AMERICAN INTERNET COMPANIES' PREDICAMENT IN CHINA: GOOGLE, EBAY, AND MSN MESSENGER jia lu Abstract This article analyses American Internet companies' predicament in China from the perspective of the Internet users. Google, eBay, and MSN Messenger were selected to represent American companies, and were compared with their major local competitors - Baidu, Taobao, and Tencent QQ. The demographic changes in the Internet user base and different choices of information have-more users and have-less users explain the rise and fall of American companies at the Chinese market. Have-more users and have-less users are respectively related to the space of flows and the space of places in Castells' (2000) notion. Moreover, this study found that have-more users have higher mobility to switch between the space of flows and the space of places than have-less users. Giddens' (1991) theory of emancipatory politics and life politics explain how individual users' self-identities affect the competition between local and American companies as well as the overall development of China's Internet. Jia Lu is Assistant Professor at the School of Journalism and Communication at Tsinghua University, China; e-mail: lujiantu@gmail.com. rN ■ LO r^. si o rN oo £ £ 3 a 0 £ 1 * (R 0 C ! LO The impact of China's transformation to a socialist market economy and quickly expanding consumer class has been nowhere more apparent than in the domain of Internet development. As of June 2010, about 420 million Chinese people use the Internet, making it the largest Internet market in the world (CNNIC 2010). Since the turn of the century, the Internet market has witnessed the drastic competition between American and local companies. The competition started with American domination and gradually turned into success for the Chinese Internet companies. Consequently, some American companies chose to give up their business in China. In 2007, eBay sold out 49 percent of its subsidiary's shares and handed over its operation to Tom.com. In 2010, Google declared to close its searching services in Mainland China. This study aims to explore this dramatic change by analysing the usage patterns of Chinese Internet users. Google, eBay, and MSN Messenger are selected to represent American Internet companies because of their leading positions at the global market. They are compared with major local competitors, including Baidu in searching engine, Taobao in online shopping, and Tencent QQ in instant messenger. A large number of market reports are used to explore Internet usage patterns. They are from three major sources: China Internet Network Information Center (CNNIC), iResearch Consulting Group (iResearch), and China IntelliCon-sulting Corporation (CIC). CNNIC is the state Internet information centre of China, which is sponsored by Ministry of Information Industry and operated by Chinese Academy of Sciences. Both iResearch and CIC are leading private Internet research and consulting companies in China. In order to improve the research validity, this study adopts a technique of triangulation, in which the data from three sources cross-validate one another in order to diminish the biases derived from different ownerships of these sources. American and Local Internet Companies at the Chinese Market y China's Internet commercialisation started in early 1997 when Info-Highway Q Network, the earliest private ICP (Internet Content Provider) and ISP (Internet £ Service Provider), launched its operation. Since then, Chinese Internet users have increased from 0.62 million in 1997 to 420 million in 2010 (CNNIC 1997; 2010). The overall Internet market scale has reached RMB 39 billion by July 2010, a 272 percent increase from RMB 9.1 billion in 2003 (iResearch 2010b). The network's transcendence of national boundaries makes the Internet economy open to all foreign and local companies. Right after the initial commercialisation in 1997, American Internet ^ companies made their presence at the Chinese market. Table 1 shows that Google and eBay have earlier access to the Chinese market than Baidu and Taobao. MSN Messenger is one year later than its local competitor - Tencent QQ. 0 American companies established their advantages at the early stage of Chinese -l......l —. —. ^ inm ....... zo/i Q ~„ A .,L?~.. m^ A 1, c Internet development. In 2003, Google (34.8 percent) and eBay (72.4 percent) took the leading positions in their respective markets, and their local competitors - Baidu (30.7 percent) and Taobao (7.8 percent) - possessed the second positions (see Figure 1 and Figure 2). However, Google and eBay surrendered the leading positions to Baidu and Taobao in 2004 and 2005. From then on, their market shares continue to drop and their local competitors' keep increasing. By the end of 2009, Baidu's share reached 72 percent and Google's dropped to 23 percent while Taobao's share reached 84.4 percent and eBay's dropped to 4.6 percent. Compared to Google and eBay, MSN Messenger has a more disappointing performance. MSN has never become the No. 1 since its access to China in 1999. In contrast, Tencent QQ steadily occupies the dominant position. Figure 3 shows that Tencent's market share has never been lower than 70 percent and MSN's has never been over 20 percent. By the end of 2009, Tencent's share increased to 86.4 percent and MSN's dropped to 4.6 percent. The figures indicated that American Internet companies are trapped in a serious predicament at the Chinese market. They are facing strong challenges from local competitors. The situation seems to be even worse in the recent years as eBay and MSN have their market shares down to a single digit. The development trajectory of American Internet companies raises several interesting questions. For example, why did Google and eBay dominate the Chinese Table 1: A Brief History of American and Local Internet Companies Searching Engine Online Shopping Instant Messenger Time Google Baidu eBay Tao-bao MSN Messenger Tencent QQ Established 1998 2000 1995 2003 1999 1998 Access to Chinese market 1998: Available to Chinese users 2005: GOOGLE. CN established 2010 April: Quit the market of Mainland China 2000 2002: Merged EACHNET.COM and launched services in China 2006 December: