Networks of Injustice and Worker Mobilisation at Walmart

This article investigates the use of Internet networks during the recent mobilisation of Californian Walmart workers. The findings of this case study suggest that Internet‐based mass self‐communication networks (Facebook, YouTube, etc.) can complement traditional organising techniques. Mass self‐communication networks ameliorate many of the weaknesses identified by previous studies of Internet networks. In particular, these types of networks can help overcome negative dispositions towards unions, increase the density of communication and the level of participation among members, create a collective identity congruent with trade unionism, facilitate organisation and spread ‘swarming actions’ which are effective at leveraging symbolic power. Moreover, unions may be well suited to providing crucial strategic oversight and coordination to wider worker networks.


INTRODUCTION
The late 20th and early 21st centuries have been marked by an extraordinary decline in private sector trade unions across many advanced capitalist countries. Of particular note has been the dramatic decline in membership and collective bargaining coverage in the United States, along with the scale and scope of labour movement collective action, most obviously strike activity (Milkman, 2013). However, in 2012, the US labour movement unexpectedly gained media prominence. In the last week of November, around 600 Walmart and fast food workers took part in widely publicised strikes (Eidelson, 2013;Greenhouse, 2012;Milkman, 2013). The striking workers were not, however, members of unions, belonging instead to worker networks (Coulter, 2013) which more closely resembled worker centres, that is, 'non-union labour-oriented advocacy groups' (Milkman, 2013: 648).
Worker centres have played an increasingly important role in the US labour movement since the early 1990s, running highly visible and often successful campaigns on the behalf of low-paid non-unionised workers, particularly those employed in the service sector (Milkman, 2013). Over the last two decades, the number of worker centres has grown at a rapid rate, increasing from only four in 1992 to over 200 in 2010 (Fine, 2011).
Unlike most worker centres, the Walmart and fast food workers' organisations were membership-based, had been set up by unions, and retained close organisational and financial ties with organised labour. 'The Organization United for Respect at Walmart' (OUR Walmart) was founded clandestinely by the 'United Food and Commercial Workers International Union' (UFCW) in 2010, and various local fast food worker associations were set up by the Service Employees International Union in 2012 (Finnegan, 2014;Uetricht, 2013).
The UFCW set up OUR Walmart, the focus of this article, after being contacted by a group of workers seeking union representation. Following this contact, the UFCW decided to form and support a network for Walmart workers rather than initiate a traditional unionisation campaign. During this initial recruitment phase, efforts focused upon face-to-face organising, principally visiting the homes of workers who had been identified as sympathetic during previous UFCW campaigns or as a result of organisers making contact with them during store visits. Around a year later in June 2011, the worker network went public and launched its website and Facebook page, and conducted a mass survey of workers. This public launching of the campaign was covered by a 1,250 word article in the New York Times by Greenhouse (2011). Though the network was created through traditional face-to-face organising, over time, mass self-communication networks (MSCNs), such as Facebook, became increasingly important in complementing physical activities.
The mobilisation of low-wage service workers maintained its momentum throughout 2013, with around 200 Walmart workers taking part in strikes alongside a campaign of civil disobedience leading to over 250 arrests ('Making Change at Walmart', 2014;OUR Walmart, 2014) while the number of fast food strikers grew to over 2,000 (Fox News, 2013;Tritch, 2013). This article explores the role of Internet-facilitated networks (IFN) and in particular MSCN (e.g. Facebook, YouTube, etc.) in the mobilisation of Walmart workers in California during 2012 and 2013.
There has been much research to evaluate the benefits, drawbacks and threats which IFNs represent to unions. However, very little research evaluates the ways in which traditional PC website and email-based IFNs differ from multi-platform MSCNs. Contemporary MSCNs are not limited to PCs but can be accessed through wireless mobile devices (principally smartphones) and have proliferated during the current decade. 1 Of particular importance is their ability to overcome the problems which have been identified for unions with regard to traditional IFNs. This article investigates the role of MSCNs in the Walmart mobilisation. More traditional elements related to workplace mobilisation also played a key role. Of particular importance were face-to-face organising through home and store visits and the holding of local worker meetings along with the principle of 'like-recruits-like', meaning that union organisers were of diverse ethnicity in order to match the make-up of the workforce. As will become evident from the analysis below, MSCNs should be understood as complementing traditional organising and facilitating particular forms rather than as independent or spontaneous. Although touched upon where relevant, it is not possible to fully detail the more traditional elements of the mobilisation here. This article will proceed by further outlining the Walmart workers' mobilisation during the period 2012-13 with particular reference to the mobilisation in California. The extant literature on unions and IFNs will then be reviewed and the benefits and problems for unions highlighted. The methods employed in this study will then be explained, the findings presented and discussed in relation to the extant literature, and conclusions drawn.

