Changes in the influential factors of China’s
outward FDI
Master thesis Xintong Li 10607641
Msc Business Studies: International Management University of Amsterdam
First Supervisor: Carsten Gelhard Second Supervisor: Dr. Niccolò Pisani Date: June 25, 2014
Abstract
Much research focuses on the influential factors of location choice of foreign direct investment (FDI). But the influential factors are not stable and vary over time. This thesis aims to identify the changes in the influential factors of China’s outward FDI during the period from 2003 to 2012. Based on FDI motives and locational attributes, two research genres of influential factors, this thesis tests seven influential factors to see whether they experienced changes in the two periods, the initial period (2003-2007) and the later period (2008-2012). According to the panel data analysis results, market size and tax policy both have effects on FDI outflows in the initial period. Chinese firms are more willing to invest in the locations with larger market and lower tax rates. Besides, market size’s influence is heavier than that of tax policy. During the later period, labor cost and technological capability influenced the FDI outflows. China’s outward FDI are more willing to flow to the locations with higher labor cost and relatively lower technological capability. This strange phenomenon is generated by the global financial crisis. Technological capability just has a little impact on the FDI outflows, smaller than labor cost’s impact.
Acknowledgements
I sincerely appreciate the suggestions and help provided by my supervisor, Carsten Gelhard. No matter in accomplishing thesis proposal or in the long process of master thesis, he kept giving me practical advice and guiding me to the right direction if it was necessary. When I had problems with my thesis, he gave me relative articles to help me solve the problems and sometimes answered me face to face patiently. In addition, I also want to thank my family and friends. They comforted me when I was frustrated and worried about my thesis. I especially want to thank my friend, Yanni Hou. She helped me a lot in the process of collecting data. Help from these people is of vital importance to this thesis.
Table of Contents
1. Introduction... 1
2. Literature review... 4
2.1 Influential factors of FDI location choice... 4
2.2 Changes in the influential factors of FDI location choice...7
2.3 Why do influential factors of FDI location choice change over time?...11
3. Theoretical framework and hypotheses... 13
3.1 Theoretical framework: resource-based and institution-based view...13
3.2 Hypotheses: influential factors in two periods...16
3.2.1 Natural resources...16 3.2.2 Market size... 17 3.2.3 Technological capability...18 3.2.4 Labor cost...19 3.2.5 Tax rate... 20 3.2.6 Agglomeration effect...21 3.2.7 Infrastructure... 22 3.2.8 Political risk...23 3.2.9 Exchange rate... 24 4. Methods... 24 4.1 Data collection...24 4.2 Measurement... 25 4.2.1 Dependent variable...27
4.2.2 Independent variable... 27 4.2.3 Control variable...28 4.3 Model...28 4.4 Pretreatment of data...28 4.4.1 Missing data... 28 4.4.2 Correlation matrix... 29
4.4.3 Unit root test and cointegration test... 30
4.4.4 Hausman test and redundant fixed effects test... 31
5. Results...32 6. Discussion...36 6.1 Implications...40 6.2 Limitations...41 7. Conclusions... 41 8. Appendix... 43 9. References... 44
1. Introduction
Nowadays, thanks to the development of regionalization and globalization, the cross-border expansions of firms, by means of foreign direct investment (FDI), happen actively around the world, no matter in developed countries or in developing countries. Location choice is the first and a very significant challenge firms have to meet when they decide to expand overseas. Scholars in international business have concentrated on this topic and have made a lot of important findings, so that firms can draw lessons from these findings when they conduct investments abroad (Fetscherin et al. 2010). In addition, accompanied by enormous economic benefits, FDI also attracts attention from governments. Policy makers expect to attract FDI to help develop local economy. As a result, grasping the influential factors of location choice is the first step for policy makers to attract inward FDI (Cheng & Kwan, 2000a).
The influential factors of location choice include two parts, external factors and internal factors (Dunning, 1981). External factors refer to the locational attributes of host countries to attract inward FDI, such as labor cost (Wheeler & Mody, 1992; Cheng & Kwan, 2000a), labor quality (Cheng & Kwan, 2000b; Belderbos & Carree, 2002), market (Cheng & Kwan, 2000a; Zhou et al., 2002), policy incentives (Head & Ries, 1996; He, 2002), infrastructure (Head et al., 1995; Du et al., 2008a), and agglomeration effects (Head et al., 1995; Belderbos & Carree, 2002; He, 2002). Internal factors refer to the motives of MNEs to carry out FDI. Dunning (1998) summaries four motives of FDI, which include resource-seeking, market-seeking, efficiency-seeking and strategic asset-seeking motive. In addition to these
four, there is some extension of Dunning’s (1998) findings, covering asset-exploiting and asset-exploring FDI (Makino et al., 2002).
The aforesaid factors influencing location choice of FDI are based on static analysis. They all consider influential factors as static things. In fact, the influential factors are dynamic and experience changes over time. According to some previous research, traditional influential factors, such as resource and market, become less important than what they were (Dunning, 2004). Infrastructure and agglomeration replace the status of resource and market and become the main consideration of MNEs in some locations (Nunnenkamp, 2002). Labor cost was considered to have a negative relationship with inward FDI. However, nowadays labor cost does not have obvious influences on location choices, and high labor cost even attracts inward FDI in the US because high labor cost means high labor quality (Flores & Aguilera, 2007). In contrast to the condition of traditional influential factors, there are emerging factors playing increasingly significant roles. Institutional-cultural factors get to have a remarkable effect on location choices (Flores & Aguilera, 2007). Moreover, MNEs prefer locations with less government intervention in spite of preferential policies (Du et al., 2008a).
To investigate changes in the influential factors of location choice of FDI, researchers mainly analyze the variations in the recipients of MNEs’ FDI. Such variations consist of two types, one is variations in the host countries over time (Culem, 1988; Flores & Aguilera, 2007) and the other one is variations in the regions or cities within a specific country (Chung and Alcacer, 2002; Crozet et al., 2004).
developing and needs more and further insights. The research aim of this thesis is to identify the changes of influential factors of FDI location choice in a specific country. China has kept a relatively high-speed development of economy and has been the second largest recipient of FDI in the world (Chen & Yeh, 2012). At the meantime, the outflows of FDI from China also experience an obviously upward trend and gradually take a significant position in the global market. Chinese firms perform much more animatedly and actively. Since 2003, the outward FDI from China has increased remarkably (Ramasamy et al., 2012). In 2012, the sum of the outflows of China’s FDI reaches 87.8 billion US dollars, increasing by 17.6% compared to that of 2011 (Statistical Bulletin of China’s Outward Foreign Direct Investment, 2012). According to World Investment Report 2013, China’s outward FDI in 2012 occupy 6.3% of the aggregate flows and 2.3% of the aggregate stocks of the world, ranking the third in term of global FDI flows and the thirteenth in term of global FDI stocks (UNCTAD, 2013). Due to the dramatic performance of China’s outward FDI in recent years, much research has focused on this enormous and developing economy, such as Buckley (2007), Chen and Yeh (2012) and Ramasamy et al. (2012). However, many articles concentrate on identifying the determinants of location choice of the outward FDI from China. They do not further analyze the changes in the locational determinants (Broadman & Sun, 1997; Cheng & Kwan, 2000a; Belderbos & Carree, 2002). This thesis aims to fill this research gap. By this way, this thesis will make three contributions. For theory development, this thesis makes a contribution to explaining FDI activities in different periods by identifying the changes in the influential factors. For Chinese firms, this thesis will analyze their choices of locations when investing abroad and their investment motives to find out what changes have happened to the
influential factors of FDI location choice. In this way, Chinese firms can draw lessons from the findings and utilize the investment opportunities more effectively. Besides, for Chinese governments of all levels, they can get information about the current trend of FDI and what Chinese firms value in their location-decision making and thus get to know how to help these firms to behave more competitively in the global market.