Sold 49 percent of Eachnet's shares to Tom.com and quit Eachnet's operation 2003 1999: Available to Chinese users 2005: MSN.COM. CN established MSN Messenger in Chinese version officially released 1998 Figure 1. The Market Share Comparison between Google and Baidu Source: ¡Research (2006a, 2007b, 2009a,) and CNNIC (2006b, 2007b, 2009b, 2009c). Figure 2. The Market Share Comparison between eBay and Taobao Source: iResearch (2006b, 2008a, 2009a) and CNNIC (2006a, 2008, 2009d). Figure 3. The Market Share Comparison between MSN Messenger and Tencent QQ u S 5 0 £ * i * (R G C 00 rv Source: iResearch (2004, 2005, 2006b, 2007c, 2008b) and CNNIC (2006c, 2009e). market in early years? Why did they gradually lose the leading positions? How did Baidu and Taobao catch up and finally beat Google and eBay? Why did MSN Messenger have no chance to surpass Tencent QQ? The answers reside in the evolution of Chinese Internet user base. Chinese Internet Users: Have-more and Have-less Over one decade's development, Chinese Internet witnessed substantial changes in users' demographics. CNNIC (2007a) summarised these changes into five tendencies: (1) younger age (under 25 years: 41.9 percent in October 1997 to 57.1 percent in January 2010); (2) lower education (below college diploma: 11 percent in January 1999 to 76.7 percent in July 2010); (3) lower income1 (7 percent in October 1997 to 58 percent in July 2010); (4) less developed areas (middle and west regions: 26.5 percent in October 1997 to 42 percent in March 2009); and (5) students (13.9 percent in October 1997 to 30.7 percent in July 2010) (see CNNIC 1997; 1999; 2009a; 2010). Demographic changes affect users' Internet access and usage. For example, the ac- C^ cess at Internet café has increased from 3 percent in January 1999 to 33.6 percent in ^ July 2010 (CNNIC 1999, 2010). Internet café access has exceeded workplace access to become the second most significant access location after home. Meanwhile, the entertainment function becomes dominant in the Internet usage. In January 1999, 35 percent of the users go online for entertainment, such as gaming and music (CNNIC 1999). In July 2010, 70.5 percent play online games, 82.5 percent listen to music, and 63.2 percent watch videos (CNNIC 2010). Recognising these demographical changes of the Internet users in China, Cartier, Castells, and Qiu (2005) purposed the rise of "information have-less" as a new class of information users, including millions of rural residents, rural-to-urban migrants, manufacturing and low-end service sector workers, laid-off or unemployed workers, pensioners, young people with low incomes, and some members of the new urban middle class who retain old consumer habits. Qiu (2007) argued that information have-less users generally are people with lower socio-economic status, such as lower education and lower income. The term of information have-less users was often used interchangeably with working-class users, low-end users, or grass-root users. The definitions of information have-less users slightly vary in a number of market reports and can be synthesised into two conditions: 1) monthly income below RMB 1,500 and 2) high-school diploma or below (see CIC 2006; 2008; CNNIC 2006a; 2006b; 2006c; 2007b; 2008; 2009b). Demographic changes in the Chinese Internet user base indicate that information have-less users have been gradually developing into the majority of the Internet users with 58 percent in low income and 76.7 percent in low education. The other three tendencies also support Cartier et al.'s (2005) position that information have-less users are more likely to be found in younger age, students group, and less developed areas (rural regions or middle and west regions). The opposing term of information have-less users is information have-more users, which was often used interchangeably with high-end users, middle/upper class users, and white-collar users. The definitions of information have-more users can be synthesised into three conditions: 1) monthly income above RMB 2,000 on a national basis or RMB 3,000 in central cities (i.e., Beijing, Shanghai, and Guangzhou) and first-level cities (i.e., provincial capitals), 2) between 23/25 and 40 years old, and 3) college diploma degree or above (see CIC 2006; 2008; CNNIC 2006a; 2006b; 2006c; 2007b; 2008; 2009b). Different Choices of Have-more Users and Have-less Users Research indicated that have-more users like to choose the services provided by American Internet companies and have-less users like to choose the services of local Internet companies. For searching engine, CNNIC (2006b) reported that 46.5 percent of high-end users in Beijing, Shanghai, and Guangzhou choose Google and 39.4 percent choose Baidu. Among the users under the age of 23, 52.7 percent use Baidu and 25.6 percent use Google. Among the users with high school diploma or below, 61.7 percent use Baidu and 38.8 percent use Google. Among the student users, 46.2 percent use Baidu and 25.8 percent use Google. Among the professional users, 32.6 percent use Baidu and 48.5 percent use Google. Among the users with the monthly income below RMB 3,000, 63.5 percent use Baidu and 20.3 percent use Google. According to CIC (2008), Baidu is more frequently used in middle and western cities (91.2 percent and 90.5 percent) than in Beijing, Shanghai, and Guangzhou (85 percent), and Google is more frequently used in Beijing, Shanghai, and Guangzhou (37.2 percent) than in middle and western cities (23.7 percent and 22.7 percent). For instant messenger, iResearch (2007e) reported that the average income of QQ users is RMB 2,247 per month, and MSN users' is RMB 2,818 per month. By age, MSN users concentrate between 25 and 30 years old (44.2 percent), and QQ users concentrate between 18 and 24 years old (36.5 percent). By location, MSN is more frequently used in Beijing and Shanghai (77.2 percent) than in the other regions (51.3 percent), and QQ is more frequently used in the other regions (95 percent) than in Beijing and Shanghai (88 percent). According to CNNIC (2006c), 54.3 percent of high-end users in Beijing use MSN and 42.2 percent use QQ while 66 percent of high-end users in Shanghai use MSN and 31.9 percent use QQ . For online