Introduction

A substantial body of literature has grown on the prominence of China as a recipient of foreign direct investment (FDI) and its consequences for national economic development and management practice (Branstetter/Lardy 2006). By contrast, much less attention has been paid to China’s position as an FDI source. Given that China attracted an annual average FDI inflow of around US$29bn (or more than 7 percent of the world’s total) in the 1990s, but contributed less than US$2.5bn (around 0.6 percent) to global outflows, this is perhaps not surprising (UNCTAD 2006). However, the sharp growth in Chinese outward direct investment (ODI) evident since 2002 (illustrated in Figure 1), combined with a number of recent high profile attempts by Chinese enterprises to acquire North American and European firms, have brought into relief China’s rising status and potential as an investor nation. This potential is recognised in a recent UNCTAD survey of investment promotion agencies which predicts that China will become a ‘top three’ source country for FDI before the end of 2008 (UNCTAD 2005). It is also highlighted by the Director-General of UNIDO, Kandeh Yumkella, who suggests that annual flows of Chinese outbound investment are likely to reach US$60bn by 2010 (MOFCOM 2006). If growth rates in Chinese ODI continue and these predictions are realised, China’s contribution to global FDI flows is likely to approximate current outflows of the leading industrialised countries.

In this exploratory study, we identify historic and emergent trends detectable in official aggregate data and individual FDI project level data on Chinese ODI for the period 1991 to 2005 with regard to investment destination, activity type, entry mode choice and investment motivation. Our aim is to assess whether or not Chinese ODI conforms to the general model of ODI and to the special case of emerging country ODI in general, and Asian countries in particular, with respect to the character and evolution of its recent ODI. To do this, we review in the next section some selected contributions to the literature on developing country ODI in order to establish a ‘received wisdom’ or base model against which we can contrast our empirical data from China. We also include some evidence from other Asian countries to control for cultural and regional interactions.Footnote 1 We go on to consider how the evolving institutional framework within which Chinese ODI is conducted and, especially, how adjustments to the administrative system and the engagement and disengagement of government at various times, notably following the launch of China’s zou chu qu or ‘go global’ policy in 1999, have influenced the internationalisation decisions and motivations of Chinese firms. After providing further evidence for the rise of China as an FDI source country, we examine trends in respect of (i) aggregate Chinese ODI stocks and flows; (ii) the spatial distribution of Chinese ODI; (iii) the sectoral distribution, and (iv) the dominant entry mode employed. This is done by reviewing data on accumulated Chinese ODI by host economy as published by the Chinese Ministry of Commerce (MOFCOM) and by analysing previously unpublished data from China’s State Administration of Foreign Exchange (SAFE), a government agency that administers, via the banking system, foreign exchange-related matters.Footnote 2 In the second part of the paper, we relate detected trends to emergent motivations advanced in the literature as driving the outward FDI activities of Chinese firms. We propose that Chinese ODI is indeed distinctive with respect to a standard model of developing country ODI, which itself is distinctive with respect to industrialised country ODI.

Statistics on Chinese ODI are compiled by MOFCOM based on a summation of individual firm’s direct investment amounts.Footnote 3 This aggregation masks the motives of the firms and reflects their choices of entry mode to foreign countries – direct investment is included, while licensing, technology transfer deals and other non-equity modes are excluded by definition. In this paper, we supplement these statistics with unique project level data from SAFE. Official statistics usually disaggregate the total by industry/sector and by destination country, but disaggregation by type of motive usually has to be conducted by analytical techniques such as regression analysis, which is an imperfect method working by inference. In addition, the time factor complicates the analysis (Buck et al. 2007. Firms often proceed by gradualism in foreign market entry, following a sequence of exporting, then non-equity modes such as licensing, then direct investment. A second type of sequential entry is from culturally and physically close countries to progressively more remote ones. These time series effects are only partially visible in cross section data. These limitations need to be borne in mind in our analysis.

Theoretical Explanations of Developing Country FDI

Firm and Industry Level Theory

Mainstream international business literature generally explains the strategy of the multinational enterprise (MNE) using the concepts of internalisation (Buckley/Casson 1976), transaction costs (Hennart 1988) and monopoly advantage (Hymer 1960). Together with location advantages, these concepts are synthesised by Dunning (2001) in his eclectic or OLI paradigm. This posits that the decision to internationalise production is predicated upon the interaction of ownership (O) advantages, location (L) advantages and the gains associated with hierarchical (I) over arm’s length transacting. Since this theory was developed to explain MNEs from the industrialised countries, its ability to account for developing country FDI has been debated. One view is that an alternative framework to explain late-comer MNEs is needed (e.g., Mathews 2002, Moon/Roehl 2001). However, the majority view is that mainstream theory does work, but that special theories nested within the general theory are needed as well (Buckley et al. 2007, Lall 1983, Wells 1983, Khan 1986, Lecraw 1993, Zin 1999, UNCTAD 2006).

Special Explanations for Asian Developing Country Firms

Lecraw (1993) identifies two key issues that could contribute to a special theory on the internationalisation of developing country firms; namely, how do they compete internationally (that is, what is their source of competitive advantage) and where do they invest (that is, what drives their location decisions)? In this paper, we also recognise as pertinent issues concerning entry mode choice, the role of home country government and cultural distance between home and host countries.

First, developing country MNEs are said to hold particular ownership advantages over established MNEs, in addition to competitively priced labour (an advantage which normally diminishes as the home economy develops) and that these derive from their experiences and knowledge of operating at home. In other words, the capabilities that firms gain to cope with home country conditions (i.e., ‘home country embeddedness’) can be leveraged as competitive advantage in similar markets abroad. Erdener and Shapiro (2005), for example, assert that overseas Chinese firms are able to penetrate Asian markets unattractive to industrialised country firms because they are adept at operating successfully in environments characterised by uncertain economic development, opaque regulatory conditions and weak market-enhancing institutions. Similarly, Scott (2002) observes that the ability to exploit culturally-dependent relational assets in Asian countries through personal relationships is a significant source of competitive advantage for overseas Chinese firms (Yin/Bao 2006). Wells (1977) and Kumar and Kim (1984) demonstrate that developing country firms in general possess older technology which is best exploited in less developed country markets. Developing country firms may also be better able than industrialised country firms at customising particular technologies, products and processes appropriately for other developing country markets. This may be accomplished by downscaling production, by simplifying or substituting local inputs or by increasing the labour intensity of production (Shenkar/Luo 2004). Developing country firms may also be more flexible and adaptable than industrialised country firms because scale economies are forsaken (Wells 1983, Erdener/Shapiro 2005). Lau (2003) argues that this is evidenced by the investments of Hong Kong-based textiles firms in other developing countries. It follows that developing country firms are often found to be involved in manufacturing activity abroad, beginning with labour-intensive production and then graduating over time into more technology and marketing-intensive production, often based on imported technology (Lall 1983, Wells 1983, Lecraw 1993, Zin 1999).Footnote 4 In short, home country embeddedness may enable developing country firms to compete successfully with established MNEs in third markets, as well as with local firms, especially in other developing countries (Aggarwal/Agmon 1990).

