This study has set out to explore the conditions that enable and constrain the capacities of local unions in garment producing countries to build sustained bargaining power vis-à-vis capital and state actors and thereby bringing about lasting improvements for workers. To this end, I have integrated theoretical concepts from research on labour control and labour agency in GPNs with the analytical perspective of relational and practice-oriented approaches within economic geography. Building on the concepts of the labour control regime and of labour’s networked agency as central heuristics, this study has been guided by two central research questions:

  1. 1.

    How do labour control regimes at specific nodes of the garment GPN—constituted through place-specific articulations of processual labour control relations at the horizontal and vertical dimension of the GPN with localised labour processes—shape and constrain the terrain for the agency of workers and unions in garment producing countries?

  2. 2.

    Which relationships and routinised interactions allow unionists and workers in garment producing countries to develop strategic capacities and power resources that enable them to shift the capital-labour power balance in favour of workers lastingly?

In this chapter, I first summarise the central findings of this study in relation to each research question before highlighting the main empirical and theoretical contributions. The chapter concludes by reflecting on the challenges and pathways for improving working conditions in the global garment industry and indicating directions for further research.

1 Answering the Research Questions: Summary of Central Findings

1.1 Labour Control Regime in the Bangalore Export-garment Cluster and Constraints for Union Agency

To answer the first research question, I have developed a practice-oriented, relational approach for analysing labour control regimes in GPNs. I have conceptualised labour control regimes at specific nodes of a GPN as emerging from place-specific articulations of various ‘network’ and territorially embedded processual relations of labour control with localised labour processes (see Sect. 3.2). In this context, I have identified six processual relations of labour control that fulfil exploiting or disciplining functions and/or contribute to securing the broader conditions for capital accumulation at specific nodes of a GPN. These processual relations of labour control are: (1) sourcing relations that link global lead firms within a GPN with local suppliers at the various nodes of the GPN; (2) wage relations linking workers, employers and state actors within a specific region, state or country; (3) workplace relations constituted through the interactions between workers and management in a specific site of production; (4) industrial relations constituted through territorially embedded relationships between employers and their organisations, workers and their organisations, and the state in a specific state, sector or country; (5) employment relations, i.e. the relationship between employers and workers, in which workers sell their labour power to an employer; and (6) labour market relations linking employers, workers and (potentially) third-party actors such as contract labour or recruiting agencies and training organisations. Through the lens of this relational, practice-oriented heuristic framework of labour control regimes in GPNs, I have then explored how the labour control regime in the Bangalore export-garment cluster constrains and shapes the agency strategies of three local garment unions (see Chap. 6).

The results of the empirical analysis highlight that major challenges for the agency of unions result from the complex intersections and interdependencies between the above-mentioned six processual relations of labour control. Due to the intricate intertwinings of more localised and spatially more extensive processual relations, employers’ exploitation and disciplining practices are often shaped or enabled by the practices of other actors located at more or less geographical distance. For example, the practices through which Bangalore export-garment manufacturers construct labour process, workplace, wage and industrial relations are shaped directly by retailers’ predatory purchasing practices at the vertical dimension of the GPN: To remain profitable in the face of retailers’ practices of squeezing prices, placing irregular orders and demanding shorter lead times, Bangalore garment manufacturers exercise tight control over workers’ productivity, rely on wage theft practices to keep production costs low, engage in practices of ‘hiring and firing’ to flexibilise employment relations, and perform union-busting practices to mitigate collective worker organisation. These practices are, in turn, enabled through pro business state practices, undermining unions’ capacities to contest illegal exploiting and disciplining practices through legal channels. Moreover, the exploiting and disciplining practices deployed by Bangalore garment manufacturers intersect with and deliberately exploit broader social power asymmetries along the lines of age, gender and geographical origin. For example, manufacturers deliberately recruit predominantly young women from rural areas within Karnataka and increasingly also from North-East India since these workers are less likely to resist exploitation practices.

In a nutshell, the empirical analysis has highlighted three ways in which the labour control regime in the Bangalore export-garment cluster constrains the terrain for the agency of the three local case study unions. First, unions’ opportunities for leveraging structural power resources are constrained by the specific practices through which retailers construct sourcing relations, and through which manufacturers organise the labour process. By constructing spatially asymmetric relationships with suppliers, retailers are able to play manufacturers in different locations off against each other and pressurise suppliers into offering lower prices (Sect. 6.2). Consequently, garment manufacturers in the Bangalore export-garment cluster also continuously seek novel strategies to keep wages low. Most export-garment companies operating in Bangalore maintain regional factory networks, enabling employers to pass price pressures on to workers. Employers argue that due to retailers’ price pressures, they cannot increase wages and would be forced to shift production to lower-wage locations in India if workers attempted collective bargaining (see Sect. 6.3). Spatial asymmetries between transnational retailers and local manufacturers are hence reproduced at a smaller scale in the relations between manufacturers and workers. As a result, workers’ and unions’ capacities to exercise workplace bargaining power vis-à-vis employers through production stoppages are limited. Workers constantly face the risk that either manufacturers or retailers shift production orders to another factory. Besides limited workplace bargaining power, workers and unions in the Bangalore export-garment cluster also possess limited marketplace bargaining power. Since manufacturers organise the labour process in an assembly line system, with more complex operations being increasingly automated, the majority of tasks in Bangalore garment factories can be carried out by unskilled or semi-skilled workers (see Sect. 6.1). This rather low-skilled work profile, in turn, enables managements to balance local labour shortages by recruiting migrant workers from rural areas that have undergone a three-month training in one of the many private and public training centres set up under the Integrated Skill Development Scheme (see Sect. 6.7). As a result, workers possess limited workplace and marketplace bargaining power, which in turn constrains unions’ abilities to implement proactive collective bargaining strategies that could achieve improvements for workers beyond the legally prescribed minimum standards.

