The Relationship between Economic and Social Upgrading in Global Apparel Production Networks and the Role of Institutions
By Henk Oosterhof
University of Groningen Faculty of Economics and Business
Nettelbosje 2 9747 AE Groningen
Newcastle University
Newcastle University Business School 5 Barrack Road
Newcastle upon Tyne NE1 4SE
Student numbers: S2408155*/B4017401**
Degree courses: MSc. International Business & Management* MSc. Advanced International Business
Management & Marketing**
2
ABSTRACT
A growing body of research has paid attention to economic and social upgrading in global production networks (GPNs) and the conditions under which the former translates into the latter. Their relationship, however, remains unclear and the role of institutions has been under-researched so far. This research analyzes the relationship between economic and social upgrading by examining how economic upgrading affects buyer-supplier relationships, whether possible changes in buyer-supplier relationships resulting from economic upgrading lead to social upgrading, and by investigating the role of institutions in the relationship between economic and social upgrading. It does so by using a multiple case study design in Bangladeshi and Indian apparel GPNs. Three types of economic upgrading are being addressed in this research: process, product and functional upgrading. It is found that process upgrading and its associated higher leverage over competitors can lead to minor improvements in terms of social upgrading, that there is no direct relationship between product upgrading and social upgrading and that functional upgrading may lead to higher wages, resulting from skill upgrading. Furthermore, this study argues that institutions play a constraining rather than a promoting role in the translation of economic upgrading into social upgrading. This research, additionally, identifies other factors that might influence economic and social upgrading.
3
ACKNOWLEDGEMENTS
I would like to thank several people who provided me with help, advice and support during my dissertation process. First of all, I want to express my appreciation to the organizations and the people that participated in my research. Without their time and effort, this study would not have seen its completion. Additionally, special thanks go out my supervisors Dr. Hanna Bahemia and Dr. Miriam Wilhelm. I am thankful for their guidance, their constructive feedback and the fact that they were in good communication with me, as well as with each other. Finally, I want to thank my family for their support during my studies.
4
The Relationship between Economic and Social Upgrading in Global Apparel Production Networks and the Role of Institutions
TABLE OF CONTENTS 1. INTRODUCTION... 6 2. LITERATURE REVIEW ... 9 2.1 GPNs and GVCs ... 9 2.2 Economic upgrading ... 11 2.3 Social upgrading ... 16
2.4 Factors accounting for economic and social upgrading ... 17
2.5 Institutions; an under-researched dimension in GPN and GVC analysis ... 21
2.6 Research model ... 23 3. METHOD ... 25 3.1 Research strategy ... 25 3.2 Case selection... 25 3.3 Data collection ... 26 3.4 Data analysis ... 28 3.5 Ethical issues ... 28
4. WITHIN-CASE ANALYSIS AND CROSS-CASE ANALYSIS ... 30
4.1 Within-case analysis Bangladesh ... 30
4.1.1 History and development ... 30
4.1.2 Economic upgrading in Bangladesh ... 32
4.1.3 Social upgrading in Bangladesh ... 34
4.1.4 The role of institutions in Bangladesh ... 36
4.2 Within-case analysis India ... 40
4.2.1 History and development ... 40
4.2.2 Economic upgrading in India ... 41
4.2.3 Social upgrading in India ... 43
4.2.4 The role of institutions in India ... 44
4.3 Cross-case analysis ... 48
4.3.1 Economic upgrading ... 48
4.3.2 Social upgrading ... 49
4.3.3 The role of institutions ... 51
4.4 Other factors influencing economic and social upgrading ... 55
5. DISCUSSION ... 60
5.1 Working propositions and interpretation of the findings ... 60
5.2 Contributions... 66
6. CONCLUSIONS ... 68
6.1 Conclusion ... 68
5
REFERENCES ... 72
APPENDICES ... 84
Appendix A: General interview guide Bangladesh ... 84
Appendix B: General interview guide India ... 87
Appendix C: Interview Participant A1 ... 90
Appendix D: Interview Participant A2 ... 100
Appendix E: Interview Participant B ... 109
Appendix F: Interview Participant C1 ... 115
Appendix G: Interview Participant C2 ... 123
Appendix H: Interview Participant D ... 131
Appendix I: Interview Participant E ... 136
Appendix J: Interview Participant F ... 150
6
1. INTRODUCTION
Markets have increasingly become globalized (Baldwin, 2006) and supply chains have become geographically dispersed to a growing extent (Baldwin & Lopez-Gonzalez, 2013). This globalization of markets has been driven by trade liberalizations and reductions in transportation, communication, and coordination costs (Baldwin, 2006). These developments have led to the emergence of global production networks (GPNs) (Henderson, Dicken, Hess, Coe & Yeung, 2002; Coe, Hess, Yeung, Dicken & Henderson, 2004; Dicken, 2005). By participating in GPNs, developing country firms carry out the low-skill tasks of developed country firms. By offshoring these tasks, developed country firms transfer knowledge, technology and personnel, required to perform these tasks effectively, to developing country firms. As a result, developing country firms that are now doing those low-skill tasks industrialize at a remarkable pace and this has various implications for developing countries (Rossi, 2010; Bernhardt & Milberg, 2011).
The implications that GPNs have on developing countries can be divided into economic and social implications. From an economic perspective, participation of developing country firms in GPNs can lead to improvements in terms of innovation and learning, resulting from interactions with developed country firms (Rossi, 2011). This is called economic upgrading, which has become synonymous with the massive shift of production to developing countries (Milberg & Winkler, 2011). Much research in the field of GPNs has been done on economic upgrading, linking economic upgrading to international trade performance (Milberg & Winkler, 2010). However, not much attention has been paid to what economic upgrading means for the employees who are instrumental to enhanced economic performance (Milberg & Winkler, 2010; Rossi, 2011).
7
In order to help understand the relationship between economic and social upgrading in GPNs and the conditions under which economic upgrading leads to social upgrading, this research analyzes the relationship between economic and social upgrading by examining how economic upgrading affects buyer-supplier relationships in global apparel production networks (GAPNs), whether possible changes in these relationships lead to social upgrading, and by investigating the role of institutions in the relationship between economic and social upgrading. The outcomes of GPNs in terms of economic and social upgrading are determined by numerous factors (Rossi, 2011). The institutional environment is one of them (Rossi, 2011) and has not adequately been addressed in GPN analysis (Coe, Dicken & Hess, 2008b; Levy, 2008; Bair, 2009a; Smith, 2014). For comparative purposes and so as to understand the different (institutional) factors shaping GPNs and playing a role in the relationship between economic and social upgrading, this research uses a multiple case study design in which Bangladeshi and Indian GAPNs are analyzed.