OUR WALMART
Walmart is the world's largest private sector employer with a global workforce of 2.2 million, 1.4 million of whom are hourly paid workers in the United States . Walmart is also the biggest US corporation by revenue (Fortune, 2013). This employer provides an interesting case with which to study the potential for networking labour through IFNs because it exemplifies many of the difficulties facing contemporary labour movements. For example, workers have little access to 'structural economic power' as the work does not require 'high-skill' levels; workers are drawn from loose labour markets and cannot easily disrupt their employer's operations or other strategic sectors of the economy (Silver, 2003). Moreover, workers cannot easily compensate for this lack of structural power through traditional forms of associational power (Silver, 2003;Wright, 2000) as Walmart has repeatedly proved its ability to defeat unionisation attempts in the United States (Lichtenstein, 2009).
Nevertheless, by 2012, thousands of Walmart workers had joined OUR Walmart, and these workers then initiated Walmart's first ever national strike. The strike, which took place on 23 November 2012, 'Black Friday', the busiest US shopping day of the year, mobilised 400 workers (Eidelson, 2013) and 30,000 supporters across the United States .
In the year that followed, further strikes took place involving around 200 workers alongside a campaign of civil disobedience leading to over 250 arrests (Making Change at Walmart, 2014;OUR Walmart, 2014). The exact number of workers who joined OUR Walmart is confidential, but it was claimed to have been in the thousands nationally, across 700 stores (each with a workforce of around 300). The data discussed below suggests that, in California alone, there were over 1,000 members. The handful of strongest stores in the Los Angeles (LA) and San Francisco Bay (Bay) Areas were reported as having a membership of around a hundred workers. However, more commonly, the most active stores had a membership in the range of a dozen to 50 workers, and there were many stores with only half a dozen or fewer members.
Mobilisation at Walmart was facilitated by what were claimed to have been two separate organisations: the UFCW, a union with a broad membership of more than 1.3 million across the retail, food processing and meat packing industries; and the Walmart workers' network, OUR Walmart. In practice, however, the two organisations worked in tandem. The UFCW provided the majority of financial resources, as well as organisational and legal experience and expertise. For example, in the LA Area, the UFCW funded 10 organisers, meaning that the UFCW committed hundreds of thousands of dollars in wages alone to mobilising Walmart workers in the LA Area.
That the UFCW attempted to mobilise workers to join OUR Walmart rather than the UCFW itself can be partly explained by the fact that US labour law requires unions to hold formal workplace certification elections. This process is extremely one sided owing to the principle of 'employers' right to free speech' meaning that they can hold 'captive-audience assemblies' while denying unions access to the workforce. Even if a union does win the election, employers have become well practised in delaying the signing of a legally binding collective agreement long enough to hold a decertification election. Only one in 20 union campaigns now ends in the signing of a collective agreement (Lichtenstein, 2009).
The mobilisation at Walmart should not, however, be understood as a 'pre-union' campaign (Heckscher and Carré, 2006), for even if OUR Walmart were to become a sustainable organisation, there would still be little realistic likelihood of translating this into union recognition. Therefore, the mobilisation explicitly did not seek formal collective bargaining rights but, rather, followed the worker centre model of calling public attention to poor terms and conditions (Milkman, 2013). This approach represents a significant challenge to traditional concepts, such as the Webbs (Webb and Webb, 1896), of what unions are and how they operate. The rationale behind this mobilisation will, therefore, be contrasted to this traditional conception of unions in the discussion below.

