آیا حد بالایی در سرمایه گذاری مستقیم خارجی مهم است ؟: مورد از تایوان

In this paper, we present a partial least squares (PLS) path model developed to investigate foreign direct investment (FDI) by Taiwan in China. The main purpose of the study is to answer the question, “Has the Taiwanese government's upper limit on investments interfered with Taiwanese firms’ decisions about whether to undertake FDI in China?” The question was answered by testing six hypotheses derived from the model. Using data on Taiwanese manufacturing firms from the Integrated Circuit industry for the years 1998–2007, we found no significant evidence supporting the effectiveness of the upper limit. The most influential of the model's five determinants of Taiwan's FDI in China are factors specific to individual firms. The second most influential is the macroeconomic environment of the host country. Previous studies have paid little attention to the parent country when analyzing FDI, a deficiency we remedied in the present study. Our study reflects an integrated perspective on the FDI literature by including the host country, the parent country, and firm-specific factors as determinants of FDI.

The deregulation of capital outflow by the Taiwanese government in 1987 created a watershed in the pattern and amount of foreign direct investment (FDI) by Taiwan. The policy permitted a business or an individual to annually send up to 5 million US dollars abroad without governmental approval. As a result, FDI surged. Between 1987 and 1988, both the number of Taiwanese FDIs and their total value surpassed those of the FDI coming into Taiwan. (In the remainder of this paper, unless specified otherwise, FDI refers to outgoing FDI, that is, FDI flowing from Taiwan to another country.) Taiwan has since become a net capital exporter. However, after the Asian Financial Crisis of 1997, Taiwan's FDI destinations changed dramatically. This is particularly evident in the rebalancing of FDI funds between crisis-affected Southeast Asian countries on the one hand and China on the other. Not only has the FDI flow from Taiwan to China increased over the last 10 years, but since 1997 so has the ratio of Taiwan's FDI in China to its FDI in Southeast Asia. The growth of FDI in China since the beginning of China's economic reforms in 1978 has been striking. Since 2002, China has become the largest recipient of foreign capital in the world. After Taiwan and China started to exchange visits across the Taiwan Straits in the 1980s, direct investment by Taiwanese businessmen in China began to rise rapidly. Even though Taiwan and China share a very similar cultural background, and despite their different economic and political systems, China has a distinct advantage over Taiwan in attracting FDI; this advantage cannot be attributed solely to economic factors. It is well recognized that the electronics industry is a key driver of Taiwan's economic growth. Since 1983, this industry has transformed itself from an original equipment manufacturer (OEM) to an original design manufacturer (ODM). With the emergence of China as a more attractive low-cost production and exporting platform, many companies have established production sites in China as a way to become more involved in global logistics management (GLM). According to Taiwan's Mainland Affairs Council, the cumulative number of Taiwanese FDIs in China, which began in the 1980s, reached 36,459 by 2007. The aggregate value of this FDI was 63.3 billion US dollars. In fact, China has now become the primary destination for Taiwanese enterprise funds. The FDI for manufacturing from Taiwan's various industries is distributed as follows: 15.8% for electrical equipment; 15.4% for computers, electronics, and optical products; and 6.8% for basic metals. These data show two things: (1) the ties between manufacturing in Taiwan and China are strong; and (2) Taiwan's electrical equipment manufacturing industry is the primary contributor to this capital outflow. In recent years, issues related to FDI have attracted much attention from scholars in international business and economics. It has been well documented that FDI provides various benefits to the host county. These include productivity gains, technology transfers, and economic growth (Ang, 2008, Baltagi et al., 2007, Chowdhury and Mavrotas, 2006 and Gholami et al., 2006). Numerous studies have found that the identity of the host country is the key factor driving FDI (Ang, 2008, Cheng and Kwan, 2000, Eichengreen and Tong, 2007, Hooper and Kim, 2007, Jinjarak, 2007, García-Herrero and Santabárbara, 2007, Giner and Giner, 2004, Mina, 2007, Xu et al., 2008 and Zhang, 2005). However, the role of the parent country's government as another determinant has been largely ignored in these studies. It is commonly noted that governments impose various types of regulation on FDI. This has been particularly true for Taiwan with respect to China. In response to the political tensions between China and Taiwan prior to 2008, Taiwan's government took steps to limit capital and technology outflow and to protect the country's employees in manufacturing and related areas. Specifically, it set an upper limit on FDI, while at the same time listing products made in China that Taiwanese firms were allowed to invest in. The upper limit for investment by any Taiwanese firm in China was defined as 40% of the investing firm's available capital or net value, whichever was lower. Both business leaders and academia continually complained about these regulations. On the other hand, unskilled workers and some political factions supported the government's FDI policies. These disputes remain unresolved. In any event, no solid empirical evidence has been offered thus far concerning whether