Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Widely used from the 1960s to the 1980s in formulating industrial policies, the concept of filière was also called upon to accompany the agricultural sector’s transformations in the North and to orient development policies in the South. The filière was thus constituted as an analytical framework for the vertical structuring of production systems, the production of intermediate and final goods, and forms of coordination between economic agents. It has been promoted historically by governments to define their intervention methods. If the use of the term filière in the francophone academic world was at the core of social science research programmes, in the English- and Spanish-speaking worlds the concept is manifested through variant terms (supply chain, value chain) that are relevant to different user communities. These terms are, in turn, increasingly finding use in French without always a clear explanation of whether this semantic shift corresponds to new analytical frameworks and new issues, linked to new forms of public and private actions.

We therefore define here these terms, their uses, and their analytical bases so that we can better analyze how agri-chains are transforming in the face of the challenges of sustainable development and examine how the changes brought about address these issues. To this end, we offer an historical exploration of the emergence of this concept in the literature and the diversification of contexts of its use. We analyze their effects on the evolution of the concept of filière not only into an analytical framework, but also into various operational tools that are now part of public and private modes of intervention.

1 The Concept of Filière

1.1 A Bridge Between Micro- and Macro-economics: A Disaggregation of the Economic System

Paradoxically, although the French term filière for value chain or supply chain is connected to the francophone academic world, it was used for the first time in 1947 at a convention of the Econometrica journal in the United States when B. Chait presented a theory of the relationships between agents to explain the flows connecting them (the output of an agent corresponding to an input for his client) and inventory levels needed to regulate the system. The initial objective was therefore to provide an analytical framework that overcame one of the main limits of the pure and perfect competition model by introducing inventories, and thus time, in the design of a general model to explain economic coordination. This initial conceptualization of a filière on the basis of an input/output system is based on separating out the economic system’s components (agents). The objective is to define subsets of agents in which the internal relationships (interdependences) are stronger than this subset’s relationships with the rest of the economy. Implications for the formulation of industrial policy or development strategies were not discussed at the time.

In France, the term was gradually defined in a very close sense through studies of industrial policies that used the tools of national accounts (Supply and Use tables; Aujac 1960). The concept of filière was then developed further to overcome the limits of an input/output analysis, whose division into sectors and branches was not sufficient for defining strategies to support economic activities. The filière was thought of here as a tool for subdividing the economic system on the basis of the relationships shown in the table of inter-industry trade, by highlighting the backward linkages effects that an agent has on the next. This concept was then applied to the implementation of development and industrialization policies in developing countries as a tool to orient the strategy of ‘industrializing’ industries, as argued by Perroux and Destanne de Bernis in particular. The filière approach also allows the creation of added value, and its distribution between agents or between production locations, to be analyzed.Footnote 1 It played an important role in the development of the ‘effects method’, a method for the economic evaluation of investment projects, which was a key tool of French cooperation for 30 years, from the 1970s to the late 1990s (Chervel and Le Gall 1976).Footnote 2

In addition to helping identify the relationships that structure an economic system, the filière was also a framework for representing the sequence of techniques that lead to the development of ever more complex products (for example, the graphite-gas filière or supply chain). The concept of filière gradually became more structured as analyses of power relationships and technological changes were incorporated into it. The Revue d’économie industrielle (Industrial Economics Journal) founded in 1977 was the main medium of this enrichment, with the filière becoming an analytical framework shared by a growing community of economists working in different fields of application, such as those of the agricultural and rural economy and of development economics (Hugon 1988).

1.2 The Production of Food Goods: A Privileged Topic of Study and Conceptualization

Authors such as Davis and Goldberg (1957) in the US have used the term ‘agribusiness’ to conceptualize the growing role of processes of transforming agricultural products into food and of their distribution, as well as the industrialization of agricultural inputs with the rise of the agrifood and agrochemical sectors after the Second World War. Their approach has highlighted the role of the agricultural sector as a supplier of raw materials and, therefore, the necessity of analyzing the links between agricultural production and upstream and downstream functions.Footnote 3 In France, the work of Milhau (1954) on the vertical links within agricultural markets (similar to the work of Chait) also led to the recognition of vertical economic coordination. This logic of organization into filières has been central to the development of agricultural policies that have structured the processes of specialization by product. A similar approach was proposed in the 1970s by the Food Research Institute at Stanford University. It advanced the concept of the commodity system to help understand the processes of transformation of agriculture in the South, which had started facing increasing competition from international markets in a rapidly globalizing agrifood system.