THE PROMISE OF IFNs
The expansion of IFNs provides unions with a means with which to become networking organisations (Carter et al., 2003;Fitzgerald et al., 2012). The formation of networks, argue Heckscher and Carré (2006), represents a solution to the inability of contemporary labour movements to mobilise workers into large centralised and disciplined organisations. Networks have the advantage over traditional unions insofar as workers are less fearful that involvement will lead to retaliation and are not burdened by the negative impressions which many workers have of unions (Saundry et al., 2006;2007). By tapping into these networks, unions can, even in hostile environments, gradually introduce young workers to ideas surrounding unionism and demonstrate their relevance. It is argued that this increases the likelihood that a new generation of workers will engage in union activities in the future (Panagiotopoulos, 2012;Saundry et al., 2006;. Moreover, Saundry et al. (2006; find that networks can be effective at generating consciousness and awareness of workplace issues and can, therefore, play a role in the defence of workers' terms and conditions. The use of the Internet in the formation of networks has a number of additional potential benefits, an important one being the ability to increase the density of communication between members (Carter et al., 2003). It can also economically reduce time-space barriers to participation and services, something which is especially beneficial to women workers and workers who work non-standard shifts (Carter et al., 2003;Diamond and Freeman, 2002;Fitzgerald et al., 2012;Greene et al., 2003;Saundry et al., 2006).
IFNs can thus aid the generation of a sense of collectivism and consciousness of workplace issues by providing a space, beyond formal union structures, in which workers can discuss workplace issues (Saundry et al., 2006;2007). They also assist in the organisation of campaigns (Diamond and Freeman, 2002;Saundry et al., 2007) and, when combined with a high level of communication technology, can enable 'swarming' collective action-the strategic pulsing of action from all directions (Arquilla and Ronfeldt, 2000). It is argued that swarming, which is a tremendous force multiplier, could provide labour with an effective new means of undertaking collective action (Heckscher and Carré, 2006). Importantly, swarming is also seen as being well suited to reputational damage (Arquilla and Ronfeldt, 2000).
Despite the significant promise which the extant literature suggests IFNs represent to worker organisation, there are also a number of drawbacks, limitations and threats which have been identified. A major drawback is the sustainability of IFNs and their lack of solidity, which can limit their ability to achieve hard industrial relations outcomes (Carter et al., 2003;Fitzgerald et al., 2012;Greene et al., 2003;Heckscher and Carré, 2006;Saundry et al., 2012). Many commentators suggest that there exists an incongruence between the horizontal and de-centred nature of networks, and the more bureaucratic, hierarchical and centralised nature of unions (see e.g. Diamond and Freeman, 2002;Heckscher and Carré, 2006;Saundry et al., 2007;. However, effective networks are not in fact based upon total autonomy, but rather, they need an 'orchestrator' to provide strategic oversight, while still enabling members of the network to quickly and easily share information (Arquilla and Ronfeldt, 2000;Heckscher and Carré, 2006;Heckscher and McCarthy, 2014). Unions often fail to grasp this and use IFNs in traditional hierarchical ways based upon vertical downwards communication such as employing websites and email to post materials and information for members to take up and use (Fitzgerald et al., 2012;Panagiotopoulos, 2012). The result is the tendency that such resources are only used by existing members, rather than attracting new members (Saundry et al., 2007).
A further limitation is that Saundry et al. (2006;2007; argue that they do not lead to the formation of a broad political identity compatible with contemporary trade unionism. Rather, networks foster a narrow collectivism based upon collegiality. Users of networks, it is argued, do not see themselves as members of an organisation, making it difficult for IFNs to adopt a clear political identity or a more substantive organisational form. This incompatibility means that unions cannot take over networks or set up their own. To do so would be self-defeating for it would inevitably erode the network's identity. Alternatively, by linking with trade unions, networks can gain the organisation, institutional identity and expertise to achieve concrete industrial relations outcomes. Saundry et al. (2012: 274), therefore, suggest that, instead of setting up networks, unions should tap into existing networks. However, networks themselves represent a threat to unions as they are alternative means of workplace collectivism and service provision (Diamond and Freeman, 2002;Panagiotopoulos and Barnett, 2014;Saundry et al., 2007).
The foregoing discussion is drawn from research into traditional IFNs as there is currently little research which specifically investigates MSCNs (Geelan, 2013). One important way in which MSCNs differ from IFNs is that, unlike traditional IFNs, they can be closely monitored and shut down from a central source (Bryson et al., 2010). However, evidence from the 2010 Chinese Honda strikes demonstrates that, as MSCNs became infiltrated, they can easily be abandoned and new ones set up at little cost (Jianhua, 2011). Jianhua's findings indicate that MSCNs share many of the benefits of traditional IFNs.

