‘I don’t care about the city’: the role of connections in job-related mobility decisions of skilled professionals

Abstract Many cities in industrialized countries engage in competition for talent. While scholars have produced rich accounts of why some cities are better equipped than others in attracting skilled workers, the individual decision-making and agency of this group remains understudied. Contributing to this debate we examine how skilled professionals make their decisions in favour of locations. It focuses on the role of connections, which comprise social networks and take into account a range of professional, social and cultural ties between individuals and places. Our argument is based on empirical material from two qualitative case studies on work-related migration. The first studies professionals who moved to Frankfurt on a temporary basis from within Germany; the second focuses on high-skilled international migrants in Hamburg. We show that these supposedly privileged professionals often address limited job opportunities and personal constraints. Based on our findings and in addition to hard location factors, we argue that different forms of connections may further explain work-related moves. Despite similarities between the groups, internal migrants are more reliant on professional networks, while the international group is more likely to benefit from institutional connections. In our conclusion, we discuss how our research reveals inadequacies of current modes of urban policy-making.


INTRODUCTION
In industrial nations, global economic restructuring was accompanied by a profound transformation of labour markets and work organization (Boltanski & Chiapello, 2007). As a result of the increasing deregulation and flexibilization of labour, work-related mobility within and across national borders has become more commonplace.
In the emerging new global division of labour, the competitiveness of developed nations is increasingly associated with knowledge-intensive economic sectors. With regard to the labour market, this means an increasing focus on skills and qualifications. As a result, some professions, such as those in information technology (IT), engineering or research and academia, have become Regional StudieS, Regional Science increasingly mobilized (e.g., Bauder, 2012). For skilled professionals, mobility has become an inherent characteristic of their job descriptions and career biographies (e.g., Bröckling, 2015). For the purposes of this paper, mobility comprises both work-related migration towards sources of employment as well as the 'normalization' of mobility as a job requirement (Kesselring, 2015).
In order to secure a sufficient supply of skilled workers for their economies, and sometimes -as in the case of Germany -in response to ongoing demographic change, many industrial countries have implemented skills-based migration policies. Cities and regions, often adopting neoliberal policy agendas and following a rationale of competition, are also engaging in this ʻwar for talentʼ (Zenker, Eggers, & Farsky, 2013). Whereas economic development policies throughout the 1990s aimed mostly at attracting companies and capital, since the 2000s the focus has shifted towards attracting human capital or alternatively skilled professionals, the creative class, knowledge workers or simply talent (Hansen & Niedomysl, 2009;Musterd & Murie, 2010;Storper & Manville, 2006).
The questions about why people move to specific cities and regions and why some economies perform more successfully than others have since been addressed exhaustively within the spatial sciences, particularly in economic geography (e.g., Chen & Rosenthal, 2008;Storper & Scott, 2009). While this research analyses the complex relationship among different contextual factors such as institutions, human capital, path dependencies and amenities (Chen & Rosenthal, 2008;Clark, Lloyd, Wong, & Jain, 2002;Storper, 2010;Storper & Scott, 2009), there is little knowledge about the individual decision-making practices and further mechanisms that shape the work-related mobility of skilled professionals (Brown, 2015).
Our paper thus focuses on the question of why skilled professionals eventually move to specific places for work-or career-related reasons and what influences their decision-making in favour of a particular city. In our analysis, we focus on the role of connections, which comprise social networks and go beyond these in a broader definition and also comprise a set of further ties that existed between individuals and places before the move.
We draw from and compare two case studies on temporary work-related mobility of middle-class professionals in two large German cities. The first examines skilled professionals who moved to Frankfurt from within Germany on a temporary basis due to project work or other fixed-term employment conditions. The second focuses on high-skilled international migrants who moved to Hamburg for a period of a few years, mostly with the intention of advancing their careers. Both studies use similar conceptual approaches to analyse work-related mobility, and both employ qualitative research designs, thus emphasizing the agency of these mobile professionals.