2. Literature review
Foreign direct investment has always been a significant and interesting research field of economists and international business scholars. Many FDI theories have come into being so far. For example, Hymer (1976) starts to explain FDI phenomenon by using the industrial organization theory and moves the object of research to multinational enterprise (MNE). Another stream of FDI theory considers FDI as an international exchange of assets (Aliber, 1970; Dunning & Rugman, 1985). There are also many branches of FDI research, such as FDI’s risk (Bitzenis & Marangos, 2008; Chang & Lu, 2012), entry modes and establish modes choice (Kim & Hwang, 1992; Morschett et al., 2010), management across border (Guimon & Filippov, 2012), performance evaluation (Chang & Rhee, 2011; Jean, 2011) and so on. Among these branches, location choice is a hot spot of research and the success of location choice is of vital importance for firms to gain profits from FDI (Chen & Yeh, 2012).
2.1 Influential factors of FDI location choice
with its development, two main genres of theories take shape (Hong, 2007). The first one indicates that firms choose the locations, which can provide local advantages, as the recipients of their FDI (Blonigen et al., 2003). This view is derived from traditional industrial location theory and new location theory (Wei et al., 1999). It assumes that firm specific advantages are given (Caves, 1971). When firms carry out investments overseas, they mainly consider locational attributes in the location-decision-making process (Wei et al., 1999). The locational attributes do not need to provide assets that firms do not own. If a firm expects to consolidate its existing advantages by foreign expansion and a certain place has a distinct locational attribute that is beneficial to enhance the firm’s advantages, the firm is rather likely to invest in this place. In this case, the recipient of FDI does not provide the firm with something new, but just helps the firm enhance its existing advantages. However, if a firm expects to explore new assets and internalize them, the recipients of FDI are better to have the locational attributes that are not owned by the firm and easy for the firm to utilize and absorb.
Locational attributes of places are the external factors that influence the location choice of FDI (Dunning, 1981). The existing research identifies many external factors. Firstly, labor factor has a significant effect on the distribution of FDI (Cheng, 2006). It consists of labor cost (Wheeler & Mody, 1992; Cheng & Kwan, 2000a) and labor quality (Cheng & Kwan, 2000b; Belderbos & Carree, 2002). Labor cost is intuitively thought to negatively affect the location decision of FDI and research has demonstrated this statement (Wheeler & Mody, 1992). However, some other studies have the opposite results: high labor cost is an attractive factor to firms in some conditions. This can be explained by the phenomenon that high labor
cost means high labor quality (Cheng & Kwan, 2000a). Secondly, market size or market demand is also a critical condition of location choice of FDI (Cheng & Kwan, 2000a; Zhou et al., 2002). The findings about market are more uniform than those about labor factor. Generally, firms are more willing to invest in the places where market demand or the potential of market demand is large (Zhou et al., 2002). Thirdly, infrastructure plays an important role in the location-decision making process (Head et al., 1995; Du et al., 2008a). A relative good condition of infrastructure make it easier (reduce cost or enhance efficiency) for firms to carry out FDI. Fourthly, preferential government policies are big attractiveness to FDI (Head & Ries, 1996; He, 2002). The case of policy is similar to that of infrastructure. Preferential policies can help firms to cope with difficulties in doing business abroad and thus attract inward FDI. Last but not least, agglomeration effect also influences location decisions (Head et al., 1995; Belderbos & Carree, 2002; He, 2002). Belderbos and Carree (2002) find that agglomeration effect helps establish good environment for foreign firms to allocate resources, reduce costs and create scale economy. Its positive effect on FDI is not direct but becomes increasingly obvious in the process of FDI.
The other genre of theories advocates that location choice of firms depends on their motives of FDI and thus the motives are considered as internal factors influencing the location decisions (Chen & Chen, 1998). The most famous finding about this genre is Dunning’s (1998) four primary motives of FDI, which include resource-seeking, market-seeking, efficiency-seeking, and strategic asset-seeking motive. Resource here does not only refer to natural resource but also covers something else. For instance, labor is a kind of important resource which firms search in a worldwide scope (Bitzenis et al., 2007).
Market-seeking motive does not only mean finding big-size markets. Some firms pay more attention to markets’ potential of growth, compared to the size of markets (Bitzenis et al., 2007). As for efficiency-seeking FDI, firms want to make use of factor endowments, preferential policies and so forth to reduce costs or increase profits (Rugman, 2010). The strategic assets that firms tend to seek include knowledge, technology, experience and the like. This kind of FDI mostly happens in developed economies, because such strategic assets are abundant there (Dunning, 1998). Another research also focuses on the motives of FDI. It asserts that firms carry out FDI either for exploiting existing asset or for seeking assets and distinguishes the motives of FDI in less developed countries from those in developed countries (Makino et al., 2002). Firms generally invest in developed countries to seek asset that they do not own, such as technology. Asset-exploitation FDI mainly occurs in developing countries (Makino et al., 2002). Firms expect to transfer their assets or advantages and exploit them in the host countries (Chen & Yeh, 2012). In fact, these two motives (Makino et al., 2002), asset-seeking and asset-exploitation, are an extension of Dunning’s (1998) statements.
In all, the research on the influential factors of FDI location choice consists of these two genres. However, these two genres are not completely irrelevant but have overlapped points (Chen & Yeh, 2012). Some of the locational attributes, such as cheap labor, big-size markets, and advanced technology, are the motives of firms to carry out FDI.
2.2 Changes in the influential factors of FDI location choice
Although many influential factors come into being automatically in the initial history of FDI, they do not keep constant during the long development. With the increasingly deeper
extent of regionalization and globalization, the influential factors of FDI location choice have experienced changes. These changes attract some scholars’ attention. Dunning (2004) indicates that seeking resource has not been the main motive for firms to carry out FDI recently, but the numbers of mergers and acquisitions (M&A) and efficiency-seeking FDI are rapidly increasing. These two types of FDI occupy a dominant part of overall FDI. Firms expect to acquire local advantages, such as capital capability, R&D intensity and preferential policies, by means of M&As (Dunning, 2004). That is to say, the better institutional environment of host countries is more and more important for firms (Dunning, 2004). Efficiency is the other critical factor that firms care about in the location decision-making process. This is a consequence of regionalization and the increasing openness of worldwide market (Dunning, 2004). In addition, strategic assets gradually replace the position of resource in overseas expansion, and this indicates the development of FDI has gone up to a higher level (Dunning, 2004).
