Flying Geese or Trapped Geese:
Assessing the footwear industry in Asia with production
fragmentation
MSc Thesis
International Development Studies
Graduate School of Social Sciences
University of Amsterdam
August 16, 2019
Ching-Chun Chen
11750774
ching-chun.chen@student.uva.nl
Supervisor: Niels Beerepoot
Abstract
The flying geese paradigm describes the spill over and industrial migration of industry. However, the prescriptions of the paradigm may be faced with the challenges brought by the production fragmentation. This research aims to examine the validity of flying geese paradigm with the framework of Global Value Chain Governance and the mixed model analysis. A research with mixed methods approach, which combines the interviews of major Taiwanese footwear firms and quantitative analysis with global value chain input-output table, indicates that local firms in China and Southeast Asia have limited benefits from the relocation of the researched Taiwanese firms as it is difficult for them to join the global value chain and Taiwanese firms relocate to other countries as the local cost rises. At the same time, Taiwanese firms are also subject to exploitation by footwear brands in the global North as they can only participate in manufacturing which is the lowest value-adding activity in the chain. The findings suggest that although the prescription of flying geese paradigm about the migration of industry remains valid to this date, whether such migration can generate sustained economic growth for the local firms may need further examination. The result concludes that by only introducing foreign direct investment may not guarantee sustainable economic development, it is also vital to facilitate local firms to join the global value chain and build up their capacities to capture more value-adding activities in the value chain.
Table of content
1 Introduction... 4
1.1 Problem Statement & Research Aims ... 5
1.2 Thesis Outline ... 6
2 Theoretical Frameworks... 8
2.1 Introduction... 8
2.2 Flying Geese Paradigm ... 9
2.2.1 Introduction ... 9
2.2.2 Stages of Industrial Development in the Original Model by Kaname Amakatsu ... 10
2.2.3 Pro-trade-oriented foreign direct investment ... 14
2.3 Production fragmentation from the perspectives of Global Value Chain 16 2.3.1 Introduction ... 16
2.3.2 Organizational & international production fragmentation ... 17
2.4 The Challenges faced by FGP in the Era of Production fragmentation .. 22
3 Methodology ... 25
3.1 Introduction... 25
3.2 Research Question ... 27
3.3 Conceptual Scheme ... 27
3.4 Operationalization of Major Concepts ... 28
3.5 Units of Analysis ... 31
3.6 Research Design ... 31
3.6.1 Sampling Strategy & data collection ... 32
3.6.2 Data Analysis ... 37
3.6.3 Ethics ... 43
3.6.4 Limitations and challenges ... 44
4 Past, Present and the Future of Global Footwear Value Chain in the Region of East Asia and Southeast Asia ... 46
4.1 Previous Development of Footwear Industry in Taiwan, China, and
Vietnam ... 46
4.2 How does contemporary global footwear value chain work in the region? 48 4.3 The relocation strategies and the challenges for the footwear intermediaries ... 50
4.4 The implication of organization fragmentation on the FGP ... 53
5 International Production fragmentation... 55
5.1 Introduction... 55
5.2 Is international production fragmentation increasing overtime? ... 56
5.3 Is FGP still relevant in the footwear industry today? ... 59
5.4 How is FGP affected by the international production fragmentation in the footwear industry? ... 62
6 Conclusion ... 66
References: ... 68
Appendices ... 75
I. Interview Questions ... 75
1 Introduction
In the past three decades, East Asia has been the one of the leading regions of economic growth in the world and has experienced exceptional economic development one after another by adopting export-oriented industrialization (Kimura & Obashi, 2016). The path of economic development which the region has gone through is unprecedented and unique to the world, and it is consistently argued whether such trajectory of economic development can be replicated to other parts of the world (Cline, 1982). Such momentum is contributed by the sophisticated and extensive regional production networks, making the region the factory of the world (Kimura & Obashi, 2016). The contemporary integration of East Asia into world production systems dominated by the West can be traced back after WWII, when U.S. established a trade friendly global environment in favor of its allies in the region, namely Japan and the four NIEs (South Korea, Taiwan, Hong Kong, Singapore) (Ozawa, 2003). Under the umbrella of the U.S. hegemony, which provided the region with industrial knowledge, capital and market, macro-scale clusters of industries then formed in the region. The hegemon-led macro-scale clusters enables these countries to achieve fast development through Export Oriented Industrialization (EOI) (Gereffi, 2014; Ozawa, 2003). In 1960s, the Japanese economist Akamatsu introduced the Flying Geese Paradigm (hereafter referred to as FGP) which explains the pattern of the industrialization in the region (Akamatsu, 1962). In his theory, the trend of industrialization is originated from the center of innovation, which is Japan, then gradually spreading toward its neighboring countries in the region and onto other countries. Japan plays as the industrial pioneer in the region, when its economy upgrades to more advanced industries, the original ones would migrate to NIEs, then to ASEAN 4 (Malaysia, Thailand, Indonesia, and Philippines), and China (Kojima, 2000; Ozawa, 2003). The FGP is not only discussed within academia, but also by the development institutes such as Asia Development Bank and the World Bank (Kimura & Obashi, 2011; Lin, 2011).
In the perspectives of the FGP, late industrializers could follow the developmental trajectory of lead geese through the migration of industries, which allow them to accumulate the wealth by exporting manufactured goods (Kojima, 2000). Nevertheless, the global environment which allows the paradigm to function properly, is faced with challenges as China’s rapid-expanding global influences and the financial crisis in 2008-09 which stumbled the economies in the global north (Gereffi, 2014).
Furthermore, the development of information and communication technology (ICT) allowed a more dispersed production network. As a result, the production network is no longer limited within one country but can be distributed across different countries and firms based on the capacities and costs of the locations and organizations (Kimura & Obashi, 2011; Venables, 1999). Apart from the changes of global environment and the advancement of technology, the new patterns of production that brought by these changes also challenge the promises of FGP for the follow geese. Firstly, as more and more countries are imposing liberalization policies to the attract the foreign direct investment, the competition among countries to offer better terms for foreign investors has led to the compromise of the welfare for the local workers and local firms, undermining the benefits of foreign investment to the local economies (Mosley & Uno, 2007). Secondly, in the FGP, the export of finished products is a vital approach to accumulate the capital, however, the spatial disperse of production network has made the export of intermediate goods more and more dense while the proportion of finished goods has less share overtime (Los, Timmer, & de Vries, 2015). This implies that the late comers can join the process of production without building the complete production networks, which makes it a lot easier to participate in the global value chain, and makes the EOI more achievable, while the economic benefit of doing so is becoming less significant (Gereffi, 2014).