METHOD
In order to investigate the role of MSCNs in workplace mobilisation, a broadly ethnographic approach, combining both participant and non-participant observations with semi-structured interviews, was used. California, particularly the LA and Bay Areas, was identified as having the highest concentrations of mobilisation and was therefore chosen as a primary research site. Data collection took place during two intensive fieldwork trips. The first of these took place from mid-February to mid-March 2013 and was timed to allow reflection upon the 2012 'Black Friday' strikes, as well as observation of mobilisation attempts for a mass fortnight long strike to coincide with Walmart's annual shareholders' meeting on 7 July 2013. The second took place for two weeks at the start of December 2013, timed to allow reflection upon the 2013 'Black Friday' protests and direct actions. The fieldwork included participation in, and observation of, organising drives at six stores, which also proved an excellent opportunity to speak with workers who were not mobilised, and attendance at two weekly union organiser meetings, two worker activist meetings, two worker activist national video conference calls and three community ally meetings. This experiential and observational data were supported by 42 semi-structured interviews with 33 informants consisting of 24 non-managerial hourly paid employees (four of whom had recently left Walmart), one recently terminated salaried assistant manager who was active in OUR Walmart, seven UFCW union organisers and one senior UFCW official. The research was not covert, and all participants gave informed consent. All names have been anonymised.

NETWORKING INJUSTICE
In order to elucidate the ways in which MSCNs influenced the mobilisation, the analysis is undertaken with reference to the influential approach known as 'Mobilisation Theory' (MT) (See Kelly, 1998). According to this theoretical perspective, dissatisfaction at work, while necessary, is not sufficient to trigger collective action. This dissatisfaction must also be understood by actors as the result of an illegitimate situation and thus an injustice. Injustice is defined by Kelly (2005: 66) as a 'breach of legal or collective agreement, rights or of widely shared social values'. Different actors in the workplace will attempt to influence the understanding of workplace issues by framing them in a particular way. Hence, framing injustices represents one potential avenue through which MSCNs may influence mobilisations.
Organisers and activists emphasised the disparity in Walmart's profits, the owners' wealth and the executives' pay relative to that of the workers, as well as comparing worker pay and conditions at Walmart to those of workers at unionised retailers. Moreover, Walmart attempted to legitimise these issues by stressing that the average pay of an associate was well above a living wage, and by highlighting the claimed pro-worker values of Walmart, as personified by the deceased founder Sam Walton. Additionally, managers sought to reduce opportunities for organisers to undertake this framing by having them ejected from stores, in some cases by the police. The ability of workplace activists to frame issues was also curtailed by managers separating them from their workmates and increased managerial surveillance of their activities in the workplace.
In this context, online spaces, or what Castells (2012)  MT holds that, for mobilisations to take place, feelings of injustice must transform into a sense of shared collective interests. The process by which injustices fuse into collective interests can be broken down into two elements: social identification with a group and attribution of blame. Social identification with a group requires a shared perception of being located in a particular group and distinct and defined in opposition to an out-group who have different interests and values. Moreover, a sense of injustice can be strengthened if shared by a substantial number of co-workers as this reinforces and legitimates it (Kelly, 1998).
At Walmart, discussions in the workplace and meetings outside of the workplace with other workers from the local area played an important role in creating a perception of a distinct social group. But MSCNs aided this process by providing a discursive space in which union organisers and workers could further co-produce a collective identity of exploited workers. MSCNs also enabled high levels of communication, enabling dispersed workers, either spatially in different stores or temporally on different shifts, to connect with each other and organisers. Through engaging in discussions over Facebook, workers were able to learn of situational similarities at each other's stores and provide each other with practical and emotional support and thus fostered identification with each other's situation. Akira, a recently terminated worker who was working as an organiser, explained this process particularly clearly: It is basically an outlet for, not only, frustration but also networking . . . seeing . . . what Walmart is doing now to other associates and comparing our similarities . . . just being there for one another so you know that you're not the only one going through what you're going through and spreading the word about trying to change Walmart and get others to join in.
