UK retail concentration, Chilean wine producers. and value chains

Transcription

1 The Geographical Journal, Vol. 174, No. 2, June 2008, pp UK retail concentration, Chilean wine producers Blackwell Publishing Ltd and value chains ROBERT N GWYNNE School of Geography, Earth and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT r.n.gwynne@bham.ac.uk This paper was accepted for publication in March 2008 This paper seeks to examine how value chains impinge upon firms within the Chilean wine sector. The value chain analysis will further link the production and export of wine in Chile with the import and retailing of this wine in one key core economy market, namely that of the UK. The analysis is divided into three sections. First, the political economy of value chains in agro-industry is discussed, particularly in relation to the distinction between network or quasi-hierarchical relationships. Then the paper examines the theme of retail concentration in the UK and the impacts that this has on global value chains which incorporate Chilean wine firms. The focus then moves to value chains and the nature of upgrading within Chilean wine firms by examining the strategic example of the lead firm and firm upgrading as a response to demands of UK retailers and through the flying winemaker model. Broad conclusions on the comparative nature of value chains and the scramble for value within them are finally made. KEY WORDS: Chile, value chains, retail concentration, firms, wine, upgrading Introduction McMichael (1996) argued that the events of the 1980s and early 1990s in the Soviet Union, Eastern Europe, Latin America and East Asia were of such import that the development project, evident since post-second World War reconstruction, gave way to the globalisation project. Gwynne et al. (2003) saw this as coinciding with the abandonment of import-substituting or state-led industrialisation schemes in favour of more exportoriented strategies amongst what they termed the countries of the global semi-periphery. These notable policy shifts initiated changing relations between countries of the global semiperiphery and the core economies. Academics required new concepts to investigate the economic processes behind these changing relations. One significant conceptual contribution was that of the global commodity chain (GCC) and the first substantial treatment of commodity chains appeared in 1994 (Gereffi et al. 1994). As Gereffi et al. (2001, 1) noted, an important part of global trade is now conducted within transnational enterprises or through systems of governance that link firms together in a variety of sourcing and contracting arrangements. The key theme is that global trade cannot be understood merely as the result of arms-length and market-based transactions. Global trade is increasingly being organised through ever more complex inter-firm and intra-firm relationships and contracts (Gwynne 2008). Bair (2005) sees the GCC concept as primarily emanating from work on world systems theory. Others have argued that the origins of the concept of the global commodity chain lie in dependency theory (Gibbon 2001). Global value chain (GVC) is now another commonly used term used within the social sciences for those studies that investigate how firms link producers and producing spaces with consumers and consuming spaces at the scale of international trade in goods and services (Humphrey 2006). According to Bair (2005), there are significant differences between the GCC and GVC approaches, with the latter more closely related to international business studies, the meso-level of sectoral dynamics and the micro-level of firm upgrading. However, GVC analysis uses much of the spatial categories inherited or adapted from world systems /07/ /$00.20/0

2 98 Chilean wine producers and value chains and dependency theories. Gibbon (2001, 346) argues that chain coordination is still typically directed from northern countries, since it is usually associated with those links (or nodes ) in a chain which have particularly high barriers to entry, and because international income distribution remains extremely uneven. Gwynne et al. (2003, 18) attempted to revise the world systems framework by dividing the world into core, semi-periphery and periphery with the semi-periphery category being determined by whether key global economic actors perceived countries as constituting emerging markets. This article will pursue the Gwynne et al. approach by referring to countries as being part of the core, semi-periphery or periphery in terms of value chain analysis. This paper seeks to examine how value chains impinge upon firms within an agro-industrial sector of one Latin American country in the semi-periphery, namely Chile. The sector chosen for analysis in this paper is that of wine in Chile, notable for considerable export growth over the past two decades (Gwynne 2006). Value chain analysis will link the production and export of wine in Chile with the import and retailing of this wine in one key core economy market, namely that of the UK. The empirical focus in this analysis will concern the scramble for value along the chain. As Sturgeon (2001, 9) argued, studies of industry value chains reveal the concrete actors at the global scale as well as the linkages that bind them into a larger whole. Themes of governance, or the non-market coordination of economic activity between firms, can be explored in particular. The Chilean wine sector has developed an impressive export record over recent years. In the last year of the Pinochet dictatorship 1989, Chilean wine exports were a meagre US$35.4 million (Banco Central de Chile 1990); by 2006 they had reached virtually US$1 billion (Banco Central de Chile 2007). The Chilean wine sector has become completely restructured as a result. In the early 1990s, Chile was producing around 300 million litres of wine annually, of which only 20% was exported; by 2006 around 800 million litres were being produced with 75% for export; by 2014 the Chilean Ministry of Agriculture forecasts production of 1.2 billion litres with 85% destined for export (Richards 2006). One key research issue that value chain methodologies facilitate is the process of technological upgrading, both in terms of product and process (Humphrey and Schmitz 2002). Firms in countries of the semi-periphery need to respond to information flows going through the value chain and the requirements of supermarket and other retail buyers in core economy markets. Within the wine sector of a country in the semi-periphery, upgrading can occur through improved raw material supply (the quality