INNOVATION IN DEVELOPING COUNTRIES

Yesim Sungu-Eryilmaz University of Pittsburgh

SUNGU-ERYILMAZ 92 PROJECTIONS 7 INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

LOCAL AND NON-LOCAL GEOGRAPHY OF TECHNOLOGICAL INNOVATIONS IN DEVELOPING COUNTRIES

ABSTRACT

In recent years, two subjects have received increasing attention from both researchers and policy makers in the industrial and regional development arena: technological innovation, i.e., developing new or improved products or processes, and the interactive or network model of an innovation process, i.e., ties with other organizations such as suppliers, universities, or customers. Especially, it is argued that local ties play an important role in facilitating knowledge exchange among fi rms and local organizations, which in turn facilitate innovation. Consequently, the majority of regional studies tend to focus on fi nding data at the local level, neglecting the importance of non-local networks. While models of local networking have emerged in the policy agendas of developing countries, these models fail to identify potential benefi ts that peripheral regions could gather from non-local (inter-regional and international) linkages. This paper argues that empirical and policy research should take into account the mixed networks of local and non-local ties. Both types of ties play an important role in building innovative capability in developing economies. It would make more sense, analytically as well as politically, to distinguish between diff erent types of networks and their role in building innovative capability in developing countries.

SUNGU-ERYILMAZ 93

PROJECTIONS 7 INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

SUNGU-ERYILMAZ

94 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT INTRODUCTION

In recent years, two subjects have received increasing attention from both researchers and policy makers in the industrial and regional development arena: technological innovation, i.e., developing new or improved products or production processes; and the interactive or network model of an innovation process, i.e., knowledge exchange with suppliers, universities, customers and other organizations during the product or process development. Especially, local network ties are viewed to play an important role in facilitating knowledge exchange among fi rms and local organizations, which in turn facilitate innovation (Audretsch, 1998; Camagni, 1991; Storper, 1997). There is a large volume of empirical studies available that describes the innovation activities of fi rms, and confi rms the links between innovation and networking in developed countries. This is not true for developing countries where the characteristics and scope of the innovation processes and networking behaviors are still largely unknown.

This paper reviews the current literature; investigates the relevance of currently used innovation and network concepts in the context of developing countries; and proposes a framework to study innovation and innovation networks in developing countries. It argues that empirical and policy research should take into account incremental innovation and the mixed networks of local and non-local ties. Both types of ties play an important role in building innovative capability in developing economies. While models of local networking have emerged in the policy agendas of developing countries (Altenburg and Meyer-Stamer, 1999; Bell and Pavitt, 1992; Cooper, 1991), these models fail to identify potential benefi ts that fi rms and regions could gather from non-local (i.e., interregional and international) linkages. This weakness may hinder an eff ective implementation of innovation policies in regions of developing countries (Ernst, 2002).

The paper is divided into six sections. After the introduction, the next two sections examine the interactive innovation process and the local geography of innovation. Section four discusses the changing view of the innovation in developing countries. Section fi ve identifi es the gaps in the literature and discusses the role of non-local networks in innovation and its relevancy to developing countries. The last section introduces the framework to study innovation and innovation networks in developing countries.

INTERACTIVE PROCESS OF TECHNOLOGICAL INNOVATION

Technological innovation is today defi ned as the introduction of new or improved products or production processes, or the introduction of new organizational structures, that have direct or indirect economic impact (Schumpeter, 1934). Examples include a

new organizational structure that makes a fi rm more competitive in domestic or international markets, an improved production process that decreases the production costs of an existing product, or the introduction of a new product that creates a new branch of an industry. Economists have always recognized the importance of innovation in economic development1. However, it was Schumpeter who fi rst defi ned innovation as used today and gave a central place to the role of innovation in his theory of economic development (Schumpeter, 1934). Schumpeter (1934) defi ned economic development as a qualitative change in the nature of production rather than quantitative changes in savings or investment patterns.

There have been two approaches to innovation process: linear and interactive models. In the 1950s and 1960s, linear models dominated the thinking about innovation. These models are called science-push and demand-pull. The science-push model assumed that innovation is a linear process and was thought of as a series of sequential steps leading directly from basic research, through to applied research, to development and commercialization2. This hierarchical approach emphasizing basic research became the principal model for innovation and science policies in the 1950s (Nelson, 1959). The demand-pull model, on the other hand, stressed the importance of the demand side and markets as the source of ideas for innovation (Schmookler, 1966). In this model, the emphasis shifts from researchers to users. Users defi ne the problems and ask researchers to conduct research specifi c to these problems (Weiss, 1979). In both views, innovation was defi ned as new machinery. The implication of this view was refl ected in ‘learning’ models. Technology mastery was achieved through ‘learning by doing’ (Arrow, 1962). Arrow (1962) argued that production costs decrease as productive experience increases. Learning was an automatic process.