Second, limitations for unions’ capacities to exercise power over employers through industrial and workplace action also result from the various constraints that the labour control regime places on unions’ opportunities for building associational power. These constraints result, on the one hand, from Bangalore export-garment manufacturers’ disciplining practices directed at preventing collective worker organisation. These disciplining practices include constructing factories as tightly controlled spaces, reproducing gendered structures of domination on the shop floor (Sect. 6.4), and a range of union-busting practices directed at discouraging workers from engaging with unions (Sect. 6.5). On the other hand, challenges for unions to build and leverage associational power resources also result from the spatial restructuring from Bangalore garment manufacturers’ practices of expanding the labour market frontier: In the face of an increasing shortage of unskilled labour in the Bangalore urban area, garment manufacturers are moving factories to the outskirts of the city or neighbouring rural areas within the State of Karnataka and are also increasingly recruiting workers from villages near Bangalore. Consequently, there is an increasing spatial separation between workplaces and workers’ living areas, with many workers being transported to the factory in company buses from distances of up to 80 kms (Sect. 6.7). As a result, unions face severe challenges for organising workers outside of the factory after their shifts or in their communities—two organising strategies that unions have for a long time relied on in the face of the tight management control inside factories.

Third and last, pro-state business practices constrain workers’ and unions’ institutional power resources. India has traditionally possessed strong labour legislation offering workers various means to challenge illegal exploiting and disciplining practices such as requesting labour inspections, filing complaints with the labour department or filing a lawsuit in the labour court. With the general shift towards neoliberal politics in India’s post-liberalisation area, these traditional sources of institutional power have, however, been dwindling. In the context of a general political climate that prioritises the creation of a business-enabling environment over the implementation of labour rights, many labour officers refrain from taking an active stance for workers in industrial dispute settlement mechanisms, thereby paving the way for an increasing employer dominance in industrial relations (Sect. 6.5). Moreover, due to the chronic understaffing of labour courts, processes are often dragged on for several years until a ruling is made. As a result, unions’ capacities to use institutionalised dispute settlement or legal mechanisms as institutional power resources for challenging employers’ illegal exploiting and disciplining mechanisms are severely constrained.

Despite these constraints for the agency of unions on the Bangalore export-garment cluster, the empirical analysis has also shown that workers and unions have, nevertheless, achieved to stop or transform specific practices of exploitation or disciplining through networked agency strategies. These strategies simultaneously target multiple actors through combined actions at various levels. GATWU, for example, achieved to stop employer and state practices of delaying minimum revisions by targeting retailers, the state and employers at the same time through combined workplace action, local public protests and transnational consumer campaigns. Unions have also used pressure from transnational worker and consumer campaigning networks to achieve interventions by retailers and, thereby, to stop employers’ illegal union-busting strategies. In doing so, unions have opened up spaces for organising and collective dialogue in selected factories. In these factories, unions were able to stop particularly harsh exploiting practices such as verbal abuse of workers or wage theft through ‘giving comp-offs’. Hence, the analysis has also showcased that the labour control regime as an institutionalised framework for capital accumulation is not only unilaterally imposed on workers and unions by state and capital actors. Instead, workers and unions also challenge and transform and thereby co-shape the practices and relations that constitute the labour control regime through their everyday actions and struggles.

In the next section, I turn towards the second research question and summarise insights into the different ways in which unions in the Bangalore garment unions construct networked agency strategies. Specifically, I recapitulate the most important findings regarding which relationships and routinised interactions have allowed workers and unionists in the Bangalore export-garment cluster to develop strategic capacities and power resources.

