Abstract: The literature on innovative regions has stressed the role of spatial clustering in the endogenous formation, accumulation, and sharing of knowledge in technology hubs in advancedcountries. Little has been written, however, on how spatial clusters may also affect technological dynamics in developing countries where external actors, such as multinational corporations (MNCs),are the main source of new technology. Drawing on studies of innovative regions and technology transfers, we analyzed the interactive patterns between MNCs and local technology actors in China’sleading information communication technology (ICT) service cluster in Zhongguancun, Beijing. We found that the relationship between MNCs and local firms is hierarchical, but also interdependent andevolutionary. Local firms’ collaboration with MNCs provides them with vital technological and organizational training, which the local firms use strategically to develop their market networks andinnovative capacity in the home market. The learning capacity of local firms is vastly improved by the presence of other related enterprises, the research and development facilities, and adevelopmental state in a market-oriented spatial cluster.
Key words: innovative region, spatial cluster, information technology sector, China, multinational corporations.
In the past 20 years, the literature on economic geography has gained remarkable insights into the technological dynamism in the high technology-oriented regions in advanced countries. In particular, this literature has stressed the processes of the endogenous formation, accumulation, and sharing of knowledge, enabled by formal and informal networks among spatially clustered firms, in sustained technological innovation. Since these studies have been conducted primarily within developed countries, it is unclear whether a similar framework can also be applied to developing regions. In such cases, the accumulation and dispersion of knowledge is often characterized by the strong influence, if not dominance, of the external players who provide the main sources of new technology.
There is a growing literature on technological changes in developing countries as a matter of technology learning and transfer. These studies have investigated the role of multinational corporations, institutional structure, strategies of firms, industrial policies, and cultural differences in the effectiveness of technological transfer (Malecki 1991; Scheraga, Winston, Tellis, and Tucker 2000; Kim and Nelson 2000; David 1997). They have rarely taken a regional perspective, however, and have tended to concentrate on the manufacturing and exporting sectors, with little regard for the distinct dynamics in the services sectors.
This article seeks to synthesize the literature on innovative regions with that on the technology transfer to developing countries by analyzing the interactions between multinational corporations (MNCs) and local firms in a region that specializes in high- technology (hereafter “high-tech”) services. We argue that the reliance on external sources of technology does not diminish the role of a region in influencing technological changes. The empirical case is based on the development of a cluster of information communication technology (ICT) in Beijing. Rather than export- oriented manufacturing activities, the article focuses on ICT services oriented to China’s domestic market.1 It shows that the relationships between MNCs and local firms are hierarchical, but also fluid and increasingly contested. On the one hand, the superiority of MNCs in capital, technology, and management may to a great extent undermine local incentives for innovation. On the other hand, the difficulties of operating under a less technologically sophisticated market for high-tech services have forced MNCs to adopt a collaborative approach with local firms. As a result, local firms have the opportunity to receive vital technological and organizational training from MNCs and to develop strategically their market networks and innovative capacity in the home market. The learning capacity of local firms is vastly improved by the presence of other related enterprises, research and development (R&D) facilities, and a developmental state favoring local firms. Technology transfer, therefore, is accomplished not as a deliberate process or through a particular organizational form, but as a result of market logic and the division of labor between MNCs and local firms in a regional cluster.
Insights from the Literature on Innovative Regions
Regions and Innovations
In the past 20 years, a tremendous amount of literature has been devoted to understanding the role of regions or spatial clusters in stimulating, facilitating, and sustaining industrial competitiveness and technological dynamism. The literature has moved beyond looking at the presence of major factors (such as human resources, universities, and corporate R&D facilities) to examining the role of social, cultural, and organizational synergy unleashed and the unique institutional culture created by spatial agglomeration (Porter 2000; Krugman 1995; Henry 1995; Storper 1997; Kenney 2000; Amin 2000; Amin and Thrift 1992).
Spatial clustering is seen as facilitating the deepening interfirm division of labor and the formation of a flexible production network, which, in turn, speeds up the innovation process (Scott 1988a, 1988b, 1993; Storper and Walker 1989; Scott and Paul 1990; Amin and Cohendet 1999; Harrison 1992; Saxenian 1994). Storper (1997) suggested that regions gain their competitiveness through localized input-output relations for the development and exchange of information, and through a process of economic organizational learning and coordination. Innovations are seen as collective and interactive processes and are thus evolutionary and path dependent (Nelson and Winter 1982; Castells 1996; Malecki 1991; Dahlman and Aubert 2001; North 1990). Various terms had been used to capture the importance of region for innovations, such as industrial districts, innovative milieus, regional systems of innovation, technopoles, and learning regions. While each term offers a somewhat different conceptual twist, all stress how regional institutions and a localized technology and knowledge base are critical for such regions’ distinct trajectories (Andersson 1985; Rigby 2000; Malecki 1991; Porter 2000; Amin and Cohendet 1999; Gertler 1995; Amin and Thrift 1992; Bunnell and Coe 2002).
Since this literature has been based primarily on case studies in advanced countries, it is not clear whether similar arguments can also be made for technology regions in developing countries. One key distinction is that technology regions in developing countries typically depend on external technology or markets,2 which seems to contradict the local embeddedness argument that is central to the literature on innovative regions. Park and Markusen (1995), for example, found it difficult to apply the theory of new industrial space to South Korea and proposed relaxing the definition of industrial district to accommodate those satellite industrial districts created by nonlocal-based corporate branches. This solution is problematic, since it requires the significant dilution of the endogenous-driven technology central to the theory of new industrial space without demonstrating much benefits of using this framework in those satellite regions.
Region, MNCs, and Technology in Developing Countries
In contrast to the abundant literature on technologically innovative regions in advanced countries, region as an analytical concept has not had such a prominent role in understanding technological diffusion or changes in developing countries. The growing literature on spatial clustering in developing countries has suggested that spatial clustering is indeed important in the Third World context, especially for smallscale production and during the incipient stage of development (Schmitz and Nadvi 1999). Yet much of the work tends to be based on more traditional sectors. Studies of high-tech regions tend to be more concerned with industrial policies than with regional networks or institutions.