By the mid 1980s and 1990s, the linear innovation model and learning type was questioned in the search of new ways of conceptualizing technological innovation. The linear model overemphasized research as the only source of innovation (Smith, 1994) and innovation policies were based on introducing internal, formal R&D-based products and processes. But these policies could not reach small and medium-sized enterprises (SMEs), which have relatively fewer fi nancial and human resources compared to larger fi rms. However, SMEs were able to develop new products or processes and they kept up with larger fi rms in the fi eld of innovation (Acs and Audretsch, 1988; B Noteboom, 1994; Rothwell, 1989). Research showed that linear conceptualization of the innovation process and R&D only represented a portion of the entire set of activities that fi rms had to take to innovate (Malecki and Oinas, 1999). Innovation did not have to be sequential

SUNGU-ERYILMAZ 95

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT PROJECTIONS 7

SUNGU-ERYILMAZ

96 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

and did not have to start from basic research in academia (Nelson and Winter, 1982) and feedback and trials were essential (Kline and Rosenberg, 1986).

Based on these criticisms, the interactive innovation model was developed. Interactive innovation is a non-linear and independent process (Kline and Rosenberg, 1986). It may stem from many sources, both inside and outside of the fi rm. Innovation is created and sustained by knowledge inputs generated not only within the fi rm, i.e., the feedbacks across all stages of the production chain (Kline and Rosenberg, 1986), but also knowledge inputs from suppliers, universities , competitors, or customers (B.-A. Lundvall, 1992).

In the interactive innovation models, knowledge is increasingly regarded as the critical source and ”at the core of production and innovation activities” (Archibugi & Michie, 1995). Two forms of knowledge are emphasized in the literature: codifi ed and tacit knowledge (Nonaka and Takeuchi, 1995; Nonako, Toyama, and Nagata, 2000). Codifi ed knowledge refers to knowledge that can be organized and codifi ed by its holder so that it can be easily saved and communicated. Codifi ed knowledge is transmittable in a systematic way (Nonaka and Takeuchi, 1995). Examples of codifi ed knowledge include software, databases, operating manuals, patents, best practices, and procedures. Some of this knowledge can be purchased in the market place. Tacit knowledge refers to intuitive, unarticulated, and implicit knowledge. In most of the innovation studies, tacit knowledge is identifi ed as an important component of the knowledge used in innovation (Dosi et al., 1988; Howells, 2002; Kline and Rosenberg, 1986). Tacit knowledge is embedded in a person or in organizational routines (Johnson and Lundvall, 2001). It is diffi cult to transfer, communicate, and assimilate (Cohen and Levinthal, 1990) since it is personal and context specifi c-temporal, spatial and social (Lam, 1998). Sharing tacit knowledge between individuals requires social interaction, shared understanding and trust (Gertler, 2003; Lam, 1998; B. Lundvall and Johnson, 1994; Maskell and Malmberg, 1999b; Storper, 1999). Therefore, it entails co-presence and co-location between the transmitter and the receiver ( Noteboom, 1999). Therefore, local geography is important to our understanding of knowledge sharing in the innovation process.

THE LOCAL GEOGRAPHY OF INTERACTIVE INNOVATION PROCESS

Diff erent local development models promoted the locality as the best level for the occurrence and diff usion of innovation. These models were labeled in a variety of ways in the literature: Industrial Districts (Brusco, 1982; Piore and Sabel, 1984; Pyke, Becattini, and Sengenberger, 1990), Innovative Milieu (Aydalot and Keeble, 1988; Camagni, 1991), New

Industrial Spaces (Saxenian, 1996; A J Scott, 1988), and Regional Innovation Systems (RIS) (Braczyk, Cooke, and Heidenreich, 1997; Kaufmann and Todtling, 2000; Morgan, 1997).

These local innovation models share common concepts that can be grouped under four headings. First is the emphasis on locality or endogenous development which emphasizes mobilization of the resources available locally. While diff erences in innovative capability among fi rms are in part attributable to their organizational capabilities (Cohen & Levinthal, 1990), it is in part attributable to properties of their local economies (Camagni, 1991; Maskell and Malmberg, 1999b; Porter, 1990; Storper, 1997). Local or endogenous resources include social, economic, technical, and political resources, such as regional entrepreneurship, human capital, existing industrial structure, R&D infrastructure, and the existence of professional associations. An essential characteristic of endogenous development is the broad involvement of local groups and individuals in the planning and policy process (Coff ey and Polese, 1984; Friedmann and Weaver, 1979; Moulaert and Sekia, 2003). The success of high technology clusters, such as Silicon Valley, and places making traditional products, such as Emilia-Romagna in Italy, emphasizes the use of local resources for competitiveness.

Second, all models acknowledge externalities associated with the spatial clustering of fi rms or agglomeration economies. Two viewpoints exist regarding externalities. One is localization economies, which refer to the agglomeration economies of similar industries (Brusco, 1982). The second is urbanization economies, which refl ect externalities associated with the presence of complementary fi rms and organizations in a variety of relevant industries and services (Harrison, et al., 1996; A. J. Scott, 1990). This is particularly relevant for some SMEs that undertake little R&D themselves, yet contribute considerable innovative activity in newly emerging industries such as biotech and computer software (Audretsch 1998). SMEs make use of universities, trade associations, and other knowledge-generating institutions. The knowledge spillover from the fi rm conducting R&D or the research lab of a university stimulates innovative capability in a region.

