Bowling Alone and Working Together ? Social Capital at Work

Robert Putnam’s books Making Democracy Work (1993) and Bowling Alone (2000) have been influential in promoting the relevance of networks and norms of civic engagement. Within this tradition of research, civil society represents the fundamental sphere for creating social capital and thereby a high level of generalized trust with implications for democracy (Rothstein & Stolle, 2008; Rothstein & Uslaner, 2003; Wollebæk & Strømsnes, 2008). The empirical evidence of the suggested link between participation in voluntary organizations and level of trust, however, is neither wholly nor partly supported (Wollebæk & Segaard, 2012, p. 40). Despite all the interest that has been shown in social capital and the political and social merit many see in the perspective, the empirical data are not as obliging; we are not finding what we want to find. With regard to Norway, this cannot be explained by a generally low level of trust. On the contrary – similar to the rest of Scandinavia, Norway is depicted as a high-trust society, which means that people’s confidence in each other and in society’s institutions is very high. This is considered crucial because generalized trust – that is, trust in people with whom one is not familiar (Wollebæk, 2016) – is believed to motivate people to take an active part in democracy and in society more generally. Generalized trust is therefore a community cornerstone in terms of cohesion, coordination and solidarity. But if voluntary organizations do not appear to play the role expected of them theoretically, it is useful to ask whether other institutions could be pivotal in this respect. In the search for candidates, the labour market stands out as a particularly evident arena. In Norway, the substantive and socio-psychological significance of the labour market is understood in terms of the Norwegian model, which, briefly put, represents a compromise in the form of reciprocal bonds of trust between the government, employer and employee organizations and the public – all of which are expected to provide security and equal opportunity to every member of the population. The objective of this chapter is to investigate the significance of the workplace as an arena for generating social capital and thereby generalized trust. A workplace comprises a small system of people who meet on a regular basis and, to a certain extent, share a common goal. Consequently, the question in this chapter is whether Putnam’s perspectives on the importance of face-to-face meetings in civil society should be expanded to include face-to face-meetings in the workplace. We ask: ‘To what extent can the workplace be understood as a sphere for creating generalized trust?’ This chapter is mainly empirical, but there are theoretical implications of the discussion to come. The empirical part of the chapter is based on interviews of young retail apprentices working on the shop floor – frequently depicted as atomized and alienated. As early

as 1951, C. W. Mills described the working conditions of sales girls with a dismal undertone in terms of estrangement (Mills, 2002). Considering a substantial proportion of the young employed in the service sectors whose retail experience is provisional in character, supporting the development of trust and networks in the workplace might suggest a specific challenge. Furthermore, the retail sector is characterized by low union density and thus weakly established institutions of employee participation and collective commitment. Consequently, this constitutes an extreme case for inquiring into whether generalized trust and subsequent participation in democratic processes can be created in the workplace.
The chapter is organized as follows. First, we give a brief presentation of social capital within civil society and working life. Subsequently, the data and methods underpinning the analysis are presented. Thereafter, we analyze the ways in which trust and networks are generated in retail work. Finally, we discuss the shop floor as a work-based community before presenting our conclusions and implications for further research.

Generalized Trust and the Workplace
According to Putnam (2000), two factors are crucial for understanding the significance of social capital: network, which is an individual trait but a collective phenomenon in virtue of being a characteristic of a society, and trust, a cultural phenomenon that regulates norms of cooperation. Research inspired by Putnam discusses mainly how, in various ways, individuals and collectives generate and challenge the link between trust and networks.
Social capital is relevant to understand processes of generating democratic engagement. Putnam (2000) assumes that citizens have arenas where they can meet in person and from there build a mutually binding sense of trust. Putnam's understanding is based on the assumption that the network produces and is regulated by mutual norms, a process which in turn is believed to lead to value-based communities. Putnam (1995, pp. 664-665) defines social capital as 'networks, norms and trust that enable participants to act together more effectively to pursue shared objectives'. Various forms of social capital have been disputed, but the literature seems to agree on three main aspects: bonding, which links people with a sense of common identity such as family, friends and close colleagues; bridging, which stretches beyond a shared sense of identity, that is, people in other organizations; and linkages, which connect people across social strata.
