The roles of family and friends in the immobility decisions of university graduates staying in a peripheral urban area in the Netherlands

Abstract Highly educated individuals constitute great assets for regional development and economic growth. Nevertheless, young university graduates are relatively geographically mobile and less likely to stay in peripheral regions. Based on semi‐structured, life‐calendar interviews, this study explored the immobility decisions of graduates who have stayed in a peripheral urban area in the Netherlands where they completed their university education. The study specifically focused on the roles of family and friends in the staying processes of these young adults. The results indicate that the decision to stay was frequently and consciously re‐evaluated by some, whereas for others, it resulted from a ‘lack of triggers’ for moving elsewhere. Notably, the interviews revealed that family and friends act as more than motives for staying or deterrents to migration. On various occasions, family and friends had played crucial roles as advisors, influencers, triggers, exemplars and facilitators in the staying processes of highly educated young adults.

This study responds to a call for more thorough investigations into the dynamics between staying behaviour and social networks (Hjälm, 2014) required for strengthening the theorisation on the roles of family and friends in residential decisions (Faist, 1997;Mulder, 2018). It does so by investigating the roles of family and friends in the staying processes of highly educated young adults in a peripheral urban area in the Netherlands. Notably, the study treats spatial immobility as an active process in one's residential trajectory, signalling a newly emerging perspective within the migration literature (e.g., Coulter, van Ham, & Findlay, 2016;Haartsen & Stockdale, 2017;Hjälm, 2014;. Data for this study stem from semi-structured, life-calendar interviews with 15 Maastricht University alumni who have stayed in Maastricht since completing their tertiary education trajectories. The city of Maastricht is located in the province of Limburg, which is a peripheral area in the south of the Netherlands. The city and the province typically experience out-migration of highly educated young adults (e.g., Hooijen, Meng, Reinold, & Siegel, 2017;Venhorst et al., 2010). The interviews revealed a diverse range of factors that may trigger reevaluations of the decision to stay as well as the role of a 'lack of triggers' for continued staying behaviour. More specifically, the interviews confirm that living close to family and friends can be an important motive for staying, as well as a deterrent to migration. The most significant contribution to the literature may be found in the descriptions of family and friends serving as advisors, influencers, triggers, exemplars and facilitators in the staying processes of highly educated young adults.