1.1 Problem Statement & Research Aims
In this context, the validity of the FGP as a reference for late comers to follow the early ones’ path is questioned. Therefore, the aim of this research project is to analyze whether FGP still serves as a channel for late comers to converge. The research will incorporate the framework Global Value Chain Governance proposed by Gereffi, Humphrey, & Sturgeon (2005) to analyze the production fragmentation in the dimension of inter-organization, as well as the analysis of the World Input-Output tables (M. P. Timmer, Dietzenbacher, Los, Stehrer, & de Vries, 2015) to focus on international production fragmentation. The research will focus on the case of footwear industry in the region, especially the development of the footwear industry in Taiwan, which is considered a classic case that follows the pattern of FGP (Fan, 2008; Lin, 2011).
1.2 Thesis Outline
The thesis program will be conducted with mixed method, which place the FGP as the subject to be assessed and analyzed. On the other hand, the Global Value Chain Governance and the World Input-Output table are being utilized as agencies which reflect the patterns of the production fragmentation in the contemporary global environment. In the following sections, the key prescriptions of FGP will be identified first and operationalized in to indicators, which will be set as dependent variables and to be evaluated in the context of production fragmentation. And the concept of production fragmentation will be introduced with the framework of Global Value Chain Governance and The World Input-Output Table. The major indicators to measure the presence of production fragmentation will also be identified as well. The indicators of production fragmentation then will be applied as the factors that has influences on the indicators of FGP to examine whether the prescriptions of FGP still remain valid in the presence of production fragmentation.
In terms of the production fragmentation, two dimensions will be covered by this thesis project. The first dimension is the organizational production fragmentation, which means the final product may not be manufactured and retailed solely by only one firm. Instead, the production processes and materials are outsourced to subcontractors, such outsourcing processes can take place in multiple levels, both vertically and horizontally, which then come into a complex production network, such complexity of production network can be observed on the production of footwear in the region(Cheng, 2011; Tasi, Chang, & Yang, 2000). In this thesis, three major multinational footwear firms are interviewed in order to have a glimpse of organizational production fragmentation in the footwear industry in the region and their relationships with international fragmentation. All of the firms have their headquarters based in Taiwan, while they take order contracts from the brands in the global north and have most of their production lines in China or Southeast Asia. Furthermore, these firms also function as intermediaries in the footwear global value chain, which are positioned in the middle of brand clients and the complex networks smaller suppliers (Fan, 2008). The special position of these Taiwanese firms in the global footwear value chain allows this thesis project to investigate not only the
development of footwear industry in Taiwan, but also the region of East Asia and Southeast Asia, even beyond, to the global scale. However, the organizational production fragmentation does not certainly result in international fragmentation (Venables, 1999), which is vital to the development of the industry in the perspectives of a country. Hence, the dimension of international fragmentation will also be covered to analyze the impact of production fragmentation on the FGP. The quantitative method will be adopted to investigate the aspect of international fragmentation with the support of The World Input-output table. The World Input-output table consists not only the trade of final goods but also contains the domestic and foreign inputs of intermediate goods, which makes it a suitable source of data to examine the international production fragmentation.
The thesis consists of six chapters with the first one giving the introduction to the background of the overarching debate, which in concerning whether the prescription of FGP still works in the time of production fragmentation. The following chapter of theoretical framework will further deliberate the concept of FGP and the two aspects of production fragmentation, making the linkage and comparison between the two concept. In the third chapter of methodology, the details of how the research project is designed and conducted will be given. And the contextual background of footwear industry in the region of East Asia and Southeast Asia, particularly Taiwan, will be introduced in the fourth chapter. Then the findings of how the international production fragmentation have impacted the validity of Flying Geese will be given in the chapter six. In the concluding chapter of number six, the conclusion and discussion will be made.
2 Theoretical Frameworks 2.1 Introduction
In this chapter of theoretical framework, two concepts will be focused on, which are respectively FGP and the production fragmentation. In the first section of FGP, the background of its development will be presented. The first section is not limited to the economic drives of the FGP but also including the historical and geopolitical background, which is also of great importance for the emergence of the paradigm and why it may face challenges in the future (Gereffi, 2014; Mozingo, 1967; Ozawa, 2003; Stubbs, 2006). Furthermore, the relevance of the footwear industry in the region and the paradigm will be touched upon while more details of the context regarding the footwear industry will be elaborated in the chapter four of contextual background. The second session will involve the concept of production fragmentation, which can be examined through the aspect of organizational and international dimensions. For the paragraphs regarding the organizational fragmentation, the framework of Global Value Chain Governance will be utilized as an instrument for us to understand how the tasks of an industry are distributed to different levels of manufacturers with different degree of power and autonomy, and how different types of production network are shaped according to the power dynamics between buyers and suppliers (Gereffi, 2014; Haworth, 2013). In addition, the dimension of international production fragmentation will be demonstrated with the works that show how a final product exported from a country is in fact consists of inputs from a complex international production network (Los et al., 2015; M. P. Timmer et al., 2015). And finally, a section will conclude how the promises FGP makes for the latecomers might be undermined by the production fragmentation in terms of theoretical assumption, which then will be examined with the case in the footwear industry in the following chapters.
2.2 Flying Geese Paradigm 2.2.1 Introduction
In late 1960s, Japanese economist Kaname Akamatsu proposed the developmental model of FGP, which metaphorically describes the process of regional development in East Asia as wild geese flying pattern of inverse V shape (Akamatsu, 1962; Kojima, 2000). In this model, Japan is placed at the leading role followed by Newly Industrialized Economies (Hong Kong, Singapore, S. Korea, Taiwan), then by ASEAN4 (Indonesia, Philippines, Malaysia, and Thailand), and finally the latest comers of China, India, and Vietnam (Ruan & Zhang, 2014). The catch-up process is driven by the rising costs of factors of production which lead to internal structural reconfiguration in more advanced economies, the declined industries then move countries with lower costs which function as receptacles for the more advanced economies (Bernard & Ravenhill, 1995; Kojima, 2000). It is worthwhile to mention that although the FGP addresses mostly on the pattern regarding the development of economy and the migration of industries, it is widely considered a product of international political economy and geopolitics instead of a model purely stem from the field of economics (Kasahara, 2004; J.-H. Wang, 2004). This is because the countries which took off before 1990s, namely Japan, Korea, Taiwan, Hong Kong, and Singapore, are all close allies of western bloc led by the United States. And these countries are at the frontline of containment policy, which was an instrument of the United States to prevent the expansion of communism (Mozingo, 1967). In this historical context of the cold war, the United States provided assistance to develop the industries in the region and ensured the security of the region with its military powers to build up a safe environment for these economies to grow (Mozingo, 1967; Ozawa, 2003; Stubbs, 2006).
