بررسی دوباره سوابق محل و روند سرمایه گذاری مستقیم خارجی: مدارک و شواهد از سرمایه گذاری های تایوانی در چین

The objectives of this study are to examine variations in the preferences of location antecedents and the pace of foreign direct investment (FDI) by multinational enterprises’ (MNEs) in relation to increased FDI experience and evolved strategic intentions. Using a rank-ordered logit model (ROLM), this study analyzes a total of 2688 investments in China from 731 Taiwanese-listed firms (TLFs) during 1997–2007. The results show that TLFs demonstrate distinct preferences for location antecedents at different stages of their FDI in China. In the early stages, TLFs attempt to exploit their existing competency and favor locations with advantages in infrastructure, labor cost, market size, and openness. In the later stages, locations with better production efficiency, labor quality, and R&D capability are more attractive because TLFs have become accustomed to the local business environment and intend to explore capabilities that create sustainable competitiveness in the face of competitive challenge. The results show that two antecedents, agglomeration and policy incentive, have consistent influence on location choice during the entire FDI process. The results also show that greater FDI experience contributes to a more intensive inter-investment time span and accelerates a TLF's investment pace.

Location choice is an imperative that multinational enterprises (MNEs) consider when expanding business outside home boundaries. In his eclectic paradigm, Dunning (1977) demonstrates that location is one of the three pillars that determines foreign direct investment (FDI). A number of studies investigate where MNEs decide to locate and articulate the influence that FDI location has on MNEs’ performance (Vanhonacker & Pan, 1997), their competitiveness in the global market (Dunning, 1998), and their intra-organizational configuration (Dunning, 1995). Despite these efforts, the location choice of MNEs deserves greater academic attention given the ongoing shift in FDI patterns in the world economy over the last decade (Chadee, Qiu, & Rose, 2003). The reasons why MNEs allocate business in a specific country or region are of great interest to researchers in the international business field. A review of the literature indicates that some specific characteristics of host destinations are attractive to MNEs in both cross-country settings (Billington, 1999) and within-country settings (Ma & Delios, 2007). Yet, most empirical studies treat each FDI decision by the same firm as an independent case (Benito & Gripsrud, 1992) and overlook the fact that MNEs can adjust their criteria for choosing a FDI location by basing it on their accrued FDI experience. With long-term experience in running businesses in international markets, MNEs are able to develop and accumulate experience that affects their subsequent strategy-making (Johanson & Vahlne, 1977). Given this, MNEs can show diverse location preferences during the FDI process (Dunning, 1998), and the location advantages emphasized in early FDI cases might differ from those in later ones. The effects of, and changes in, location antecedents over time are thus in need of greater research (Flores & Aguilera, 2007). In addition, the tenure of MNEs’ international operation can impact not only their location preference but their FDI pace as well. Johanson and Vahlne (1977) point out that MNEs’ knowledge that is relevant to a specific market derives from cumulative local experience and determines the speed of subsequent investments. Based on research on the internationalization of the Hong Kong garment industry, Lau (2003) asserts that the time span between investments for firms shortens over time. Also, Davidson (1980) mentions that MNEs increase entry frequency into the same country in light of accumulated location-specific experience. These observations imply that the continuous investments of MNEs enable them to develop a better understanding of their investment destinations and this learning effect can then facilitate their future location decisions. However, this contention lacks robust supporting evidence (Nadolska & Barkema, 2007). Taken as a whole, the objectives of this study are to examine how the learning effect influences MNEs’ dynamic preferences with regard to FDI-location antecedents and the time span between successive FDI cases. To this end, this study analyzes a sequence of FDI cases of Taiwanese manufacturing firms in China. According to the latest report of the United Nations Conference of Trade and Development (UNCTAD, 2010), China has become the second largest recipient of FDI after the U.S. and received investments valued at US$95 billion in 2009. Of the numerous countries investing in China, Taiwan, which officially opened investment in China in 1991, accounts for a large portion of FDI in China in spite of the political hostility between these two parties (China Statistical Yearbook, 1997–2007 editions). Notably, this paper delves into the FDI patterns of TLFs in China at the provincial level and explicitly focuses on the economic incentives, like market size and labor cost, each province offers (Buckley et al., 2007 and Coviello and McAuley, 1999). Though China is a unitary nation with a uniform legal system, the institutions that contribute to a well-operating market economy (e.g., property rights protection and contract enforcement) can vary across provinces and influence a MNE's location choice (Du, Lu, & Tao, 2008). Given that Taiwan and China share the same language and similar cultures, very little cultural distance exists between them (Zhang, 2005). These close cultural ties enable TLFs to quickly adapt to the distinct local business environment by connecting to the local authorities and to minimize the influence of institutional variation. Hence, this study excludes location antecedents such as culture, language, and regulations (O'Grady and Lane, 1996, Schroath et al., 1993 and Sun et al., 2002). The structure of this paper is as follows. Section two begins with a review of literature on FDI location choice and associates the MNEs’ FDI motives with their preferred location advantages. The remainder of this section introduces nine major location antecedents that MNEs consider for FDI decisions in China. Section three delineates the potential effect of learning on the motive-preference relationship and on the FDI expansion pace. The ROLM and measurements of the focal variables adopted are in section four. The final two sections present the empirical results and discussion.