Nunnenkamp (2002) also advocates that the importance of determinants of location choice of FDI has changed. Traditional factors, such as market and labor, are not as crucial as what they were. Conversely, infrastructure condition, cost differences between locations, and agglomeration effect have become more important when firms choose recipients of their overseas investments (Nunnenkamp, 2002). However, though traditional and emerging factors have such opposite phenomena, the trend of traditional factors’ decline is not so strong. Firms still cannot neglect the role of traditional factors when making location decision (Nunnenkamp, 2002). After investigating US MNEs’ international activities from 1980 to 2000, Flores and Aguilera (2007) find that institutional-cultural factors moderate the
influence of market factor on FDI location choice. In other words, traditional factors’ influence on location choice may be moderated by some emerging factors. These emerging factors begin to work, though their significance has not been so strong. In addition to the above finding, large-size markets are found to be not that attractive to firms. Besides, US MNEs prefer to choose high-wage countries rather than low-wage ones, inferring that low labor cost may be also decreasingly critical (Flores & Aguilera, 2007). Since high wage can be a symbol of well-educated labor (Flores & Aguilera, 2007), it indicates labor quality becomes more critical in location choice. Although previous study has stated that a friendly policy environment is a big attraction for FDI (Head & Ries, 1996; He, 2002), firms currently prefer to choose locations with less government intervention and corruption rather than choose locations with perhaps preferential but complex policy conditions (Du et al., 2008a). In their views, complex policy environment can increase liability of foreignness and multinationals and they are not capable enough of distinguishing preferential and adverse policies.
The findings mentioned above, though not covering all of the changes, reveal the general trend of changes in the influential factors of FDI location choice. In addition to the direct findings of such changes, variations in the FDI recipients can also indicate indirectly the changes in the influential factors. The variations in the locations of accepting FDI consist of two types. The first one is that due to firms conduct more and more overseas investments over time, they choose different countries or regions as the recipients of their FDI. This type of variations is popular in the researchers to analyze the relevant topics. To identify the location determinants of FDI in industrial countries, Culem (1988) investigate 30 pairs of
counties from 1969 to 1982 and find that market growth rate and tariff barriers are both becoming increasingly critical. Flores and Aguilera (2007) use this type of variations to conduct their research too. They divide the regions which accept FDIs from US MNEs from 1980 to 2000 into three parts, Ohmae’s (1985) TRIAD region, Rugman and Verbeke’s (2004) core region and UN’s (United Nations’) region. By investigating the different amounts of FDI from US MNEs in those three regions, they find the decline of market and low labor cost and the rise of institutional-cultural factors and labor quality as above mentioned (Flores & Aguilera, 2007). This type of variations is useful to reveal the changes in the influential factors, but previous research mainly stops with the analysis of determinants of location choice rather than goes further in analyzing the changes (Cheng & Kwan, 2000a). Consequently, this gap of research requires more attention. Moreover, to investigate the influential factors of FDI location choice of a country, this type of variations is easy to collect data and analyze. This thesis uses this type of variations to analyze the influential factors of the China’s outward FDI location choice.
The second type of variations in FDI recipients is that with the process of conducting FDI, firms choose different regions or cities within a specific country. Chung and Alcacer (2002), by examining the FDI location choices from OECD countries to the US between 1987 and 1993, find that knowledge-seeking FDI plays a more important role in FDI location choice. Moreover, to demonstrate the agglomeration effect on location choice, based on a sample of FDI in French 92 locations during 10 years, Crozet, Mayer and Mucchielli (2004) find that in contrast to agglomeration effect, preferential policy lose its positive impact on location choice.
2.3 Why do influential factors of FDI location choice change over time?
There is some research on FDI explaining the reasons why determinants of location choice change over time. Firstly, such changes can be explained by organizational learning theory. Organizations are routine based and history dependent and their previous experience can influence their following decisions (Zhu et al., 2012). In the case of FDI, scholars also find that the previous FDI experience can affect the subsequent location choices (Lin & Yeh, 2004). The experience of FDI can moderate the impacts of some factors on location choice. The previous experience here does not only refer to the firms’ own experience, but also include the experience of other firms, like early movers (Zhu et al., 2012). Firms face the liability of foreignness and multi-nationalities when they conduct FDI outside home countries. Learning from their own experience, local companies and early movers helps firms get to know about local market, industry competition, and customer preferences and thus overcome the liability of foreignness and multi-nationalities (Greve, 1998). If a place is attractive to firms for its big market size but the market competitions are out of order, firms may not choose this place after knowing about the local market chaos. In this way, the learning effect leads to different extents of importance of locational influential factors of FDI (Davidson, 1980). Besides, FDI location choice is a sequential process (Mataloni, 2011). Benito and Gripsrud (1992) advocate that the sequent location choices of FDI are not independent but interrelated. This also demonstrates the influence of learning effect on the change of influential factors.
Secondly, the factors influencing location choice of FDI are based on a static analysis. In fact, firms’ advantages and capabilities keep varying over time. With the development of
firms, some locational attributes may be less attractive to them because they acquire advantages of these attributes by exploiting own assets and capabilities. Dunning’s (2000) eclectic paradigm explains that firms carry out FDI to exploit their ownership advantages and acquire local advantages in host countries. There is a trade-off between firms’ own assets and capabilities and locational attributes. The development of firms’ capabilities can cause the decreasing importance of some determinants of location choice of FDI. For instance, if an MNE improves its R&D capability and gains a lot progress on technology, it may not invest in locations where technological advantage is outstanding.
The above two causes of changes in influential factors are internal causes. It is the firms’ own knowledge and capabilities changes that lead to their different evaluations of influential factors. In addition to internal causes, the changing institutional environment, as an external cause, also can bring about changes in locational determinants of FDI. Institutional environments consist of all levels of government institutions (Du et al., 2008b). Traditional institutional factors, such as preferential policies, though still having impacts on location decisions of FDI, get less significant than ever before (Du et al., 2008a). The changing institutional environments generate some emerging influential factors and these emerging factors lead to the decreasing significance of other factors. For example, though there are preferential policies to attract foreign investors, the serious government corruption will impede inward FDI. Moreover, less government intervention are more preferential for foreign firms. The overall policies and regulations may be helpful for foreign firms to invest in local markets, but the complexity may confuse them and slow down the pace of conducting FDI (Du et al., 2008a).
In summary, it is obvious that the factors influencing location choice of FDI are changing over time. The above-mentioned research reveals the general changes of determinants of location choice of firms. However, the research on how influential factors change focusing on outward FDI of a specific country, particularly in China, is limited. China has increasingly opened its door to the world since it joined WTO (the World Trade Organization) in 2001. The economic activities and collaborations between China and other countries and regions have become much more frequent and larger-scaled. The outward FDI from China’s firms also make an obvious progress under such condition. This thesis will concentrate on the influential factors of location choice of outward FDI of China to analyze what changes in these factors have arisen over time. In this way, the research gap can be filled and Chinese firms can integrate the local business conditions of host countries with their own assets and capabilities, and thus choose the locations which are more appropriate to their international expansions.
3. Theoretical framework and hypotheses
3.1 Theoretical framework: resource-based and institution-based view
In the last section, two main genres of research on influential factors of FDI location choice are mentioned, one is locational advantages (Blonigen et al., 2003) and the other is FDI motives (Dunning, 1998; Makino et al., 2002). Although these two genres represent two angles to investigate influential factors, they share one theoretical background, which consists of resource-based view (RBV) and institution-based view (IBV) of firms.
