Review of International Political Economy12:1February2005:78–104
The governance of global value chains
Gary Gereffi
governanceDuke University
John Humphrey
Institute of Development Studies
and
Timothy Sturgeon
Massachusetts Institute of Technology
ABSTRACT
This article builds a theoretical framework to help explain governance pat-terns in global value chains.It draws on three streams of literature–trans-action costs economics,production networks,and technological capability andfirm-level learning–to identify three variables that play a large role in determining how global value chains are governed and change.These are:(1)the complexity of transactions,(2)the ability to codify transactions, and(3)the capabilities in the supply-base.The theory generatesfive types of global value chain governance–hierarchy,captive,relational,modular, and market–which range from high to low levels of explicit coordination and power asymmetry.The article highlights the dynamic and overlapping nature of global value chain governance through four brief industry case studies:bicycles,apparel,horticulture and electronics.
KEYWORDS
Global value chains;governance;networks;transaction costs;value chain modularity.
The world economy has changed in significant ways during the past sev-eral decades,especially in the areas of international trade and industrial organization.Two of the most important new features of the contemporary economy are the globalization of production and trade,1which have fueled
Review of International Political Economy
ISSN0969-2290print/ISSN1466-4526online C 2005Taylor&Francis Ltd
uk
DOI:10.1080/09692290500049805
GEREFFI ET AL.:GLOBAL VALUE CHAINS
the growth of industrial capabilities in a wide range of developing coun-tries,and the vertical disintegration of transnational corporations,which are redefining their core competencies to focus on innovation and product strategy,marketing,and the highest value-added segments of manufac-turing and services,while reducing their direct ownership over‘non-core’functions such as generic services and volume production.Together,these two shifts have laid the groundwork for a variety of network forms of gov-ernance situated between arm’s length markets,on the one hand,and large vertically int
egrated corporations,on the other.The purpose of this article is to generate a theoretical framework for better understanding the shifting governance structures in sectors producing for global markets,structures we refer to as‘global value chains’.Our intent is to bring some order to the variety of network forms that have been observed in thefield.2
The evolution of global-scale industrial organization affects not only the fortunes offirms and the structure of industries,but also how and why countries advance–or fail to advance–in the global economy.Global value chain research and policy work examine the different ways in which global production and distribution systems are integrated,and the possibilities for firms in developing countries to enhance their position in global markets. One of our hopes is that the theory of global value chain governance that we develop here will be useful for the crafting of effective policy tools related to industrial upgrading,economic development,employment creation,and poverty alleviation.
1.FRAGMENTATION,COORDINATION,
AND NETWORKS IN THE GLOBAL ECONOMY
For us,the starting point for understanding the changing nature of inter-national trade and industrial organization is contained in the notion of a value-added chain,as developed by international busines
s scholars who have focused on the strategies of bothfirms and countries in the global economy.In its most basic form,a value-added chain is‘the process by which technology is combined with material and labor inputs,and then processed inputs are assembled,marketed,and distributed.A singlefirm may consist of only one link in this process,or it may be extensively ’(Kogut,1985:15).The key issues in this literature are which activities and technologies afirm keeps in-house and which should be outsourced to otherfirms,and where the various activities should be located.
Trade economists are also concerned with how global production is or-ganized.Arndt and Kierzkowski(2001)use the term‘fragmentation’to describe the physical separation of different parts of a production pro-cess,arguing that the international dimension of this separation is new. Fragmentation allows production in different countries to be formed into
79
REVIEW OF INTERNATIONAL POLITICAL ECONOMY
cross-border production networks that can be within or betweenfirms. Feenstra(1998)takes this idea one step further by explicitly connecting the‘integration of trade’with the‘disintegration of production’in
the global economy.The rising integration of world markets through trade has brought with it a disintegration of multinationalfirms,since companies arefinding it advantageous to‘outsource’an increasing share of their non-core manufacturing and service activities both domestically and abroad. This has led to a growing proportion of international trade occurring in components and other intermediate goods(Yeats,2001).3
If production is increasingly fragmented across geographic space and betweenfirms,then how are these fragmented activities coordinated?For Arndt and Kierzkowski,the options are clear:‘Separability of ownership is an important determinant of the organizational structure of cross-border production sharing.Where separation of ownership is not feasible,multi-national corporations and foreign direct investment are likely to play a dominant role.Where it is feasible,arm’s-length relationships are possible and foreign direct investment is less important’(Arndt and Kierzkowski, 2001:4).
