Global Value Chains Meet Innovation Systems: Are There Learning Opportunities for Developing Countries?

World Development Vol. 39, No. 7, pp. 1261–1269, 2011� 2011 Elsevier Ltd. All rights reserved

0305-750X/$ - see front matter

www.elsevier.com/locate/worlddevdoi:10.1016/j.worlddev.2010.05.013

Global Value Chains Meet Innovation Systems: Are There

Learning Opportunities for Developing Countries?

CARLO PIETROBELLIUniversity Roma Tre, Italy

Inter-American Development Bank, Washington DC, USA

and

ROBERTA RABELLOTTI *

Universita del Piemonte Orientale, Italy

Summary. — The Innovation Systems (IS) literature tends not to emphasize the crucial impact of international knowledge and innova-tion exchange and collaboration through, for example, inter-firm and intra-firm networks and Global Value Chains (GVC). In devel-oping countries this aspect is crucial, with integration in GVC playing a growing and very important role in accessing knowledgeand enhancing learning and innovation. However, there is no agreement in the literature about how innovation systems and GVC inter-act, and how this interaction is likely to affect enterprise learning.Three main conclusions emerge from the theoretical analysis and evidence presented in this paper. First, learning mechanisms can varywidely within the various forms of governance of GVC: they can be the result of the pressure to achieve international standards, or befacilitated by direct involvement of the value chain leaders when the suppliers’ competence is low and the risk of failure to comply is high.When the competences of the actors in the value chain are complementary, learning is mutual and is based on intense face-to-face inter-actions. Second, as we “open” IS to foreign sources of knowledge, the relationship between GVCs and IS is nonlinear and endogenous,and mutually affecting. On the basis of our model, we would expect a well-structured and efficient innovation system would help to re-duce transaction complexity and enable transactions based on relational forms of GVC governance. Third, the internal governance ofGVC is a dynamic phenomenon that is subject to continuous adjustments and changes, and the nature of the IS affects this co-evolution.� 2011 Elsevier Ltd. All rights reserved.

Key words — global value chain, innovation system, learning

* The authors thank Cristina Chaminade, Jorge Katz, Rajah Rasiah, Alan

Rugman, Jan Vang and the participants at the SLPTMD Oxford Confe-

rence (May 2008) and the Globelics Mexico Conference (September 2008),

plus two anonymous referees, for their comments. The opinions expressed

in the paper are those of the authors and do not necessarily reflect the

views of their respective organizations. Final revision accepted: May 20,2010.

1. INTRODUCTION

These days, no one would disagree that learning and innova-tion are the key to competitiveness and growth of countries,regions, and firms. In addition, some observers emphasize thatinnovation and learning in their turn are affected by firm-spe-cific attitudes and actions, and the meso and macroeconomiccontexts in which firms operate. In advanced countries, theconcept of Innovation System (IS) was developed to accountfor the role played by the institutions and the organizationsthat systemically interact and have an effect on the rate anddirection of technological change in an economic system(Lundvall, 1992; Nelson, 1993).

However, it is being stressed increasingly that the ISD ap-proach would be enriched by the incorporation of an interna-tional dimension (Asheim & Herstad, 2005; Bunnell & Coe,2001; Carlsson, 2006; Fromhold-Eisebith, 2007). The IS liter-ature often plays down the crucial impact of internationalinformation exchange and collaboration on the generationand diffusion of knowledge and innovation, for example,through inter-firm and intra-firm networks.

This argument in relation to less developed countries(LDCs) is even more important. The extra-national influenceson the innovation process are particularly crucial given thatfrontier innovation is rarely achieved in LDC and most ofthe knowledge and technology has to be imported. Variousstrands in the literature analyze the impact of foreign firmsin the process of innovation and learning in developing coun-tries, 1 the most recent focuses on Global Value Chains

1261

(GVC). 2 For firms in developing countries inclusion in GVCnot only provides new markets for their products, it also playsa growing and crucial role in access to knowledge and en-hanced learning and innovation.

In taking account of the linkages of IS to foreign sources ofknowledge, we have to recognize that the relationship betweenGVC and IS is nonlinear and endogenous, and mutuallyaffecting. For example, in terms of enterprise learning, GVCsmay contribute to improving the local IS, which, in turn, willaffect decisions about local sourcing of inputs, and support forlocal firms’ learning and innovation.

This paper addresses two research questions: how do learn-ing mechanisms operate in different types of chains (i.e.,through pressure to learn or through explicit support anddeliberate knowledge transfer)? And, what is the supportingrole of the IS in GVC-driven learning and innovation pro-cesses?

We use empirical evidence from our own research and sec-ondary sources. Given the highly differentiated reality behind

http://dx.doi.org/10.1016/j.worlddev.2010.05.013

1262 WORLD DEVELOPMENT

the sketchy and simplistic term “developing countries,” wefocus on a particular group of middle-income developingcountries: Brazil, Taiwan, and Mexico. In Section 2 we discussthe concept of IS in developing countries and review the mainissues addressed in the GVC literature in order to highlight thedifferent mechanisms of learning that may prevail in differenttypes of chains. Section 3 explicitly links GVC and their differ-ent governance patterns to the notion of IS, and discusses ave-nues of mutual interaction. Section 4 concludes.

2. IS AND GVC IN DEVELOPING COUNTRIES

(a) Innovation systems

Application of the IS 3 concept to developing countries isrelatively recent, but is increasing rapidly. With regard to itsapplication to developing countries, there are a number of rea-sons why this is not straightforward. 4

First, innovation processes differ from those in developedcountries: incremental innovation and absorption of knowl-edge and technologies new to the firm are more frequent andrelevant than radical, new to the world innovation. While theanalysis of IS in industrialized economies increasingly focuseson R&D and frontier innovation, in most LDCs, the natureof the technological effort is quite different, and is based mainlyon firm-level activities which are not included in formal mea-sures of innovation. In developing countries, most innovationis based on non-R&D activities which consist of operationaliz-ing technology that is new to the situation of application (Bell,2007). Second, the main science and technology organizationsanalyzed in developed country contexts, such as universities,research and development (R&D) laboratories, and researchinstitutes, may not exist in some developing countries or maybe inadequate, and linkages among them, and with local firmsmay be nonexistent or very weak. The organizations that aremore important in the systems in developing countries arethose providing technology diffusion and extension servicessuch as metrology, standards, testing and quality (MSTQ),and technical and organizational consultancies (or Knowl-edge-Intensive Business Services—KIBS). Third, inflows ofknowledge and technology from external sources are essentialcomponents of the innovation and learning processes in LDCs.Thus, it is the policies and institutions affecting internationalflows of equipment and services, human capital and foreigninvestments, as well as the GVC that matter.