The analysis of connections between agricultural production and other sectors of the economy led to an increasing reliance on this concept. Thus, in France, a specific scientific programme in the agricultural and food sector was set up within INRA and national higher agricultural engineering schools. In the South, filière analyses played a major role in the formulation of national agricultural policies such as those for dealing with challenges of international competition (policy analysis matrix). In francophone Africa, they formed the basis for the organization of the agricultural sector.

The major agri-chains (livestock meat, milk, grapes and wine, etc.) have thus been characterized within the framework of policies of agricultural transformation under the impetus of the common agricultural policy. The filière or value chain, defined as the linked sequence of technical, logistical, and commercial operations necessary to produce and distribute a food or agro-industrial product, from production to consumption, thus became a useful reference for the research community as well as for policymakers. Filière and value chain economists then more explicitly mobilized systemic reference bases to account for intended consistencies of the proposed analytical division. Filière or value chain analysis thus seemed to have been accepted as a new concept able to take into account intermediate categories (mesoeconomics) that structure the coordination of economic agents between the micro- and macro-economic dimensions, thus breaking with the restrictive assumptions of neo-classical economics.

The food system was then defined in agriculture, in complement to the concept of filière or value chain, as the set of all food chains that link producers to consumers (Malassis 1979). In developing countries, these two concepts were used to analyze the performances of different export crops. They were also subsequently used in studies on food supply channels to cities (Baris and Couty 1981). Due to the way CIRAD used to be organized – by broad commodity chains – this work was further advanced in this context by many of its researchers and those of its partners (Griffon 1989; Fabre et al. 1997) for analyzing tropical export chains (coffee, cocoa, oilseeds, cotton, banana, etc.) and food supply to urban markets (rice, maize, plantain, vegetables, milk, meat, etc.). During the severe economic crisis in the African countries of the CFA franc zone in the early 1990s, several studies on the competitiveness of agri-chains were carried out at the behest of or by the French government (for example, Jouve and de Milly 1990; Daviron and Fousse 1993; Ministry of Cooperation 1994). French public aid contributed to the dissemination of studies of agri-chains at the FAO and the European Commission, and remained focused predominantly on the ‘efficiency’ of agri-chains (Griffon et al. 2001). Although starting in the 2000s, tools for analyzing the place of agricultural systems of the South in social and economic dynamics became widely diversified, CIRAD continued to use the filière as a reference (Temple and Lançon 2008; Temple et al. 2011).

It is not by chance that the concept of filière or value chain acquired a predominant role in analyzing and defining public policies for the integration of agricultural and livestock production into the economy. Historically, the processing of agricultural products into food is based on a small number of stages. The processes involved consist primarily of extraction and sorting. Very rarely are combination and assembly also part of the processing; these steps are usually left to the consumer or the cook at the time of preparing the dishes – though in recent decades the agrifood industry has also started producing ready-to-eat products. Even if the use of this concept is not exclusive to the food sector and extends to some industrial sectors (textiles at the beginning of the century and then energy, etc.), agrifood processes are especially well-suited to a value chain or supply chain analysis since the initial agricultural product used serves as a frame at every stage. In contrast, manufacturing of industrial products by assembling components is less conducive to this method of analysis. The value chain is also an important instrument for analysis and a method of intervention in the South because it can be used in contexts of limited or incomplete information. It also gained legitimacy through the strong polarization and dependence of the dynamics of development on a few agro-industrial chains in which the processes of accumulation concentrate (Hugon 1985).