Third is the emphasis on networking. All territorial innovation models use network concepts as key characteristics. The industrial district literature stresses the role of personal relations, and networks of such relations in the innovation process. Similarly, innovative milieu theory argues that fi rms innovate through relationships with other agents of the same milieu. New Industrial Spaces argues for inter-fi rm transactions and a culture of networking and social interaction as the characteristics of new industrial spaces. Last,

SUNGU-ERYILMAZ 97

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT PROJECTIONS 7

SUNGU-ERYILMAZ

98 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

Regional Innovation System (RIS) sees the innovation and learning process as an interactive process and the network as an organizational mode of interactive learning.

Fourth, all territorial innovation models emphasize that fi rms are embedded in local networks. These local networks constitute a valuable resource in the conduct of economic activity and for innovation activities. Tacit knowledge in fi rms and organizations comes into existence in local networks of fi rms and organizations (Storper, 1997). Local networks are dense (Storper, 1997) and contain a diverse set of organizations including fi rms, customers, suppliers, universities, fi nancial organizations, research institutions and professional associations (Kaufmann and Todtling, 2000; Perrin, 1991). However, customers and suppliers are specifi cally emphasized. While a fl ow of incremental innovations is generated through localized interaction with customers (Von Hippel 1988), embodied technologies are imported into the fi rm through the exchanges with suppliers as knowledge spillovers (Audretsch, 1998).

These ties can be formal or informal, market or non-market (Torre & Gilly, 2000). Informal or personal relationships depend on trust, while formal relations are based on contractual agreements. The informality of interrelationships is viewed as being a potential strength rather than a weakness of local networks. Because of trust-based relationships, fi rms are willing to undertake risky co-operative and joint ventures and to act as a group (Gordon and McCann, 2000). The duration of networks is argued to be longer in local networks (Oerlemans, Meeus, and Boekema, 2000). It is argued that a long duration of a tie enhances mutual understanding and trust (Nooteboom & Gilsing, 2004). Tacit knowledge is best transferred via frequent face-to-face interaction and can best be managed in local proximity (Audretsch, 1998).

The basic assumption in the literature is that geographical distance aff ects the ability to receive and transfer knowledge. In general, innovation is assumed to be more dependent on local networks. This local networking pattern has been closely linked to the best examples, those which built their competitive advantage from localized learning. These include Silicon Valley and Emilia-Romagna in Italy. However, these examples are unique and non-transferable although policymakers try to do the same in developing countries.

CHANGING VIEW OF TECHNOLOGICAL INNOVATION IN DEVELOPING COUNTRIES During the 1950s and 1960s, there was little mention or interest in understanding innovation in developing countries, partly because innovation was assumed to be absent

in developing countries. This assumption was based on a notion of innovation at that time that was considered to be embodied in durable, capital goods, in other words, the development of new kinds of machinery (Solow, 1957). While the developed countries were the producers of new kinds of machinery, developing countries were the users of this machinery. Any problem about achieving technological change and economic growth for developing countries was largely seen as the acquisition and installation of new machinery which had already been developed elsewhere (innovation as technology diff usion). Therefore, the issue was viewed as generating level of savings (Domar, 1957) and capital accumulation (Solow, 1957) through international capital fl ows, that needed to acquire externally provided machinery. Local industry in developing countries was essentially seen as passive, involving only the adoption and routine operation of externally supplied technologies. Consequently, policy concerns about innovation tended to focus on the choice of appropriate technology, technology transfer, or fi nancial and informational gaps that hindered the fl ows of capital embodied technology

This view changed in the 1980s, partly because the defi nition of innovation changed. Innovation was not seen specifi cally as “new machinery” anymore. Subsequent research has extended Schumpeter’s theory and addressed the scope and scale of innovation. The basic dichotomy of product and process innovation informed the empirical research agenda on the scope of innovation. While product innovations are usually associated with the creation of new markets or the quality enhancement of existing products, process innovations are typically introduced to reduce costs and/or increase the fl exibility and performance of production processes (Edquist et al., 2001; Simonetti et al., 1995). Regarding the scale of innovation, radical (new) vs. incremental (improvement) innovation was also researched. According to several fi rm level empirical studies, both radical and incremental innovation proved to be important (Cohen and Levinthal, 1990; Nonaka and Takeuchi, 1995; Porter, 1990). Case studies in developing countries revealed that learning was not an automatic process. Learning was a conscious, systematic, and frequent eff ort made by the concerned actors (Bell and Pavitt, 1992; Cooper, 1991; L. Mytelka and Ernst, 1998; Westphal et al., 1984). While R&D existed only in small scale in developing countries, their innovative capabilities have been gained mainly through reverse engineering and shop-fl oor level incremental processes such as resolving production line bugs and suggesting product improvements (Amsden, 1989; Westphal et al., 1984). In addition, machinery suppliers and multinational corporations were not the only sources of technological change (Bell and Albu, 1999).

SUNGU-ERYILMAZ 99

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT PROJECTIONS 7

SUNGU-ERYILMAZ

100 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

Dosi et.al (1988) defi ned incremental and new innovations as follows: Incremental innovations occur more or less continuously in any industry or service activity although at diff ering rates in diff erent industries and countries, depending on a combination of demand pressures, socio-cultural factors, technological opportunities and trajectories… Although their combined eff ect is extremely important in the growth of productivity, no single incremental innovation has dramatic eff ects, and they may sometimes pass unnoticed and unrecorded. However, their eff ects are apparent in the steady growth of productivity.