Discussing the history of the concept, Wollebaek and Seegård (2011) argue that social capital, on one hand, is a label of an individual's resources in terms of networking and trust and, on the other, a description of those collective relationships. With this understanding, they object to Bourdieu, who conceives of social capital as a mere individual resource. According to Wollebaek and Seegård (2011), the Bourdieusian perspective diminishes our possibility to understand the consequence of social capital for collective action at the aggregated level (p. 29). They therefore argue in favour of a combination of social trust and networking, which, they claim, allows for non-coercive collective action.
Putnam's point of departure entails different people who meet face to face. Interaction is a precondition for generating networks, which in practice implies that individuals practise a type of trust, which in turn they develop into a more generalized trust. According to Wollebaek and Seegård (2011, p. 38) 'non-governmental organizations are the main arenas where such interactions take place: here we meet people who are different from ourselves, and try to achieve something together with them'. What matters is the link between individuals in the sense that an actor can be the bearer of more or less social capital -to be a description of a given collective.1 Scandinavian researchers have pointed to dissimilarities between the United States and the Nordic countries with regard to the welfare state and general trust in institutions. According to Rothstein and Uslaner (2006), social inequality is a crucial factor in understanding variations in social trust. Rothstein and Stolle (2003) have also noted that strong welfare institutions are crucial for understanding the high level of trust in socially supportive institutions in Scandinavia. For this kind of trust, Rothstein and Stolle (2008) have coined the term 'institutional social capital', implying that the formation of generalized trust is embedded in the structure and characteristics of political institutions. In contrast to Putnam, who views regular social interaction through face-to-face meetings in voluntary associations as the prime mechanism for the generation of social capital, Rothstein and Stolle emphasize that institutions and policies create, channel and influence social capital. Thus, for social capital to flourish, it needs to be embedded in the political context and the formal political and legal institutions. In other words, while Putnam highlights the horizontal dimension through face-to-face meetings, Rothstein and Stolle draw attention to the vertical dimension and the importance of institutions as a precondition for generalized trust.
Our ambition here is to pursue the Scandinavian perspective, where institutions are given attention. Our approach involves examining the significance of working life as an arena entailing solidarity, coordination and participation. It is therefore not unreasonable to assume that participation in the workplace can have a very important role to play alongside the voluntary sector if we want to understand why Scandinavia comprises nations with particularly high levels of confidence or trust.
That a positive correlation exists between participation in various arenas and democratic involvement is nothing new. A standard reference is Carole Pateman's Participation and Democratic Theory (1970), where the author justifies her perspective by referring to John Stuart Mill. The 'democratic principle', she says, must be seen in connection with spheres initially located outside the purely political sphere. Particularly interesting in this context is Pateman's reference to the labour market, which -since it is an institutionalization of hierarchy and inequality -is a particularly important arena for political involvement (pp. 42-43). This is because, first of all, democratic institutions alone cannot give citizens the necessary training in democracy nor the necessary 'individual attitudes and psychological qualities' (ibid.). While volunteer researchers focus exclusively on voluntary organizations as schools in democracy, we can make use of Pateman to underpin our hypothesis that schools in democracy are also to be found in the business sector.
Moreover, Pateman believes that a business or commercial enterprise can be understood as a political system in 'its own right' (ibid.), providing opportunities for active participation outside the purely political system. Now it is nothing new that companies can be understood as political systems. Several empirical studies have explored democratic organizations with employee representation on boards, union density and the activity of personnel representatives (e.g. Hagen, 2014;Trygstad, 2013). But these are studies where the objectives are somewhat different from ours. This chapter is concerned with the role of work in democracy, but not in the sense of employee participation and involvement (Levin et al., 2012).
As suggested, little scholarly attention has been devoted to the ways in which the workplace may generate social capital.2 However, an important initiative has already been taken. Researchers at the John F. Kennedy School of Government, Harvard University, have established an ongoing network, arranged seminars, a webpage and a research programme emphasizing the relationship between work and social capital.3 The main argument is that the workplace generates social capital in three broad ways. First, the job is where people build trust relationships based on mutual assistance.
