Bringing social institutions into global value chain analysis: the case of salmon farming in Chile

Introduction

Globalization is a multidimensional and contested phenomenon, which has been conceptualized in different ways in diverse literatures. One important feature of globalization has been the disaggregation and geographical dispersion of value creating activities and the rise of new organizational forms. These innovations in the organization and internationalization strategies of firms have led to new approaches to the conceptualization of multinational firms which, instead of focusing on the ownership of internationally based assets, stress the power to coordinate and control international operations even when no ownership rights exist (Dicken, 2007). While much of the literature on new organizational forms has focused on the processes of vertical disintegration of activities of lead firms via the outsourcing of value creating activities (Flecker and Meil, 2010; Hitt et al., 1999; Marchington et al., 2005; Taylor and Bain, 2005) a global value chain (GVC) perspective also shows that opposite processes of vertical integration have often characterized the supplier base (Gereffi, 1999; Sturgeon, 2002).

The concept of global value chains has been developed to explore the functional integration of internationally dispersed value-adding activities in a variety of inter-organizational networks (Coe et al., 2008; Gereffi et al., 2001, 2005; Henderson et al., 2002). Thompson argues that these changes in organizational form mean that an analytical focus solely on the company or the workplace – the traditional focus of the sociology of work – may neglect the role of markets, on the one hand, and ‘the extended hierarchies that constitute forms of co-ordination within and between firms on the other’ (Thompson, 2003: 366). He argues that ‘[n]ew strategies address the overall productivity of the whole value creation chain, largely by reorganizing the relations between focal and dependent companies, and by innovation in supply chains, logistics and quality management’ (2003: 367).

A central concern of the GVC approach has been the extent to which insertion in the global economy offers developing countries and firms opportunities for learning, skills upgrading and the enhancement of technological capabilities. These international networks have the potential to develop into vehicles for knowledge diffusion, learning and upgrading as large, powerful buyers and producers provide information, technical and managerial assistance, as well as training to their network suppliers and partners. Demands on suppliers by buyers, such as major supermarket chains, and the lead producers which dominate the value chain have resulted in some suppliers upgrading their operations to more complex and higher-value-added activities. However, as Lloyd and James (2008) argue in their study of the impact of supermarkets on occupational health and safety in food manufacturing, there are often contradictions between price and delivery demands and espoused commitments to the improvement of working conditions.

Despite this potential for upgrading, the GVC literature only partially explains processes of innovation and labour upgrading (Ramirez and Rainbird, 2010). While the importance of national institutions for the development of GVCs has been acknowledged (Ernst and Kim, 2002; Gereffi, 1999; Henderson et al., 2002; Humphrey and Schmitz, 2002), few authors incorporate national social institutions into their analysis, Lane and Probert (2004, 2009) being the principal exceptions. Moreover, as Barrientos et al. (2010: 3) note, the benefits to firms do not necessarily result in benefits to workers, and they identify the need to make linkages between economic and social upgrading.

The ability of national institutions to support the upgrading efforts of firms and their workforces is a critical factor in establishing the benefits from participation in GVCs. Debates on skill formation tend to focus on the role of national institutions and organized interest groups in supporting the quantity and quality of skills in the economy, on the one hand, and their development and deployment in workplaces, on the other (Whitley, 2000). While national training systems are seen as responding to and evolving in the context of the international economic environment (Bosch and Charest, 2010), globalization tends to be treated as an external force to which national policy makers and companies respond. The institutional literature on skill formation rarely focuses explicitly on innovations in organizational form (for an exception, see Grugulis et al., 2003); or on the intersection between national institutions and the activities of multinational companies (Heyes and Rainbird, 2010); or refers to developments in the international location of production facilities. Herod et al. (2007: 251) argue that bringing a geographical perspective on the changing location of production with the varieties of capitalism literature raises the question ‘how […] do the particularities of place impact on subcontracting and decentralization?’ Local institutions for innovation and skill formation may affect local economic actors’ participation in global value chains, their ability to upgrade (Gereffi et al., 2001: 2) and their relationships to other organizations. In other words, social institutions, particularly those relative to skill and knowledge formation contribute to comparative advantage and specialization within the overall context of insertion in the global economy.