Based on RBV, firms’ competitive advantages are generated from their heterogeneous resources (Barney, 1991). These resources should meet four criteria: they should be valuable, rare, inimitable and non-substitutable (Barney, 1991). The resources which meet these four criteria can provide firms with advantages which cost other firms too high to copy them. And thus, such competitive advantages keep firms living in the fierce market competition or even help them win a dominant position in the market. Some research on FDI corresponds with RBV. The heterogeneous resources are not only owned by firms, but also owned by their home country. Firm specific advantages (FSA) and country specific advantages (CSA) are two significant sources of firms’ competitive advantages (Rugman and Li, 2007). Generally, firms decide to invest abroad to acquire FSAs which they lack and CSAs which their home countries lack (Rugman and Li, 2007). Hence, according to RBV, four influential factors are involved in this thesis to investigate their changes, natural resources, labor cost, market size and technological capability. These four influential factors also overlap with the two mentioned theory genres. Natural resources, markets size and technological capability (as a strategic asset) belong to the four primary motives summarized by Dunning (1998) and labor cost obviously is considered as a locational attribute (Wheeler & Mody, 1992; Cheng & Kwan, 2000a).
IBV has a different perspective of firms’ decisions from that of RBV. Firms obtain legitimacy from obeying the rules and beliefs in the institutional environment (DiMaggio and Powell, 1983). Firms must firstly think about rules and beliefs instead of their assets and resources when they make a decision (DiMaggio and Powell, 1983). This is obviously distinct from RBV. When IBV is extended in the field of FDI, research indicates that both
home countries’ and host countries’ institutional environments influences the activities of internationalization (Zhang et al., 2011). Home countries’ institutional environment can attract inward FDI and host countries’ institutional environment can attract outward FDI (Zhang et al., 2011). Governments can make local institutional environments more preferential for foreign firms by changing rules and regulations and can also help create a more open environment. In some cases, an open and convenient business environment is more important than resource-based factors, because firms need to acquire legitimacy from the institutional environment at first and then carry out business activities. If the institutional environment of a country is hard for foreign firms to get legitimate permission to make investments, they are definitely not willing to go there no matter how attractive resources the country has. What is more, the institutional barriers can increase the costs of acquiring the specific advantages in the locational firms and host countries. Firms may give up the investing opportunities after they balance the costs and value of the resources. In sum, institution-based factors deserve attention of scholars. Hence, based on IBV, three influential factors are involved in this thesis, tax rate, agglomeration effect and infrastructure. As is known to all, these three factors are typical locational attributes to attract FDI (He, 2002). Besides, they are also useful to increase efficiency for firms and thus are consistent with the efficiency-seeking motive (Dunning, 1998).
Table 1 Theoretical model
Theoretical background
Resource-based view Institution-based view
Influential factors Natural resource Market size Techno-logical capability Labor cost Tax rate Agglomeration effect infrastructure Genres of research on influential factors FDI motives: Resource-seeking Market-seeking Strategic asset-seeking L oc ati on al att rib ute Locational attributes/ FDI motive: efficiency-seeking
3.2 Hypotheses: influential factors in two periods
3.2.1 Natural resources
Seeking resources is one of the four primary motives of FDI defined by Dunning (1998). Firms expect to get access to the natural resources which are scarce or costly in their home countries through making investments abroad. Based on Dunning’s (2000) eclectic paradigm, natural resources, as an immobile location attribute, are generally used by firms jointly with their own advantages in FDI. Natural resources have been one of the most significant firms’ advantages because of its value, rarity and inimitability. As a result, they are irreplaceable in the field of location choice of FDI.
since the country needs natural resources to guarantee its economic development (Ramasamy et al., 2012). According to the current national condition of China, the high-speed economic growth still needs cheap and a large amount of resource. For instance, Chinese firms (most are supported by government) make many investments in Africa to ensure the supply of natural resources in spite of political unrest (Ramasamy et al., 2012). But, because of the transformation of China’s economy and the development of Chinese medium-sized multinational firms, natural resources are not as important as what they used to be (Ramasamy et al., 2012). So it is logical to predict that natural resource gets decreasingly significant as an influential factor of FDI location choice of China.
Hypothesis 1a: China’s outward FDI is influenced by natural resources in the initial period of FDI.
Hypothesis 1b: China’s outward FDI is influenced by natural resources in the later period of FDI.
3.2.2 Market size
Seeking market is another primary motive of FDI (Dunning, 1998). Market is a traditional and significant influential factor of FDI flows. To increase sales volume and revenue, firms race to take over foreign markets. In such competition, the size of market is of vital importance. The larger the size of market, the more likely can resource be allocated efficiently and can scale economy be established (Buckley et al., 2007). Besides, a large size of market can help firms broaden their business scopes.
Market-seeking outward FDI previously did not occupy a considerable part of the overall outward FDI of China. The local market has enough capacity to consume the products and services of Chinese firms. However, many findings indicate that Chinese firms have begun to carry out FDI for exploring overseas market and these investments seem to get more and more active (Deng, 2004; Buckley et al., 2007). This phenomenon suggests that many Chinese firms begin to expect to develop and improve themselves and join in the international competition by FDI (Deng, 2004; Buckley et al., 2007). Hence, market size is considered to be an increasingly important influential factor of China’s outward FDI.
Hypothesis 2a: China’s outward FDI is influenced by market size in the initial period of FDI.
Hypothesis 2b: China’s outward FDI is influenced by market size in the later period of FDI.
3.2.3 Technological capability
Technological capability is a significant strategic asset to firms both from developed countries and from developing countries. Innovation is now thought to be the core of firms’ vitality and technological capability is the principle factor to keep innovation (Florida, 1997). Due to the limitations of home countries or the remarkable advantages of other countries, firms explore technological assets to increase their R&D capability by FDI (Florida, 1997). Chung and Alcacer (2002) consider this kind of FDI as knowledge-seeking FDI and in their view, this kind of FDI has become a trend of foreign firms’ inward FDI in the United States.
carry out FDI. At that time, Chinese firms which are able to carry out outward FDI are mostly state-owned firms and natural resources are what they want most (Ramasamy et al., 2012). But as globalization develops, more Chinese firms want to participate in global competition and FDI is their main approach to enter global market. To shorten the gap between strong multinational firms from developed economies and themselves and transform themselves to technology-intensive firms, Chinese firms tend to explore technological assets in other countries to enforce their R&D capabilities (Ramasamy et al., 2012). Therefore, technological capability is increasingly significant to Chinese firms in the field of FDI.
Hypothesis 3a: China’s outward FDI is influenced by technological capability in the initial period of FDI.
Hypothesis 3b: China’s outward FDI is influenced by technological capability in the later period of FDI.
3.2.4 Labor cost
Labor is an important input factor to firms, especially to the labor-intensive industry. Labor cost is a critical part of production cost of firms. If the wage level is higher than firms’ expectation, firms will transfer their labor-intensive production segments to countries where labor cost is lower, as long as the cost of investing abroad is less than the labor cost gap between two countries. Low-cost labor is also a locational advantage that firms expect to gain by FDI, just as natural resources.