The literature on developing countries has instead emphasized the role of MNCs. Since MNCs often stand at the apex of the technology landscape in developing countries, it is critical to analyze their relationship to local firms to unpack the dynamics of technology in these regions. It has been well recognized that multinational firms have little incentive to transfer core technologies to developing countries unless they are pressured to do so by a host country that has a strong bargaining position3 (Dicken 1998; Porter and Sheppard 1998). In addition, the transfer may include “know-how” (production engineering), but not “know-why” (basic design and R&D) (Lall 1984, 10), so it often only reinforces the dependence of receiving countries.
Economists who have studied the experiences of Southeast Asia, particularly South Korea and Taiwan, have argued that technology catch up can occur if developing countries adopt sound policies on investment in human capital, export orientation, and industrial policies (Kim 1980; Lall 2000; Pack 2000). Export strategie\s, either through original equipment manufacturing, as practiced in Taiwan, or through encouraging subsidiaries of MNCs, as practiced in Malaysia, are seen as generating superior results on technology transfer than is import substitution practiced by Brazil (Hobday 2000). Yet, scholars have also acknowledged that the catch up does not eliminate the dependence on MNCs. Even when economies like South Korea and Taiwan are technological powers in their own right, they are still subordinate to MNCs for core technology and market.
Although this literature has yielded important insights on the process of technological catching up, it has shed little light on the role of regions in technology development beyond being sites that provide cheap labor and policy incentives. It is even questionable that region as an endogenous knowledge base even matters. Pack (2000), for example, contended that extreme openness to foreign technology is the key to upgrading technology, while building domestic R&D capacity has not been proved beneficial. This view is indeed in stark contrast to the literature on developed regions.
A number of successful high-tech regions exist in developing countries. The Hsinchu Science-based Industrial Park in Taiwan is an example of a dynamic, yet indigenous company-based innovative region in the 1990s (Saxenian and Hsu 2001). The companies were able to keep their technology edge by forging close personal networks with firms in Silicon Valley, California, thereby internalizing the MNCs’ knowledge through social networks. Taiwan in the 1990s, however, had an income level comparable to some countries in the Organization for Economic Cooperation and Development (OECD)-hardly typical of the developing world. Bangalore in India is a truly successful high- tech region for software export in a developing country. But its dependence on external markets may arguably limit its long-term growth and its contribution to India’s transformation (Parthasarathy 2002; Schware 1992).
This article argues that the reliance on external sources of technology in developing countries need not be a prohibitive factor in studying a region’s influence on innovation. Insights generated in developed countries can provide a useful framework for understanding the dynamics of technology in developing countries, provided we take good account of the role of MNCs and their relationship to local firms. Scholars have recently called for more consideration to be given to extra-local connections in technological changes even in advanced countries (Bunnell and Coe 2002; Oinas and Malecki 2002). While MNCs provide the main source of new technology in developing regions, the relationship between MNCs and local firms varies from region to region and cannot be understood simply as a matter of domination.
In fact, local firms in developing countries may occupy strategic positions in their technological market. Castells (1996, 32) argued that the current technological revolution is characterized by a “cumulative feedback loop” between the innovation and the use of innovation. Since local firms are often the agents that deal directly with the technology users in developing countries, they have considerable marketing power and are favorably situated to gather information on the market. Bell and Albu (1999) argued that a spectrum of technology improvements unaccounted for by formal R&D existed in developed countries. Chew and Yeung (2001) found that “reverse transfer” existed as Singapore’s small- and medium-sized firms transferred their local expertise to giant MNCs. This finding suggests that the relationship between MNCs and local firms can also be interdependent even though it is hierarchical. Regional resources, networks, and institutions are thus not entirely passive or irrelevant contexts, but constitute a critical platform that shapes the potential for learning and innovation in local firms.
High-Tech Services and Technology Dynamics
We decided to focus on the information technology service sector not only because it is a rapidly growing area in both developing and developed countries, but also because it allowed us to target the technology application process and examine the dynamics involved. The literature on technology has generally paid far greater attention to the production of technology than to its consumption and application (see Naughton 1997, for example). Given that export orientation is seen as the strategy for technological catching up, even less attention has been paid to technology learning in serving developing countries’ domestic markets. Indeed, such sectors may even be dismissed as failed routes of import substitution with the implication of protectionism. But import substitution can occur in a relatively open system and represents a viable venue for technology learning. Some scholars have contended that mainstream economists have underestimated the technological learning under import substitution in countries such as Argentina, Brazil, and Mexico (Katz 2000). We argue that understanding the process and agents involved in applying technology in developing countries will not only fill a glaring gap in our knowledge of global technology diffusion, but also allow us to address how the application process may affect technology advances in local firms.
There is a significant interdependence between MNCs and local firms in technology application processes. Unlike the manufacturing of standard products for export, MNCs in service sectors need substantial local expertise and channels to market their technology. They also tend to work in partnership with local technology firms and thus have much greater incentives to promote, share, and modify their technology to address market needs that may be different from those of their home countries, provided that the market is attractive enough. In addition, the technical frontier of the ICT industry is fragmented and rapidly changing. The fast-moving technology makes MNCs relatively generous in transferring mature technology to local partners.
Compared with MNCs, local firms have the advantage of their local expertise and connections, as well as lower costs, so they tend to thrive on the application end of the work. Local firms have a great interest in moving up on the ICT value chain, however, to earn higher profit margins, especially when the competition heats up in their field. The entry barriers in many parts of information services can be low (e.g., in specialized software development or system integration), so it is not unusual to have rapidly increasing competition in profitable areas of ICT services. By being closer to the users, local firms are in a favorable position to understand the needs of the market and the potential of technology. If they are able to incorporate the vital technology and management expertise from MNCs, they can advance, gain a measure of autonomy, and even construct the “cumulative feedback loop” in technology in their home country. However, whether local firms can succeed depends on the competition with MNCs, the availability of resources (R&D, human, and capital), and the quality of the market (size and sophistication).