1.2 Building Unions’ Strategic Capacities and Power Resources in Relational Spaces of Labour Agency

To answer the second research question addressing the potential of different agency strategies for enabling unionists and workers to build strategic capacities and power resources, I have developed a relational heuristic framework for analysing union agency (see Sect. 3.3). Building on the concepts of ‘Networks of Labour Activism’ and worker and union power resources, I have developed a heuristic framework for analysing unions’ strategic approaches through the lens of three intersecting ‘spaces of labour agency’. These spaces are: (1) spaces of organising constituted through unions’ practices of building relationships with workers as (potential) union members; (2) spaces of collaboration constituted through unions’ collaborative relationships with external organisations such as other unions, NGOs or consumer organisations; and (3) spaces of contestation constructed by unions around specific labour struggles through practices of targeting employers, lead firms and (in some cases) state actors, and through practices of ‘drawing’ allies such as other workers, consumer or labour rights groups into struggles. Spaces of labour agency are hence relational networks of processual relations and routinised interactions, within which workers and unions can potentially develop the strategic capacities and power resources that ultimately enable unions to build lasting bargaining power vis-à-vis capital and state actors.

The empirical analysis of the three Bangalore-based garment unions has highlighted that unions construct spaces of organising, collaboration and contestation through different routinised interactions and relationships—with varying implications for developing workers’ and unionists’ strategic capacities and power resources (see Chap. 7). In the analysis, I have identified two stylised strategic agency approaches characterised by different practices of constructing spaces of organising, collaboration and contestation. I have labelled these two approaches as (1) the ‘community organising, fire-fighting’ approach and (2) the ‘strategic factory organising, collective bargaining’ approach. In the following, I will summarise the main practices and relations that characterise each approach as well as the potential and limits of each approach for building sustained union bargaining power.

Under the community organising, fire-fighting approach, unions construct spaces of organising mainly at the community level. Relationships with workers are not constructed primarily around workplace issues but also around issues of workers’ everyday life in the household and the community through organising saving groups and area committees, or providing counselling for family problems. To this end, unions usually collaborate closely with local community organisations. Community organising strategies were prevalent especially in the early phase of the three garment unions, which emerged as independent unions from an NGO-led support project for garment workers. In this context, the community organising approach responded to the specific composition of the workforce in the Bangalore export-garment cluster. Being mostly female, first-generation industrial workers from rural areas, many garment workers in the cluster were unfamiliar with the concept of unionisation and unions as collective organisations. This lack of awareness, combined with tight employer control in the workplace and patriarchal power structures, made traditional workplace organising unviable. Against this backdrop, building relationships with workers through community organising strategies enabled unions to gain workers’ trust, foster collective experiences and thereby build associational power at the local level. However, this organising strategy also had limits in building bargaining power vis-à-vis employers. While organising workers in their communities allowed unions to mobilise relatively large numbers of workers for punctual public protests, it did not enable unions to build a strong membership base inside specific factories. A strong membership at the factory level is, however, a precondition for the exercise of workplace bargaining power through industrial action.

Unions’ early community organising strategies need to be understood as interrelated with the practices of constructing spaces of collaboration that characterised unions’ early stages. In their early years, unions constructed collaborations with external actors, mainly around projects funded by NGOs from the Global North. As mentioned before, the three unions emerged from an NGO-led garment worker support project funded by Oxfam International, through which the unions’ activities and full-time organisers were paid. As for community organising strategies, the empirical analysis has highlighted mixed effects of project collaborations with Northern NGOs for unions’ capacity to build sustained bargaining power. On the one hand, funding from NGOs provided unions with the financial means to sustain their organising and community work when unions were unable to sustain themselves through membership fees. On the other hand, collaborations with NGOs for funded projects had constraining effects on unions’ capacities to engage in strategic factory organising and foster internal union democracy. Constraints for strategic factory organising resulted from the primary objective of these NGO-funded projects, which was usually not to strengthen union-building processes but instead to provide aid for garment workers. In this light, the project’s success was not measured against its contributions to strengthening a unions’ bargaining power but instead against the number of workers reached through the project. This ‘institutional logic’ of maximising outreach (c.f. Egels-Zandén et al. 2015) hence requires unions to create loose ties with a large number of workers rather than strong ties with a smaller number of workers from specific factories. Moreover, funded project collaborations with Northern NGOs tended to bind significant personnel resources for documenting and research tasks—resources that, in turn, could not be invested in union building and worker organising.

The focus on building relations with many workers at the community level and the resulting lack of a strong associational power base inside factories also influenced how unions constructed spaces of contestation under the community organising, fire-fighting approach. As the term ‘fire-fighting’ suggests, unions mainly constructed spaces of contestation in a reactive manner by addressing labour rights violations reported to the union by individual workers. In the face of the constraints for unions to exercise leverage on employers through industrial action, unions relied primarily on the moral power of transnational consumer campaigns to harness the leverage of retailers over garment manufacturers and thereby achieve corrections of labour rights violations in specific factories. However, as the empirical analysis has shown, relying on the borrowed moral power of transnational consumer campaigning networks had limited potential in building lasting union bargaining power vis-à-vis employers. The scope of issues that could be addressed through transnational consumer campaigning strategies was limited to reacting to particularly harsh labour rights violations. Therefore, transnational consumer campaigns alone did not enable unions to pressure employers into collective bargaining processes and achieve improvements for workers beyond the mere implementation of minimum labour standards.