Existing research has examined whether economic upgrading leads to social upgrading. However, their relationship remains blurred and the role that institutions play in the translation of economic upgrading into social upgrading is neglected. In order to contribute to filling these research gaps, the following research question needs to be answered: What is the
relationship between economic and social upgrading in Bangladeshi and Indian apparel GPNs, what is the role of institutions in this relationship and what other conditions might influence this relationship?
8
lot of criticism on the offshoring of production to developing countries and the conditions employees work in. However, this criticism may be unwarranted if offshoring leads to social upgrading. In this case, offshoring may be a first step to improving people’s lives. Insight into the conditions under which social upgrading occurs is therefore of utmost importance. Furthermore, policy makers have the obligation to take into account the conditions that need to be present in order for economic upgrading to be translated into social upgrading. An understanding of how institutions influence this relationship will help governments, non-governmental organizations, and businesses design policies that boost living standards of employees in developing countries. Institutional reform may be required in order for economic upgrading to be translated into social upgrading, therefore, insight into the role that institutions play in this relationship is necessary.
9
2. LITERATURE REVIEW
This chapter introduces and defines the concepts involved. First, the concepts that have been used to analyze global production are discussed, arguing that both GPNs and GVCs are relevant to this research. Second, economic and social upgrading, two implications resulting from participation in GPNs, and their linkages, are described. Third, institutional theories are discussed. Finally, this section ends with a figure illustrating the core concepts and relationships that will be investigated in this research.
2.1 GPNs and GVCs
In today’s world of globalization, economic and social upgrading has become associated with developing country firms participating in global networks (Bernhardt, 2013). These networks account for a rising share of GDP and employment, and link firms, workers and consumers from all over the world and provide means for developing country firms to gain foothold in global markets (Gereffi & Fernandez-Stark, 2011). By participating in global networks, developing country firms are able to move up the value chain and their employees, subsequently, might benefit from an increase in firm performance (Gereffi & Fernandez-Stark, 2011; Rossi, 2011). The literature on global production uses several closely related terms to describe firm networks in which economic and social upgrading take place. This section discusses these concepts and clarifies why this research uses the GPN concept as its main analytical framework and why global value chains (GVCs) still matter as well.
Fernandez-10
Stark, 2011: 4). By replacing the term “commodity” by the term “value”, GVCs allowed the analysis to go beyond commodities (Bair, 2009a). Additionally, the GVC framework involves more complex governance modes (Sturgeon, 2009). Originally, in his GCC framework, Gereffi made a distinction between two chain types (producer-driven and buyer-driven); however, a new typology was required which would incorporate new types of chain modes (i.e., markets, modular modes, relational modes, captive modes and hierarchy) (Gereffi, Humphrey & Sturgeon, 2005). Most studies using the GVC framework have paid attention to input-output structures and governance modes (Bazan & Navas-Alemán 2004). Therefore, the global production network (GPN) emerged as a critique of the GVC framework’s lack of attention to institutions (Bair, 2009a). The institutional dimension has been incorporated in the GVC framework, however, little indication was provided of how it should be perceived (Thomsen, 2007). Gibbon, Bair & Ponte (2008) argue that actors external to the chain and directly influencing the actions of economic actors operating within the chain should be regarded as key factors in GVC analysis (Gibbon et al., 2008). GPNs aim to touch on this criticism by integrating the GVC framework into the social and institutional context in which social actors, being part of the value chain, are embedded. The GPN’s definition is based on the recognition that people’s behavior and economic activity are strongly affected by the social and institutional context (Granovetter, 1985). GPNs differ from GVCs in that “GPNs do not only connect firms functionally and territorially, but they also connect aspects of the social and spatial arrangements in which firms are embedded and which influence their strategies and the values, priorities and expectations of managers, workers and communities alike” (Henderson et al., 2002: 451). Besides vertical relations between lead firms and their suppliers, GPNs also take into consideration horizontal relationships and all the actors impacting global production such as governments, NGOs, trade unions and workers (Bair, 2009b).
11
a network maps both the vertical and horizontal linkages between economic actors, i.e., recognizing that various value chains often share common economic actors and are dynamic in that they are reused and reconfigured on an ongoing basis”. Despite the differences between GPNs and GVCs, there is also a noticeable similarity. Both concepts argue that power asymmetries and governance modes have a significant influence on the upgrading prospects of actors operating within networks and chains (Coe, Dicken & Hess, 2008a).
Despite the theoretical differences, GPNs and GVCs are often used interchangeably and research spawned by GPN literature does usually not differ in practice from GVC literature (Coe et al., 2008b; Levy, 2008; Bair, 2009a). In order to provide clarification, this research argues that the term GPN should be considered an all-encompassing framework, which includes multiple GVCs (Sturgeon, 2000) and takes into account the social and institutional context and all the economic actors operating in it such as governments, NGOs, trade unions and workers (Bair, 2009b). It should be noted, though, that economic upgrading directly takes place within GVCs and only indirectly within GPNs, as GVCs are a subset of GPNs (Sturgeon, 2000). However, as the scope of GVCs is limited to the dyadic level between buyer and supplier firms and does not take into consideration the implications for workers (Rossi, 2011), social upgrading only takes place within GPNs. Therefore, both concepts are relevant in this research; GVCs are useful for exploring the dynamics in vertical relationships between lead firms and suppliers and GPNs are useful for exploring the implications for workers. However, as the GPN concept constitutes an all-encompassing theory, this research selects the GPN construct as its main analytical framework.
2.2 Economic upgrading
Since economic upgrading directly takes place within GVCs and only indirectly in GPNs, as argued in the previous section, economic upgrading will be further clarified by referring to the GVC literature.
12
resulting from reorganizations of the production system, and/or the use of new technologies. The shift toward the production of a more sophisticated product, which is of higher value, is referred to as product upgrading. Functional upgrading is a transition in the mix of tasks carried out by the supplier, subsequently leading to a higher value added supply relationship. This could, for instance, imply that a supplier moves from assembly to manufacturing or from manufacturing to design or branding. Of all types of economic upgrading, functional upgrading is perceived to be the most important step in order to gain a higher share of value added. Finally, chain upgrading is the movement toward a higher value added production chain which occurs when a supplier applies the competences and knowledge gained in one chain to another chain. For instance, a supplier can use knowledge gained in producing TVs so as to move into the computer industry by producing monitors (Humphrey & Schmitz, 2002).