shopping, CNNIC (2008) reported that eBay is more frequently used by the people with a long-time Internet experience, and Taobao is more attractive to the new users whose Internet experience is below three years. Among the users under 24 years old, 42.6 percent use Taobao and 33.5 percent use eBay. Among the student users, 30.3 percent use Taobao and 25 percent use eBay. Among the users with the monthly income below RMB 2,000, 48.9 percent use Taobao and 43.1 percent use eBay. Among the users with the monthly income above RMB 5,000, 22.1 percent use eBay and 15.8 percent use Taobao. By location, Taobao enjoys the higher brand awareness in the other cities (91 percent) than in Beijing, Shanghai, and Guangzhou (87.4 percent), and eBay retains higher brand awareness in Beijing, Shanghai, and Guangzhou (44 percent) than in the other cities (40.1 percent). Have-less ICTs The figures above indicate demographic distinctions between have-more users y and have-less users and their different choices of Internet services. Cartier et al. Q (2005) referred to the technologies and services used by information have-less users as "have-less ICTs," which have three important characteristics: low cost, critical functions, and limited mobility. This study examined local and American services against these three characteristics and found that they meet the respective needs of have-less and have-more users. 0 £ C > Low Cost Cost is critical for have-less users to choose Internet services. This concern is more obviously expressed in online shopping. According to iResearch (2009b), 59.2 yj percent of low-end netizens expressed that low price is the primary reason of their online shopping and 53.6 percent of high-end netizens chose fast and convenient delivery as the primary reason. Taobao, since its beginning, sticks to a free-charge policy. In Taobao, neither buyers nor sellers are required to pay any fee. However, eBay followed its successful model in the United States and insisted charging a variety of fees, including transaction fee, inventory fee, search ranking fee, registration fee, and window show fee. This policy has not been changed until January qq 2006 when eBay's market share has been largely taken over by Taobao. Gong (2006) suggested that eBay's model is based on an assumption that users, mostly have-more users, are willing to pay for high quality services and their payments consist of an important income source for the company. eBay's model failed at the later stage of Chinese Internet development when the majority of the user base turned to be information have-less people who are concerned with low cost. The cost distinction between Taobao and eBay includes not only access expense but also the products sold at their websites. According to CNNIC (2006a), eBay and Taobao are greatly different in the sales of several product categories. For example, Taobao beats eBay in garment (29 percent vs. 24 percent), cosmetics and jewellery (26 percent vs. 17 percent), and phone cards and virtual money (15 percent vs. 10 percent). eBay beats Taobao in computers (32 percent vs. 12 percent), books (18 percent vs. 12 percent), audio and video (12 percent vs. 7 percent), and paintings, music instruments, artefacts, and antiques (11 percent vs. 8 percent). In general, eBay's products are more valuable than Taobao's. For example, Taobao's phone cards and virtual money are considered typical have-less ICTs by Cartier et al. (2005). In contrast, eBay's computers allow their consumers to possess advanced ICTs at home instead of going to Internet cafés. Product categories reflect not only the cost concern but also lifestyles and tastes. Featherstone (1987) suggested that the people in high economic and educational status are more likely to engage in the consumption of high cultural goods, services and experiences (i.e., museum visit, concert going, and reading) than the people in low economic and educational status. Bourdieu (1984) pointed out that taste in cultural goods, including in high cultural practices and in lifestyles and consumption preferences, functions as a marker of cultural identities of a class. eBay focuses on knowledge intensity and high cultural products (i.e., computers, books, audio and video, and artifacts) and Taobao emphasises daily necessities (i.e., garments and cosmetics). iResearch (2009b) reported that white-collar netizens are different from the rest in product types they purchase online. White-collar netizens like to buy books and audio-video products (54 percent vs. 44.9 percent) and IT digital products (39.7 percent vs. 30.8 percent) while other netizens prefer garments (47.9 percent vs. 38.8 percent), virtual cards (37.4 percent vs. 29.3 percent), and cosmetics (29.4 percent vs. 23.8 percent). This observation supports Yao's (2007) position that eBay's target customers are white collar workers who are middle-aged, well-paid, internationalised, have technology knowledge, and pay attention to high cultural practices, and Taobao's are blue collar workers, young, less-educated, and cost-sensitive. Critical Functions The second distinction between American and local is about critical functions of their services - the professional function for have-more users and the entertainment function for have-less users. According to iResearch (2007e), 77.8 percent QQ users adopt QQ for the purpose of entertainment and 60.8 percent MSN users for the purpose of working. 