The paper begins with a brief review of work examining the spatial variations in the flow of skilled professionals. It then turns to studies that investigate the decision-making practices of work-related mobility and what it tells us about the role of connections to particular locations. As both strands of research have remained largely separate, our aim is to bring them together. We then provide background on the two research projects. The following section comprises the analysis of our empirical data. First, it explores the motivations behind the work-related mobility; and second, it focuses on the role of existing ties and other connections to place. We conclude by arguing that the role of these ties during the decision-making process of mobile professionals with regard to particular locations has not yet been fully acknowledged. We argue that urban policy approaches merely aiming at attracting this sought-after group fail to understand the underlying mechanisms behind their locational choices.

CITIES AND SKILLED WORKERS
The question about why and how cities attract new inhabitants has received wide scholarly attention. Some studies have highlighted the role of specific location factors, which may influence the decisions of mobile professionals for selecting a specific city (e.g., Storper & Manville, 2006;Turok, 2004).
Regional StudieS, Regional Science Hard factors typically include local and regional economic structures that shape labour markets and job opportunities. Most literature continues to view hard factors as the most significant influence on individual decision-making (Chen & Rosenthal, 2008;Martin-Brelot, Grossetti, Eckert, Gritsai, & Kovács, 2010;Musterd & Gritsai, 2013).
In contrast, soft factors are more place-specific and can potentially be influenced more by local policy makers, but they are also more elusive. A prominent example of stressing soft factors is the work on the creative class (Florida, 2002) and other skills-based approaches. This literature points towards the role of overall tolerance towards newcomers, as well as diversity, for example (Florida, 2002). While these arguments have gained broad attention since the 1990s and have clearly had an impact on urban and economic development policies on city and regional levels, they have also received strong criticism for a variety of reasons (e.g., Evans, 2009;Kloosterman, 2013;Krätke, 2010;Markusen, 2006;Peck, 2005;Scott, 2006;Storper & Manville, 2006). In the European context, soft factors need to be regarded with particular caution because the creative class in Europe is less mobile than its counterpart in the United States (Hansen & Niedomysl, 2009;Martin-Brelot et al., 2010;Musterd & Gritsai, 2013;Pareja-Eastaway, Bontje, & d'Ovidio, 2010, p. 193). Studies on high-skilled migrants within the European Union show that they prefer locations that allow quick access to family and friends (Andreotti, Le Galès, & Moreno Fuentes, 2013;Favell, 2008;Viry & Kaufmann, 2015). Through their personal trajectories, high-skilled migrants are not free-floating subjects drifting towards the best opportunities but rather remain attached to specific places.
Mobility decisions are never simply outcomes of rational choice decision-making for or against specific places; they also need to be understood as ʻreactions to pressure and social or economic constraintsʼ (Kesselring, 2008, p. 85). Mobility practices are thus embedded in economic, spatial and political structures, cultural norms and values, discursively generated knowledge, social structures, and further social practices and processes (Manderscheid, 2009). An example of these contextual frames is that, in many cases, moves are determined by limited options, sometimes turning them into 'no-choice decisions' (Martin-Brelot et al., 2010, p. 865). We argue that more attention should be given to the process and scopes of decision-making. In an extended conceptualization of networks and based on our empirical work, for the purpose of this paper we focus on the role of ties and connections that relate people to place(s).

PLACE AND CONNECTIONS
Connections and place ties are, of course, related to social networks; and in addition to hard and soft factors, they have in fact been stressed as additional location factors by recent research in European cities (Musterd & Gritsai, 2013). We agree and see potential for further research. Social networks have been studied widely, and their function of relating people has been highlighted as a key feature of late modern societies. For a variety of reasons, work on transnational migration has taken particular interest in social networks' potential to connect people across geographical distances (Portes, 1995;Pries, 2010).