This binary view of how global production might be organized,either through markets or within transnationalfirms,is explained by transac-tion costs economics in terms of the complexity of inter-firm relationships and the extent to which they involve investments specific to a particular transaction–asset specificity(Williamson,1975).Arm’s-length market re-lations work well for standard products because they are easily described and valued.Coordination problems are reduced not only
because their ease of description makes contracts simple to write,but also because standard products can be produced for stock and supplied as needed.At the same time,because standard products are made by a variety of suppliers and bought by a variety of customers,problems arising from asset specificity are low.
Conversely,the transaction costs approach offers various reasons why firms will bring certain activities in-house.First,the more customized the product or service,the more likely it is to involve transaction-specific in-vestments.This raises the risk of opportunism,which either rules out out-sourcing altogether,or makes it more costly because safeguards have to be put in place.Second,even without opportunism,transaction costs increase when inter-firm relationships require greater coordination.For example, non-standard inputs and integrated product design architectures involve more complex transfers of design information and therefore intense in-teractions across enterprise boundaries.Integral product architectures are more likely to require non-standard inputs,and changes in the design of particular parts tend to precipitate design changes in other areas of the system(Fine,1998;Langlois and Robertson,1995).Similarly,coordina-tion costs increase for parts whose supply is time-sensitive,as separate
80
GEREFFI ET AL.:GLOBAL VALUE CHAINS
processes have to be better coordinated in order to synchronize theflow of inputs through the chain.
Nevertheless,recognizing the importance of transaction costs need not lead to the conclusion that complex and tightly coordinated production systems always result in vertical integration.Rather,asset specificity,op-portunism,and coordination costs can be managed at the inter-firm level through a variety of methods.Network actors in many instances control opportunism through the effects of repeat transactions,reputation,and social norms that are embedded in particular geographic locations or so-cial groups.Network ,Jarillo,1988;Lorenz,1988;Powell, 1990;Thorelli,1986)argue that trust,reputation,and mutual dependence dampen opportunistic behavior,and in so doing they make possible more complex inter-firm divisions of labor and interdependence than would be predicted by transaction costs theory.
Furthermore,the literature onfirm capabilities and learning,which has its roots in the resource view of thefirm pioneered by Penrose(1959),pro-vides other reasons whyfirms are prepared to buy key inputs in the face of asset specificity and therefore construct relatively complex inter-firm relationships.According to Penrose,how and whetherfirms can capture value depends in part on the
generation and retention of competencies (that is,resources)that are difficult for competitors to replicate.In practice, even the most vertically integratedfirms rarely internalize all the techno-logical and management capabilities that are required to bring a product or service to market.Transaction cost economics acknowledges this fact by employing the variable of frequency.If an input,even an important one,is required infrequently,then it will likely be acquired externally.This is essentially an argument about scale economies.The literature onfirm capabilities and learning,by contrast,argues that the learning required to effectively develop the capability to engage in certain value chain activities may be difficult,time-consuming,and effectively impossible for somefirms to acquire,regardless of frequency or scale economies.Thus,firms must in certain instances depend on external resources.The doctrine of‘core competence’takes this a step further,arguing thatfirms which rely on the complementary competencies of otherfirms and focus more intensively on their own areas of competence will perform better thanfirms that are ver-tically integrated or incoherently diversified(Prahalad and Hamel,1990). These issues,while often discussed at the local or national level,or in the context of‘a dense network of social relations’(Granovetter,1985:507),can equally be applied to the structuring of global-scale production and distri-bution.The recent work of geographers such as Hughes(2000),Henderson et al.(2002)and Dicken et al.(2001)has emphasized the complexity of inter-firm relationships in the global economy.The key insight is that co-ordination and c
ontrol of global-scale production systems,despite their complexity,can be achieved without direct ownership.
81
REVIEW OF INTERNATIONAL POLITICAL ECONOMY
The theories of industrial organization discussed here,when considered cumulatively,suggest that different ways of dealing with the problem of asset specificity,and different motivations for constructing complexfirm-to-firm relationships in the face of asset specificity,result in three modes of industrial organization:market,hierarchy,and network.But empirical observation tells us that not all networks are alike.In the next section we develop a theory that can help to specify and explain this variation.