The migration pattern of footwear industry matched the prescription of the FGP in the past five decades (Fan, 2008; Lin, 2011), and the thresholds to enter labor-intensive sectors like footwear industry is relatively low, it is a suitable entry point for the late comers to join the global production networks (Scott, 2006).
Nonetheless, the validity of the paradigm could be confronted with the uncertainty brought by the challenges of declining power of the U.S., technological progress of ICT and automation (Gereffi, 2014). Apart from its implication for the firms’ strategies, these challenges could also undermine the opportunities for the global south to catch up with the global north. Since the migration of industries is identified as a pathway to bring in employments and introduce innovations to the global south, and hence serves as an approach to generate sustainable economic growth and converge with the global north.
2.2.2 Stages of Industrial Development in the Original Model by Kaname Amakatsu There are three stages of industrial development in Amakatsu’s model with ‘wild-gees-flying’ pattern which were generated from his empirical observation from the textile industry in Japan during the Meiji era (from late 19th century to early 20th century) (Chen & Huang, 2009). And each stage can be defined by the sequences consisting of three phases which are respectively import, domestic production and export (Akamatsu, 1961, 1962; Kasahara, 2004; Kojima, 2000).
The first stage takes place when a traditional nation which is economically predominantly depending on the agricultural sector come in to the global trade system. In this stage, the nation possesses comparative advantages in the primary products while the comparative advantages held by other more industrialized counterpart are in the manufactured consumer goods for these products come with higher qualities and lower prices. In this condition, it would be a sensible choice for the relatively underdeveloped nation to export its agricultural goods in exchange to the manufactured consumer goods from more industrialized countries. In addition, Akamatsu (1961) also pointed out that the soaring growth in trade of the nation is mostly led by the trade which take place between the underdeveloped nation and the more industrialized , usually distant, nations. The trade between the nation and its bordering countries would remain rather stable, since the bordering countries often have similar economic structures and therefore share similar advantages with it. Thus, it would not benefit that much from having trade with its neighbors than with other industrialized nations.
After the nation has been joining the global trade system, it gradually steps into the second stage prescribed by Akamatsu (1961). The most noticeable development in this stage is that the production of consumer goods stars to emerge in the domestic
economy with the domestic consumers as the major destination to sell these products. The domestic production of consumer goods is made possible by the gradual concentration and accumulation of the purchasing power in the economy with the help of global trade, which makes it profitable to produce such products domestically as a consequence and therefore generate the incentives for the producers to manufacture consumer goods within the country. Furthermore, national industrial policies play an important role, since the industrial policies function as vital drivers to stimulate the movement of capital from the agricultural sector to manufacturing sector. Therefore, the developmental state, which the government of the state is heavily participating in the managing and planning of direction of industries and offering assistance to the selected firms (H. W. Yeung, 2017; H. W. C. Yeung, 2009), are considered one of the crucial factors for the success of the catching-up process in the FGP (H. Jiang, 2016).
As the domestic capital is increasingly attracted to the production of consumer goods, the country is homogenizing with the more industrialized ones steadily in terms of the structure of the economy. The import of consumer goods from the more industrialized countries stop increasing and start to decrease afterwards. However, though the import of consumer goods from the more industrialized countries has decreased, other goods which are required in order to produce consumer goods, such as machineries, still need to be imported from other countries. As a result, the economic transformation from primary industries to light manufacturing industries are leading to the shift of imported articles from consumer goods to capital goods in the second stage. One interesting pattern revealed by Akamatsu is the apearerance of “advanced differentiation”, which refers to the gap of different industries between countries in different stages of industrial development. Although the less advanced countries are becoming more capable of producing consumer goods and thus do not import as much consumer goods as it did prior to this stage, it still relies heavily on more advanced countries in terms of accessing capital goods. As a consequence, the countries are homogenizing in the production of consumer goods, but the gap in the production of capital goods remains and it even widens over time (Akamatsu, 1961).
In the final stage of Amakatsu’s flying geese model, the less developed country is not only competent enough to produce consumer goods to meet the needs of the market within the country, it is also building up its productivity in pursuance of exporting such products to other countries. According to his prescription, the destination of its exported consumer goods would be its less industrialized
neighboring countries for the most part, and in return it imports the primary products form its neighbors which were previously homogenous with it in terms of economic structure for they share more or less the same natural conditions.
Moreover, the country does not solely rely on more industrialized countries for the capital goods in the third stage. Instead, the production of capital goods can be observed now in the country as well. At this moment, the cycle of consumer goods has gone through the three stage of import, domestic production, and export respectively. On the other hand, the stage of capital goods is in the second stage of domestic production with the target market remaining with in the country, in this context, it can be predicted that the country will start to export its capital goods in the next stage and such cycle would go on and on for more advanced products (Akamatsu, 1961).
The upgrading of industry is the ultimate goal of the FGP, especially for the lead geese who are faced with the pressure of being caught up by the follow geese (Tsai, 2004). The Japanese Economist Terutomo Ozawa summarizes four phases of the industrial upgrading that will take place in the FGP from the experience of Japan’s development of industries. The first phase is the “Heckscher-Ohlin” industries, which means the labor-intensive industries such as textiles, footwear. Then the scale economies and heavy industries, such as steel and chemical industries, are built in the second phase. In the phase three, the assembly of the consumer durable goods like automobiles and televisions will take over, the Ford-style mass production can be seen in this phase. And the final phase is dominant by the “Schumpeterian” industries, which focus on the research and development of cutting edge technologies, new materials, and electronic products (Ozawa, 2001). The process of industrial upgrading and the migration of old industries are shown in the graph below.
Picture 1: migration of old industries (Okita, 1985)
This characteristic echoes with the product cycle theorem, which prescribe that the more advanced countries play as the major sources of innovation, while the latecomers can only take over the outdated industry that is no longer suitable to produce domestically in the country of origin (Vernon, 1966). And some see the relations between the lead geese and the follow geese in the process of industrial development as an extended and reginal specific version of the world system theory, which views the relations between the lead geese and the follow geese as that of the “core” and the “periphery” (Tsai, 2004).
To sum up, the original FGP has three major traits. The first one is that the cycles of not only consumer good but also capital goods follow the sequence of import stage, domestic production stage and export stage. Furthermore, when a country completed a cycle of a certain product, it would move forward to produce more advanced ones, therefore a pattern which a country’s production would move from primary products toward more sophisticated products could be observed. And finally, the position of more industrialized countries in comparison to less industrialized countries is consolidated through the pattern of “ advanced diffrenciation ” (Akamatsu, 1961;
Bernard & Ravenhill, 1995).