This view acknowledges that mobility and employment decisions, whether relating to internal or international migration, are always embedded in the individual's wider social context, including family, kinship, friendships and ethnic communities (Bunnell, Yea, Peake, Skelton, & Smith, 2012;Guarnizo & Smith, 1998;Haug, 2008;King, 2012).
With regard to their spatiality, social networks are often diffuse and the rise and widespread use of new telecommunication technologies and digital social media have arguably further undermined the significance of place (Wellman, 2001). However, Kesselring also notes that mobilities are ʻby no means undirected, that is chaotic, aimless or anomicʼ, but follow ʻa connective logic of integration into meaningful contexts and social bonds and interactionsʼ (Kesselring, 2015, p. 583). Social networks of highly skilled people have been described as more spatially disperse Regional StudieS, Regional Science because of former educational and career stages in various places (Ohnmacht, Frei, & Axhausen, 2008). Places are related to people and vice versa and are thus unequally important in people's lives (Nadler, 2014, p. 368).
For the group of middle-class mobile professionals studied here, professional networks are also of relevance because they provide the space for the exchange of information about employment opportunities, work conditions and other aspects related to work-related moves (e.g., Grabher, 2004). For the media sector, a typical way of recruiting project staff is to ask former project members to collaborate again (Apitzsch & Piotti, 2012). We thus argue that relevant weak ties to a place are established through not only different social networks but also professional contacts established throughout career trajectories (Granovetter, 1973). Having access to such networks is a resource that constitutes a specific form of social capital (Bian & Huang, 2015;Ong, 1999). We assume that highly skilled, highly mobile professionals can draw from a spatially dispersed web of social ties. Thus far, however, there is little knowledge on how these connections are used during the job search in other places.
While previous work-related mobility has often been associated with low-skilled and sometimes elite groups, the heterogeneous middle-class has arguably been ʻmobilizedʼ in recent decades. Our focus lies on the work-related mobility of high-skilled professionals from middle-class backgrounds, including both internal and international migration. This group has often remained ʻunder the radarʼ because of a stronger tendency to self-organize their mobility, as opposed to being sent as expatriates, for example, or being channelled into specific jobs through intermediaries (Ryan & Mulholland, 2014) and because they have not moved to the most global cities (Brown, 2015;Van Riemsdijk, 2014).

TWO CASE STUDIES ON MOBILE PROFESSIONALS
This article draws on empirical material from research projects about two distinct groups of mobile professionals in two large German cities, both of which are interested in the relationship between mobility and place. Both focus on the group of middle-class, skilled mobile professionals (e.g., Iredale, 2001;Van Riemsdijk, 2014).
The first study focuses on professionals in IT, science, consulting, media and the construction sector who moved from within Germany to Frankfurt for temporary work assignments. In this case, and enabled by shorter distances, about half of these professionals maintained dual-residence arrangements (e.g., Reuschke, 2010;Van der Klis & Karsten, 2009). The second study analyses the different modes of spatial incorporation of highly qualified international migrants who moved to Hamburg for a limited number of years (Plöger & Becker, 2015). Both studies examine forms of temporary high-skilled migration and comprise individuals working in professions that have become 'mobilized', such as those in IT, engineering, corporate services or academia. They thus address ʻthe mobilisation and activation of working subjects and their human capitalʼ ( Jeanes, Loacker, Śliwa, & Weiskopf, 2015, p. 705).
Since the 2000s, Germany, in an attempt to attract skilled workers, reformed its migration policies, gradually facilitating the labour market access of non-European Union professionals (Organisation for Economic Co-operation and Development (OECD), 2013). Nonetheless, the national and supra-national migration regimes constituted bureaucratic barriers for the international migrants in Hamburg before the actual move. In their case, we also assume that the move was generally better planned due to larger geographical and cultural distances to the receiving context. In the Frankfurt study, moves generally occurred on a rather short-term basis and were thus less planned.