the government's upper limit has had any effect on Taiwanese FDI in China. If there is an effect, how large is it? If there is not an effect, why not? Previous studies of FDI by Taiwanese industries have addressed issues such as performance evaluation, technology forecasting, and location selection (Chen and Ku, 2000, Lee et al., 2007 and Li and Hu, 2002). Deng (2007) examined the motivation underlying China's FDI from an asset-seeking perspective. Demirbag, Tatoglub, and Glaister (2007) adopted an integrated perspective incorporating both the host country and firm levels to examine the factors that influence perceptions of FDI success. Hsiao and Hsiao (2004) designated regional distribution, geographic proximity, and cultural similarity as important reasons why Taiwanese industry considers China to be such a good investment opportunity. Zhang (2005) considered the primary determinants of direct investment in China by Hong Kong and Taiwan (HKT) to be their export-promotion strategy. Compared to European Union, the US, and Japan their advantages in terms of export-oriented FDI, their unique linkage with China, and China's cheap labor. Ng and Tuan (2006) studied the geographical concentration of firms in China, especially the impact of this concentration on China's economic growth and how the decision of where to locate is related to institutional factors, such as government preferential or regional FDI-led policy. Xu et al. (2008) argued that the FDI chaos in China might be governed by the intervention of the Chinese government (host country policy). Unfortunately, none of these studies addressed FDI from the parent country's perspective. Although García-Herrero and Santabárbara (2007) incorporated capital flows, the home country, the host country, and global factors into their FDI model, they considered the impact on FDI primarily from the viewpoint of the host country. In this paper, we introduce a partial least squares (PLS) path model for FDI. This integrated model includes factors related to the host country, the parent country, and the individual firm as determinants. Hypotheses regarding the effects of the Taiwanese government's FDI policy on firms’ investment decisions are then developed and tested. Using industry data from the Taiwanese Integrated Circuit for the years 1998–2007, we found no evidence that the Taiwanese government's upper limit on FDI affected integrated circuit (IC) firms’ decisions for FDI in China. During the period sampled, firm-specific determinant was found to have the greatest effect on FDI. The remainder of the paper is organized as follows: Section 2 covers hypothesis development; Section 3 describes the methods and data used in applying the PLS path model; Section 4 presents the results of testing the model; Section 5 presents a discussion of these results.

In this study, we found that the Taiwanese government's policy of an upper limit on FDI, the macroeconomic environments of the parent and host countries, comparative advantage, and firm-specific factors have had a strong influence on Taiwan's FDI. The R2 for our proposed model is 0.921, lending it strong support. Structural relationships from the proposed model are shown in Table 9. Most previous studies have focused only on the host country when analyzing FDI, while ignoring the parent country. The results of our study show that firm-specific factors have a much stronger impact on FDI than the upper limit (0.781 vs. 0.273). Further, we found that the macroeconomic environment of the host country (0.431), the macroeconomic environment of the parent country (0.235) and the comparative advantage (0.227) have also affected FDI. Thus, our study contributes to the literature by demonstrating that firm-specific determinants, the parent country determinants and the comparative advantage determinants also matter in outward FDI. Given that the standardized path coefficients indicate the strengths of the direct effects, it is worth noting that the firm-specific factors (0.781) have the greatest effect on FDI. As shown in Fig. 2 and Table 8, the weights for RD-R, Royalty and K are 0.833, 0.620, and 0.955, respectively. These values indicate that IC manufacturing firms exist in a highly capital-intensive, highly technology-intensive industrial environment, in which the availability of working capital, the expenses for royalties and revenue associated with R&D, or technology transfers are inevitable. The weight for Export Order is 0.955, indicating that the higher ratio of business orders (in Taiwan) compared to overseas production (in China) implies a higher incentive in Taiwan for FDI; this conclusion is based on the factors of geographical clustering of industries, low costs of operation, and short times to market. Due to technology protection and ownership control, Taiwanese IC manufacturing firms tend to engage in FDI instead of outsourcing. However, the weight of only −0.436 for ASG suggests that as the rate of total asset growth in Taiwan decreased, IC firms faced a bottleneck in their attempts to increase their capacity. This situation accelerated their capital outflow to China. The weight for the PB Ratio is notable. A positive PB Ratio would mean that firms’ actions are consistent with market expectations. However, the negative PB Ratio of −0.386 indicates that firms’ actions ran counter to investors’ expectations; although the firms may have wanted to engage in FDI, they did not do so because of the upper limit. The path coefficient for the host country (0.431) ranks second. According to Fig. 2 and Table 8, the weights for C-CPI, C-GDP, C-Save, C-Trade, and C-Invest are 0.858, 0.917, 0.979, 0.990, and 0.978, respectively. These values indicate that FDI is positively related to the