2.TYPES OF GOVERNANCE IN GLOBAL VALUE CHAINS If a theory of global value chain governance is to be useful to policymakers, it should be parsimonious.It has to simplify and abstract from an extremely heterogeneous body of evidence,identifying the variables that play a large role in determining patterns of value chain governance while holding oth-ers at bay,at least initially.Clearly,history,institutions,geographic and social contexts,the evolving rules of the game,and
path dependence mat-ter;and many factors will influence howfirms and groups offirms are linked in the global economy.Nevertheless,a simple framework is useful because it isolates key variables and provides a clear view of fundamen-tal forces underlying specific empirical situations that might otherwise be overlooked.Our intention is to create the simplest framework that gener-ates results relevant to real-world outcomes.
In the1990s Gereffiand others developed a framework,called‘global commodity chains’,that tied the concept of the value-added chain directly to the global organization of industries(see Gereffiand Korzeniewicz, 1994).This work not only highlighted the importance of coordination acrossfirm boundaries,but also the growing importance of new global buyers(mainly retailers and brand marketers)as key drivers in the forma-tion of globally dispersed and organizationally fragmented production and distribution networks.Gereffi(1994)used the term‘buyer-driven global commodity chain’to denote how global buyers used explicit coordination4 to help create a highly competent supply-base upon which global-scale pro-duction and distribution systems could be built without direct ownership. By highlighting explicit coordination in dis-integrated chains and con-trasting them to the relationships contained within vertically integrated,or ‘producer driven’chains,the global commodity chains framework drew attention to the role of networks in driving the co-evolution of cros
s-border industrial organization.However,the global commodity chains framework did not adequately specify the variety of network forms that more recent field research has uncovered.While,research on the horticulture indus-try(Dolan and Humphrey,2000)and the footwear industry(Schmitz and Knorringa,2000)reinforced Gereffi’s notion that global buyers(retailers, marketers,and traders)can and do exert a high degree of control over spatially dispersed value chains even when they do not own production,
82
GEREFFI ET AL.:GLOBAL VALUE CHAINS
transport or processing facilities,recent research on global production has highlighted other important forms of coordination.
Work on the electronics industry and contract manufacturing by Sturgeon(2002)and by Sturgeon and Lee(2001)contrasted three types of supply relationships,based on the degree of standardization of prod-uct and process:(1)the‘commodity supplier’that provides standard products through arm’s length market relationships,(2)the‘captive sup-plier’that makes non-standard products using machinery dedicated to the buyer’s needs,and(3)the‘turn-key supplier’that produces customized prod
ucts for buyers and usesflexible machinery to pool capacity for different customers.This analysis emphasized the complexity of infor-mation exchanged betweenfirms and the degree of asset specificity in production equipment.Sturgeon(2002)referred to production systems that rely on turn-key suppliers as‘modular production networks’because highly competent suppliers could be added and subtracted from the global production arrangements on as as-needed basis.Around the same time, Humphrey and Schmitz(2000,2002)distinguished between suppliers in quasi-hierarchical relationships with buyers,whose situation corresponds to‘captive suppliers’,and network relationships betweenfirms that coop-erate because they possess complementary competences.5Humphrey and Schmitz emphasized the role of supplier competence in determining the extent of subordination of suppliers to buyers.If global buyers needed to invest in supplier competence,they would need both to specify the prod-uct and process parameters to be followed by suppliers and to guard this investment in the supplier by remaining the dominant,if not exclusive, customer.6
Using the approaches outlined above and empirical reference points taken from many studies of global value chains,7we propose a more com-plete typology of value-chain governance.We acknowledge,as do most other frameworks that seek to explain industry organization–from trans-actions costs to global commodity chains to organizational theory–that market-based relationships a
mongfirms and vertically integratedfirms (hierarchies)make up opposite ends of a spectrum of explicit coordination, and that network relationships comprise an intermediate mode of value chain governance.What we add to this conceptualization is an extension of the network category into three distinct types:modular,relational,and captive.Thus,our typology identifiesfive basic types of value chain gover-nance.These are analytical,not empirical,types,although they have been in part derived from empirical observation.They are:
1.Markets.Market linkages do not have to be completely transitory,as is
typical of spot markets;they can persist over time,with repeat transac-tions.The essential point is that the costs of switching to new partners are low for both parties.
83

版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。