of wine grapes), investment in process technology (stainless steel fermentation tanks and oak barrels, for example), the increasing utilisation of flying winemakers (Lagendijk 2004) and the more studied formulation of the final product in terms of targeted markets. This paper argues that the favourable export trajectories for Chilean wine to global markets in general and the UK market in particular are partly due to the nature of the insertion of wine-producing firms into global value chains. As a result, this paper is divided into three sections. First, the political economy of value chains in agro-industry will be discussed, particularly in relation to the distinction between network or quasi-hierarchical relationships. The paper then moves on to examine the theme of retail concentration in core country markets, such as that of the UK, and the impacts that this has on global value chains which incorporates Chilean wine firms. The paper then analyses how value chains give context to the nature of upgrading within Chilean wine firms. Much of the data come from a two-year British Academy research project (2005 7) which examined the impacts of globalisation on export-oriented wine firms in Chile s Colchagua Valley and the record of collaboration between these firms and key purchasing companies within the UK market. Twenty-one wine firms were interviewed in Chile and five buyers of Chilean wine from leading UK supermarkets and specialist retail chains. This permitted a more detailed perspective of the nature of the relationships that Chilean export-oriented wineries established with the major world market for its wines. Value chains and agro-industry Value chain analysis has inherited some of the concepts and methodologies from commodity chain analysis. For example, Gereffi (1994, 97) introduced the dimension of territoriality and described it as the spatial dispersion or concentration of production and distribution networks, comprised of enterprises of different sizes and types. In value chain analysis this dimension could focus on understanding the links and functional integration between producing districts on the one hand, and consuming spaces on the other (Cook et al. 2006). These can be geographically highly dispersed locations as in the case of Chilean wine-producing districts and wine-consuming spaces in the UK. Another dimension that value chain analysis examines is the governance structure the authority and power relationships that determine how financial,

3 Chilean wine producers and value chains 99 material, and human resources are allocated and flow within a chain. This could be seen as examining the nature of power relations that exist along a chain and the search for how key actors within the chain generate and attempt to appropriate value. Three forms of governance have been identified in this context: inter-firm networks; quasi-hierarchical relationships between powerful lead firms and independent but subordinate firms in the chain; and vertical integration. Humphrey and Schmitz (2002, 1023) developed the crucial distinction between value chains as networks and quasi-hierarchical relationships. In networks, firms cooperate in a more informationintensive relationship, frequently dividing essential value chain competences between them. Knowledge transfer up and down the chain is encouraged for the potential benefit of all actors. The buyer may specify certain product and process standards to be attained, but should be confident that the supplier can meet them. Humphrey and Schmitz (2002, 1023) commented that chains characterized by even networks offer ideal upgrading conditions but are the least likely for developing country producers because of the high level of (complementary) competencies required. Meanwhile, in the quasi-hierarchy, the lead firm exercises a high degree of control over other firms in the value chain. This can involve specifying the characteristics of the product to be produced, the processes to be followed and the control mechanisms to be enforced. There can be some doubt about the competence of the supply chain and hence the strong involvement of the lead firm. Knowledge transfer is very much orchestrated through the chain from the lead firm. Are, then, the value chains in the Chilean wine sector more similar to the even network or the quasi-hierarchy? In Gereffi s (1999) work on value chains in the global clothing sector, the main pattern observed was distinctly quasi-hierarchical. Furthermore, quasi-hierarchical relationships have been found in value chains in the horticultural sectors, such as in the relationships between large supermarkets in the UK and fresh vegetable producers in Africa (Barrett et al. 1999; Dolan and Humphrey 2000). What then is the relationship between governance and upgrading in agro-industry in general, and the wine industry in particular? There are at least two issues that agro-industrial firms from countries of the semi-periphery should consider as they develop transactional relationships with powerful global buyers from the core economies. First is the issue of market access. Even when core countries dismantle trade barriers to the import of agricultural and agro-industrial goods, producers in the semi-periphery and periphery do not automatically gain market access (Humphrey and Schmitz 2001). This is because the value chains that producers feed into are often governed by a limited number of core country buyers intent on generating value. Through a number of annual contracts, this normally signifies the core country buyers requiring products with increasing quality at lower, or at least, similar prices. Nevertheless, in order to participate in export production for the key global markets of North America and the European Union, producers need access to the lead firms and buyers of these chains. Secondly, becoming part of a value chain could offer firms a fast track to the acquisition of production capabilities. Those producers that gain access to the chains lead firms tend to find themselves on a steep learning curve. The lead firms are very demanding with regard to reducing cost and raising quality, but they also transmit best practices and advice. Thus local producers can learn a great deal from global buyers about how to improve their production processes and attain consistent and high quality. Research into agro-industrial value chains must be informed and framed by global value chain studies both in manufacturing on the one hand and agriculture on the other (Le Heron 1993; Marsden and Arce 1995). However, the key area of analysis must be the relationship