Meeting the challenge of China: the Vietnamese garment industry in the post MFA era

Meeting the challenge of China: the Vietnamese garment industry in the post MFA era KENTA GOTO, * KAORU NATSUDA AND JOHN THOBURN * Faculty of Economics, Kansai University, Japan gotoken@kansai-u.ac.jp College of International Management, Ritsumeikan Asia Pacific University, Japan natsuda@apu.ac.jp School of International Development, University of East Anglia, UK j.t.thoburn@uea.ac.uk (corresponding author) Abstract Although China has diversified into sophisticated, higher value-added exports, it is still a formidable competitor in global markets for basic labour-intensive products. It is the world s largest exporting country of textiles and garments, the archetypical driver of industrial growth both in developed countries in the past and in most newly industrializing countries more recently. When the export restrictions under the Multi-Fibre Arrangement (MFA) ended at the start of 2005, it was predicted that China would greatly increase its market shares at the expense of most competitors, except perhaps India. Vietnam has proved to be an effective competitor in the garment industry in markets where China is dominant. In this article, we investigate how key export-oriented garment suppliers of Vietnam have been coping with competitive challenges in the post MFA era at a time when global buyers have been reorganizing their international production networks. We emphasize the influence of different global value chains on upgrading since Vietnamese suppliers switched to the US market after the implementation of the US Bilateral Trade Agreement in 2001. We note the uneven performance of Vietnamese garment suppliers, with some lagging behind others in upgrading and competitiveness, and their different responses to Vietnam s growing labour shortages. We base the article mainly on interviews conducted over the 2001 2008 period with garment companies and global buyers in Vietnam, Hong Kong and China. Keywords GARMENT INDUSTRY, TEXTILES, GLOBAL VALUE CHAINS, MULTI-FIBRE ARRANGEMENT, CHINA, VIETNAM Global Networks 11, 3 (2011) 355 79. ISSN 1470 2266. 2011 The Author(s) Journal compilation 2011 Blackwell Publishing Ltd & Global Networks Partnership 355

Kenta Goto, Kaoru Natsuda and John Thoburn In this article, we investigate how Vietnamese export-oriented garment suppliers are coping with challenges from international integration. Garments are Vietnam s largest manufactured export item, and the abolition of the Multi-Fibre Arrangement (MFA) has increased international competition, especially from China. Although China has diversified into more sophisticated, higher value-added exports (Lall and Albaladejo 2004), it is still a formidable competitor in global markets for basic labour-intensive products. It is the world s largest exporter of textiles and garments (WTO 2009). From 1974, the MFA s elaborate quota system strictly controlled international trade in textiles and garments. Under this, the largest garment markets in the world, including the USA and EU, restricted their imports to protect their domestic suppliers. They tightly specified the restrictions on volume by product category and export origin. In 1994, during the Uruguay round of the GATT negotiations, an Agreement on Textiles and Clothing (ATC) gradually abolished the MFA quota system over a ten-year transition period that would last until 31 December 2004. Thus, after 2005, trade in textiles and garments was to be integrated into normal WTO rules (Yamagata 2007). The abolition of the MFA 1 provided competitive suppliers with opportunities to expand their exports further. At the same time, it also meant that less competitive exporters would face threats because supplies from more competitive sources could now substitute previous exports guaranteed under the quota system. The predictions were that China would triple its share of the US garment market and increase its share in the EU by more than half, with other producers shares falling, except for India s (Nordås 2004). The easiest way for a country like Vietnam to counter such threats was to reduce costs. As the garment industry is very labour intensive, much of this would translate into lowering wages when productivity levels remained stable. However, reducing costs has its drawbacks and to realize sustainable development of the industry, it is important to raise productivity continuously. In this article, we look at what kinds of qualitative changes the abolition of the MFA has brought to the Vietnamese garment industry, what strategies key suppliers in the industry have applied in response to such changes, and what this implies for its development trajectory. We shall analyse the industry in terms of the positioning of Vietnam s major garment suppliers within international production and distribution networks. We use the global value chain (GVC) framework to illustrate how individual firm level strategies have resulted in process, product, or functional upgrading. We combine an overview of the Vietnamese garment industry with a longitudinal study of ten large garment producers, mostly state-owned or formerly state-owned, interviewed in 2001 and again in 2007. The 2001 survey included 59 garment companies, of which 23 were export oriented. Of these 23, seventeen were state-owned enterprises (SOEs) and six private firms. As a comprehensive list of garment suppliers was unavailable at the time of interview (and still is, as far as we know), we drew the companies from a publication, Vietnam Textile Garment Industry the Present and Future, which the state holding company Vinatex (Vietnam National Textile and Garment Group) produced in 1997. This book included a list of garment (and textile) companies that were operating at the 356 2011 The Author(s)