China has the largest population in the world and thus has huge labor force reserves. There is much previous research that demonstrates low labor cost is a highlighted locational
attribute of China (Broadman & Sun, 1997; Cheng & Kwan, 2000a; Kang & Lee, 2007). However, labor cost of China has increased with the rising level of economy. Compared to the labor costs of some Southeast Asian countries, the low-cost labor forces of China is as attractive as before (Chen & Yeh, 2012). For example, many Japanese firms are observed to transfer their Chinese factories to Southeast Asian countries to reduce costs (Chen & Yeh, 2012). Consequently, Chinese firms are also likely to invest in the above-mentioned countries for lower labor cost. Besides, the short geographical distances between China and these countries increase this possibility. So the importance of labor cost is considered to become larger.
Hypothesis 4a: China’s outward FDI is influenced by labor cost in the initial period of FDI.
Hypothesis 4b: China’s outward FDI is influenced by labor cost in the later period of FDI.
3.2.5 Tax rate
Tax rate can influence many decisions of firms in various aspects and one of these decisions is FDI location choice. Findings of Bartik (1985) demonstrate that corporation tax rate affects business location choice within the US. Although this article focuses on business investments within a specific country, it also indicates the influence of tax rate on overseas investments. Taxation on firms is a significant influential factor, since it is related to gains of firms from overseas investments. Tax policy, as a local attribute, is often implemented by governments to attract inward FDI. Low corporation tax rate can reduce the loss of profit
from FDI and thus enhance the willingness of firms to invest. Buettner and Ruf (2007) investigate the relationship between tax incentives and German firms’ FDI location choice. They find that taxation on corporation generally has an important impact on location choice. But they divide tax incentives into various indicators and finally have different results (Buettner & Ruf, 2007). This thesis does not dig so deeply in tax rate. Here tax rate just consist of two types, high and low ones. Locations with lower tax rate are considered more attractive for FDI.
Hypothesis 5a: China’s outward FDI is influenced by tax rate in the initial period of FDI. Hypothesis 5b: China’s outward FDI is influenced by tax policies in the later period of FDI.
3.2.6 Agglomeration effect
Agglomeration is a kind of phenomenon that firms concentrate in a particular location and thus generate positive externalities (Head et al., 1995). This is a very significant factor to attract FDI. Agglomeration does not only reduce transport cost but also ensure enough demand to high specialized components (Head et al., 1995). In addition, agglomeration may induce positive knowledge spillover among firms (Myles & Flyer, 2000). Similar to large market size, agglomeration effect can allocate resources more efficiently and make it easy to establish scale economy. This can help firms overcome the liability of foreignness and adjust to the local market quickly. Most of Chinese firms, which begin to carry out FDI recently, are not state-owned firms or large multinational firms (Ramasamy, et al., 2012). They lack experience to cope with FDI problems and may not have enough financial and human capital
to address big problems. They are more willing to choose FDI-concentrated places so that they can learn from early movers. As a result, Chinese firms without enough FDI experience will prefer to invest in locations which have agglomerate multinational firms. Thus, it is logical to consider agglomeration to be more important.
Hypothesis 6a: China’s outward FDI is influenced by agglomeration effect in the initial period of FDI.
Hypothesis 6b: China’s outward FDI is influenced by agglomeration effect in the later period of FDI.
3.2.7 Infrastructure
The influence of infrastructure on FDI is not as direct as that of other factors, like market size and labor cost. But its influence should not be ignored in the FDI process. Infrastructure is related to information cost (He, 2002). Firms come across liability of foreignness in unfamiliar markets. Such liability of foreignness refers to a lack of information about local markets or a lack of approaches to the information. Therefore, the cost of obtaining information to foreign firms may be very high (He, 2002). Despite all this, information cost can be decreased considerably by good infrastructure condition. He (2002) indicates that business information of local markets is easy to obtain in the locations where infrastructure is developed and the information cost is also relatively low. Hence, good infrastructure is attractive to FDI projects.
In the initial period of going out, Chinese firms’ overseas investments were mostly motivated by taking advantage of natural resources of other countries, so infrastructure was
not an important consideration at that time (Ramasamy et al., 2012). But as more Chinese firms begin to invest abroad, various motives come into being and thus make Chinese firms pay more attention to the infrastructure condition of host countries. In this sense, infrastructure is also considered to become more significant as an influential factor of China’s outward FDI.
Hypothesis 7a: China’s outward FDI is influenced by infrastructure condition in the initial period of FDI.
Hypothesis 7b: China’s outward FDI is influenced by infrastructure condition in the later period of FDI.
3.2.8 Political risk
Buckley et al. (2007) state that in general, high political risk will discourage the inward FDI flows, because substantial sunk cost is hard to take back if the political situation is not stable. However, high political risk sometimes means a high return rate (Buckley et al., 2007). In the case of China’s outward FDI, political risk is highly related to natural resource suppliers. Many Chinese FDI recipients are located in Africa, since these countries can provide various metal and ore resources. Meanwhile, the political situations in these countries are rather unstable and foreign investment projects cannot be well protected. Despite this condition, these countries attract a huge flow of Chinese FDI. This is because resource-seeking FDIs are mostly carried out by Chinese state-owned firms. These firms are more acceptable to invest in highly risky countries due to their governmental background (Ramasamy et al., 2012). Hence, political risk is also worthy to take into account as a control
variable.
3.2.9 Exchange rate
Previous research demonstrates that a low exchange rate is beneficial to export but not to FDI (Buckley et al., 2007). A low exchange rate means that it costs firms more to acquire host countries’ locational preferential attributes. In addition, the cost of transferring their own specific assets to host countries may also increase. However, if the exchange rate rises (home country’s currency appreciates), the prices of host countries’ locational attributes will be lower, and thus firms will be more willing to carry out FDI to those countries. China adopts a semi-floating exchange rate regime (Hall, 2003). Although its currency, Renminbi (RMB), has kept a peg against the US dollar, the real exchange rate of RMB has risen by about 20% since 1995 (Hall, 2003). Hence, the influence of exchange rate on China’s outward FDI cannot be neglected. So exchange rate in this thesis is involved as a control variable.
4. Methods
4.1 Data collection
This thesis includes two main datasets to explore the relationship between China’s outward FDI and the above-mentioned influential factors. The first dataset is the Statistical Bulletin of China’s Outward Foreign Direct Investment. The statistical bulletins are issued by Ministry of Commerce of the People’s Republic of China, National Bureau of Statistics of the People’s Republic of China and State Administration of Foreign Exchange (China). These
bulletins are now the most authoritative dataset of China’s outward FDI. They have been issued by 10 times and record the data of China’s outward FDI from 2003 to 2012. As what is stated in the hypotheses section, to identify the changes, the time period of China’s outward FDI studied in this thesis is divided into two parts, the initial period (2003-2007) and the later period (2008-2012). Since the sub-prime crisis occurred in 2007 and it had an enduring and impeditive influence on the following cross-border commercial activities around the world, China’s FDI outflows are separated into the outflows before 2008 and those after 2008. The other significant dataset is World Development Indicators from the World Bank. These indicators reveal various characters at the country (or region) level, which can represent most of the influential factors of FDI. Eight of the variables included in this thesis use World Development Indicators as their proxies. The proxy for other variables (technological capability and political risk) is mentioned in the next section.