The workplace is an especially important source of social capital in an increasingly fragmented and diverse society. As growing numbers of people lack the comfort of a nuclear family, a religious or fraternal community, or even a tight-knit neighbourhood, the job site has become a place to build stable, caring, long-term relationships.… They are more diverse, on average, than our neighbourhoods, houses of worship, and voluntary organizations. It is on the job that one is most likely to encounter, and work closely with, someone of a different race, ethnicity, religious affiliation, sexual orientation, social class, political ideology, or regional heritage. In short, American workplaces represent the ripest venue for bridging social capital. (A Civic Nation at Risk)4 2 One researcher who has mentioned the link between employment and social capital, however, is Gudmund Hernes (2006). In Den norske mikromodellen. Virksomhetsstyring, partssamarbeid og sosial kapital [The Norwegian micro model. Corporate governance, collaboration and social capital], Hernes makes use of the term but not in a Putnamian way. 3 http://wcfia.harvard.edu/files/wcfia/files/762_bettertogether.pdf 4 http://docplayer.net/42998702-Introduction-a-civic-nation-at-risk.html Second, workplaces act as recruiting grounds for individuals and community organizations outside the office and factory walls. Third, employers contribute as organizations by sponsoring volunteer groups. The main focus of this chapter is the workplace as an arena for generating social capital. We deploy an institutional perspective, understanding an institution as a set of rules or norms regulating the behaviour of individuals as well as of organizations and corporate actors. It is a framework for action with relatively high stability (Thelen, 1999in Engelstad et al., 2017.

Retail Work -An Extreme Case
Apprentices in the retail trade are a particularly interesting group if we want to learn how the workplace acts as an arena for fostering a sense of generalized trust. The retail sector, which constitutes the highest number of jobs after the public sector, is to many young people the first labour market entry point. The retail sector therefore has the role of socializing young labour market entrants into the adult responsibilities, norms, rights and duties of working life. And insofar as the retail trade is a significant employer in the modern economy, the retail sector is interesting in itself. At the same time, it serves as an extreme case if we look at conditions in the sector. The retail sector is characterized by high turnover rates and the extensive use of young people in part-time positions (Høst, Seland, & Skålholt, 2013;Høst & Reegård, 2015). In the Norwegian labour market, approximately 162,000 report having another main activity besides traditional wage labour (primarily education among young persons). Onethird of these are to be found in the retail sector. Hence, the retail sector is by far the largest employer of this type of labour (Jordfald & Mühlbradt, 2015). The average sales assistant is therefore a young person who views his/her retail job as a stopgap on the way to graduate destinations.
It is therefore not unreasonable to assume initially that these are organizational factors which discourage a collective orientation among workers. That notwithstanding, other factors point in the opposite direction, with the collective assuming importance.
A key aspect is related to the young age of the workers in the retail trade, implying substantial homogeneity along one significant dimension: the young employees, it is probably safe to say, have certain common interests and a shared sense of identity, albeit not in an institutionalized way such as in a labour union. Another shared attribute is that a job in the retail sector is very likely the first taste many of these young people have of the labour market. It would therefore be reasonable to assume the existence of informal but coordinating norms of cooperation in the individual workplace. These may involve the purely practical performance of the work but also be about increasing sales and dealing with customers. In brief, it does not necessarily follow from the low rate of unionization that a sense of community is absent. What interests us is whether we can identify some form of collective identity -a sense of 'us' -within the main structures of the sector.
If we raise our eyes a little, the question is whether the labour market, particularly work in the retail sector in Norway, differs significantly from the characteristic features of the voluntary sector. Much of the research on changes in the labour market paints a picture of ongoing brutalization and individualization in society at large and the workplace in particular. One example is the book Organisasjonsformer. Kontinuitet eller forandring? [Organizations. Continuity or Change?] (Skorstad, 2002). The workplace is increasingly fashioned according to a techno-economic ideal, Skorstad contends, where a sense of the collective is weak. The crucial factor is the new perception of competition and market customization, where employers are much more likely to define the terms of interaction than before. If Skorstad is right, the social capital-generating role of organizations is under threat.