| The decision to stay
One of the first qualitative studies on lifelong staying behaviour put forward the idea that 'staying is by no means just something that "happened to" the informants' (Hjälm, 2014, p. 577). In fact, that study revealed that the decision to stay is commonly 'an active and informed choice' and may be re-evaluated and renegotiated over the life course (Hjälm, 2014, p. 577). Since then, other studies have confirmed that immobility decisions are not one-off events but consciously revisited (e.g., Haartsen & Stockdale, 2017;. For example, some informants in a study by Haartsen and Stockdale (2017) were 'convinced stayers' , yet they anticipated re-evaluating their staying decision at the onset of future life-course changes or life events, such as transitioning to an empty-nest life phase or reaching old age and becoming less physically abled. These findings provide evidence for an idea from the migration literature that residential processes are fluid and never complete (Halfacree, 2004(Halfacree, , 2018, but they are novel in treating nonmigration as an active component in these processes. The findings further suggest that nonmigration experiences should be positioned more fully within one's biography to take into account the role of the life course and key events, as has been argued for migration experiences (Barcus & Halfacree, 2018;Halfacree & Boyle, 1993).
By positioning staying processes more fully within one's biography, the existing immobility literature has framed past, present and future life events as potential triggers for a re-evaluation of the decision to stay (e.g., Haartsen & Stockdale, 2017;Ye, 2018). Haartsen and Stockdale (2017) have further suggested that a history of migration may trigger more conscious re-evaluations of one's residential situation because movers are already familiar with the re-evaluation process. To this end, it should be noted that each re-evaluation, now and in the future, may result in either staying or moving behaviour. Hence, Stockdale et al. (2018, p. 6) have argued that conscious renegotiations of the decision to stay signal that staying processes are 'in a state of flux'.
2.3 | The roles of family and friends in the decision to stay At the same time, Hjälm (2014, p. 579) has argued that the decision to stay 'does not occur in isolation but is connected to other life projects and people'. In order to understand how immobility decisions or re-evaluations thereof are influenced by others, the existing framework regarding the decision to stay should include the notion of 'linked lives'. The 'linked lives' approach encourages abandoning the individual as the sole unit of analysis and, instead, views life events as experienced together with others, such as family and friends (Elder, 1994;Elder, Johnson, & Crosnoe, 2003). Life events that are typically experienced with others during young adulthood include leaving the parental home, starting to cohabit, getting married, having children, leaving a relationship, experiencing loss and parental divorce.
Some of these collective experiences may well affect the relationship between young adults and their parents, siblings, partners or housemates (e.g., Bidart & Lavenu, 2005). Notably, leaving the parental home may decrease the frequency of face-to-face contact and change the type of support that is given and received between young adults and their parents. It could be argued that such changes to the parent-child relationship may be more prominent among young adult migrants than those who stay in their hometown. Nevertheless, Hjälm (2011Hjälm ( , 2014 notes that a decision to stay may affect the social network too, because staying may lead to extensive longstanding social networks as well as feelings of isolation if family and friends move away. These contrasting immobility experiences led Hjälm (2011Hjälm ( , 2014 to call on future research to explore the dynamics between staying behaviour and the social network in more detail. The 'linked lives' approach further suggests that family and friends may influence the course of one's life events (Elder, 1994;Elder et al., 2003) and this raises the question of how linked lives affect one's residential trajectory. Thus far, the literature on family ties in relation to mobility behaviour has found that living close to family increases the likelihood of staying and of returning, as well as decreasing the propensity to migrate (Clark et al., 2017;Ermisch & Mulder, 2018;Michielin et al., 2008;Mulder & Malmberg, 2011, 2014Niedomysl & Amcoff, 2011;von Reichert, Cromartie, & Arthun, 2014;Zorlu, 2009). Similarly, living close to friends decreases the likelihood of migrating (Belot & Ermisch, 2009).
Thus, living close to family and friends is expected to constitute a motive for staying and act as a deterrent to migrating in the immobility decisions of young, highly educated adults.
Another way in which family and friends have been found to exert influence on one's residential behaviour is through socialisation.
Socialisation theories suggest that one's preferred behaviour is informed by choices, preferences, attitudes and examples set by others. Residential experiences during childhood, experienced together with household members, are found to be positively related to similar residential behaviour in adulthood (AErø, 2006;Bernard & Vidal, 2020;Blaauboer, 2011). On the family level, the transnational migration literature has shown that migration behaviour is passed down through generations in a family (e.g., Guveli et al., 2016). Similar findings have been presented regarding passed-down migration intentions of young adults in peripheral regions (Thissen, Fortuijn, Strijker, & Haartsen, 2010). On a more aggregated scale, migration has been found to trigger further migration, known as the 'cumulative causation of migration' (Massey, 1990). The immobility decisions of the highly educated young adults are thus expected to be informed by the residential behaviour of their social environment, including family and friends.