Meeting the challenge of China time of publication, which in 2001 seemed exhaustive. We selected the companies from this list and gave priority to large export-oriented garment producers, which tended to be SOEs. We contacted almost all the large SOEs on the list (which were under the Vinatex umbrella), but only 23 agreed to in-depth interviews. To be representative, we included smaller ones too, which were entirely domestic oriented. For the smaller ones we used a more or less randomized sampling (systematic random sampling method). Getting appointments for these firms was more difficult, but we interviewed 36, which were located in Hanoi, Hai Phong, Nam Dinh, Tay Nguyen, Dong Nai, Bing Duong, and Ho Chi Minh City. Data for 2007 include interviews with ten garment suppliers, predominantly SOEs or former SOEs selected from the 23 companies interviewed in 2001. A purely randomized sample would be difficult to implement for these types of companies, and we cannot reveal more details for reasons of preserving anonymity. Note that ten companies is a large sample in relation to the number of companies operating under Vinatex. In the early 2000s, Vinatex had 42 member textile and garment companies, including all the centrally run textile and garment SOEs. At that time, Vinatex accounted for between 30 and 40 per cent (the amount varied according to different sources) of Vietnam s garment and textile exports (Nadvi and Thoburn 2004b: 114). In addition, our chosen companies would be unlikely to face the same difficulties gaining state help to contact global buyers that Thomsen (2007) documented for some private companies in Vietnam. Interviews in 2008 with other company and key informants provide an additional perspective. It proved impossible systematically to interview the same companies at the time of the 2008 interviews, though we did visit some of them again. In the next section, we briefly discuss global value chains in relation to entry barriers and upgrading before going on to present an overview of the Vietnamese garment industry. In the core of the article, the fourth section, we detail the longitudinal study of ten large Vietnamese garment producers. Global value chains, entry barriers and upgrading Analysis of GVCs looks at the process of selling a product from the supply of raw materials to its final distribution and marketing (see Gereffi and Memedovic 2004; Palpacuer et al. 2005). A GVC, however, is more than a set of input output relations spread over a number of countries. Economic agents at certain stages of the chain control the entry and distribution of activities over the chain; in other words, they exercise governance and they can earn rents (surplus profits see below). Garments are buyer-driven GVCs, where barriers to entry and rents are concentrated at the retail end. In clothing GVCs, a number of buyers without factories of their own generally organize production. The needs of global buyers generate barriers to entry, in the sense that they include requirements about production volume capabilities, product quality and the services (including compliance with international social and environmental standards) that some suppliers are able to offer and others not. These barriers are 2011 The Author(s) 357

Kenta Goto, Kaoru Natsuda and John Thoburn over and above those necessary simply to enter the industry to produce for the domestic market, which are low in the case of garment production, based on low capital costs and low wage labour. Access to particular markets, such as available trade preferences, for firms from the country in question also influence buyers sourcing patterns. Lead times that suppliers can offer are also an important determinant of a buyer s sourcing. The requirements of different buyers, particularly from different markets, may differ, with some buyers requiring more in the way of services their suppliers can perform. There is also evidence that buyers requirements have been tightening, focusing on a smaller number of supplier countries and on a smaller number of suppliers within each country (Nadvi and Thoburn 2004a: 260 1). Barriers to entry generate economic rents, in the sense of profits arising from the scarcity that the barriers create, insulating the producer from competition, although most rents in garments occur at the retail end and are associated with control of brands and market outlets (Kaplinsky 2005: 62). However, though increased buyer requirements generate barriers to new entrants to the chain, they do not always increase rents for existing producers who upgrade to meet them: the upgrading may be necessary simply to maintain a foothold in the chain. The concept of upgrading is a central part of GVC analysis. Upgrading is a form of innovation that generates rent if it occurs ahead of that of rivals (Kaplinsky 2005: 63). There is a wide consensus in the GVC literature that one can most usefully categorize upgrading within a GVC into three broad areas product upgrading, process upgrading and functional upgrading (see for instance Gereffi and Memodovic 2004; Kaplinsky 2005; Kaplinsky and Morris 2001; Palpacuer et al. 2005). Product upgrading involves producing new products or improving the design or specification of existing ones; process upgrading involves reducing costs or shortening lead times; and functional upgrading involves successfully taking on new functions, such as design, labelling or materials sourcing. Some writers have introduced additional categories. For example, Palpacuer et al. (2005: 412) write of volume-based upgrading where costs can be lowered if orders are large enough to generate economies of scale. We think this is best included under process upgrading, although in some circumstances it can be useful to separate it out. Producers sometimes are able to learn from buyers about how to upgrade (Schmitz and Knorringa 2000). Overview of Vietnam and its export oriented garment industry Vietnam has been growing fast since the implementation of the Doi Moi economic reform policy in 1986. Its economic growth has been particularly strong since it started establishing trade relations with Western economies in the 1990s, when the economy had the fastest export growth rate in the world. While growth slowed 358 2011 The Author(s)