We argue that a regional cluster of diverse, but interrelated, technology firms and R&D institutions provides an ideal setting to address the last two difficulties that local firms necessarily confront. The presence of diverse firms and research institutions in spatial proximity can assist firms to advance technology by adding or solving necessary pieces of the puzzle that are not provided externally. The clusters can also enhance firms’ access to the market by providing information and entrance to the network. Overall, clusters strengthen the learning capacity of local firms even if the sources of new technology are external.
China’s Zhongguancun High-Tech Cluster
The empirical case presented here is based on China’s most prominent high-tech center, Zhongguancun (ZGC), in north-western Beijing (see Figure 1). Beijing is China’s most important center of R&D in the ICT sector. In 2001, 16 of China’s top 100 electronic hardware manufacturing companies were headquartered in Beijing, as were 23 of the top 100 software companies (Ministry of Information Industry 2002a, 2002b). Beijing’s revenue from software sales accounted for 32.8 percent of the national revenue in the same year (Beijing Statistical Bureau 2001). The core of Beijing’s high-tech power is nested in ZGC. ZGC emerged in the mid-1980s as a cluster of computer stores, and it is considered the birthplace of China’s nongovernmental computer sector (Francis 1997). It has grown rapidly since the mid-1990s, and in 2000 hosted the headquarters of many of China’s leading information technology companies, as well as over 6,000 other small- and medium-sized companies in a wide range of high-tech sectors, 75 percent of the firms in the IGT sectors (Zhongguancun Administrative Government 2001) (see Table 1). Popularly known as China’s Silicon Valley, ZGC was expected by the Chinese central government to become the next pivotal region in China, akin to ShenZhen in the 1980s and PuDong district in Shanghai in the 1990s (N. Li 1999). Despite its glowing domestic fame and ambition, ZGC has received little international recognition as a high-tech hub. It was not on the list of 46 locations deemed to “matter most in new digital geography” by Wired magazine, while Hsinchu and Taipei, Taiwan; Hong Kong; Seoul; Albuquerque, New Mexico; and Gauteng in South Africa were (Hillner 2000).4 In fact, none of the locations within mainland China made the list.
We decided to focus on ZGC not only because we believe that it deserves a place on the global map as a center of technological changes, but also because it is a region that specializes in high- tech services for China’s domestic market and therefore is appropriate for studying the interactions between MNCs and local fi\rms in the diffusion and application of technology. ZGC is a unique high-tech region whose remarkable feature is neither its position in the global production chain nor its potential to generate core ICT technology, but its ability to learn and translate foreign technology into the institutions and marketplace of China.
The research was based on 10 months of fieldwork in Beijing from September 2001 to June 2002 and subsequent data gathering through the Internet. We collected the secondary information from the ZGC Administrative Government, marketing research companies inside and outside China, and Chinese ICT industrial journals. The primary source was over 80 interviews with company executives; industrial journalists; and officials in governmental, university, and nongovernmental organizations (see Table 2). The firms vary greatly in size, from the largest ICT firm in China to software companies with about a dozen people. Most of the executives we interviewed were vice presidents of the companies or higher. Sometimes multiple interviews were conducted in large firms. The interviews inquired about the history of the firm; its technological development and key technology sources; its subcontracting relationships; and its interactions with other firms, especially MNCs, local research institutions, and governments. We found that although getting interviews was no easy matter, the persons we encountered were surprisingly open and candid in sharing their opinions. Most interviews were taped and transcribed in Chinese and a few in English, depending on the language spoken in the interviews.
Figure 1. Zhongguancun Science Park is composed of five zones: Haidian zone, Changping zone, Yizhuang electronics and technology development zone, Jiuxianqiao electronics city, and Fengtai zone. The earliest and core region is Haidian zone, which we refer to in this article as Zhongeuancun.
The next section gives a brief account of the evolution of ZGC. It is followed by a discussion of the interactions among ZGC’s main actors, with a particular focus on the dynamic relationships between MNCs and local firms. The last two sections describe the characteristics of the region and the role of the Chinese state in technology development.
Development of Zhongguancun as a High-Tech Cluster in China
ZGC was originally designed as Beijing’s scientific research and educational zone. Over time, the Beijing municipal government either moved or established research institutions and universities in this area, creating the largest enclave of top scientists and engineers in China. In the late 1990s, the vicinity of ZGC counted 39 universities, 213 research institutes, and 378,000 engineers and scientists (ZGC Administrative Government 2000). The most prominent institutions include Peking University and Tsinghua University, China’s “Harvard” and “MIT,” and the Chinese Academy of Sciences with its 41 research units, in addition to other leading technology universities, such as Beijing Polytechnic University, Beijing University of Aeronautics and Astronautics, and Beijing University of Post and Telecommunication. Under China’s centrally planned regime of separation of scientific research and industrial enterprises, this enclave of China’s scientific crown jewels yielded little benefit for the local region (Baark 2001). In 1980, Beijing’s Haidian district, where ZGC is located, was known simply for its good schools and agricultural produce.
In the mid-1980s, under the pressure of severe budget cuts for public research institutions, also inspired by China’s reform policies, a successive number of scientists began leaving China’s Academy of Sciences and universities to create enterprises. Many of these companies were so-called Guo-You-Min-Ying firms, which literally translates into state-owned, people-run firms, or just Mm- Ying (people-run) firms to distinguish them from dominant state- owned enterprises. Because the state provided little initial funding, the companies were granted considerable autonomy in allocating capital and hiring employees. Drawing from the technological accumulation, expertise, connections, and name recognition of their mother institutions, these companies cultivated an emerging marketplace on technology. Initially, many ZGC firms engaged in technical services, such as testing and assembling computers for Chinese organizational users. In the late 1980s and early 1990s, a number of firms also developed their own products in the areas of Chinese language processing, based on the technological accumulation of their mother institutions. The profit made by these firms, in return, provided desperately needed financial relief for the financially deprived public institutions (Zhao 1998). Wang and Wang (1998) found that from the 1980s to early 1990s, ZGC was similar to Silicon Valley in its small firms, formation of new firms, and innovativeness. The dense academic institutions in the ZGC area provided the initial human resources, expertise, and market for information technology products, thus constituting an ideal launching point for technology diffusion in China.