Second, it follows from this discussion that developing country MNEs generally concentrate their investment strategy on other developing countries. In the case of Asian firms this is often in markets geographically close to home (Lau 2003). In a study of small internationally active South Korean firms, Tallman and Shenkar (1994) found that investments were preferentially sought in Asian countries less economically advanced than Korea, where investing firms often acted as intermediaries in technology flows from developed to less developed host countries. Pang and Komaran (1985) report that Singaporean firms in the 1970s were slow to venture further abroad than Southeast Asia and when they did it was mainly to other, more distant, developing countries such as India. Chen (2003) reports that Taiwanese electronics firms preferentially invested first in Malaysia and Thailand because this enabled them to maintain important links with business networks in Taiwan. Only latterly did they relocate production to mainland China, once requisite industrial networks of buyers and suppliers were in place, to benefit from agglomeration effects and psychic and geographical proximity to home. The trend for Asian firms to preferentially invest in Asia is also evident in aggregate FDI data. For example, UNCTAD (2006), in an of intra- and inter-regional FDI flows between developing countries for the period 2002 to 2004, reports that as much as 96 percent of Asian FDI (with an annual average value of US$49.8bn) was directed to other Asian developing economies.

Third, the “stages” (or Uppsala) theory of incremental internationalisation may help to understand the distribution of developing country ODI over time. It proposes a gradual deepening in the engagement of the firm with individual host countries and a gradual widening of the host countries entered on a ‘closest first’ basis (Johanson/Vahlne 1977). Thus psychic distance between home and host markets bears upon managerial decision-making. Firms generally invest first in countries that are psychically proximate (that is, culturally similar) to their own because local market knowledge is more readily obtainable. As the firm’s international experience, knowledge and opportunities for learning grow, so too does its commitment to more culturally distant countries, since better local market knowledge raises the value of resources to be committed to the market (Brewer 2007, Dow/Karunaratna 2006). At the same time, FDI increasingly substitutes for ‘arm’s length’ agency and license contracts. A number of studies provide partial support for this proposition in relation to developing country firms in Asia by showing that they often invest preferentially in countries with strong historical ties or cultural similarity to the home region. For example, Pang and Komaran (1985) found that a large number of Singaporean firms initiated overseas investment activities in Southeast Asia, and later extended this to China, India and other developing countries before making debut entries in Australia and North America. Yang (1997) and Chen (2003) report that one of the main reasons why Taiwanese firms have invested heavily in mainland China is because of the short psychic distance between home and host country. Similarly, Erdener and Shapiro (2005) find that overseas Chinese owners and managers are preferentially attracted to investment opportunities in their ancestral home towns on mainland China because personal connections (that is, relational capital) can be exploited for competitive advantage. To test this theory fully, however, time series data at firm level are required.

Fourth, entry mode choice is an important aspect of the internationalisation of developing country firms. Though little researched, it is generally considered that a large proportion of MNEs from developing countries have preferred the international joint venture (IJV) entry mode (often with minority equity share) because this helps to reduce entry costs and increases the opportunities for learning from the foreign partner (Wells 1977, 1983, Kumar/Kim 1984). By contrast, more recent evidence suggests that, when possible, developing country MNEs in Asia (in particular, those from Taiwan and Singapore) choose wholly-owned subsidiaries and majority IJVs over minority ones (e.g., Yeung 1994). The use of higher equity modes appears positively correlated with later stages in the economic development of the home country, as a consequence of, for example, accumulated experience, greater managerial capacity and improved competitiveness of national firms. It may also reflect improvements in the ability to protect and enforce intellectual property and other proprietary assets abroad. However, industry effects may also be at work. For example, Tallman and Shenkar (1994) report that small Korean companies operating in technology intensive industries generally favour non-equity modes of cooperation to sell explicit technology, while equity IJVs are used in low technology-intensive fields. Perceptions of risk may also impact on the choice of entry mode by Asian firms. For example, a study of Malaysian multinational firms found that “low risk perceptions were associated with high control modes of entry and high risk perceptions were associated with low control modes of entry” (Ahmed et al. 2002).

Fifth, at a more aggregate level, research indicates that national governments in developing countries often play a critical role in determining the level and direction of ODI (Aggarwal/Agmon 1990). On the one hand, control and, effectively, the restriction of ODI has been a major strand of economic policy for many developing countries (UNCTAD 1996, 2006). Key objectives include prioritising domestic investment levels, preventing capital flight, strengthening foreign exchange reserves and maintaining control of state-owned assets abroad (Sauvant 2005). Such controls are generally relaxed over time once an adequate current account surplus has been achieved (UNCTAD 2006). On the other, developing country MNEs also commonly enjoy high levels of home government support, which may help them to ameliorate certain ownership and location disadvantages (Buckley et al. 2007, Aggarwal/Agmon 1990, Lecraw 1993). Typically, this takes the form of privileged access to raw materials, cheap capital, government subsidies and other benefits. It has been argued that one of the main reasons why South Korean chaebols have been able to invest abroad across a diversified range of industries is the soft budget constraint they enjoy from their close relationship with government and domestic financial institutions (Chow et al. 2004). Similarly, Lau (2003) asserts that strong government support has enabled Korean firms to invest heavily abroad at an early stage in the internationalisation process in a manner not predicted by the ‘stages’ model. The instrumental role played by developing country governments in setting the institutional framework for ODI activity is confirmed by a number of other studies, notably on Taiwan, Singapore and India (e.g., Pang/Lomaran 1985, Yeung 1994).

Aggregation and Explanation

This review enables us to establish a standard theoretical explanation of ODI from developing countries, in particular that from Asian economies, against which we can compare trends observed in Chinese ODI (see Table 1). This theoretical characterisation suggests six main dimensions. First, developing country firms have special ownership advantages that derive from their home country-embeddedness; second, developing country ODI is generally directed towards other developing countries; third, developing country firms invest preferentially in psychically and geographically close locations where relational assets can be exploited most effectively; fourth, developing country firms over time increasingly target investment opportunities in more advanced economies; fifth, international joint ventures are the preferred entry mode, especially early in the internationalisation process; and sixth, home country government has a strong influence on the level and direction of ODI.

Table 1. Chinese ODI Compared to a Standard Model of Developing Country ODI
Full size table

The Institutional Setting for Chinese ODI

Institutional factors are likely to be an important influence on any country’s aggregate ODI flow as at least part of the direction and nature of that ODI will be determined by source nation factors (Buckley/Casson 1976). However, institutional factors are dynamic and government policy changes over time. This section shows the influence of the Chinese institutional framework on Chinese ODI.

Since the late 1970s the Chinese government has determined to a considerable degree the legal, regulatory and financial framework of ODI, either directly, by administrative fiat (via the approval process and foreign exchange controls), or indirectly, using economic policy implementation and other measures (Buckley et al. 2008). Moreover, as the ultimate owner of state-owned enterprises (SOEs) (which dominated Chinese ODI prior to 2003), the government (at various levels) has effectively been the key operational decision-taker in the majority of formally approved investment projects. However, policy has often been ambivalent and inconsistent, with national and sub-national government at various times supporting, pushing and constraining Chinese ODI (Buckley et al. 2008). Key stages in the evolution of China’s official FDI approval process and some concomitant changes to the character of Chinese ODI are presented in Table 2.