favorability of the host country's macroeconomic environment. Since 2002, China has become the largest recipient of foreign capital in the world, showing that China has become a magnetic force attracting FDI from abroad. The path coefficient for the parent country (0.235) ranks third. According to Fig. 2 and Table 8, the weights for NC, Re-GNP and Trade are 0.929, 0.973, and 0.951, respectively. It is well known that the IC industry was the foundation of Taiwan's economic miracle. Consistent with the theory of the international product life cycle, following a period of growth there was a demand for IC firms to invest and further enlarge their capacity and product diversification. The weight for Depend (0.951) suggests that the greater Taiwan's export dependence on China and Hong Kong, the greater the incentive for Taiwanese FDI in China. Again, investment and production in China has provided the incentive for IC firms. This incentive is based on geographical clustering, low cost of operation, and short time to market. The weight of G-Invest is −0.572, suggesting that the Taiwanese government's investment in the IC industry is insufficient. As the preferential treatment provided by the parent government declines and environmental impact protection gradually increases, IC firms are pressured to increase capital outflow to China. The path coefficient (0.227) for the comparative advantage determinants ranks fourth. This result suggests that the electronics industry exists in an era of microprofits, a time when cost reduction is an IC manufacturing firm's primary goal. As shown in Fig. 2 and Table 8, the weights for Wage Ratio, VAT Ratio, Rate Ratio, LVIT Ratio, and Exchange Ratio are 0.987, 0.858, 0.774, 0.974, and 0.887, respectively. Because IC manufacturing firms exist in a highly labor-intensive and cost-sensitive environment, as the endowment costs (including relative wages, relative increments in the tax on land values, the relative lending rates of banks, and relative value-added taxes) increase in Taiwan, the incentive for Taiwanese FDI in China increases. The path coefficient for the parent country's upper limit is 0.273. As shown in Fig. 2 and Table 8, the weights for A and M are 0.983 and 0.981, respectively. Given that the significance level for the upper limit fails to reach 0.05, H1 (the amount of FDI is negatively related to the upper investment limit of the parent country) is rejected. This result seems consistent with the free-market economic philosophy; that is, other firm-related factors affecting FDI deter the effect of the upper limit. Because of the trends toward globalization and liberalization, the government's policy is shown to be insufficient. Because the available capital or net value of most IC manufacturing firms is generally high, an investment of 40% of this available capital may suffice for Taiwan to maintain the technological status quo in China. The rejection of H1 suggests that the upper limit is not a barrier to investment. Likewise, technology protection and control have been delayed by the Taiwanese (parent) government, as illustrated by the case of the Taiwan Semiconductor Manufacturing Corporation's desire to build a 12″ wafer fabrication plant in China. In a similar vein, the path coefficient for the host country's macroeconomic environment is 0.949, which suggests that China's macroeconomic environment indeed affects Taiwan's. Further, this path coefficient for the host country is higher than that for the parent country (0.431 vs. 0.235), possibly due to a greater comparative advantage determinants for production in China. This result suggests that the relationship between Taiwan and China is very close, and the pressure for Taiwan to provide FDI comes more from China than from Taiwan. Above all, these results highlight the importance of considering not only the influence of the host country but also that of the parent country. They also have important policy implications for the Taiwanese government's investment authority, because they show that the upper limit on investment is not a key factor in determining whether Taiwanese IC manufacturing firms decide to invest abroad. The government's policy of keeping Taiwan's capital and technology in Taiwan, or encouraging Taiwanese companies located in China to return or reinvest their capital in Taiwan, can be achieved more effectively by improving the macro- and microeconomic circumstances that most firms face at home. Recognizing that the competitive nature and unique characteristics of many other industries in Taiwan is quite different from that of the IC industry, we stress that the implications of our results cannot, and should not, be extended to the whole spectrum of Taiwanese FDI. Moreover, the results of our study do not undercut the objectives of the regulations: after all, a parent country's laissez-faire policy on FDI is effective only if the relations between the parent and host countries are friendly. However, it is well known that there was severe tension between China and Taiwan when the upper limit was established. Fortunately, relations between China and Taiwan have improved significantly since the inauguration of Taiwan's new president, Mr. Ma Ying-Jeou, in May 2008. The results of this study contribute to the literature from an integrated perspective, as they address the influence of the host country, the parent country, and firm-specific factors. In contrast, previous studies have paid little attention to the parent country when analyzing FDI. We encourage further research applying our FDI PLS model to other Taiwanese industries, so as to better understand the effects of the government's FDI policy on them, thereby allowing the development of propositions that are generalizable.