| Other motives for staying
Of course, staying behaviour may be motivated by factors other than family and friends. Some studies have highlighted the role of personal factors, the physical environment and the cultural environment as motives for staying (e.g., Alberts & Hazen, 2005;Haartsen & Stockdale, 2017). Most notably, numerous empirical studies have found that the length of stay increases the likelihood of continuing to stay (Clark et al., 2017;P. A. Fischer et al., 2000;P. A. Fischer & Malmberg, 2001;Thomas, Stillwell, & Gould, 2016). This has been attributed to the idea that one feels more resistance towards moving as the duration of stay increases, also known as 'cumulative inertia' (e.g., Huff & Clark, 1978). This is also why continued staying behaviour has been associated with increased feelings of place attachment (Clark et al., 2017;Di Masso et al., 2019), rootedness (Cooke, 2011;C. S. Fischer, 2002) and the accumulation of local ties and location-specific capital (Blaauboer, 2011;David, Janiak, & Wasmer, 2010;P. A. Fischer & Malmberg, 2001;Mulder & Malmberg, 2014). Others have identified staying behaviour as the result of a 'lack of triggers' or a lack of reasons for moving (Mulder, 2006), which may reflect the absence of push factors or 'residential stress' in the current location and of more attractive places elsewhere (Hjälm, 2014;Huff & Clark, 1978). Finally, applying a life course approach to long-term immobility, Hjälm (2014) found that stayers mention multiple reasons or rationales for the decision to stay at various phases of the life course. Indeed, Haartsen and Stockdale (2017) have since then showed that there may be a temporal dimension to one's stated motives.
2.5 | Context: Spatial immobility among university graduates in Maastricht The motivation for this study's focus is based on the premise that spatial immobility is relatively uncommon among highly educated, young adults. In part, this was deduced from the universal age profile of migration, which has revealed that migration propensities are highest during young adulthood (Bernard et al., 2014;Rogers & Castro, 1981).
In addition, more highly educated individuals show higher migration frequencies (Chudnovskaya & Kolk, 2017;Lundholm, 2007), tend to move over longer distances (Champion & Shuttleworth, 2017;van Ham, 2002) and towards centres of job opportunities, also known as escalator regions (Fielding, 1992;Findlay et al., 2009;Venhorst et al., 2010Venhorst et al., , 2011. In the Netherlands, access to jobs tends to be worst in the peripheral regions ( van Ham, Mulder, & Hooimeijer, 2001) towards the southern borders with Belgium and the eastern borders with Germany. Access to jobs and further opportunities for upward job mobility are highest in the Randstad (van Ham et al., 2001), the central urban region that includes some of the country's primary cities: Amsterdam, Rotterdam, The Hague and Utrecht. In Fielding's (1992) terminology, the Randstad can be considered the Netherlands' primary 'escalator region' (van Ham et al., 2001). Accordingly, a significant number of students and graduates from the Dutch peripheral regions move to the Randstad (Latten et al., 2017;Venhorst, 2013;Venhorst et al., 2010Venhorst et al., , 2011. The spatial context for this study is Maastricht: a university city located in a peripheral area in the south of the Netherlands on the border with Belgium. It is the provincial capital and the largest urban conurbation in the Dutch province of Limburg. The city houses a young, internationally competitive university, which has frequently ranked in the top 10 of the Young University Rankings (Times Higher Education, 2020). Despite this, most of its graduates move towards the Randstad or abroad upon graduation (Hooijen et al., 2017;Venhorst et al., 2010). The university's alumni who decide to stay in Maastricht embody strong human capital that is crucial for Limburg's development and economic growth (e.g., Venhorst et al., 2010).

| The recruitment process and sample
This study set out to explore the immobility decisions of Maastricht University alumni who have stayed in Maastricht since completing their tertiary education. The author requested members of an alumni community on social media to respond if they were interested in taking part in the study. The request contained a decision-tree, which ensured that potential interviewees met the following five selectionrequirements: (1) to hold a master's degree, or equivalent, 1 from Maastricht University; (2) to have lived in Maastricht while completing their master's degree; (3) to have stayed in Maastricht since graduating from Maastricht University; (4) to be no more than 40 years old; and (5) to be born in the Netherlands or the bordering Belgian province of Limburg. The author received over 30 responses of which 15 were from young adults who met all the requirements and were able to meet for an interview. Some self-selection bias in the recruitment process may have been created by targeting members of an alumni-group as these tend to foster particularly positive feelings towards their former university or its city. We targeted young adults who were born in the Netherlands or the bordering Belgian province of Limburg to ensure that all interviews could be held in one language, namely, Dutch. This reduces validity issues related to the interpretation and translation of language-specific expressions (e.g., van Nes, Abma, Jonsson, & Deeg, 2010). Nevertheless, the staying processes of non-Dutch-speaking graduates in Maastricht are equally relevant and may be investigated in future studies. Table 1