This has led some authors to propose the term “NationalTechnology System,” since the bulk of technological activityin developing countries concerns absorption of and improve-ment to existing technologies rather than frontier innovation(Lall & Pietrobelli, 2002, 2003, 2005). 5 Most importantly, theconcept of a National Technology System emphasizes that indeveloping countries what is essential is the ability to absorbtechnology and knowledge produced elsewhere and to incre-mentally generate innovation. 6 In the emerging global patternof industrial organization, GVC represent an increasinglyimportant opportunity for firms in LDCs to learn and toinnovate.

(b) GVC and their patterns of governance

It is quite common for enterprises to outsource a number ofactivities that previously were handled internally, and to keepin house those activities in which they have core competences.Different parts of the production processes are becoming

increasingly dislocated across various developed and develop-ing countries. 7

A common feature in this new global division of labor isthat lead firms, often from developed countries, engage incoordinating the activities of their business partners upstreamand downstream. They may prefer different forms of gover-nance for different strategic reasons. A very useful typologyof GVC governance patterns was proposed by Gereffi,Humphrey, and Sturgeon (2005), who discuss the conditionsunder which different patterns can be expected to emerge.According to these authors, three factors determine the leadfirm’s choice of governance: the complexity of the informationinvolved in the transactions; the possibility to codify thatinformation; and the competence of the suppliers along thevalue chain.

The GVC literature stresses the role played by the leaders inthe chain in terms of transferring knowledge to their suppliers.For small firms in LDC, participation in value chains is a cru-cial means of obtaining information on the type and quality ofproducts and technologies required by global markets, and ofgaining access to those markets. However, this informationneeds to be combined with local technological capabilitiesand this requires substantial technological and learning efforts(Morrison, Pietrobelli, & Rabellotti, 2008). What role theleaders of GVC actually play in fostering and supporting thisprocess is one of the focuses of this literature (Giuliani et al.,2005; Pietrobelli, 2008).

The dynamics of governance patterns is crucial for under-standing the opportunities for suppliers “to move up the valueladder,” moving out of the “low road” to competitivenesswhere competition is based mainly on price and squeezingwages, and the barriers to entry are low (Pietrobelli & Rabe-llotti, 2007). By building and deepening their technologicalcapabilities, small suppliers in LDC can exploit opportunitiesfor different types of upgrading: process upgrading which istransforming inputs into outputs more efficiently by reorganiz-ing the production system or introducing superior technology;product upgrading which is moving into more sophisticatedproduct lines in terms of increased unit values; functionalupgrading implying the acquisition of new, superior functionsin the chain, such as design or marketing, or abandoning exist-ing lower-value-added functions to focus on higher-value-added activities; and inter-chain upgrading which is applyingthe competence acquired in a particular function to move intoa new chain. The challenge is not always about moving intomore advanced functions “along the value chain,” but is oftenabout deepening the specific capabilities required to explorenew opportunities offered “on the side,” in the value chainstage in which the firm is currently engaged (Morrison et al.,2008). Moving from natural resources to their exploitation,manufacturing, packaging, distribution, and branding is veryimportant and can be described as climbing the ladder. Butdeepening capabilities to explore new original features andvarieties at each stage of the GVC (e.g., from new flower vari-eties via biotechnological research, to new packaging with ori-ginal, highly-valued characteristics) is also important, andclearly requires learning, creation and acquisition of higherlevel skills, and more complex technological capabilities.

(C) The learning mechanisms within GVC

Integration in GVC is increasingly common among firms inLDCs that get access to knowledge, learn, and innovatethrough participation in these chains. To satisfy requirementsrelated to product quality, delivery time, efficiency of pro-cesses, environmental, labor, and social standards imposed

Table 1. Learning mechanisms within GVC

Governancetype

Complexity oftransactions

Codification oftransactions

Competenceof suppliers

Learning mechanisms within GVC

Market Low High High� Knowledge spillovers� Imitation

Modular High High High� Learning through pressure to accomplish inter-national standards� Transfer of knowledge embodied in standards,codes, technical definitions

Relational High Low High� Mutual learning from face-to-face interactions

Captive High High Low� Learning via deliberate knowledge transfer fromlead firms confined to a narrow range of tasks—for example simple assembly

Hierarchy High Low Low� Imitation� Turnover of skilled managers and workers� Training by foreign leader/owner� Knowledge spillovers

Source: Adapted from Gereffi et al. (2005).

GLOBAL VALUE CHAINS MEET INNOVATION SYSTEMS 1263

by GVC, firms specialized in different functions have to learnand to innovate. How learning takes place is influenced by thegovernance of the GVC, and different mechanisms of learningand innovation are likely to dominate in different types ofchains. Table 1 presents some empirical evidence on the differ-ent learning mechanisms discussed in the rest of this section, inorder

In arm’s-length GVC, which are more usual for relativelysimple and easily codifiable transactions involving competentsuppliers, only suppliers with the required capabilities canenter the value chain. Inclusion in a GVC provides a win-dow—and related information—on the global market’srequirements in terms of products, processes, technology andstandards. The main learning mechanisms are spillovers andimitation which allow small LDC firms to capture knowledgefor adaptive change and innovation needed to stay in the valuechain.

Schmitz (2004) provides some examples of market-basedchains characterized by the small size of buyers. In Brazil, buy-ers selling in the domestic market purchase ready-designedshoes and sell them either under their own labels or underthe supplier’s own brand. Similarly, in Ludhiana (India) knit-wear firms sell to small foreign traders and also develop theirown products (Tewari, 1999). Based on this empirical evi-dence, Schmitz (2004) concludes that advances in functionalupgrading seem to be facilitated by dealing with small ratherthan large customers. It is the different capabilities of firmsto make the required investments in design, product develop-ment, and marketing that may explain why some firms succeedand others do not.