In the face of the limitations of the community organising, fire-fighting approach for building union bargaining power vis-à-vis employers, the three Bangalore garment unions have—to varying extents—implemented a different strategic agency approach over the past years. This approach prioritises the factory as space for organising and seeks to foster collective bargaining with employers. Under this strategic factory organising, collective bargaining approach, unions have started to concentrate their organising activities on selected target factories to build strong ties with a core group of workers from each factory. These workers then act as worker representatives and organisers inside the factory. Complementary to this shift in spaces of organising from the community to the workplace, two of the three Bangalore garment unions have constructed new spaces of collaboration. Instead of focussing their resources on constructing alliances with transnational consumer networks and NGOs, these unions are now building collaborations with labour organisations and networks. In this context, the empirical analysis has exemplified that such collaborations between labour actors can represent important spaces for union organisers and workers to gain strategic knowledge—for example, about the value chain structure—and to develop strategic capacities. These capacities include, for example, strategy development capacities for designing networked agency strategies that use transnational solidarity only as an instrument to open up spaces for organising and collective bargaining in target factories.

As a result, with the shift to a factory organising, collective bargaining approach, unions have shifted the practices through which they construct spaces of contestation as well. With increased associational power inside factories and enhanced capacities to engage in collective action, the workplace has become the central arena in struggles for collective bargaining agreements. Transnational campaigning strategies and labour solidarity, in turn, are employed by unions increasingly only as a secondary source of leverage. Following such a networked agency approach, the Bangalore garment union GATWU signed in 2019 the first collective bargaining agreement in the history of the union. As opposed to previous victories under the ‘fire-fighting’ approach, with this bargaining agreement, Avery Dennison not only committed to refrain from union-busting practices but also to grant workers benefits and wages beyond the legally prescribed minimum (see Sect. 7.1.2).

However, it is important to note that in the everyday practices of all three garment unions, up to date, tensions exist between practices rooted in the community organising, fire-fighting approach and unions’ declared strategic goals of factory organising and collective bargaining. The two presented stylised models of strategic union agency approaches, hence, need to be understood as opposite ends of a spectrum, with Bangalore garment unions currently finding themselves somewhere in between. Nevertheless, the findings from the empirical analysis have provided important insights into the potentials and limitations of the two agency strategies for building sustained union bargaining power.

Together with the insights from the analysis of the labour control regime and resulting constraints for union agency, this study, hence, makes several critical empirical contributions to debates in labour geography and GPN analysis about how to build sustained union power in garment producing countries. These empirical contributions will be discussed in the next section.

2 Empirical Contributions of This Study: Lessons for Building Local Union Power in Garment Producing Countries

Which lessons can we learn from the analysis presented in this study for building local union power and improving working conditions in the global garment industry? Four essential teachings shall be pointed out here that are valuable for local unions in garment producing countries as well as for labour rights and consumer organisations in the Global North concerned with improving labour conditions in the garment industry.

First, the empirical analysis has highlighted that a significant challenge for improving working conditions in the global garment industry results from the highly complex structural frameworks for labour exploitation and capital accumulation at the various nodes of the garment GPN. These structural frameworks—designated in this study as labour control regimes—are constructed and reproduced through the intertwined practices of a multitude of actors located in more or less distant places, who all seek to extract and appropriate surplus value from living labour (c.f. Cumbers et al. 2008). As a result, to challenge institutionalised frameworks for labour exploitation, local unions need to develop networked agency strategies, which allow unions to target capital and state actors simultaneously at multiple levels. This lesson contradicts the arguments made by earlier studies in labour geography that different types of labour control regimes in garment producing countries—e.g. market, state, employer control regimes—are conducive to different agency strategies by local unions, e.g. wildcat strikes, engaging in multi-stakeholder initiatives and transnational labour organising (see e.g. Anner 2015a). Instead, the findings from this study highlight how local unions—in the face of complex, networked labour control regimes—need to intertwine all of the aforementioned strategies to contest and transform structural relations of exploitation (see also Tufts 2007; Wills 2002). In this context, this study has exemplarily illustrated how the local garment union GATWU has achieved stopping employer and state practices of undermining and circumventing statutory minimum wage revisions through a networked minimum wage campaign (see Sect. 7.1.1). Strategic actions within the campaign comprised (1) conducting symbolic protests in the workplace, (2) holding public protests at the local level, (3) filing a lawsuit to contest illegal state practices of withdrawing already issued minimum wage notifications and (4) harnessing the leverage of lead firms over manufacturers through transnational consumer campaigns. It was only through the combination of all these actions that GATWU was able to stop the interrelated set of employer and state practices that had prevented the implementation of mandatory minimum wage increases. Hence, this study makes a case for a heightened sensitivity towards the networked character of capital and state-produced labour control structures. It has shown that no scale can be singled out as particularly dominant within the local labour control regime in the Bangalore export-garment cluster. Consequently, unions must develop agency approaches that combine and strategically intertwine actions at various scales.