13
Table 1: Key determinants of GVC governance
Source: Gereffi et al. (2005)
Captive value chains, for instance, provide opportunities for process and product upgrading but offer little prospects for functional upgrading (Bazan & Navas-Aleman, 2004; Humphrey & Schmitz, 2000; Schmitz, 2006). Also, lead firms might be more willing to foster upgrading in GVCs where intense interaction is necessary to transfer complex knowledge (Giuliani, Pietrobelli & Rabellotti, 2005), for example in relational value chains. In market value chains, there is little interaction between lead firms and suppliers, thereby limiting process and production upgrading. However, the occurrence of functional upgrading seems more likely in this type of value chain (Humphrey & Schmitz, 2002). In modular value chains, lead firms do not contribute directly to the learning processes of suppliers. However, these suppliers are required to comply with fully specified standards, which indirectly provide an inducement for learning (Morrison, Pietrobelli & Rabellotti, 2008).
14
Figure 1: Curve of Value-Added Stages in the Global Apparel Value Chain
Source: Frederick (2010)
15
Table 2: Upgrading trajectories in the Global Apparel Value Chain
Source: Frederick (2010)
16
2003). This is, however, not necessarily the case as firms might, for instance, move directly to functional upgrading. Please note that this section has not elaborated on chain upgrading as this research only focuses on one particular GVC, i.e., the GAVC. Taking into account chain upgrading would require a wider perspective and necessitate the study of multiple types of GVCs and GPNs. Therefore, chain upgrading does not fit within the scope of this research.
2.3 Social upgrading
GPNs provide a wider perspective than GVCs and include implications for workers as well (Bair, 2009b; Barrientos et al., 2010). These implications are referred to as social upgrading/downgrading. Social upgrading is “the process of improvements in the rights and entitlements of workers as social actors by enhancing the quality of their employment” (Rossi, 2011: 61). These improvements can manifest themselves in many forms such as access to better work, healthier working conditions and the improvement of social rights. In GPN studies, the definition of social upgrading has mainly been based on the ILO decent work framework (Barrientos & Smith, 2007), which comprises four pillars: employment creation, standards and rights at work, social protection, and social dialogue. These four pillars promote work carried out under conditions of human dignity, security and freedom, in which social coverage and adequate remuneration are provided and rights are protected (ILO, 1999).
Social upgrading (decent work) consists of two dimensions; measurable standards and enabling rights (Barrientos & Smith, 2007). Measurable standards are easier to quantify and include employment type, working hours, physical wellbeing, wage level, and social protection. Measurable standards can be perceived as the outcome of bargaining processes, enabled by the rights of employees (Barrientos, 2011). The second element, enabling rights, is more difficult to quantify and includes the right to collective bargaining, freedom of association, the right to choose type of employment, non-discrimination, and voice and empowerment. Lack of access to enabling rights reduces employees’ ability to negotiate over working conditions that could improve their wellbeing (Barrientos, 2011).
17
provide an explanation as to why economic upgrading does not lead to social upgrading per se. Nevertheless, it may be the case that increases in certain aspects of decent work, such as employment, deteriorate other elements of decent work such as wages (Lee et al., 2011). Mather (2008) for instance, inferred in his study on the Global Squid Value Chain that, although economic upgrading led to a significant increase in the amount of jobs, working conditions did not ameliorate. Another study on apparel and electronics industries in Romania concluded that, even though process upgrading occurred in both industries, working conditions, wages and opportunities to access trade unions remained poor (Plank, Staritz & Lukas, 2009). Furthermore, Bernhardt and Milberg (2011) found a positive relationship between economic and social upgrading in apparel and horticulture industries, but did not find such a relationship in the mobile telecom industry. Besides the facts that their analysis was based on narrowly defined measures of upgrading and that there were substantial data gaps (especially for social indicators), Bernhardt and Milberg (2011) did not offer an explanation for these divergent findings. As research provides a mixed picture and since an explanation as to why economic upgrading does not necessarily lead to social upgrading is lacking, there is a need to advance the understanding of the relationship between economic and social upgrading by investigating the conditions under which economic upgrading leads to social upgrading (Barrientos et al., 2010; Barrientos et al., 2011; Rossi, 2011).
2.4 Factors accounting for economic and social upgrading
18
permanent and skilled work force, may lead suppliers to invest in employees and proper working conditions. However, pressure to cut costs and shorten delivery time may lead suppliers to economize on working conditions (Plank et al., 2009; Barrientos et al., 2011).
Linkages between economic and social upgrading are complex, with mixed results and outcomes for different kinds of workers. A case study on the garment industry in Morocco (Rossi, 2011) concluded that functional upgrading results in social upgrading and downgrading at the same time, depending on skill-level and employment type. Whereas regular workers (often higher-skilled workers) experienced social upgrading, irregular workers experienced social downgrading. Employees who had permanent contracts were offered better social protection so as to prevent them from taking a job somewhere else and in order to comply with buyers’ codes of conduct (Rossi, 2011). On the contrary, irregular workers were employed so as to quickly respond to buyers’ orders and to ensure short lead times. These unskilled workers often were generally young women, discriminated by getting paid below the minimum wage, poorly treated and not socially protected (Rossi, 2011). Regular workers were also represented by trade unions more frequently. This improved their enabling rights, leading to higher benefits accruing from measurable standards (e.g., higher wages, less overtime work etc.) (Rossi, 2011). Irregular workers, often low-skilled, were less likely to be represented by trade union organizations. Their limited access to enabling rights also restricted their chances of acquiring benefits derived from measurable standards. Moreover, irregular workers often experienced double discrimination as many of them are women or part of minority groups (Rossi, 2011). This suggests that gains resulting from social upgrading are unevenly distributed along gender lines. Women often disproportionately occupy jobs in lower-skilled positions, thereby limiting their access to decent work (Goger et al., 2014). The simultaneous use of regular, high-skilled employees, who guarantee quality, and irregular employees, who can be laid off easily and recalled when necessary, also implies that many developing country firms, so as to ensure their participation in GPNs/GVCs, take on a mixed strategy in order to comply with buyer standards and increase flexibility, thereby combining high quality (a high road strategy) with low costs (a low road strategy) (Barrientos et al., 2011).