In the 1990s, the growth of China’s technology market attracted an influx of MNCs. Table 3 shows the timetable of some of the most important MNCs’ operations in China. To reach Chinese markets and to comply with the Chinese government’s regulations, MNCs recruited Chinese firms as their sales agents. While ZGC firms grew substantially in size and number as a result, the incentive for autonomous R&D diminished as frontier technological development was monopolized by the MNCs. Even local firms’ core strength in Chinese language processing was largely eliminated by the release of the Chinese version of Windows. By the late 1990s, ZGC changed from an indigenous innovative region to a satellite marketing platform for MNCs. The loss of its innovative edge was widely bemoaned, but was perhaps inevitable, since the previous innovativeness in ZGC was achieved only within an isolated economy. Such innovativeness could not be sustained under the far-superior technological, capital, and management power of MNCs in an unsheltered market.
Table 3
Selected Multinational Technology Corporation Establishments in Beijing
Although there was little self-generated technological innovation by ZGC firms in the 1990s, the speed of adaptation of foreign technology quickened significantly. A number of local firms were able to use the marketing expertise learned from being representatives of foreign companies to produce and market their own products. Legend Group was the most successful one.5 Legend realized that the PC products that MNCs introduced to China were one generation behind those in the world market. With the knowledge of China’s users and by taking advantage of its established marketing channels for selling MNCs products, Legend introduced the fastest PC chip into its models and sharply reduced its price at below the level of foreign brands in 1996. The dramatic gain in its market share made it the best seller in China in 1997 (Zhu 2000). Legend has since maintained the lead in the Chinese market at about a 30 percent share and has become one of the largest PC makers in the Asia/Pacific Rim. Its success has widely been credited not to its technological superiority, however, but to its market strategies, efficient management, and massive distribution and service networks across China. Here is an initial case of how being close to the users and home market can be translated into superior business performance and technological advancement. A number of Chinese PC makers have also followed. Now Chinese brand-name PCs control roughly two-thirds of the market in China (China Center for Information Industry Development 2002).
Legend’s success in the PC market was instrumental not only in ending the domination of foreign-brand PCs in the Chinese market, but also in suggesting the possibility of the growing roles and autonomy of Chinese high-tech firms under intense global competition. Since then, the quest to establish domestic brand names and to master core technology in the hardware and software fields has been an ever-present theme in the Chinese ICT industry, as well as a key consideration in state policy.
Chinese Firms and MNCs: Relationships and Dynamics
Entering the twenty-first century, China has emerged as one of the largest ICT markets in the world, particularly in cell phones, PCs, and telecommunication equipment (G. Li and Wang 2001). In 2002, it also had the world’s second largest and rapidly growing Internet population. The market growth has caused MNCs to be increasingly serious about China. It has also provided unprecedented new market opportunities for ZGC’s hardware and software firms.
Cooperation Between Chinese Firms and MNCs
Figures 2 and 3 summarize the relationships and links among some of the key actors in ZGC. They depict hardware and software separately, with the understanding that many firms are involved in both. While Figures 2 and 3 were constructed separately for technical reasons, our comments are presented in relation to the commonalities and differences in both areas. In both hardware and software, Chinese firms occupy an intermediate position between more technologically advanced MNCs and the Chinese market. In the hardware market, MNCs are dominant suppliers of high-end products, such as high-end telecommunication network equipment, advanced integrated circuit (IC) chips, and mainframes and microcomputers. For PCs and the lower-end communication equipment market, Chinese brand names are rapidly gaining ground.
In the software field, MNCs are also dominant suppliers of system and application software, including office applications, enterprise management software like enterprise resource planning (ERP), database, CAD, and middleware. Because of China’s high piracy rate, the consumer software market is barely profitable for anysoftware makers (“Reining in China’s Pirate Software Market” 2000). The most profitable fields are in enterprise software, such as database, middleware, and management software. Although hardly a formidable force in the past 15 years, software is expected to grow much faster than China’s hardware ICT sectors in the near future (China Software Industrial Association 2000). Most Chinese software companies are small and privately owned. They tend to survive on niche markets that are either sheltered from foreign competition or not affected severely by piracy, such as antivirus software, educational software, and accounting software (“Reining in China’s Pirate Software Market” 2000). There are a few large state-owned software companies, such as Chinese National Computer Software & Technology Service Corporation (CS&S), a spin-off from the Ministry of Information Industry. State-owned firms are involved in many areas of software development, from providing network system solutions and outsourcing work for foreign corporations to developing embedded software for microchips. But their most distinguished field is the development of software for alternative systems, based on Linux, for example.
Both Figures 2 and 3 reveal that Chinese firms are the key agents for delivering technology to Chinese consumers and organizations. Table 4 shows that the largest concentration of ZGC firms is in software developer/system integration (i.e., providing network installation, service, or solutions for enterprises and institutions). The bulk of such work consists of distributing and installing the hardware and software products for Chinese organizational users, but it typically also involves some software development and service consulting to make a network operational for its users. A number of large system integrators, like CS&S, also incorporate their own products, such as data-base and network security, into their contracts.
Table 4
ZGC IT Firm Subsectors and Size of Firms, 2001
As one can see, although MNCs and Chinese firms do compete in certain areas, they generally occupy different positions in the ICT value chain. MNCs are at the core or high end of product development, and Chinese firms are in the area of lower-end products, marketing, services, and system integration. Although MNCs may have been forced at the beginning to depend on Chinese firms for sales and distribution, they gradually discovered the merit of recruiting and relying on Chinese partners to do the applied end of the jobs, and they have made major efforts to train and develop Chinese partners. Some companies like Sun and Cisco estimated that they sell nearly 100 percent of their products through partnerships with Chinese firms.
The marketing executive6 of Oracle explained the reason why Oracle relies on Chinese partners.