When the complexity of the transactions is high and thereare capable potential suppliers, then modular chains prevailwith highly codified links and transactions. Technical stan-dards, requiring suppliers to make products to a customer’sspecifications and to take full responsibility for process tech-nology, contribute to codification. In modular chains supplierslearn how to produce components and modules to fully spec-ified technical standards. The need to adhere to these stan-dards is important for inducing learning; lead firms imposepressure on their suppliers to innovate and keep abreast oftechnological advancements, but do not become directlyinvolved in the learning process. In other words, lead firms

represent a crucial external stimulus for learning and innova-tion among suppliers, and are the spectators and final judgesof the process. In addition, upgrading within modular chainsmay result in positive externalities for the rest of the economybased on spill over to other sectors served by the samesuppliers. 8

Firms involved in modular chains need to undertake specificinvestments, build specialized production capabilities and con-stantly update in order to remain in the GVC. Their learningefforts must be accomplished independently since they are notsupported by the GVC leaders. Quadros (2004) analyzing thecase of GM and Volkswagen in Brazil, shows that GVC localsuppliers improved their production quality and achieved ISO9000 certification, but the leading firms in these chains playedlittle part in assisting suppliers to meet these standards. Tech-nical support came mostly from consultancies and accreditedcertification institutions. Similar evidence was found for theautomotive sectors in Argentina (Albornoz, Milesi, & Yoguel,2002) and Mexico (Dutrenit, Vera-Cruz, & Gil, 2002).

In relational chains transactions are complex and not easilycodified. Relationships tend to be idiosyncratic and thus diffi-cult and time-consuming to re-establish with new value chainpartners (i.e., switching costs are high). In relational chainsmutual dependence is regulated through reputation, socialand spatial proximity, long-term commitment and reputationand, in some cases, is based on family and ethnic ties. Trustis a deliberate strategy to enhance economic performance(Sako & Helper, 1998).

Given the complexity of tacit information and knowledge,the linkages in relational chains are very tight, and often in-volve a high proportion of face-to-face interaction and mutuallearning. The firms in these types of chains have highly com-plementary competences. LDC suppliers must be able to main-tain and strengthen their production and linkage capabilitiesto interact with lead firms in the GVC. Learning efforts imply(sunk) costs and take time, which binds parties into continuedinteraction.

An example of the evolution of a chain from captive to rela-tional is the apparel industry in East Asia, which upgradedfrom assembly to “full package” production, requiring thedevelopment of capabilities to interpret designs, produce sam-ples, monitor product quality, and meet buyers’ price and time


conditions (Gereffi, 1999). According to Gereffi et al. (2005:92), the main opportunity in these chains is that they allow“. . .local firms to learn how to make internationally competitiveconsumers goods and generates substantial backward linkages tothe domestic economy.”

Another case of a local supplier that progressed from pro-ducing to buyer’s specification to own design manufacturingcomes from the Taiwanese computer industry (Kishimoto,2004). The knowledge was transmitted through the supply ofblueprints from MNCs to local suppliers, and interactionsamong personnel to transfer the tacit dimensions of technol-ogy creation (Guerrieri & Pietrobelli, 2006). The technologyand technical expertise acquired through manufacturing with-in a GVC, are transferred via the products manufactured forother multinationals and/or in the production for own de-signed and branded products. Taiwanese computer firms oftenparticipate in more than one GVC, and “leverage competencesacross chains” (Schmitz, 2006: 561).

In the Brazilian State of Espırito Santo, learning occurred ina relational chain, in which local small and medium sized firms(SME) benefited from interacting with larger firms, who actedas anchors for the local cluster. The process was fostered bythe activities of intermediary institutions—matching the inter-ests of small and large firms—and by the active role of the lo-cal government to give the authority and credibility for thesefirms to negotiate with large firms and create better linkagesand collaboration with SMEs (Villaschi, Cassiolato, & Las-tres, 2007).

When suppliers lack competences, there are also alternativepatterns: hierarchy, which is vertical integration and occurswhen are difficult to codify, and captive, buyer-driven 9 chainswhere small suppliers are dependent on larger, dominant buy-ers that can exert high levels of monitoring and control, andwhere transactions are easier to codify.

In captive chains, lead firms intervene actively in the learningprocesses of suppliers that lack competences. Their support isusually confined to a narrow range of tasks—for example,simple assembly. However, there is a risk of lock-ins becauselead firms do not sustain the development of strategic, corecapabilities. The case of the shoe industry in the Sinos Valleyin Brazil is an example of how inclusion in a GVC can facili-tate product and process upgrading, but prevent functionalupgrading, leaving firms dependent on a small number of pow-erful customers (Bazan & Navas-Aleman, 2004; Schmitz,2006). Local shoe suppliers in the Sinos Valley were discour-aged from engaging in design, marketing and sales becausethese were the core competences of the United States buyers,the leaders in the GVC. Brazilians have been members of foot-wear value chain mostly as producers, with their buyers keento maintain the status quo. Other empirical evidence on theBrazilian sports shoe sector shows that local suppliers havedeveloped the capability to adapt designs to local conditions(tropicaliz c�ao), but have not been involved by lead firms innew design development (Lemos & Palhano, 2003).

Over time, the direct involvement of United States buyers inassisting the product and process upgrading of Brazilian shoeproducers has diminished: in the 1980s, most support camefrom specialized United States technical staff, which weregradually replaced by local staff. The activity was moved toChina in the 1990s, because the risk of supplier failure in Bra-zil was much higher (Schmitz, 2006).

The Sinos Valley shoe industry provides insights on thelearning mechanisms that occur across (inter-) GVC. The func-tional upgrading in design, branding, and marketing, whichwas discouraged by the United States buyers, was achieved,and allowed the firms to sell to buyers in the domestic and

regional markets in Latin America (Bazan & Navas-Aleman,2004). A similar process of experience being transferred acrosschains occurred in the Mexican footwear sector, where produc-ers also began selling in the domestic market and in some otherparts of Latin America (Rabellotti, 1999).

At the opposite end of the typology is vertical integration,which is where the lead firm takes direct ownership of someof the operations in the chain and transactions are not easyto codify. This is similar to the case of intra-firm trade betweena trans-national company and its subsidiaries, and implies var-ious potential learning mechanisms analyzed in the literatureon foreign direct investment in LDCs, such as transfer of man-agement, skilled labor turnover, training of the local work-force, knowledge spillovers, and imitation (Barba Navaretti& Venables, 2004).

3. IS AND THEIR INTERACTION WITH GVC

GVC analysis is limited because of the lack of attention tothe institutional context within which local firms interactingin GVC are embedded. This limitation is highlighted in the lit-erature on Global Production Networks (GPN), which dealswith how actors in the various networks are embedded in dif-ferent places. This includes the geographical dimension fromthe national to the local scale (Ernst, 2002; Hess & Yeung,2006). The work of geographers and planners on local indus-trial agglomerations stresses the spatial embeddedness of tacitknowledge and the importance of tight interdependencies be-tween geographically clustered firms (Storper, 1995).