Despite the eroding tendencies in modern working life, there is, as we have mentioned, still reason to look at the potential of the collective community -not only by virtue of the level of unionization but also in terms of identity communities and, not least, because politicians often see the labour market as a key arena of integration and thereby self-realization and social recognition. In other words, work is a major source of trust and social networking. The rationale is partly because the work is a source of livelihood but, not least, that participation in employment is an admission ticket to the social community.

Data and Methods
We conducted interviews with managers and apprentices in nine retail shops in Norway. The apprentices were enrolled in the sales vocational education and training programme. This entails two years of school-based learning followed by two years of apprenticeship in a training company, leading to a trade certificate (International Standard Classification of Education [ISCED] level 3).
A qualitative approach was chosen to gain insight into the work and learning environment based on the managers' and apprentices' subjective frames of reference. Detailed data were collected in nine shops with apprenticeships. The shop selection criteria were twofold. First, pre-interviews were conducted with three of the apprentices as part of an ongoing qualitative longitudinal study on service sector vocational education and training (VET) in Norway. Three of these shops were where these apprentices had commenced apprenticeships. Additional apprentices and their managers were interviewed to ensure a more robust empirical foundation. The selection criteria for these additional shops were chosen to encompass variation in the type of retail being considered subsector and to provide information about the different products on sale. The shop types with their apprentices and managers are presented in Table 12.1. Note: Two of the managers declined to participate in the study.
The recruitment of interviewees proved difficult. Context characteristics of the shops were collected through limited observation. This provided valuable insight into the nature of the apprentices' responsibilities, workload, customer/team interaction and atmosphere in the shop. The shops represented a variety of retail sub-sectors, but there were great similarities regarding the apprentices' task characteristics, the shops' organizational structures and the degree of guidance and autonomy available to the apprentices. The research sample comprised 11 apprentices (six men and five women) and seven managers (three men and four women). Individual semi-structured in-depth interviews were conducted with each of the apprentices and managers during the fall of 2013. In the interviews with the managers, attention was on motivations for assuming the responsibility for apprentices, how learning and work were organized and how they evaluated the apprentices' progress. The topics covered in the apprentice interviews were how they perceived and coped with the demands of work, engagement in different tasks, opinion of the learning progress and self-perception of becoming and being a sales assistant. The interviews lasted between 45 and 60 minutes. All interviews were audio taped and transcribed verbatim. Following the observations and interviews in each shop, immediate impressions were written down in the form of field notes. The data were analyzed by thematic text-close coding and the clustering of data in terms of key topics in identifying patterns within the data. All participants were given pseudonyms, and the shops have been anonymized. The authors have translated all quotations.
In the subsequent sections, we analyze the data and the potential for building work-based social capital within the retail sector based on the analytical framework presented.

Generating Trust in Retail Work
One main aspect of social capital theory is that trustful, face-to-face encounters are transformed and generate generalized trust in institutions. While Pateman focused on the importance of participation, Rothstein and others highlight the importance of institutions (Rothstein & Stolle, 2008). Accordingly, the latter concerns the importance of variables at the macro level for generating trust and confidence at the micro level. We also indicate how trust is institutionalized but in a way that differs from the matters concerning Rothstein. We found that the young apprentices were socialized by and into trustful and supportive communities of practice. They were given a great deal of responsibility and thus became a part of the interdependent work team. However, not all employees in the retail sector enjoy high levels of generalized trust, network affiliation or coordination to solve their collective problems. The sector scores low on union density, and employees exhibit low collective action ability. They are therefore unable to transform trust in physical encounters into a resource for political mobilization in defence of their interests. In the following, we delve into the 'micro grammar' of the ways face-to-face interaction produces trust within different settings.

Networks in Retail Work
Data were collected in both urban and rural areas. In the most remote rural areas, where the video rental and candy shop and supermarket were located, the stores provided a neighbourhood meeting spot. This characteristic of retail work involving customer interaction increased the diversity of the work-related network. Taken together, these factors help prevent the decline of community networks as depicted in the literature. The empirical material shows young persons of different backgrounds who interact face to face to achieve common goals, allowing them to work as a team and members of a community that trust one another -a precondition for generating networks. Moreover, most of the apprentices reported a high level of satisfaction with the working environment -this despite the registration of sales figures, which potentially could increase competition and individualistic attitudes among the staff.