4.2 Measurement
Table 2 Proxies for variables
Influential factor
Proxy Variable
type
Source
Outward FDI Annual outward FDI flows by country and region
Dependent Statistical Bulletin of China’s Outward Foreign Direct Investment (2003-2012) Natural
resource
NR: Ores and metals exports (% of merchandise exports)
Independent World Development Indicators (the World Bank)
Market size MS: GDP (current US$) Independent World Development Indicators (the World Bank)
Technological capability
TC: Annual patents
application by residents and nonresidents
Independent World Intellectual Property Organization
Labor cost LC: GNI per capita based on PPP (current international $)
Independent World Development Indicators (the World Bank)
Tax rate TR: Taxes on income, profits and capital gains (% of revenue)
Independent World Development Indicators (the World Bank)
Agglomeration effect
AE: Foreign direct investment, net inflows (current US$)
Independent World Development Indicators (the World Bank)
Infrastructure I: Goods transported via air, roads and railways (million ton-km)
Independent World Development Indicators (the World Bank)
Exchange rate ER: The values of the ten years’ official exchange rates
Control World Development Indicators (the World Bank)
Political risk PR: The values of the ten years’ political risk indexes
Control International Country Risk Guide
4.2.1 Dependent variable
The dependent variable is the annual China’s outward FDI flow by country and region from 2003 to 2012. 51 countries (and regions) are involved in this thesis as the host countries of China’s outward FDI (see Appendix 1). These 51 countries (regions) are ranked in top 55 recipients of China’s outward FDI based on their stocks of Chinese FDI in 2012 (Statistical Bulletin of China’s Outward Foreign Direct Investment, 2012). Four countries (regions) are not included in the study, because they lack too much data in independent variables.
4.2.2 Independent variable
The independent variables are the proxies for the seven influential factors: natural resource, market size, technological capability, labor cost, tax rate, agglomeration effect, and infrastructure. The selection of proxies for natural resource, market size and technological capability is suggested by Buckley et al. (2007) and can be seen from Table 2. To measure labor cost of host countries, this study transfers labor cost to income level so as to find an appropriate proxy. The World Bank uses PPP (purchasing power parity) rate to convert GNI (general national income) per capita to US dollar to measure income level and thus this study select this indicator to represent labor cost. To measure the effect of tax rate on FDI, the ratio of taxes to revenue is chosen as proxy, since the amount of profits that firms can finally gain from foreign countries is what they most care about. Agglomeration effect means a lot of foreign firms invest in a particular location. This can be seen as a huge inward FDI flow located in a host country, so net inflows of FDI to host countries are chosen to represent agglomeration effect. To measure infrastructure condition, this thesis uses the weights of
goods transported via air, roads and railways during the ten years. The larger the weight of goods is, the more goods are transported and the better the infrastructure of this country is.
4.2.3 Control variable
As to exchange rate, the values of the ten years’ official exchange rates are used as proxy. International Country Risk Guide provides a list of political risk index every year and the values of data from 2003 to 2012 are used in the study.
4.3 Model
Previous research on FDI has demonstrated that there is no linear relation between the dependent variable and the independent variables (Buckley et al., 2007). So the data of all the involved variables are transformed into their natural logarithms. In this thesis, a log-linear model is implemented as follows:
LOFDI = α+ β1LNR + β2LMS +β3LTC+β4LLC +β5LTP +β6LAE + β7LI + β8LER + β9LPR + eit
i = 1, 2, ……, 51 (stand for countries) ; t = 2003, 2004, ……, 2012 (stand for years)
4.4 Pretreatment of data
4.4.1 Missing data
The data of this thesis are analyzed by using Eveiws version 7.0. There are some missing values in the variables. Since most of the variables are continuable variables, hotdeck imputation is not suitable for the data under study. The number of selected countries is 51 and
the sample size is not very large. This thesis chooses to replace the missing values with the serious means so as to ensure the sample size.
The descriptive statistics of data in the two periods can be seen in Appendix 2a and 2b. The values of Skewness (0.072 and -0.064) and Kurtosis (3.311 and 3.986) of outward FDI show that the data are nearly normally distributed. The standard deviation values of exchange rates are the biggest, 2.393 and 2.419. This suggests a big scale of fluctuation in the recipient countries’ exchange rates. Finally, the observations of all variables are 255, indicating that all missing values are substituted by serious means. Besides, to get rid of the influence of different dimensions, the data are standardized.
4.4.2 Correlation matrix
Table 3a Correlation matrix for data in the initial period (2003-2007)
LOFDI LNR LMS LTC LLC LTP LAE LI LER LPR LOFDI 1.000 LNR 0.083 1.000 LMS 0.223 -0.130 1.000 LTC 0.334 -0.020 0.873 1.000 LLC 0.141 -0.131 0.512 0.479 1.000 LTR 0.209 0.202 0.328 0.356 0.037 1.000 LAE 0.291 0.015 0.773 0.644 0.622 0.227 1.000 LI 0.206 0.094 0.723 0.692 0.423 0.100 0.617 1.000 LER -0.245 0.351 0.071 -0.131 -0.014 0.143 0.110 0.165 1.000 LPR 0.043 -0.008 -0.320 -0.084 -0.420 -0.137 -0.461 -0.279 -0.403 1.000
Table 3b Correlation matrix for data in the later period (2008-2012)
LOFDI LNR LMS LTC LLC LTP LAE LI LER LPR LOFDI 1.000 LNR 0.179 1.000 LMS 0.246 -0.086 1.000 LTC 0.342 -0.028 0.888 1.000 LLC -0.038 -0.099 0.429 0.389 1.000 LTR 0.217 0.164 0.216 0.246 -0.139 1.000 LAE 0.327 0.168 0.666 0.623 0.477 0.117 1.000 LI 0.129 0.132 0.662 0.628 0.288 0.030 0.589 1.000 LER 0.012 -0.062 -0.336 -0.140 -0.414 -0.132 -0.362 -0.247 1.000 LPR -0.077 0.134 0.302 0.196 0.647 0.134 0.410 0.235 -0.579 1.000
In table 3a and 3b, there are high correlations between some pairs of independent variables (e.g. LMS and LAE/LI/LTC, LTC and LAE/LI, LLC and LAE, and LAE and LI). This can cause multi-collinearity in the regression estimation (Hsiao, 1985). However, these pairs of independent variables are still involved in this thesis, since the data used in this thesis are panel data. Different from time serious data’s analysis, the nature of panel data analysis takes multi-collinearity into consideration and handles this problem in the estimation process (Hsiao, 1985; Ranjan & Agrawal, 2011). Thus this will not have a bad influence on the results.
4.4.3 Unit root test and cointegration test
There are conflicts on whether all panel data should do unit root test and cointegration test. It is generally considered that if the time scale of data is small, then it is not necessary to do these two tests (Banerjee, 1999). The time scale of each group of the data under study is only five years, so the data do not have to pass these two tests. But in case the validity of the regression results was impaired, this thesis still conducts these two tests to the two groups of
data. The unit root test includes five test types in Eviews, LLC, IPS, Breintung, ADF-Fisher, PP-Fisher. All the ten variables pass four out of the five tests. Data are considered to have unit roots unless they pass none of the five tests (Banerjee, 1999). So the variables in this thesis can be thought to have no unit root. As to cointegration test, Eviews includes seven test types. The p values of five tests out of these seven are close to zero, indicating that there exist cointegration relationships between the dependent variable and the independent (control) variables. So the two groups of data can be used directly in the next step.