This value chain approach based on a combination of the various functions necessary to produce the final product has been gradually revised with new analytical references that reflect the growing importance of the globalization of production systems, a process in which companies, initially part of industrial groups and now increasingly of financial ones, occupy a prominent place. Thus, in the work of Bandt, and its extension into the contexts of the countries of the South (Hugon 1994), the concept of structuring an economy in terms of value chains has been advanced, with the theory of regulation and the changes in modes of accumulation in countries of the North and the South as frames of reference. The economics of conventions is also mobilized, in particular to clarify the role of quality standards in the control exerted over the value chains by the companies that define them (Ponte and Gibbon 2005).

2 The Value Chain: An Instrument for Analyzing Business Strategies in a Context of Globalized Competition

Major technological advances in long-distance transportation, the liberalization and financialization of trade, and the increasing mobility of non-tangible goods such as information intensified competition in the 1980s in the industrialized economies and the expansion of international markets to new products. Value chain analyses were frequently called upon at the time to determine and assess the determinants of competitiveness of industrial as well as agro-industrial companies. The performances of these companies were no longer determined solely by the cost and productivity of factors of production as argued by the standard theory of comparative advantage. They also depended on the ability of these companies to leverage a set of external resources, to create a framework for institutionalized coordination between economic agents, notably through conventions of quality, and to position themselves at strategic points in production systems (Porter 1985).

The need to consider the configuration of value chains not solely as a process of division and specialization of productive tasks, but also as arising out of the strategies of companies to manage their position in these systems had already been addressed in the methodologies for analyzing filières. Nevertheless, it was not until the 1990s that companies began to be widely considered as entry points for analysis.Footnote 4 This development arose from two different strands of thought. The first belongs primarily to the management sciences and shows how a company’s competitiveness is determined by its position and its relationships with other actors of the same production system. The second strives to understand how economic globalization leads to a reorganization of production systems at the planetary scale. Both unite in confirming the importance of the strategies of agro-industrial companies in the governance of globalized value chains.

2.1 Value and Supply Chain: A Strategic Framework for the Competitiveness of Companies

If sociologists of development questioned very early on the global logic of organization of production chains as a way to capture value, they did so with the aim of explaining the inequalities of development. Other authors, however, conceptualized value chains as a tool for optimizing the strategy of firms in a rationale of supporting decision-making processes. The value chain then designates the boundary within which the firm optimizes its relationships, not only with upstream and downstream actors but with all the services it mobilizes to become and remain competitive (research and innovation, maintenance, transportation, training, etc.). The concept of the value chain was thus initially conceived of as a management tool. The notion of value is here understood in a wider sense than that of added value of filière analysis; it is thus often associated, in an ex ante situation, to the difference between the price the consumer is willing to pay and the producer’s reserve price.Footnote 5

The management sciences also offer a tool similar to that of value chain but more operational in nature: the supply chain. The supply chain, formalized as far back as the 1950s (Forrester 1958) and which then gave rise to the supply chain management approach (Harland 1996), focuses on the efficient flow of goods and services between a company, its suppliers, and customers in order to coordinate production chains that involve several industrial entities. This analytic framework is an application of the concerns first advanced by Chait (1949) to conceptualize inventories and time in a framework of pure and perfect competition. It uses the tools of information and of network theories, with its applications multiplying as new information and communication technologies develop (Omta et al. 2008). It is the counterpart of just-in-time production methods, i.e., business-to-business computer applications aimed at reducing inventories and ensuring efficient coordination of multiple production-related decision making.