On the other hand, the sales assistants' temporal orientation to retail work and the employer could pose barriers to the formation of company loyalty and subjective identification with the occupation. The high turnover rate might slow group formation, referring to the membership solidarity of a group of people with common interests and collective strength with the power and resources to achieve collective benefits and the unionization of members. The data indicate that the apprentices made good friends with their colleagues and even the regular customers.
The sales vocational educational programme is weakly established. This means that 'the skilled retail assistant' does not constitute a widely recognized social category, nor do retail employers demand, or even value, the trade certificate. Thus, the labour market currency of the education is low. Yet, the apprentices displayed traces of personal and institutional trust. Despite facing uncertain employment prospects upon completion of training, the apprentices displayed an apparently confident and trusting attitude towards the future. This attitude has possible roots in two aspects of the apprentices' experience: first, they interpreted their difficulty in finding employment post-apprenticeship within a general discourse in which continuous transitions have become the norm. Their peers commonly postponed the transition to paid employment through part-time work or a gap year. Consequently, their accounts of 'try-out-to-find-out' are seemingly common and accepted. Second, their trustful attitudes might be explained by the current Norwegian economic climate, with high levels of social capital, a generous welfare state and generally low youth unemployment rates.
I mostly try to learn from myself. In case I need assistance, then I can ask anybody who works here, not only the manager. (Peter,19) It appears as if the store serves as an important institutional arena for generating trust. This finding corresponds with Putnam's initial findings regarding the importance of the bowling mall as a place where people meet face to face and through these meetings generate mutual trust. Our argument is that the same mechanism is at work in the store context. Salaman (1974) identifies two key features of what he refers to as 'occupational communities'. The first refers to readily observable phenomena such as the concrete group and the tendency to associate with, and make friends with, members of one's own occupation. The second refers to the process whereby practitioners subjectively perceive themselves in terms of their work role and internalize such a role and the norms and values associated with it to the extent that it becomes central to their self-understanding. The literature commonly presupposes a sense of subjective emotional engagement with work as a necessary condition for the development of occupational community, a sense of belonging and mutual trust (Sandiford & Seymour, 2007;Van Maanen & Barley, 1984). There are three main ways in which workplaces generate social capital: as communities, as networks and, finally, as community institutions.
One theoretical point of departure is that work-based social capital is not built through isolated tasks or customer interaction but is instead mediated by membership in so-called communities of practice. This concept resembles Putnam's (2000) bridging, a type of social capital stretching beyond a shared sense of identity -that is, the ways people of diverse backgrounds, identities and ethnicities come together and work to achieve a common goal. Based on Putnam's perspectives, we can further particularize trust to various relations in the retail trade. First, there is a type of trust between the young apprentices and others working in stores. It is usually young people who work part time and spend the majority of their time elsewhere, at high school or in higher education. For them, a job in a retail business is a temporary stopover on the road to their real goal. We describe the relationship between apprentices and the temporal situation with the help of Putnam's concept of bridging -meaning ties that connect across differences. Second, there will be relationships between apprentices and their managers. These we refer to as linkages. The third and final type of relationship is what Putnam refers to as bonding. Bonding relationships are based on equality. We use the concept for apprentices who find themselves in similar jobs and consequently may have strong reciprocal and community ties that effectively keep others on the outside.
The question is what the empirical data reveal about apprentices in retail work when we take a closer look based on the conceptual framework outlined above.

Bridging and Bonding: The Significance of Us and Them
It emerged from the interviews with our informants that they formed a community, which could be referred to as a mobilizational resource -that is, a resource one could call on if and when needed. Another interesting feature was that this type of solidarity upon which the sense of community rested was not dependent on whether the apprentices worked full time or had retail work as their main activity. Rather, it was the position of the person in the retail business. A particularly interesting example is the collegial commitment that existed between apprentices and relations between apprentices and incidental others who happened to stop by.