forged between key buyers (and/or owners) in the core country markets and the producers and suppliers in the countries of the global semi-periphery. For this reason, analysis should begin by focusing on the nature of the key buyers in core country markets. Retail concentration in the UK and impacts on the Chilean wine value chain Humphrey (2006, 574) argues that concentration in the retailing of fresh and processed food has led to a substantial reorganization of agribusiness value chains. Large buyers have transformed themselves from resellers of products made by others into firms that go out to find suppliers for the products that they want for their customers. This would also appear to be very much the case in the wine value chain. This section will examine the links between Chilean wine firms and the large supermarket chains in the UK market; these are playing an increasing role in product development, branding and supplier selection. Figure 1 demonstrates the broad framework of the downstream value chain of Chilean wine in terms of the UK market. Chilean wine firms can be divided into three: wine company groups; newly established companies (both those aiming for highquality production from the beginning and those with the strategy of upgrading the quality of wine

4 100 Chilean wine producers and value chains Figure 1 Chilean wine: value chain downstream to UK market through time); and old established wine firms trying to restructure in order to supply global markets. In terms of gaining access to the UK market, they are mostly represented by UK distributors. Only Chile s largest wine company, Concha y Toro, has set up its own distribution company in the UK. Although this decision was seen as risky initially, it soon proved commercially viable; supply-chain management within the Concha y Toro group and its UK network has improved and the 6 10% for the distributor now stays with Concha y Toro (Davis 2006). The role of the distributors (whether Concha y Toro UK or a long list of UK distributors) is then to coordinate with the UK ontrade and offtrade. The ontrade consists of pubs, restaurants and hotels and data from the International Wine and Spirit Record (IWSR) indicate that about 21% of Chilean wine by volume is destined for this sector leaving 79% for the offtrade (IWSR 2007). Here the supermarket chains dominate with 72% of offtrade wine sales. Within these supermarket chains, four supermarkets (Tesco, Sainsburys, Asda and Morrisons) are responsible for 70% of total supermarket wine sales, with Tesco alone responsible for 33%. Retail concentration has also occurred within specialist wine retailing groups only three major groups remain at the national scale (Majestic, Oddbins and the Thresher group) and they record about 9% of offtrade sales. Hence, small-scale wine-retailing companies throughout the UK (local independent wine merchants, internet and mail order wine retailers) now only account for around 19% of wine sales. Thus, increasing concentration in the UK wine retailing sector has produced a small number of wine buyers with a significant combined market power; the top 10 supermarket and specialist wine retail chains account for around 80% of the UK offtrade wine market. The wine buyers of these 10 key retailing companies are clear targets for export-oriented wine firms in Chile and their purchasing (and tasting) decisions have significant impacts on those firms. Over the past two decades, supermarkets have regarded sales of wine as critical for attracting higher-spending customers and have developed competitive strategies based on increasing the range and quality of wine. The larger supermarket groups outperform many of the smaller supermarket groups in terms of wine sales (compared with total sales). Over the past decade, Tesco has considerably widened its range of Chilean wines. The Chilean red wine portfolio gives an idea of the increasing range and product differentiation that this has involved. By the end of 2006, Tesco sold six Chilean Cabernet Sauvignons (ranging in price from 2.99 to 9.99), four Merlots, three Pinot Noirs, two Carmenères, one Shiraz, five red blends and one 3 litre Cabernet Sauvignon box (Davis 2006). Another element of UK supermarket strategy has been to develop brand image and increase the significance of own-label products. This has meant that supermarkets have taken an active role in product innovation and supply chain management. Humphrey (2005, 3) shows that own-label penetration of retailing in the UK rose from around 22% in 1980 to around 43% in Pursuing the Tesco red wine product range, a stage further is illustrative.

5 Chilean wine producers and value chains 101 Of the 22 Chilean red wine products sold in Tesco at the end of 2006, seven were own label: The role of own-label brands for Chilean wine is fundamental. In terms of possible consumer confusion about price and quality there is always more consumer confidence in buying a Tesco branded wine. As a result Tesco branded wine accounts for nearly 50% of Chilean wine sales in Tesco branches Davis (2006) Tesco s own-label quality Chilean brand (Tesco Finest) has greater than half of the total Chilean wine sales in the higher price range, indicating that Tesco consumers believe in the supermarket brand image more as the quality increases. Supermarket chains have thus succeeded in establishing credibility in their own brands so that consumers do not just perceive the own-label brand as a cheap alternative, but as a worthy competitor to the manufactured brand (Chaney 2004, 5). This growth in own-label branding has led to customised, complex relationships between supermarket buyers and wine firm suppliers. However, through time, suppliers can acquire competences and diversify their customer base. Retail concentration in the UK market has provided a key target for the upgrading strategy of many Chilean wine firms. Humphrey and Schmitz (2002, 1022) argue that task complexity increases as products become more customized and cannot be obtained readily from alternative suppliers. When the risk of supplier failure to cope with this complexity is perceived to be low, Humphrey and Schmitz (2002) see the coordination of the value chain as more resembling a network than a quasihierarchy. The upgrading strategies of wine producers are often linked to a highly competitive business model. One medium-sized Chilean winery, Luis Felipe Edwards (LFE), demonstrates the advantages and disadvantages of having the UK market at the end of its own value chain. It has achieved rapid export growth by focusing on the UK as its key strategic market. In 2005, 