| Semi-structured, life-calendar interviews
The author opted for combining traditional, qualitative semistructured interviews with life-calendar grids. Qualitative interviewing methods help to reveal in-depth information about interviewees, whereas calendar-based interviewing methods have been developed to increase the richness of data in quantitative life-course research (Belli, 1998;Belli, Shay, & Stafford, 2001;Freedman, Thornton, Camburn, Alwin, & Young-DeMarco, 1988). More recently, life-calendar grids have been used to provide direction and chronological structure to qualitative studies and to capture the timing and interlinkages of life events more accurately (e.g., Barbeiro & Spini, 2015Kõu, van Wissen, van Dijk, & Bailey, 2015). These two methods were combined to allow a biographical approach to the staying processes of the interviewees, which was deemed crucial for treating nonmigration as an active component in their residential trajectories. In this respect, the life-calendar grid was utilised as a tool for stimulating long-term memory (Barbeiro & Spini, 2015Bradburn, Sudman, & Wansink, 2004). This was expected to aid the recollection of past, present and future life events, which are known to trigger re-evaluations of immobility decisions (e.g., Haartsen & Stockdale, 2017;Hjälm, 2014;Ye, 2018). Nevertheless, interviewers are cautioned about the many types of recall bias and recall errors that may arise in retrospective recollections (e.g., Auriat, 1991;Bell, 2005). for instance, about reflections on the staying period, motives for the decision to stay and residential intentions for the future. Following this, a life calendar was designed that corresponded to the subthemes in the interview guide (see Table 2 for an example of a blank life-calendar grid). The author conducted three test interviews before finalising the interview guide and the outline of the life-calendar grid; to enhance legibility and allow the interviewee to read and correct the interviewer's interpretations, it was found that the life-calendar grid was to spread across two landscape-oriented A3 sheets. At the start of each interview, the life-calendar grid was personalised by entering the interviewee's age and corresponding calendar years in the first two rows. As such, each completed lifecalendar grid represented a personalised timeline that was annotated according to the interviewees' retrospective recollections of life events and experiences, thus containing extensive information about the interviewee's life course and staying period.
The interviewees were offered a choice of where to meet and eight chose to be interviewed in the author's temporary home in Maastricht and seven were interviewed at their workplace in the area of Maastricht. The interviews were recorded and lasted between 50 and 105 min. Here, one should note that the author studied at Maastricht University and lived in Maastricht several years prior to the interviews. This shared identity, residential location and preexisting knowledge of location-specific aspects may have influenced the statements given by the interviewees. For example, shared knowledge can help the interviewer understand the interviewees' locationspecific references, but it can also negatively affect the richness of the recorded data if interviewees refrain from offering detailed information because they assume the information is already known to the interviewer.

| Analytical approach
The interviews were transcribed verbatim and the life-calendar grids were photographed and digitised. Except for 'Maastricht', all names, locations and other identifiers were changed to pseudonyms or general geographical areas to ensure privacy and anonymity. Following this, the digital transcripts and life-calendar grids were coded and analysed in Atlas.ti software. The biographical approach included linking the narratives of the interviewees to specific moments and events in their life courses or staying periods; visualising and mapping out the decision-making processes; and adding in-depth information about the various factors that influenced mobility and immobility decisions. As recommended for non-English qualitative data (see van Nes et al., 2010), relevant extracts were translated to English with the help of a professional language editor towards the end of the analytical process.
Three main themes arose from deductive and inductive coding (i.e., through a thematic analysis), namely, the decision to stay; the roles of the microcontext, mesocontext and macrocontext in the staying process; and the stated motives for staying. The last theme was the direct product of the answers to the question, 'What is your main motive for staying in Maastricht?' , revealing a wide range  (Mulder, 2006). Moreover, comparing the transcripts of Levi (LIM, 26) and Lucas (NL, 38) revealed the possible effects of having previous migration experiences on re-evaluating the decision to stay. This relates to a finding by Haartsen and Stockdale (2017), who suggested that people with a history of migration may reflect on their staying behaviour more frequently or more consciously than natives and long-term stayers because they are familiar with exercising such decisions and such re-evaluations.

| Open to moving, but likely to stay
The recurring re-evaluations provide evidence for the idea that the decision to stay is not necessarily a one-off event. The interviewees seemed to be aware that each evaluation of the decision to stay-now and in the future-may result in either moving or Yes, you see, you tend to strengthen ties to an area.
Or, you grow more and more roots there as it were.
Those things make it more difficult to consider leaving and uprooting your life again. (Steffie,LIM,33) And that's when we knew we were in the right place, and also when we bought the house. That's the moment we decided that we were going to stay here for the long-term. Those things are interdependent and make you feel like settling down a little bit more. (Lucas,NL,38)  Living close to friends has also been found to deter migration (Belot & Ermisch, 2009

| Family and friends in combination with other motives
Besides being mentioned on their own, family-and friend-related motives would commonly be raised in combination with other factors.