Table 2. Key Stages in the Development of Chinese ODI Policy
Full size table

Even before the introduction of China’s ‘Open Door’ policy reforms in 1978, numerous small-scale investments by Chinese SOEs could be found in major trading hubs around the world, mostly in service sectors such as international trade, transportation and financial services. After 1979, and in hand with the ‘Open Door’ policies, the Chinese government cautiously sought to encourage ODI as a means to better integrate the country into the global economy and to improve access to domestically scarce raw materials (Zhang 2003). The government promoted international trade by permitting, and later encouraging, export-oriented FDI by state-owned import and export corporations. However, in the 1980s and 1990s tight centralised control of outward FDI was reimposed amid concerns that it was detrimental to national development. Outward direct investment was seen as a substitute for domestic investment (Sauvant 2005). It was also feared that control of state property held overseas might be lost because of both the cost of supervising international projects at a distance and the inexperience of Chinese firms at competing internationally (Zhan 1995, Ding 2000). However, a few selected SOEs, like China International Trust and Investment Corporation (CITIC) and Shougang, were granted the freedom to expand abroad as ‘experimental’ MNEs (Zhang 2003).

In the late 1980s and 1990s, it is generally acknowledged that Chinese firms internationalised mainly in pursuit of certain national and provincial economic goals and policy objectives, in particular: (i) to support the export function of state-owned manufacturers; (ii) to help stabilize the supply of domestically-scarce natural resources; and (iii) to acquire information and learning about operating abroad for the benefit of other domestic enterprises (Lu 2002, Ye 1992, Zhan 1995, UNCTAD 2006, Sauvant 2005). State-owned enterprises also undertook FDI to meet aspects of the government’s political agenda, not least in establishing and strengthening diplomatic relations with other developing countries through the building of economic links. For these reasons, research has generally stressed the importance of state engagement in the business affairs of Chinese firms, either through direct ownership of productive assets or indirectly, through various kinds of regulatory control and intervention (Sauvant 2005). From the late 1990s onwards, however, Chinese firms are increasingly portrayed in the literature as internationalising in order to achieve other objectives, in particular (i) to improve access to foreign proprietary technology, immobile strategic assets and capabilities; (ii) to exploit new markets for products and services; and (iii) to enhance overall firm competitiveness through the diversification of business activities (e.g. Taylor 2002, Child/Rodrigues 2005, Pei/Wang 2001, Deng 2003, Deng 2004, Zhang 2003, Zhang 2005, Warner/Hong/Xu 2004, Sauvant 2005, Beebe 2006). Ostensibly, these motivations are attributable as much to market forces, industry dynamics and discretionary, autonomous, managerial decision-taking as to government intervention and fiat. As UNCTAD (2006) comments, state-ownership does not necessarily invoke state-directed international strategy. At the same time, however, there remains a presumption held by some that the Chinese authorities continue to exert considerable influence over the investment activities of Chinese MNEs (e.g., Deng 2004, Deutsche Bank Research 2006). In this somewhat paradoxical milieu, it is interesting to investigate the extent to which engagement and disengagement of various levels of government has influenced the internationalisation decisions of Chinese firms (Voss 2007).

There is little doubt that state control over the international activities of Chinese firms has been relaxed considerably since the late 1990s. Perhaps the most prominent and clearly articulated policy has been the introduction of the ‘go global’, or zou chu qu, policy in 1999. This was subsequently formalised in China’s 10th five year plan, 2001-2006, and re-emphasised in the latest 11th five year plan, 2007-2010. Its objective is to encourage ODI through various means with a view to improving the international competitiveness of domestic companies and thus strengthen the national economy (Sauvant 2005, UNCTAD 2006).Footnote 5 It is partly in response to marketization of the Chinese economy and the country’s World Trade Organisation (WTO) accession commitments (Sauvant 2005), both of which have combined to heighten domestic competition, amongst other things. Accordingly, since 2001, policies towards ODI have been liberalised (mainly through the easing of investment restrictions,Footnote 6 simplification of approval procedures and relaxation of foreign exchange controls) and with indirect, ‘hands-off’ economic policies increasingly substituting for direct, ‘hands-on’ management (see also Table 2). To illustrate, government agencies like MOFCOM and the National Development and Reform Commission (NDRC), which were previously instrumental to the formal approval process, now purport to provide mainly advisory, information and support functions to international investors. A further important aspect is the treatment of private Chinese enterprises, which were prevented from investing abroad officially (with a few notable exceptions, like the white goods manufacturer Haier) before this restriction was lifted in 2003.

In future, it seems likely therefore that the individual investment decisions of Chinese firms will be shaped more by commercial considerations and less by political ones. The partial nature of the privatisation of SOEs may also influence ODI. In the early years of privatisation SOEs were given the opportunity to invest overseas and were encouraged to do so but they were not strictly governed as for-profit enterprises. This led to a serious agency problem. Top managers in SOEs could increase their income by positioning themselves overseas as managers of the companies’ foreign operations. This perverse incentive (together with round-tripping to exploit tax incentives) induced excessive ODI and may account for some of the unique patterns of China’s ODI. As institutional reform proceeds, we would expect these perverse incentives to subside.Footnote 7 However, the picture is complex and the challenge for researchers is to disentangle the role of national and sub-national government from other determinants (such as demand conditions and competition) of the level and direction of Chinese ODI flow.

China’s Outward Direct Investment Position

In Box One, we outline a number of shortcomings inherent with data from MOFCOM and SAFE. In particular, the data are for approved outward FDI only (typically undertaken by SOEs and large private or quasi-private Chinese firms). This excludes direct investments made by those (typically private, small and medium sized) Chinese firms using ‘informal’ (and often illegal) routes to international expansion beyond the government approval process. In practice, MOFCOM and SAFE data probably undervalue China’s outward FDI position, but to an indeterminate extent.

When firm level and individual FDI data are aggregated in source country statistics, some issues become blurred. However, from the above account we expect that Chinese ODI will be aimed at developing countries, especially in Asia. In line with theory, we expect proximity effects to decline over time as more global strategies are developed and as investing firms become more experienced. We note that the stage of development of firms in this trajectory will slowly feed into the aggregate data. To address this, we analyse project-level data collated by SAFE to draw inferences about the behaviour of single foreign direct investments. Ecological fallacy issues are avoided by analysing MOFCOM data at an aggregate level and SAFE data at the level of the individual foreign investment project.

Trends in China’s Aggregate Annual ODI Flow

Between 1982 and 1991, ODI from China increased year on year, but only slowly and never exceeding US$1bn annually (see Figure 1). In addition to restrictive investment approval procedures and tight foreign exchange control, the poor competitiveness of many Chinese firms was a contributory factor. Outward investors were generally large SOEs investing in projects of national importance, typically resource-oriented ones. Between 1991 and 1993 outward investment policies were relaxed and ODI surged, only to slow again in 1994 as the government sought to cool the rate of domestic economic expansion. New project proposals were subjected to more exacting approval procedures.

Figure 1.
figure 1

China’s Approved Outward FDI Flows (current prices), 1991-2005 (US$mn)

Full size image

After 1994, ODI flows recovered, accelerating modestly between 1995 and 1998 in parallel with further foreign exchange and trade liberalisation, growth in Chinese exports (which promoted trade-complementing FDI) and the handover of Hong Kong. In 1999, official Chinese investment outflow again declined, partly in response to the re-imposition of foreign exchange controls and the economic slowdown of several neighbouring countries, both outcomes of the Asian Financial Crisis. This continued to 2000, after which Chinese ODI accelerated sharply. This was concomitant with the introduction of the government’s formal ‘go-global’ (zou chu qu) policy initiative. At the time, this growth caused some commentators to assert that Chinese companies were on the threshold of becoming major foreign direct investors in Asia and beyond (e.g., UNCTAD 2003). By contrast, between 2002 and 2003 levels of Chinese approved outward FDI was in decline. This mirrored trends in global FDI flows more generally, with the economic downturn in the US and in the broad global economy being contributory factors. More recently, Chinese ODI has accelerated once more: in 2006, it amounted to some US$16bn, a seven fold increase on 2004 levels (MOFCOM 2007, NBS 2006).