65.4 percent users talk about the work-related issues on MSN and 54.6 percent users talk about the issues in their daily lives on QQ. 62.3 percent users adopt QQ to communicate with their classmates, friends, and family members, and 58 percent users adopt MSN to communicate with their colleagues and the people they encounter in work. QQ is frequently used at home (88.1 per- cent), Internet cafés (42.3 percent), and schools (31 percent), and in evenings (59.7 percent), weekends (50.9 percent), and holidays (45.7 percent). MSN is frequently used at workplace (81.5 percent) and in working hours (66 percent). iResearch (2007e) identified QQ as an entertainment instant messenger and MSN as a professional instant messenger. For searching engine, 83 percent users adopt Google for the purpose of working and 76 percent adopt Baidu for the purpose of entertainment (CIC 2006). According to CNNIC (2009c), Baidu exceeds Google in searching for music (40.5 percent vs. 32.7 percent), movie and television (31 percent vs. 25.3 percent), and gaming (18 percent vs. 8.7 percent). Google exceeds Baidu in searching for professional materials (30.2 percent vs. 28.5 percent), travel information (11.2 percent vs. 6.6 percent), information related to enterprises and products (7.4 percent vs. 6.3 percent), software programs (4.2 percent vs. 3.8 percent), and information in foreign languages (3.9 percent vs. 1.2 percent). Limited Quality The third distinction between American and local services refers to limited mobility, including three layers: limited capability of transcending geographical boundaries, limited quality of services and limited protection of users' interests, and local orientation (Cartier et al., 2005). Because of the Internet's inherent capability of transcending geographical limits, limited mobility for Internet services is mainly expressed in the last two layers. First, American companies are better than local companies in providing high-quality services and protecting users' interests. For instant messenger, MSN is better to protect online security than QQ. 60.6 percent users experienced virus infection on QQ, and 56.7 percent on MSN. 85.1 percent QQ users once received disturbing messages or advertisements sent by strangers, and 78.5 percent for MSN users (CNNIC 2006c). QQ has lower ratings of users' satisfaction than MSN in a number of security preventions from strangers (3.3 vs. 3.6), virus infection (3.0 vs. 3.7), advertising (3.0 vs. 3.7), and username stealing (2.9 y vs. 3.9) (iResearch 2007e). For searching engine, CIC (2008) reported that Google excels Baidu in all major service indicators, including relevance of searched results (34.6 percent vs. 29 percent), source diversity of searched results (35.4 percent vs. 21.6 percent), ranking of searched results (29.6 percent vs. 22.1 percent), security of searched results (29.1 percent vs. 11.9 percent), technological innovation (35.4 percent vs. 17.2 percent), and enterprise images (34.8 percent vs. 17.5 percent). 0 £ C > Local Orientation Local services are more locally-oriented than American ones. As mentioned before, QQ is used primarily to maintain users' local relations with their classmates, friends, and family members. Meanwhile, both Baidu and QQ allow users to build up online local communities. Baidu has a section named "Local Friendship," which includes all the provinces in China. Each provincial board contains a number of regions in this province. The people coming from the same region can get together and discuss local issues they are concerned with. Another local section on Baidu is called "Campus" where users can find their elementary schools, middle schools and universities, and talk about the issues related to their schools and classmates. qq The similar discussion boards are found on QQ: "Alumni" and "Same City." They more or less resemble Baidu's "Campus" and "Local Friendship." Postill (2008) suggested that the Internet becomes more local as online communities strengthen local identities and local connections of their users (also see Davies and Crabtree 2004). No such local communities are found at Google and MSN. Google and MSN focus on providing services for the professional purpose. The professional function of Google and MSN constitutes Giddens' (1990) notion of "expert systems." Expert systems refer to the systems of technical accomplishment or professional expertise on which individuals can develop their trust, for example, the building standards that allow trust in the integrity of our houses, the engineering practices that allow trust in the integrity of our cars, and the medical standards that allow trust in the integrity of our doctors. Giddens (1990) suggested that an expert system is a disembedding mechanism because it removes local or immediate spatial-temporal contexts from social relations that impart expert knowledge. The development of expert systems transforms what would once have been the expertise of one known local expert into an arcane body of knowledge and rules that are systematically produced and impersonally tested to provide guarantees of universal expectations across large time-space distance. The professional use of Google and MSN looks for technical applications and professional expertise across geographical distances and transcends users' local identities to meet their universal needs derived from their professions. Taobao's local orientation is enabled by "Taobao Wangwang," an instant messenger to allow the direct communication between sellers and buyers. On "Taobao Wangwang," users can see and talk to one another at real time. The direct communication is strongly forbidden by eBay, because it worries that sellers and buyers bypass its platform to complete their transactions offline. In contrast, Taobao has a free-charge policy and does not care about offline transactions very much. According to CNNIC (2006a), 89 percent of Taobao users use "Taobao Wangwang" and 62.6 percent have offline transaction experiences while 61 percent of eBay users use email to communicate and 24.9 percent have offline transaction experiences. The communication channels allowed by Taobao and eBay reflect different trust relations between sellers and buyers. Giddens (1990) proposed two dynamics to build up trust relations - facework commitments and faceless commitments. Facework commitments refer to trust relations that are sustained by or expressed in social connections established in local circumstances in which participants are co-presented. Trust in persons is developed on the basis of facework commitments, in which participants seek perceptual indicators of the integrity of others through some kind of one-on-one realtime communication. Faceless commitments, however, develop faith in abstract systems, for example, expert systems. Trust in systems is developed on the basis of faceless commitments, in which faith is sustained in the workings of technical accomplishment and professional expertise. Faceless commitments do not require co-presence