The occupational communities the apprentices were socialized by and into consisted largely of young, unskilled people working part time. They were typically still at school and viewed retail work as a temporary job on their way to graduate destinations. This implied a clear difference between the apprentices and the other employees -'us' versus 'them'. This boundary work is considered crucial to the development of a vocational sense of self. The apprentices were given work of great organizational value, which improved their sense of self-confidence. One of the apprentices, Stine, aged 18, described the learning process and in doing so made a distinction between the part-time staff and herself as a skilled retail assistant in the making: When I first started, I even asked the part-time staff, but they don't know very much, just a little bit. (Stine,18) In distinguishing herself from the part-time staff, Stine constituted a sense of self as learning and becoming skilled.
The suggestion that a significant barrier exists between 'us' and 'the others' is relevant for understanding the conditions of generalized trust in retail work. It is, as we have pointed out repeatedly, a feature of the industry that many have a rather casual connection to the labour market. Inasmuch as many have part-time jobs and will probably switch to a new job before long, it would not be surprising if the sense of community and mutual trust was small. The interviews show, however, that it is not that straightforward.
The interview with Stine revealed other interesting correlations in terms of the importance of possessing knowledge and an ability to reciprocate. It is useful to conceive knowledge as a scarce resource, something everyone is interested in acquiring in order to do a good job. In an untrammelled market transaction, knowledge as a resource is traded in money and has consequently become something one negotiates for and with. In a workplace community, like a store, other rules prevail -not least, it is precisely the fact that people share knowledge and some of the mutual trust within the system.
In the beginning, I felt like they knew a little bit more than I about the products and stuff, but now, it's more like me giving them assignments. (Stine,18) If we compare Stine's first and last statements in this respect, a most intriguing difference emerges. In both statements Stine reports having observed a difference between the apprentices and the others. At the same time, the attribution of an us and of the others is not essential for the sharing of knowledge. First it was she who learned; then it was she who taught. If we look at the statements as a whole, we see a type of reciprocity that is tied neither to a person nor to a position. In other words, there is a type of reciprocity that exists between positions within a wider system. Given our understanding of bridging, our findings indicate that a type of trust is created between actors in very different positions of seniority in retail work.
This finding is even more important if we acknowledge the presence of some very specific features characterizing the retail sector which one initially would not imagine as encouraging a sense of unity and solidarity. But as we have said, it would not be unreasonable to see in the fact that many are young -a driving force in the creation of a sense of community that traverses positions, whether the individual has envisioned a career in the retail sector or the job is a temporary stopgap on the road to something else. Some of the trainees said they had made new friends at work, regardless of position.
I'm friends with everybody here. We look out for each other. Like, when one of the others is stuck with a dissatisfied customer, I try to help out. (Peter,19) The sense of community can consequently be linked to performance at work, but at the same time friendship extends beyond the fellowship in the execution of specific tasks. And while we lack statements we can interpret as expressions of generalized trust, it is reasonable to assume that the kind of reciprocity we have discovered helps build trust. The ties we find among young persons in retail work are just one type of mutual trust, a form of mutuality Putnam wants to see in American society when he writes about the lonely bowler: a society that disintegrates as social venues disappear. What our interviews tell us is that for many young people a job as a sales assistant can provide a social meeting place where they get to know many others and are dependent on sharing knowledge, experiences and how to work together to perform in the best possible way. It is through this partnership that work in the retail sector, which initially may appear to be very individualistic, looks and is perceived as a system of solidarity and reciprocity

Linkages: Trust and Autonomy
In the previous section we looked at relations between employees, whether they were apprentices or were working in retail for a short period. The next question is how relations between the manager and workers affect levels of confidence and trust. While employee relations can be based to a large degree on a shared experience of youth and friendship, when we bring managers into the equation the situation changes. Similar to bridging, this is also a relationship based on difference, but when it comes to employed managers the main differences are linked to pay, authority/ power and prestige.