Third-labor contractors also play a significant role in the occurrence of social upgrading. A
19
increase workers’ ability to continuously have a job by sending them to production sites where there is work. However, it also provides opportunities for unethical agencies to exploit their workers, thereby eroding working conditions (Barrientos & Kritzinger, 2004; Theron & Godfrey, 2000). Thus, third-labor contractors have a strong influence on the relationship between economic and social upgrading (Barrientos, 2011).
Furthermore, industry standards have been an important driver of social upgrading (Goger et al., 2014). By using standards, lead firms set minimum expectations for quality, hygiene and safety, etc., that developing country suppliers have to comply with (Bettiol, De Marchi, Di Maria & Micelli, 2011; Fernandez-Stark, Bamber, & Gereffi, 2011), thereby increasing the wellbeing of employees. However, adherence to standards cuts two ways; on one hand, it provides developing country firms with opportunities for economic upgrading as they are required to comply with certain quality levels, implying that they add more value to the product; on the other hand, it restricts participation only to those firms able to make the necessary investments required for compliance.
Additionally, the occurrence of economic upgrading is higher under certain governance
modes (Humphrey & Schmitz, 2001), which could mean that social upgrading is also more
likely to occur under certain governance modes. Different GVCs produce different kinds of products, perform different activities and require different combinations of low-skilled and high-skilled workers (Barrientos et al., 2010). Based on Rossi’s (2011) study on the Moroccan garment industry, these different ratios of skilled labor are likely to lead to different levels of social upgrading. Furthermore, these different ratios of low-skilled and high-skilled employees, which ultimately determine the capabilities in the supply base, influence, along with the complexity of knowledge transactions and the ability to codify these
transactions, the relationships between lead firms and their suppliers. These relationships
subsequently influence the link between economic and social upgrading. Research confirms that governance modes, employed to implement standards in GVCs, have significant consequences for economic and social upgrading (Evers, Opondo, Barrientos, Krishnan, Amoding & Ndlovu, 2014).
20
pressure on price and quality, implying they do not have to economize on working conditions as much as smaller suppliers. Intense competition makes it more plausible that suppliers take the low road, thereby economizing on wages for workers, social protections and rights. This means that it is not easy to achieve social upgrading in the apparel industry as apparel manufactures experience strong pressure to reduce labor costs (Pickles, 2012; OECD, WTO & IDE-JETRO, 2013) and operate under conditions of intense competition and contract uncertainty (Pickles, 2012).
Barrientos et al. (2010) argue that the position of the firm within GVCs is also an element that significantly influences economic and social upgrading. The closer the supplier is to the lead firm, the more likely it is that social improvements take place. This implies that employees of suppliers performing more complex tasks are more likely to socially upgrade as these firms are more closely interacting with the lead firm, meaning that skill-level has an influence on economic and social upgrading. A summarizing figure of the foregoing discussion is provided below.
Figure 2: Factors accounting for economic and social upgrading
21
The foregoing discussion provides insight into the factors accounting for economic and social upgrading. However, there is still a limited understanding of the conditions under which economic upgrading translates into social upgrading (Barrientos, Gereffi & Rossi, 2010; Barrientos, Gereffi & Rossi, 2011; Rossi, 2011). Therefore, more research on this relationship is required. Rossi (2011) suggests researching the role of the institutional environment, while institutional economics also argues that the relationship between economic and social upgrading is mediated by institutional arrangements (Milberg & Winkler, 2011). Rossi (2011) herself aimed to shed light on the role of institutions in the relationship between economic and social upgrading. However, her study did not sufficiently explore the role of institutional actors such as governments, trade unions and NGOs. It rather focused on how firm location and different cultural contexts impact social upgrading levels. Besides, her study only focused on one country (i.e., Morocco), thereby disallowing for cross-country comparisons. Besides, institutional structures differ nationally (Kostova, 1997) and may bring about different levels of social upgrading. Therefore, gaining insight into the role of different national institutional contexts in the relationship between economic and social upgrading may advance existing knowledge on the conditions under which economic upgrading translates into social upgrading and may give rise to the identification of other factors influencing economic and social upgrading.
2.5 Institutions; an under-researched dimension in GPN and GVC analysis
22
structures cannot be treated independently from institutions as institutions impact firms’ ability to cooperate and interact, thereby influencing transaction costs (Mudambi & Navarra, 2002), which is a key variable codetermining the governance modes of GVCs. Institutions and governance modes are mutually dependent; implying that they influence each other (De Pereny, 2012). Consequently, institutions cannot be treated as external to the GVC framework. Different institutional environments have a different influence on the distribution of economic, politic and social power, which determine the degree of cooperation and conflict between GVC participants (De Pereny, 2012). Thus, the way institutions are designed shape the position of GVC participants, their power, and the distribution of profits along chain members.
Although the GPN framework emerged as a critique of the GVC framework’s lack of attention to institutions (Bair, 2009a), the institutional environment has not adequately been addressed in GPNs either (Coe et al., 2008b; Levy, 2008; Bair, 2009a; Smith, 2014). Despite the theoretical differences between GPNs and GVCs, both terms are often used interchangeably and research spawned by GPN literature does usually not differ in practice (Coe et al., 2008b; Levy, 2008; Bair, 2009a), meaning that GPN studies have not sufficiently addressed the lack of attention to institutions in GVC studies (Coe et al., 2008b; Smith, 2014).
23
poor working conditions are the norm among organizations and when institutional arrangements are weak, change in favor of employees being exploited is unlikely to occur. Also, employees’ ability to access enabling rights is strongly determined by the institutional environment of a society (Amengual, 2010; Nathan & Posthuma, 2010).
The foregoing discussion makes clear that institutions are likely to play a significant role in the relationship between economic and social upgrading. As the role of institutions has been neglected in both GPN and GVC studies, this research puts the institutional dimension more on the foreground by empirically studying its role in the relationship between economic and social upgrading in Bangladeshi and Indian apparel GPNs. This study adopts the definition of institutions provided by North (1994), according to which institutions comprise two types of institutions; formal and informal institutions. With reference to social upgrading, this research considers formal institutions to consist of governments, trade unions and NGOs. Informal institutions are considered to be made up of culturally embedded systems and norms and values such as attitudes toward the social position of men and women and attitudes toward child labor.