If one is to consider the challenges for workforce skills in the global economy, one needs to look not just at those faced by advanced economies, but also those faced by developing economies. In this article, the authors focus on Chile, a middle income, natural resource-based economy, which represents an interesting case study for a number of reasons. First, it is an example of a middle income developing country, which Blomström and Kokko (2002: 15) argue is a category which is most likely to benefit from the spill-overs from foreign direct investment provided host country conditions meet certain thresholds in relation to economic stability and human and physical infrastructure. Second, it has been argued that Chile’s rapid growth, based on its open economy, low inflation and safety for foreign investments, make it a ‘Latin American tiger’ comparable to Taiwan, Japan, South Korea and Singapore (Yaňez, undated: 111) and a model for Latin American countries. This would appear to be confirmed by its acceptance as the first South American member of the OECD in January 2010 (www.oecd.org/chile). Third, the ‘paradigmatic importance’ of the Chilean economic model in the development debate has been identified by Reinecke (2000: 4) and Peru, Mexico, Bolivia and Argentina have followed its policies of privatization and integration into the global economy. However, although Chile’s economic performance has been perceived as superior to that of the rest of Latin America this has not been based on a strategy to develop human capital. On the contrary, under the Pinochet military regime (1973–90) production was concentrated on primary products with low value-added, facilitated by authoritarian management styles, low wages and a high level of labour market flexibility (Sehnbruch, 2006).

In order to integrate the analysis of social institutions into global value chain analysis the authors have chosen to analyse salmon farming in Chile. From craft-based origins, the industry evolved over a period of 20 years to become by 2008 the second largest global producer of farmed Atlantic salmon, after Norway (Katz, 2004; Maggi Campos, 2006). In 2005, Norway was responsible for 39 per cent of farmed salmon globally, Chile for 38 per cent, followed by the UK and Canada with 8 and 7 per cent respectively (Boston Consulting Group, 2007: 75). The significance of this for employment is that in 2005 this export sector generated approximately 28,000 direct and 8000 indirect jobs, concentrated primarily in the X Lakes Region in the south of Chile (Boston Consulting Group, 2007: 97).

This article is divided into five sections. In the first section the authors analyse the relationship between companies’ participation in GVCs and the potential for upgrading, highlighting the fact that the upgrading of firms’ capabilities does not necessarily bring gains for labour. In the second section the research methodology is discussed. In the third section, the authors outline the role of social institutions in the historical development of salmon farming in Chile, identifying how this affected local producers’ insertion into particular parts of the GVC and their ability to capture value. In the fourth section the authors discuss the role of institutions in shaping labour conditions and skills supply and the extent to which the involvement of foreign owned firms constituted a source of dissemination of ‘good practice’. In the Conclusion the authors assess the extent to which Chilean institutions have supported the development of innovation and skills in the industry and identify the need for comparative study of the role of social institutions in shaping countries’ and firms’ participation in GVCs.

Participation in GVCs and the potential for upgrading

Empirical studies of a number of GVCs have shown that these international networks have the potential to develop into important vehicles for international knowledge diffusion, learning and upgrading (Bair and Gereffi, 2003; Ernst and Kim, 2002; Gerefffi, 1999; Gereffi et al., 2005; Humphrey and Schmitz, 2002; Sturgeon, 2002) as large buyers and producers provide information, technical and managerial assistance, as well as training, to their network suppliers and partners. At the same time, by making increasingly more stringent demands with respect to quality, consistency and speed of response on their supply base, they have spurred the development of selected suppliers towards more complex and higher-value-added activities of the value chain.

These global production and trade networks have various forms of governance and a key concern of the literature is how changes in supplier capabilities may change power relations within GVCs (Gereffi et al., 2005; Humphrey and Schmitz, 2002). Despite the upgrading potential of GVCs, the organizational, technological and managerial capabilities of suppliers to learn, internalize and adapt new knowledge is central to the ability of these firms and their workforces to benefit from international linkages (Coe et al., 2008; Ernst and Kim, 2002; Humphrey and Schmitz, 2002). As Ernst and Kim (2002) point out, participation in GVCs is no substitute for domestic upgrading efforts and this requires the existence of an effective system of national institutions that support knowledge creation and diffusion, skills formation and the development of organizational, technological and managerial capabilities. The character of national business and innovation systems and the ability of these national institutions to engage with and support the upgrading efforts of firms and their workforces therefore becomes a critical factor in the capacity of countries, regions and firms to benefit from participation in GVCs.