45% of its exports went to the UK and its percentage to the four leading supermarkets (70%) was exactly the same as the industry average. They developed a networking relationship with all four of the leading supermarkets in the UK, and had supply contracts with Marks and Spencer as well. The company saw the UK market as the best barometer of international markets, being five years ahead of the rest of world markets in terms of taste, price and requirements (Edwards 2005). Furthermore, LFE saw the UK as a very open market, giving relatively easy access to new firms (LFE started in 1975) that are able to compete in tough conditions. However, the power of supermarkets in the UK wine market has led to very competitive pricing conditions and supermarkets tend to swallow the profits in the Chilean wine value chain (Edwards 2005). In 2005, LFE sold cases of premium red wine to Tesco of Cabernet Sauvignon and Carmenère. The list price was 5.99 but Tesco insisted on two or three promotions a year when the price went down to 3.99; 75% of LFE cases were sold during these promotions. As with the networking model, LFE had managed to diversify its customer base so that it could walk away from contracts where the prices were too demanding; for example, it had decided not to pursue an own-label production contract with Tesco in 2005 (Edwards 2005). Overall, the UK supermarket is perceived, even by Chilean wine firms that have targeted the UK market, as key to their global strategy as appropriating excessive value in the wine value chain. Chilean wine producers realise that they have to cope with tight value margins in selling to UK supermarkets and the associated high export volumes that can develop. Thus, in the case of LFE the UK market provided 45% of export volume, but only generated 30% of export value. Other global markets (and their retailers) provided slower market growth but higher retention of value 55% of export volume but 70% of export value in the case of LFE. Retail concentration in the UK market has thus provided Chilean wine producers with the opportunity of rapid growth of sales but often combined with low per unit profitability. This links into the interesting relationship between geographical sourcing of wine and supermarket shelf space, particularly in such specialist retailers as Majestic which organise their wine sales by country (and region in the case of France). The link between different wines and shelf space is reviewed every six months. Chilean wine has been on a steady upward path for the last four years (Pym 2006). Hence the geographical differentiation of shelf space in any supermarket chain gives a snapshot of how that particular firm views its geography of wine supply at any particular time. In other words, the relationship between national market share and shelf space in leading UK supermarkets and specialist retail chains is very close. Between 2003 and 2007, the proportion of UK supermarket wine shelf space selling Chilean wine crept up by an average of 0.25% a year from 6.3 to 7.3% (Wines of Chile 2007). In the case of Majestic, the two main elements of the rapid growth in Chilean wine shelf space were for Pinot Noir and Sauvignon Blanc products (Pym 2006). This changing geography of UK supermarket shelf space is perhaps the key indicator of export success for the wine industries of exporting countries.

6 102 Chilean wine producers and value chains There are different shelf space indicators for different supermarkets. In 2006, Majestic over-indexed on shelf space for Chilean wine compared with the national average 8% as opposed to 7% (Pym 2006). In contrast, by 2006, Threshers was underindexing quite severely 4% as opposed to 7% of shelf space for Chilean wine (McEvoy 2007). This had been caused by a deliberate strategy to remove shelf space from Chile and give this space to other countries such as New Zealand (McEvoy 2007). However, such a wide difference between Threshers shelf space allocation and that of the national average for Chilean wine prompted a change of strategy in Chilean wines are being allocated more shelf space as we believe that Chile offers the customer a good value for money alternative to most other New World countries (McEvoy 2007). Associated campaigns to brief staff and communicate about Chilean wine to Threshers customers have been put in place. This introduces the concept of the geography of shelf space allocation. If the allocation of shelf space for a country s wine by a leading retail firm falls too far below the national average, the firm is soon prompted to develop strategies to reverse this trend. Market share takes time to expand (though decline can be quicker) due to inertia and the large number of lobbying influences attempting to maintain existing allocations of shelf space. The value chain approach is also applicable to many countries of the semi-periphery as there has been a very rapid process of retail concentration in these countries as well (Reardon and Berdegué 2002; Reardon and Hopkins 2006). In many Latin American countries there has been a rapid shift to centralised procurement in such countries as Brazil, Mexico and Colombia, leading to stronger bargaining power with suppliers and reductions of per unit fixed costs of transaction. This may reduce the per unit value received by the Chilean wine producer in these markets, but can offer the potential for significant growth in sales. These retailing transformations have facilitated Chilean wine producers in sending an increasingly differentiated range of products to a more diversified set of global markets. In 2004, the Concha y Toro group sent wine exports to 110 countries, more than three times the number of a decade earlier (Concha y Toro 2005). Perhaps the most dynamic Chilean exporter in this sense is Montes. This was a firm that only started in 1988, but its key growth strategy has been to produce quality wines (second highest export price of all firms see Table 1) and to export them to as many markets as possible; by 2005 it was exporting to 68 countries and had developed a good supply record to both the ontrade (restaurants and hotels) as well as the offtrade in each market. Retail concentration in core countries, and an increasing number of countries in the global semiperiphery, has meant that Chilean wine firms have a wide range of options in terms of developing their export profile. The value chains in which they insert themselves should be seen more as a network than a quasi-hierarchical relationship. Each firm has the potential to supply a range of supermarkets in an increasingly large number of countries