At the national level, the relevance of rules, values, and insti-tutions (e.g., financial system, corporate governance, educa-tion, and training system) that affect the character andevolution of industries and firms is highlighted in the literatureon types of capitalism (e.g., Berger & Dore, 1996). Especiallyremarkable among these rules and institutions are those “. . .elements and relationships which interact in the production,diffusion and use of new, and economically useful, knowl-edge. . . and are either located within or rooted inside the bor-ders of a nations state” (Lundvall, 1992). These institutionsand organizations can have profound effects on value chaingovernance and the innovation and learning strategies of firmsin developing countries. In this section we explore the role ofinnovation systems in the GVC-driven learning and innova-tion process in developing countries.

We focus on two aspects of innovation systems: 10 technol-ogy policies and technology organizations. Technology poli-cies cover aspects such as technology imports via licensingand FDI, and incentives for local R&D and training. Technol-ogy organizations are those bodies that provide services suchas MSTQ, R&D, training, and KIBS. This differs in industrialcountries where the emphasis is much more on basic researchand the creation of new knowledge. These organizations canbe government-run, started by government but run autono-mously, or started and managed by industry associations, orprivate actors. In developing countries, government-run orga-nizations often play leading roles to counter the weaknesses inand precariousness of the private productive sector.

Many services provided by these organizations are the essen-tial “public goods” of technological effort. Public researchinstitutes and universities undertake basic research that doesnot yield commercial results in the short term, but providesa long-term base of knowledge for enterprise effort. Quality,standards, and metrology institutions provide the basic frame-work for firms to communicate on technology and maintainbasic standards for the industry. Extension services alleviate


informational, technical, equipment, and other SME handi-caps. KIBS are consultancy services related to technical andorganizational issues and, so far, their investigation is confinedto the developed country context; 11 however, as Schmitz andStrambach (2009) show, they are become increasingly relevantin some developing countries. The provision of these servicesremedies market failures that all governments face, regardlessof the national level of development.

How do different IS affect the determinants of GVC gover-nance—and through this, the opportunity for enterprise learn-ing and upgrading? The relationship between the form ofgovernance and the nature of the system is intrinsically dy-namic and cannot be univocal (one-to-one), given the varietyof possible systems and the endogeneity of most of the eventsoutlined above, and the frequent two-way causality and con-tinuous feed-back. 12 The nature of the IS affects the rangeof possible modes of governance of value chains. Table 2 be-low shows the relationship between GVC governance andthe nature of the IS. We discuss how the latter impacts onthe three key determinants of governance: complexity of trans-actions, extent of codification, and suppliers’ capabilities. Thelast column in Table 2 presents some possible dynamic trajec-tories from the different patterns of GVC governance that mayemerge from a well functioning IS.

(b) Complexity of transactions and IS

A well-structured and efficient IS can help to reduce thecomplexity of transactions and enable transactions based onarms’ length or on weak hierarchical forms of GVC gover-nance—the risk of falling into a captive relationship, or beingacquired by a leader is diminished. In other words, the lowerthe complexity of the transactions the less need there is foran effective IS—but an effective system increases the capabili-ties to cope with complex transactions.

When investors engage in make-or-buy decisions, they facea trade-off between lower production costs and higher

Table 2. GVC and their interact

Governance type Determinants

1 Market Low complexityHigh codification MSTQ organization

High supplier competence Education, training2 Modular High complexity

High codification MSTQ organizationHigh supplier competence Education, training

3 Relational High complexity “Local” systems anknowledge matter

Low codification MSTQ are perhapsHigh supplier competence Education, training

4 Captive High complexityHigh codification MSTQ organizationLow supplier competence

5 Hierarchy High complexity Local R&D organizfrom interaction

Low codificationLow supplier competence GVC is expected to

technical skills

Source: Authors’ elaboration.

transaction costs. In countries with weak institutions, weakcontract enforcement, pervasive corruption, cumbersomebureaucratic procedures, multiple barriers to trade and poorinfrastructures, it is difficult to capitalize on the benefits ofinter-firm specialization (Altenburg, 2006).

The weaker the institutional framework, the costlier andriskier will be contract enforcement, inter-firm coordinationand transactions will be more difficult, which favors nonmar-ket forms of governance, possibly vertical integration. The re-lated bureaucratic procedures and high administrative costs ofregistration may exclude small firms from doing business,“emerging” out of informality, and linking up with globaland national value chains.

In terms of science and technology, if the system offers effi-cient and homogeneous standards, testing, and quality assur-ance institutions and organizations, the costs of technologyand learning-related transactions will be lower, and relationalforms of governance will be smoother. Local firms’ learning incaptive value chains may extend beyond simple tasks to designand planning of activities for example, the experience of theindustrial and technological development of Taiwanese firmsand clusters is an insightful example of an IS supporting thetransition from hierarchy and captive chains led by foreignleaders, to local innovation, functional upgrading, and domes-tic firm-led value chains. Taiwan’s IS strengthened over timethanks to substantial investments in human capital and scien-tific and technological research, institutions and rules reward-ing innovation, and organizations such as science andtechnology parks that further facilitated efficient inter-firmand university-industry collaborations in high-tech activities(Guerrieri, Iammarino, & Pietrobelli, 2001; Saxenian & Hsu,2001; Tsai & Wang, 2005; Lee & Yang, 2000).

The establishment of relational value chains is also facili-tated by a well functioning IS with active technical bodies,where the chain leaders and their local partners can meet to ex-change complementary knowledge and to reduce the complex-ity of transactions. The development of specialized technical

ion with innovation systems

Innovation systems

s matter A well-structured, complete, smooth systemmakes 1–2–3 more likely to occur

organizations matter 4–5 may prevail also with “poorer,”fragmented systems. The chain leader maycompensate system weaknesses, butupgrading is restricted

s matter Possible dynamicsorganizations matter "

d complementary� From 5 and 4 to 2: Thanks to improvementin MSTQ� From 5 and 4 to 3: Thanks to improvementin “local” systems� From 5 and 4 to 2 and 3: Thanks to IS sup-porting the co-evolution of suppliers andGVC competences

less crucialorganizations matter

s matter

ations may benefit

improve human


institutions is more common in local clusters, where they sup-port local generation of innovative processes and practices(Bell & Albu, 1999). Several authors show that agglomerationsare associated with the relational portions of GC (Schmitz,2004; Sturgeon, 2003). The existence of these supportingbodies may be an attraction for GVC, promoting relationalforms of governance and enabling the transition from hierar-chical or captive chains to relational chains (last column inTable 2 the shift from 5 and 4 to 3).

(b) Codification of transactions and IS

In market-based transactions within efficient markets, all therelevant information is conveyed by the market price becausethe complexity of transactions is low. However, when com-plexity increases, enterprises in developing countries are unli-kely to have the internal skills and capabilities to operatewithin a context of codified transactions. The IS can simplifytheir efforts and enhance their effectiveness, especially throughthe MSTQ infrastructure.