Figure 1: change of import, production and export of consumer goods and capital goods over different time of development. source: Akamatsu (1961)
2.2.3 Pro-trade-oriented foreign direct investment
Upon the original model of Amakatsu’s FGP, Kiyoshi Kojima address the importance of pro-trade-oriented foreign direct investment as well (Kojima, 2000). In the perspective of Kojima, foreign direct investment from the lead geese plays as a vital instrument to introduce capital goods, more advanced technologies and skills to the follower geese. In Kojima (2000), He also stated that the role of foreign direct investment is particularly important in the industries which require a large amount of labors. The incentives to relocate the production in the lead geese is especially higher in the labor-intensive industries.
In the figure 2, t1 represents the time that a new consumer good is introduced to a less industrialized country, since it does not have the technology and skills needed to produce the product, it imports the good from the more industrialized countries which are capable of manufacturing such product. Then the time proceeds to t2, it is the starting point where the less industrialized country begin to acquire the capacity required to produce the product, therefore, the import of such product start to decrease as the country can domestically provide this product to meet the demand of its market. The time of t3 is where the production capacity is high enough not only to satisfy the demand from the domestic consumers but also having a surplus of production, which allows it to begin exporting the product to other countries, probably with a more competitive price. While the import of the product continues to decrease and the export in increasing, the trade in balance is reached at the time point t*, which implies that the implementation of catching-up process is successful and the original model
by Amakatsu is finished at this time point (Kojima, 2000). But in Kojima’s model with pro-trade-oriented foreign direct investment, the country, which has become a lead goose in this product, will transfer capital, technology and other factors of production to the follow geese at t4, when the cost of production is too high in the orinal country. Therefore, a situation which local capital is put into a foreign country to produce the product abraod can be observed. It is illustrated that after the production in an industry reached its peak, it starts to invest in other lower-cost countries and produce in foreign countries which is represented by the curve Pf. The product manufactured in the abroad has three markets to serve. The first one is the market where the product is produce, the second one is exporting to other foreign countries. and finally, parts of the product are imported back to the home country which the industry was based previously, the phenomenon is referred to as “reverse import” and is shown by the curve M’.
Figure 2. source:(Kojima, 2000)
The development of footwear industry in Taiwan fits the trajectory
described by figure 2 (Cheng, 2011; Fan, 2008; Jou & Chen, 2001). The footwear industry started booming in Taiwan in 1960s, and later it reached its golden years in 1970s, when Taiwan took over Italy and become the largest footwear export in the world with 330 million pairs of shoes per year. Then the export of footwear peaked in the year of 1986 with 793 pairs, and started to decline each year gradually (Cheng, 2011). However, the decline of footwear export does not lead to the end of the industry in Taiwan, the Taiwanese footwear firms began to invest in China and Southeast Asia since late 1980s as a result of rising cost of production in Taiwan, which follows the prescription of t4 in Kojima’s model.
To conclude, the model modified by Kojima emphasize the production
cycle of consumer goods, which start with the import of the manufactured product, then production and export, then the additional curves of foreign
production and reversed import. In my opinion, the foreign production and reversed import curves reflect the cycle of production in the industry of real world more precisely then the production cycle of capital goods in Amakatsu’s original model. The curves of foreign production and reversed import represent the export of capital goods, but the ownership of means of production and the majority of economic profit are still in the hand of firms in the original country. In addition, from the Kojima’s model with foreign direct investment we can observe some patterns of production fragmentation already, which will be introduced in the following section.
2.3 Production fragmentation from the perspectives of Global Value Chain
2.3.1 Introduction
The production fragmentation is not a new phenomenon, however prior to 1990s, the production fragmentation existed mostly between different firms while the production was spatially limited within the same national border (Sven W. Arndt & Kierzkowski, 2001, p.4). But since the 1990s, the development of information and communication technology has made the long-distance coordination and commutation much easier than it was before. And the liberalization of international trade, which has led to the significant decrease of tariffs and non-tariff barriers, made the international trade unprecedentedly accessible. These two major factors have motivated the manufacturers to seek both production bases and market beyond their national border, and therefore gradually create a spatially fragmented production network (Gereffi, 2014; Sven W. Arndt & Kierzkowski, 2001, p.4).
In terms of international trade, the comparative advantages of final product possessed by a nation have been shifted to comparative advantages in different stages and components of productions (Sven W. Arndt & Kierzkowski, 2001, p.2). For instance, one country may manufacture the whole car from the engine to the frame to assembling of all components in the past. but in the present days, the engine would be designed and made in the country with the knowledge and technology, while the frame could be made in another country with abundant recourses like steel, and these items would be shipped to the third country with more affordable labor resources to assemble, and in the end the automobile is imported to the first two nations as the
final product (M. P. Timmer et al., 2015). Such shift in the different concept of comparative advantages has allowed more sophisticated products to be manufactured with the most economically efficient manners (Sven W. Arndt & Kierzkowski, 2001, p.2). In this context, examining trade in parts and components would be a better option to investigate which countries benefited more from the global trade than examining trade only in final products.
Nonetheless, the production fragmentation does not only occur in the dimension of internationality, it also took place in the aspect of organization. In other words, the process of production is no longer being taken care of by only one firm, multiple tasks and production processes are outsourced to different firms with different niche in the production network (Tasi et al., 2000). Such complexity embedded in international and inter-organizational production networks requires unconventional system of governance, which cannot be completed by the government or single firm along, but with the multiple entities with different level of power in the production network (Gereffi et al., 2005). In this context, the framework of Global Value Chain Governance is proposed, which enables us to investigate the how the global network of production is organized through different capacities and power dynamics between the buyers and suppliers (Gereffi et al., 2005; Humphrey & Schmitz, 2002). The Global Value Chain Governance can serve as an instrument for us to investigate how the organizational production fragmentation are taking place in the footwear global value chain, and most importantly what firms are having more power and benefiting the most in the process of production fragmentation.
2.3.2 Organizational & international production fragmentation
The framework of Global value chain (GVC) governance serves as a perspective to analyze the process of organizational production fragmentation by conceptualizing what value-added activities the firms participated, the power relations among the firms and their positions in GVC (Gereffi, Humphrey, Kaplinsky, & Sturgeon*, 2001; Gereffi et al., 2005). A GVC consists of all value-adding activities involved in the process of production, ranging from design to manufacture and to retail which take place in multiple firms and countries. Hence, it is a useful tool to assess the network of production in the increasingly interdependent world (Gereffi et al., 2005).