The rapid recent growth in Chinese ODI is reflected in a number of key indicators. First, the contracted value of Chinese-owned outward FDI stock (at current prices) increased from US$1.4bn in 1991 to US$73bn in 2006. By 2006, Chinese ODI was distributed across some 160 countries (MOFCOM 2007). Second, although China’s ODI stock position is still quite modest relative to industrialised country norms, it already compares favourably to a number of smaller developed economies. To illustrate, by 2004 it was greater in value terms than that of Israel and Ireland and was only slightly smaller than that of Norway and Portugal, for example (UNCTAD 2004). Third, between 1996 and 2003 the number of Chinese MNEs grew from 103 to 510, while the number of Chinese-owned affiliates abroad rose from 1008 in 1991 to 8259 in 2004 (MOFCOM 2005). Fourth, the number of mainland Chinese firms among the world’s top one hundred non-financial MNEs from the developing countries rose from three in 2000 to ten in 2004 (UNCTAD 2002, 2006).

Although Chinese ODI is dominated by state-owned enterprises, since private Chinese firms were not permitted to invest abroad officially prior to 2003, at least a proportion of the recent improvement in China’s ODI performance will have been due to the international expansion of firms outside of direct state-control. However, data on this development are sketchy. Estimates suggest that 12 percent of Chinese ODI in 2004 was undertaken by private sector firms, but the overall contribution of the private sector to the accumulated stock of Chinese ODI remains minimal, standing at around 2 percent in 2004 (MOFCOM 2005). Although these data hint to the fact that private Chinese enterprise is likely to contribute increasingly to annual FDI outflows, research is needed to establish just how much of the recent expansion of Chinese ODI is attributable to this or to the relaxation of ODI controls enjoyed by state firms.

Geographic Distribution

MOFCOM statistics allow us to characterise changes to the geographic distribution of Chinese outward FDI since 1990. The MOFCOM Almanac of China’s Foreign Economic Relations and Trade (now the China Commerce Yearbook) reports an individual annual stock position for each host country in respect of approved Chinese FDI. These data are cumulative flow statistics based on new project approvals (but does not distinguish by ownership form or industry). We present these data in Table 3, using a rolling three year annual average to smoothen the irregularities commonplace in annualised FDI flow statistics. It also shows the number of investment projects made by Chinese firms per country per period so inferences can be drawn about average project value. Informal transactions that circumvent official FDI approval procedures (e.g., those associated with the round-tripping phenomenon) inevitably are excluded from the data.Footnote 9

Table 3. Approved Chinese ODI by Host Region and Economy, 1990-2004 (US$ 10,000 and percent)
Full size table

In the period 1990 to 1992, the majority of China’s outward stock of approved FDI (almost 70 percent of the total value) was located in the developed countries. Indeed, the bulk of Chinese ODI was concentrated in just three countries: Australia (host to an annual average of 23.3 percent of the total), the USA (22.2 percent) and Canada (19.4 percent). Western European and other developed countries received negligible amounts. Developing countries, by contrast, hosted collectively just under a third (30.5 percent) of China’s ODI, with the South, East and Southeast Asian region receiving almost half of this amount (15.4 percent). Of this, Hong Kong (the fourth ranked recipient), Thailand (6th) and Macao (8th) were the main destinations, the balance being more or less evenly distributed amongst the others. The remaining share of Chinese ODI to developing countries was divided almost equally between three regions, namely Africa (4.0 percent), Latin America and the Caribbean (4.9 percent) and Central and Eastern Europe (4.8 percent). No individual developing country in these regions was host to more than one percentage point of Chinese outward FDI stock except the Russian Federation (4.1 percent) and Chile (1.6 percent).

The distribution of Chinese ODI by value observed in the early 1990s is at odds with aspects of the received view of developing country FDI and its development path as predicted by the stages theory. In particular, the prominence of the industrialised countries and the comparatively weak positions of developing countries as hosts to Chinese ODI, especially in Southeast Asia, Africa, South America and, notably, India, suggest that Chinese firms were generally slow to invest substantial funds in other developing countries. In respect of geographic distance, the sixteen countries that physically border China attracted only 13.5 percent of Chinese ODI by value between 1990 and 1992.Footnote 10 This provides some evidence for the assertion that, when viewed in aggregate, Chinese ODI at this time was not influenced much by geography. Moreover, the fact that developed countries like the USA, Australia and Canada (all physically and, arguably, psychically distant from mainland China) figure so highly as hosts to approved Chinese FDI in the 1990s,Footnote 11 and that Hong Kong and Macao (both with large ethnic Chinese populations) were not major recipients, suggests that psychic distance and relational location advantages were also not key determinants. This interpretation contrasts to that of Zhan (1995), for example, who, among others, identifies ethnicity as a major determinant of the location decision of Chinese MNEs in the early 1990s. We attribute the distribution to the high degree of government involvement in the internationalisation decisions of Chinese firms.

A somewhat different picture emerges when we examine project numbers, however. Between 1990 and 1992, Australia, the USA and Canada were host to 24.2 percent of the investment projects made by Chinese firms, while the developing countries were host to 63.6 percent, with Hong Kong and Macao together accounting for 13.2 percent of the total. This hints at the possibility that a combination of physical proximity, cultural affinity and relational location advantages did indeed contribute to strengthening the presence of Chinese MNEs in these locations relative to others, but only when smaller scale investments are concerned. This is recognised by Fung (1996), who reports that Chinese FDI to Hong Kong in the early 1990s was largely motivated by the need to access new sources of finance for mainland operations. Besides a few large-scale investments in Hong Kong by China Resources, China Merchants International and the Bank of China, amongst others, this involved the formation of numerous shell companies – Sung (1996) estimates more than 14,000 – with minimum registered capital and limited commercial activities (Sung 1996). Such investment is not modelled well by the ”stages“ theory. However, as Fung (1996) comments, some Chinese ODI in Hong Kong at this time was to gain early internationalisation experience in a location with a contrasting institutional setting but cultural similarities to home, a behaviour which is predicted by the ‘stages’ model.