New innovations are discontinuous events and usually the result of a deliberate R&D activity in enterprises and/or university and government laboratories… Whenever they may occur, they are important as the potential springboard for the growth of new markets and for the surges of new investment associated with booms. They may often involve a combined product, process, and organizational innovation. Over a period of decades radical innovations… may have fairly dramatic eff ects… but in terms of their aggregate economic impact they are relatively small and localized unless …… radical innovations are linked together in the rise of new industries and services, such as in the synthetic materials industry or the semiconductor industry.

Innovation is now seen as one part of the capabilities of developing countries. Four main categories of capabilities were defi ned for developing countries (Dahlman et al., 1987; Kim, 1997; Lall, 1992; Mytelka and Ernst, 1998; Westphal et al., 1984):

1 Production capabilities include the day-to-day, shop fl oor activities such as monitoring production and output quality.

2 Investment capabilities relate to the knowledge and skills needed to establish or extend production facilities. These capabilities include feasibility analysis, evaluation, and selection of technology as well as setting up equipment.

3 Innovative capabilities imply the capability to adapt, change or create technologies in response to changing needs (Kim, 1997). The term covers a wide range of activities including incremental vs. radical and product vs. processes and organizational innovation.

Incremental innovation involves the ability to adapt and improve existing products or processes. Incremental innovation is argued to be key for developing countries (Forbes and Weild, 2000; Kim, 1997). These innovations add value, especially if they are continuous (Evenson 1995; Dahlman 1987; Tushman and Nelson 1990). Developing countries in Asia, in particular, succeeded in developing considerable innovative capability and export success via incremental innovations such as imitation, reverse engineering, and resolving production line bugs (Bell and Pavitt, 1992; Ernst, et al., 1998; Kim, 1997; Nelson, 1993). New or radical innovation involves the ability to create new products or production processes and to develop patentable ideas (Ernst et al., 1998). It could be new for a fi rm or the market it serves. Even if a fi rm introduces a technique that is already used by other fi rms, this still represents a new innovation for that fi rm.

4 Marketing capabilities refer to the ability to understand user needs; to keep track of changing market demand; to create new markets; to establish distribution channels; and to provide customer services.

Forbes and Weild (2000) examined basic similarities and diff erences in the nature of innovative activities between developing and developed countries. They reached the following conclusions. First, incremental innovation is key for both developed and developing countries. “As developed countries continue to improve the technology, keeping up requires incremental innovation, and catching up requires incremental innovation at a faster pace than in the developed countries. Incremental innovation is thus the primary source of long-run competitiveness in developing countries” (Forbes & Weild, 2000, p.1099). Second, radical or new innovation can be a new technological paradigm for developed countries, but for developing countries this could be new to a fi rm. Third, both product and process innovation are important for developing and developed countries. However, product and process innovations are diff erent at diff erent stages in industrial development. Wong (1999) suggests that fi rms in developing countries deliberately choose to focus on either the product or process side while some fi rms may focus on both simultaneously. Last, shop-fl oor innovations, occurring in the day-to-day operations, contribute signifi cantly to the competitiveness of developing countries in cost-sensitive markets. However they are not captured by formal innovation indicators (Forbes and Weild, 2000).

NON-LOCAL GEOGRAPHY OF TECHNOLOGICAL INNOVATION

While the phenomenon of local networking has excited considerable interest in developing countries, there are also important weaknesses that need to be addressed

SUNGU-ERYILMAZ 101

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT PROJECTIONS 7

SUNGU-ERYILMAZ

102 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

to broaden the acceptance of network theory and to improve its policy relevance in developing regions. An important weakness is the neglect of the non-local dimension. The importance of proximate relationships may be overstated by failing to take into account non-local forms of networking (Alderman, 1999; Amin, 1999; Oinas, 1999; Hendry, 2000; Markusen, 1999; Malecki, 1999; Harrison, 1994; Simmie, 1998; Simmie, 1999; Amin, 1992; Amin, 1999; Ernst, 1999; Staber, 1996; Amin, 2005). Non-local network relationships have been mentioned by innovative milieu and in new industrial districts literature, which states that local systems are not self-contained, but are linked to the outside world by various sorts of connections. The role of these non-local connections in innovation, however, has not been emphasized. These counter debates and evidence can be grouped under three headings:

1 The dichotomy of knowledge: The network theory identifi es tacit knowledge as an important component of the knowledge used in innovation. Local networks are hypothesized to be of particular importance to innovation due to the production and diff usion of tacit knowledge, requiring proximate relations. However, research shows that the simple tacit vs. codifi ed dichotomy and its local and global implications are problematic (Bathelt, et al., 2004). Tacit and codifi ed knowledge are not alternatives but complements for competitive advantages at diff erent stages of a fi rm’s or product’s life cycle (Amin and Cohendet, 1999, 2005; Gertler, 2003; Howells, 2002; Lawson and Lorenz, 1999). Firms depend on diff erent knowledge types and adopt diff erent approaches to learning (Amin 1999; Nonaka 1995; Polanyi 1967). The relative importance of tacit vs. codifi ed knowledge and their role in learning and innovation can vary greatly between fi rms in diff erent societal contexts (Nonaka 1995). Moreover, accepting the superiority of tacit knowledge over codifi ed knowledge would come at the expense of denying not only the role of global codifi ed knowledge, but also of denying the role of local sources based on formal research, and the development eff orts within fi rms, universities, and research institutions (Amin and Cohendet, 1999). Malmberg and Maskell (2002) argue that in some cases the process explaining spatially concentrated innovation has less to do with tacit knowledge and more to do with local opportunities to share and monitor codifi ed knowledge ( Malmberg and Maskell, 2002).