Meeting the challenge of China somewhat after the Asian financial crisis in 1997, overall it remained robust (Thoburn 2007, 2009). Until the late 1980s, Vietnam s textile and garment (T&G) exports were predominantly to the former Soviet bloc, but this trade declined precipitously after the collapse of the USSR. Vietnam s entry into export garment production for Western and Japanese markets has required access to GVCs. The tendency to source from established East Asian foreign investors within GVCs the well-known triangular manufacturing system (Appelbaum 2008; Gereffi 1999) is breaking down to some extent as more and more buyers establish local offices in major producing countries and become more expert in assessing local companies. Foreign-owned production remains important, however, although its importance varies considerably from country to country. In the extreme case of Cambodia, for example, almost all production is in the hands of foreign-invested enterprises, mainly from Korea, Taiwan and Hong Kong (Natsuda et al. 2010). In Vietnam, in 2008 foreign firms accounted for 32 per cent of textile and 45 per cent of garment output, compared with 26 per cent and 25 per cent respectively in 2000 (figures from www.gso.gov.vn). In Vietnam s main competitor, China, in 2008 foreign invested companies produced 42 per cent of garment gross output and 23 per cent of textile gross output. 2 The share of foreign companies in both Vietnam s and China s exports would probably be higher than the production share in the case of garments, and lower (for direct exports) in the case of textiles. Unlike China, Vietnam s (direct) exports of textiles are minimal, and its T&G export expansion has focused almost exclusively on garments. 3 In Vietnam, foreign garment companies are almost wholly export oriented (for further details see Thoburn 2010: 10 11). In both Vietnam and China, state-owned or semi state-owned producing enterprises have been important in maintaining a local presence. In 2008, Vietnam s state owned enterprises still produced 27 per cent of the country s textile output and 10 per cent of its garment output. This was despite those proportions being considerably lower than in 2000, when the shares were 51 per cent for textiles and 32 per cent for garments, for SOEs have been (semi-)privatized and the private sector has grown. On average, garment manufacturing SOEs in Vietnam have been better equipped in terms of capital and access to foreign markets compared with private suppliers, and have been playing important roles in the export oriented industry (Goto 2003). The state holding company Vinatex was generating between 30 and 40 per cent of Vietnam s T&G exports in the early 2000s (Nadvi and Thoburn 2004b: 114). In China, in the late 1990s, SOEs produced 36 per cent of textile gross output, though only 7 per cent of its garment output (UNCTAD 2002: 152). 4 By 2008, however, the privatization of SOEs had brought that figure down to 3 per cent and 1 per cent respectively. 5 The private sector accounted for 50 per cent of textile and 41 per cent of garment output. In Vietnam, textile and garment SOEs undertook major economic reform in the 1990s, raising output per worker considerably (Nadvi and Thoburn 2004b). In China too, the state-owned textile and garment sector was at the forefront of economic reform in the 1990s (Eberhardt and Thoburn 2007). Household firms primarily catered for the domestic garment market in Vietnam in the past. They accounted for 2011 The Author(s) 359

Kenta Goto, Kaoru Natsuda and John Thoburn 15 per cent of garment output in 2007 (the latest figure available) and are predominantly micro enterprises, often under informal subcontracting arrangements for larger private firms (Goto 2005). Vietnamese SOEs have also entered the domestic retail market with brands of their own. The private economic sector accounted for 30 per cent of apparel output in Vietnam in 2007 (the latest available figure), up from 17 per cent in 2000. However, some segments of the private sector find it difficult to enter export markets since they lack necessary state support in connecting with buyers (Thomsen 2007). The importance of the garment industry in exports Figure 1 summarizes the main export items of Vietnam based on the HS2002 trade classification. Unlike neighbouring Cambodia, for example, where the overwhelming bulk of export earnings come from garments (Natsuda et al. 2010), Vietnam has a more diversified export portfolio. Garments were only 14 per cent of total exports in 2008, 6 but they are still the second largest export item, after crude oil, and the largest manufacturing export. Figure 1: Major export items of Vietnam in 2008 20% 8% 6% 6% 7% 14% 47% 6% Others Mineral fuels, oils, distillation products, etc. (HS2002; 27) Footwear (HS2002; 64) Fish & crustacean, mollusc & other aquatic invertebrate (HS2002; 03) Electrical machinery and equipment and parts (HS2002; 85) Apparel and clothing (Woven fabric, HS2002; 62) Apparel and clothing (Knitted fabric, HS2002; 61) } Note: Numbers do not add up to 100 per cent due to rounding, and similarly for total of HS 62 and 61. Source: UN Commodity Trade Statistics Database (UNcomtrade). 360 2011 The Author(s)

Meeting the challenge of China Figure 2 depicts the trends of garment exports between 1997 and 2008 in terms of value and share. The figure shows that exports have, apart from 1998, increased rapidly, while the share of garments in total exports has remained more or less the same. Figure 2: Garment exports from Vietnam 10000 9000 8724 20.0% 18.0% 8000 7000 7400 16.0% 14.0% Million US Dollars 6000 5000 4000 3000 Exports in million US$ Share of garment 2633 3465 4250 4681 5579 12.0% 10.0% 8.0% 6.0% Share of total exports 2000 1384 1302 1622 1821 1867 4.0% 1000 2.0% 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 0.0% Note: Export data are based on SITC Rev.3 84 (Clothing and accessories). Source: UN Commodity Trade Statistics Database (UNcomtrade). Changes in value chain orientation Vietnam s orientation towards Western markets dates from its bilateral trade agreement with the European Union in 1992, which gave it MFA export quotas into the EU market effective from 1993, as well as GSP (Generalized System of Preferences) access to the EU. This gives Vietnam a discount on the normal MFN tariff into the EU, although it is subject to rules of origin requirements, which Vietnamese exporters often cannot meet when they import their fabric. However, unlike its garment exporting rivals Cambodia and Bangladesh, Vietnam is not classed as a least developed country and so does not qualify for zero import duty access under the EU s Everything but Arms (EBA) programme, but then neither does China. While Vietnam also enjoyed normal trade access to Japan in the 1990s, the US market was effectively closed. Before 1994, there was a US embargo on imports from Vietnam, and subsequently a prohibitive set of tariffs on Vietnam as a nonmember of the World Trade Organization. Under the US Vietnam Bilateral Trade Agreement (USBTA) signed in 2000, Vietnam was at last given normal trade relations access to the USA, although Vietnam did not join the WTO until 2007 (Thoburn 2009). 2011 The Author(s) 361