Between 1992 and 2001, there was a steady, three-fold increase in the annual average value of Chinese ODI. However, this growth was distributed unevenly. Although Australia, the USA and Canada continued to attract increasing amounts of Chinese outflows (in terms of both value and number), this was outstripped by that recorded for developing countries. Contrary to expectation, perhaps, the improvement in the relative position of developing countries in Chinese ODI is not accounted for by Hong Kong SAR and Macao: the proportion of total annual approved Chinese ODI destined to these countries over this period remained confined to the narrow range of 8 to 10 percent and 1 to 3 percent, respectively. Similarly, the position of the Russian Federation changes little over this period, consistently hosting between 3 and 6 percent of outflow. Overall, this pattern suggests that geography, geo-political and culture-related factors continued to play only a limited role in determining the destination of new Chinese ODI at the time. Again, this suggests that the stages theory has only limited explanatory power when applied to the internationalisation of Chinese firms prior to 2001, particularly of investments made by state-owned enterprises via the formal, approved route, with government influence a key contributory factor.Footnote 12

Instead, the greatest growth in China’s ODI position took place in Africa, and especially sub-Saharan Africa (notably Zambia, South Africa, Mali and Tanzania), Latin America and the Caribbean (especially Peru, Mexico, and Brazil) and in the South, East and Southeast Asian region generally (notably Cambodia and Indonesia). This distribution mirrors the findings of a survey on leading Chinese TNCs conducted by UNCTAD (2003). Since many of these countries at the time were characterised by comparatively high levels of political and economic risk, this distribution raises a number of interesting issues concerning the management of risk by internationalising Chinese firms. First, Chinese firms may have drawn on their home-country embeddedness (beyond those culturally-derived relational advantages) to exploit opportunities abroad in countries that industrialised country MNEs might regard as risky and where international competition levels are therefore low. Scott (2002), for example, points to the ability of Chinese-owned enterprises to manage risk as being a key source of ownership advantage (though his analysis is confined mainly to Chinese family-owned enterprises). Second, soft budget constraints and access to cheap capital (that is, domestic capital market imperfections) arising from high levels of state involvement in overseas projects may also have led Chinese MNEs to demonstrate a perverse attitude to risk management in comparison to industrialised country firms (Buckley et al. 2007, Antkiewicz/Whalley 2006). Third, the conclusion by the Chinese government of bilateral investment treaties, double taxation agreements and other initiatives designed to build strong economic and diplomatic relations with developing countries will also have helped Chinese investors to mitigate certain aspects of investment risk in the countries concerned (Buckley et al. 2008). This is especially relevant in the case of state-directed FDI that advances the political agenda of the Chinese government: for example, when it is used to develop connections with countries ideologically or politically distant from the west, for whatever reason. In such instances, the normal commercial considerations associated with risk, psychic distance and exposure to the liabilities of foreignness, for example, will have had little bearing upon the decision-taking of Chinese firms.

Given its population, geographic proximity and developing-country status, India continues to be a notable absentee from the list of principal destinations for Chinese firms throughout the period under investigation (mirrored also in India’s relatively modest position in global FDI flows generally). Also evident is continuation in the comparatively weak position of the western European countries, which attracted less than 1.7 percent of annual global Chinese ODI flow in the period 1999-2001, a decrease from the 2.5 percent or so generally observed in the 1990s. Given the relative openness of the European Union (EU) countries to inbound FDI, this is worthy of comment. First, despite the Single European Market and other economic harmonisation initiatives, Chinese firms may have been discouraged from investing because the EU is viewed by them as comprising distinct and separate national markets, each with their own set of standards, regulations, employment laws, immigration and visa requirements, language and so forth, unlike other attractive markets such as the USA. Second, because Chinese firms are required to negotiate separately with different national and regional investment agencies in Europe, this may have served as a disincentive by comparison to investing in other large markets. In sum, there may be a perception among Chinese firms that investing in the EU is more complex and bureaucratic than investing elsewhere.

Notwithstanding the dominant position of the USA as a host, figures suggest that the distribution of Chinese ODI in recent years has begun to approximate that of developing countries more generally in that other developing countries are increasingly being targeted as investment locations by Chinese MNEs. Indeed, the widening distribution of Chinese FDI by country over the period is striking. The number of individual countries host to Chinese ODI rose from 95 in 1990, to 139 in 1996 and to 162 in 2003, by when Chinese MNEs had invested in forty-six Sub-Saharan countries, thirty Latin American countries and eighteen Central and Eastern European countries. While this distribution can be attributed to responses to market opportunities and other factors endogenous to the firm, another interpretation is that it is in response to the government’s preference for a spatially diversified overseas production portfolio that minimises exposure to political and other risks, especially in places like Africa and West and Central Asia.

In the 2002-2004 period, when Chinese ODI continued to increase sharply, this again was mainly to the developing rather than developed countries, in terms of both value and project number. Much of this is attributable to Hong Kong SAR, which has become a major destination for Chinese investors. This suggests that geo-cultural affinity is an increasingly important driver of Chinese ODI, although Hong Kong’s position as a financial centre and as a pathway for investing elsewhere is also an important locational advantage for Chinese MNEs.

Sectoral Distribution

Using SAFE data, we are able to discern certain changes to the sectoral distribution of Chinese ODI. The SAFE dataset, the most detailed available on the subject, comprises approved investment project information by host country, industry and entry mode for the years 1991 to 2001 (after which project level data were no longer formally collected). Annually collected data on projects by sector reveals that the bulk of Chinese ODI by value in the early to mid 1990s was predominantly engaged in the tertiary and manufacturing sectors (see Table 4). Historically, Chinese FDI in services has generally involved a large number of small scale investments in trade-supporting activities by Chinese trading companies, with investments in the banking sector (notably by the Bank of China but also by CITIC and China Merchant Holdings, both in Hong Kong), insurance and construction, communication, real estate and restaurants also are significant. Many of these investments served as vanguard operations for later Chinese entrants in addition to providing important overseas trade support to firms in China. Similarly, Chinese manufacturing ODI has generally involved relatively small-scale and labour-intensive production of undifferentiated and low-value-added goods using simple product and process technologies. In more recent years, SAFE data reveal that Chinese ODI has occurred mostly in the manufacturing sector. The presence of foreign invested enterprises in China and two decades of market opening have yielded spillover benefits that have enhanced domestic and international competitiveness (Buckley/Clegg/Wang 2002, Buckley et al. 2002). Together with greater familiarisation with operating internationally, this is likely to have assisted in the international expansion of Chinese-owned manufacturing activity. In addition, Chinese government policy to shift international expansion away from ‘one-track’, trade-related activity to more diversified, ‘multi-purpose’ operations and the support provided to Chinese manufacturers as part of China’s ‘go global’ policy are also key driving forces. Viewed in aggregate, the growth in Chinese manufacturing FDI in recent times is congruent with the received view of developing country ODI in that a greater propensity to invest in manufacturing activity abroad is observed over time. We note, however, that official ODI stock data published by MOFCOM suggests that the greatest proportion of Chinese ODI has occurred in the primary sector. Taylor (2002), for example, uses MOFCOM data to report that the manufacturing sector accounted for only 11.5 percent of China’s outward FDI in the late 1990s, compared with 19.4 percent for resource development and extraction, 1.8 percent for communications and transport, and 66.4 percent for other categories (see also Zhan 1995, Chan 1995). MOFCOM (2005) report a similar sectoral distribution of Chinese ODI stock for 2003 as follows: 6.2 percent in manufacturing, 19.2 percent resources in development and extraction, 6.1 percent in transportation and warehousing, 19.7 percent in wholesale and retail trade and 48.8 percent for the remaining categories (construction, business services, information technology-related sectors and other industries). It is clear from these data that investment in the extractive industries is an important contribution to Chinese ODI, especially in mining, fisheries and forestry exploitation and petroleum and natural gas exploration. However, regular reclassification and procedural revisions to the reporting of ODI by activity on the part of MOFCOM makes it difficult to draw further conclusions from a longitudinal examination of the data.