MSTQ institutions form the basic infrastructure for nationaltechnological activities. The use of recognized standards andtheir certification by internationally accredited bodies orGVC leaders, is increasingly demanded in world trade. 13 Stan-dards can reduce transaction costs and information asymme-tries between seller and buyer, and minimize the uncertaintieswith respect to quality and technical characteristics.

The importance of industrial standards has increased andthey make a major contribution to the diffusion of technologywithin and across industries. In developing countries, stan-dards organizations disseminate best practices in an industryby encouraging and helping firms to understand and applynew standards, which is likely in turn to improve suppliers’competences. Redundant experimentation with new technolo-gies is reduced, and enterprises are introduced to a commonlanguage that is shared across the international market. Thisreduces the complexity of inter-firm technical linkages and col-laboration.

The existence of well-structured MSTQ institutions andorganizations has important implications for GVC, for theirgovernance and for developing countries’ innovation andtechnology systems. They make the handling of complextransactions and the organization of the GVC web of localrelationships easier. In principle, modular and relationalchains are more likely to prevail, provided that local suppliersare competent, and understand and use technical codes andstandards. The choice of either form may depend on the differ-ent degrees of knowledge codifiability.

Standards matter increasingly for natural resource-basedactivities. In Southern Chile a very successful salmon clusterhas been developed since the early 1990s, and in this clusterthe process of standards setting and compliance offers remark-able insights (Katz, 2006; Maggi, 2007). Compliance withinternational standards has allowed the Chilean salmon indus-try to progress from passive to active learning, with moreinvolvement of local firms as value chain leaders, and suppliersin foreign-led chains (Iizuka, 2009). A meso-level institution,the Association of (Chilean) Salmon Industries, played a cru-cial role in this process.

An explicit account of the dynamics involved allows a betterunderstanding of the implications of different systems on GVCgovernance and the opportunities for learning: better MSTQorganizations will enhance the probability of a transition fromhierarchical and captive value chains to modular forms of gov-ernance (last column in Table 2, the shift from 5 and 4 to 2).

(c) Supplier competence and IS

The IS includes all the institutions and organizations thatcontribute to improving suppliers’ competences. They consistof the organizations responsible for education and technicaltraining, and the set of incentives that induce individuals to in-vest in improving their knowledge and competence. As suppli-ers learn and acquire greater competence, GVC governance islikely to change. In very general terms, we would expect thatincreased capabilities in the supply-base will help to push thearchitecture of GVC away from hierarchical and captive net-works toward more relational and modular chains (Gereffi etal., 2005). However, better capabilities among suppliers arealso likely to have affect on the prevailing mode of value chaingovernance, and ceteris paribus, enhance learning mechanismswithin all value chains, allowing suppliers to benefit more fromparticipation in a value chain.

A co-evolution of suppliers and GVC leaders can be envis-aged since if suppliers acquire new competences, then the chainleaders, often buyers, need to change and adapt their core com-petences to the new governance patterns (Sturgeon & Lee,2001).In order to support and fit in with the acquisition ofnew competences by suppliers, chain leaders will need to receivesome benefits from these developments (Humphrey, 2006).

The case of the wine industry in South Africa, reported inPonte and Ewert (2009) is an example. The main foreign mar-ket for South African wine is the UK and the way that theGVC is organized has undergone profound changes. Underpressure for shorter lead times, UK agents and marketers havehad to increase their control over logistics (some importers areselling to retailers on the basis of delivery from the UK ware-house rather than “free-on-board” export from Cape Town).At the same time, UK agents and marketers have increasedtheir role in product innovation, new packaging, new presen-tations and styles, while retailers are increasingly becomingshelf-space providers. South African producers’ cellars haveimproved in terms of guaranteed wine quality, and improvedcapability to innovate production in response to consumermarket demand.

In the electronics sector, value chain leaders are happy tooutsource increasing amounts of production, including pro-cess-related design, to suppliers. In the two electronics clustersof Jalisco (Mexico) and Penang (Malaysia) analyzed by Rasi-ah (2007), chain leaders encouraged and supported the devel-opment of local technical competences. Local human capitaland suppliers’ competences, and the specific differentiationand division of labor that emerged in Penang and Jalisco, haveallowed remarkable integration with multinational corpora-tions and GVC.

Although, initially this generated improved economic andexport performance, the lack of technical and R&D scientistsand engineers, combined with relatively underdeveloped high-tech and R&D infrastructures in Malaysia and Mexico, haveundermined the capacity of multinational corporations and lo-cal firms to achieve functional integration. Thus, this has notresulted in the horizontal integration necessary to enter highervalue-added segments in value chains.

The cases of the electronics clusters in Malaysia and Mexicopoint to the difficulties involved in upgrading to high value-added functions. However, there are some examples of successin emerging countries (Schmitz and Strambach, 2009). Themost obvious one is software in Bangalore, but there are inter-esting cases in very different sectors such as the wine industryin Chile. Cusmano, Morrison, and Rabellotti (forthcoming)describe one of the main competitive factors in Chilean wineproduction, which has a well integrated international GVC,


is the strong link with university research, which has allowedquality upgrade and scaling up in the global wine market. An-other example is Tesco in Thailand, which has developed a no-vel, low build-cost store format that is essentially a smallhypermarket core surrounded by a local fresh food vendormarket (leased space) and a farming supplies area, in an at-tempt to circumvent the threat of tightened development con-trol in low-income provincial “up country” towns whereconventional large-format hypermarket development is con-sidered infeasible politically and unviable commercially (Coe& Wrigley, 2007).

(d) Learning across different chains (IS can help)

There are some significant learning mechanisms in differentvalue chains. In Taiwan in the 1990s, Taiwanese firms embed-ded in a developed IS, were frequently participating in morethan one GVC (Guerrieri & Pietrobelli, 2006), and leveragedcompetences across chains (Schmitz, 2006). The same thinghappened in the Sinos Valley in Brazil, where suppliers learnedand employed various competences through working with twoor more value chains (Bazan & Navas-Aleman, 2004).

Public policy can support diversification of value chains andlearning across chains. For example, an information-bargain-ing organization to identify emerging/promising markets andvalue chain leaders could help by holding information andmotivation events, subcontracting exchange schemes, and sup-plier fairs and exhibitions (Altenburg, 2006).

We have described in this section the multiply ways that ISinteract with GVC governance, and suppliers’ learning andinnovation. We discussed some possible forms of interactionand mutual effects, issues that further analytical and empiricalresearch could throw further light on. Most importantly, fu-ture research could explore systematically the dynamics ofGVC and the co-evolution of suppliers and buyers, and of re-lated IS.