in which the process of retail concentration has taken place. The Chilean wine producer has a certain amount of bargaining power if a supermarket chain attempts to impose too demanding a contract, the wine firm has a number of other opportunities at the global scale. Nevertheless, in any particular contract the large potential sales offered by the large supermarket chain will mean that expectations over the retention of per unit value within the value chain will have to be kept low. The combination of needing to improve quality and reduce price does mean that Chilean wine firms should have clear upgrading strategies for their export products. Value chains, upgrading and the Chilean wine firm How have UK supermarkets created contractual and sourcing arrangements with Chilean wine firms? Supermarkets in core country markets have become one of the key agencies in the global value chain of any New World or wine-exporting country. One could refer to Gereffi s (1999) requirements thesis of key actors in this context. Although supermarkets and other end clients do not have direct access to the technology of their suppliers, they draw up requirements in this case for the wine-exporting companies. Requirements can be either long term (development of a brand) or short term (development of a new or upgraded product to see how the development of a new taste, grape variety or blend coincides with consumers). Unlike the Chilean fruit sector, leading UK supermarkets and specialist retail chains have mainly had to deal with Chilean, as opposed to transnational, firms. Of Chile s 20 major firms in 2005, only two were foreign owned the relatively small Casa Lapostolle and Los Vascos firms (Table 1). Referring back to Gereffi s (1999) analysis of upgrading in his value chain work, two themes can apply to such agro-industrial sectors as that of wine product and process upgrading. Product upgrading is normally defined as moving into more sophisticated product lines, with more value-added. In the wine sector, this could signify the firm starting with the production of basic wine and then gradually improving the product and moving up through the various quality categories varietal, premium, super-premium and even ultra-premium.

7 Chilean wine producers and value chains 103 Table 1 Access to the UK market of Chile s 20 major wine companies by value of exports, 2005 Wine firm Year firm started Export value US$ million 2005 Proportion of export value to UK market (%) Export volume (million litres) 2005 Price per litre (US$) 2005 Concha y Toro San Pedro Santa Rita Cono Sur Errazuriz Undurraga Montes * Via Santa Carolina Santa Helena (Vinex) Tarapacá Carmen Maipú Viñedos Emiliana La Rosa Ventisquero Los Vascos Valdivieso Casa Lapostolle Montgras Total/average *Estimate Source: Wines of Chile database: Duijker (1999) This would involve a slow but sustained strategy of upgrading by the wine firm. It means developing new higher-quality products alongside the production of the more basic categories with which the firm started supplying international markets. Within export-oriented wine enterprises, this can also involve the role of flying winemakers (Lagendijk 2004) who can bring knowledge of which products are best suited for which markets. Process upgrading constitutes a second related theme. This involves transforming inputs into outputs more efficiently, by reorganising the production system or introducing superior technology. The introduction of stainless-steel fermentation tanks with strict temperature control and an effective policy towards the supply of oak barrels are two key examples in the wine sector. Flying winemakers are significant here as well, as they introduce the knowledge of how to best organise the process in order to achieve the marketable product required. One example would be the slow and low-temperature fermentation of white grapes (such as Sauvignon Blanc) in order to maximise the fruit-driven component of the final product. Furthermore, in agro-industry there is often the need to improve the quality of the raw material input. This has led to a distinct process of vertical integration as large wine firms purchase large amounts of land (to plant new vines) and sometimes existing vineyards in order to better control the quality of the raw material input. The wine sector contradicts the lessons of most agro-industrial firms as the wine firm emphasis has to be on reducing wine grape yields in the vineyard in order to maximise the subsequent quality of the wine. It can be easier to control yields in vineyards owned and managed by the winery, rather than purchasing wine grapes from contract farmers or through the spot market at harvest time. As a result, there has been an interesting shift in vineyard location in Chile. When the Chilean wine sector was oriented towards the domestic market, vineyard locations on the flat plains of the longitudinal Central Valley and transverse valleys were favoured. This produced high yields of most grapes, but reduced the quality and complexity of the final product. Since the 1990s, major Chilean wine firms have been planting vineyards in new zones, such

8 104 Chilean wine producers and value chains as slopes and hillsides along the transverse valleys, along the Coastal Range and on the Andean piedmont. Here, yields are much reduced but, if these vineyards are carefully managed in terms of access to water, they can produce the raw material for much improved wines. Chilean wine firms have been distinguished by their record of success in both product and process upgrading. This record of upgrading has first of all been set by the strategic directions taken by the lead firm, Concha y Toro. Secondly, there has been a group of firms that have upgraded as a response to key retailers in core country markets. Thirdly, there has been what could be called the flying winemaker model to upgrading. The strategic example of the lead firm The Concha y Toro group consists of a number of companies linked by ownership (and common technological and marketing knowledge). The parent company, controlled by the Guilisasti family, has developed a number of new subsidiary firms with distinctive roles in the overall group. The Concha y Toro company, which was formed back in 1883, had by 2005 become the dominant wine-exporting enterprise in Chile, with nearly three times the exports of its nearest rival (Table 1). In addition, it has set up or acquired three other subsidiaries since 1986, all within the top 20 firms in terms of exports Cono Sur, Emiliana and Maipo. Santa Emiliana (later renamed Viñedos Emiliana) was set up in 