Table 4. Sectoral Distribution of Outward Chinese FDI: 1991-2001 (percent of total value)
Full size table

Entry Mode

Literature on the internationalisation of developing country MNEs suggests that minority IJVs are the preferred mode of market entry (Wells 1983, Yeung 1994). One reason is that such firms seldom possess the level of proprietary technology and firm-specific know-how to necessitate internalisation via majority or full ownership (Buckley/Casson 1976, Dunning 1993). This is also evident in the early years of Chinese ODI development: project level SAFE data reveal that, in the early 1990s, around 70 percent of overseas projects of Chinese firms took the IJV form (see Table 5). Zhan (1995) also reports that Chinese firms tended to opt for majority equity shareholdings in overseas projects, typically in the range of 40 to 70 percent equity participation, especially in natural resource-oriented and manufacturing-related projects. A number of explanations can be envisaged. From a governmental perspective, the formal investment approval process generally required Chinese MNEs to adopt the IJV entry mode. The Chinese authorities had become familiar with the economic gains associated with the promotion of inward FDI in the form of IJVs, the promotion of which was a cornerstone of China’s ‘Open Door’ policy. The JV form was seen as a vehicle for promoting the inflow of foreign-owned technology, management know-how and other skills to China. The authorities were also now adept and comfortable with at administering foreign invested enterprises in China. It is likely that equivalent advantages were sought when Chinese enterprises invested abroad. Familiar cost and risk-minimising features of IJVs will also have been important to the investment approval agencies (Zhan 1995, Taylor 2002, Wang 2002). From an enterprise perspective, inefficient domestic capital markets and budget constraints meant that many Chinese enterprises, including state-owned ones, often found it difficult to obtain sufficient funds to purchase overseas assets outright, compelling them to opt for the IJV alternative. The JV form also allowed Chinese MNEs to exercise a degree of control over local operations whilst avoiding outright ownership and the concomitant exposure to political and commercial risk. Chinese enterprises could tap foreign partner contributions, such as improved access to market intelligence, knowledge of the local operating environment, opportunities for reputation riding and better access to local distribution channels through the IJV (Taylor 2002). When established with other ethnically-Chinese enterprises (in Hong Kong and elsewhere), the JV form also allowed relational assets to be optimised, reducing perceived risks and costs associated with psychic distance, especially for smaller and less experienced Chinese investors (Zhan 1995). Mutual trust would also have been easier to establish. Thus, we see both institutional and firm-specific factors influencing the choice of IJV by Chinese firms at this time.

Table 5. Entry Mode in Chinese Outward FDI: 1991-2001 (percent of number of foreign affiliates)
Full size table

From the mid-1990s onwards, however, SAFE data at individual project level reveal that wholly-owned FDI projects have increasingly substituted for jointly-owned ones in the international expansion of Chinese enterprise, with 61 percent of overseas affiliates in approved projects taking this form in 2001 compared to 30 percent in 1991. We note that this contrasts somewhat with the findings of Taylor (2002), who reports much greater use of IJVs in the recent internationalisation of Chinese firms, especially in manufacturing-related activity. A number of reasons explain the more frequent use of the wholly-owned entry mode in recent years. First, more frequent approval of wholly Chinese-owned projects reflects growing confidence among the regulating authorities that managers of state-owned Chinese MNEs have become sufficiently experienced and skilled to take control of, and co-ordinate effectively, the activities of geographically-dispersed affiliates. It is also a reflection, at least in part, of the strategic importance placed on particular projects by the Chinese authorities. Theory asserts that, by internalising markets, the internationalising firm is able to reduce its dependency on independent intermediaries; militate against the threat of technology and know-how leakage; reduce the risk of opportunistic behaviour by alliance partners and allow for full appropriation of returns on investment (Buckley/Casson 1976). Both the investment approval agencies and enterprises will have found such advantages attractive, despite the costs and risks associated with full ownership. Second, greater use of wholly-owned affiliates may reflect improved availability of investment funds. Government initiatives under the ‘go global’ (zou chu qu) policy have released capital to state-owned firms (often at below market rates) in the form of loans and improved access to hard currencies, to help them finance the outright purchase of foreign assets (Antkiewicz/Whalley 2006). Many Chinese enterprises are also now skilled at raising investment funds on international capital markets, especially in Hong Kong (Buckley et al. 2007, Chan 1995). Thus, many Chinese firms are no longer obligated to reduce investment cost by undertaking an IJV. Third, the growth in international market entry by acquisition will have led Chinese enterprises to establish more wholly-owned subsidiaries in foreign markets rather than jointly-owned projects.

The standard theoretical model of “Asian ODI” suggests that China is not unusual among Asian countries in using wholly owned subsidiaries more frequently over time (Pang/Komaran 1985, Euh/Min 1986, Yeung 1994). However, caution should be exercised in assuming that this mirrors improvements in the managerial capacity and competitiveness of Chinese MNEs: greater deployment of majority and wholly-owned foreign operations may also be more a function of the government’s desire to retain effective control of state assets abroad and a growing confidence in its ability to do so than of purely firm-specific or market-related considerations.

Motives for Chinese Outward FDI

In this section, we relate historic and emergent trends in aggregate Chinese ODI data identified above to changes in the motivations driving the internationalisation of Chinese MNEs. Dunning (1993) identifies four basic motivations that provide the impetus for foreign-owned production and are discussed below: namely, natural-resource seeking, market-seeking, efficiency-seeking and strategic asset-seeking motives.

Natural Resource-Seeking FDI

Backward integration to acquire or secure the supply of specific location-bound resources and commodities abroad for domestic consumption has been the predominant driver of Chinese outward FDI since the late 1970s (Taylor 2002). More recently, China’s rapid economic growth over the past decade has fuelled what some say is an almost insatiable demand for raw materials and other inputs in many sectors (Economist 2004). Footnote 13 The dual objective of further improving the supply of natural resources from abroad while ameliorating (at a national level) exposure to political and commercial risk has seen Chinese enterprises recently investing in natural resources-oriented projects across a broad range of resource-rich countries, especially in Africa and East and Central Asia (see Table 3). Leading recent recipients are Zambia (for copper), Peru (iron ore), and western and central Asian countries like Kazakhstan (oil exploration and extraction). Most investors are state-owned enterprises which enjoy strong support from the Chinese government in the form of direct financial assistance; the negotiation of bilateral investment treaties and trade agreements with host countries and the close inter-governmental relationships that China is now reviving across certain parts of the developing world. Exemplar companies include China Natural Petroleum Corporation (CNPC), the joint owner of a Sudanese oil production plant (together with Canadian, Malaysian and local interests), Sinopec, Shanghai Baosteel (the owner of six joint ventures in Australia, Brazil and South Africa in iron-ore mining and steel trading), Sinochem and China National Offshore Oil Corporation (CNOOC). There is some evidence to suggest that official development aid provided by China to developing countries (for example, concerning telecommunications and transportation infrastructure development, project-specific inter-governmental loans, education packages and so forth) is predicated upon market access or exploitation and extraction rights being granted to Chinese MNEs (Pan 2006, Evans/Downs 2006).

It has been argued that MNEs from emerging countries are most likely to invest in the industrialised countries when looking to access technology and learning (Monkiewicz 1986, Ye 1992, Deng 2003). However, this may not be the case for China. Whilst knowledge-acquisition has become increasingly important to Chinese MNEs in recent years, much of Chinese ODI by value was invested in the industrialised countries primarily for natural-resource seeking reasons, especially in the early 1990s. Good examples include the investments by CITIC and Huaguang Forest Co. Ltd in timber plantations in New Zealand, CITIC’s investment in forestry in the USA, CITIC and China National Nonferrous Metal Industrial Corporation’s 10 percent (US$120mn) investment in Portland Aluminium’s smelter operations in Australia. Canada is also now host to a number of timber and fisheries related Chinese investments (e.g., CITIC’s investment in the Celgar pulp mill and sawmill project) (Zhang 2003).