Where supplier upgrading takes place, employee remunerations, work conditions and employment show a mixed picture in terms of improvements (see Bair and Gereffi, 2003; Knorringa and Pegler, 2006; Nadvi, 2004; Palpacuer, 2000; Palpacuer and Parisotto, 2003; Reinecke, 2006). When local firms participate in GVCs, many of the advances made by labour are precarious as a result of intensification in international competitive pressures. Similarly, the impact of GVCs on the informal economy can be both positive and negative, with differential outcomes on workers and producers with different employment statuses, in different sub-sectors and regions. These impacts are strongly gendered, whereby unpaid domestic work and childcare responsibilities in the ‘reproductive economy’ differentiate the options of women and men to participate in market activity (Barrientos et al., 2003).

If social institutions are to be incorporated into GVC analysis, this needs to be understood as a two-way relationship. Social institutions influence the supply and regulation of labour, impacting on local firms’ capabilities, skills and labour practices. In turn, national and multinational firms inserted in GVCs have the potential to serve as conduits for the diffusion of new technologies and new forms of work organization (Ramirez and Rainbird, 2010). This has the potential to create contradictory pressures on governments: to support infrastructural development (research capability, training and skills development) which has the potential to support skills upgrading, on the one hand, and to encourage labour market flexibility to reduce labour costs, on the other. Pressures for the upgrading of labour and management practices may come through lead firms’ (either producers’ or buyers’) interest in ensuring the existence of suppliers with the capabilities of meeting international management and product quality standards; from interest in corporate social responsibility; or through labour clauses embodied in trade agreements, for example, the labour chapter of the 2004 Chile/USA Free Trade Agreement (Díaz Andrade et al., 2007). Nevertheless, Robinson argues that retailers’ corporate social responsibility policies and commitment to voluntary labour initiatives do not necessarily result in improved labour conditions for producers (2010). They require the involvement of other actors in the value chain, such as trade unions, non-governmental organizations and consumer groups in the monitoring of labour codes (Robinson, 2010). Sustainable improvements in the conditions of labour may still depend in an essential way on the role of national and local institutions (Palpacuer and Parisotto, 2003), alongside coordination between union representatives in different parts of the value chain.

Methodology

This research draws primarily on a literature review, which brings together publications in English and Spanish on the salmon farming industry and the Chilean context. This includes a number of reports prepared for the Ministry of Labour and non-governmental organizations on labour conditions in the industry (Díaz Andrade, 2003, 2007, 2009, 2010) and on the role of foreign owned companies in Chilean natural resource-based industries (Díaz Andrade et al., 2007). In addition, the authors have drawn on secondary data and semi-structured interviews with representatives from the main national institutions associated with the development of the Chilean salmon farming GVC. These are the Corporación para el Fomento (CORFO), the main government body for promoting industrial development; Fundación Chile, a public-private institution formed in 1974 that concentrates on technology development and transfer and has played a significant role in adapting imported technology to Chilean conditions; and SalmonChile, the salmon-producers’ industry association and lobbying group. The interviews explored their approaches to the challenges facing the industry in relation to technology, innovation and skills, including the identification of areas with important skills gaps. Primary data were collected through semi-structured interviews with senior technical directors of the five largest salmon farming firms operating in Chile: Marine Harvest (Norwegian), Ewos (a subsidiary of the Norwegian state-owned company Cermaq ASA), and the three largest Chilean companies: AquaChile, Multi-Export and Camanchaca. In the month of December 2008 10 interviews were undertaken in Santiago and in Puerto Montt, in the south of Chile where salmon farming is concentrated. The interviews focused on the development of firm specific knowledge and capabilities as well as the development of workers’ skills, differences and similarities between Chilean-owned and foreign owned GVCs in terms of knowledge base, skills and capabilities, the nature of links between firms and science and technology institutions in Chile and abroad, the role of international technological standards and the demand of buyers in the development of firm capabilities.