Empirically, Lawson and Lorenz (1999) explored the relationship between codifi ed and tacit knowledge in the innovation process. They showed the importance of the regional capability in combining and integrating diverse knowledge, based on a case study of Minneapolis, US, and Cambridge, UK (Lawson and Lorenz, 1999). They observed that both tacit and codifi ed knowledge seems to be crucial for product development in both

regions. Similarly, in the case of developing countries, Ernst (2002) showed the need to blend diverse international and domestic sources of knowledge to compensate for initially weak national production and innovation systems (Ernst, 2002). He further argues that the key to success is to facilitate the concurrent leveraging of multiple and diverse sources of knowledge—the global production networks of buyers and suppliers of both foreign and domestic origins, as well as the diverse carriers of national innovation systems (Ernst, 2002).

2 Lock-in vs. stay tuned: Local networks are hypothesized to be of particular importance in enhancing interactive learning due to the frequent, face-to-face, and durable local ties. However, local networks may be harmful for interactive learning and innovation because they may create spatial lock-in situations. The lock-in situations occur when the local structures become so narrowly focused on a particular economic activity (technology or market organization and technology) that they are unable to shift to another development track (A. Malmberg and Maskell, 1997, p.38). Amin and Cohendant (1999) argued that business networks that are largely dependent on local tacit knowledge may be inadaptable in the face of radical shifts in markets and technologies. For example, research on Italian districts showed that they were not well-equipped to cope with radical changes in product or the technological trajectory, and their preference towards local tacit knowledge hindered districts’ performance (Amin and Cohendet, 1999). Similarly, Glasmeier (1999) argued that local networks may create a lock-in situation in small areas with a limited infl ow of external knowledge, a resistance to change and a delay in generating response to changing economic conditions (Glasmeier, 1999). In that case, spatial lock-in situations may be prevented by establishing non-local ties. Non-local ties help fi rms and organizations to stay tuned with what happens in the market, among producers (both competitors and collaborators), and among consumers (Britton, 2004). Of course, this requires that local fi rms have the capabilities to absorb non-local knowledge, which necessitates organizational proximity (Gertler, 2003; Malecki and Oinas, 1999; Oinas, 1999).

In the case of developing countries, local linkages may not be suffi cient (Ernst, 2002). This is because most newly industrializing countries and second-tier OECD countries have an incomplete set of domestic linkages (Sanjaya Lall, 1990, 2000; L. Mytelka and Ernst, 1998; L. K. Mytelka, 1999). Therefore, the ability of fi rms to select and connect to relevant local, regional or international ties becomes increasingly critical.

SUNGU-ERYILMAZ 103

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT PROJECTIONS 7

SUNGU-ERYILMAZ

104 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

3 Beyond locality: It is argued that most fi rms’ network ties – personal or business - are not only embedded within social relationships (Granovetter, 1985; Uzzi, 1997), but also embedded in their local environment (Braczyk et al., 1997; Brusco, 1982; Cooke, 2001; Storper, 1997). However, some research argues that many local economic actors have relationships outside the region rather than within it (Britton, 2004; A. R. Markusen et al., 1999).

Research in places such as Silicon Valley (Harrison 1994), Baden Wuerttemberg (Staber 1996), Hertfordshire (Simmie 1999; Simmie 1998) has also indicated that fi rms have networks outside the region. These studies showed that local ties are less eff ective in the later stages of growth due to increasing competition. In addition, fi ndings from research in the UK, Germany and US suggest that the role of international and national relationships is found to be much stronger than local ones (Henry et al., 2000). Also, Alderman’s (1999) fi ndings from research in engineering in three regions argued that local networks for technical development are not important. In fact, in many instances they appear to be irrelevant. In his analysis of manufacturing establishments of the electronics cluster in the Toronto region, Britton (2004) concluded that fi rms do not constrain their knowledge inputs to opportunities found in their industrial cluster. Rather, fi rms developed strong non-local ties to meet their input and output requirements (Britton, 2004). Doloreux (2004) showed from case studies of the Ottawa and Beauce regions of Canada that fi rms make use of regional, national and international knowledge sources to sustain innovation (Doloreux, 2004). Similarly, Asheim (2002) studied three regional clusters in Norway dominated by shipbuilding, mechanical engineering and the electronics industry. His fi ndings supported the claim that non-local ties were crucial to the innovation process (B. Asheim and Isaksen, 2002). In their case study of three industrial districts in Germany, Grotz and Braun (1997) showed that while local networks are important for general business issues, non-local networks are important for innovation and technology-oriented information (Groties, 1997).