Kenta Goto, Kaoru Natsuda and John Thoburn Figure 3: Major export markets for Vietnamese garments 100% 90% 20.6 17.7 15.3 12.2 8.2 8.8 8.4 9.2 6.4 4.9 2.7 4.3 80% 70% 60% 22.3 18.6 18.4 10.4 20.9 15.1 17.4 19.2 18.5 50% 40% 30% 20% 10% 0% 32.9 35.9 29.0 57.1 56.2 60.5 38.7 59.0 2.5 1.8 2.4 30.8 26.3 25.4 17.8 13.2 12.7 9.2 9.0 1997 1999 2001 2002 2003 2005 2007 2008 Japan US EU Taiwan and Korea Others Note: Others include all countries ranked outside the top 20 major importers. Source: UN Commodity Trade Statistics Database (UNcomtrade). Figure 3 tracks changes in the export destinations of Vietnamese garments. In 1997, the three major destinations Japan, Korea/Taiwan and the EU were each responsible for roughly a quarter of Vietnamese garment exports, with Japan the largest foreign market up until 2001. Following the implementation of the USBTA in 2001, exports expanded rapidly to the USA, which became Vietnam s largest market for garments in 2002. Vietnam s exports to the EU have been increasing as well, but Japan remains Vietnam s second largest country market for garments, with sales roughly double those to Germany and the UK, the largest customers in the EU-27. Between 2000 and 2004, Japan s clothing imports from Vietnam stagnated in terms of dollars, and increased only marginally in terms of yen, while sales to the USA boomed. Between 2004 and 2009, however, Japan s clothing imports from Vietnam increased 84 per cent in dollar terms and 40 per cent in yen terms. 7 Asian importers besides Japan, particularly Korea and Taiwan, remained major export destinations until 2002, but since then exports to those countries have dropped in terms of share, particularly for Korea. The falls to Korea and Taiwan represent the possibility that, prior to 2001, some exports to those countries were subsequently rerouted to the USA, which no longer was necessary after the USBTA. 8 Table 1 indicates Vietnam s position in major markets in relation to China as of 2008. China s dominant position is clear, but the table also shows Vietnam s strong position in the world, US and Japanese markets. In all cases, including the EU-27, Vietnam s market share has increased since 2004. 362 2011 The Author(s)

Table 1: China and Vietnam in the world garment economy, 2008 Meeting the challenge of China World Exports US Imports EU-27 Imports Japan Imports US$ billion 361.9 82.5 177.8 25.9 Market share of China (%) 33.2 34.6 22.7 84.2 Rank of China 1 1 2 1 Market share of Vietnam (%) 2.5 7.1 1.1 3.5 Rank of Vietnam 7 2 10 3 Sources and notes: Export data from WTO (2009), Eurostat for EU-27 market share and rank of China in EU, and Comtrade for ranks and shares in Japan and USA. Note that EU-27 countries are the largest suppliers of clothing to the EU market; otherwise, China would be the largest supplier with a market share of 42.6 per cent (and Vietnam s would be 2.1 per cent). Vietnam world rank assumes Hong Kong as 2, which includes Hong Kong s re-exports; otherwise Vietnam would be number 6. Trade distortions that have continued since the end of the MFA have greatly affected Vietnam s positioning within different value chains, operating not only as they apply to Vietnam directly, but also, and equally importantly, indirectly via their impacts on China. After the end of the MFA, Vietnam was still subject to MFA-style quotas into the US market as a non-member of the WTO. After Vietnam s WTO accession in January 2007, to check for dumping, the USA started a Vietnam Textile and Apparel Import Monitoring Program. The Bush administration took the view that the large Vietnamese state-owned textile and apparel sector was distorting trade and harming US domestic producers (Ellis 2007). In the event, the USA failed to identify sufficient dumping by the Vietnamese to invoke sanctions, and announced that the programme would not be continued beyond January 2009. 9 Nevertheless, the monitoring programme had important short run effects. When interviewed in August 2007, Vietnamese suppliers and international buyers saw the programme as a major risk because of the uncertainties related to possible future export restrictions to the USA. The Vietnamese government s reaction was to introduce a system of export licensing to restrict export growth to 30 per cent after WTO entry in 2007, which would avoid provoking the US government. To facilitate this, it placed much of export production in the hands of a small number of large SOEs/ex- SOEs over which control could easily be imposed (interviews February 2008). The US restrictions also led in 2007 to some Vietnamese suppliers adopting a strategy of reducing export shares to the USA and, in turn, increasing exports to other major markets, particularly the EU and Japan. Several key exporters emphasized this as their new strategy, and the most competitive and productive suppliers were even claiming to be starting to select buyers by rejecting some of the new incoming orders from the USA. 10 2011 The Author(s) 363