The salmon farming value chain and its implications for knowledge and skills

The salmon farming GVC includes a number of production stages alongside a range of support services. The initial stage is breeding where fish are selected for characteristics such as growth rate, skin colour and disease resistance. This is followed by a process of ova production involving the artificial fertilization of eggs. The hatching and juvenile phase of the value chain involve various stages of freshwater development with the aim being to maximize the yield of quality fry before the fish are moved to larger freshwater tanks or open net cages in lakes until they develop the characteristics that will enable them to survive in seawater. It is estimated that the early stages of the value chain represent approximately 15 per cent of direct jobs in the Chilean industry (calculations based on Maggi Campos, 2006: 119). The next stage, growing and fattening, is the seawater phase where the fish are kept in net pens for fattening till they have reached the market weight. This stage of the value chain accounts for approximately 33 per cent of direct jobs (calculations based on Maggi Campos, 2006: 119).

Knowledge of genetics and biology are important for all stages of the value chain up to processing but are, above all, critical in the breeding and ova production stages. At the same time, skilled personnel capable of managing sophisticated quality control and environmental protection routines are required since factors such as water quality, temperature, salinity and a number of ecological variables related to micro-organisms vary according to each particular lake (Katz, 2004: 19). Furthermore, according to the representative of SalmonChile, expertise in marine sciences is important as marine ecosystems as well as currents and waves have important implications for fish health (the spread of disease, for example) and growth rates. The development and use of process technologies such as cultivation tanks, automatic feeding mechanisms, computerized censoring and control systems are essential during cultivation stages (the hatching/juvenile and growing/fattening stages).

The harvesting and processing stages involve stunning, bleeding and gutting fish, leading to various processing activities ranging from primary processing which involves cutting the head and packing; secondary processing which prepares products (e.g. fillets, steaks and portions) for retail and food services; and the value-added products which involve smoked or ready-to-eat and ready-to-heat products. The treatment for processing fish varies significantly depending on whether the final market destination is the EU, the USA or Japan. The processing stages of the value chain are the most labour intensive, accounting for 50 per cent of all direct jobs, but represent the less knowledge and skill intensive parts of the production process.

A number of services support these activities. These relate to fish health; the maintenance of nets, pens and cages; water circulation and oxygen machinery and equipment; laboratory testing; waste disposal; logistic services involving packing, crating and various forms of transport; research and development; consultancies; and technical training. Many of these products and services are provided by specialist suppliers with a high demand for professionals and skilled workers such as biologists, chemists and laboratory technicians, software engineers and technicians, as well as specialists in engineering and environment sciences. They often locate close to the main salmon farm producers resulting in the development of regional clusters (Maggi Campos, 2006).

In order to understand the ability of producers in Chile to capture value, it is necessary to identify where they are located in relation to other producers and buyers in the GVC. There are two particularly powerful nodes where actors have a high degree of power: the production of salmon feed and retailing. Salmon feed accounts for nearly 50 per cent of the direct production costs of farmed salmon (Vidal, 2002: 106) and has become one of the most research and development intensive nodes of the GVC. Three firms, Skretting (a subsidiary of Dutch firm Nutreco), Ewos (a subsidiary of Norwegian state-owned company Cermaq ASA) and BioMar (Danish) are responsible for 92 per cent of world production (Marine Harvest, 2008: 14). These firms keep research and development centralized in their country of origin, though testing facilities are distributed in the producing countries.

Power in this GVC also rests with large retailers and food distributors such as Wal-Mart, Tesco, Costco and Sysco who purchase between 60 and 90 per cent of farmed salmon (Tveterås and Kvaløy, 2004: 8) and play a key role in setting industry product and process standards. These powerful buyers often buy directly from producers establishing contractual relations that differ from pure arms-length market transactions. As a result, they are in a position to buy large quantities of farmed salmon, often buying the whole production run of a specific producer, and exert constant pressure on prices. In this respect the salmon farming GVC is a good example of Gereffi’s (1994) concept of a ‘buyer-driven’ value chain.