2.2 From National Production Systems to Transnational Production Systems: The Global Value Chains

As mentioned above, another approach was proposed for an improved understanding of the internationalization of business activities, which accelerated in the 1980s with the advent of neoliberal policies that promoted free trade and the financialization of the economy and with the emergence of new countries and the associated restructuring of the international division of labour. This change in ambit within which the companies defined their strategy became visible in the 1970s when large industrial groups began reorganizing themselves. Their international departments were eliminated and replaced by the grouping of their various business activities by major world regions according to available resources (raw material, technology, other production factors) or end-product demands (Michalet 1976). This transformation was theorized by Gereffi and Korzeniwicz (1994), first under the concept of ‘global commodity chain’, then as the ‘global value chain’ (Gereffi et al. 2005). The concept of the global value chain refers to the same subdivision of production systems as in filière analysis but it highlights the role of ‘driver’ in the chain who have the ability (in the form of technology, financial capital, network) to govern this process. Global value chains differentiate themselves depending on the link in which the governance capacity resides, upstream or downstream, according to the types of more or less sophisticated products manufactured therein. As far as public policies are concerned, if globalization calls national industrial policies into question, the conceptual framework provided by the global value chain allows governments to target their actions (on the type of value chain or level of intervention within the value chain) in order to anchor global value chains in national territories and induce backward linkages on their development (Palpacuer et al. 2005).

Agriculture occupies a paradoxical position in this globalization process. Colonial trade in tropical products (coffee, tea, cocoa, sugar, cotton, banana, etc.), which combined production locations in the South with those of processing and value addition in the North, can be considered the first forms of global value chains. However, their rationale of accumulation was limited to the ambit of the colonial empires. The current phenomenon of globalization of agri-chains is more recent than that of the industrial sector. Globalization has really come into its own since the 2000s; it corresponds to a process of dissemination of food consumption patterns around the world (processed cereals, chicken, vegetable fats, dairy products, chocolate) and of domination by internationally renowned brands, whether in the supply of upstream inputs (seeds) or the downstream distribution of finished products (brands, ready meals, fast food restaurant chains). ‘The supermarket revolution’, as T. Reardon called it, is one manifestation of this dynamic, especially in emerging countries (Reardon et al. 2003). It is resulting in a restructuring of domestic forms of distribution, based in varying degrees on the development of specialized wholesalers, the consolidation of logistics and distribution centres, contracts with preferred agricultural suppliers, and private standards (Reardon and Berdegué 2002). By modifying the relationships between producers and the downstream parts of the value chain, this dynamic has generated substantial effects of exclusion for small producers (Weatherspoon and Reardon 2003). Another manifestation is the emergence of globalized companies in the South, for example in the production of instant rice noodles which, though they originated in Southeast Asia, are now being widely distributed in Africa.

2.3 An Integrated Conceptual Framework of the Value Chain, a Predominant Reference for the Development of Rural Areas

Since the early years of this century, the scientific community and international institutions (for example, the World Bank) have most frequently relied on the concept of the value chain for analyzing relationships between actors in a production system. The business strategies approach dominates and is enriched and articulated by transaction cost theory and its qualification of modes of coordination (market, contract, and hierarchy) (Gereffi et al. 2005).

The value chain serves as a reference for strategies of development in the South, especially those pertaining to agriculture, formulated a decade ago by international institutions (Werner et al. 2014). This reference base renews the initial concept of filière by introducing the study of governance mechanisms, while retaining the methodological fundamentals. Value chain developments offering stable outlets for farmers, particularly the smaller ones amongst them, are considered a preferred vehicle for inclusive growth, resulting in a permanent reduction of rural poverty (Vermeulen et al. 2008). These strategies have been formulated in response to the limited successes in alleviating poverty, if not the counterproductive consequences, of the liberalization policies of the 1980s (Casadella et al. 2015). These policies were based on the strong assumption that a spontaneous development of all-inclusive markets would connect rural producers to the globalizing economy, an assumption belied by 20 years of increased differentiation.

The value chain is understood as a framework for constructing markets that allows the formulation of integration strategies consistent with a market economy, and hence as a framework for defining interventions in pursuit of development. The intervention of the State is acknowledged in producing an institutional framework for market functioning, but it remains limited. This context explains why value chain R&D programmes assign significant importance to forms of coordination such as contracts, considered a controlled form of trade, which overcomes the disadvantages of trade outside a formal framework (asymmetrical position, risk, uncertainty, etc.). By the same logic, studying the effects of standardization and codification of exchanges and trade is necessary in order to determine the steps to take so that rural producers can meet the demands of the dominant actors of the agrifood systems (agro-industries, distribution centres, consumers, etc.) (see for example, Dolan and Humphrey 2000). The same causes underpin the emphasis on the development of market information systems (bids, availability, etc.).

