants.
As with other forms of lifestyle migration, their search for a better quality of life is ultimately reflected in the perceived satisfaction with their 'new' life abroad. Results show that most British retirees in our sample have succeeded in maintaining their international bonding social capital by regularly 'keeping in touch' and visiting (and being visited by) their families and friends back in their home country. As well as providing them with a highly valuable social resource in terms of social support, the migrants' transnational practices and networks enable them to maintain a sense of belonging and place attachment and, ultimately, to keep an 'open' channel with their country of origin in the case of an eventual return. It is the migrants' ability to mobilise these resources for the pursuit of advantage or the mitigation of disadvantage (i.e. keeping an 'open door' in the case of an eventual return) which gives them value as 'capital' (Anthias, 2007). More specifically, the analysis of the data reveals both the strength of the retirees' international bonding social capital and the role of VFR travel and communication technologies in sustaining these transnational social practices and, ultimately, these 'transnational lifestyles' among amenity-seeking older migrants. These findings are in line with previous research on the importance of VFR tourism to international retirement destinations (Casado-D ıaz, 2009;Gustafson, 2008;Huber & O'Reilly, 2004;. The analysis suggests that the older the respondents, the fewer number of visits they took to the UK and that those with less knowledge of Spanish spent more weeks in their home country (that was also the case for those who still maintained a property in the UK). These results might indicate that knowledge of the local language acts as a 'bridge' for contact with the local community, hence reducing the frequency of the visits to the home country. Similarly, the migrants' social ties with the UK do not necessarily weaken over time, as might be expected with international migrants, and this is reflected in the results regarding age, except when the migrants have reached a high level of 'integration' in the local community as a result of their ability to speak Spanish. To this effect, our sample of British retirees seem to be participants of an ongoing migration process by which they are successfully maintaining a varying degree of attachment to both their home countries and their new places of residence.
As for international social contacts, a large proportion of the respondents stated that they were regularly in touch with friends and/or family back in the UK. They also felt that they had successfully maintained 'close personal relations with friends back in the UK' and expressed a high degree of satisfaction with these personal relationships 'at a distance'. These results suggest a deliberate effort of 'keeping up' their international bonding social capital and, as argued by Ryan et al., their 'transnational sources of emotional support' (2008, p. 673). The analysis of the data suggests that these social contacts are independent of gender or age, except for the use of the Internet which is lower in the oldest group. Additionally, results show that those migrants who have lived in Spain for longer have fewer social contacts with the UK than the 'youngest' group (those living for less than 3 years in Spain).
Similarly to the findings reported by Portes (1998, p. 14), it would seem that British retirees have compensated for the absence of bridging forms of social capital at the local level by putting an emphasis on kin and friendship ties as a source of social support, one of the basic functions of social capital. While previous studies have noted the salience of friendship among international retired migrants, particularly for the creation of bonding types of social capital in their new communities of residence (Casado-D ıaz, 2009), the results from this study acknowledge the relevance of the migrants' social ties 'back home'. As argued by Gustafson (2009), the seasonal mobility of the retirees gives an additional meaning to the concept of lifestyle migration since their recurrent migration becomes part of their lifestyle in an attempt to achieve the best of both worlds. Results suggest that this particular form of migration could be seen as an ongoing process rather than a one-time permanent move (Gustafson, 2008). Regular (and reciprocal) visits enable migrants to sustain and reinforce their existing international bonding social capital, a capital derived from their existing strong kinship and friendship ties 'back home' and the social resources attached to them and, in Putnam's words, generally good for 'getting by ' in life (2000, p. 22).
In terms of limitations of our study, it is important to bear in mind that those interviewed through the U3A meetings are prone to report higher levels of sociability than others who might not belong to similar leisure-based associations. Future studies will also need to widen sampling frames in order to include not only those migrants who are members of leisure-based associations, but also those who might not. Similarly, this study has not focused on other groups of older European migrants with fewer resources and who might not have been so successful in their search for a better lifestyle (O'Reilly, 2007). Another area of comparative research that requires further consideration is the extent to which international retired migrants have succeeded in getting 'the best of both worlds', that is, whether they have both managed to create new social ties in the destination country (local social capital) and maintained the existing ones in their country of origin (international social capital) and how this is reflected in their overall level of satisfaction with their new post-retirement lifestyle abroad. Migrants might access social support through a combination of established and newly formed networks involving a mixture of contacts in both the country of origin and destination (Ryan et al., 2008, p. 673), so much research is needed into the ways in which the social networks established by the migrants, both locally and internationally, and the resources attached to them, will evolve in the future. Similarly, the impact of IRM on the receiving communities remains significantly underresearched. This includes more traditional issues, such as the ways in which the spatial concentration of large (and highly mobile) communities of retired EU migrants might alter economic, social, political and demographic structures at the destination level, together with the demands that they might impose on local state services (particularly health and welfare services due to the migrants' old age).