In the case of developing countries, empirical studies show that substantial networking takes place between technology-related actors in some regions (Fromhold-Eisebith, 1999; Razavi, 1997). However, these ties include not only local ties but also non-local ties with machinery suppliers, customers (Katz, 1987; Lall, 1987; Bell, 1999), the state, and other international linkages (Ernst, 2002; Fromhold-Eisebith, 1999; A. Markusen et al., 1999). The state sets the political framework for development with a wide range of instruments regarding industrial and regional policy, science/technology policy, and educational policy (Fromhold-Eisebith, 1999). In a case study of

TABLE 1 The Structure of Local and Non-local Networks in Innovation

Characteristics Local Networks of Innovative Firms Non-local Networks of Innovative Firms

Boundary Spatial proximity, co-location (Pyke, Becattini, and Sengenberger 1990; Storper 1997; Ratti 1991; Cooke 2001) Decentralized, i.e. interregional and international relations (Amin and Cohendet 1999) Organizational proximity (Oinas 1999; Malecki and Oinas 1999)

Size

Diversity

Type of Resources Dense, as in the higher number of interactions (Torre and Gilly 2000; Staber 2001)1) Emphasis on customers and suppliers (Von Hippel 1998; Audrestch 1988) Diverse networks including fi rms, customers, suppliers, universities, research organizations and other (Kaufmann and Todtling 2000; Perrin 1991) Customers and suppliers, state organizations, universities

Emphasis on tacit knowledge (Storper 1997) Emphasis on codifi ed knowledge (Amin and Cohendet 1999; Maskell and Malmberg 1999a; Asheim and Isaksen 2002)

Stability (Duration)

Formality

Communication frequency and media Longer duration of ties, i.e. the number of years relationship existed (Oerlemans, Meeus, and Boekema 2000) Shorter duration of ties (Nooteboom and Gilsing 2004)

Formal and informal relations, market and non-market (Storper 1997; Torre and Gilly 2000; Gordon and McCann 2000)

Formal relations, regulated by market

Face to face and frequent relationships (Storper 1997; Torre and Gilly 2000; Gordon and McCann 2000) Communication media including phone, e-mail, fax Less frequent

SUNGU-ERYILMAZ 105

PROJECTIONS 7 INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

SUNGU-ERYILMAZ

106 PROJECTIONS 7 INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT Korean fi rms, Dahlman et al. (1987) show that it is the co-evolution of international and domestic knowledge ties that explains Korea’s extraordinary success. Table 1 summarizes the main characteristics of local and non-local networks mentioned in the literature.

FRAMEWORK TO STUDY INNOVATION AND INNOVATION NETWORKS IN DEVELOPING COUNTRIES

Any study about technological innovation in developing countries should consider several issues. First, studies in developing countries should take into account incremental innovation and their eff ect on economic development. Shop-fl oor or incremental innovations, occurring in day-to-day operations, contribute signifi cantly to the competitiveness of developing countries. However their eff ects are not captured by formal innovation studies and indicators. As to the new or major innovation, the defi nition of these types of innovation should be broad. Even if a fi rm introduces a technique that is already used, this still represents a new innovation for that fi rm.

Second, the geographical manifestation of innovation is a more complex phenomenon for developing countries. While local networks are important in innovation activities of fi rms, they may also have non-local ties in the form of inter-regional and international ties. Regions are important units of analysis in innovation because diff erences in economic, business, social and institutional infrastructure infl uence the type and intensity of local networks (Storper 1997; Amin 1999; Visser and Boschma 2004), and its mixture with non-local relations. The degree of mixedness of local and non-local ties plays an important role in building innovative capability (Sungu 2006). For example, comparison of innovation networks in two regions (Ankara and Istanbul) in Turkey revealed that these regions diff er in the degree of mixedness and characteristics of local and non-local networks. Istanbul as a global city has more ties to international and national organizations while Ankara has mostly local ties due to specialization in defense industry and government as the main customer (Sungu 2006). Local and non-local networks are not substitutes but complements. Non-local networks may be used to access capabilities that are not present locally. The region may not contain all the resources, especially in the case of developing countries. Firms have to stay tuned with what happens in the market, what happens among other producers, customers, and suppliers (ibid).

Third, the characteristics of local and non-local ties should be considered in regional innovation studies. These characteristics include their size, diversity (type of organizations), multiplexity (type of resources provided), stability (duration of ties), and formality. Territorial innovation models identify local networks as strong networks in the

innovation process as they are larger in size (Storper 1997); they facilitate production and diff usion of tacit knowledge which is emphasized as an important component of the knowledge used in innovation; and they are built on longer, informal, and personal relationships depending on trust. However, the characteristics of non-local networks are not generally known.

CONCLUSION

Territorial innovation models highlight the local dimensions of networks, but these models have not assessed the non-local dimension and have not considered the relative importance of local versus non-local innovative networks. It is true that interactions have a spatial nature, but they also have an organizational nature. Non-local networks might represent organizational proximity (Malecki and Oinas, 1999; Oinas, 1999). Similarly, local networks are important due to production and easy diff usion of tacit knowledge, which both require proximate relations. However, tacit and codifi ed knowledge are complements for competitive advantages in diff erent stages of a fi rm’s life cycle (Bathelt et al., 2004). Local networks may not be eff ective in places where resources and knowledge infl ow are limited. In those cases, local networks should be complemented by non-local resources.