Because the sale is not finished when you sell the software. We are in the area of e-business service. License is only the first step. You have to help the enterprise to change its flow procedure to work under the Internet environment. So you need the internal knowledge of the firm. China is so big. We cannot handle it all by ourselves. So we decide to educate our partners so they can help us cover the majority of the market. This process was difficult. At first, the sales rep did not care, just sold the license. Later, the competition became intense; they found they have to provide the value-added service. Now we have over a dozen good general contractor[s], including Legend and AsiaInfo. We are still developing more such contractors.
The Chinese market differs considerably from the American market, not only in its lower level of infrastructure and development, but also in its unevenness. A number of interviewees from MNCs commented that the fragmentation of the Chinese market is more comparable to the market in Europe than to the relatively homogeneous American market. For the MNCs, the high cost prevents them from being deeply involved in the complex and uneven application process that demands the liberal use of programmers and inside knowledge and connections to the organizations. They can do so only with selected large organizations. Chinese firms, on the other hand, have much lower labor costs and generally have an easier time establishing connections or communicating with users. It is not surprising, then, that many Chinese firms have moved into the application area.
MNCs and Chinese firms also cooperate in other areas. MNCs subcontract portions of peripheral software development to local firms. Some companies, such as IBM, have more substantial subcontracting work, but others, such as Sun, subcontract only the translation part (interviews). MNCs also encourage Chinese firms to develop applications that are based on their platforms. Toward this end, large MNCs have devoted considerable effort to train qualified Chinese engineers and users. Intel, Microsoft, Cisco, and Sun all provide training programs, called something like network universities (Cisco), to encourage Chinese users to attain certain levels of technological sophistication. Certification by Microsoft and Cisco is hotly pursued in Beijing as a prerequisite for landing good jobs.
The largest MNCs, such as Intel, Microsoft, Lucent, Sun, IBM, Motorola, and Oracle, also established or are planning to establish R&D centers in Beijing (see Table 3). The work in such R&D centers varies from basic research, like that at Microsoft’s Asia Research Institute, to work on localization. Many of the R&D projects concentrate on technology that is most significant in the Chinese market, such as voice recognition and wireless communication. For MNCs, R&D centers tap into inexpensive Chinese engineers and can score them political points, since the Chinese mass media inevitably interpret this step as making a deep commitment and bringing prestige to China. However, these R&D centers are generally part of the R&D in headquarters, having little connection with Chinese local firms (interviews).
One conflict between MNCs and Chinese firms is the competition for talent. MNCs are known to provide much better pay for skilled engineers and hence have attracted a continuous flow from Chinese local firms to MNCs, known as the “brain drain.” An interviewee at CS&S commented that CS&S is like a prep school for large MNCs because new graduates often work at CS&S for a few years before they move on to MNCs. Since 2001 the flow became less one way as a number of more successful Chinese companies were able to lure experienced executives from MNCs by providing comparable benefits. But this only says that it is now merely conceivable that employees in MNCs can move to Chinese firms, not that the trend has been reversed.
Beyond foreign technology corporations, overseas financial capital has also become more active in China’s ICT landscape, thereby projecting a different kind of influence. Many of the leading Chinese IT firms are listed on the stock markets in Hong Kong and New York. Foreign venture capital has also been active in establishing and acquiring Chinese local firms (Sheff 2002).
Overall, the relationships between Chinese local firms and MNCs have largely been mutually beneficial thus far. MNCs use Chinese local firms to open their markets and to deal with the more tedious and costly application work in the form of system integration or network solution. The arrangement benefits Chinese firms, since MNCs help to train Chinese firms in the technical, managerial, and organizational areas. Even the competition for talent can be seen as a way to force improvement in Chinese human resource management systems and to promote the exchange of information between MNCs and local firms. At the same time, however, the growth of Chinese markets and the maturing of Chinese ICT enterprises also means that competition and conflict between Chinese and foreign firms are likely to increase in the future.
Moving Up the Technology Value Chain
As the Chinese ICT market develops, specialized demands emerge, generating incentives and pressure for firms to move up the value chain. In the field of system integration, for example, simply installing computer networks has given way to providing total solutions for industrial-specific management and information-flow systems under the new network environment. As the number of system integrators increased, the entire sector began to stratify into general contractors and subcontractors. On the one hand, the overall profit margin declined for local firms as competition intensified. On the other hand, market specialization started to push the more capable companies to think about advancing to more value-added activities. The interviews with representatives of Chinese firms revealed that upgrading is a prevailing mode of thought. Some, such as Legend, have moved or plan to move into network equipment manufacturing, while others are looking at the software and service development area. This upward move, however, is filled with uncertainty. A lot depends not only on the technical ability and resources of the firms, but also on the competition with MNCs. Two cases illustrate the trajectories that local firms are likely to choose.
TongTech started as a system integrator in 1992. As the company encountered the use of middleware in a number of projects in the Chinese financial sector, it decided to focus on this area and direct its R&D to this field. In 1998, it abandoned the system integration business to become an independent software developer of middleware. It received venture capital from the American IDG group and hoped to be listed on China’s stock market. By the end of 2000, it had a total of 100,000 installations and accounted for 30 percent of China’s market share (company website: http://www.tongtech. com/ English). However, major American companies, including IBM, Hewlett- Packard, and BEA have been aggressively moving into the middleware fi\eld in China, viewing it as having great growth potential. Interviews with a senior official at TongTech suggested that TongTech’s advantages are in its lower cost and experience dealing with uneven infrastructures in the Chinese market. The official claimed that TongTech’s products are more stable than MNCs’ products because its development took into account their uneven network conditions and poor maintenance of networks in China. On the other hand, MNCs have a better business reputation overall, and interviews with China’s bank IT officers often suggested that they may purchase MNC products just to be safe. In addition, MNCs have been paying ever-greater attention to localization by establishing local R&D centers, and China’s infrastructure keeps improving, so an argument based on the technical merits of Chinese domestic products may not be valid for long, if at all. MNCs can also reduce their costs or make arrangements to preinstall their software to kill off local competitors. BEA, for example, entered the Chinese market with a price tag 10 times as high as TongTech’s, but has reduced its price to a similar level. IBM, the best seller of large network equipment in China, can bundle its middleware with the hardware. As a result, TongTech’s growth has not been spectacular since it gave up the system integration business. IDG group withdrew its venture capital in late 2000, a decision attributed to the conflict between a short- term goal of venture capital and TongTech’s long-term goal of becoming China’s indigenous brand of middleware (Xing and Chen 2001). TongTech argued that successful or not, by being in the game, it has at least forced MNCs to reduce their prices to a far lower level, therefore saving Chinese institutions millions of dollars (Ge 2001). TongTech’s experience suggests that by working in the application end of technology, Chinese local firms do have the opportunity and incentives to move up the ICT value chain. Yet direct competition with MNCs in their fields can be difficult and risky for local firms, despite their lower cost and other home market advantages.