In Frankfurt, the employment conditions comprised a range of constellations including fixedterm contracts or projects, secondment, transfers within a company and agency work, all of which Regional StudieS, Regional Science arguably represent late-capitalist outcomes of the ongoing transformation of work (Grabher, 2004;Rosa, 2003) and have been referred to as portfolio careers.
Hamburg and Frankfurt are among the most economically dynamic cities in Germany and function as centres of relatively affluent agglomerations. Frankfurt is Germany's fifth largest city (700,000 inhabitants) and the main centre of the fourth largest metropolitan area in terms of population, as well as Germany's most important -and a major European -financial centre. Hamburg (1.75 million inhabitants), Germany's second largest city and the core of the country's sixth largest metropolitan area, has a diversified economic structure and hosts the second largest European seaport. With their specific economic and institutional structures, these cities' labour markets attract a constant supply of skilled professionals from within and outside of their regions.
Both studies used similar designs with in-depth semi-structured interviews and do not claim to be representative. The qualitative approach enabled us to explore the nuances of individual decision-making processes of mobile professionals with regard to particular jobs and locations.
The sample selection of mobile professionals in both cities was achieved through the 'snowball' method based on personal networks, social media and Internet platforms (e.g., InterNations, Meetup), as well as through larger international companies and academic institutions.
The Hamburg case study (fieldwork in 2013) comprises 18 highly qualified international migrants from nine non-European Union countries. In Frankfurt (fieldwork in 2012/13), the sample consists of 21 professionals who have moved to the city from outside of the region yet within the country. Both studies have a slightly higher share of male respondents. The age range varies from approximately 25 to 60 years in both studies.
The interviews in both case studies were transcribed and coded using the MaxQDA software. For the purpose of this paper, two main codes were chosen, which are regarded as constitutive for locational choices: 'job search' and 'previous connection to place'. Unlike the Hamburg interviews, those in Frankfurt were conducted in German. For the purpose of this paper, any direct quotations from the latter were translated into English by the authors.

WORK-RELATED MOBILITY: LIMITED SCOPE FOR DECISION-MAKING?
Drawing on our empirical material, we now look at the mechanisms of how the mobile professionals in Hamburg and Frankfurt found their new jobs. First, we look at the decision-making in favour of a job and, subsequently, a location. Second, we specifically focus on the role of ties and connections during this process.
For both internal and international migrants, the decision to relocate was primarily work related. Criteria such as finding an attractive job, salaries or career prospects were thus important during the job search.
During the job search, the respondents used a range of filtering mechanisms, both conscious and unconscious, to narrow down their selection. However, although personal preferences and opinions shape the eventual outcome, our findings suggest that they are usually subordinated to job-related factors. In addition to professional motivations, additional aspects were considered to be important. In particular, this includes social (e.g., relationship, family, friends) and, to a lesser degree, locational (e.g., attractiveness of a particular location) aspects. Max, a 35-year-old freelance management consultant, describes how these aspects are taken into consideration when he chooses between different project offers: I had to make a decision between a project in Stuttgart or Frankfurt. From an earlier short-term stay in Stuttgart, I knew that this was not an option. It's too far away and the mentality is different. And in Frankfurt, I knew people like a friend of mine who moved there seven years ago and a few more people. And it's not too far away.
Regional StudieS, Regional Science In his case, and similar to the findings of Musterd and Gritsai (2013) in a range of European cities, existing weak ties were also mentioned as having an influence on selecting the location.
Before applying to job vacancies, internal migrants in Frankfurt usually narrowed down their job search to particular regions, particularly those where they have lived before, where they have strong personal relationships and those near their current place of residence, particularly when it comprised a household or family. Internal migrants generally had sound first-hand knowledge of the selected regions and were thus able to provide informed opinions about specific location factors.