4. CONCLUSIONS

Questions have been raised about whether the spatialembeddedness of learning and knowledge creation could bechallenged by alternative organizational forms. 14 Accordingto this view, organizational or relational proximity is moreimportant than geographical proximity to support the produc-tion, identification, appropriation and flow of tacit knowledge.Thus, multinational firms and GVC with dispersed, but care-fully organized knowledge bases and sites of innovation evenin developing countries, and use of “communities of practice,”could compensate for lack of geographical proximity.

This paper shows that IS interact with GVC in multipleways, and influence whether and how developing country

firms learn and innovate through entering and interacting inthese value chains. The relational proximity created withinGVC does not replace, but rather interacts with IS.

The first main conclusion is that the different characteristicsof value chains have an impact on the mechanisms of learningprevailing in the chain. In general, LDC firms learn and inno-vate based on their participation in the GVC because theyhave to satisfy the product quality, delivery time, process effi-ciency, environmental, labor and social standards require-ments of these chains. The learning mechanisms within GCVvary according to the form of governance that is adopted: theycan be the result of pressure to match international standardsor may be facilitated by direct involvement of the value chainleaders if the competence of suppliers is low and the risk non-compliance is high. When the actors in the value chain havecomplementary competences, learning is mutual and basedon intense face-to-face interactions.

The second conclusion of this paper is related to the multipleforms of interaction between IS, GVC governance and suppli-ers’ learning and innovation. On the basis of our model, wewould expect a well-structured and efficient IS would help toreduce the complexity of transactions, enabling arms’ lengthtransactions and weaker hierarchical forms of GVC gover-nance. In other words, the risk of falling into a captive rela-tionship, or being acquired by a leader, diminishes with astronger IS. The less complex the transactions the less is theneed for an effective IS, but an effective system also increasesthe capabilities to cope with complex transactions. The systemof organizations in charge of MSTQ plays a central role, andmay influence the form of governance adopted for developingcountry firms.

The third conclusion is that internal governance of the GVCis dynamic and subject to continuous adjustments andchanges. We have explored some of these changes. Future re-search should consider the dynamism of innovation and ex-plore systematically the co-evolution of suppliers and buyers,and the related IS. There is a large body of research showingthat buyers have evolving strategies toward their supply chains(Sturgeon & Lee, 2001), and their strategies can differ, forexample, in terms of knowledge transfer—some welcomeopportunities to transfer parts of their activities to the supplychain, others obstruct this. The type of IS that prevails locallywill affect this co-evolution.

Numerous avenues for further research are opened by thisstudy. More quantitative analyzes of value chains, their formsof governance, and their impact on local firms are needed. 15

More analysis of IS in developing countries is also needed, withspecific emphasis on the features highlighted in this paper.Studies of the dynamics of GVC, and the policy strategies ofdeveloping country governments and suppliers are also re-quired.

NOTES

1. For example, “learning from exporting” (Wagner, 2007), ForeignDirect Investments (FDI) and spillovers and imitation (Barba Navaretti &Venables, 2004; UNCTAD, various years).

2. See, for example, Gereffi (1994), Gereffi (1999), Giuliani, Pietrobelli,and Rabellotti (2005), Kaplinsky (2000), Humphrey and Schmitz(2002a,b), Pietrobelli and Rabellotti (2007).

3. On IS see Freeman (1995), Lundvall (1992), Nelson (1993), Metcalfe(1995) and Edquist (1997).

4. On IS in developing countries see Arocena and Sutz (1999), Cassio-lato, Lastres, and Maciel (2003), Edquist (2001); Intarakamnerd, Chaira-tana, and Tangchitpiboon (2002) and Lundvall, Joseph, Chaminade, andVang (2009).

5. However, we should stress that in LDC there are huge differences ininnovation and technological capacity. A small number of developingcountries have begun to make the difficult transition from beingeconomically successful in industrial production to building up innovationcapabilities (Schmitz & Strambach, 2009). Thus, China and India, or some


parts of these countries, have very similar IS to those in developedcountries and in some sectors are world class standard (Altenburg,Schmitz, & Stamm, 2008).

6. This view of technology system as open and deeply embedded inglobal flows of knowledge and technology is shared by scholars such asErnst (2002), who believes that IS theory fails to address the disruptivechanges imposed by globalization on the geography of IS.

7. www.globalvaluechains.org provides a synthetic and clear presenta-tion of these concepts.

8. We thank an anonymous referee for pointing this out.

9. Gereffi (1999) introduced the useful distinction between buyer-drivenchains, dominated by large retailers, branded marketers, and brandedmanufacturers and producer-driven chains in which large, usuallytransnational, manufacturers play the central roles in coordinatingproduction networks.

10. Although we argued in Section 2.1 that technology system is themore accurate term to describe systems in developing countries, forsimplicity we use the term IS.

11. The literature on KIBS is vast. Among others, there are Miles (2005),Strambach (2008) and Wood (2002).

12. To simplify the discussion, there is the temptation to classifyinnovation systems along a linear dimension, from “good” to “bad”

systems. Nevertheless, non-linearity, and idiosyncrasies are especiallyrelevant and frequent in this analysis. Clearly, there is simply not a singlebest way to organize an innovation system and a system which can begood in a certain point in time, it will not be so in a different moment,given the intrinsic dynamism in the innovation process.

13. Standards are the set of technical specifications that become the rulesand guidelines and describe the characteristics of products, services,processes and materials. Metrology (the science of measurement) providesthe measurement accuracy and calibration required for standards to beapplied. The application of standards and the certification of productsnecessarily imply (accredited) testing and quality control services. TheInternational Standards Organisation (ISO) has introduced the bestknown quality management (not technical) standards in use today: theISO 9000 series. ISO 9000 certification has become a requirement forpotential exporters, and signals quality and reliability to foreign buyers,value chain leaders, and transnational corporations seeking local partnersand subcontractors.

14. See Asheim and Gertler (2005) for a discussion.

15. See Pietrobelli and Saliola (2008) for a recent attempt to develop amethod to measure GVC governance.

REFERENCES

Albornoz, F., Milesi, D., & Yoguel, G. (2002). New Economy in OldSectors: Some issues coming from two production networks inArgentine. Paper Presented at the DRUID Summer Conference.<www.druid.dk>.

Altenburg, T. (2006). Governance patterns in value chains and theirdevelopment impact. The European Journal of Development Research,18(4), 498–521.