Second, the study has illustrated that in countries or regions where the garment industry is characterised by a highly feminised workforce or by a high share of migrant, first-generation industrial workers, traditional workplace organising strategies focusing exclusively on economic demands may not be conducive. In these contexts, it can be helpful for unions to construct initial relationships with workers through organising practices that address workers not only as wage workers but also as women, mothers, daughters, migrants and community members (see also Doutch 2021; Jenkins 2013). Such organising practices can help to raise workers’ awareness of intersecting lines of structural exploitation along categories of class, gender and geographical origin and thereby to foster workers’ collective mindset—an essential precondition for building associational and organisational power (c.f. Lévesque and Murray 2010). Therefore, the findings from this study align with the arguments of prior work by labour geographers and researchers stressing the potential of ‘community’ or ‘social movement’ unionism approaches for building union power in countries of the Global South (see e.g. Moody 1997; Nowak 2017). In contrast to prior work, this study, however, also highlights the risks that come with organising strategies that prioritise the community as space for organising at the expense of more targeted workplace organising strategies: While community organising strategies can enable unions to build associational power at the local level, they do not enable unions to build a strong membership base inside the factory. However, such a membership base inside the factory is necessary for unions to be able to exercise workplace bargaining power through industrial action. As a result, unions that rely exclusively on a community organising strategy are likely to limit their scope of action to correcting individual labour rights violations since they do not possess the necessary workplace bargaining power to engage in proactive collective bargaining. Therefore, to achieve improvements for workers beyond the implementation of basic minimum labour standards, community organising practices need to be combined with a targeted workplace organising strategy.

Third, the results of this study call for heightened sensitivity to the mixed effects that collaborations between local unions in garment producing countries and NGOs in consumer countries from the Global North have on local unions’ capacities for building sustained bargaining power vis-à-vis employers. This lesson is particularly relevant since the global garment industry has seen a proliferation of such North-South collaborations since the 1990s with the rise of transnational anti-sweatshop movements and multi-stakeholder initiatives (Zajak et al. 2017: 914; see also Anner 2015b; Esbenshade 2004; Fütterer and López Ayala 2018; Hauf 2017; Merk 2009). At the same time, this study shows that while such collaborations can help local unions to access and leverage different types of coalitional power resources (e.g. financial resources from funded projects, moral power resources from consumer campaigns), North-South collaborations can also hamper unions’ abilities to develop organisational and associational power resources. Financial flows from Northern NGOs to local unions in garment producing countries are often linked to strict accountability regimes requiring unions to document all activities and expenses. As a result, union organisers invest significant time into activities that do not directly contribute to the union-building process. Moreover, when unions sustain themselves primarily through external funding and not through membership fees, the union leadership becomes primarily accountable to external donors instead of being primarily accountable to the union members. This shift from internal accountability to external accountability of the union leadership, in turn, hampers internal union democracy. This study has hence provided further evidence for arguments from past studies that external funding for local unions can be a double-edged sword (see e.g. Banse 2016; Fink 2014).

Similarly, this study has pointed out potential hampering effects for building local unions’ associational and organisational power resources linked to local unions’ engagement with transnational consumer campaigning networks. Especially when unions rely exclusively on moral power resources from consumer campaigns and on the leverage of retailers to compensate for a lack of associational power in the workplace, this strategy may create a path dependency that limits unions’ capacities to engage in workplace organising and collective bargaining. In transnational campaigns, NGOs from the Global North take on strategic planning and decision-making capacities while the role of unions and workers is reduced to providing information about and documentation of labour rights violations. As a result, neither unionists nor workers develop the strategic capacities that are necessary to build the unions’ associational and organisational power resources in the long term. However, as this study and others have pointed out, victories achieved through transnational campaigning strategies tend to be temporary and limited to correcting particularly harsh labour rights violations (see e.g. Anner 2015b; López and Fütterer 2019; Ross 2006). Therefore, building associational and organisational power resources is central for unions to lastingly shift the capital-labour power balance in favour of workers and achieve lasting and comprehensive improvements for workers (see also Kumar 2014, 2019; Oka 2016).

As a result of the third lesson, this study has, fourth, highlighted the importance of transnational collaborations and union strategies that prioritise the development of unionists’ and workers’ strategic capacities and collective experiences as a vehicle for building sustained union bargaining power. In this regard, this study has, on the one hand, showcased the potential of transnational collaborations with other labour organisations that—as opposed to collaborations with consumer organisations—prioritise union building and collective bargaining processes as a long-term goal. In doing so, these collaborations can provide network spaces for mutual learning, planning and strategy development as well as for the exercise of transnational solidarity to combat union-busting practices, and thereby open up wiggle room for local union building and collective bargaining in garment producing countries (see also Lohmeyer et al. 2018; López and Fütterer 2022). On the other hand, this study has highlighted the importance of unions in garment producing countries to actively involve workers in planning and decision-making processes in specific labour struggles and to foster collective experiences of resistance. Such experiences are, in turn, essential to build the associational and organisational power resources that will allow a union to shift the power balance between employers and unions in the long term. In this context, this study has shown that in cases where unions do not have sufficient associational power at the factory level to engage in industrial action, symbolic protests in the workplace can be an instrument for fostering workers’ active participation and collective resistance experience in a specific labour struggle. Developing workers’ strategic capacities is particularly important to mitigate professionalisation processes that lead to the increasing centralisation of decision-making and planning competencies on full-time union staff (see e.g. Choudry and Kapoor 2013; Fink 2014; Fütterer and López Ayala 2018). Such a centralisation of strategic capacities weakens unions since they hamper the development of a strong second-rank leadership in the workplace, which can serve as a nucleus for organising and which can negotiate everyday problems with the management independently.