Nevertheless, it is reasonable to assume that relations between managers and staff can be particularly important for achieving an understanding of trust. Put differently, if one experiences good managers, it is likely to be generalized to others in central positions of power. Accordingly, these linkages may well prove to be sources of generalized trust. In studies of workers and working life, the degree of worker autonomy is always given particular attention. This is not surprising, given that autonomy usually requires trust on the part of the management.
The empirical finding is that retail managers give apprentices personal responsibilities from day one. The non-hierarchical organization of retail work where the manager works alongside the apprentice in the store might have helped the manager decide whether to give the apprentice a specific responsibility and independence from an early stage. The manager of the automotive retailer described how he threw the apprentice into the work head-first, giving him a great deal of responsibility: He [the apprentice] is thrown into it, with his own office and phone. Now, we are introducing a new car, and he's the one getting the responsibility. (Manager) The manager's trust in the apprentice is a deliberate strategy. He clearly considers it the most fruitful learning strategy. It was his understanding that throwing the apprentice into the work meant showing him trust, giving him responsibility and treating him with respect. Compared to the other informants, this retailer was characterized by few customers and low levels of stress. This may have eased his decision to delegate responsibility since he had the time to monitor the apprentice. Each car sales transaction represented much-needed revenue to the automotive retailer, and, conversely, each failed sales transaction had severe negative impact on the business. The apprentice's job was therefore of the utmost organizational significance and the trust shown in the apprentice correspondingly high. The manager of electronics retailer A, on the other hand, gave his apprentice independence because he lacked the time to guide and supervise him:

It's really busy here, so we cannot stand over him [the apprentice] all the time. (Manager)
The manager of the sports and leisure equipment store also pointed to lack of time to supervise but felt that independence and responsibility 'made them better'. As long as the young apprentices do their work well and show they can shoulder the responsibility they were given, the practice, we found, supported the building of work-based social capital. The apprentices' subjective appreciation of the freedom, trust, and responsibility invested in them provided rich learning affordances, as the apprentice in electronics retailer A says. He believed the manager saw him as increasingly able to undertake responsibility: They [management] trust you more. (Arne,20) Assuming responsibility for jobs encouraged personal and professional development and the satisfaction of fulfilling the trust placed in them, leading to increased selfconfidence and self-efficacy.
The desire expressed by the managers to turn the apprentices into independent employees can be seen as emphasizing their interdependent role in the workplacethat is, not their independence of the work team or of management control, guidance and support but rather the interdependence of the work community. The apprentices were expected to work as hard as the regular employees. The workplace as an arena for building democratic spirit, trust and network affiliations is embedded in complex interdependencies between the staff and management, work practices, production goals and organizational rules.

Discussion
Studies underscore the relevance of Putnam's (2000) concept of bonding networks linking people with a sense of common identity, such as family and close friends, but also linking people with a common culture or ethnicity. Research shows that many of us have close friends at work. Moreover, most of us feel part of a community at work and look forward to being with co-workers each day. The work communities in the stores consisted mainly of young people working part time, with the apprentice in most cases being the only full-time employee besides the manager. Furthermore, the apprentices report high degrees of well-being and contentment with their work and learning experiences in the stores. Similarities in age and stage of life might bridge divides within the workplace. However, employees with a diversity of ethnicities, genders, political and ideological beliefs, social classes, religious affiliations and temporal orientations to retail work (long-term career perspectives versus 'just a job') regard the workplace as a sound and significant venue for building and bridging social capital.
One argument put forward to enhance social capital through work is to bridge occupational divides within the workplace. This can entail bringing professional ethnic groups closer together and narrowing levels of hierarchy and gender gaps. The shops studied in this chapter were flat organizational structures with few or no middle managers located between the sales assistants and the store manager. For much of the day, the manager worked in the store alongside the employees. The main work involved customer interaction and sales transactions. There was task equality in the sense that all employees, the manager included, undertook the same tasks, implying a non-hierarchical socialization process.