However, when making an actual decision, this knowledge often appeared of little relevance either because they simply had no other option or because job-related aspects such as the attractiveness of jobs in terms of career prospects or incomes overshadowed other aspects. Furthermore, when internal migrants were unable to realize their preferred locations, they eventually widened the scope of their search. Lucy (30 years old) had originally applied for a job in a company at a particular location. However, when this was unsuccessful, the only job offer she received and eventually accepted was a traineeship in the media sector at another location: Actually, I set my sights on Cologne [where the respondent went to university] and Hamburg [where the partner of the respondent lives]. The vacancy was actually in Düsseldorf, but I was asked if I would also move to Frankfurt. When you're in such a situation, you haven't got much choice. If you get ten rejections a day, one says, ok, I'm willing to move.
This account highlights the difficulties of labour market access for younger people after graduation from higher education (Rudolph, 2005;Ryan & Mulholland, 2014). Taking up the job offer and moving to Frankfurt essentially became a 'no-choice decision' (Martin-Brelot et al., 2010). In contrast to some of the literature suggesting that skilled people are able to choose between different options, this also serves as an illustration of the rather subordinated role of locational aspects during the job search.
Another group with little influence on the eventual location of their job comprised those professionals who were either sent away for project work by their company or as freelance consultants took up short-term projects for customers. The following quotation, again from Max, illustrates the fine line between independence and constraints with regard to job decisions when none of the available options is in the home region or otherwise a favourite: Well, in general, I am not able to select the location of my projects. It really depends on the customers. When I finished one of those projects, there was nothing in Hamburg. I then had the choice to say, I'll wait until a project appears in a city that I like, but this isn't something one should do. […] I can't afford to wait.
Scope of deciding the eventual place of work is further limited when companies assign project work to their staff as illustrated by the following quotation from Norbert, a 44-year-old employee in an IT consultancy firm: My company can basically place me according to demand. Of course, I have a bit of a say in it, but when projects are in Munich or Stuttgart, Cologne or Frankfurt, or Madrid and everything fits, well, then you are there. This is very typical in the consultancy sector.
Against a backdrop of dependency on economic cycles and flexible working arrangements, several interviewees in Frankfurt referred to the structure of the labour market as a decisive factor for their location decisions. For some, the decision to take a job in Frankfurt was also related to the expectation that it would facilitate finding a follow-up job or project due to the strong metropolitan economy and dynamic labour market. In summary, for many mobile professionals, the most attractive and, in many cases, only offer came from Frankfurt.

Regional StudieS, Regional Science
Due to the often short-or medium-term duration of jobs, several interviewees in Frankfurt -particularly the 'project workers' -used dual-residence arrangements. These are anchored by a 'normal' place of residence where respondents would spend the weekends with family and friends and an additional place focused mostly on work (e.g., Van der Klis & Karsten, 2009). In such constellations, the move is almost entirely based on the employment opportunities, while soft location factors are practically irrelevant.
In both studies, work-related motives clearly dominated the decision-making. However, whereas the specific job opportunity generally over-shadowed further motives in Frankfurt, additional motives were more common for the international mobile professionals. Several interviewees in Hamburg also associated the move with the desire to move with or closer to a partner, which the literature refers to as 'trailing spouses' (Van der Klis & Mulder, 2008) or 'tied movers' (Mincer, 1978). In a few cases, respondents also alluded to a general desire for change and for the experience of living in different places abroad.
For the most part, the locational choice of the interviewees in Hamburg was driven by the desire to work in a specific sector or for a particular employer, such as a prestigious international company or research institution, irrespective of the geographical location. Unlike the internal migrants, this group usually had no or only limited knowledge of the city and its specific location factors and amenities. First-hand experience of the city was the exception and -if any -was gained during the initial job interview, when a partner already lived there or when the interviewee had biographical links to Germany in general. The following quotation from Lew, a male, 25-yearold Armenian-born Russian physicist in Hamburg, illustrates not only the limited significance of place-specific factors but also the incorporation of mobility demands throughout mobilized professions, such as research and academia (e.g., Bauder, 2012;Mahroum, 2000): I don't care about the city. When we go somewhere, we don't take into account where it will be, because it's more about the job. If you have a job, normal financial support and you see that it will lead to something, then you don't care about the place where you live.