4.4.4 Hausman test and redundant fixed effects test
The panel data analysis has three different models, fixed effects (FE), random effects (RE) and mixed effects model (same as pooled ordinary least squares POLS). Hausman test can decide which effects model should be chosen from fixed and random effects model (Hauman, 1978). The two groups of data conduct Hausman test and the p values of the two groups are both nearly zero (see Appendix 3a and 3b). This means the null hypothesis is rejected, and the fixed effects model should be used (Hauman, 1978). So the two groups of data should use fixed effects model (Hausman, 1978). Redundant fixed effects test decides which effects model should be chosen from fixed and mixed effects model. The two groups of data conducts redundant fixed effects test and the p values keep being close to zero (see Appendix 3a and 3b). This suggests that compared to POLS, fixed effects model is more suitable for the two groups’ data. Hence, the statistic model used in this thesis is fixed effects model.
5. Results
Table 4a Estimation results for the data in the initial period (2003-2007) FE
Coefficient Std. Error P value
LNR 0.026 0.228 0.910 LMS 3.289*** 1.053 0.002 LTC 0.117 0.316 0.711 LLC -1.977 1.305 0.132 LTR 0.385* 0.221 0.083 LAE 0.143 0.122 0.241 LI -0.053 0.126 0.674 LER 1.580 2.144 0.462 LPR 0.282 0.336 0.403 Adj. R2 0.606
Table 4b Estimation results for the data in the later period (2008-2012) FE
Coefficient Std. Error P value
LNR 0.068 0.274 0.804 LMS -0.698 0.617 0.259 LTC -0.656* 0.340 0.055 LLC 2.996*** 0.893 0.001 LTR 0.118 0.201 0.556 LAE 0.034 0.085 0.688 LI -0.138 0.122 0.262 LER 0.421 0.907 0.643 LPR -0.387 0.360 0.283 Adj. R2 0.699
Note: ***, **and * mean the coefficient is significant at the 1%, 5% and 10% levels.
Table 4a and 4b include the panel data estimation results of fixed effects model. Firstly, the values of adjusted R2of the two groups’ data are 0.606 and 0.699. Both values are more than 0.6, indicating the independent and control variables included in the model can explain most of the changes in the dependent variable. Secondly, the estimation results of independent variables show obvious difference between the two periods. From the two tables
it can be seen that none of the independent variables are significant simultaneously in both periods. Some of them are significant in the initial period and others are significant in the later period. This indicates that different influential variables affect the outward FDI in the two periods. The coefficients of LNR (natural resources), LAE (agglomeration effect) and LI (infrastructure) are not significant in the two periods. Hence, hypothesis 1a&1b, 6a&6b and 7a&7b are all rejected. This means that in this study, natural resource, agglomeration effect and infrastructure condition are demonstrated not to have an influence on China’s outward FDI in the period from 2003 to 2012. As to LMS (market size) and LTR (tax rate), their coefficients are significant in the initial period but not in the later period and both of them are positively related to the outflows of FDI. Hence, hypothesis 2a and 5a are supported, instead of hypothesis 2b and 5b. That is to say, the influences of market size and tax policy vary over time. These two are important influential factors for China’s outward FDI from 2003 to 2007. But their importance decreases over time and they do not influence outward FDI from 2008 to 2012. The variables LTC (technological capability) and LLC (labor cost) show an opposite trend. Their coefficients are not significant in the initial period but become significant in the later period. As a result, hypothesis 3a and 4a are rejected and 3b and 4b are supported by the estimation results. In other words, there are also changes in these two influential factors, technological capability and labor cost. China’s outward FDI are not influenced by these two factors during the period from 2003 to 2007. But as time passes, their significances gradually increase. In the period from 2008 to 2012, they are critical influential factors of the outflows of FDI. However, the relationships of these two factors between the outward FDI show a difference. Technological capability is negatively correlated to the outward FDI but labor cost
has a positive relationship.
Next, the above findings are discussed in detail. Market size and tax policies both have a positive influence on China’s FDI outflows during the initial period. The coefficient of LMS (3.289) indicates that the outflows of FDI increase 3.289% with a 1% increase on the recipient’s market size. The absolute value of this coefficient (3.289) is relatively high and this suggests that the effect of the market size of FDI recipients on FDI outflows is considerable. China joined in WTO (World Trade Organization) in 2001. From then on, Chinese market has been more open to the world and conversely Chinese firms have become more willing to go out and explore the giant global market. In addition, one of the dominant industries of China is labor-intensive industry. This kind of industry is characterized as small profit margin. In general, this industry should increase its sales to offset costs and further obtain profits. Hence, expanding markets around the world is an important motivation of Chinese firms in this industry. Based on the Statistical Bulletin of China’s Outward Foreign Direct Investment (2003-2007), firms in commercial service, manufacturing, and wholesale & retail industry conducted the great majority of FDI outflows in this period. Market is a significant factor for these three industries to win the fierce competition. So it is reasonable to say that the outward FDI of China in this period is highly motivated by market seeking. As to tax rate, this factor is positively related to the outward FDI, indicating FDI is more likely to flow to the countries with higher tax rates. But the absolute value of this factor’ coefficient is 0.385, lower than that of LMS. It means that the influence of tax rate is not as important as that of market size. To expand the global market, Chinese firms invest in countries with large market size. This kind of countries simultaneously has high tax rates. But since the main
motivation of FDI is market seeking, Chinese firms accepted their relatively high tax rates. This explains the positive relation between tax rate and the outward FDI from 2003 to 2007.
Then the outward FDI in the later period shows a different case. Market size and tax rate lose their influences on FDI and another two factors begin to affect the overseas investments from Chinese firms. There is a positive relation between labor cost and the FDI outflows and the absolute value of the coefficient (2.996) represents a rather heavy effect of labor cost. This is contrary to the assumption in the hypotheses section and is likely to be caused by the change of global environment. In the latter half of the year of 2007, a large-scale sub-prime crisis occurred in the US. Gradually, this crisis led to a lot of severe outcomes and spread to North America, Europe and other continents. A world-wide financial crisis was generated and had a lasting influence on the global investments during the period from 2008 to 2012. Though the financial markets of North America and Europe suffered serious defeats from the financial crisis, Chinese firms were not heavily influenced by the crisis and gradually found many opportunities of investments in the overseas financial markets. The Statistical Bulletins of China’s Outward Foreign Direct Investment (2008-2012) show that there are two obvious upward trends in the sum of aggregate investments in financial industry and the sum of investments in European area. One of the characteristics of financial industry is that the income of employees in this industry is very high. In other words, the labor cost of financial industry is relatively high. Since Chinese firms substantially increased their investments in financial industry, there is no doubt that China’s outward FDI was more likely to flow to countries’ with high labor cost. Besides, the coefficient of technological capability is negative and its absolute value (0.656) is smaller than that of labor cost. This also indicates a condition
contrast to the expectation in the hypotheses section. The change in the investments in European countries can explain this condition. During the later period (2008-2012), Hong Kong and EU (European Union) occupied over 60% of the sum of aggregate outward FDI from China. Especially in Europe, Chinese firms have found a lot of investment opportunities during these five years and the amount of investment to EU countries has increased substantially at that time (Statistical Bulletin of China’s Outward Foreign Direct Investment, 2008-2012). Hong Kong and most of the countries in EU do not have many annual patents applications but have a high number in GNI per capita. So these regions and countries show a relatively low level in technological capabilities and high level in labor cost. Because Chinese firms emphasized on financial investments in the later period, technological capability was not their first consideration. They were willing to invest in locations with favorable financial markets even though their level of technology was low. As a result, labor cost had a heavy and positive influence and technological capability had a slight and negative influence on China’s outward FDI in the later period.