In Chile the early stages of development of the salmon farming industry were characterized by high levels of vertical disaggregation with different firms specializing in distinct parts of the value chain (Katz, 2004; Tveterås and Kvaløy, 2004). The 1990s saw a process of vertical integration whereby large firms integrated the three main stages of the value chain: hatching, growing and harvesting. Some Chilean firms also started egg production activities. Industry consolidation took place during the 1990s and early 2000s and was marked by the exit of many small firms from the industry and the entrance of Norwegian producers via mergers and acquisitions, a process that was facilitated by their links with Chilean producers through sales of salmon eggs and feed. By 2001, 36 per cent of salmon farms located in Chile were in the hands of foreign firms (Fazio, 2001). A number of factors have been given to explain this process of vertical integration, among them the highly perishable and fragile nature of the product which demands high degrees of coordination in order to preserve product quality and extend the shelf life (Tveterås and Kvaløy, 2004). Added to this are retailers’ and buyers’ increasing demands for traceability and quality standards in a broad range of areas including worker training, hygiene, health and safety, environmental management and animal welfare, which have led to a need for information and control of the production and distribution chains (Iizuka, 2006; Tveterås and Kvaløy, 2004). It has also been suggested that the reduction in the international price of salmon and profit margins following the entrance of Chilean producers into world markets in the 1990s induced processes of vertical integration as large producers attempted to gain stricter control of world production (Katz, 2004; Muños Goma, 2009).

Institutions for innovation, labour and skills

National institutions played a significant role in the emergence of the salmon farming industry. Salmon is not an indigenous species to Chile and so major scientific and technological inputs were required. The emergence and initial development of the industry was a state-led initiative with public national institutions playing a catalytic role in processes of learning, technological adaptation and capability development (Katz, 2004). The learning and upgrading efforts of private firms were strongly subsidized and supported by public institutions, notably CORFO and Fundación Chile, which also played a key role in establishing the regulatory framework. This took place during the period of the Pinochet regime where the discourse against state intervention and in favour of the ‘free market’ and ‘private enterprise’ was at its height.

As the industry expanded between 1986 and 1995, complexity increased and there was some technological upgrading with the introduction of automatic feeding systems and computerized censoring and control systems. Many suppliers of technologically sophisticated inputs, such as the genetic material (eggs), automated machinery and industrial robots, were foreign owned, primarily Norwegian. In cases where inputs developed for the Norwegian industry failed to work in the Chilean environment (e.g. vaccines) domestically owned firms supported by state institutions such as CORFO often emerged to fill the gap.

Though Chile has developed relatively strong natural resource-based industries, knowledge, skills and technology do not represent the main source of competitive advantage (OECD, 2007). This is reflected in the very low research and development intensity of the Chilean economy where ‘[t]he modest role played by the business sector in the financing and performance of R&D is the feature that most visibly and measurably distinguishes Chile’s innovation system from those of more advanced economies’ (OECD, 2007: 11). The OECD report found that only a small proportion of firms placed the development of new products and processes at the centre of their competitive strategy and that successful, export-oriented firms showed innovativeness mainly in non-research and development-based product differentiation strategies, business models and marketing. Therefore for the majority of Chilean firms, the purchase of capital goods is the main means of technological adoption. This view was confirmed in interviews with CORFO, Fundación Chile, SalmonChile and five leading firms (including the two subsidiaries of Norwegian-owned firms) which showed that in salmon farming, research and development activities mainly focus on the adaptation of imported technologies and know-how.