1986 and has been given the strategic focus of developing and exporting organic wines. Cono Sur was established in 1993 and was given the strategic mission of being a specialist export-oriented firm, particularly targeting European markets. Vina Maipo, meanwhile, specialises in the cheaper, varietal market segments hence its relatively low average price per exported litre of US$1.89 (Table 1). Together these four inter-connected firms are responsible for more than 26% of Chilean wine exports. In this way, the Concha y Toro group has developed the characteristics of the lead firm in Chile s wine export sector. For example, the group has significant power and influence upstream in the wine value chain as its constituent wineries purchase wine grapes off farmers throughout much of Central Chile. The prices that Concha y Toro offers for wine grapes on the open market during the harvest in March and April tend to be taken as the key reference point for all other firms purchasing wine grapes from grape farmers on the spot market at harvest time. The role of Concha y Toro as lead firm in the Chilean wine sector can also be seen in its strategy of technological upgrading. Since the 1980s, the Concha y Toro group has invested strategically in capital goods and new technologies in order to upgrade vineyards and wineries. The aim was to set up a process of continual upgrading in order to improve wine quality throughout the price range. At least three strategic areas have been significant: raw material supply; investment in process technology; more studied formulation of the final product in terms of the requirements of targeted markets. The example of its Cono Sur subsidiary is instructive. Cono Sur was set up in 1993 to target the European and particularly the UK markets. In the UK, it started its export strategy by supplying wine for own-label brands to supermarkets. This allowed Cono Sur to gain knowledge of consumer taste and supermarket requirements in the UK market, despite surviving on low margins. After having acquired competences in supplying own-label wines to supermarkets, its strategy shifted to developing its own brand image. It was able to develop two brands Cono Sur and Isla Negra the former focused at the premium quality category and the latter at the varietal. By 2005, Cono Sur had become the fourth largest exporting company of Chilean wine and had come to rely more on the UK market (receiving 55% of firm exports) than any other company see Table 1. Within the Cono Sur brand, it also developed a number of higher quality levels (Ocio and 20 Barrels signifying super premium categories). The formulation of these quality levels and products was designed to meet the requirements of leading supermarkets and specialist retail chains in the UK in particular. By 2005, Cono Sur sold to all leading UK supermarkets and specialist retail chains though the combination of products varied between them (Downes 2005). Product upgrading relied on important improvements in its Chimbarongo winery. The technological upgrading of the winery required significant investment as with the installation of large numbers of temperature-controlled stainless-steel tanks for fermentation with smaller tanks required for the fermentation of those wine grapes expected to produce higher quality wines. In addition, the annual purchase of large numbers of new oak barrels was important in raising quality, particularly for the premium and super-premium red wines. Cono Sur has not relied on flying winemakers most of its winemaking expertise has been developed within Cono Sur and from collaboration with the winemakers of the wider Concha y Toro group (Padilla 2005). Upgrading has also been evident in the supply of its raw material from the vineyards. This has had at least three elements. First is the issue of increasing vertical integration into upstream supply and control of vineyard management. When Cono Sur started in 1993, it relied mainly on buying in wine grapes

9 Chilean wine producers and value chains 105 from grape farmers (both on contract and through the spot market). Since then, Cono Sur has been acquiring more vineyards so that it can have more control on the quality of raw material supply. By 2005, Cono Sur owned 300 ha of vineyards which meant that it was sourcing about one-third of its wine grapes from its own vineyards. By 2009, it is due to own and control 850 ha of vineyards, which should mean that it is fast approaching the position of closely controlling the production of most of its wine grape supply (Padilla 2005). Second, Cono Sur sources its wine grapes from an impressive range of Chile s growing spaces. In contrast to wine producers in other New and Old World countries, there is a 1200 km distance between its most northerly vineyards in the Elqui valley from its most southerly in the Bio-Bio valley. It has used the huge north south range of Chile to locate the growing of the 11 grape varieties that it uses in the most appropriate valleys in terms of environment, terroir and climate; Riesling and Gewurtztraminer wine grapes, for example, are mostly sourced from the cool Bio-Bio valley. The third element is the nature of technological upgrading. Cono Sur has developed an advanced technological approach for its vineyard management very different to that of Old World vineyards. This involves use of drip irrigation systems and fertilisers to maximise grape quality, significant pruning to restrict yields per vine and foliage management to improve the uniformity of grape bunch ripening. The use of drip irrigation systems is informed by regular measuring of soil humidity throughout the vineyards (which are continually mapped via satellite for vegetative stress) so that each vine neither receives too much moisture on the one hand nor suffers stress through lack of moisture on the other (Padilla 2005). Concha y Toro has thus been a good example of a lead firm in terms of its export-oriented strategy in general and its record of technological upgrading in particular. The fact that it has four distinct companies within the overall wine group means that there are distinct differences in setup and winemaking styles between them. It maximizes the company s potential to be all things to all men (O Halleron 2007, 143). Some have argued that, as a result, it is one of the few global wine brands that has enjoyed both critical and commercial success (O Halleron 2007). It could also be argued that its strategy of upgrading has been the template for many of the other successful exporting wine companies from Chile. Firm upgrading as a response to demands of