A more empirical problem in the literature is that the majority of regional studies have a tendency to focus on fi nding data at the local level, and consequently neglect the importance of non-local networks. This problem is especially relevant for regions in developing countries. While models of local networking have emerged in the policy agendas of developing countries (Altenburg and Meyer-Stamer, 1999; Bell and Pavitt, 1992; Cooper, 1991), these models fail to identify potential benefi ts that peripheral regions could collect from non-local (inter-regional and international) linkages. Therefore, empirical and policy research should look for a more complete model of the networking behavior of innovative fi rms in developing countries by combining local and non-local networks.

SUNGU-ERYILMAZ 107

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT PROJECTIONS 7

SUNGU-ERYILMAZ

108 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT ENDNOTES

1 Adam Smith, in his book Wealth of Nations (1776), Chapter 1, discusses the importance of improvement in machinery. Marshall (1890), in his book of Principles of Economics, describes knowledge as the chief engine of production and growth.

2 Godin (2005) argues that the precise source of the linear model of innovation is unknown. However, it can be argued that in his second model, Schumpeter argued that innovation process had become endogenous with the emergence of R&D departments.

3 Please see the special issue of World Development (March 1974) on the subject of the choice of appropriate technology and technology transfer.

REFERENCES

Acs, Z. J., and Audretsch, D. B. (1988). Innovation in Large and Small Firms: An Empirical Analysis. American Economic Review, 78, 678-690.

Altenburg, T., and Meyer-Stamer, J. (1999). How to Promote CLusters: Policy Experiences from Latin America. World Development, 27(9), 1693-1713.

Amin, A. (1999). An Institutionalist Perspective on Regional Economic Development. International Journal of Urban and Regional Research, 23(2), 65-79.

Amin, A., and Cohendet, P. (1999). Learning and Adaptation in Decentralized Business Networks. Environment and Planning D: Society and Space, 17, 87-104.

Amin, A., and Cohendet, P. (2005). Geographies of Knowledge. Industry and Innovation, 12(4), 465-486.

Amsden, A. (1989). Asia’s Next Giant. South Korea and Late Industrialization. London: Oxford University Press.

Archibugi, D., and Michie, J. (1995). The Globalization of Technology: A new taxanomy. Cambridge Journal of Economics, 19, 121-140.

Arrow, K. J. (1962). The Economic Implications of Learning by Doing. Review of Economic Studies, 29.

Asheim, B., and Isaksen, A. (2002). Regional Innovation Systems: The integration of local sticky and global ubiquitious knowledge. Journal of Technology Transfer, 27, 77-86.

Asheim, B. T. (1996). Location, Agglomeration and Innovation: Towards Regional Innovation Systems in Norway? In A. Isaksen (Ed.). STEP Group.

Audretsch, D. (1998). Agglomeration and the Location of Innovative Activity. Oxford Review of Economic Policy, 14(2), 18-29.

Aydalot, P., and Keeble, D. (1988). High Technology Industry and Innovative Environments. London: Routledge.

Bathelt, H., Malmberg, A., and Maskell, P. (2004). Clusters and knowledge: Local buzz, global pipelines and the process of knowledge creation. Progress in Human Geography, 28(1), 31-56.

Bell, M., and Albu, M. (1999). Knowledge Systems and Technological Dynamism in Industrial Clusters in Developing Countries. World Development, 27(9), 1715-1734.

Bell, M., and Pavitt, K. (1992). Accumulating Technological Capability in Developing Countries. Paper presented at the Annual Conference on Development Economics, Washington, D.C.

Braczyk, H., Cooke, P., and Heidenreich, R. (Eds.). (1997). Regional Innovation System. An Evolutionary Approach. London: UCL Press.

Britton, J. (2004). High technology localization and extra-regional networks. Entrepreneurship and Regional Development, 16(September), 369-390.

Brusco, S. (1982). The Emilian Model: productive decentralization and social integration. Cambridge Journal of Economics, 6, 167-184.

Camagni, R. (Ed.). (1991). Innovation Networks: Spatial Perspectives. London: Belhaven Press.

Coff ey, W., and Polese, M. (1984). The concept of local development: A Stages Model of Endegenous Regional Development. Papers of Regional Science Association, 55, 1-12.

Cohen, W., and Levinthal, D. (1990). Absorptive Capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128-153.

Cooke, P. (2001). Regional Innovation Systems, Clusters and the Knowledge Economy. Industrial and Corporate Change, 10(4), 945-974.

Cooper, C. (1991). Are Innovation Studies on Industrialized Countries Relevant to Technology Policy in Developing Countries? : UNU/INTECH.

Dahlman, C. J., Ross-Larson, B., and Westphal, L. E. (1987). Managing Technological Development : Lessons from the Newly Industrializing Countries. World Development.

Doloreux, D. (2004). Regional Innovation Systems in Canada: A Comparative Study. Regional Studies, 38(5), 481-494.

Domar, E. (1957). Essays in the theory of economic growth. New York: Oxford University Press.

Dosi, G., Freeman, C., Nelson, R., Silverberg, G., and Soete, L. (1988). Technical Change and Economic Theory. London: Pinter.