Tim, an activist in his late 20s, explained the profound effect of the realisation that a sense of injustice is shared by a wider collective: It's a great way for people, especially when they are first starting, to be introduced to the larger scope of things, because there is a difference when you're used to dealing with your individual store and then when you see it is nationwide and you're talking to other people-it kinda blows your mind away. A lot of workers think that the problems they are experiencing are just this store or it's just that manager, but everything else is great. But when they . . . hear or see the same problems they are dealing with being expressed by people in Washington State or New York or Texas they are just, like, wow! As suggested in the quote above, a sense of group identity was further fostered visually by the uploading of videos on Facebook and YouTube of speeches by charismatic leaders and totemic actions. These connections were possible despite the network being geographically dispersed across a vast country. Bill, a senior UFCW official, explained how mass self-communication massively expanded social interaction and social network density: It's been transformative . . . there's thousands of conversations happening every day amongst members of OUR Walmart . . . It's totally widely open, people are building their own groups, they are learning from each other, they're supporting each other . . . this campaign wouldn't have been possible five years ago . . . it breaks down the barriers and the walls that people face in life and it's also a place where people can support each other whether they are in the same store or across the nation and, lastly, it's got a natural way that people can become engaged.
MSCNs were easily combined with traditional IFN communication, such as online video conference calls, which enabled workers from across the country to link together in order to discuss major issues, provide feedback and make decisions. Furthermore, being a network provided a discursive platform from which, by stressing the fact that their social group was not a union but an association, they could circumvent the prevalent anti-union prejudices of much of the workforce. This enabled easier identification with OUR Walmart by workers who would otherwise have been hostile to a 'union'.
'Attribution' refers to pinpointing a tangible target for action by placing blame upon an agency, normally the employer or the government, rather than impersonal forces such as 'the market' or 'global competition' (Kelly, 1998). Again, MSCNs aided this process by providing a discursive space in which union organisers and workers could foster the blaming of the owners. Initially, Walmart workers attributed blame to 'greedy' and 'criminal' owners and directors of Walmart who were deemed to be exploiting them. This can be seen particularly clearly in Pamela's statement that Walmart has been lying to them [the workers] because their slogan is 'Save Money to Live Better' but who's actually living better? The Waltons, not their associates.
However, following the strikes by fast food workers, this identity coalesced into being a part of a wider 'low-wage workers' movement'. The workers thus came to see their mobilisation not as a sectional fight to improve their own conditions, but as part of wider battle to improve the conditions of low-paid workers. This was a common theme of the second batch of interviews carried out in December 2013 and was elucidated particularly clearly by a worker called Joe: What we are doing isn't just for Walmart workers, it's for everybody because everyone's getting the shit of the deal right now . . . we are all allies, we need to stand together, you know what I mean-solidarity.
Clearly, the collective identity fostered here was not, as suggested in the literature above, one based upon notions of a collegial or occupational identity which would be incompatible with the broad political identity of contemporary trade unionism. Rather, it led to a general class-based identity of exploited workers.

A MOBILISATION FOR 'DIFFERENT TIMES'
According to MT, for collective interests to translate into collective action, there must be an organisation that can provide the necessary resources. At Walmart, both a union, the UFCW, and a worker network, OUR Walmart, played key roles. Despite the union's close links and heavy support for OUR Walmart, it remained committed to trying to mobilise workers to join OUR Walmart, rather than the UFCW. OUR Walmart, for its part, explicitly did not seek to engage in the formal collective bargaining which is central to classic conceptualisations of trade unions. Although the union played a vital role in the decision making of the mobilisation, it did not do so in a bureaucratic manner. Instead, the network's meetings, whether on-line or in person, were run in a participative manner, departing from the formal process-heavy manner typical of union meetings. Union organisers played the role of facilitators, actively seeking out workers' views and encouraging participation. That the workers found the experience empowering was clear and they undeniably felt a sense of ownership of the network and the decisions it made. This was something Rachel, a worker, felt particularly strongly about: The relationship between union and network was not marked by the cultural conflict between a bureaucratic union and the horizontal network that the literature suggests. Although it was too early to investigate the long-term sustainability of the network, it had lasted two years in the open without fragmenting and had managed to maintain enough solidity to undertake some significant actions.
With a sense of collective interests and an organisational structure in place which is able to support mobilisation, MT suggests that an important element influencing collective action is the opportunity for different forms of action to be successful. The mobilisation at Walmart lacked the economic and political sources of power which the US labour movement has traditionally relied upon, such as strikes and stateenforced recognition. But Wright and Brown (2013) argue that employers' increasing sensitivity about their public reputations may present an alternative 'opportunity structure'. Indeed, Chun's (2009) ethnographic studies of union organising of cleaners and domestic care workers in the United States and cleaners and golf caddies in South Korea demonstrate the importance of such symbolic power. The workers she studied were successful owing to their ability to manipulate socially accepted concepts and norms in order to legitimise their struggle.