As to the two control variables, LER (exchange rate) and LPR (political risk), are significant neither in the initial period nor in the later period. This indicates that in this study, exchange rate and political risk are not demonstrated to affect China’s outflows of FDI in both two periods.
6. Discussion
study in this thesis is based on two main genres of influential factors, FDI motives and locational factors (Blonigen et al., 2003; Dunning, 1998). Seven hypotheses are presented as the independent variables, covering both two genres of influential factors. Besides, the study also includes two control variables to ensure they cannot affect the validity of results. To explore the changes in influential factors, the data are divided into two groups in the initial and later periods. The changes can be seen from the differences in the independent variables in the two periods. The analysis model in this thesis is demonstrated as reasonable by previous research and is suitable for the data after passing various tests. The data under study are all collected from the authorities and thus their reliability can be guaranteed.
From the data analysis in this thesis, there indeed exist some changes in the influential factors of China’s outward FDI. In the initial period (2003-2007), market size and tax rate both had a positive relation with the outward FDI. Market size especially had a heavy influence on the location choice of FDI. This is consistent with many previous findings (Cheng & Kwan, 2000a; Zhou et al., 2002; Buckley et al., 2007). In other words, market-seeking FDI occupied an important role in the Chinese FDI activities in the initial period. However, market size lost its impact on FDI outflows in the later period. This is opposite to the previous assumptions. Based on previous research on China’s outward FDI, market-seeding FDI in the first five years were encouraged by reform and opening-up policy and the position as a member country of WTO to a great extent (Deng, 2004; Buckley et al., 2007). But as time passes, the influence of the policies and WTO is gradually weakened and compared to other assets, market size are not as attractive as before. This downward trend of market-seeking FDI agrees with the statement of Nunnenkamp (2002). Tax rate is a critical
factor of institutional environment. The analysis result of tax rate is opposite to the expectation in hypotheses section. The sum of FDI outflows increased with rising tax rates in this period. But the absolute values of coefficients indicate the influence of market size was heavier than that of tax rate. Same as market size, tax rate did not affect the outward FDI during the latter five years. As an institutional factor, tax rate’s important role during the initial period corresponds to the above-mentioned emerging force of institutional-cultural factors (Flores and Aguilera, 2007). Though its decline in the later period does not agree with Flores and Aguilera’s (2007) findings, it only means tax rate, this specific institutional factor, does not influence China’s outward FDI. And this may be generated by some other spring-up attractive locational attributes during the latter five years.
As to the later period (2008-2012), labor cost and technological capability affected the FDI outflows from China. But these two influential factors had converse effects and both of them are contrast to the expectations. Labor cost was positively related to the FDI outflows and the influence was considerable. Due to the financial crisis and the changing financial markets around the world, Chinese firms increased substantially their foreign investments in financial industry. The obvious feature of financial industry, high labor cost, finally led to the positive relation between labor cost and FDI flows. Flores and Aguilera (2007) state that FDI have been more willing to flow to the locations with high labor cost in recent years. They explain that is because high labor cost means high labor quality, and high labor quality is gradually emphasized by multinational firms. The rising of labor cost to China’s outward FDI is consistent with the two authors’ findings. Technological capability is also demonstrated to have an influence on FDI outflows, but their relation was negative in the later period. This
phenomenon can be explained by the changes in the composition of host countries. There was a dramatically upward trend of the proportion of investments in European countries because of the increasing financial investments. Besides, Hong Kong and European countries occupied nearly two-thirds of the aggregate outward FDI from China during the later period. Such regions and countries did not have many annual patents applications and thus showed a relatively low level in technological capability. As a result, the relation between technological capability and the FDI outflows was slightly negative. Generally, the upward trend of technological capability during the ten years conforms to the findings of Dunning (2004). As an important strategic asset, technological capability gradually replaces the dominant position of traditional factors, such as market size (Dunning, 2004). But though technological capability is significant during the later period, its negative relation with FDI outflows is contrast to Dunning’s (2004) statements. This is a new finding of this thesis and just applied to the condition of China’s outward FDI temporarily.
In conclusion, consistent with the expectation of this thesis, there exist some changes in the influential factors of China’s outward FDI from 2003 to 2012. Market size and tax rate are decreasingly important and conversely labor cost and technological capability have increasing significances. Resource-based factors (market size, labor cost, and technological capability) keep affecting the FDI outflows. However, institution-based factors (tax rate) lose their effects over time. Maybe this is generated by the changing global institutional environment. Compared to the inconstant institution-based factors, preferential and stable resources are more favored by Chinese firms.
6.1 Implications
There are two main implications of this thesis, one in theory and the other one in practice. Firstly, this thesis fills the gap of research on studying the changing influential factors of a specific country’s FDI. Though there has been plenty of research on the influential factors of FDI, the changes of the influential factors over time lack enough attention (Lei & Chen, 2011). Influential factors and FDI activities are not stable, but keep varying. The previous findings of influential factors may not be applied to explaining the current FDI activities (Lei & Chen, 2011). So it is meaningful to make up this blank in the field of international business. This thesis makes a contribution to explaining FDI activities in different periods by identifying the changes in the influential factors. What is more, though some articles have mentioned the changes of influential factors, this thesis has some new findings, which are different from those of previous research, and provides some new insights. Thus, this thesis moves a step forward in this field.
Secondly, this thesis also provides a guideline for Chinese firms to do foreign investments. By learning about the changes in influential factors of outward FDI, Chinese firms can get to know about the changing political and economic environment of China and the whole world. Since financial investment has been the hot spot of FDI in recent years, they can search their investing opportunities in the financial industry in foreign market and make appropriate investing decisions by integrating their own competitive advantages. Moreover, the resource-based factors are demonstrated to be more stable than institution-based factor. Hence, Chinese firms should pay more attention to resource-based factors when they invest abroad, if they prefer steady and safe investments. Besides, Chinese firms should also pay
attention to the institutional environment of financial industry so as to guarantee their legitimacy of investment abroad. In addition, the findings of this thesis also enlighten Chinese government. To encourage domestic firms to go out, the government should keep noticing the changes in the influential factors of FDI and adjust their policies to the changes to provide domestic firms with more convenience. Also, the government can make use of the changes of influential factors to influence the composition of industries of outward FDI and further improve the domestic industry structure.
6.2 Limitations
There are still some limitations about the study of this thesis. Due to the limitation of length, this thesis does not include a complete set of influential factors of FDI. Besides, there may be other control variables that also influence the location choice of FDI. This thesis just covers a small part of them. Besides, some missing values are replaced with serious means. This can lead to artificial deflation of variation and further affect the estimated values. Thirdly, some regions and African countries attract large FDI flows from China. But they lack too many data of the independent variables and thus cannot be included in the study. This results in a limited sample of the study.
7. Conclusions
The number of articles focusing on identifying the influential factors of FDI is large enough. However, the global political and economic situations have experienced so many changes in recent years. These environmental (or institutional) changes can lead to changes in