SUNGU-ERYILMAZ

109

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT PROJECTIONS 7

SUNGU-ERYILMAZ

110 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

Edquist, C., Hommen, L., and McKelvey, M. (2001). Innovation and Employment: Process Versus Product Innovation. Cheltenham, UK: Edward Elgar Publishing.

Ernst, D. (2002). Global Production Networks and The Changing Geography of Innovation Systems. Implications for Developing Countries. Economics of Innovation and New Technology, 11(6), 497-523.

Ernst, D., Ganiatsos, T., and Mytelka, L. (Eds.). (1998). Technological Capabilities and Export Success in Asia. United Nations: Routledge.

Forbes, N., and Weild, D. (2000). Managing R&D in technology-followers. Research Policy, 29(9), 10951109.

Friedmann, J., and Weaver, C. M. (1979). Territory and Function: The Evolution of Regional Planning. London: Edward Arnold.

Fromhold-Eisebith. (1999). Bangalore: A Network Model for Innovation-Oriented Regional Development in NICs? In E. Malecki & P. Oinas (Eds.), Making Connections: Technological Learning and Regional Economic Change (pp. 231-260). UK: Aldershot.

Gertler, M. S. (2003). Tacit Knowledge and the economic geography of context, or the undefi nable tacitnessof being (there). Journal of Economic Geography, 3, 75-99.

Glasmeier, A. (1999). Territory- Based Regional Development Policy and Planning in a Learning Economy. The Case of “Real Service Centers’ in Industrial Districts. European Urban and Regional Studies, 6(1), 73-84.

Godin, B. (2005). The Linear Model of Innovation: The Historical Construction of Analytical Framework. Canadian Science and Innovation Indicators Consortium (CSIIC).

Gordon, I. R., and McCann, P. (2000). Industrial Clusters: Complexes, Agglomeration and/or Social Networks. Urban Studies, 37(3), 513.

Granovetter, M. (1985). Economic Action and Social Structure: The problems of Embeddedness. American Journal of Sociology, 91, 481-510.

Harrison, B., Kelley, M. R., and Gant, J. (1996). Innovative Firm Behavior and Local Milieu: Exploring the Intersection of Agglomeration, Firm Eff ects, and Technological Change. Economic Geography, 72(3), 233-258.

Howells, J. (2002). Tacit Knowledge, Innovation and Economic Geography. Urban Studies, 39(5-6), 871-884.

Johnson, B., and Lundvall, B.-A. (2001). Why all this fuss about codifi ed and tacit knowledge? DRUID Winter Conference.

Kaufmann, A., and Todtling, F. (2000). Systems of Innovation in Traditional Industrial Regions: The Case of Styria in a Comparative Perspective. Regional Studies, 34(1), 29-40.

SUNGU-ERYILMAZ

112 PROJECTIONS 7

INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT

Mytelka, L., and Ernst, D. (1998). Catching up, Keeping up and getting ahead: Korean Model under pressure. In D. Ernst, T. Ganiatsos & L. Mytelka (Eds.), Technological Capabilities and Export Success in Asia. London: Routledge.

Mytelka, L. K. (Ed.). (1999). Competition, Innovation and Competitiveness in Developing Countries. France: OECD.

Nelson, R. (Ed.). (1993). National Innovation System: A Comparative Analysis. New York: Oxford University Press, Inc.

Nelson, R., and Winter, S. (1982). An Evolutionary Theory of Economic Change. Cambridge: The Belknap Press of Harvard University Press.

Nelson, R. R. (1959). The Simple Economics of Basic Scientifi c Research. Journal of Political Economy, 67(3), 297-306.

Nonaka, I., and Takeuchi, H. (1995). The knowledge Creating Company: How Japanese companies create the dynamics of innovation. Oxford: Oxford University Press.

Nonako, I., Toyama, R., and Nagata, A. (2000). A Firm as a Knowledge-creating Entity: A New Perspective on the Theory of the Firm. Industrial and Corporate Change, 9(1), 1-20.

Nooteboom, B., and Gilsing, V. A. (2004). Density and Strenght of Ties in Innovation Networks: A competence and governance view. Erasmus Research Institute of Management.

Noteboom, B. (1994). Innovation and Diff usion in Small Firms: Theory and Evidence. Small Business Economics, 6(5), 327-347.

Noteboom, B. (1999). Innovation, Learning and Industrial Organization. Cambridge Journal of Economics, 23, 127-150.

Oerlemans, L., Meeus, M., and Boekema, F. (2000). On the Spatial Embeddness of Innovation Networks: An exploration of the proximity eff ect. Paper presented at the European Congress of the European Regional Science Association, Barcelona.

Oinas, P. (1999). Activity-specifi city in Organizational Learning: Implications for Analysing the Role of Proximity. GeoJournal, 49, 363-372.

Perrin, J. (1991). Technological Innovation and Territorial Development. An Approach in terms of networks and milue. In R. Camagni (Ed.), Innovation Network: Spatial Perspectives. London, New York: Belhaven Press.

Piore, M. J., and Sabel, C. F. (1984). The Second Industrial Divide. New York: Basic Books.

Porter, M. E. (1990). The Competitive Advantage of Nations. New York: Free Press.

SUNGU-ERYILMAZ 115

PROJECTIONS 7 INSTITUTIONAL INNOVATIONS FOR DEVELOPMENT