In summary, this study has highlighted that to achieve lasting improvements for workers, unions not only need to construct collaborations with external actors in ways that foster unionists’ and workers’ strategic capacities but also need to construct internal union relations in horizontal and democratic ways. In local contexts characterised by a highly feminised workforce and strong patriarchal social relations—as in many Asian garment producing countries—fostering horizontal and democratic internal union relations may, therefore, also require unions to actively combat internal gendered power asymmetries by fostering women leadership (see also Doutch 2021; Evans 2017).

3 Theoretical Contributions of This Study: Producing New Insights Through a Relational Analytical Perspective

Besides offering important empirical findings regarding the challenges and strategies for building sustained union bargaining power and improving working conditions in garment-producing countries, this study has made several theoretical contributions to current debates within economic and labour geography. Specifically, the relational analytical approach presented in this book contributes to advancing theoretical concepts and debates within three strands of research in economic and labour geography: (1) research on labour in GPNs; (2) GPN analysis more generally and (3) practice-oriented research in economic geography.

3.1 Contributions to Research on Labour Control and Labour Agency in GPNs

Most importantly, by introducing a relational approach as an alternative to dominant scalar approaches, this book centrally advances the theoretical discussion of labour control and labour agency in GPNs. As illustrated in the literature review (Chapter 2), the dominance of scalar heuristics has limited past studies’ capacity to recognise the deeply relational nature of the multi-scalar ‘labour control architectures’ underpinning GPNs and of workers’ multi-scalar agency strategies in GPNs. The relational analytical approach developed in this book achieves to overcome these limitations by shifting the analytical focus from pre-defined scales to networks of relationships as a central heuristic.

As a result, the relational analytical approach developed here firstly, achieves to overcome a crucial limitation of past studies on labour control regimes in GPNs regarding their ability to grasp the socio-spatial relations underpinning specific local labour control regimes. These studies have tended to presuppose a universal, hierarchical nested scalar order as characteristic of all labour control regimes in GPNs. This presupposition has limited past studies’ capacity to map the empirically existing socio-spatial relations that constitute labour control regimes. In contrast, the relational heuristic framework for studying labour control in GPNs introduced in this book leaves analytical space for carving out the individual and place-specific socio-spatialities of labour control regimes at different nodes of a GPN. Instead of seeking to fit dynamics and relations of labour control into pre-given scalar categories, the proposed relational framework takes empirically existing practices, relations and their interrelations as an analytical point of departure and maps their spatial extensions and characteristics. As a result, the here-developed relational approach is more sensitive to the different ways, in which geographically more delimited and spatially more unbounded processes and relationships of labour control shape and enable each other. It, therefore, provides an apt tool for addressing recent calls from labour geographers who have called on researchers to pay increased attention to the “mix of geographically distant and proximate relationships across different scales” (Wickramasingha and Coe 2021) that characterise labour control regimes in GPNs.

Second, the relational approach developed in this study sheds light on another aspect that has remained understudied in past research on labour control regimes, namely the dialectical relationships between labour control and labour agency (see e.g. Hastings and MacKinnon 2017; Wickramasingha and Coe 2021). Past studies working with scalar heuristics have tended to conceptualise labour control regimes as structural contexts at various levels that are unilaterally imposed on local workers (see e.g. Pattenden 2016; Smith et al. 2018). In contrast to this dominant ‘top-down’ conceptualisation, the here-developed relational approach stresses that labour control regimes as structural contexts are constructed through practices and relationships that are situated in space and time and can be challenged and transformed by workers and unions. Consequently, the here-developed relational, practice-oriented analytical approach can shed light on the ‘small transformations’ (Latham 2002) of labour control practices and relations achieved by workers and unions through strategies of reworking (Cumbers et al. 2010). Even though such strategies may not challenge hegemonic power and capitalist exploitative relations per se, they may still recalibrate local power relations and thereby redistribute resources in favour of workers. Therefore, the relational, practice-oriented framework developed in this book allows producing analyses that are sensitive to how labour control regimes as institutional frameworks for capital accumulation are produced and continuously transformed in a “dialectical process of interaction” between capital, state and labour actors (Hastings and MacKinnon 2017: 104).