Vulnerability to reputational damage explains why the Walmart mobilisation emphasised high-profile confrontational direct action aimed at the raising of public awareness. This was summed up by UFCW organiser, Ali: We are much more about taking direct action . . . you know we are taking action now and doing something about it now rather than waiting for the law to do something. Hyman (1989) states that strikes are, by their very nature, collective, but when Walmart workers took strike action, it was as part of a small workplace group of around five workers at most and sometimes even lone individuals. It was only because of their interconnectedness through MSCNs, particularly Facebook, that it came to be seen as truly a collective act. The other major form of action was similarly dramatic civil disobedience which resulted in workers and supporters being led away in handcuffs. The first time this tactic was used was during the 2012 'Black Friday' strike when workers, family and supporters in the LA Area blockaded the road outside a store for two hours before being arrested. The spectacle of these direct actions was heightened by the presence of large numbers of supporters.
MSCNs played two crucial roles in these actions. First, they were an effective means of spreading word of the dozens of disparate but simultaneous actions through Facebook 'Events', while the lack of rigid organisational or communicative boundaries enabled the expansive networking of support for these actions. This support included other low-wage workers and labour and community groups and meant that significant solidarity was mobilised both physically and financially. For example, the sheriff's department reported that at the main 2012 LA Area demonstration, there were around 1,000 supporters (Fox, 2002).
Second, MSCNs enabled traditional and self-generated coverage of these actions to be widely disseminated. For example, OUR Walmart (2013) claims there were over 300,000 posts on Facebook and 60,000 tweets on Twitter about the 2012 'Black Friday' strike. The swarming of these actions and their representation through MSCNs increased media interest. Nevertheless, this required the union to play the role of network orchestrator in order to bring together the actions effectively and strategically. The force multiplier effect of swarming can be seen by the fact that the 2012 'Black Friday' strike involved a relatively small number of workers stopping work for a single shift but created a significant level of media coverage of working conditions at Walmart. For example, during November 2011, the only coverage relating to working conditions at Walmart in the New York Times amounted to just 57 words in one article, whereas during November 2012, there were 2,089 words across six articles. 2 According to a senior UFCW official, the print and website coverage generated by OUR Walmart alone was equivalent to $24 million of advertisements in 2012 and $31 million in 2013. 3 As Michael Bender (2012), President of Walmart West, put it: 'the media coverage created the illusion that Walmart's associates were protesting instead of serving customers'.
Finally, that workers did not have access to a PC at work was not a barrier to MSCNs playing an important role. Workers carried smartphones with them, enabling easy access to the OUR Walmart network. The use of MSCNs was not confined to younger workers or those with higher levels of Internet proficiency. In fact, of the five worker informants aged over 60, three made frequent use of MSCNs. The benefits of MSCN use were consistent with the expectations derived from the IFN literature. The data also supports Jianhua's (2011) finding that the ease with which MSCNs can be closely monitored is not an insurmountable problem. Though it was no doubt easy for managers to observe workers on the open pages of social networking sites such as Facebook, what actually matters in terms of participation is not whether they were under surveillance or not, but whether they felt they were. Moreover, this sense of freedom from surveillance could be heightened with little cost with the use of more secure online spaces such as video conferencing and private local Facebook pages.

DISCUSSION: THE EFFICACY AND LOGIC OF MSCN MOBILISATION
This investigation into MSCNs replicates many of the findings of previous studies, outlined above, which identify ways in which IFNs are advantageous to workplace organisation. In particular, MSCNs were found to: overcome workers' negative dispositions towards unions and engage a new generation of workers. Increasing the density of communication, level of participation and internal vibrancy aided the creation of a collective identity, and facilitated the organisation and propagation of swarming actions. Importantly, the above findings demonstrate that MSCNs do not suffer from the barriers to access which have been found to exist for traditional IFNs. Workers did not need the use of a PC at work in order to participate as they had smartphones which enabled them to access the network. MSCNs in this context were not only used by young workers or those with a higher level of Internet proficiency. MSCNs and smartphones have a high level of interface simplicity, which does not require specialist computer-related knowledge, compared with earlier eras of the Internet and computer software. This is the very reason that they have been so widely adopted.