.1 | Linked lives, advisors and influencers
Beyond the motives for the decision to stay, the transcripts revealed a number of other roles that family and friends played in the staying processes of the interviewees. Firstly, the concept of 'linked lives' (Elder, 1994;Elder et al., 2003) provided clarity in understanding how family and friends had been involved in the decision-making processes regarding the housing trajectory. The interviewees would consider the possible effects that their personal staying decisions could have on the wellbeing of their family and friends who would be on the receiving end of the decision-making process. These effects were particularly emphasised when parents or in-laws required care or when the interviewees had experienced parental divorce or loss in the family. The above quotes by Anne (NL,26) and Nina (LIM, 37) reiterate that they felt, as also did Thomas (LIM,27) and Hugo (LIM,27), that many fewer people had stayed than had moved away and those who had moved took social capital with them. At the same time, these responses reveal the investments and efforts that the interviewees had to make to facilitate their own staying behaviour because, although they are stable, the environment is changing (e.g., Hjälm, 2014). In turn, such investments in social capital may strengthen the ties to the area and increase the likelihood of continued staying behaviour. Therefore, the findings also highlight the active roles that the interviewees played in their personal staying processes (e.g., Ye, 2018). These findings are in line with previous findings in the immobility literature (e.g., Haartsen & Stockdale, 2017;Hjälm, 2014;Ye, 2018) and highlight the position of nonmigration experiences in the interviewees' biographies (Barcus & Halfacree, 2018;Halfacree & Boyle, 1993).

| DISCUSSION AND CONCLUSIONS
Moreover, the interviews have revealed that some young adults frequently and consciously reflected on the decision to stay. For others, spatial immobility had been the result of a 'lack of triggers' or a positive feedback between settling down and continuing to stay.
These findings highlight that staying processes are active (Haartsen & Stockdale, 2017;Hjälm, 2014) and affected by the length of stay (P. A. Fischer & Malmberg, 2001;Huff & Clark, 1978). Moreover, the interviewees with a history of migration may have re-evaluated their residential situations more frequently than those who had always lived in Maastricht. This cannot be claimed with certainty using this sample, but the finding is in line with suggestions of earlier studies (Haartsen & Stockdale, 2017). Using larger samples, the literature on transnational migration has indeed discovered a link between previous and intergenerational migration experiences and the future migration behaviour of individuals (e.g., Guveli et al., 2016). Similarly, the immobility literature could extend its explorations by investigating the effects of previous and intergenerational staying experiences on future staying behaviour.
Furthermore, by focusing on the notion of 'linked lives', this study has reemphasised earlier findings regarding the importance of geographic proximity to family and friends in mobility decisions (e.g., Belot & Ermisch, 2009;Clark et al., 2017;Ermisch & Mulder, 2019;Michielin et al., 2008;Mulder & Malmberg, 2011, 2014Niedomysl & Amcoff, 2011;von Reichert et al., 2014;Zorlu, 2009 (Faist, 1997;Mulder, 2018 Coding the staying processes was not straightforward. In part, the difficulty in coding was caused by the lack of existing conceptualisations for the various decisions that are made at different moments in a staying process. For example, the author was expecting to code various primary and secondary motives for staying, as is common for migration motives. However, the author soon realised that the interviewees did not rank their motives according to importance but stated time-dependent motives. This led to codes such as 'earlier motive', 'later motive' and 'motive as a student'. Such temporal dimensions have been reported in the literature and are in line with a biographical approach to the study of nonmigration (Haartsen & Stockdale, 2017;Hjälm, 2014) but have not led to clear conceptualisations for qualitative studies on nonmigration experiences compared with, for example, the decision-making processes of moving behaviour. Thus, a necessary condition for future immobility studies is to further unpack the staying process by means of conceptual exercises that enable the exploration of nonmigration experiences in a more comprehensive and biographical manner.
The main conclusion from this study is that immobility decisions, just like migration decisions, are diverse. A factor that greatly contributes to this diversity is the influence of 'linked lives'. Geographic proximity has been found beneficial for strong relationships and indeed presents one way in which family and friends played a role in immobility decisions. Nevertheless, regardless of geographic proximity, family and friends were found to fulfil other crucial roles in the staying processes of highly educated, young adults. In terms of the relevance of the findings, one should recognise that the interviewees constitute strong human capital, which is a vital element for regional growth and economic development in peripheral areas such as the Dutch province of Limburg. The interviewees are examples of highly educated individuals who have stayed in a peripheral urban area despite the high out-migration rates present among their peers.
On various occasions, possessing good social capital had facilitated some interviewees to stay despite a lack of suitable jobs or housing.
As such, the findings may help policymakers to understand for what reasons highly educated young adults come to stay in peripheral areas. Understanding what facilitates continued staying behaviour may contribute to policy ideas aiming to retain strong human capital.

ACKNOWLEDGEMENTS
The interview data were initially collected to develop the author's

CONFLICT OF INTEREST
The author has no conflict of interest to declare.