Third, the relational approach to union agency developed in this book can generate enhanced insights into how different processes and relationships of labour control at various stages enable and shape each other and how unions can strategically intertwine actions at various levels. Thereby, the relational approach presented here has expanded the scope of past studies of labour agency in GPNs, which have tended to adopt a one-sided focus on the ‘up-scaling’ of local labour struggles to the international level while neglecting other scales of agency (see e.g. Anner 2015b; Merk 2009). By giving visibility to how unions deploy strategic actions at various levels and how these interplay in building unions’ bargaining power, the here-introduced relational approach facilitates understanding labour’s networked agency “as constituted by interdependent scales of action that are not nested in a hierarchy privileging one scale over another” (Tufts 2007: 2387). Such an understanding heightens researchers’ sensitivity towards the structural effects that collaborations with external actors at various levels may have on internal union relations and practices (see also Zajak et al. 2017). As has been shown in the previous section, only by analysing different scales of action as interrelated can researchers evaluate which types of union collaborations have enabling effects and which types of collaborations have constraining effects for fostering workers’ and unionists’ strategic capacities.

Fourth and last, the relational approach to labour agency developed in this book allows tackling a blind spot in past research on union agency in GPNs. In past research, internal union relations and their intersections with broader social relations have largely remained a black box (see also Cumbers 2015). By conceptualising internal union relations as a vital dimension of union agency, the relational approach to union agency developed in this book opens this black box. Thereby, it allows for a critical analysis of unions’ everyday practices not only with regard to constructing alliances with external actors but also with regard to constructing internal relations between union leadership and members as well as between union members and non-members. As a result, the here-developed relational approach also sharpens our understanding of the intersections of internal union relations with other social relations, such as relations of gender or geographical provenience and the power structures enshrined in these relations. It can, therefore, refine our understanding of the embeddedness of labour agency not only within the structural-relational formations of capital and the state but also within broader, place-specific socio-cultural relations (Coe and Jordhus-Lier 2010; Doutch 2021; Hastings 2016).

3.2 Contributions to GPN Analysis

Beyond providing new insights into the dynamics of labour control and labour agency in GPNs, the relational approach developed in this study contributes to reviving and reinvigorating a relational perspective within GPN analysis more broadly. As laid out in Sect. 2.1, early work with the GPN approach was underpinned by a profoundly relational understanding of the global economy as constituted through networked vertical and horizontal relations of production, exchange and consumption (c.f. Dicken et al. 2001). However, in the further evolution of GPN analysis, there has only been sporadic engagement with this incipient relational analytical perspective (Cumbers 2015). As a result, GPN scholars have recently voiced critique towards many (self-attributed) GPN-studies for focussing exclusively on territorial dynamics at the horizontal dimension without systematically exploring their interconnections with “the configuration and operation of the global production network in question” (Coe and Yeung 2019: 788; see also Yeung 2020).

Against this background, the relational approach developed in this book firstly provides an innovative analytical framework for producing empirically grounded demonstrations of the “causal links between […] network dynamics and territorial outcomes” (Coe and Yeung 2019: 778). The focus of this study has been on analysing the conditions and role of labour within GPNs. Nevertheless, the here-developed conceptual approach of analysing place-specific territorial outcomes within GPNs through the lens of interwoven network and territorial relations laden with power can enrich other research areas in GPN analysis as well, such as research at the intersection of GPN analysis and political ecology (see e.g. Bridge and Bradshaw 2017; Dorn and Huber 2020; Irarrázaval and Bustos-Gallardo 2019) or research on the development effects of integrating specific places into GPNs (see e.g. McGrath 2018; Tessmann 2018; Vicol et al. 2019).

Second, the relational approach developed in this book also contributes to a more nuanced understanding of “the relational, networked and institutional qualities of how power is generated and ultimately exercised” in GPNs (Hess 2008: 456; see also Arnold and Hess 2017; Raj-Reichert 2020). Whereas the GPN framework has traditionally conceptualised power predominantly as a static resource held by specific actors within a GPN (Henderson et al. 2002), this study has highlighted that power in GPNs is profoundly dynamic, relational and networked. Power in GPNs is relational, since it only becomes effective in shaping material outcomes when actors exercise it in relation to other actors. Moreover, power within GPNs has a networked character when actors strategically direct flows of power within networks of relationships to exercise leverage over other actors. This study has illustrated networked power on the example of Bangalore-based unions’ strategies of leveraging the influence of geographically distant lead firms, consumer and worker groups from the Global North to shift the local power balance with employers. By foregrounding the relationships through which power flows and is exercised, the here-developed relational approach enhances our understanding of the complex power flows within GPNs that “reach and stretch across distances where the lives of others far away are shaped by those nearby and vice versa” (Raj-Reichert 2020: 654).