The findings also provide no evidence for the suggestion by Saundry et al. (2006;2007; that the type of collective identity which develops through networks is incompatible with the broad-based political identity of contemporary trade unions. In fact, the collective identity forged among Walmart workers coalesced into a classbased notion of exploited low-wage workers. There are two potential reasons for this. First, the findings of Saundry et al. may have been overly influenced by the specific workers whom they studied. These workers were highly skilled and professionalised freelance media workers. In fact, Saundry et al. (2012: 271) themselves state that the identity which developed through the networks they studied were 'mediated by common experiences and a discourse rooted in the nature of the labour process'. The narrow collegial collective identity which they identify perhaps reveals more about freelance media workers than it does about networks. This might suggest that while IFNs facilitate the generation of greater levels of collective identity, the form they take will be rooted in the labour process and workplace organisation.
Second, the broad-based collective identity of the Walmart mobilisation could also, in part, be a reflection of the 'expansive networking' which is a result of MSCNs being de-centred and lacking in defined boundaries, compared with traditional website forum and email-based networks (Castells, 2012). This means that broader sets of connections are likely to form as anyone can connect to the network. This is what happened as the Walmart mobilisation continued over time. As the identities of Walmart and fast food workers were increasingly linked together through MSCNs, the OUR Walmart collective identity transformed from one focused on being exploited workers of a specific employer to a general sense of being exploited lowwage workers.
Saundry et al. go on to argue that the collegial identity which they associate with networks means that users of networks will not see themselves as members of an organisation, will wish to remain independent from unions, and that any attempt by unions to create, organise and take over networks will be self-defeating as it will inevitably erode the network's collegial collective identity. These propositions are also not supported by this study. Not only did MSCN users identify as members of an organisation, but that organisation was set up, and its direction heavily influenced by a broad-based union.
The findings support the view that networks do not require total autonomy but rather an orchestrator who can provide strategic oversight (Arquilla and Ronfeldt, 2000;Heckscher and McCarthy, 2014), as long as the orchestrator does not attempt to bureaucratise communication and instead allows the quick and easy sharing of information across the network. Despite being bureaucratic themselves, it is possible for unions to interact with networks in this way. In fact, in this case, MSCN activities complemented the initial traditional union organising, which then facilitated the development of the mobilisation but which was not independent or spontaneous. Saundry et al. then are wrong to suggest that unions are limited to tapping into existing networks. The incongruence which much of the literature (see e.g. Diamond and Freeman, 2002;Carré, 2006: Saundry et al., 2007; suggests exists between the bureaucratic, centralised and hierarchical nature of unions, and decentralised horizontal networks is not inevitable. Unlike the findings of Fitzgerald et al. (2012) and Panagiotopoulos (2012), the union did not attempt to use the IFN in a traditional hierarchical way based upon vertical downwards communication. Instead, the union acted as a facilitator of network participation, seeking to increase the bottom-up density of network communication. This mirrors findings that successful union organising requires 'managed activism'-leadership commitment to building membership-led organising campaigns while coordinating strategy and structural change from the centre (Simms, 2007).
Let us now turn to the more difficult question of whether this MSCN-based mobilisation had the solidity to achieve the concrete industrial relations outcomes necessary to justify the high level of resources it consumed? Hobsbawm (1952) demands that the effectiveness of the labour organisations of the 18th century be judged according to the actually existing alternatives. Likewise, given the high level of surveillance, fear of retaliation and lack of structural power of these workers, even if a more traditional union structure were seen as superior, it seems highly unlikely that a mobilisation would be able to take such a form in the foreseeable future. For better or worse, mobilisations taking place in similar contexts to Walmart may have to partly adopt network forms as the only realistic manner in which they are possible. The big question, then, is whether such networks can achieve influence despite having low density and a lack of structural economic power?
Sidney and Beatrice Webb's (Webb and Webb, 1896) classic definition of a union is mostly concerned with the improvement of conditions through formal collective bargaining. The Webbs also highlight how collective bargaining is dependent upon the collective threat of withholding labour through strikes. But the Webbs' analysis is problematic in terms of union renewal. Wright and Brown (2013) argue that the internationalising of product markets and corporate ownership has undermined the possibilities for multi-employer bargaining, while the intensification of competition and greater fragmentation of labour and product markets have, in turn, undermined the possibilities for single-employer bargaining. It also requires employer 'recognition' of union bargaining rights. However, over the past two decades, there has been substantial 'procedural individualisation' (Brown et al., 2000) of employment relations in countries such as the United States, and this can be equated with de facto de-recognition.