Kenta Goto, Kaoru Natsuda and John Thoburn This new strategy implied some shift in the balance of power between suppliers and buyers within a chain that is fundamentally buyer driven. The increased market power of Vietnamese producers depended largely on global buyers views of the risks of sourcing from China, its main rival. Following initial surges in 2005 to both EU and US markets, after the lifting of MFA quota restrictions from China, the EU restricted important categories of apparel imports from China until the end of 2007, and the USA until the end of 2008. Company B (refer forward to Table 3), for instance, noted a sharp fall in orders in early 2005, after China was freed of MFA restrictions, and then a rise in export orders when restrictions were reimposed on China in late 2005. Other companies noticed falls in prices for export orders during the 2005 China surge. In 2008, despite the beginning of declining US market demand resulting from the sub-prime crisis, there was a strong sense among companies interviewed that buyers were shifting purchases away from China towards Vietnam, and that Vietnamese suppliers could be more selective. Indeed, Vietnam was increasingly becoming the one in buyers China plus one sourcing policies. Buyers and not only US buyers sought to reduce the risks of over-dependence on Chinese supplies, while China was not only restricted in the USA but was facing rising domestic costs in its main producing areas in the south (HKTDC 2007; SCMP 2008). This was despite Vietnam having only a limited supply of textiles from its own SOEs (and from Korean and Taiwanese inward investors); in fact, Vietnam s physical proximity to China has helped Vietnam s garment industry base its exports on imported Chinese fabrics. We should note, however, that not all Vietnamese suppliers were able to be selective about orders. The more competitive suppliers, whose demand was typically larger than their supply capacity, mainly adopted the new strategy. Less competitive suppliers have become more dependent on the growing export businesses to US markets in conditions where they were unable to negotiate terms or reject orders. Differences of market positions according to export destinations In the world garment industry, buyers determine product specification through the demands and conditions of the markets they serve. Important for the purpose of this article, however, are the differences in the types of garments that Vietnamese suppliers produce for different export destinations. 11 Figure 4 attempts to classify Vietnamese products with respect to the level of value-added and production volume per order for the Japanese, US and EU markets respectively. Note that this classification is in relative terms based on perspectives of Vietnamese suppliers and some international buyers who cater for multiple export destinations. Later, we discuss this in terms of import unit values into Vietnam s three main markets. Most Vietnamese garments produced for Japan, Vietnam s largest single country market after the USA, are on the higher value-added end, with relatively complex design and product specification compared with the EU and USA. The size for Japanese orders is typically small, and with wide variations in size and colour. 364 2011 The Author(s)

Figure 4: Market segment classification for Vietnamese garments Meeting the challenge of China High Japanese Market Value Added EU Market US Market Low Small Quantity per Order Large Source: Based on interviews conducted in 2007. While cost pressures are increasing from Japanese market oriented value chains, buyers still regard quality as extremely important. Those who coordinate production and distribution for this market see Vietnamese suppliers key sources of competitiveness as their capacity to produce products with complex specifications subject to stringent quality requirements. This, according to them, is what differentiates Vietnam from other exporters such as India and China, which they perceive as better suited for market segments in the relatively cheaper volume zone categories. According to Japanese buyers, however, China s garment industry is much wider in scope and thus able to produce everything, including items that range from low value-added, price competitive products to high value-added men s and women s suits. They are therefore also able to supply similar items to those where Vietnam has competitiveness. However, Vietnam has comparative advantages in catering for smaller and more complex niche market orders, relative to China. By contrast, garments for the US market are mostly supplied to segments in the lower price range, where competition in prices is extremely fierce. Reducing costs becomes most important, and is done by producing garments with very simple design specifications allowing suppliers to minimize operational losses. Most orders for the US market are also large in terms of quantity, which allows Vietnamese suppliers to gain productivity from being on the same learning curve for a longer period, helping to achieve the economies of scale that Palpacuer et al. (2005: 412) call volume-based upgrading. The EU market is between the Japanese and US markets in terms of both the levels of value-added and order quantity. Such differences based on market destinations have important implications for the 2011 The Author(s) 365

Kenta Goto, Kaoru Natsuda and John Thoburn Vietnamese garment industry since they define its expectations from buyers and its key international competitors. To a certain extent, they also define the main areas in which competitiveness will be built up over time. Japanese trading companies or brand apparel firms normally coordinate value chains that cater for Japanese markets. To minimize the business risks of retailers rejecting products that fail to meet their precise specifications, Japanese buyers are strongly committed to raising the technical capacity of Vietnamese suppliers. The most common method of achieving this is to dispatch their technical staff to suppliers on a relatively long-term basis, at the buyers expense. Another is to train Vietnamese production line managers at Japanese garment manufacturing firms in Japan. Through such arrangements, they transfer relatively advanced knowledge and technology, particularly on configurations of production lines and management methods, to Vietnamese suppliers. As such, arrangements are costly to Japanese buyers and are de facto context specific investments; they also tend to prefer to establish a long-term relationship with selected suppliers. Value chains oriented towards Japanese markets are thus typical quasi-hierarchical in terms of their governance structure (Humphrey and Schmitz 2000). The transfer of knowledge and technology through such interfirm relationships has helped to increase the productivity of Vietnamese suppliers (Goto 2003). Garments for the US and EU markets, on the other hand, are coordinated in value chains typically governed by traders from Hong Kong, Taiwan and Korea, who function as agents for American and European retailers and apparel firms. Transfers of technology through production arrangements are much more limited compared with production arrangements by Japanese buyers, primarily because the level of quality required is lower. As price competitiveness is exceedingly important for the majority of exports to these markets, orders are placed competitively, which may result in less stable business relationships with value chain coordinators compared with Japanese market oriented value chains. The possibility of different development trajectories according to value chain orientation Differences in export destinations and value chain orientation can play out in terms of potential upgrading. For some types of upgrading, particularly of processes and products, technology transfers through the linkages with production and distribution networks that international buyers coordinate are important (Schmitz and Knorringa 2000). Functional upgrading, however, is about shifting towards more knowledge and skill intensive functions in the value chain, which is more difficult and requires more time (Giuliani et al. 2005; Goto 2007; Humphrey and Schmitz 2000). There is a danger for Vietnam that competitive pressures may compel firms to compete by cutting production costs. This could lead to a race to the bottom, in which firms lower costs at the expense of social conditions, particularly labour standards (Kaplinsky and Morris 2001: 31). In this context, firms connected to value chains that Japanese buyers coordinate have a better chance of process and product 366 2011 The Author(s)