Rather than the city, the country was in many cases a more important scale of reference for the international group. In several cases, job hunting was restricted to Germany, which is associated with hosting leading international companies and research institutions, as well as being a favourable environment for new technologies and innovation. For those in a research or higher education institution, access to federal subsidies or grants was sometimes also important, thus constituting a national-level locational factor. In contrast to the internal migrants, any narrowing of the search radius occurred at a higher scale and was influenced by specific national or supranational location factors, particularly migration policies. Again Lew: There are difficult problems for people who are going to the U.S. because, if you go there it's difficult for your family to join you. […] I mean, for us Russians, the rules are very strict. […] So, for the U.S. you have to wait around three months and you don't know whether we will get the visa or not […] and if you're rejected once it's like forever.
However, the eventual location within that frame is then determined by the availability of companies and jobs that are both attractive and accessible to skilled professionals from abroad. Both Hamburg and Frankfurt, with their diverse and dynamic economic base and institutional landscape, thus become viable options for these groups.
The following section examines how different forms of connections mediate the migration outcome.
Regional StudieS, Regional Science

THE ROLE OF PERSONAL NETWORKS AND CONNECTIONS
We were interested in not only where and why mobile professionals move to new employment but also how. In this paper, we focus on the role of a range of connections that often constitute a mechanism that channels people to particular jobs and, thus, places. In addition to hard and soft location factors, such connections constitute a third type of location factor. Our research reflects the relevance of these network factors (Musterd & Gritsai, 2013). Most respondents mentioned some form of connection, for example, either towards the respective employer (or customer) in Frankfurt or Hamburg or the respective city (internal migrants) or country (international migrants). Figure 1 provides an overview of these connections and differentiates between internal and international professionals. Based on our empirical material, we distinguish between three categories, professional, social and cultural, which we refer to below.
Professional connections comprise links between institutions such as companies or research institutes, as well as those between individuals. Whereas the former are shaped by existing networks within an institution or company (e.g., office network, business partners and exchange programmes), they are tied to individuals and expressed by their professional networks in the latter case. Many respondents mentioned previous personal and institutional contacts, which points towards the level of international connectedness of major employers in both Hamburg and Frankfurt.
Respondents were able to use institutional connections in the form of fixed-term secondments and transfers, training, scholarship programmes or new projects. Staff transfers or secondments, which can involve sending staff to another office within the company or sending experts for project work to other companies (Kels, 2008), were particularly relevant in the Frankfurt case study. The latter were often highly specialized experts (e.g., software development) employed by consulting firms.  Figure 1. Previous connections to the city.
Regional StudieS, Regional Science In both cases, these professionals could draw on their professional connections, including former fellow students, work colleagues or supervisors, during their job search, for example, some of these connections made recommendations or noted interesting vacancies. This sort of support is usually detached from specific locations. In the Frankfurt study, co-workers on projects recommended each other for further project assignments. Max again: Interestingly, most of the projects that I have done in the past were through people I personally know and had worked with together. They called and said: 'Listen, we got something, are you interested.' This relevance of such connections within specific professional networks as a characteristic way of recruiting project staff has also been highlighted elsewhere (Apitzsch & Piotti, 2012;Grabher, 2004).
For the international sample, recommendations by former colleagues or other persons in their professional networks were less significant, while formalized and institutional links were more important. Those in academia often benefited from access to the extended network of higher-ranking persons, such as university professors. For this group, links to supervisors and colleagues were established through scholarship programmes, summer schools, international degree courses or collaborations between universities. Yaroslav, a 25-year-old doctoral student from Ukraine mentioned how his current job at a renowned research institute in Hamburg resulted from a previous stay there while being a student: I used to be a summer student here, in DESY, before.