of participants in specific local contexts. According to Giddens (1990), facework commitments are bounded by local contexts to represent local orientation in the modern society, and faceless commitments rely on the disembedding mechanism of expert systems to transcend local identities. For eBay's users, their trust relations depend on faceless commitments that choose to trust the expert system under the brand of eBay, which has already been well known for advanced technologies and excellent expertise in the field of online shopping. For Taobao's users, their trust relations depend on facework commitments that choose to trust individual persons and seek the indicators of people's integrity through one-on-one communication enabled by "Taobao Wangwang" or in offline transactions. CNNIC (2006a) reported that about 40 percent of Taobao's buyers are most concerned with their feelings or impressions in communication with sellers, and only 20 percent for eBay's users. The Space of Flows and the Space of Places The correlations between users and services reflect Castells' (2000) notion of the space of flows and the space of places. The space of flows refers to the material organisation of time-sharing social practices that work through "flows" that are defined as "purposeful, repetitive, programmable sequences of exchange and interaction between physically disjointed positions held by social actors in the economic, political, and symbolic structures of society" (Castells 2000, 442). The space of flows is constituted by three major layers of material supports - a circuit of advanced electronic exchanges, nodes and hubs in the circuit, and spatial organisation of dominant elites. Electronic exchanges include a variety of advanced ICTs, such as micro-electronics-based devices, telecommunications, computer processing, broadcasting systems, and high-speed transportation enabled by information technologies. Nodes and hubs function to coordinate the smooth interaction of all the elements of the network as well as link the locality-based activities and organisations with the whole network. By nodes and hubs, Castells (2000) referred to global cities with advanced knowledge and services, which constitute the production site of the informational, global economy and direct the development of local societies and economies. Spatial organisation of dominant elites organises the elites around the world into a cohesive social group that is unified by a shared culture superseding the historical specificity of each locale. Castells (2000) revealed that the space of flows is limited to technocratic-finan-y cial-managerial elites and does not permeate down into the whole realm of the network society. The majority of people are still living in the space of places where people's life and experience are rooted in their places, cultures, and histories. Elites in the space of flows are cosmopolitan and people in the space of places are local. In this study, information have-more users are related to the space of flows. First, high quality of American Internet services facilitates have-more users' access to the global circuit of advanced electronic exchanges, for example, MSN's good protection of online security and Google's excellence in searching quality. Second, have-more users concentrate in global cities, such as Beijing, Shanghai, and Guangzhou, which constitute nodes and hubs in the space of flows and possess necessary resources to support complex time-sharing activities in the space of flows. Meanwhile, high education of have-more users contributes to the advanced knowledge base in global Airi Af I Wirrl ki^tta nfrtrf t^ o t*ri /^i « -^rn in f Ki^ n ri n I /^v/ytmi n ri rtr* rtr c > cities. Third, have-more users participate in the spatial organisation of dominant elites. On one hand, the disembedding mechanisms supported by American services lift have-more users out of local contexts and build up their connections with the other elites around the world, for example, eBay's faceless commitments in trust building and the professional function of Google and MSN. On the other hand, qq have-more users are engaged in pursuing a shared lifestyle to unify the symboli- cally secluded community of global elites (see Castells 1997; Featherstone 1990; Giddens 1991; Tomlinson 1994; 1999). LO Hassan and Katsanis (1994) noted the rise of global elite lifestyle that is featured with professionalism, leadership, exclusivity, high quality, and status. Global elites often differentiate themselves through buying and using luxurious products with prestige images that fit the expectation of being recognisable and universal on a global scale. iResearch (2007a) reported that Chinese white collar netizens (54.9 percent) favor personalised, distinguished, and recognisable products and services. Among them, 81.5 percent believe that vacation is an indispensible expense and 70.3 percent believe that overdraft consumption is very natural. Meanwhile, white collar netizens are more likely to purchase luxurious products than the other netizens. iResearch (2010a) reported that white collar netizens exceed the other netizens in terms of the experience (92.2 percent vs. 89.2 percent) and the intension (90.5 percent vs. 85.3 percent) of purchasing luxurious products. The consumption of luxurious products represents a distinctive lifestyle Chinese white collar neti-zens pursue. According to iResearch (2010a), 44.7 percent of white collar netizens expressed that personal interest/pursuit of quality lifestyles is the primary reason for them to purchase luxurious products. This number is only 25.7 percent for the other netizens. 