Meeting the challenge of China upgrading through the transfer of new and advanced technology. Exports to the US market, by contrast, compete primarily on costs, so technological transfers from buyers to Vietnamese suppliers are much more limited. In exporting to the USA, Vietnamese firms may end up competing directly on costs with garment exporting countries such as Bangladesh and India, where wages are lower. When this happens, adding new technological content would become important. Product analysis Some 54 per cent of Vietnam s garment exports were woven fabric based garments (such as shirts, trousers and jackets) and the rest knitted garments (sweaters, polo shirts and dresses). However, this proportion differs considerably between Vietnam s main markets. Less than half of Vietnam s exports to the USA in 2008 were woven, but for Japan, Vietnam s largest single country market outside the USA, the figure was nearly 70 per cent. Similarly, for the EU-27, nearly three-quarters of Vietnam s garment exports were woven. 12 In the USA, the years from 2004 to 2009 saw a substantial rise in the shares of knits (HS 61) to the US market in relation to wovens (HS 62): the share of wovens in imports from Vietnam fell from 57 per cent in 2004 to 42 per cent in 2009. Between 2004 and 2009, total garment exports in current dollar terms rose 1.95 times, while knits increased 2.6 fold and wovens 1.45 fold. Vietnam rose to be the number two supplier of knits to the USA after China, and went up from sixth to fourth most important supplier for wovens. Vietnam s performance, while impressive, is less remarkable than that of China. Between 2004 and 2009, China increased its share of US imports of knits from 13 per cent to 34 per cent, compared with Vietnam s 3 per cent to 9 per cent; and, for wovens, China increased its share from 19 per cent to 42 per cent, compared with Vietnam s 4 per cent to 7 per cent. By contrast, the share of wovens in EU imports in 2009 is only slightly down from the 2004 (pre-mfa removal) figure of 77 per cent. Nonetheless, over the same period, China s share of wovens among its exports to the EU decreased from 63 per cent to 55 per cent (figures from Eurostat). In an Appendix (not published here but available on request to the corresponding author), we conduct analysis of market shares and export unit values for the top six HS (2002) 4-digit imports into each of Vietnam s three main markets. The top six products in each market account for nearly two-thirds of each market s imports of apparel from Vietnam (though only three products are in the top six in all three markets). Table 2 provides support for our earlier contention that the Japanese market is for higher value products than the USA, for the three 4-digit HS products, which are in the top six apparel imports of all three of Vietnam s major markets. In addition, for two of the three products, the EU-27 occupies an intermediate position in terms of unit values. Evidence of process upgrading is best sought in the US data, from which one can calculate unit values per item (as well as per kilo). 13 In all six cases, Vietnam s market share in the USA increased over the period 2004 9. In terms of unit values based on 2011 The Author(s) 367

Kenta Goto, Kaoru Natsuda and John Thoburn the numbers of items, Vietnam s UVs fell for five of its top six 4-digit products. 14 For all but one (HS 6203, men and boys suit, trousers, jackets), however, Vietnam has experienced unit values falling by much less than in the case of China, suggesting that Vietnamese products have been better able to maintain their value at a time when Chinese export appears to have been driving down apparel prices in most cases. Table 2: Import unit values for Vietnam in its major markets, 2009 ($ per 100 kilos) HS 2002 4-digit main imports from Vietnam USA Japan EU-27 6110 (jerseys and cardigans) 1863 2137 1457 6204 (W/G suits, skirts) 1611 3450 1708 6203 (M/B suits, trousers) 1472 2350 1706 Sources and notes: Calculated from Comtrade for USA and Japan, and Eurostat for EU. See Appendix for more details. HS categories are those in the top six apparel 4-digit imports from Vietnam in all three major markets. EU-27 unit values converted to US$ at average 2009 exchange rate of 1 = 1.39 (Exchange rate from http://france.usembassy.gov/irs-euro.html). W/G is women s and girls, M/B is men s and boys. Vietnamese garment suppliers did upgrading happen? In this section we presents a longitudinal case study over the period 2001 to 2007 of ten large domestic garment companies, average size more than 6600 workers, mostly in (or recently in) the state sector. We attempt to analyse whether these key garment suppliers were able to upgrade in the post MFA era, with special reference to the changing domestic economic environment. Our focus on domestic companies reflects the fact that foreign-invested firms, in contrast, have the opportunity to upgrade with help from their parent companies. This focus also reflects the fact that SOEs and former SOEs are likely to receive state help in establishing buyer contacts, help that is unavailable to parts of the private sector (Thomsen 2007). The situation of garment suppliers in 2007 in comparison with 2001 In Table 3, we provide a summary overview of the firms interviewed in 2001 and 2007. We interviewed ten firms in both years, of which five were located in Hanoi and five in Ho Chi Minh City. While most of the firms have expanded in terms of output since 2001, three suppliers (A, G and J) experienced almost no change, and for one (E) output actually decreased. In terms of export ratio, most of the firms were already primarily focusing on the export market in 2001, and this had not changed in 2007. It was very common to see suppliers like D and G that produce garments solely for the export market. F was the only one that reduced its export share (and thus increased production for the domestic garment market), but the shares had fallen slightly from 90 per cent in 2001 to 88 per cent in 2007. 368 2011 The Author(s)