Despite a policy discourse on innovation, technology and the development of human resources, it has been argued that Chile relies on ‘a moderate stock of human capital whose accumulation has been slow, whose distribution is unequal, whose renovation is scarce and whose quality and effort are inadequate to confront the requirements of globalization’ (Brunner, 2003, cited by Departamento de Competitividad Regional, 2004: 5 – authors’ translation). The main mechanism for skill development in employment has been through the provisions of the Statute of Training and Employment of 1976, (Estatuto de Capacitación y Empleo) which set up the National Service for Training and Employment (Servicio Nacional de Capacitación y Empleo - SENCE), a decentralized body responsible to the Ministry of Labour, which administers a training rebate scheme and monitors standards of private training providers. Its objectives are to develop labour force skills in order to contribute to an adequate level of employment, improve the productivity of workers and firms, as well as the quality of processes and products. SENCE manages a franquicia tributaria, a tax rebate system whereby companies can reclaim up to 1 per cent of the wage bill for training and resources are provided for vulnerable workers (the unemployed, young unskilled adults and the extremely poor). At the same time, the military regime privatized INACAP, the state training provider. INACAP was a national training body set up in 1966 to support import substitution policies, linked to the national development agency, CORFO (Sehnbruch, 2006). Provisions were made to encourage two types of group training arrangements, funded by membership dues which are tax deductible. These are Technical Intermediary Training Organisms (Organismos Técnicos Intermediarios de Capacitación – OTICs) which are non-profit organizations set up to manage joint training funds and the Technical Training Organisms (Organismos Técnicos de Capacitación – OTCs), training providers.

The training tax rebate system has encouraged an increase in training activities, especially since 1990 as the Concertación governments emphasized the importance of investment in human resources (Sehnbruch, 2006: 190). This increase has followed market forces rather than strategic priorities set by national or sectoral training bodies. Large companies have taken advantage of this system and training has been concentrated on managerial and professional workers. In 1997, executives and professionals received 16.6 per cent and supervisors and administrators 46.7 per cent of SENCE funded training (Reinecke, 2000: 325). In other words, it has not acted to redistribute training within companies or to develop the skills of production workers, but has been directed towards employees who already had higher levels of qualification and salaries. Reinecke argues that this demand is indicative of a managerially driven rather than participatory management style in companies (2000: 330–31).

In response to criticisms of the training rebate, new measures were introduced in 1997 which aimed to make training subject to more participative decision making; to create incentives for companies to train less skilled and lower paid workers; and to make it easier for small companies to claim the rebates. A tripartite Consejo Nacional de Capacitación, involving government, employer and union representatives, was set up to foster national social dialogue on training, alongside similar provisions at regional level. Companies with more than 15 employees were obliged to set up bipartite training committees (Sehnbruch, 2006). Where training programmes are jointly agreed by these committees, greater rebates can be claimed. Despite these incentives, there is scepticism concerning the enforceability of these voluntary provisions, given employers’ preference for unilateral decision making on training (2006: 181). Sehnbruch maintains that in Chile ‘technical and vocational training is essentially regulated and run by free market forces’ (2006: 207).

Turning to the salmon industry in particular it is worth first exploring labour force characteristics before discussing the capacity of the sector for skill development. According to SalmonChile, employment in the sector grew fourfold between 1993 and 2002: from 8500 direct employees in 1993 to 31,500 in 2002; and from 2500 indirect employees to 13,500 in 2002 (Pinto and Kremerman, 2005: 8). The features of labour conditions are: low wages; the physical nature of the work; a tendency to replace permanent workers with temporary workers; firings due to new technology; seasonal fluctuations in labour demand; cold, wet, unhygienic working conditions; and lack of bargaining rights (Barrett and Caniggia, 2002: 1956–7). Díaz Andrade et al.’s (2007: 60–61) study of 15 processing plants found that women and men were distributed evenly between the plants, though often with a numerical preponderance of one sex or the other. Some processing plants use temporary and subcontract labour, often to cover peak harvesting periods, as reported by interviewees at Camanchaca and AquaChile.

In Chile collective bargaining is only legal at the company level and this affects union organization and bargaining power. Many workers on subcontracts are employed by empresas de papel (paper companies) which have been set up to weaken trade union organization (Alvarez Vallejos, 2009). Case studies of 10 of the largest salmon companies, involving a total of 14 plants, were conducted by the Labour Directorate (Dirección del Trabajo) in 2007. They found levels of unionization at plant level which ranged from 7 per cent to 73 per cent, with higher levels of unionization among women (Díaz Andrade et al., 2007: 87). Despite this, collective rights were weak: 10 plants had been denounced for anti-union practices and between 2004 and 2005 there had been nine formal complaints against five of the plants for infringing rights relating to the right to organize and collective bargaining (2007: 85–6). The report also noted weaknesses in the management of health and safety, high levels of accidents and a lack of management concern about anti-union practices (2007: 93–4).