37.1 percent of white collars reported that purchasing luxurious products has become a regular behaviour in their life, and this number goes up as age, income, and education increase. While have-more users actively participate in the space of flows, have-less users are attached to the space of places. First, low quality of local Internet services, to some extent, limits have-less users' access to the global circuit of advanced electronic exchanges, for example, QQ's security problems and Baidu's struggles for improving searching quality. Second, have-less users are widely spread in rural and less developed regions that do not have necessary resources to support complex time-sharing activities in the space of flows. Meanwhile, low education prevents have-less users from participating in the knowledge-based space of flows. Third, have-less users are allowed to maintain local connections and identities by locally-oriented services, for example, "Taobao Wangwang" and online local communities of Baidu and QQ. Linkages between the Space of Flows and the Space of Places So far, the study found that have-more users who like American services are related to the space of flows and have-less users who like local services are related to the space of flows. The distinctions between Internet services and between spaces seem to suggest a dichotomised view. However, Castells (1999) rejected the simplistic separation and pointed out the linkages between the space of flows and the space of places "under which people are finding ways to enter the space of flows without leaving the space of places, or to be more precise, to go back and forth easily" (Stalder 2006, 151). In this study, the linkages between the space of flows and the space of places are expressed in the fact that a certain number of have-more users simultaneously use local and American services. The simultaneous use is indicated by low loyalty of American services. Users' loyalty is a concept to assess the degree to which users stick to a particular brand of 0 c > Internet services. It is calculated by comparing the change in the number of users who adopt a brand as the first choice within past six months. Higher loyalty means fewer users who change first-choice Internet services in past six months. For searching engine, Baidu has the highest loyalty of 94.8 percent exceeding Google's 79.2 percent (CNNIC 2009c). In addition, 26.1 percent users in Beijing, Shanghai, and Guangzhou simultaneously use Baidu and Google on a regular basis, and this number drops to 17.3 percent and 16.7 percent in provincial capitals in middle and western regions (CIC 2008). For online shopping, Taobao has the highest loyalty of 94.6 percent exceeding eBay's 65.7 percent (CNNIC 2009d). 79.3 percent of Baidu's users only shop at Taobao and 45.6 percent of eBay's users only shop at eBay. In addition, about 27 percent of eBay's users expressed that they also go to Taobao (CNNIC 2008). For instant messenger, QQ has the highest loyalty of 97.3 percent and MSN's loyalty is 88.1 percent (iResearch 2006). According to CNNIC (2009d), 36.6 percent of users adopt only one brand of instant messenger that is predominantly QQ, and 28.3 percent of users simultaneously adopt two brands of QQ and MSN. High loyalty of local services indicates that have-less users constantly stick to local services. Low loyalty of American services indicates that some have-more users switch between different services for their first choices. Here, what needs to be mentioned is that the concept of loyalty only describes the situation of users' first choices and low loyalty of American services does not mean that some have-more users already give up American services. Instead, the figures above show that a certain number of have-more users choose both local and American services at the same time. The different loyalty rates suggest that have-more user have higher mobility than have-less users to switch back and forth between the space of flows and the space of places. High mobility constitutes the linkages between two spaces and allows have-more users to enter the space of flows without leaving the space of places. Emancipatory Politics and Life Politics: Lifestyle Choices and Identities of Users J The mobility difference between have-more and have-less users reflects Giddens' position about identity formation in the modern society: The more tradition loses its hold, and the more daily life is reconstituted in terms of the dialectical interplay of the local and the global, the more individuals are forced to negotiate lifestyle choices among a diversity of options (Giddens 1991, 5). In the present study, have-less users are attached to the space of places where their life are largely shaped ^ by local conditions, such tradition, culture, and history. On the other hand, have-more users are related to the space of flows where a globally shared elite culture tends to supersede local traditions. Thus, local conditions have less control over have-more users than have-less users. According to Giddens' theory, have-more users are forced to negotiate their lifestyles between local and global options. The result of the negotiation is high mobility of have-more users between local and American services. Giddens (1991) suggested that lifestyle choices represent people's self-identities that are shaped by emancipatory politics and life politics. Emancipatory politics qq is a politics of life chances that liberate people by giving them the conditions to make choices and life politics is a politics of lifestyle choices that build up people's identities through self-actualisation. In this study, the wide adoption of ICTs creates r^ necessary conditions of emancipatory politics for Chinese users to have equal access ^ to different Internet services. With a certain level of autonomy, life politics enable users to make their choices more freely. Their identities are formed in the process of lifestyle choices as a means of self-actualisation. As a result of life politics, have-more users choose American services and have-less users choose local services. However, Giddens (1991) suggested that emancipatory politics is not merely the preparation for the emergence of life politics. Instead, their relation is dialectic. First, all choices of life politics reflect corresponding problems of emancipatory politics. Although have-less users are liberated by enhancing chances of equal access to Internet services, their choices are restricted to some extent by the other material inequalities in emancipatory politics, such as income and education. For example, the cost concern makes have-less users choose Taobao for free services, and low education prevents have-less users from the professional use of the Internet. Second, life politics allow free