2.6 Research model
24
Figure 3: Research model
25
3. METHOD
This chapter provides details of the research design, explains how the research was done, which data were collected and how the data were analyzed.
3.1 Research strategy
Jonker and Pennink (2010) argue that a qualitative research design is considered to be appropriate for exploring phenomena in a specific context. Therefore, a qualitative approach was warranted as this research focused on two countries in particular, namely Bangladesh and India. Besides, Barrientos et al. (2011) also state that social upgrading is difficult to measure quantifiably. Therefore, so as to advance the limited existing knowledge on economic and social upgrading, this study warranted and employed a qualitative approach (Eisenhardt, 1989).
One way of doing qualitative research is the case study approach. “Case studies have become one of the most common ways to do qualitative inquiry” (Stake, 2000, p. 435) and according to Milberg and Winkler (2011:345), “case studies bring a deeper understanding of the process of upgrading, the role of each of the key actors and the obstacles that upgrading firms face in GVCs”. Hence, the research strategy used was the case study approach. Yin (1989:23) refers to a case study as “an empirical inquiry that investigates a contemporary phenomenon within its real-life context, when the boundaries between phenomenon and context are not clearly evident”. This study employed an instrumental and multiple case study approach. It used an instrumental case study approach in that it focused on particular cases in order to obtain a better general understanding of the relation between economic and social upgrading (Bassey, 1999). The study employed a multiple case study approach in that it attempted to explore similarities and differences between the case of Bangladesh and the case of India (Stake, 1995). The multiple case approach was preferred over the single case approach as theory building from several cases generally yields more robust and generalizable results, thereby enhancing external validity (Eisenhardt & Graebner, 2007).
3.2 Case selection
26
(Nathan & Posthuma, 2009; Gereffi & Fernandez-Stark, 2011; Bernhardt, 2013) and that no other industry has received more attention in GPN analysis (Bair & Peters, 2006). Historically, the apparel industry has been a significant gateway to economic development (Bernhardt, 2013); technology, capital and skill requirements are relatively low and the production of apparel is labor-intensive, meaning that developing countries, with their relative abundance in low-skilled workers, have a comparative advantage in this industry (Bernhardt, 2013). Reasons for choosing Bangladesh and India were sevenfold: (1) both countries are among the top five apparel exporting countries in terms of world export market share (Bernhardt & Milberg, 2011; Frederick & Staritz, 2012; Bernhardt 2013); (2) both countries have rapidly developing economies, predominantly resulting from the phenomenal growth of the apparel industry (Asia Foundation, 2010; Textile Industry India, 2012); (3) their participation in global apparel production has provided numerous jobs to men and women (Bernhardt, 2013; Pickles & Godfrey, 2013); (4) Bangladesh and India were differently affected by the Multi-Fibre Arrangement (Bernhardt, 2013; Frederick & Staritz, 2012); (5) there has been a considerable wage gap between Bangladeshi and Indian apparel workers (Bernhardt, 2013; Frederick & Staritz, 2012); (6) they (at least) differ in terms of informal institutions as traces of the Indian caste system are still visible in India’s modern society (Deshpande, 2010); (7) Bangladesh and India differ in terms of industry dependence; Bangladesh is much more dependent on the apparel industry for economic growth (Frederick & Staritz, 2012). This difference in industry dependence may have an influence on how institutions operate and on how buyer-supplier relationships are governed in the apparel industry.
3.3 Data collection
27
mediator between the state and business, protecting people’s interests (Crane & Matten, 2010). This implies that NGOs have insight into organizations’ behavior and the institutional forces that impact their behavior, meaning they were suitable data sources.
Interviews were the main method of data collection as this method is particularly appropriate for exploring new relationships and theory building studies (Daniels & Cannice, 2004). Interviews were semi-structured and in-depth in nature. Two interview guides, one for Bangladesh and one for India, were designed consisting of, respectively, twenty-five and twenty-seven open-ended questions (see Appendix A and B). In an early stage, fifty organizations were identified and approached through email and phone calls. The email explained the purpose of the study and posed the question whether the organizations were willing to participate in this research. After fifteen days of sending them a request by email, a first reminder was sent to the potential participants, reminding them of this research. Fifteen days later, another reminder was sent. If there were still any cases of no response, the concerning organizations were called with the request to participate. Eventually, seven organizations and nine individuals were willing to participate. Table 3 provides more information on these organizations and their participants. Please note that a participant coding system was used in order to ensure anonymization of the data.
Table 3: List of expert interviews
28
Interviews with Dutch organizations were conducted face-to-face, held in Dutch and took place at the organization’s site or in a public place. Interviews with organizations located outside the Netherlands were held in English and took place via Skype, mainly due to geographical constraints. The interview questions were formulated in such a way that they did not contain any technical terms or jargon, which could have led to misunderstandings. The interview guide was sent to the participants in advance so as to enable them to be prepared for the interviews. All interviews were audio-recorded, allowing the researcher to stay focused on the interview and the participant. Detailed notes would have been taken in instances where participants did not allow the recording of the interview. However, all participants agreed to the recording of the interview. The interviews typically lasted thirty-five to ninety minutes. After conducting the first three interviews, the interview guide was evaluated and revised. Afterwards, all interviews were transcribed and sent back to the participants in order to allow for data verification.
3.4 Data analysis
All interviews were transcribed. However, there is little consensus as to how interview transcripts should be analyzed (Thomas, 2004). According to Thomas (2004), in the case of qualitative research, data need to be categorized and coded. The aim of this technique is to reduce the data by summarizing and assigning it to certain categories. This method of data analysis was also employed in this study. All coding was done manually. Color-coding and highlighting were used so as to organize the data. The interview questions and the related literature were used as a guide to develop a preliminary list of possible coding categories. However, new categories emerged from the data set during the coding process. This study started analyzing the data by looking at each case individually. This is called within-case analysis. Thereafter, a cross-case analysis was conducted in order to find similarities and differences between both cases and to search for patterns.
3.5 Ethical issues
29
30
4. WITHIN-CASE ANALYSIS AND CROSS-CASE ANALYSIS
This chapter presents the findings from the empirical research. The first subsection addresses the case of Bangladesh, describing the industry’s history and development, elaborating on the relationship between economic and social upgrading and the role of institutions herein. The second subsection discusses the same issues for the India case. Thereafter, the third subsection compares both cases. Finally, the last subsection touches upon other factors that may influence economic and social upgrading, emerging from interviews conducted in this research. Results derived from interviews are presented throughout this chapter. Please note that statements made by respondents need to be interpreted with caution and should not be generalized.