Because the retail workforce is typically characterized by weak professional segmentation, there are fewer inter-professional power struggles in the retail sector than in, for example, in the field of health and medicine (doctors, nurses, skilled health workers and unskilled assistants). On one hand, there is less concern with bridging occupational divides within the workplace; on the other, a wealth of literature points to the common orientation and the potential for collective action that strong professional identity can produce (e.g. Abbott, 1988;Larson, 1977). The data indicate low levels of a common identity as 'sales assistants'. Rather, one might develop sentiments of belonging in the retail chain or the work team. Thus, communities of practice in retail work might constitute rich arenas for generating social capital by bringing diverse types of people together centred on shared practices and mutual engagement.

Conclusions
An apparent consensus exists according to which social capital is exclusively connected to civil society or, more specifically, that participation in voluntary organizations is the linchpin of democracy. The point of departure in this chapter was not to challenge whether an active civil society can be good for democracy but that we lack empirical evidence of the link between participation in voluntary organizations and level of trust (cf. Wollebaek & Seegård, 2011, p. 40). This fact has led us to explore other explanations.
Obviously, one possibility is that researchers of social capital have misunderstood the link between social capital and democracy. Another possibility is that social capital is important but that the wrong venues have been studied. Wollebaek and Seegård (ibid.) agree with the latter proposition in stating that the traditional social capital perspective is not supported empirically. While 'schools and workplaces may be more important', they neither pursue nor refute the idea, arguing instead that the 'role of voluntary organizations is not exaggerated but misguided and wrongly specified'.5 In their wide-ranging anthology Sosial kapital i Norge [Social Capital in Norway], a book which presents the state of the art in this area of sociology, no contribution examines social capital in the workplace. Consequently, when researchers of the voluntary sector find only limited empirical support for the Putnam perspective in Scandinavia, it could appear as a response to the factual situation -that the perspective is misleading. Another explanation is that insufficient focus has been placed on the most important institutional setting where trust is generated -the workplace.
The main argument in this chapter has been the relevance of the workplace in understanding arenas where generalized trust can be generated, while the implications of the discussion are relevant to understand the link between generalized trust and democratic engagement. Our main findings are that networks, cooperation and possible personal and generalized trust are being generated in the workplaces. Since it takes place on the shop floor, it can, and possibly will, happen in other workplaces as well. This argument is in line with Pateman, who claims that democratic practice within working life has far-reaching implications for democracy. Should this conclusion be correct, we can argue that trust created in working life is an important contribution to the quality of the democracy, even when other institutions which are supposed to lead to political engagement are missing, such as labour unions, employee representation on boards and other democratic processes. However, in line with Rothstein and Stolle's (2003) notion of institutional social capital, the retail apprentices are embedded within broader institutional arrangements which frame their experiences. First, the apprentices and the employers are subsumed within the well-institutionalized framework of the vocational education and training system in Norway. This includes established rules, norms and legal indentures, providing the apprentices and the employers with specified rights and duties. Second, working life in Norway is regulated by established relationships in terms of tripartite cooperation, as part of the Nordic model -specifically, institutionalized networks based on reciprocity and trust of each other and society (Levin et al., 2012). This implies rules and norms prohibiting apprentice exploitation. Third, the retail sector is subsumed within consumer protection legislation, regulating the relation between the seller and the purchaser. Taken together, these institutional arrangements have implications for vertical face-to-face relations and, thus, the type of trust generated. The importance of institutions as a precondition for generalized trust addresses the vertical dimension, whereas Putnam's perspective of face-to-face interaction highlights the horizontal dimension. In this chapter, we have demonstrated the relevance of combining the two dimensions -the vertical within the horizontal -to understand the ways in which social capital is generated in working life.
By using young apprentices in the retail sector as an extreme case, we have shed light upon the ways work practices produce trustful communities of practice. We argue that the level of trust invested in the apprentices and the responsibility they shoulder foster emotional engagement, self-confidence and trust. Young people are gradually recognized as members of the workplace network by their co-workers. Insofar as the store is often the first point of entry into the labour market for a significant proportion of youth today, the ways in which this type of labour market functions as a 'school of democracy' is of great importance. This chapter has pointed to initial ideas for generating new research questions while issuing a call for a new research agenda for social capital scholars. The potential of the workplace as an arena where social capital can be built appears non-ambiguous yet strongly understudied.