choices and affect emancipatory politics. In this study, have-less users' choice of local services creates the conditions of emancipatory politics to enhance the chances for local companies to develop at the Chinese market. The increasing access of have-less users to the Internet gradually transforms power and resource distribution between local and American companies. Consequently, local companies dominated the market and American companies were marginalised until the exit from China. A Preliminary Assessment of Internet Development in China Emancipatory politics and life politics explain how Internet users' self-identities transformed the overarching structure of China's Internet market. The relation between self-identities and the market structure reflects Castells' (2000) conception of the primary conflict in the networked society - the bipolar opposition of the Net and the Self. The Net refers to the placeless network logic around which dominant social processes are organised, for example, the Internet market in China. The Self denotes the practices a person uses in reaffirming social identity and meaning in a continually changing cultural landscape, for example, Internet users' choices as a means of building up self-identities. The complex interaction between the Net and the Self establishes the distinction between the space of flows and the space of places. Castells (2000) argued that the space of flows and the space of places constitute two major dynamics of social process to jointly shape the networked society. The current situation at the Chinese market illustrates that the space of places exceeds the space of flows to dominate the Internet development in China. This observation supports Stalder's (2006) claim that the dynamics of the space of flows no longer always supersede the dynamics of the space of places as the space of flows is becoming a part of everyday life that is subject to the dynamics of the space of places. As a result, Goldsmith and Wu (2006) argued that what we once called a global Internet has been reduced to a collection of nation-state networks - networks still linked by the Internet protocol but separated by national borders. They explained: Internet is splitting apart and becoming bordered. Far from flattening the world, the Internet - its language, its content, its norms - is conforming to local conditions. The result is an Internet that differs among nations and regions that are increasingly separated by walls of bandwidth, language, and filters (Goldsmith and Wu 2006, viii). Goldsmith and Wu (2006) introduced two types of forces to form the bordered Internet. One refers to the top-down pressures from the governments that are imposing national laws and technological applications on the Internet within their borders. The top-down pressures were often used to explain the difficulties of American Internet companies at the Chinese market, particularly in the recent case of Google's quit (see Einhorn, Elgin, and Hof 2005; Gutmann 2004; Magnier and Menn 2005; Thompson 2006; Nakashima Mufson and Pomfret 2010; Krishnan 2010; Conway 2010). Another force includes bottom-up pressures from individual users who demand an Internet that corresponds to local preferences, and from the Internet companies who provide services to satisfy these demands. That is what this study presents. These two forces work together to shape the Internet development in China. Firstly, the Chinese government actively control information flows on the Internet in order to maintain social stability under the name of building up a harmonious society. The market domination of local companies facilitates the government's control, because the government has more methods, latent or manifest, to manipulate local companies than American ones (see Liu and Lee 2010; Synovitz 2010; Young and Adegoke 2010). Secondly, Chinese users' strong preference to online entertainment realises imperial Rome's strategy of "bread and circuses," which greatly reduces the chances for Internet users to collectively participate in political activism and social movements by keeping them entertained at home and off the streets (see Lagerkvist 2008; Mertha 2005; Wang and Zhu 2003). Thirdly, the local orientation of Chinese Internet services together with the emerging bordered Internet serves as an effective and useful communication tool for the government y to promote nationalistic sentiments that are often appealed to for building up the legitimacy of the government and the Party (see Nathan 2003; Gries 2005; Wu 2007; Lagerkvist 2008). As a matter of fact, a reciprocal relationship exists between top-down pressures and bottom-up pressures. The pervasive criticism about the Chinese government's control of the Internet only sees one side of a coin. Meanwhile, it overlooks Chinese users' choices, which objectively provide the ground Q for the governmental control. £ > (6 Note: 1. CNNIC (1997) and CNNIC (2007a) adopted different methods to measure the Internet user's income. CNNIC (1997) used per capita monthly income of the Internet user's household, and CNNIC ^ (2007a) used monthly personal income of individual Internet users. CNNIC (2007a) defined low 0 income as monthly personal income below RMB 1,500. CNNIC (1997) did not define low income. According to the National Bureau of Statistics (n.d.), however, per capita monthly income of urban household was RMB 430 in 1997. Thus, low income in the 1997 report can be roughly defined as per capita monthly income of the Internet user's household below RMB 430. According to CNNIC (1997), 7 percent of the Internet users had per capita monthly income below RMB 400. So 7 percent was used as the percentage of the Internet users with low income in the 1997 survey. 00 00 References: cn Bourdieu, Pierre. 1984. Distinction: A Social Critique of the Judgment of Taste (R. Nice, trans.). 00 Cambridge, MA: Harvard University Press. Castells, Manuel. 1997. The Power of Identity. Maiden, MA: Blackwell. Castells, Manuel. 1999. Grassrooting the Space of Flows. Urban Geography 20, 4, 294-302. Castells, Manuel. 