But Hobsbawm (1952: 58) demonstrates that the Webbs' focus upon strikes and formal collective bargaining is empirically and conceptually myopic. During the mid-18th and early 19th century, workers' capacity to influence their employers lay in 'machine-wrecking, rioting and the destruction of property in general (or, in modern terms, sabotage and direct action)'. This direct action could usually be counted on to improve workers' conditions and created a minimum floor for standards which employers understood needed to be maintained in order to avoid negative sanctions (the destruction of their property).
In addition, OUR Walmart's power cannot be understood via the direct economic damage of its strikes but rather by its ability to undertake direct action which can damage the reputation of Walmart. Hence, it should be understood as an example of what might be termed: 'raising labour standards through reputational damage'. Chun (2009) argues that symbolic power is most effective when used against institutions which are sensitive to public opinion, a category which Walmart, as a low-cost discount retailer, does not seem to fit into. Additionally, Walmart is also unusual for a major multi-national in that it remains family controlled (Lichtenstein, 2009); consequently, reputational risk in terms of share price is reduced. The tactics adopted in this mobilisation are, therefore, likely to be more effective when utilised against other employers such as governments, brand-driven companies and universities.
The mobilisation at Walmart was not, however, necessarily misdirected. With reference to workers' direct action in the 18th century, Hobsbawm (1952) explains that the state can play an important role in raising labour standards as a result of direct action. There were two ways in which the Walmart mobilisation might have influenced the state in order to achieve industrial relations outcomes. First, Walmart had reached market saturation in its traditional markets, and same-store-sales had been declining in recent years. This made expansion into metropolitan markets, which in the past it had avoided owing to the presence of unions, crucial (Lichtenstein, 2009). Yet this expansion could be delayed and even blocked by concerned local authorities, as recently happened in New York City (Greenhouse and Clifford, 2013). Thus, symbolic sources of power could be translated into a political source of power by compelling Democrat politicians to believe that it would be untenable for them to support new Walmart developments. Potential evidence of the success of this approach is that in 2015, Walmart announced that it would raise its starting pay rate to $9 an hour, improving the pay of 500,000 workers, and then to $10 an hour in 2016 (Isidore, 2015). This equates to around 12 and 25 per cent more than what many workers in this study were being paid at the time.
The second route through which the mobilisation might have influenced the state in order to help achieve its goals is through minimum wage laws. For example, in California, before the mobilisation, the minimum wage had remained at $8.00 per hour since 2008. However, in September 2013 (17 months before Walmart increased its starting pay nationally), Democrats increased it to $9.00 per hour from July 2014 onwards, and to $10.00 per hour from January 2016 onwards. Additionally, in 2014, Democrats in San Francisco and Oakland both raised their minimum wages respectively to $15.00 per hour from 2018 and $12.25 per hour from March 2015. Los Angeles is expected to enact a similar raise in the near future (Hirsch, 2014). This suggests that reputational damage can achieve concrete outcomes regardless of the total denial of recognition by employers.

CONCLUSION
This article has demonstrated that MSCNs have the potential to significantly benefit workplace mobilisations. They can help to overcome negative dispositions towards unions and engage a new generation of workers, increase the density of communication and the level of participation, aid the formation of collective identity, facilitate organisation and spread swarming actions. MSCNs overcome many of the weaknesses of traditional IFNs identified in previous studies. Having limited access to a PC at work, being older or less Internet proficient were not barriers to MSCN participation. Importantly, the collective identity which developed through MSCNs was compatible with the broad political identity associated with trade unions. Similarly, no barriers to the successful establishment of networks by unions were identified, casting doubt on the oft-expressed belief in the inevitable clash between the hierarchical nature of unions and the horizontal nature of networks. In fact, unions may be well suited to playing the crucial role of network orchestrator in order to provide strategic oversight and coordination.
The raising of labour standards through reputational damage rather than formal collective bargaining is also identified as the mobilisation's rationale. This is arguably better suited to contemporary socio-economic conditions marked by the internationalisation of product markets and ownership; increased competition and fragmentation of labour and product markets; and the procedural individualisation of employment relationships, than formal collective bargaining. It is also a strategy which is likely to be more appropriate to mobilisations targeting organisations which are more concerned with risks to their reputations, or in situations in which the state is more directly involved in the employment relationship. Although it is not currently possible to determine whether the mobilisation at Walmart is sustainable, the use of MSCNs to raise labour standards through reputational damage may provide a fruitful approach for other workers.