In Gereffi et al. (2005), five types of global value chain governance are identified according to three factors, namely complexity of transactions, ability to codify
transactions, and capabilities in the supply-base (See Table 1). Different types of GVC governance reflect different levels of power asymmetry between firms. The type of governance that requires barely any coordination is market, where the transactions and the specification of the products are simple and require barely any coordination between the buyers and the suppliers. When the complexity of the product rises, the modularity in the value chains also emerges. In the value chain that the codification and information of the product can be exchanged easily, the modular type of governance is formed. While in the case which the costs and difficulty to exchange the codification and information of the product is high, the governance would become the relational one. In both the modular and relational types of governance, the capacity of the suppliers are high, but it is easier for the buyers to change the suppliers in the case of modular value chain because the codification can be coordinated with more ease. And the buyers in the relational value chain have to make more effort to coordinate the information and codification to the new suppliers, therefore, it is more difficult for the buyers to change the suppliers in the relational value chain. When it comes to the value chain which both the complexity of the product and the power to codify the product is high, but the suppliers do not have high capacity like the ones in the modular and relational value chains, a captive type of value chain governance can be seen. The buyers have more power over suppliers as they can easily switch the suppliers whereas the suppliers do not have enough capacity to opt for other buyers. In this sense, the captive value chain governance represents that it is the suppliers whom are kept “captive” instead of the buyers. And Finally, when the complexity of the product is high but it is difficult to exchange the information and codify the product specifications and the suppliers with enough capacity cannot be find, the hierarchy type of governance is shaped. In the hierarchy type of value chain governance, the buyers have to establish the integrated production lines by themselves. The visualization of the five types of value chain is illustrated in the figure 3 below.
Figure 3: The visualization of five value chain governance type. (source: Gereffi et al., 2005)
The global footwear value chain was a captive system, in which lead firms have more power and influence over the suppliers in terms of codifying the standards since the products are highly complex and capacities of the suppliers are generally low (Schmitz, 2006; Thi et al., 2016). However, as some of the suppliers build up their capacities through innovation and vertical integration, some firms emerged as the intermediaries to help their clients to arrange the value chain consisting of small and medium sized suppliers, therefore, a transition from captive type of governance to relational type is observed in the footwear industry (Gereffi et al., 2005; Mao & Fan, 2010; Schmitz, 2006). Interestingly, such emergence of big contractors can be observed in the region across different kinds of consumer goods (Appelbaum, 2008). In the process of production relocation, intermediaries play as the pioneers to relocate the production lines for they require the most amounts of low-cost labors. Afterwards, the suppliers of these intermediaries follow them to the new location to maintain the spatial proximity (Mao & Fan, 2010)
Governance type Complexity of transactions Ability to codify transactions Capabilities in supply-base Degree of explicit coordination and power asymmetry
Market Low High High Low
High
Modular High High High
Relational High Low High
Captive High High Low
Hiearchy High Low Low
Table 1: Key determinants of global value chain governance (Gereffi et al., 2005) The activities the firms are participating in may not stay static, they can take part in higher value-added activities by coining upgrading strategies. In the GVC framework, upgrading strategies are categorized into four types (See Table 2). Process upgrading refers to increasing output with the same amount of input, this could be achieved by introducing more advanced technology, reorganizing the production management, or relocation the production lines. Product upgrading means the upgrading from the original products to more sophisticated and more valuable product lines. Functional upgrading could be accomplished by acquiring new functions or dropping the existing functions in the industry, which increase the skill content of activities. And the final type of inter-sectoral upgrading, or chain upgrading, refers to movements from current industry into a new but often related one (Gereffi, 2014; Humphrey & Schmitz, 2002).
Type of
upgrading Definition
Example in the case of footwear industry
Process upgrading
transforming inputs into outputs more efficiently by reorganizing the production system or introducing superior technology
Automation of production Relocation to a lower-wage country Product upgrading
moving into more sophisticated product lines (which can be defined in terms of
firms increased unit values)
Specialization in Gore-Tex certified footwear Customization of footwear Functional upgrading
acquiring new functions (or abandoning existing functions) to increase the overall skill content of activities Development of new materials Establishment of private brands Chain/ inter-sectoral upgrading
firms of clusters move into new productive activities
Moving to garment
industry Table 2: Typology of upgrading strategies (Humphrey & Schmitz, 2002)
On the other hand, the work of Timmer et al. (2015) clearly showcases how production fragmentation took place in the German automobile industry with the calculation form the world input-output table. It analyses the final automobile output of Germany from the year of 1995 to 2008 with the origin of intermediate input. The result indicates that the domestic input has decreased by 12.8% from 1995 to 2008, while the input originating from other European Union countries has increased by 5.4% in the span of 13 years. This finding aligns with the Baldwin & Lopez-Gonzalez (2015), which observes a growing trade of intermediate goods within the region, forming a spatially dispersed production network and deepening the interdependence within the region. Similarly, the work of Suder, Liesch, Inomata, Mihailova, & Meng (2015) the production networks are highly integrated in the region of East Asia, especially among Japan, Taiwan, South Korea, and China, making it one of the most interdependent regions in the world. The findings of this study imply that the impacts the production fragmentation has on the validity of FGP are not necessarily negative,
it could strengthen the effect of the paradigm as well. However, each industry is exposed to the different level of production fragmentation, and the level of production fragmentation may be significantly different across different industries (Feenstra & Jensen, 2012). And a study on the footwear industry in Italy from 1996 to 2000 shows that most value adding activities still take place domestically (Amighini & Rabellotti, 2006). Correspondently, the analysis of conducted by Los et al. (2015) suggests that textile products has experienced only 3.7 percent of increase in foreign value added share from the year 1995 to 2008, while the same figure is only 1 percent for the leather products. Therefore, it would be interesting to investigate whether the footwear industry in East Asia and Southeast Asia is facing significant production fragmentation in the international dimension.
2.4 The Challenges faced by FGP in the Era of Production fragmentation
This section will put the theoretical framework of FGP in the context of production fragmentation to make linkage and comparison between the two concepts to illustrate what challenges FGP could be face with and where the potential of the paradigm may lie in terms of theoretical prescription.
First of all, the frameworks of FGP and the production fragmentation are not completely incompatible and rival to each other. As previously disclosed, the concept of production fragmentation is implicitly addressed in Kojima’s model with pro-trade-oriented foreign direct investment, which depicts the international movement of capital looking for more cost efficient locations to manufacture the product (Kojima, 2000). On the other hand, it is observed in multiple studies that the production fragmentation is often accompanied with the deeper integration within the region, making the countries in the region more dependent on one another to distribute different tasks in the process of production (R. Baldwin & Lopez-Gonzalez, 2013; Suder et al., 2015; M. P. Timmer et al., 2015), which could amplify the effect of FGP. Another comparable feature between the two concepts is the idea of upgrading. In the global value chain governance, there are four channels to upgrade in order to participated in more value adding activities, while the FGP only address the upgrading from one industry to another, which is similar to the chain upgrading in global value chain governance. Also, the upgrading strategies in the global value chain governance are meant for the firms, and the FGP address the industries
upgrading in the unit of the country. Though the upgrading strategies in the global value chain governance does not aim at the convergence of latecomers, it would be interesting to investigate whether the strategies can reach have convergence effect as prescript in the FGP.