Meeting the challenge of China Table 3: Outline of key suppliers (in comparison to the situation in 2001) Supplier Change in Outputs Share of Exports Share of CMT Number of Workers Average Wages Labour Shortage Expanding and/or Relocation A About the same About the same - -- + Major Problem Thai Nguyen, Bac Ninh B + About the same - ++ + Major Problem Hai Phong, Thai Binh, Quang Binh C + No Change - + n.a. Major Problem Vinh Phuc D n.a. No Change + - + Problematic Nam Dinh, Thai Binh E - + - - + F ++ - - ++ + Major Problem Not a Problem Nam Dinh, Ha Nam Ben Tre, Vinh Long, Tien Giang, Ninh Thuan, Binh Thuan G About the same + - + + Major Problem Long An, Dong Thap H + No Change - ++ + Problematic An Giang, Tien Giang, Da Lat, Gia Lai, Binh Duong I + About the same No Change + + Not a Problem Long An, Laos, Cambodia J About the same No Change n.a. No Change + Problematic Currently planning Notes: ++ means that the annual increase since 2001 was more than 10%, + is an annual increase less than 10%. Likewise, -- means an average annual decrease of 10% or more, and - is a decrease less than 10% a year. On labour shortages, answers are categorized into either Major Problem, Problematic, or Not a Problem. n.a. means that data was not available at the time of interview. Source: From interviews in 2001 and 2007. 2011 The Author(s) 369

Kenta Goto, Kaoru Natsuda and John Thoburn Most garments produced in value chains that international buyers coordinate are under the contractual modality CMT (cut-make-and-trim). Under CMT, Vietnamese suppliers only carry out the labour-intensive assembly functions, and these are of low knowledge intensity; buyers supply Vietnamese firms most of the input materials, such as fabrics and accessories, free of charge. Vietnamese suppliers produce garments based on buyers specifications and export all products under buyers arrangements in exchange for processing fees. Vietnamese producers carry no responsibility for knowledge intensive functions such as product design, distribution arrangement and marketing. Most of the business risks are concentrated in these knowledge intensive functions, which is where the level of value-added is highest (Goto 2003, 2007; Nadvi and Thoburn 2004b). The share of CMT arrangements had decreased for most suppliers apart from D and I. Instead, such firms have increased their shares in the FOB (free-on-board) type orders, in which Vietnamese suppliers procure fabrics and accessories on their own financial account. 15 However, such FOB can include a wide range of functions. Most types of FOB in Vietnam involve arrangements whereby Vietnamese suppliers only purchase input materials when the buyers instruct them to do so, which is, in essence, the equivalent of CMT. There were also no suppliers using Vietnamese inputs, except for some minor accessories including carton boxes, nametags and plastic bags. However, we know that a number of state textile companies have integrated forwards into garments using their own fabrics (Nadvi and Thoburn 2004b: 114). In terms of numbers of workers, which has direct implications for changes in operational scale and output, the outcome has been quite diverse with six suppliers increasing, three decreasing, and one being roughly the same. On wages, however, all suppliers reported an increase. In Table 4, we provide a summary of average wages and compare those for the garment industry with the average for all manufacturing based industries for 2000 to 2008. According to this, wages in all industries, including the garment sector, have increased, particularly after 2006 with growing inflation. The average wage level in the garment industry is lower than in the manufacturing sector in general. All suppliers in 2007, apart from F and I, faced labour shortages. Most of the suppliers we interviewed had their factories in Hanoi or Ho Chi Minh City. The factories of the few that did not were nevertheless located fairly near the city centres. Wage rates in the large city areas were increasing rapidly, thus making the Vietnamese garment suppliers wages relatively unattractive. The labour turnover rate was high and retaining or hiring workers was becoming increasingly difficult. Some of the workers who left those firms went to garment suppliers that offered higher wages, but many were moving to better-paid jobs outside the garment industry, often in the more capital-intensive manufacturing sector, or in the service sector. By 2008, these difficulties had intensified. This squeezed garment companies between inflationary pressure on wages and other costs on the one hand, and the customers unwillingness to raise prices, on the other. All the companies interviewed in 2001 and 2007 had plans either to build additional production capacity in rural areas, or to relocate all production facilities to 370 2011 The Author(s)

Meeting the challenge of China rural areas where labour was more abundant and cheaper. There is an important difference, however, between companies that established plants based on the first reason and the second. Suppliers that became more competitive and thus were able to increase outputs, such as I and F, generally had no serious problems in either securing workers or hiring new ones. However, the amount of land they had imposed restrictions on attempts to expand operations, so they had to move outside the city centre to build additional manufacturing capacities. The weaker ones, on the other hand, had not only problems in retaining and hiring workers, but they were also struggling just to keep their businesses running. The entry barriers had risen in terms of buyer requirements, and such companies were finding it hard to maintain a foothold. Table 4: Wage comparisons, 2000 2008 (Average wage VND1000 per month) 2000 2001 2002 2003 2004 2005 2006 2007 2008 Manufacturing average 0953 1001 1049 1161 1259 1390 1581 1832 2295 % annual increase -05.0-04.8-10.7-08.4-10.5-13.7-15.9-25.2 Textiles 0866 0846 0887 0964 1064 1198 1408 1614 2300 % annual increase 0-2.3-04.9-08.7-10.3-12.6-17.6-14.6-42.5 Apparel 0872 0846 0849 0995 1076 1143 1338 1522 1908 % annual increase 0-2.9-00.4-17.1 0-8.1-06.2-17.1-13.7-25.4 All activities State sector 1025 1126 1274 1590 1671 2146 2617 3202 4054 % annual increase -09.9-13.1-24.8-05.1-28.4-22.0-22.3-26.6 Non-state sector 0689 0750 0835 0956 1045 1215 1408 1802 2203 % annual increase -08.9-11.2-14.5-09.3-16.3-15.9-27.9-22.3 Foreign sector 1655 1558 1637 1606 1641 1810 2003 2242 2886 % annual increase 0-5.9-05.0 0-1.8-02.1-10.3-10.6-12.0-28.7 With 100% foreign capital 1282 1295 1217 1343 1412 1572 1786 2016 2548 % annual increase -01.0 0-6.0-10.4 0-5.1-11.4-13.6-12.9-26.4 Joint ventures 2533 2325 3090 2657 2742 3082 3287 3692 5294 % annual increase 0-8.2-32.9-14.0 0-3.2-12.4 0-6.6-12.4-43.4 Source: GSO Enterprise Survey 2009 (http://www.gso.gov.vn/). Note: Apparel wage data are for Manufacture of wearing apparel; dressing and dyeing of fur. Calculated for total employee compensation divided by employment. In addition, the opportunity cost of land was simply too high for such companies, which thus decided to relocate everything from the city centres to rural areas. On the land that would become available through relocation, most planned to establish hotels, 2011 The Author(s) 371