2000. The Rise of the Network Society (2nd ed.). Malden, MA: Blackwell. Cartier, Carolyn, Manuel Castells, and Jack Linchuan Qiu. 2005. The Information Have-less: Inequality, Mobility, and Translocal Networks in Chinese Cities. Studies in Comparative International Development 40, 2, 9-34. Conway, Edmund. 2010. Google Will Quit China Unless Web Censorship Ends. Telegraph, January 12, 2010. CIC. 2006. CIC Report on Chinese Search Engine Users, September 2006. CIC. 2008. CIC Survey Report on Chinese Search Engine Users, September 2008. CNNIC. 1997. The 1st Statistical Survey Report on the Internet Development in China, October 1997. CNNIC. 1999. The 3rd Statistical Survey Report on the Internet Development in China, January 1999. CNNIC. 2006a. Survey Report on C2C Online Shopping in China, May 2006. CNNIC. 2006b. Survey Report on Search Engine Market in China, September 2006. CNNIC. 2006c. Survey Report on Instant Messenger Market in China, December 2006. CNNIC. 2007a. The 20th Statistical Survey Report on the Internet Development in China, July 2007. CNNIC. 2007b. Survey Report on Search Engine Market in China, September 2007. CNNIC. 2008. Survey Report on Online Shopping in China, June 2008. CNNIC. 2009a. The 23rd Statistical Survey Report on the Internet Development in China, March 2009. CNNIC. 2009b. Chinese Search Engine Users Research Report, March 2008. CNNIC. 2009c. Survey Report on Chinese Searching Engine Users, September 2009. CNNIC. 2009d. The Market Report of Online Shopping in China, November 2009. CNNIC. 2009e. Survey Report on Chinese Instant Messenger Users, December 2009. CNNIC. 2010. The 26st Statistical Survey Report on the Internet Development in China, July 2010. Davies, William and James Crabtree. 2004. Invisible Villages: Technolocalism and Community Renewal. Renewal 12, 1, 40-47. Einhorn, Bruce, Ben Elgin and Robert D. Hof. 2005. The Giant Web Race; Internet Use is Exploding in China, so U. S. Giants like Yahoo! and eBay are Hunting Harder for Deals. Business Week, June 13. Featherstone, Mike. 1987. Lifestyle and Consumer Culture. Theory, Culture & Society 4, 1, 55-70. Featherstone, Mike. 1990. Global Culture: An Introduction. Theory, Culture & Society 7, 2, 1-14. Giddens, Anthony. 1990. The Consequences of Modernity. Stanford, CA: Stanford University Press. Giddens, Anthony. 1991. Modernity and Self-identity: Self and Society in the Late Modern Age. Stanford, CA: Stanford University Press. Gong, Wenxiang. 2006. Ten Reasons for American Internet Companies' Failure in China. IT Times Weekly, December 11, A1. Goldsmith, Jack and Tim Wu. 2006. Who Controls the Internet: Illusions of a Borderless World. Oxford: Oxford University Press. Gries, Peter Hays. 2005. China's New Nationalism: Pride, Politics, and Diplomacy. Ewing, NJ: University of California Press. Gutmann, Ethan. 2004. Losing the New China: A Story of American Commerce, Desire and Betrayal. San Francisco, CA: Encounter Books. Hassan, S. Salah and Leas Prevel Katsanis. 1994. Global Market Segmentation Strategies and Trends. In S. S. Hassan and E. Kaynak (eds.), Globalization of Consumer Markets: Structures and Strategies, 47-61. New York: International Business Press. iResearch. 2004. China Instant Messenger Research Report, January 2003. iResearch. 2005. China Instant Messenger Research Report, February 2004. iResearch. 2006a. China Online Search Annual Report, January 2005. iResearch. 2006b. China C2C E-commerce Research Report, March 2005. iResearch. 2006c. China Instant Messenger Research Report, March 2005. iResearch. 2007a. White Collar Netizen Research Report, January 2007. iResearch. 2007b. China Online Search Research Report, March 2006. iResearch. 2007c. China Instant Messaging Research Report, March 2006. iResearch. 2007d. China Internet Industry Research Report, March 2007. iResearch. 2007e. Aliwangwang's Competitiveness Research Report, August 2007. iResearch. 2008a. China Online Shopping Research Report, 2007-2008. U iResearch. 2008. China IM Research Report 2007-2008. iResearch. 2009a. China Search Engine Research Report 2008-2009. iResearch. 2009b. Survey Report on Chinese White Collars' Online Shopping, October 2009. iResearch. 2010a. Survey Report on White Collars' Consumption of Luxurious Products in China, Q March 2010. iResearch. 2010. The Core Data of the Internet Economy in China in the Second Quarter of 2010, June 2010. Krishnan, Ananth. 2010. Google Threatens to Quit China, Citing Cyber Attacks. The Hindu, (January 13). ^ Lagerkvist, Johan. 2008. Internet Ideotainment in the PRC: National Responses to Cultural Globalization. Journal of Contemporary China 17, 54, 121-140. Liu, John and Mark Lee. 2010. Google China Exit would Hand Baidu Control of Market. Bloomberg News, January 14. Magnier, Mark and Joseph Menn. 2005. As China Censors the Internet, Money Talks. Los Angeles Times, June 17, A1. Nakashima, Ellen, Steven Mufson and John Pomfret. 2010. Google Threatens to Leave China after ^ Attacks on Activists' Email. Washington Post, January 13. G C o Nathan, Andrew J. 2003. Authoritarian Resilience: China's Changing of the Guard. Journal of Democracy 14, 1, 6-17. _| Postill, John. 2008. Localizing the Internet beyond Communities and Networks. New Media & Society ^ 10, 3, 413-431. Qiu, Jack Linchuan. 2007. The Accidental Accomplishment of Little Smart: Understanding the Emergence of a Working-class ICT. New Media & Society 9, 6, 903-923. Stalder, Felix. 2006. Manuel Castells: The Theory of the Network Society. Cambridge, MA: Polity. Synovitz, Ron. 2010. Google-China Row could Reflect a New Era of Internet Control. Thompson, Clive. 2006. Google's China Problem (and China's Google Problem). New York Times, April 23. Tomlinson, John. 1994. A Phenomenology of Globalization? Giddens on Global Modernity. European Journal of Communication 9, 2, 149-172. Tomlinson, John. 1999. Globalization and Culture. Chicago: University of Chicago Press. Wang, Shujen and Jonathan Zhu. 2003. Mapping Film Piracy in China. Theory, Culture & Society 20, 4, 97-125. Wu, Xu. 2007. Chinese Cyber Nationalism: Evolution, Characteristics, and Implications. Lanham, MD: Lexington Books. Yao, Y. 2007. Choose eBay or Taobao: Globalization Lost to Localization. Shanghai Securities News, (April 6). Young, Doug and Yinka Adegoke. 2010. Global Media Titans Hit China Wall, Take Local Route. Reuters, August 13.