Nevertheless, some assumptions and the global environment which allows FGP to be realized are facing the fundamental changes. The first change is that the hegemony of the United State in the international arena is facing the challenges brought by the rise of China, both in terms of politics and economics (Gereffi, 2014; Liu, 2017). The FGP are developed and conceptualized under the global environment of cold war, in which the United States provided nurturing conditions for the East Asian economies to prosper by implementing the containment policy to prevent the proliferation of communism, providing military protection and development assistances (Mozingo, 1967; Ozawa, 2003; Stubbs, 2006). And the United States has been the major destination of the export form the region, which accounts more than 40 percent of the total export (Tsai, 2004). Therefore, the crisis faced by the hegemony of the United State are also challenging the preconditions for FGP. The rise of China is challenging Japan’s position as the lead goose in the formation, and the enormous scale of China’s economy is also absorbing the economic potential of other countries in the region, leading the restructure of order in East Asia and South East Asia (He-Hua, 2015). Moreover, some scholars in the region has claimed the arrival of “Post-Flying Geese” era due to the emergence of China’s economic and political power in the region of East Asia as well as in the world, implying that FGP can only explain the trajectory of development in the past and it is losing relevance in the global environment of the present days (H.-H. Jiang & Huang, 2016; Tsai, 2004).
Apart from the changing global environment, the roles of governments are also becoming less important. In the prescription of FGP, the developmental state, which is deeply involved in the planning of the development industries, is considered an important precondition of the catching-up process (Tsai, 2004). Although the heavy intervention of state in the development of industry is regarded as contradicting the basis of Washington Consensus, which views free and open market and deregulation as the most important doctrines, it is considered an effective approach in terms of promoting the development of industries (Stiglitz, 2002, p201). However, the emergence of multinational enterprises has led to the compromises of state sovereignty (Kobrin, 2009), undermining the potentials of the state to lead the direction of development of industries. The countries have to meet the needs of the
multinational enterprises in order to attract the foreign investment, which often implies the government cannot provide adequate protection for its labors (Mosley & Uno, 2007). Most importantly, the export of product serves as the most vital momentum for the economic development and the source to accumulate the capital. But most of labor-intensive industries, which most of the less industrialized countries have comparative advantages, are affected by the emergence of the buyer-driven commodity chains in the early 1970s. And the rise of the buyer-driven commodity chains has granted more bargaining power to the global buyers in the global north (Gereffi, 2014). In other words, the restructure of power dynamics in the global value chain has made it more difficult for the latecomers to converge through the implementation of export-oriented industrialization. Finally, the FGP only address the export of final goods, assuming most of the value adding activities are taking place within the country where the final product is exported from. However, in the era of production fragmentation, trade in intermediate goods, i.e. parts and components, has been increasing significantly in the world, and the region of East Asia is no exception (Athukorala & Yamashita, 2006). The increase of trade in intermediate goods implies that the final product exported from a country is in fact consists of input from multiple countries, which diluted the actual gain of the exporting country of the final goods in terms of values and employment generated (Helg & Tajoli, 2005; M. P. Timmer et al., 2015)
From the comparisons stated above we can tell that there are a number of challenges face by the FGP. The changing global environment with the emergence of China has disturbed the nurturing condition and economic orders that allows FGP to be realized. And private sector is gradually taking over the role of the state, which is vital to the success of FGP, in terms of industrial development, leaving less instruments for the state to impose. And the export of the final products as a channel to accumulate capital and converge with lead geese has been undermined by the increasing proportion of trade in intermediate goods. However, there are some new opportunities emerged from the restructure of production as well. The deepening trade integration of the region implies that the industrial upgrading in the lead gees may require the participation of the follow geese in the process as well. And the shift of economic dominance from the state to the private sector may leave more flexibility to the private sector, expanding the possibilities for the private sector to join higher value adding activities. In the following chapter, the research design of the study will
be elaborated to show how this thesis project will investigate the interaction of the concept of the FGP and the production fragmentation.
3 Methodology 3.1 Introduction
The chapter of methodology concerns the research questions, the operationalization of concepts, what and how data are collected and analyzed. The methodology is aimed at examining the impact production fragmentation have on the validity of FGP, which is made clear in the research questions. To investigate the interactions between the two concepts, key features of the concepts, which are identified in the chapter of theoretical framework, need to be operationalized into indicators in order to measure the changes in the figure. For the operationalization of production fragmentation, different approaches are used to measure the dimension of organizational fragmentation and international fragmentation. The work on the global value chain governance by Gereffi et al. (2005) is adopted to compose the indicators for the organizational production fragmentation, which intend to inspect how the tasks and the provision of materials are distributed among the firms in the global footwear value chain, and which firms have the power and capacity to manage the production network and codify the specification of the products. As for the international production fragmentation, the indicators used in the work of Timmer et al. (2015) is utilized to calculate how much proportion of the final product consists of intermediate goods from foreign countries. Furthermore, several studies address the increase of trade in intermediate goods, so the proportion of output as intermediate goods is taken into the indicators to measure international production fragmentation as well (Athukorala & Yamashita, 2006; Cingolani & Tajoli, 2018; Helg & Tajoli, 2005). For the indicators of the FGP, the indicators are extracted from the major features of the works of Akamatsu (1962) and Kojima (2000) Since the quantitative studies to examine the effectiveness of the FGP usually implement RCA method, which is aiming at figuring out the comparative advantages of a county across a set of industries (Chen & Huang, 2009), the existing quantitative indicators of the FGP do not fit the need of this study. Therefore, in order to create the quantitative indicators that fit into the context of this research, I translated the qualitative prescription into the quantitative indicators based on The World Input-Output table, which is the same
source as the indicators for the international production fragmentation.
The study is conducted with the mixed method design with interviews and mixed model analysis. For the organizational production fragmentation, 3 Taiwanese firms are interviewed, all of the firms are intermediaries which severs as the central point of communication and the management of production network between lead firms in the global north and production side in the global south. And the international production fragmentation is assessed with the quantitative method of mixed model analysis. The data required for the analysis of international fragmentation are retrieved from The World Input-Output Database constructed by Groningen Growth and Development Centre, University of Groningen (Groningen Growth and Development Centre, n.d.), Asian Development Bank. The details of data selection and the method selected will be elaborated following sub-sections.
3.2 Research Question
Main Question
How is the validity of FGP affected by the production fragmentation in the case of footwear industry in the region of East Asia and Southeast Asia?
Sub Questions
What is the level of production fragmentation in the footwear industry throughout time in different countries in the region?