Turning to training and skill development, closer examination of skill usage suggests a polarized pattern which combines the underutilization of formal qualifications at higher levels in the jobs hierarchy and a lack of investment in operative skills at the lower end. According to SalmonChile, since the 1990s regional universities and private professional training institutes have provided courses for specialized professionals in areas such as biochemistry or ichthyopathology and for technicians and administrators specializing in fish farming. The result has been a broadening of the supply of technical and professional specializations (Maggi Campos, 2006). The SalmonChile representative reported that although workers with higher level skills can be recruited from the education system, they are often working in technician roles and this represents ineffective use of qualifications. This partly reflects the fact that young people and their families pay for higher education and so the costs do not fall on firms.

There are shortages of qualified non-professional labour, but this requires firms to recruit students from technical school and to provide training and education once they are employed (Maggi Campos, 2006: 133). The absence of intermediate or technician skills was identified as a problem by representatives of SalmonChile and CORFO. The interviewee from CORFO commented, ‘It’s very common to have a mechanical engineer and a bloke with a screwdriver and nothing in between.’ This is perceived partly as a problem of labour supply: parental aspirations for their children’s education results in a preference for higher status university studies, which have higher returns than professional and technical education. This polarization also reflects patterns of labour demand and the relative absence of intermediate occupations and progression routes into more qualified work. In response to this situation, CORFO is developing a programme to meet technical gaps, drawing on private sector technological institutes which provide training through SENCE.

At the lower end of the job hierarchy, there is little demand for qualifications in recruitment and little access to training once workers are employed. According to estimates given by Fundacion Chinquihue (a training and industrial services facility created with Japanese cooperation) 76 per cent of the workforce employed in the processing plants is either completely unqualified or possesses a low level of qualification (cited in Maggi Campos, 2006: 119). In their study of aquaculture on the island of Chiloé, Barrett and Caniggia found that no previous knowledge, education or experience was required of newly recruited process workers (2002: 1959). According to interviewees at SalmonChile and Marine Harvest, high levels of turnover, exacerbated by poor conditions of employment, have contributed to employers’ interest in developing a system of competence assessment, so that training received in one company, for example in environmental, health and safety and labour regulations, is recognized by others. According to the Marine Harvest representative, there are also gaps in managerial competence: managers often lack specialist understanding of fish biology and physiology.

To what extent do the Norwegian companies, which dominate the industry locally and internationally, serve as a mechanism for transferring ‘home country’ practices, based on their experience of the ‘Norwegian model’ with its emphasis on a highly educated and unionized workforce, and participative management styles? Díaz Andrade’s (2003) study of Norwegian multinationals shows that investments in Chile were driven primarily by the need to achieve higher levels of growth, market penetration and economies of scale which could not be achieved in their country of origin, due to the operation of production quotas (2003: 29–30). While the Norwegian trade representative, Sjur Holsen, argued in 2002 that Norwegian companies would be in the vanguard of improving Chilean environmental and social standards, in practice investors have tended to comply with, rather than improve on, local norms. Moreover, the multinational companies (MNCs) prioritize the appointment of Chilean nationals as executives, and line managers are often poorly informed of companies’ policies which propose management by ‘harmonizing competitiveness with the humanization of work’ (Díaz Andrade, 2003: 47 – authors’ translation). In two of Díaz Andrade’s case study companies, there had been greater clarity in the vertical integration of the production chain, allowing national and international peers to be identified (2003:49). In contrast, in another MNC there was a well publicized ethical code of conduct but it was not applied equally in all stages of production. Unions at the cultivation and processing plants reported that its diffusion was only partial, whereas at the salmon feed plant, its application was rigorous (2003: 51). While these differences could be attributed to the presence of distinct management teams and differences in the skill composition of the workforces, the fact that the code of practice was only available in English affected the ability of the workforce to understand its contents (2003: 51). Another of her case study companies had received the Investors in People Award for its operations in the UK, prompting Díaz Andrade to ask, ‘Why is it that what is good for the workforce in Europe is not also good for the workforce in Chile?’ (2003: 52, authors’ translation).