4.1 Within-case analysis Bangladesh 4.1.1 History and development
The global apparel industry has experienced significant restructuring recently, especially after the expiration of the Multi-Fibre Arrangement (MFA). Under the MFA, developed countries could set quotas on developing countries’ apparel exports (Robertson, 2012; Bernhardt, 2013). The arrangement put major restrictions on the opportunities of countries such as China and India (Bernhardt, 2013; Robertson, 2012). However, the arrangement did not have a negative influence on all developing countries. When manufacturers reached quota limits in one country, they would import from adjacent countries where the limit had not yet been reached. This process, called quota hopping, enabled many other countries, such as Bangladesh, to reinforce their position in the apparel export market (Robertson, 2012; Bernhardt, 2013).
31
Table 4: Rapid development of the Bangladeshi apparel industry
Source: Frederick & Staritz (2012) Note: Data for 1983 are not available
32
At the moment, Bangladesh is the second largest apparel exporter in the world, employing approximately 3.5-4 million people and making the industry responsible for around 80% of the country’s total exports. Around 80% of the people working in the Bangladeshi apparel industry are women (Frederick & Staritz, 2012; FWF, 2013).
4.1.2 Economic upgrading in Bangladesh
GVCs are often characterized by considerable power asymmetries and local firms are vulnerable to the exercise of power by multinationals. The comparatively strong bargaining position of multinationals as opposed to supplier firms was confirmed by all respondents. Respondent C1, for example, stated the following:
“[A]pparel firms in Bangladesh rely on brands and retailers for income and orders. They are very vulnerable to the directives of those brands and retailers whether it be to offer lower prices or offer faster shipping and turnaround times. Whatever it is that the brands and retailers are demanding of their suppliers, Bangladeshi supplier firms understand they have to provide those things, otherwise the brands will go somewhere else.”
33
delivery times shorten. On the other hand, respondent A2 stated that supplier firms can produce at a lower cost and are able to shorten their lead times by using better machinery, which provides them with a higher leverage over competitors. Furthermore, respondent F explained that multinationals are more likely to source from those factories that can deliver on time and are able to do a good job, which is more likely when firms use better machinery. However, despite their advantage over other firms, these factories are still largely dependent on exports to advanced countries (interview Participant A2), meaning that process upgrading may provide firms with an advantage over others, but may not strengthen their bargaining position toward buyers. When looking at product upgrading, respondent A2 indicated that some relationships may become more relational over time, depending on product quality and working conditions:
“Leading companies realize that when they produce high quality products under good working conditions, their relationships with buyers may evolve into a long-term relationship. Relationships between buyers and factories have changed in recent years. Buyers prefer to enter into long-term relationships nowadays, lasting 10-20 years.”
Respondent E stated the following with reference to product upgrading:
“Usually, firms that produce higher quality products get more orders. But this is not always the case. During the economic crisis you saw that cheaper clothing was more popular. Bangladesh was also barely affected by the crisis.”
This indicates that product upgrading may be beneficial to factories, leading to more orders and long-term relationships. However, when economic times turn tough, buyers may switch to cheaper, lower quality products, meaning that buyers may switch suppliers. Regarding functional upgrading, some Bangladeshi firms have been able to move from mere assembly to OEM production. Respondent E indicated that firms being responsible for OEM production (complete manufacturing) have stronger bargaining power than those who only carry out assembly tasks:
34
The same respondent also pointed out that it might be difficult for Bangladeshi firms to further upgrade functionally. Bangladeshi firms have limited opportunities to adopt a multi-chain strategy. A multi-multi-chain strategy implies that a firm maintains OEM in one supply multi-chain, i.e., the chain controlled by the global buyer, while simultaneously leveraging competences developed in this particular chain to develop ODM/OBM capabilities in another chain, geared toward domestic and regional markets (Schmitz, 2006). The success of this strategy, thus, depends on domestic market size. As respondent E pointed out:
“[W]hen you have a big market, you are less dependent on exports as you can make sales domestically as well. Bangladesh […] does not have a large domestic market and is entirely dependent on exports and has therefore less bargaining power [toward multinationals].”
To sum up, all three types of economic have taken place in Bangladesh. Process upgrading may provide factories with an advantage over competitors, product upgrading may lead to long-term relationships and functional upgrading may improve the bargaining position of factories; buyers find it easier to source from one factory than to source from multiple factories, perhaps due to transaction costs. Besides, a large domestic market provides more opportunities for functional upgrading and may provide factories with a stronger position toward buyers.
4.1.3 Social upgrading in Bangladesh
The next question is whether process, product and functional upgrading and the, respectively, resulting competitive edge over other supplier firms, long-term relationships and increased bargaining power lead to social upgrading, i.e., better working conditions. The influence that process upgrading has on working conditions might be limited to improvements of facilities:
35
Thus, process upgrading may result in better health and safety. However, improvements in conditions such as leak-proof roofs, secure wiring and clean floors may only emanate from the self-interest of owners. Respondent C1 argued that there is no one-on-one relationship between process upgrading and social upgrading and that working conditions have become worse in many instances:
“Working conditions for apparel workers in Bangladesh have not improved over the years. In fact, in many cases the situation has gotten worse. When you look for example at health and safety, the largest and deadliest disaster in the apparel industry happened a year and a half ago. Brands and retailers have been aware of the health and safety issues in the Bangladeshi apparel industry. For more than a decade workers have been dying in these very preventable accidents. And yet nothing was done, despite the alarms being raised by all kinds of organizations. Mainly because it was still so cheap to produce there and to make it safer would have cost money. So, I don’t think you will find anybody who will tell you that conditions have improved as the result of better production machinery or higher productivity.”
Regarding product upgrading, respondent A1 stated the following:
“Those who produce for global buyers, with higher value added, are more likely to produce higher quality clothing. Generally, working conditions are better in those factories, mainly because global buyers call for better working conditions, as they share responsibility. It is assumed that factories producing for global buyers have better working conditions, otherwise they would not have gotten the order.”