How does the inter-dependency between follower geese and the lead geese change in the process of production fragmentation?
How are countries in different tiers of FGP affected by the production fragmentation in different ways and different degree, and what are the possible reasons for such difference?
3.4 Operationalization of Major Concepts
Concept Dimension Variable Indicator
International production fragmentation Trade in Value added input
Percentage of input by other countries (foreign input / total input)
output
Percentage of intermediate goods in total output
(intermediate goods/ total output)
Prescription of FGP
Domestic industrial development
The industry would follow the import, production, export development pattern
The export of final goods in the footwear industry would grow overtime before the industry start to migrate to other country
The industry would be more developed along the time and therefore would manufacture more
Value added by the domestic footwear industry compared to the total output (value added/ total output)
sophisticated
products. Percentage of higher skilled labor
in the industry would grow overtime. Interdepende nce with other countries in the region
The migration of the industry
The industry would migrate to a new location with lower costs of production (follow geese) when the costs in the original country is too high
follower geese would rely on capital goods from lead geese
Foreign direct investment from the firms in lead geese countries to the follower geese countries
Intermediate goods imported from other countries in the region (intermediate goods from other flying geese country/ Total import of intermediate goods)
The flow of finished goods in the region
The percentage of finished goods to other flying geese country.
(finished goods to other flying geese country/ Total export of finished goods).
Global Value Chain Position
Lead firms
Do they have the most power to codify the standards of production?
Intermediaries
Do they make final products for lead firms?
Do they manage the suppliers for the lead firms?
Do they have limited power to codify the standards of production?
Second tier suppliers
Do they directly receive orders from intermediaries?
Do they produce manufactured materials of footwear?
Third tier suppliers
Do they receive orders from second tier suppliers?
Do they produce raw materials of manufactured materials?
Upgrading
Process upgrading
Will the firm relocate to places with lower costs of production? Will the firm automate the production?
Will the firm improve the manage system to reduce the costs of production?
Product upgrading
Will the firm specialize in Gore-Tex certified products?
Will the firm customize its products?
Will the firm manufacture other higher-value products?
Functional upgrading
Will the firm develop new materials or design new products? Will the firm establish private brand or retail chain?
Chain/ inter-sectoral upgrading
Will the firm move to other chains other than footwear?
3.5 Units of Analysis
The study will predominantly make use of quantitative method to explore the production fragmentation in the dimension of national fragmentation. The quantitative data acquired is country-based input-output table and social-economic account. Therefore, country will be major unit of analysis in this study. However, while the main focus of this study is special fragmentation of footwear production, this study will also touch on the dimension of Inter-organizational fragmentation. It will come with the example of the footwear global value chain which major Taiwanese footwear firms are in, focusing on the arrangement of international and inter-organizational production. In other words, this research will also be facilitated with a small portion of qualitative data, in which the unit of analysis is the footwear firms. The difference in the units of analysis is because of the fundamental difference in nature of these two dimensions. However, the subject in the FGP lies in the scale of country, therefore the units of analysis will be at the level of country.
3.6 Research Design
As stated above, methodology adopted in this study is mix method while leaning more toward quantitative method. To answer the main research question, which is “how FGP is affected by the production fragmentation in footwear industry”, the countries to be analyzed and the tiers they are in have to be determined first. In V. Tran (1992), it is claimed that the migration pattern of the industry in the region starts from Japan to newly industrialized countries( South Korea, Taiwan), then to ASEAN4 (Malaysia, Thailand, Indonesia, Philippines), China and Vietnam. Therefore, the countries mentioned above, will be subjects to analyze. The reason to exclude Singapore and Hong Kong form the analysis is that both of the economies are city-based, which is not suitable for the mass production industries (Kojima, 2000). Another study observes that China actually take over the more industries directly from Taiwan and South Korea
instead of from the ASEAN4, therefore, it should be classified as the tier 3 rather than tier 4 (Chen & Huang, 2009). This finding echoes with other studies on the migration pattern of footwear industry, which also observed that the footwear industry migrated from Taiwan and South Korea directly without settling in the ASEAN4 first (Fan, 2008; Tasi et al., 2000). In this sense, it would be more reasonable to put China in the tier 3 instead of tier 4 in this study, which focuses on the development of the footwear industry in this region. The quantitative data were retrieved from Asia Development
Bank Multi-Region Input-Output Database, World Input-Output Database, and
processed with MatLab and SPSS. And in terms qualitative data for the organizational fragmentation, three Taiwanese firms are interviewed, the locations where these multinational firms are operating covers Taiwan, China, Vietnam, Indonesia, Myanmar, and Cambodia. The interviews questions focused on how the global footwear production network is organized, how the tasks of production are distributed in different locations, and whether these firms are able to participate in more value adding activities via upgrading strategies. Then assess how the dynamics lie in the global value chain can have impacts on the realization of the FGP.
To calculate the level of international production fragmentation, the quantitative
data of the world input-output tables is required. Comparing to the traditional data concerning global trade, which only contain the flow of goods among the countries, the input-output table provides us with a new approach to analyze how final product consists of inputs from other industries and other countries. The construction of the world input-output tables is an complicated task, and the simplified steps are listed as follows. The basis of the world input-output tables is national supply and use tables, which shows the flows of input and output among the industries of the country. As the national supply and use tables are not published every year, the national accounts statistics such as employment and the output of production, which are usually revised annually, are being used as benchmark to calculate the annual supply and use tables. And finally, the data across countries are harmonized with the statistics of international trade statistics (Dietzenbacher, Los, Stehrer, Timmer, & de Vries, 2013; M. Timmer et al., 2012). And once the statistics are calculated, the layout of the world input-output tables will look like the figure 4 below.
Figure 4: the layout of the world input-output tables, edited from (M. P. Timmer et al., 2015)
In the figure 3, nine blocks are marked regarding the input and output from industry 1 to industry N in country 1. Block A in light orange represents the domestic trasactions of the intermediate goods among the industires. The statistics in the block B are the export of intermediate goods from country 1, which will later be put into the industrial production of other countries. And the block C is the domestic consumption of the final goods produce within the country 1. Block D showes the export of the final goods from country 1. And the numbers in the block E2 of total use are sums of each
column in the block A , B, C, and D. The imports of final goods for the consumption of country 1 are put in the block H.As for the input side, the block F represents the intermediate goods from other countries to the production of country 1. And the real value added of the industries in country 1 is shown in the block G. Therefore, the gross output of the block E1 can be calculate by adding up the domestic intermediate input (A), foreign intermediate input (F), and the value added by the industry itself (G). One important note is that the values in the block E1 equal to the values in the block E2, which makes the gross input and gross output balace. From the layout of the world input-output tables we can make sure that all of the quantitative indicators regarding the international production fragmentation and the FGP, with the only exception of employment data, can be extracted from the world input-output tables.