Kenta Goto, Kaoru Natsuda and John Thoburn office buildings or shopping centres, with a clear strategic vision of moving out from the garment industry. Vinatex has been diversifying into real estate too, as also have some of the more successful garment exporters we interviewed in 2008. 16 Value chain orientation and upgrading at the factory level Table 5 summarizes changes in export destinations for the firms interviewed in 2001 and 2007. Compared with 2001, most have increased exports to the USA quite drastically with an annual increase in shares of 10 per cent or more, except for F and H. By contrast, the export share to Japan has been decreasing, except for D and E. Changes in the export share to the EU market are more mixed. Companies F and H are interesting cases where changes in export destinations were pursued as part of their enterprise strategies. Both firms were producing around 50 per cent of garments for the US markets prior to 2007, but following the announcement of the US antidumping monitoring programme, they reduced their export shares to the USA rapidly and instead increased it for the EU and Japan. B adopted a similar strategy, and those three are among the most competitive garment suppliers in Vietnam with very stable increases in orders. All three firms became more selective in what type of orders to accept, and in some instances they claimed to be rejecting new orders particularly from the USA and shift production capacity for orders to the EU and Japan. Most of the buyers for the US and EU markets are trading companies from Hong Kong, Korea and Taiwan, while for the Japanese market they are predominantly Japanese trading companies. In Table 6, we attempt to examine how much upgrading has happened in terms of process, product and function for the suppliers interviewed since 2001. For process upgrading, because consistent and comparable information and data on value, particularly value-added, are unavailable, we compare output per worker in terms of quantity. 17 Roughly half the firms experienced upgrading in their processes, with some increasing productivity levels by as much as between 30 and 50 per cent since 2001. Some firms, however, such as A, C, E and G, were unable to realize significant productivity gains during the same period. Most of the suppliers who were successful in process upgrading either had received (or were still receiving at the time of the interview) technological assistance from Japanese buyers; some had even taken out a long-term lease on very advanced machinery from such buyers in their production line. This group of suppliers expanded their production capacity significantly and hired more workers, whereas the unsuccessful ones had much more difficulty retaining and hiring workers and, as a result, their production capacity shrank. These would suggest that the degree of the suppliers success in process upgrading has strong implications for their ability to cope with retaining and hiring workers, which should be particularly critical in the context of rising wages and labour shortages. Suppliers that were unsuccessful in process upgrading had high turnover rates, resulting in lower average years of experience of their workers. They found it especially difficult to retain skilled workers, which had detrimental effects on labour productivity. 372 2011 The Author(s)

Meeting the challenge of China Table 5: Export orientation and buyers profiles USA EU Japan** Change* Share (%), 2007 Change Share (%), 2007 Change Share (%), 2007 Buyer Profile A ++ 60 ++ 35 -- 5 Mostly Korean trading companies. B ++ 40 + 40 -- 20 For the USA and EU: Trading companies from Hong Kong, Taiwan and Korea. For Japan: Japanese trading companies. C ++ 55 No change 20 -- 7 For the USA and EU: Trading companies from Hong Kong, Taiwan and Korea. For Japan: Japanese trading companies. D ++ 50-20 Almost no change 30 (Japan, Canada, Taiwan) 90% of the buyers are trading companies from Korea and Hong Kong. However, there are also some direct relationships with American and European wholesalers, trading firms, and retailers. E ++ 40-20 Almost no change 40 For the USA and EU: Trading companies from Hong Kong, Taiwan and Korea. For Japan: Japanese trading companies. F + 30 Almost no change 20-30 For the USA and EU: Trading companies from Hong Kong and Korea. For Japan: Japanese trading companies. G ++ 60-20 -- 10 H + 20 + 50 -- 30 I ++ 70 - A few -- 23~26 J ++ 50-30 - A few Most are trading companies from Hong Kong, Korea, and Taiwan. Orders for one of the largest US retailers, however, goes directly through the retailer s branch office in HCMC. For the USA and EU: Trading companies from Hong Kong, Taiwan and Korea. For Japan: Japanese trading companies. For the USA: Trading companies from Hong Kong and Korea. Trading Companies from Korea and Hong Kong. Notes: *++ means that the annual increase since 2001 was more than 10 per cent, and + is an annual increase less than 10 per cent. Likewise, -- means an average annual decrease of 10 per cent or more, and - is a decrease less than 10 per cent a year. ** Export share for supplier D includes shares for Canada and Taiwan. Source: Interviews during field work in 2007. 2011 The Author(s) 373