To summarize, in terms of global value chain analysis, one can see that producers based in Chile are primarily providing natural resources and cheap labour inputs for markets located primarily in the Northern hemisphere. Though in its initial stages the industry was characterized by a relatively disaggregated value chain, in the 1990s large firms integrated various previously dispersed stages of the value chain, resulting in a more consolidated industry. Even today Chilean firms are not present in some of the more knowledge intensive nodes of the value chain, despite some state-supported efforts in the area of research and development where imported solutions are not available (these are primarily related to specific Chilean environmental conditions). Countervailing tendencies of disintegration can be observed in the outsourcing of labour and the division of production units into separate legal entities, which Chilean law permits. These have been used as a mechanism for combating union attempts to bargain at company level. While a small number of foreign owned multinationals have encouraged international communication between workers in different parts of the value chain, the majority of foreign owned investments have been through acquisitions. Alongside their preference for recruiting Chilean nationals as managers, this suggests managers prioritize adaptation to local conditions over innovation in management practices.

Conclusion

In this article the authors have attempted to bring social institutions into global value chain analysis by focusing on the historical development of the Chilean salmon farming industry. This new industry expanded rapidly with the support of state institutions which in turn encouraged employer cooperation through SalmonChile. It has been shown that there has been some limited development of Chilean technological capacity, but most knowledge intensive inputs are dependent on imported knowledge. With some exceptions (e.g. EWO’s small R&D unit and AquaChile’s Aquainnova facilities) nationally owned and MNC producers in Chile are primarily located in the lower knowledge intensive segments of the value chain. This raises important questions about the role of the state in promoting research and innovation and the significance of decisions relating to the importation of technology as opposed to the development of independent national capabilities through social institutions such as universities and their linkages with industry, technological institutes and public-private research organizations.

As far as skill development is concerned, the salmon industry has generated a demand for specialists in biochemistry, ichthyopathologists, technicians and administrators specializing in fish farming, which is met by a supply of graduates. It appears to have had relatively little impact on operative skills and career progression routes are not in evidence. Rather, Chilean labour represents a relatively cheap input into the production process and low level competence assessment is the primary initiative undertaken by the industry body for the sector. This suggests that the motivation for employer cooperation has been linked to quality assurance systems and marketing, on the one hand, and as a mechanism for dealing with labour turnover, on the other. To date, the development of a cadre of technicians and the establishment of a career path for unskilled workers have not been priorities. Institutions promoting employer investment in training and the development of worker expertise and problem-solving capabilities at the point of production are essential to increasing bottom up capacity for innovation and addressing the causes of turnover. Struggles over collective bargaining rights suggest that basic issues concerning the quality of work and employment and authoritarian management styles lie behind this failure of investment in skill development. Moreover, producers compete in international markets on the basis of price and cost reductions centered on wage squeezing and inferior employment conditions – the ‘low road’ to competitiveness – instead of adopting strategies to increase productivity and innovation based on the development of innovative capabilities and a well trained and skilled workforce – the ‘high road’ to competitiveness (Porter, 1990). Chile’s liberal market model, based on the natural resource-based model of development, makes it difficult for locally based producers to capture more value in the GVC. Moreover, multinational firms’ location decisions appear to be driven by the desire to exploit natural resources and to reduce labour costs, rather than in enhancing workers’ skills and productivity. There is limited evidence to date of their acting as innovators in HRM practices.

GVC analysis has much to offer researchers interested in the impact of globalization on work and employment and raises questions about the levers for labour upgrading which have consequences for research design and methods. This article has focused on a single national context and questions relating to innovation and skills, but it is evident that some of the actors in the Chilean context, the foreign owned companies, operate with a very different structure of labour costs, skill structures and management practices in their home countries. To understand the extent to which national social institutions create competitive advantage for companies and support labour quality requires comparative study. Comparative international study in turn needs to take into account the significance of institutional resources in companies’ ability to engage in different parts of the value chain. The value of international comparison would lie in understanding how the pressures for vertical integration and concentration, on the one hand, and decentralization and outsourcing, on the other, vary between locations. The scope that different actors have for enhancing labour quality and conditions in particular locations would therefore be informed by an understanding of the relative weight of national institutions, on the one hand, and relationships with buyers, suppliers and other producers in different parts of the value chain, on the other.