Altenburg, T., Schmitz, H., & Stamm, A. (2008). Breakthrough? China’sand India’s transition from production to innovation. World Develop-ment, 36(2), 325–344.

Arocena, R., & Sutz, J. (1999). Looking at national innovation systemsfrom the South. Industry and Innovation, 7, 55–75.

Asheim, B. T., & Gertler, M. S. (2005). The geography of innovation.Regional innovation systems. In J. Fagerberg, D. C. Mowery, & R. R.Nelson (Eds.), The Oxford handbook of innovation (pp. 291–317).Oxford: Oxford University Press.

Asheim, B. T., & Herstad, S. J. (2005). Regional innovation systems,varieties of capitalism and non-local relations: Challenges from theglobalising economy. In R. Boschma, & R. Kloosterman (Eds.),Learning from clusters: A critical assessment from an economic-geographical perspective (pp. 169–201). Dordrecht: Springer.

Barba Navaretti, G., & Venables, A. J. (2004). Multinational firms in theworld economy. Princeton: Princeton University Press.

Bazan, L., & Navas-Aleman, L. (2004). The underground revolution in theSinos Valley: A comparison of upgrading in global and national valuechain. In H. Schmitz (Ed.), Local enterprises in the global economy: Issuesof governance and upgrading (pp. 110–139). Cheltenham: Edward Elgar.

Bell, M. (2007). Technological learning and the development of productionand Innovative Capacities in the Industry and Infrastructure Sectors ofthe least developed Countries: What Role for ODA?, Background PaperNo. 10 for The Least Developed Countries Report. Geneva: UNCTAD.

Berger, S., & Dore, R. (Eds.) (1996). National diversity and globalcapitalism. Ithaca, NY: Cornell University Press.

Bell, R. M., & Albu, M. (1999). Knowledge systems and technologicaldynamism in industrial clusters in developing countries. WorldDevelopment, 27(9), 1715–1734.

Bunnell, T. G., & Coe, N. M. (2001). Spaces and scales of innovation.Progress in Human Geography, 25, 569–589.

Cassiolato, J. E., Lastres, H., & Maciel, M. (Eds.) (2003). Systems ofinnovation and development: Evidence from Brazil. Cheltenham: Elgar.

Carlsson, B. (2006). Internationalization of innovation systems: A surveyof the literature. Research Policy, 35, 56–67.

Edquist, C. (Ed.) (1997). Systems of innovation: Technologies, institutionsand organisations. London: Pinter.

Coe, N. M., & Wrigley, N. (2007). Host economy impact of transnationalretail: The research agenda. The Journal of Economic Geography, 1–31.

Cusmano, L., Morrison, A., & Rabellotti, R. (forthcoming). Catching-uptrajectories in the wine sector: A comparative study of Chile, Italy andSouth Africa. World Development.

Dutrenit, G., Vera-Cruz, A. O., & Gil, J. L. (2002). Desafıos y oportunidadesde las Pymes para su integracion a la red de proveedores: el caso de lamaquila automotriz en ciudad Juarez. Nota Tecnica No. 20, Instituto deEconomıa, Universidad Federal do Rıo de Janeiro.

Edquist, C. (2001). Systems of innovation for development. Backgroundpaper for UNIDO World Industrial Development Report 2002/3.Vienna: UNIDO.

Ernst, D. (2002). Global production networks and the changing geogra-phy of innovation systems. Implications for developing countries.Economics of Innovation and New Technology, 11(6), 497–523.

Freeman, C. (1995). The ‘National System of Innovation’ in historicalperspective. Cambridge Journal of Economics, 19(1), 5–24.

Fromhold-Eisebith, M. (2007). Bridging scales in innovation policies:How to link regional, national and international innovation systems.European Planning Studies, 15(2), 217–233.

Gereffi, G. (1994). The organization of buyer-driven global commoditychains: How US retailers shape overseas production networks. In G.Gereffi, & M. Korzeniewicz (Eds.), Commodity chains and globalcapitalism. London: Praeger.

Gereffi, G. (1999). International trade and industrial upgrading inthe apparel commodity chain. Journal of International Economics, 48,37–70.

Gereffi, G., Humphrey, J., & Sturgeon, T. (2005). The governance ofglobal value chains. Review of International Political Economy, 12(1),78–104.

Guerrieri, P., Iammarino, S., & Pietrobelli, C. (Eds.) (2001). The globalchallenge to industrial districts: SMEs in Italy and Taiwan. Cheltenham,UK and Lyme, US: Edward Elgar.

Giuliani, E., Pietrobelli, C., & Rabellotti, R. (2005). Upgrading in globalvalue chains: Lessons from Latin America clusters. World Develop-ment, 33(4), 549–573.

http://www.globalvaluechains.org


Guerrieri, P., & Pietrobelli, C. (2006). Old and new forms of clustering andproduction networks in changing technological regimes: Contrastingevidence from Taiwan and Italy. Science Technology and Society, 11(1),9–38.

Hess, M., & Yeung, H. (2006). Whither global production networks ineconomic geography?. Environment and Planning A, 38, 1193–1204.

Humphrey, J. (2006). Policy implications and trends in agribusinessvalue chains. The European Journal of Development Research, 18(4),572–592.

Humphrey, J., & Schmitz, H. (2002b). Developing country firms in theworld economy: Governance and upgrading in global value chains,INEF Report, No. 61, University of Duisburg, Duisburg.

Humphrey, J., & Schmitz, H. (2002a). How does insertion in global valuechains affect upgrading industrial clusters?. Regional Studies, 36(9),1017–1027.

Iizuka, M. (2009). Low tech industry: A new path for development? Thecase of the salmon farming industry in Chile. In F. Malerba, & S. Mani(Eds.), Sectoral systems of innovation and production in developingcountries: Actors, structure and evolution (pp. 232–257). Cheltenham:Edward Elgar.

Intarakamnerd, P., Chairatana, P., & Tangchitpiboon, T. (2002). Nationalinnovation system in less successful developing countries: The case ofThailand. Research Policy, 31, 1445–1457.

Kaplinsky, R. (2000). Globalisation and unequalisation: What can belearned from value chain analysis?. Journal of Development Studies,37(2), 117–146.

Katz, J. (2006). Salmon farming in Chile. In V. Chandra (Ed.),Technology, adaptation and exports. How some developing countriesgot it right (pp. 193–224). Washington DC: The World Bank.

Kishimoto, C. (2004). Clustering and upgrading in global value chains:The Taiwanese personal computer industry. In H. Schmitz (Ed.), Localenterprises in the global economy: Issues of governance and upgrading(pp. 233–264). Cheltenham: Edward Elgar.