Market-Seeking FDI

Chinese MNEs now conduct both defensive and offensive market-seeking FDI. It is axiomatic to state that China enjoys a comparative advantage in low-cost labour and labour-intensive production. Given the location-bound nature of labour, the international competitiveness of the majority of (both foreign and locally-owned) firms in China necessitates domestic production and foreign market servicing by exports. Chinese enterprises have long established overseas operations to facilitate trade. Certainly, in the early 1990s, the bulk of Chinese ODI in services was export trade-related. Chinese exporters have commonly confronted a range of tariff and non-tariff trade barriers abroad. Although China’s WTO accession should see these reduced, the reverse may happen in those countries with which Chinese enjoys a large trade surplus, such as the USA. For example, the imposition of protectionist measures (or its threat) presently underpins a small but significant proportion of the recent growth in Chinese ODI to the USA, Latin America and, but less so, to Europe (Taylor 2002), for defensive market-seeking reasons. Protectionist pressure also accounts for a significant share of Chinese ODI in ‘third-party’ trading countries (Taylor 2002). Increasingly, Chinese enterprises are locating ‘offshore’ manufacturing plants to those countries with which the industrialised nations set few, if any, export quotas and other ‘anti-dumping’ measures, or they invest in countries where quota rights can be appropriated readily (Lau 2002, Taylor 2002, UNCTAD 2003). This accounts for much of the recent growth in market-seeking ODI by Chinese firms in a number of countries, including, for example, Cambodia (where Chinese garment manufacturers in particular enjoy fewer quota restrictions in third markets); Mauritius (where export quota restrictions are mostly absent), Jamaica and Fiji (UNCTAD 2003). A further illustration of defensive, market-seeking FDI is provided by the purchase in 2002 of the insolvent German television maker Schneider Electronics AG by TCL, China’s second largest television and mobile-phone maker. Reportedly, this was motivated, at least in part, by TCL’s desire to negate possible accusations of dumping products in Europe (CNN 2003).

A second aspect to defensive market-seeking FDI by Chinese firms is their response to factors that combine to limit growth opportunities at home (Beede 2006). First, China is obligated under its WTO accession terms to further open domestic markets to both imports and FDI. This has inevitably increased competitive pressures in home markets (Taylor 2002). Second, supply-chain bottlenecks, restricted demand and fragmented national markets are now commonplace in certain sectors in China (e.g., domestic appliances and machinery and in textiles, clothing and footwear) and this has led to excess capacity. Third, greater regulatory transparency and superior distribution networks abroad means it is often easier for Chinese enterprises to develop foreign markets than domestic ones, especially for those located in the coastal provinces close to international transportation networks. For many, the challenges associated with supplying domestic markets are less in evidence abroad.

There is also growing evidence to suggest that Chinese enterprises are now investing abroad for offensive market-seeking reasons; that is, to develop new markets and raise brand awareness (UNCTAD 2003). Although inefficiencies and lack of competitiveness of Chinese enterprise across a gamut of industries and sectors have been highlighted (e.g., Nolan 2001, Nolan/Zhang 2002), a growing number of Chinese enterprise are now able to compete more effectively in international markets. For them, international expansion represents a proactive step, with new markets being developed overseas because of attractive demand conditions. Although many Chinese companies are able to compete by selling simple, undifferentiated, mature products in low-income countries (exemplified by Chinese-owned bicycle production in Ghana and video-player sales in South-East Asia), others are increasingly able to compete in more technology-intensive sectors in both developing and developed countries by undertaking large-scale, capital-intensive, market-seeking investments. Good examples include electronics companies such as Huawei, ZTE, Konka Electronics, Skyworth Group, Changhong Electronics Group Corp, Lenovo Corporation (formerly known as Legend Corporation) and Haier, and a number of enterprises in the plastics, chemicals and pharmaceuticals sectors (UNCTAD 2003). Chinese MNEs are now establishing sales and marketing functions in target markets to lower their dependency on intermediaries. Sinochem, for example, now has sixty foreign affiliates to develop and expand sales of chemical products in major overseas markets while Lenovo Corporation now has over twenty foreign affiliates to sell software products internationally. Typically, physical proximity to key local markets reduces transportation costs and improves access to local market knowledge and information flows to and from both consumers and suppliers. It also facilitates the adaptation of products and services to local conditions. In time, a local presence should also allow Chinese investors to be perceived as ‘insiders’. This may become important should neo-protectionist trade or political tensions grow between China and host countries as negative ‘country of origin’ effects may be reduced. To illustrate, the private Chinese autoparts supplier, Wanxiang, purchased a number of insolvent US component manufacturers to secure access to leading car assemblers in the USA and, reportedly, to help circumvent negative connotations associated with its nationality that might have been held by unionists and other stakeholders.

A second aspect to offensive market seeking FDI is the response of Chinese firms to deepening regional economic integration in some parts of the world. For example, a number of recent Chinese investments in Mexico were made in order to benefit from preferential treatment given by the USA to Mexican imports under the terms of the North American Free Trade Association (NAFTA). Similarly, a proportion of Chinese FDI bound for Cambodia and Vietnam was stimulated by the prospect of improved access to South East Asian markets as a consequence of the ASEAN Free Trade Agreement and the Asian Investment Area. However, the relatively modest amounts of Chinese ODI hosted by the European Union points strongly to the fact that regional economic integration and large markets may be a necessary but not a sufficient condition for offensive market-seeking FDI by Chinese MNEs, for reasons already discussed.

Strategic Asset Seeking FDI

A variety of foreign-owned assets, both tangible and intangible, are of potential interest to Chinese enterprise. Historically, the principal intangible resource sought by Chinese MNEs was information, especially about external economic and trade conditions. In the past, Chinese MNEs have been obligated to assimilate and disseminate experience and knowledge of foreign management practices to advance the international competitiveness of Chinese enterprise more generally. Indeed, Taylor (2002) comments that China’s outward FDI is an on-going quest for market information to improve domestic export performance. Nevertheless, there is some evidence to indicate that Chinese MNEs are becoming less interested in market information and operations-related knowledge and instead are looking to tap foreign knowledge of technology-intensive production and local markets (UNCTAD 2003). To this end, Chinese enterprise are now establishing research-oriented affiliates in high-income countries to assist in the development of high technology, knowledge intensive products manufactured in China and exported via sales affiliates. In some places, like the USA and UK, this process is supported by home country efforts to attract this type of investment (Sauvant 2005), especially in sectors that do not challenge the local manufacturing base.

A second intangible asset increasingly sought by Chinese MNEs relates to brands and complementary assets (see Table 6 for recent examples). While some companies, such as Lenovo Corporation and Haier, have extended their key brands and trade names into foreign markets themselves, with some success, others have found it quicker and more effective to simply acquire established western brands and associated marketing channels. To illustrate, a key reason for the formation in 2003 of TCL-Thomson Electronics, an IJV between the French electronics firm Thomson and TCL International Holdings (the Hong Kong-listed affiliate of TCL Group), reportedly was to enable the JV to exploit the brand portfolios of the partners in Asia and North America respectively (CNN 2003).