3.3 Contributions to Practice-Oriented Research in Economic Geography

Last but not least, the relational analytical approach to labour control regimes and labour agency developed in this book also contributes to advancing practice-oriented research in economic geography: It provides a novel conceptual tool for theorising the links between micro-scale practices and macro-scale social and economic phenomena. As practice-oriented economic geographers have reiterated: the value but also the central challenge for practice-oriented research lies in demonstrating how “higher order phenomena”, such as institutions or class structures, “are enacted, reproduced, and/or transformed through the everyday actions embedded within them” (Jones and Murphy 2010: 372; see also Everts 2016; Wiemann et al. 2019). In this vein, by emphasising the links between the manifold labour control practices and relations that constitute the structural context for worker and union agency in GPNs, the here-developed analysis shows “how context, structures, and individual agency or action come together in the doing of economic and industrial activities” (Jones and Murphy 2010: 3050).

4 Final Reflections and Directions for Further Research

This book highlights the challenges for improving working conditions in the global garment industry, which result from uneven power relations between multinational retailers and local manufacturers on the one hand, and between manufacturers and workers, on the other. Despite an increasing consolidation of supplier networks and the emergence of large tier one suppliers over the past 15 years, spatial asymmetries between retailers and suppliers persist. Retailers continue to maintain large and geographically dispersed networks of suppliers and to establish new sourcing relations with manufacturers in ever lower-wage countries, such as Myanmar and Ethiopia. Garment manufacturers, in contrast, are usually forced to concentrate the largest share of their business on a few key buyers due to variations in buyers’ technical and social standards and requirements. As a result, especially large retailers like H&M, Inditex or G.A.P are still able to ‘squeeze’ manufacturers by demanding lower prices, shorter lead times and enhanced flexibility from their suppliers.

Manufacturers pass on the pressures for reducing costs while increasing productivity to workers through a complex web of disciplining and exploiting practices. Whereas strong state control and regulation in garment producing countries could help to mitigate worker exploitation, such a regulatory role of the state conflicts with the aim of governments in many garment producing countries to boost economic development through providing a business enabling environment for capital. Due to its ability to generate employment for the unskilled, rural population and to attract foreign investments, the export-garment industry enjoys a special status in many industrialising countries. As a result, not only in India but also in other garment producing countries, the state and public institutions do not provide a counterweight to the dominance of employers over labour. Instead, they frequently tolerate or actively support illegal employer practices of exploiting and disciplining, such as wage theft or union busting (see e.g. Anner 2022; Hossain 2019; Wickramasingha and Coe 2021).

In light of these complex entanglements of intersecting relationships of domination at the vertical, ‘network’ dimension of the garment GPN and at the horizontal dimension, i.e. within individual garment producing countries, it becomes evident that ‘soft’ regulation attempts through codes of conducts and ethical trading initiatives alone can only have a limited effect with regard to improving working conditions. Without strong pressure from labour and consumer organisations, capital and state actors have little incentive for effectively putting the social standards defined in the context of such initiatives into practice. At the same time, this study has highlighted once more that interventions of consumer organisations without the presence of strong local unions can, at best, contribute to correcting and mitigating particularly cruel violations of workers’ rights. In contrast, to bring about lasting improvements of garment workers beyond minimum labour standards, strong labour movements are needed that can shift the capital-labour power balance in favour of workers. Shifting the capital-labour power balance in garment producing countries is particularly important to improve workers’ wages, which still remain below subsistence levels in most Asian countries. Where statutory minimum wages have been increased in past years, these raises have usually been the result of sustained worker campaigning and strike action, for example, in Bangladesh (Wickramasingha and Coe 2021), Cambodia (Lawreniuk and Parsons 2018) and India (see Sect. 7.1). At the same time, in the face of the highly feminised and informalised nature of work in the garment industry, unions also need to tackle internal patriarchal structures of domination and develop innovative organising approaches that address the struggles of women and informal workers beyond the sphere of production (Doutch 2021; Evans 2017).

In this light, the scope of this and other studies on labour organising in garment production countries consists of a relatively narrow focus on the struggles of workers labouring in tier one garment factories acting as direct suppliers for transnational retailers. Less attention has been paid so far to the challenges and strategies for organising workers in the subcontracted tier two and three segments of the Asian export-garment industry, where work is frequently carried out in the form of piece-based, informal homework production arrangements (see e.g. Mezzadri 2016; Neve 2014). As a result, the labour process in these lower tiers tends to be characterised by a higher level of spatial segmentation and a dilution of the employee-employer relationship, with workers being formally self-employed and relationships with factories often being mediated through a chain of intermediaries. These organisational characteristics bring along distinct challenges for collective organisation, representation and bargaining. At the same time, home-based, subcontracted workers represent the weakest link in the garment value chain, since they fall through the cracks of both social auditing regimes and state regulation. Work in the subcontracted tier two and three segments of retailers’ supplier networks is, therefore, often characterised by an even higher level of precarity and insecurity than in retailers’ direct supplier factories.

Against this backdrop, a stronger engagement of researchers and unions with workers in informal settings and the conditions that constrain and foster these workers’ collective agency is needed. The relational, practice-oriented approach for analysing labour control and labour agency in GPN developed in this book can provide a conceptual starting point for such an engagement by labour geographers and other researchers.