[…] It was the first time, when I arrived here to Hamburg. After that my boss told me, that it would not be bad if you find a PhD position here.
Professional networks thus often develop through institutionalized programmes or previous stays, which can significantly influence future job-related moves.
Social connections are identical with what is commonly referred to as social networks. The social networks of highly mobile people are characterized by consisting of weak ties and being spatially dispersed although mostly located in larger cities (Ohnmacht et al., 2008;Ong, 1999).
Several interviewees in Frankfurt mentioned that they already had some personal contacts, friends or extended family there prior to moving. Where available, these were reactivated or intensified once the decision in favour of a job there was made. Only a few professionals in Frankfurt moved with or to a partner who already lived there. In one case, a respondent mentioned how she felt a 'connection' to Frankfurt because her parents had lived there before she was born.
Contacts in situ are not only 'activated' during the job search but also access localized social networks or assist during the move, for example, by identifying suitable neighbourhoods or finding an apartment. The localized social network comprises work-related as well as personal contacts, with both spheres often overlapping. The following quotation from Gudrun, a 38-yearold architect in the public sector, illustrates how she was able to draw on her loose network of acquaintances during different stays throughout her career: Everywhere you go and start building up something, this leaves traces. So your network grows and there are always one or two people you know. Of course, the quality [of these contacts] is not like before, they are looser, but they have also been beneficial, for instance, getting this apartment. And if this doesn't work out here and I need to move somewhere else, then maybe I know someone there […].
Weak ties that are spread out enable her to cope with the normalization of mobility as a job requirement in certain economic sectors and professions (Kesselring, 2015;Ong, 2007). Moreover, mobility provides further employment options, thus contributing to the making of portfolio careers (Boltanski & Chiapello, 2007). Despite that, nearly all interviewees from the Frankfurt Regional StudieS, Regional Science study describe maintaining friendships at former places of residence and establishing new ones at their current location as a main problem associated with work-related mobility. The difficulty of forming new social ties with 'locals' after moving has been acknowledged by Ryan (2015). In these cases, a growing network of weak ties involves a weakening of strong ties.
For the international group, the move with or to a partner was slightly more significant than in Frankfurt, either when following a German partner or when moving with or to a foreign spouse. Shankar, a 38-year-old engineer from India, who moved to Hamburg from the United States explained how, although the move had to fit his career, another key objective was to reduce the distance to his partner and live together in one place: I was in a relationship. She was originally from Germany. So one of us had to make a decision or compromise.
[…] So eventually, I took the decision to come here. When I was in Texas last year, I was looking for a job somewhere in Germany and of course it had to fit my profile.
Cultural connections are not between institutions, individuals or groups but rather between individuals and specific places and can result from previous visits or, as in the case of international migrants, a proximity to German culture, such as the knowledge of the language. These links are generally more ephemeral and subordinated to professional and social connections.
Within the international sample, several respondents mentioned links to German culture. Examples include having German ancestors, having attended (private) German schools or having taken part in exchange programmes with Germany while at school or university. Arif, a 42-yearold Turkish interviewee in Hamburg, described how he had become somewhat familiar with German culture while at school because of German teachers at his school and with a specific region through a pen pal: There were the teachers who came to Turkey to work as German teachers for a few years. I also had a pen pal who came from the Altes Land [a region south of Hamburg]. This way, I got used to the German culture a bit and also learnt a little about German families. This quotation indicates that it is not so much a question of precise knowledge of a country but rather about an ease of access and positive associations as a basis for cultural ties. Being familiar with German culture in one way or another can therefore contribute to considering the country as destination for a job-related move.