36 “In CMT factories the working conditions are worse than in factories where they are responsible for complete manufacturing. But working conditions in assembly factories are similar to the working conditions of workers that are carrying out assembly tasks in larger factories, factories that are also responsible for other activities. So, there are […] differences in working conditions between workers employed by the same factory, based on activities that are being carried out and skill-level.”
The same respondent also stated the following:
“Stronger bargaining positions are more likely to lead to higher profits for factory owners, not so much to better working conditions for workers.” (interview Participant E)
This implies that stronger bargaining power, which may result from functional upgrading, do not give rise to better working conditions but may only contribute to higher profits. To sum up, process upgrading may only lead to minor improvements in health and safety, product upgrading may only lead to better working conditions through buyers’ demands and functional upgrading may lead to better working conditions as a result of skill upgrading.
4.1.4 The role of institutions in Bangladesh
37
therefore, not all firms have been able to make this move. Family relationships and friendship ties play a large role in this regard and could be reasons why some firms have better access to finance than others (interview Participant E). This also touches upon the issue of conflicts of interest where power is unequally distributed between the government, employers and workers, who are brought together by the tripartite system (Bell & Newitt, 2010). Power is unequally distributed between parties in that there is a significant overlap of interests between the government, employers and trade unions. To start with, the interests of the government and the apparel industry strongly converge as many politicians are factory owners or part of factory-owning families (interview Participant A1; A2; B; C1; E; F). This is reflected in the following statement:
“Bangladesh has around 150/175 members of the parliament. At least 40 of them are direct owner and another 60 have direct ties with someone who is an owner.”(interview Participant A2)
In fact, all the respondents indicated that the industry is the most powerful player. Many factory owners are member of the parliament or at least have linkages with political parties. While the government has to balance the interests of both the industry and the workers, they consider the industry to be more important than the workers (interview Participant E). In general, maximizing profits is what matters most, which is perhaps characterized best by the following statement:
“Just after the collapse of Rana Plaza, there where factory owners who said: “we are only here to make profits”.”(interview Participant A2)
38 “Workers who try to form unions are intimidated, harassed, fired, physically attacked and murdered. If the government were to take actions to give unions more power, that would influence the costs and the prices of apparel products in Bangladesh […], which is something the government understands that the industry will not tolerate.” (interview Participant C1)
Unionization levels are also low because there are many restrictions on forming unions. The government wants to attract investments and together with the industry they make every effort to make life for unions as difficult as possible (interview Participant A1). This is also reflected in the following statement:
“The government does not want to do anything to jeopardize the investment. […]In the
last couple of years you have seen over 200 new unions registering in the apparel industry. Before that the government had imposed a ban on registering new unions. Because of external pressure this has changed. However, we have seen a significant wave of anti-union repression. Registration may be easier now but it happens that employers sack the entire executive board of the union and in some cases there are violent situations where they bring in armed thugs to beat up trade union leaders or workers who are going with the union. And even where you don’t have this violent repression or mass sackings, employers are still not negotiating in good faith with the newly recognized unions. That is how industrial relations in Bangladesh are today. […] [I]t is still very much a brutal industry. And when workers do try to stand up to their employers, they are in many cases sent to the hospital. The government is not doing anything about it, they are not penalizing.”(interview Participant B)
There are around 200 small independent unions. However, they have very little power and are not adequately informed about workers’ rights (interview Participant A2). Workers are often not aware of their collective bargaining rights as well (interview Participant A1; B; C1). The majority of the unions are corrupt, self-interested and willing to do anything for a certain amount of money:
39 leaders money, so they won’t say anything and will keep their mouth shut.” (interview Participant A1)
NGOs do not have much power either. There has been significant retaliation against NGOs and labor unions by both employers and the government up to and including the murder of a Bangladeshi union organizer two years ago, for which no one has been prosecuted today (interview Participant C1). Moreover, it does not make sense for unions to go to court or the labor inspection. If courts take a decision, it takes them 5 years to do so (interview Participant A1). The labor inspection is very weak and unable to contribute to better working conditions, mainly because there are very few of them (interview Participant A1; B; C1; E); 30 labor inspectors are responsible for inspecting the whole country, not just the apparel industry (interview Participant A1; E). According to respondent E, the government is not planning on increasing this number as they do not have the budget to spend more money on this, nor are they willing to do so, predominantly because the apparel industry is the most important player of Bangladesh’s economy. Respondent A1 indicates that it is not the government’s priority. Only a really small budget is made available for labor inspection; labor inspectors have not been sufficiently trained and lack the resources and equipment to execute adequate inspections. The main priority of the government is mainly to ensure the industry grows and working conditions are of secondary importance (interview Participant A1). However, it goes much deeper than just a weak labor inspectorate; many politicians also do not want to see any actions taken against their own company (interview Participant B). And civil servants do not earn much money, definitely not enough to support their families. This leads to high levels of corruption and self-interest. The government only undertakes action when something bad has occurred, such as the collapse of Rana Plaza (interview Participant A1). Furthermore, when a particular ministry is assigned a higher budged, most often resulting from international pressure, other ministries will feel deprived and jealousy will arise among ministries. Eventually, as respondent A1 argued, this will lead to higher levels of corruption, which might worsen the current situation.
40
already advanced, just because the apparel industry offers them a chance to earn wages (interview Participant A1; F). They have more money, power and choice (interview Participant A1). But in general, women and men experience the same working conditions. They work in the same building and salaries are similar. However, women are treated differently. Several cases of discrimination and harassment have been reported, particularly against women and children (interview Participant A1; E). Child labor is also difficult to fight against; families depend upon children’s wages. Moreover, sending them to school costs money while sending them to work yields money (interview Participant C1; F). Low road factories are not likely to miss the opportunity to replace an adult worker by a child, simply because they cost less money, are more obedient and less likely to protest when their rights are being violated (interview Participant C1; F). Families may not be able to survive if children are pulled out of work and therefore, over time, it has become part of society (interview Participant A1).
To sum up, the government’s main aim is to protect the industry and to increase exports, partly because many politicians are simultaneously factory owners and partly because the industry is invaluable to the country’s economic situation. On the one hand, the government offers all kinds of incentives to ensure factories are using advanced machinery and are able to upgrade. On the other hand, conflicts of interest, weak trade unions and the absence of a correctly functioning labor inspectorate drive down the working conditions employees in the Bangladeshi apparel industry are subjected to.