The input-output tables are acquired from three sources. The first one is The World Input-Output Database released in 2013, which is constructed by Groningen Growth and Development Centre, University of Groningen (Groningen Growth and Development Centre, n.d.). The reason for choosing the 2013 release instead of the latest version, which released in 2016, is that the 2013 release has a specific category for leather and footwear while the 2016 release merges the footwear industry into broader category of wearing apparels. In addition to The World Input-Output Database, Groningen Growth and Development Centre also released the industry-based Socio Economic Accounts, which include the data of employments in the footwear industry of the countries mentioned above. The Socio Economics Accounts will be utilized to analyze the change in employment in the footwear industry in these countries.
Based on the 2013 release of The World Input-Output Database, The Asian Development Bank has put more Asian countries in the dataset, which include Thailand, Malaysia, Philippines and Vietnam (Mariasingham, n.d.; The Asian Development Bank, n.d.). However, the data for the additional four countries only cover the years of 2000, 2005-2008, and 2011. Therefore, they may not as complete as the ones in the original dataset.
Country Tier The World Input-Output
Database
Socio Economic Accounts (Source of the employment data)
Japan 1 1995 - 2011 1995 - 2009
Taiwan 2 1995 - 2011 1995 - 2009 Malaysia 3 2000, 2005-2008, 2011 - Thailand 3 2000, 2005-2008, 2011 - Indonesia 3 1995 - 2011 1995 - 2009 Philippines 3 2000, 2005-2008, 2011 - China 3 1995 - 2011 1995 - 2009 Vietnam 4 2000, 2005-2008, 2011 -
Table 3. the quantitative data available for analysis
On the other hand, the qualitative data was collected in Taiwan for another closely related topic, which focused on the upgrading strategies of Taiwanese footwear intermediaries, the information of the interviewees are listed below in the table 4. Despite the fact that these data are not particularly collected for this topic, they could serve as additional sources to have a preliminary understanding of inter-organizational production fragmentation in footwear industry. As the previous main theme of this project was based on the upgrading strategies in the global value chain of intermediary firms, which have headquarters based in Taiwan while having production lines in China and Southeast Asia, the qualitative data was collected from the three of the biggest footwear firms in Taiwan with semi-structured interview with the employers from the firm. And all of the three firms function as intermediaries in the footwear global value chain. The first firm is Shiang Shin Bao Corporate (SSB), the size of SSB is the smallest among the three firms interviewed for this project. The second one is Feng Tay Group, the group employs one hundred ten thousand people globally, which makes it the second largest footwear firms in Taiwan. Nike is the most important client of Feng Tay Group, the orders from Nike accounts for around 80% of its annual turnover and over one sixth of Nike’s footwear are manufactured by it. Apart from that, it also manufactures sport wears specialized for skiing and skating (Feng Tay Group, 2017). The last firm is Fulgent Sun Group, which is the third biggest footwear intermediary firms in Taiwan with 22 thousand employees in total (Fulgent Sun Group, n.d.). Unlike SSB and Feng Tay Group, which manufacture primarily sneakers, Feng Tay Group specialized in outdoor footwear and most of its production lines are certified by Gore-Tex. Some of its well-known clients include The North Face, Mammut, and Timberland.
The Taiwanese employers of these firms are interviewed, and the interview questions are composed of three major components. The first part is related to their
strategies of relocation, which intended to understand what are the main factors these firms would take into consideration when they are relocating their production lines, which locations are they going to build up more production capacities, and what factors make them retreat from a location. The relocation patterns of these firms serve as a vital indicator to determine the validity of the FGP, since the migration of outdated industries from the lead geese is a major channel to catch-up by receiving capital and technologies for the follow geese. The second part of the interview questions concerns the composition and the power dynamics within the global footwear value chain, the questions include how do they meet the needs of the lead firms and what are the compositions of their suppliers. By investigating the relations between the firms and the composition of the network, we can then infer that what firms can join the footwear global value chain and who benefit the most in the in the process of production fragmentation. Finally, the third part of the interview questions consist of the upgrading strategies of these firms. Although the relocation of production lines is also an approach of upgrading strategies, it is singled out to the first part of the interview for it is directly related to the prescription of the FGP. However, other forms of the upgrading strategies are interested as well in order to understand whether the upgrading in the FGP can take place in other forms from the perspectives of the global value chain governance.
code Name Gender Age Firms Position Base
A Kai Yu M 28 Fulgent Sun
/Decathlon
Production manager
Vietnam/ Taiwan
B Ethan Lee M 26 Fulgent Sun/
SSB
Automation Specialist
China/ Vietnam
C Yan Ning F 25 Fulgent Sun/
Feng Tai
Product
Developer Taiwan
code Name Interview date Language Types of record
A Kai Yu
2019/ Feb./ 20
Mandarin
notes
2019/ Feb./ 23 audio recording
2019/ Mar./ 27 notes
B Ethan Lee 2019/ Feb./ 23 Mandarin audio recording
Table 4: information of interviews
3.6.2 Data Analysis
Mixed model analysis, also known as multilevel models, is a linear model method that is used to analyze data that the observations are not independent, which implies that the errors may be correlated. This type of data is common in the situation where the measurements are made repeatedly on the same subject, or the subject pool can be divided into smaller groups where the individual within the group share more similarity (Goldstein, 1986; Hox & Maas, 2005). For example, in a supposed experiment to test the effect of different teaching methods, the measurement should consider the individual level as well as the upper level of different classes. It is still possible to conduct general linear model analysis without considering the effect of nesting, however, it could result in the violation of fundamental assumption of general linear model analysis that the error terms are uncorrelated. On the other hand, the multilevel model analysis does not assume that every single observation is independent from one another (Garson, 2012).
As the data is time series observation of flows of intermediate and final goods. The research adapts the mixed model analysis, in which we can place the observations of each year in the level one, and the country in the level 2. In other words, the time of observations are nested in the countries, because the observations with in each country could be highly correlated. The first step is to test the effect of time and tier on the international production fragmentation, and whether the effect of time has significant differences across different tiers. Therefore, the independent variables would by time and tier, while the dependent variable would be the indicators of international production fragmentation, namely the percentage of foreign input as intermediate goods and percentage export of intermediate goods. The different independent variables are put in different level according to the relations between their properties and the level. The structure of data in the mixed model is shown in figure 5 below, from this figure, we can tell that the first level is the observations made in different time points, and the time series data can be nested in the second level of individual country. Therefore, we shall put the independent variable of year in the first level, and put the independent variable of tier, which is a property comes with the country, in the level two of country. The results of mixed model analysis yield two groups of parameters. The first one is fixed effects, which represent overall effects of