Local firms, however, can choose a less-confrontational strategy and target areas of less interest to MNCs. One such example is ERP software for small- to medium-size firms. With the development of the Internet and e-commerce, an increasing number of Chinese enterprises have become interested in using various management software, including ERP. Many of China’s leading software and even system integration firms have declared their interest in developing in this direction. China’s largest accounting software firm, Ufsoft in Beijing, has also targeted this area for a strategic move. However, ERP software is an area that has been dominated by MNCs, particularly the German company SAP and the U.S. firm Oracle. Only large enterprises can afford ERP softtware from MNCs, though. Thus, the field is relatively open for Chinese firms to develop management software for a larger number of smaller firms. An Oracle executive commented on Ufsoft’s move:
Ufsoft has a larger market share in accounting software. The market share of SAP (the German ERP software provider) and ours together is not as big as theirs. It is natural that it wants to expand into total solution in enterprise management software, similar to what SAP has done. We are partners with Ufsoft, and we are also helping it to move to a higher level. Their database is ours, and we are talking about incorporating some of our modules into their future ERP system. However, their accounting system is still way too simple, having yet to incorporate the management flows of the enterprises. Our advantage is that we, as an American company, have accumulated long-time experience in this area. Ufsoft just started. But I think their expansion is natural, and they will need some more years of experience accumulation.
It is clear that Oracle does not see Ufsoft or other similar Chinese firms as threatening its market share yet. On the contrary, Chinese firms can be partners to help expand Oracle’s market reach by incorporating its products. Since the gap between what MNCs can offer and what most Chinese firms can afford is still large, there is some space for Chinese firms to move in without encountering major resistance from MNCs.
These cases indicate that the pressure of advancing on the ICT value chain by “import substitution” is not only present, but is also getting intense for Chinese local ICT firms. One of the key factors that has influenced the success of Chinese firms is MNCs’ cooperation, which varies depending on the strategic importance of the market and whether their market share is being threatened. Direct competition with MNCs’ established fields is risky. It is far easier to ride on MNCs’ backs or target areas that are unattractive to MNCs.
Region and Technology Change
Working with MNCs is certainly not a unique attribute of ZGC; indeed, what makes ZGC unique in China is its record of consistently and aggressively keeping up with advances in world technology and its ability to pioneer new technology and business ground in China. ZGC produced China’s first group of Min-Ying high-tech firms, the first internationally recognized patent on computer technology (the Chinese publishing system by the Founder Group; see Zhu and Huang 1999, 140), the first commercially successful domestic PC brand, and the first Internet portal (sina.com). This record could not have been achieved without the region’s deep indigenous R&D base; diverse mix of firms; and young, educated, and mobile labor force with a strong entrepreneurial tradition. Pack (2000) argued that indigenous R&D capacity has not been proved important for technological upgrading in developing countries. We found, on the contrary, that the local capacity for R&D is essential to assist local firms to solve problems and upgrade technology.
The concentration of China’s best research institutions in ZGC provides relatively abundant R&D resources for local firms to draw upon. Universities and the Academy of Sciences have been the incubators for many ZGC firms, and some of the major firms are still owned by universities. The core technology patents held by ZGC firms can be traced mainly to regions’ research institutions. Although the relationships between enterprises and universities are by no means harmonious, with the enterprises becoming increasingly independent, almost all such enterprises still rely heavily on university- trained employees, laboratories, and faculty. Other independent enterprises in ZGC also use graduate students, part-time workers, faculty consultants, and university laboratories in the area, often on an informal basis to solve technical problems. China’s decades of accumulation of R&D under the planning or market economy in ZGC have nurtured several generations of scientists and engineers in diverse technological fields. Although many of these scientists and engineers were limited by the rigid institutional culture they inherited from the planning economy, ZGC as a market-oriented region provided addi-tional opportunities to exploit their talent outside their institutional confinement.
The importance of local research facilities for local firms cannot be overstated. Even in the common cases in which the core technology came from abroad or MNCs in China, local firms still need additional R&D to piece together the puzzles if they want to deviate even slightly from the prescribed course of standard products. The inexpensive access to R&D capacity is especially significant for small firms without great R&D budgets. A professor in the Department of Computer Sciences at Peking University, who is a consultant for several software firms, complained bitterly about the lack of institutional structure governing outside work because most of the best students are almost always working for outside firms without the department’s knowledge, not to mention proper supervision. In a conversation with a not-atypical entrepreneurial graduate student at Peking University, the student claimed that he can solve any specific technical problem for outside firms because he can always find an expert in Peking University, or if not in Peking University, then through his friends in other surrounding campuses. This governance nightmare shows that formal and informal resources are indeed available, if not well organized and properly managed, for the local firms in the area.