retailers One group of the top 20 Chilean wine companies has very much followed the Concha y Toro model of basing their export strategy on sustained upgrading and supplying key supermarkets in core country economies. It is interesting to point out that these firms are normally newly established (less than 20 years old). They first inserted themselves into international markets by producing highly competitive basic and varietal wine (good quality, low prices) and have subsequently tried steadily to upgrade their quality and increase their prices. One example of this strategy is seen in the performance of Montgras. Between 2000 and 2004, nearly 50% of its production was linked to a massive export of varietal wine to Sainsburys as a Christmas special offer. Whereas this gave the firm economies of scale in terms of production, per unit profit or value was very low indeed. In 2005, Montgras decided to stop this deal; while export volumes understandably declined by 37.6%, export values were down only 18.9% as the average price per litre rose in a startling way over a 1-year period from US$2.65 in 2004 to US$3.46 in Again this shows the trade-off in terms of prioritising the supply of UK supermarkets high-volume growth alongside supermarkets receiving the greater share of the value in the chain. Other firms that have followed this upgrading strategy include Ventisquiero and Via. In 2005, Via and Ventisquero had average export prices of only US$1.42 and 1.97 per litre respectively, which implies that they are at the beginning of this upgrading process; Via and Ventisquero are the major suppliers of own-label wines to the UK s largest supermarket, Tesco. However, between 2004 and 2005 they recorded the fastest growing exports of all the top 20 Chilean wine companies. All these firms have an upgrading strategy closely linked to the product requirements of key retailers in core country markets, particularly that of the UK; Via and Ventisquero had 39 and 45% of their exports respectively go to the UK market in 2005 (Table 1). Firms upgrading through the flying winemaker model Another upgrading strategy has been followed by two foreign companies (Casa Lapostolle and Los Vascos) and one Chilean company, Montes. Casa Lapostolle, advised by the flying winemaker, Michel Rolland, is owned by the French Marnier-Lapostolle group. These companies have emphasised the pursuit of quality wine production for export since the firms were established. Upgrading is important but from a higher base than the previous group of firms. Table 1 shows that these three firms recorded the three highest average export prices for their wine in 2005 between US$4.53 and 6.89 per litre. They did not start wine exports by supplying own-label wine for supermarkets and, interestingly, have not

10 106 Chilean wine producers and value chains linked their export strategy to prioritising the UK market. Indeed their target market has been the United States (receiving between 40 and 60% of firm exports) where retail concentration has been much less than in the UK in terms of wine. This is mainly due to the institutional framework in which each US state has different laws governing alcohol consumption. In effect, this has meant that there is a lack of any nationwide specialist wine retailers or US supermarket chains selling similar wine brands in all states. The US thus provides a much more fragmented market for Chilean wine exporters but one with higher per unit value than the UK market. These firms have emphasised the importance of the flying winemaker in their operations, whether it be the foremost Chilean flying winemaker, Aurelio Montes, or Michel Rolland, the Pomerol-based star of the film about globalisation and wine, Mondovino. Lagendijk (2004, 523) sees flying winemakers as quality and marketing symbols themselves, able to impart success to their wine product through their name being on the label. There is, however, a further issue here the links between flying winemakers and wine critics a component of what Lagendijk refers to as the cultural circuit of the wine sector. In the critical US market, the most influential wine critic is Robert Parker; he has a long history of giving high (and very high) evaluations to wines in which Michel Rolland has been involved. Hence, the logic of the Casa Lapostolle upgrading strategy; it has relied on Michel Rolland to introduce modern technologies and Pomerol winemaking knowledge to its vineyards and winery, and has then focused on selling the quality wine to the US market, where such wines as Clos Apalta have achieved iconic status. In 2005, 61% of Casa Lapostolle s total exports were sold within the US market. This review of Chilean wine firms in terms of global value chains in general and the UK market in particular shows the vital importance of emphasising quality and a long-term commitment to upgrading. The most successful export company, Concha y Toro, has transformed itself in two decades from a firm dedicated to supplying the limited domestic market to a wine group with four component firms, each with clearly defined and different strategies in the global market. Meanwhile, the new companies with a firm commitment to upgrading have recorded rapid export growth Montes, Montgras, Ventisquero and Via are all less than two decades old, but have transformed themselves into the fastest growing exporters and have developed networks with supermarket chains in most leading core country markets. Concluding remarks Thus, the global value chains in which Chilean wine firms have inserted themselves should be seen more as a network than a quasi-hierarchical relationship. As Humphrey and Schmitz (2002) pointed out, network relationships offer firms clearer opportunities for upgrading than in more hierarchical arrangements. This does mean that the Chilean wine sector may be rather unusual in that it is one of the few examples of significant upgrading and bargaining power being enjoyed by developing country firms within an agricultural or agro-industrial global value chain. Indeed Watts and Goodman (1997, 14) argued that in global commodity chains in agriculture and agro-industry, capital mobility has resulted in the centralisation of power by retailers and suggested that developing country firms (and farms) within these commodity chains had even less bargaining power than developing country firms in the (clothing) manufacturing chains studied by Gereffi (1999). In contrast, many Chilean wine firms have developed the competences to supply a range of supermarkets in an increasingly large number of countries. Rapid export growth has particularly