In comparison with their international counterparts in Hamburg, the professionals in Frankfurt had, of course, a greater cultural proximity. They also had far larger knowledge about the city, in many cases based on having visited the city before. The interviewees made nuanced differentiations regarding their regional preferences and their opinions about Frankfurt. Interestingly, this knowledge had only a minor influence on the location decision, which was, as laid out above, clearly driven by the particular job offer. It did, however, affect which cities and regions the respondents considered during their job search. When these were associated with negative connotations, perceived as too distant, both geographically and culturally, they were excluded from the search. The role of positive or negative association with places during the location decision is illustrated by the following quotation from Andreas, a 36-year-old engineer: I was born in Hesse [the federal state surrounding Frankfurt]. Hesse was thus my primary focus. I also considered North Rhine-Westphalia, where I have lived two years. I also considered Lower Saxony. I have no real connection to East Germany. I had a project there for half a year, but to me, because of the mentality, I found it difficult to deal with the people there.
In summary, our research shows how different types of connections can either encourage or constrain the decision for or against specific places. Professional connections are clearly the most Regional StudieS, Regional Science important, thus matching the dominance of work-or career-related motives. These connections often mirror the relative position and functions of a city within the global urban system because they host those employers that both require and attract skilled professionals. However, social and cultural connections can also become relevant. In both samples, spouses or partners or the household context influenced the decision, but while the internal migrants often engaged in dual-residence arrangements, those in Hamburg who migrated over larger distances were more likely to either follow or bring a partner.

DISCUSSION AND CONCLUSIONS
Recent studies on location decisions, while emphasizing the ongoing relevance of hard location factors, such as economic structure and the labour market, have highlighted the role of networks, at least for skilled professionals (Musterd & Gritsai, 2013). Individual connections and ties constitute an understudied yet relevant factor that contributes to the decision-making in favour of specific locations and merits further attention. In our qualitative analysis of the work-related mobility of mobile professionals in Frankfurt and Hamburg, we have thus focused on the role of these connections and ties.
The significance of job-related motives, particularly evident in the Frankfurt case, supports the dominance of hard over soft location factors (e.g., Hansen & Niedomysl, 2009;Martin-Brelot et al., 2010;Storper & Scott, 2009). However, established social contacts and family connections are also relevant, e.g., when limiting the job search to places in spatial proximity or to those where anchor points already exist or when following or moving with a partner. In the Frankfurt case, the specific work conditions, such as short-term contracts, project work or assignments, also contributed to dual-residence arrangements, whereby giving up the original residence was not considered an option. While internal migrants do at least consider local location factors, the scale moves up to the national level for international migrants. The latter have significantly less information and knowledge about the location to which they eventually move. Their scale of reference is more often the national level, as evidenced when referring to desirable employers or institutions in Germany.
The scopes of individual decision-making are often limited. Decisions are predominantly not made for a specific location but rather for a specific job, and the mobile professionals in our studies are clearly not free floating (Colic-Peisker, 2010). Thus, when studying mobility and location decisions, the scopes of decision-making and the associated search criteria are crucial. Both studies do, however, show how job-related mobility has become an essential characteristic of their career trajectories, at least for the mobilized professions represented here. In both cases, mobile professionals acknowledge job-related mobility as a taken-for-granted requirement of their professions.
Our research reveals the relevance of professional, social and cultural connections for both internal and international migrants. During their careers, mobile professionals build up a network of weak ties at different locations. Sometimes they overlap, thus becoming 'thicker' and more powerful. Professional networks and institutional connections inform about job and project vacancies, ask for collaborations and facilitate getting the job or project. While individual connections such as professional networks are more relevant for internal migrants, institutional connections such as linkages between organizations or formalized programmes are more significant for the international group. After getting the job, social contacts at the destination provide access to these locations and can assist in everyday matters. Job-related mobility and mobile careers thus contribute to the accumulation of social, symbolic and economic capital. The influence of established personal networks on the process of decision-making is thus twofold: first, professional networks assist in finding the job, and second, social contacts become crucial when choosing