Probing the technological expertise from academic institutions was the origin of the vast majority of ZGC technology firms, and the trend continued as the number of firms and their specializations multiplied. The diverse mix of IT-related companies provide specialized services for one another, although such collaboration is yet to be well developed and used.7 Interviewees in relatively large firms in ZGC do observe a constant flow of small specialized firms coming to their doors to sell their products and services. ZGC is also known for its young, well-educated, mobile, and entrepreneurial workforce, produced by its concentration of universities. According to the most comprehensive survey of human resources in the IT industry ever conducted in ZGC (Zhongguancun Administrative Government, see Table 4), covering 103,417 individuals working in 3,404 IT firms, the average age of employees is 28.8 years and the average age of senior managers is only 36.6 years. Of these employees, 54 percent have four-year degrees or higher, and the rest have two-year associate degrees. Local institutions contributed to a substantial portion of the workforce; 66.7 percent of individuals received their highest degrees from universities in Beijing, and 70 percent received them in science or engineering. Unlike the lifetime employment of state-owned companies, ZGC’s workforce is highly mobile. On average, 7 percent of the employees of high-tech companies rotate out of t\heir companies each year, and 11 percent come in. This population also has a marked entrepreneurial tendency, stemming partly from the entrepreneurial legacy of ZGC and partly from the inspiration of other high-tech regions in the world. Their energy was particularly on display during Beijing’s dot.com fever from 1998 to 2000. Beijing is the epicenter of China’s Internet revolution and hosts China’s three top Internet portals. According to a statistical report by the China Internet Information Center (2002), Beijing, with less than 1 percent of China’s population, accounted for over one-third of the Internet domain registration, 20 percent of the web sites, and about 10 percent of the users in China. Previous reports showed even higher concentrations in Beijing. Most of China’s leading young dot.com entrepreneurs have at some point lived or studied in ZGC (Liu and Li 2000). Although the fever died out after NASDAQ crashed in 2000, the desire for technology entrepreneurship seemed to persist among the people we interviewed. In addition, the above mentioned human resource survey also found that 15 percent of the respondents stated that entrepreneurship was the main direction for their career development. Angel (2000) argued that spatial clustering of a mobile and entrepreneurial labor force increased transactional efficiencies and facilitated the diffusion of information in Silicon Valley. Similar effects can be expected in ZGC.
The concentration of universities is also conducive to networking. Most of Beijing’s universities have residential campuses, promoting intense socialization among peer groups. The networks of students, alumni, and professionals are the most prominent informal institutions in ZGC. They also extend from ZGC to some other key Chinese IT cities and overseas (Walcott 2002; Saxenian 1999, 2002). Alumni networks are not only convenient conduits for information sharing, but also tend to be vehicles for entrepreneurship. The start-ups we interviewed in ZGC show marked concentrations of alumni from various universities. The vast majority of start-ups hosted by the incubator sponsored by Tsinghua University, for example, are Tsinghua alumni or students (interview).
In addition to personal networks and a mobile labor force, Beijing’s location as the country’s capital draws unusually high frequencies of international conferences, trade shows, and media exposure that create a fertile ground for the circulation and diffusion of information for ZGC. During the authors’ fieldwork period, the second half of 2002 and first half of 2001, IT-related international events, expos, seminars, workshops, and conventions were staged weekly in Beijing. In short, the extensive and vibrant networks, composed of the region’s large number of young professionals, are as crucial to the vitality of ZGC’s knowledge economy as is the R&D capacity of top universities or the Academy of Sciences.
For young professionals, ZGC represents a rewarding environment. About 60 percent of the respondents to the human resource survey expressed relative satisfaction (selecting 4 and 5 on a satisfaction scale of 1 to 5) for the region’s cultural environment, job opportunities, and training opportunities. They were most unhappy about the congested traffic and poor quality of the housing (only 23 percent chose 4 and 5 on transportation and 30 percent on the living environment). These findings suggest that despite its infrastructure problems, ZGC provides a reasonably good environment for an educated, technology oriented, and upwardly mobile community.
The Roles of the Chinese State and Beijing Location
Research in advanced countries has contended that the role of the state, even in mature market economies, is instrumental for technology development (Markusen 1998; Markusen, Hall, Campbell, and Deitrick 1991; Leslie 2000). Markusen and Yudken (1992), and Leslie (2000) argued that military spending, particularly the defense contracts and capital investment in universities and research centers in the United States, planted the original seeds for technological innovation in many localities. The role of industrial policies from the central government and local states in promoting technology development in newly industrialized economies has also been well documented (Gu 1996; Haggard and Cheng 1987; Johnson 1987; Vogel 1991; Kim and Nelson 2000; Jussawalla, 1999; Yeung 2000; Segal and Thun 2001; Howe 1997).
China’s state, with its legacy of centrally planned economy, is certainly more important than capitalist states. China’s ICT market could not have grown without the heavy promotion of the Chinese state. The location of ZGC in China’s capital gives ZGC firms’ critical influence on state policy, preference in receiving public R&D investments, and better access to China’s national market (Wang and Zhou 2001). During the first 15 years of ZGC development, the Chinese state took a relatively sideline position in the region (Segal and Thun 2001). It tentatively endorsed ZGC’s experiment with Guo-You-Min-Ying high-tech enterprises, but made it clear that enterprises were on their own (Baark 2001). This situation changed dramatically in 1999, however, when the central government, looking for indigenous-driven technological innovative centers, found ZGC a picturesque candidate.
The government’s change of heart led to the promotion of ZGC up the administrative hierarchy from under the urban district level to directly under the municipal government of Beijing and with close ties to the State Council. The promotion means that ZGC firms now hold a far more influential position in state policy making than was previously possible. It also led to a major increase in investment in R&D (for the universities and research institutes) and in improving the infrastructure. The Chinese government also made a series of policy changes, including reducing taxes and providing incentives for software and IC companies, expanding the allocation of capital to private firms, and relaxing Beijing’s strict household registration system so that local firms can hire university graduates even if they were not originally from Beijing.
Beyond state policy and investment, the power of the Chinese state is also excised through its control of and influence on China’s ICT market. Over the past 10 years, China’s high-tech industrial policies have gradually shifted from directly managing state-owned high-tech enterprises to encouraging market competition and emphasizing the state’s role in promoting and regulating the market (Baark 2001). With the slogan of “industrialization led by information technology” in the Tenth Five-year Plan, various state programs aimed at promoting the use of information technology in the government have led to the rapid growth in demand for ICT products in the Chinese marketplace (“The Flies Swarm In” 2000). From the interviews, the large institutional users of system integrators and software developers are mostly governmental institutions and large state-owned enterprises, first in banking and telecommunication sectors and later expanding to all significant social economic systems, such as transportation, postal services, customs, social welfare, hospital, military, education, and security. Since the headquarters of these national organizations are almost inevitably located in Beijing, firms in ZGC have a handy locational advantage to the market.