Lall, S., & Pietrobelli, C. (2002). Failing to compete: Technologydevelopment and technology systems in Africa. Cheltenham: EdwardElgar.

Lall, S., & Pietrobelli, C. (2003). Manufacturing in Sub-Saharan Africaand the need of a national technology system. In M. Muchie, B. A.Lundvall, & P. Gammeltoft (Eds.), The making of African innovationsystems. Aalborg: Aalborg University Press.

Lall, S., & Pietrobelli, C. (2005). National technology systems in Sub-Saharan Africa. International Journal of Technology and Globalisation,1(3–4), 311–342.

Lundvall, B.-A., Joseph, K. J., Chaminade, C., & Vang, J. (Eds.) (2009).Handbook of innovation systems and developing countries. Buildingdomestic capabilities in a global setting. Cheltenham: Elgar.

Lee, W.-H., & Yang, W.-T. (2000). The cradle of Taiwan High-technologyindustry development—Hsinchu Science Park. Technovation, 20(1),55–59.

Wood, P. (Ed.) (2002). Consultancy and innovation: The business servicerevolution in Europe. London: Routledge.

Lemos, C., & Palhano, A. (2003). Clustering in a backward region: Thefootwear productive system in Campina Grande. In J. E. Cassiolato,H. Lastres, & M. Maciel (Eds.), Systems of innovation and development:Evidence from Brazil (pp. 331–351). Cheltenham: Elgar.

Lundvall, B. A. (1992). National systems of innovation: Towards a theory ofinnovation and interactive learning. London: Pinter.

Maggi, C. (2007). The Salmon Farming and Processing Cluster inSouthern Chile. In C. Pietrobelli, & R. Rabellotti (Eds.), Upgrading tocompete: SMEs, clusters and value chains in Latin America(pp. 109–142). Cambridge Mass: Harvard University Press.

Metcalfe, S. (1995). Technological system and technological policy in anevolutionary framework. Cambridge Journal of Economics, 19, 25–46.

Miles, I. (2005). Innovation in services. In J. Fagerberg, D. C. Mowery, &R. R. Nelson (Eds.), The Oxford handbook of innovation (pp. 433–458).New York: Oxford Innovation Press.

Nelson, R. R. (Ed.) (1993). National innovation systems: A comparativeanalysis. New York: Oxford University Press.

Morrison, A., Pietrobelli, C., & Rabellotti, R. (2008). Global value chainsand technological capabilities: A framework to study industrialinnovation in developing countries. Oxford Development Studies,36(1), 39–58.

Pietrobelli, C. (2008). Global value chains in the least developed countriesof the world: Threats and opportunities for local producers. Interna-tional Journal of Technological Learning, Innovation and Development,1(4), 459–481.

Pietrobelli, C., & Rabellotti, R. (Eds.) (2007). Upgrading to compete:SMEs, clusters and value chains in Latin America. Cambridge Mass:Harvard University Press.

Pietrobelli, C., & Saliola, F. (2008). Power relationships along the valuechain: Multinational firms, global buyers, and local suppliers’ perfor-mance. Cambridge Journal of Economics, 32(6), 947–962.

Ponte, S., & Ewert, J. (2009). Which way is ‘up’ in upgrading? Trajectoriesof change in the value chain for South African wine. WorldDevelopment, 37(10), 1637–1650.

Quadros, R. (2004). Global quality standards, and technological upgrad-ing in the Brazilian auto-components industry. In H. Schmitz (Ed.),Local enterprises in the global economy: Issues of governance andupgrading (pp. 265–296). Cheltenham: Edward Elgar.

Rabellotti, R. (1999). Recovery of a Mexican cluster: Devaluation bonanzaor collective efficiency?. World Development, 27(9), 1571–1585.

Rasiah, R. (2007). Clusters and regional industrial synergies. TheElecetronics Industry in Penang and Jalisco. In A. J. Scott, & G.Garofoli (Eds.), Development on the ground, clusters, networks andregions in emerging economies (pp. 223–250). London and New York:Routledge.

Sako, M., & Helper, S. (1998). Determinants of trust in supplier relations:Evidence from the automotive industry in Japan and the United States.Journal of Economic Behavior & Organization, 34, 387–417.

Saxenian, A., & Hsu, J. Y. (2001). The silicon Valley-Hsinchu connection:Technical communities and industrial upgrading. Industrial andCorporate Change, 10(4), 893–920.

Schmitz, H. (Ed.) (2004). Local enterprises in the global economy: Issues ofgovernance and upgrading. Cheltenham: Edward Elgar.

Schmitz, H. (2006). Learning and earning in global garment and footwearchains. The European Journal of Development Research, 18(4), 546–571.

Schmitz, H., & Strambach, S. (2009). The organizational decomposition ofthe innovation and global distribution of innovative activities: Insightsand research agenda. International Journal of Technological Learning,Innovation and Development, 2(4), 231–249.

Storper, M. (1995). The resurgence of regional economies, ten years later.European Urban and Regional Studies, 2(3), 191–221.

Strambach, S. (2008). Knowledge-Intensive Business Services (KIBS) asdrivers of multi-level knowledge dynamics. International Journal ofService and Technology Management, 10(2–4), 152–174.

Sturgeon, T. (2003). What really goes on in silicon valley? Spatialclustering dispersal in modular production networks. Journal ofEconomic Geography, 3, 199–225.

Sturgeon, T., & Lee, J. -R. (2001). Industry co-evolution and the rise of ashared supply-base for electronics manufacturing. Paper Presented atNelson and Winter Conference, Aalborg, June.

Tewari, M. (1999). Successful adjustment in Indian industry: The case ofLudhiana’s Woollen Knitwear cluster. World Development, 27(9),1651–1672.

Tsai, K. H., & Wang, J. C. (2005). An examination of Taiwan’s innovationpolicy measures and their effects. International Journal of Technologyand Globalisation, 1(2), 239–257.

UNCTAD, various years. World Investment Report. Geneva: UnitedNations.

Villaschi, A., Cassiolato, J. E., & Lastres, H. (2007). Local production andinnovation systems in Brazil: The metalworking cluster in EspiritoSanto. In C. Pietrobelli, & R. Rabellotti (Eds.), Upgrading to compete:SMEs, clusters and value chains in Latin America (pp. 175–190).Cambridge Mass: Harvard University Press.

Wagner, J. (2007). Exports and productivity: A survey of evidence fromfirm-level data. The World Economy, 30, 60–82.

Available online at www.sciencedirect.com

Documents

Global Value Chains Meet Innovation Systems: Are There Learning Opportunities for Developing Countries?