occurred with those countries where the process of retail concentration has been significant. The Chilean wine producer has been able to establish a certain amount of bargaining power within the value chain, much more, for example, than large-scale farmers within Chile s fruit value chain (Gwynne 2003; Murray 1997). Nevertheless, export-oriented wine firms have become highly dependent on the relationships they forge with large supermarket chains in core country markets. These supermarkets have been successful in the scramble for value, at least within the value chain linking Chilean producers with UK supermarkets. However, this value chain has provided a vehicle for rapid growth in export volume for many Chilean wine producers. Furthermore, key retailing actors strongly influence consumer attitudes in their respective markets. This role is closely related to what happens in the wine sector at large, in terms of trends and developments in markets and production. This is reflected in the concept of the changing geographies of shelf space within core country supermarkets. If the allocation of shelf space for a country s wine, such as that of Chile, differs too much from the national average, the supermarket chain is soon prompted to reverse this trend. Value chain analysis is very much framed by the nature of the particular sector as it develops at the global scale (Friedland 2005). Unlike global value chains in agriculture, wine provides an example of an agro-industrial value chain in which significant

11 Chilean wine producers and value chains 107 value-added can occur in the processing plant (winery). In contrast to such agricultural products as wheat or grapes, for which global exchanges tend to set prices, the wine sector has much greater flexibility in price setting and can potentially generate a huge variety of products. Prices have to be negotiated between producers and retailers rather than set in an exchange. The role of technological upgrading (through both product and process) becomes important, as this can act to differentiate the final products of an exporting firm. Unlike other manufacturing sectors, the exportoriented wine firm does not normally receive technology direct from a branded merchandiser or retailer. Rather, it provides an example of the requirements thesis. As supermarkets and other end clients do not have direct access to the technology of their suppliers, buyers of Chilean wine within these lead firms draw up product requirements for the wine-exporting companies. These require wine-exporting firms to seek out the relevant technologies from other sources. In certain wine firms, flying winemakers have become increasingly significant. Technological change has not only come from the classic Old World countries, such as France, that have always strongly influenced the style of Chilean wine production, but also from New World countries such as Australia and New Zealand (Richards 2006). Thus the geographical sources of information flows for research into new process and product technologies (such as France and Australia) are different from the main geographical sources of knowledge relating to sales and marketing (such as the UK). This is an example of Lagendijk s (2004) concept of interconnected locales. Places of winemaking in one country (which are spread along a 1200 km axis in Chile) are connected to places of exchange and consumption in another. At the same time, these places are connected to knowledge centres in other countries, such as the Pomerol region in France (where flying winemakers such as Michel Rolland are based) or the Adelaide region in Australia (home to Roseworthy, one of the world s leading wine universities, and another group of flying winemakers). The systemworld of the wine value chain thus not only links producing spaces with consuming spaces but also links these with the spaces that generate technology and the applied knowledge of wine. The systemworld also involves the cultural circuit of the wine sector wine associations, trade journals, websites, wine critics and journalists, scholars and experts. Some Chilean wine firms have been able to forge export growth by harnessing the potential of the cultural circuits as well as the value chains of the global wine sector. Countries of the semi-periphery such as Chile have long wanted to develop an export-oriented agro-industry that was based on adding value to resources, was sustainable economically in the long term and had a trajectory of technological upgrading. Over the past two decades, the wine sector in Chile has successfully developed this combination of value-added, sustainability and technological upgrading and this has brought dramatic success in export growth. Dependency theory needs to adapt to the findings of value chain research. The fact that there are many Chilean wine companies that are now inserted into global value chains has substantially changed the nature and pace of technological change of the export-oriented firms. There may still be asymmetries in the economic relations between core country economies and those of the semiperiphery as dependency theory highlighted (Kay and Gwynne 2000, 51), but the insertion of firms into global value chains can also bring new export opportunities and impressive records of technological upgrading. Acknowledgements I would like to thank the British Academy for funding this research, Michael Cox for his generous help in providing data and contacts in the Chilean wine sector, and my fellow contributors for their comments on an earlier draft. References Bair J 2005 Global capitalism and commodity chains: looking back, going forward Competition and Change Banco Central de Chile 1990 Indicadores de Comercio Exterior, Diciembre 1989 Santiago Banco Central de Chile 2007 Indicadores de Comercio Exterior, Cuatro Trimestre de 2006 Santiago Barrett H R, Ilbery B W, Browne A W and Binns A 1999 Globalization and the changing networks of food supply: the importation of fresh horticultural produce from Kenya into the UK Transactions of the Institute of British Geographers Chaney I M 2004 Own-label in the UK grocery market International Journal of Wine Marketing Concha y Toro 2005 Memoria anual Santiago Cook I et al Geographies of food: following Progress in Human Geography Davis J 2006 Interview with Chile wine buying manager for Tesco 28 November Dolan C and Humphrey J 2000 Governance and trade in fresh vegetables: the impact of UK supermarkets on the African horticulture industry Journal of Development Studies Downes S 2005 Interview with Commercial Director, Cono Sur 12 August, Santiago