Table 6. International Brand Acquisition by Chinese Companies – Some Successes and Attempts
Full size table
Table 7. Selected Acquisitions by Chinese MNEs since 2000
Full size table

While the large amount of investment finance required to effect an overseas acquisition may have precluded most Chinese enterprises in the past, this is no longer the case. Sales growth in certain sectors of the Chinese economy has meant that an increasing minority of Chinese enterprises, both private and state-owned, are accumulating sufficient retained earnings to fund major capital investment projects abroad. It is also helped by the relaxation of foreign exchange controls, by the low cost of capital enjoyed by some state-owned firms and by the strength of the Renminbi, which has lowered relative investment costs in certain markets. All this means that increasing numbers of Chinese enterprises are now able to obtain the foreign currency required to make strategic-asset seeking FDI in industrialised as well as developing countries a feasible option. In the case of industrialised country target firms, this often entails the purchase of loss-making businesses (see Tables 6 and 7 for examples). Viewed alongside the observation that many Chinese MNEs have acquired intangible and complementary assets that they have little to no prior experience of managing, this raises a question concerning the ability of some Chinese firms to generate profits from post-acquired businesses.

A third type of strategic asset sought by Chinese MNEs is improved access to capital markets. China’s domestic capital market has long been inefficient and Chinese policy has generally restricted the holding of external debt by SOEs. Project finance was often difficult to obtain, therefore, especially in non-priority sectors. In the early 1990s, a number of large Chinese enterprises responded by acquiring weak corporations in Hong Kong, transforming themselves into MNEs overnight. These were used to obtain listings on the Hong Kong stock exchange with the capital secured redirected to China to fund domestic enterprise (Liu 2001, Liu/Li 2002, Sung 1996). Many Chinese-owned stocks are now listed on the Heng Seng Index, including China Telecom, China Merchants Holdings, China Unicom, China Mobile (HK) Ltd., Sinopec, CNPC, China Everbright Ltd., Lenovo Group Limited (formerly known as Legend Holdings Limited) and Founder Holding International (Shi 2000). Tax-havens such as the British Virgin Islands, the Cayman Islands and Bermuda have also been used by Chinese MNEs to obtain venture capital (Frost 2005), to channel funds back to mainland China (and thus benefit from foreign invested enterprise status) and to circumvent restrictive outward investment approval procedures (Voss 2007).

Efficiency-Seeking FDI

When a firm internationalises for efficiency-seeking reasons, it generally does so by reorganising and rationalising established resource-based or market-oriented FDI operations (Dunning 1993). Typically, this is done to exploit the benefits of regional economic integration and the international division of labour. Firms take advantage of different factor endowments, yet converging cultures, institutional arrangements and economic systems across a regionally integrated group of countries by supplying markets from a reduced number of intra-regional plants. In contrast to much of the FDI undertaken by industrialised country MNEs, it is unlikely that greater efficiency is currently a major driver for Chinese firms. At present, Chinese enterprise have little incentive to seek production efficiencies abroad since domestic markets provide ample supplies of relatively low-cost labour, land and other factor inputs, especially away from the coastal regions. Moreover, few Chinese companies currently have sufficient numbers of overseas operations to warrant substantial reorganisation. However, as we have seen, regional integration is beginning to shape the investment strategies of Chinese MNEs, notably within NAFTA and South East Asia, though mostly for market-related rather than efficiency-related reasons. As these international operations expand in scale and scope, efficiency-seeking FDI by Chinese MNEs is likely to become more commonplace.

Conclusions: Is Chinese ODI a Special Case of Emerging Country ODI?

This paper makes a number of contributions to our current understanding of Chinese ODI. An application of several levels of explanation using aggregate (MOFCOM) data, individual project level (SAFE) data and a review of recent studies has enabled us to provide a rich picture of the phenomenon little discussed in the literature (e.g., Deng 2003, Taylor 2002). We find that Chinese ODI is similar to, yet distinct from, the standard model of emerging country ODI, as Table 1 shows.

Our analysis of the changing geographic distribution of aggregate Chinese ODI by value suggests that geographic and psychic distance were not important determinants of larger scale Chinese investment projects in the 1990s, which were generally directed towards the industrialised countries, often for resource-seeking reasons. This is at odds with aspects of the standard model of ODI from Asian developing countries and the incremental ‘stages theory’ approach which predict a tendency for Chinese ODI to be associated negatively with the level of development of the host country and with increasing geographic and psychic distance (see Table 1). We attribute this to the significant involvement of government, both direct and indirect, in the internationalisation decisions of Chinese MNEs in the 1990s. However, we observe geographic and psychic distances to have had greater influence on the international distribution of smaller scale and more recent projects by Chinese MNEs, especially after 1999. This suggests that ‘stages’ model of internationalisation has greater explanatory power for more recent Chinese ODI. Today, Chinese ODI is distributed more widely to encompass a large number of developing host countries (notably in Africa and Southeast Asia) in addition to the industrialised countries, which historically have been important destinations (with the exception of Western Europe). This provides some support for the argument that the distribution of Chinese ODI is beginning to conform to patterns predicted by the received view of ODI from developing Asian economies. However, to what extent this is attributable to the gradual disengagement of government from the direct regulation and control of ODI is a subject for further research.

It is clear that increasing numbers of Chinese MNEs (mainly government owned or controlled) are now grasping opportunities arising from deregulation and liberalisation of the ODI regime by extending their international reach. They continue to pursue strategies that fulfil certain national economic and political imperatives but other motivations are also now at work. In the 1990s, improvement to the supply of natural resources was an important driver of Chinese ODI and this continues to be the case. The development of overseas market opportunities also remains an important driver. However, firm strategy appears to be shifting away from merely support of the trade function and information gathering towards market-seeking FDI that is both defensive (i.e., to circumvent obstacles to trade with import-substituting and quota-hopping FDI or in response to competitive pressures and weak market access at home) or offensive (i.e., that seeks to improve foreign market access through the establishment of sales and manufacturing subsidiaries) in orientation. We also see Chinese enterprises attempting to raise their competitiveness by undertaking strategic asset seeking FDI. Often, but not always, this is achieved with the purchase of under-performing foreign firms. The objective is to acquire hard-to-replicate assets such as advanced technology and established foreign brands and to improve access to distribution channels and sources of foreign capital. For these reasons, manufacturing activity now takes a greater share of the sectoral distribution of Chinese FDI, and wholly-owned subsidiaries are now preferred to IJVs. These findings contrast with aspects of earlier work on Chinese MNEs (e.g., Taylor 2002, Deng 2003) and with the notion that developing country firms generally opt for the IJV entry mode because of the cost and risk-related advantages it brings, along with opportunities for learning from partners.

Table 8. Some Emergent Trends in Approved Outward Chinese FDI
Full size table

There is some evidence from our analysis, therefore, to suggest that Chinese MNEs can no longer be regarded as ‘apprentices’ on the international stage, investing primarily in the developed countries to obtain information and to support the export function, or to learn from joint venture partners. Rather, a small but growing number of Chinese MNEs are becoming truly ‘transnational’, acquiring not only the confidence but the knowledge, resources and capabilities needed to coordinate international activities and compete effectively for market share in both developed and developing countries. This gives rise to an ‘emergent’ strategic behaviour of Chinese MNEs (see Table 8) which we argue is increasingly superseding ‘historic’ behaviour under each of Dunning’s internationalisation motives.