4.2 Within-case analysis India 4.2.1 History and development
The apparel industry has been an important driver of India’s economic growth in the last decades. India has a rich history in the apparel industry. However, due to crises and strikes in the 1970s and 1980s, factory after factory closed down, which led to the Indian apparel industry being completely de-industrialized in the early 1990s (D’Monte, 2002; Davies & Nyland, 2004; Hardgrave & Stanley, 2008). After the economic reforms in 1991, the Indian apparel industry had to be entirely rebuilt from scratch, meaning that, although the Indian apparel industry has a rich history, it is relatively new at the same time.
41
gave way to developed countries to increase imports from countries such as India (Bernhardt, 2013). Investments in backward linkages and India’s domestic market have been conducive to growth in exports, overall production and the development of capabilities such as product development, design and branding. Government policies also played a significant role in the flourishing of the Indian apparel industry (e.g., the government stimulated technological upgrading by offering credit at reduced rates) (Frederick & Staritz, 2012).
At the moment, India is the fifth largest apparel exporter in the world, employing around 6 million people, making the industry responsible for approximately 12% of the country’s total exports (Frederick & Staritz, 2012; FWF, 2012). Around 60% of the people working in the apparel industry are women (FWF, 2012). The domestic market is still India’s largest market. In 2007, around 65% of apparel output was intended for the domestic market, whereas approximately 35% was intended for exports (Frederick & Staritz, 2012).
4.2.2 Economic upgrading in India
Just as was the case in Bangladesh, all three types of economic upgrading being addressed in this study have taken place in India (Fernandez-Stark et al., 2011; Frederick & Staritz, 2012; Indian Council, 2013). With respect to functional upgrading, India is mainly concentrated in OEM and ODM production (Fernandez-Stark et al., 2011; Frederick & Staritz, 2012) and to some extent in OBM production (Frederick & Staritz, 2012; Indian Council, 2013); Indian firms are selling their own brands on domestic and adjacent country markets (Indian Council, 2013). What applies to economic upgrading in Bangladesh also applies to economic upgrading in India. Process upgrading is likely to provide factories with an advantage over competitors (due to increased productivity and faster delivery) (interview Participant A2; C2; E), product upgrading is expected to lead to long-term relationships (buyers want to establish long-term relationships and are more likely to go into a long-term relationship with suppliers that can produce high quality products) (interview Participant A2; G) and functional upgrading may enhance the bargaining position of factories (interview Participant E; G). Regarding process upgrading, respondent C2 put it as follows:
42
As to product upgrading, interviewee G pointed out the following:
“I wouldn’t say that factories producing higher quality products have stronger bargaining positions. However, factories producing higher quality products do often have long-term relationships with buyers. For example, factories specialized in particular segments, such as outdoor clothing for instance, usually maintain long-term relationships with buyers. These factories produce technical clothing with high quality standards.”
With reference to functional upgrading, there tends to be a bifurcation between large and small/medium sized firms. Although GVCs are generally characterized by considerable power asymmetries where local suppliers have a disadvantage when it comes to bargaining power, in India, this is not the case for every supplier. Respondent C2 argues that there are two categories of apparel firms in India. One category consists of really large suppliers. The other category comprises medium and small suppliers. The very large Indian manufacturers adopt a multi-chain strategy; in one supply chain they supply goods to global buyers, which are subsequently sold on European and US markets under European or US brand names. At the same time and through another supply chain, these large Indian firms sell their own designs and brands to the Indian domestic market but also to other Asian markets, the Middle East market and the South-American market (interview Participant C2; G). These larger firms are, therefore, less dependent on orders from European and US based firms and have higher bargaining power than factories mainly producing for one or a few global buyers (interview Participant C2). On the contrary, relationships between medium/small suppliers and powerful multinationals are still characterized by considerable power asymmetries (interview C2; D; G). Respondent D, for instance, stated the following:
43
In India, this implies that different types of firms experience different effects, arising from economic upgrading and functional upgrading in particular.
4.2.3 Social upgrading in India
Now that economic upgrading has been discussed, a look can be taken into its influence on social upgrading. Responses on the relation between economic and social upgrading in India were similar to those on Bangladesh. As to process upgrading, interviewee C2 put it as follows:
“Large factories have slightly better conditions and are more often under the scanners. These large factories are often using better machinery, but it is not necessarily the case that workers experience better conditions just because they are using more advanced machinery. Factories outside these scanners do not have good conditions. Still, their techniques and machinery may be advanced. It just depends on how much monitoring happens in the factory. […] If you are among the top 10, having more than 30-40 production facilities across the country, you are producing for almost all international buyers and you are definitely monitored. They will have slightly better working conditions than those that are not monitored.”
Just as was the case in Bangladesh, process upgrading does not necessarily lead to better working conditions, but may lead to improvements in some respects:
“Higher productivity and the use of new machinery definitely lead to better working conditions. However, they are not passed on to the worker in the form of wages or less overtime work, but more in the form of less heavy and demanding work. Machinery replaces some manual tasks, which makes work less burdensome. Some factories also get the message that they might be able to build a sustainable relationship with the buyer by treating their workers right. Especially the larger, state-of-the-art factories get this. These factories use better machinery than those in the lower strata of the industry.” (interview Participant A2)
44 “The higher the quality of the product, the higher the price. However, higher prices and profits do not have to lead to better conditions per se.”
Respondent A2 continued to argue that companies producing high quality products under good conditions are more likely to have long-term relationships with buyers, which might imply that there is a relationship between product upgrading and working conditions as factories producing high quality products are more likely to acquire orders from buyers, on the condition that production takes place under good working conditions. At the same time, the Indian domestic market has enabled Indian factories to develop ODM and OBM capabilities. Respondent C2 said the following about this:
“[The large Indian domestic market] has […] resulted in higher profits for Indian firms compared to firms [that] do not have such a big domestic market. However, these higher profits do not get distributed to the workers. It does not matter whether they are producing for big buyers or for their own brand. Why would a firm share profits? They do not want to pay any higher wages than necessary. The minimum wage has become the maximum wage. The industry does not go above it.”
To summarize, process upgrading may only lead to minor improvements in the form of less heavy and demanding work; social upgrading may not be related to improvements in technology but to firm size, whether the supplier gets monitored and whether the supplier has a direct relationship with the buyer; and functional upgrading may lead to higher profits for suppliers but may not translate into better working conditions for employees.
4.2.4 The role of institutions in India