ZGC firms are also aggressive in lobbying the government to support their domestic brands. Kingsoft, an early private software developer, is a competitor of Microsoft in office products. Long plagued by the high piracy rate in China, it became the biggest winner in 2001, in the Beijing municipal government’s first open bid for office software in an effort to promote the legal use of copies in the government. In fact, among the seven bidding software companies, all six Chinese domestic firms gained contracts but Microsoft did not (China Business Report 2002). In this case, China’s massive bureaucratic system is seen by the local firms as the best hope for combating piracy, since governmental purchasing could guarantee a significant sales volume.
The reliance on markets controlled by the state and other state instruments does not stop ZGC firms from criticizing the state, however. With its concentration of China’s intellectuals and mostly nonstate enterprises, ZGC is arguably the most outspoken region in China. Complaints about the state’s ineffectiveness in protecting intellectual property rights and in ownership reform, on its insufficiencies in investment and R&D, and on its misguided investments in real estate are abundant. Every firm we interviewed expressed some form of dissatisfaction with the irregularities, abuses, and corruptions in the industry for which, they asserted, the government should take primary responsibility.
Complaints aside, the advantages of locating close to the state apparatus and state policy makers clearly outweigh the disadvantages of experiencing bureaucratic hurdles during the earlier periods of ZGC. These advantages, in combination with the region’s rich human resources; R&D facilities; and evolving networks among enterprises in hardware, software, and services, go a long way toward enhancing the innovative potential of firms. In the end, region matters for local firms even if the main sources of technology are external.
ZGC still has a long way to go, however, to become an internationally recognizable and competitive technology center. There are substantial technological and management gaps between ZGC firms and MNCs. Major flaws in the regional innovative environment are also apparent and cannot be expected to change in a short time (Liu 1999). In particular, ZGC faces the daunting tasks of establishing a fair and open legal framework to protect intellectual property rights, achieving information transparency, creating and stabilizing a venturecapital mechanism, and expanding business services. Many older ZGC firms are also facing massive reorganizations to avoid being left out of the rapidly evolving technological landscape. The central government’s negative attitude toward freedom of information on the Internet adds another layer of uncertainty. However, it is clear that ZGC firms have started to engage actively with their users, with MNCs, and with the government in an effort to construct the “accumulative feedback loop” in ICT technology in China. If China’s ICT market grows as it is expected to, it is not far-fetched to expect that, in time, ZGC will become internationally recognized as China’s Silicon Valley.
Conclusion
The literature on innovative regions has stressed the critical role of regional endogenous knowledge bases, networks, and institutions in sustaining innovation, but it offers little insight into the dynamics of technology in the developing countries where new sources of technology tend to be external. The literature on technology transfer has recognized the key role of MNCs, but does not accord sufficient appreciation of the variation in regional capacities to shape technology trajectories. In this article, we have argued that the literature on innovative regions provides a useful framework for understanding technological changes in high- tech regions in developing countries if we pay more attention to the relationship between external and internal players. We first analyzed the interaction between MNCs and local firms in a high- tech service cluster in ZGC, demonstrating that the relationships between MNCs and local firms are hierarchical, but also fluid and interdependent. On the one hand, MNCs dominate high-end technology, undermine the incentives for local innovation, and draw human resources away from local firms. On the other hand, MNCs depend on local firms for marketing and servicing their products and thus have transferred considerable technical and management expertise to local firms. The transfer, combined with local firms’ market expertise and R&D capacity, provides opportunities for their advance.
We then showed how regional resources and networks are critical in assisting local firms in learning technology. The presence of rich R&D institutions and the diverse mix of local firms provide complementary expertise that is not available from MNCs. The young, educated, and mobile labor force, with its strong entrepreneurial tradition, facilitates information sharing and start-ups. ZGC firms also benefit from their spatial proximity to the decision makers in the Chinese state and state-controlled organizations and enterprises.
The technology landscape in developing countries is highly uneven. Among high-tech regions in China, there are considerable differences in local resources, cultures, and institutions, as well as in their positions in MNCs’ global or China strategies. ZGC is a unique in its extraordinary endowment of indigenous R&D resources, its history as the birthplace of China’s nongovernmental technology firms, and its location in the national capital. Other prominent high-tech regions, such as Shanghai and Shenzhen in China, would demonstrate a different combination of factors involving MNCs, local firms, indigenous R&D institutions, and the local and central state. Further studies on these areas will help us to understand better the dialectical relationship between global forces and local accommodation and resilience.
Economic Geography 79(2): 129-152, 2003. (C) 2003 Clark University, http://www.clarku.edu/econgeography
This research was supported by grant SC8107 from the Pacific Cultural Foundation, a World Bank fellowship from the China Center of Economic Research at Peking University, and a Vassar College research grant. We thank the China Center of Economic Research at Peking University for providing institutional support for this project. In addition, we thank Professor Wang Jici of Peking University for her intellectual input and generous assistance. Several undergraduates at Peking University and Vassar College also provided graphic and editorial assistance. We also thank Henry W.- c. Yeung, George C. S. Lin, and three anonymous reviewers for their feedback on the first draft. Finally, we thank all the interviewees in Beijing who kindly shared their insights on China’s information communication technology industry. This article would not have been written without their open, candid, and thoughtful input.
1 IGT service companies are difficult to define, since many companies may be involved in both manufacturing and services. OECD (2001, 7) defines ICT service industries as those intended to enable the function of information processing and communication by electronic means. This definition would include companies dealing with wholesale ICT products, telecommunications services, and consultancy services, including various system integration and network solutions. This article uses the term to cover the above activities, as well as software firms, e-commerce firms, and manufacturing firms of customized equipment.
2 One has to be careful when defining innovation in the context of developing countries. Experience in newly industrialized economies in Asia, particularly Taiwan and South Korea, has shown that it is the acquisition and progressive mastering of technologies, rather than frontier innovation, that contribute to the dynamism and success of these economies (Kim and Nelson 2000). Hobday (2000), following many other scholars, defined innovation as a product or process that is new to the firm, rather than to t