The institutionalization of value chains as a development tool is widely promoted as a way to link the macro-economic level, in the form of investments and international trade, with development issues at the micro-economic level, in the form of improved livelihoods. But it is also denounced as a way for dominant actors to push and perpetuate the neoliberal development agenda (Neilson 2014) and, in the agricultural domain specifically, as a new technology of economic and ecological power for the appropriation of value produced by small farmers (MacMichael 2013).

3 Conclusion: Incorporating Sustainable Development Issues in the Concepts of Filière and Value Chain

While the various different conceptualizations of the filière show a close similarity in the way they represent it – a system of interdependent actors fulfilling complementary functions for the manufacture of a product or a group of technologically homogeneous products –, each relies on a different perspective to analyze these interdependencies. Some lay emphasis on functional interdependencies in terms of inputs, outputs, and management of flows, others on forms of coordination between the actors. These distinctions reflect the issues that these analytical frameworks are meant to address, the scale and the scope of the analysis, and, finally, the roles assigned to public policies, and forms given to them, for addressing these issues.

Since filière and value chain approaches have been used to structure the productive dynamics of agricultural specialization (territorial or sectoral), whose environmental and social impacts continue to be controversial, they have come in for criticism for their impacts on the sustainability of development. Indeed, as Bolwig et al. (2010) emphasize, analyses of value chains only partially address the impacts of these value chains’ activities on poverty, inequalities (including gender issues), food security, and the environment. As far as social impacts are concerned, these analyses are generally limited to opportunities of generation of income without a consideration of the exposure of poor populations to various risks. Furthermore, they take the environment into account only through the role of quality standards.

Indeed, the issue of sustainable development is historically quite alien to these analytical frameworks. This explains the use, since the 1970s, of life cycle assessment (Boustead 1996) to develop a method to evaluate environmental externalities and then the material benefits of recycling materials. This method has also been used to construct indicators for the consumption of material and energy in order to develop environmental standards or to meet them. In methodological terms, life cycle assessment is very similar to filière analysis. It favours an approach by technical functions, with an evaluation system in physical units on one side and in monetary values on the other. Nevertheless, life cycle assessments remain very focused on the technical goal of measuring environmental effects and developing environmental standards. The difficulties of identifying the social impacts and dynamics that govern the coordination of actors and technological choices make life cycle assessments useful but imperfect tools for determining the sustainability of production systems and for examining the relationships between a value chain and sustainable development.

Without getting into a discussion here on the definition of ‘sustainable development’, we can note that attempts to analyze sustainability are confronted by the challenges of scale and the feasibility of the methods of analysis. In its most comprehensive and rigorous sense, the sustainability of a human activity can be really appreciated only by taking into account its interaction with the entire biophysical system at the planetary scale. It is a matter of analyzing to what extent the production of a final good and the entire value chain associated with it increases entropy and reduces non-renewable resources. This would require, once the value chain’s system has been clearly defined, the qualification of relationships in terms of inputs and outputs with the system ‘Earth’ in its entirety, and this in a dynamic perspective.

Another issue is that of mapping forms of physical sustainability to forms of governance and of mapping relationships between actors. In this regard, an approach solely based on standards, which are the products of a social construction, cannot suffice to analyze the effects of different forms of organization and of the structuring of economic activities on sustainable development. From the agribusiness to the filière to the global value chain, the various analytical frameworks all aim to characterize these processes of structuring and organization of production systems and to evaluate their performances according to criteria that encompass only partially the imperatives of sustainable development.

An analytical framework that allows all three dimensions of sustainable development to be articulated – preferably one that uses a dynamic approach (modelling) – remains yet to be developed. Approaches in terms of sustainable food systems, in which the concept of the filière or value chain can characterize supply chains and their shortcomings constitutes possible directions for future research.