The majority of the proposed hypotheses have marked statistical support; and the results generally corroborate the argument that TLFs change their preferences for location antecedents in the FDI trajectory due to the organizational learning effect and altered strategic intention, which is consistent with Chang and Rosenzweig's (2001) findings on entry mode. As the knowledge related to China grows, the strategic objectives of TLFs evolve from exploitation to exploration and then generate different priorities for location antecedents in different FDI stages. As expected, the intention of TLFs to exploit their existing competency in their early investments causes them to emphasize antecedents such as infrastructure, labor cost, market size, and openness. By contrast, R&D capability and labor quality become more important when TLFs attempt to explore their future competitiveness in their later investments. This result also shows that the more FDI experience TLFs accumulate, the faster their investment pace is. Three location antecedents fail to meet our expectations. The significantly negative effect of production efficiency on the sequential location choice indicates that TLFs place more emphasis on production efficiency in the later FDI stages rather than the early stages. One reason might be that TLFs take production efficiency into account seriously if market competition intensifies (Bunyaratavej, Hahn, & Doh, 2008). In initial FDI stages in China, TLFs exploit the manufacturing capabilities they developed at home and derive competitive cost advantages from the abundant labor force. When more firms view China as a source of future growth and attempt to gain cost advantage in a similar way, then improving production efficiency is a promising means to confront this new challenge. Another possible explanation might be that TLFs show higher involvement as the initial liability of being foreign reduces incrementally. With this reduction, they are then willing to renew advanced production facilities that are usually capital-intensive and productive. Production is thereby more efficient as labor productivity increases. In this situation, sophisticated machines require skilled labor to be more effective, and this requirement corresponds to labor quality and is more crucial in later FDI stages. In addition, the positive but nonsignificant effects of agglomeration and policy incentive indicate that TLFs emphasize these two antecedents in both their exploitation and exploration stages. TLFs source knowledge not only from internal R&D activities but also from external M&A, alliances, and informal intrafirm interaction. The geographic proximity of agglomeration might contribute to TLFs’ knowledge exploration through “common buyers and suppliers, chance meetings of different firms’ scientists and engineers, and employees switching jobs (Alcácer & Chung, 2007, p. 760).” Policy incentives, such as preferential taxation for R&D activities offered by economic development zones, are a major attraction for TLFs when choosing locations as well (Zanatta & Queiroz, 2007). These incentives encourage TLFs to explore technology and procedural enhancements without worrying that R&D expenses will undermine their profits, particularly when the outcomes from R&D activities remain uncertain. 6.1. Theoretical implications The results of this study contain two theoretical implications. First, this study is consistent with the view of Hutzschenreuter, Pedersen, and Volberda (2007) that internationalization is a function of MNEs’ experience/knowledge, managerial intentionality, and other factors such as institutional forces. They contend that the three major theoretical approaches, internationalization theory, eclectic paradigm, and knowledge-based view basically underestimate the effect of strategic intent and entrepreneurship in internationalization. Undoubtedly, MNEs accumulate experience in the course of internationalization. These experiences, especially successful cases, become templates that lead to MNEs’ inertia in decision-making. To sustain their success, MNEs tend to follow a similar investment manner: namely, path dependence (Schreyögg & Sydow, 2010). However, from a long-term perspective, this dependence can restrict MNE decision options and hinder their potential growth. To cope with the pressure of a competitive business environment, MNEs have to evaluate their knowledge status and augment their knowledge base to prevent knowledge obsolescence (March, 2003). At this point, the role of MNE's strategic intent is influential in internationalization (Volberda, Baden-Fuller, & van den Bosch, 2001), and MNEs appear to develop an internationalization trajectory that involves path creation or exploration (Hutzschenreuter et al., 2007). The main contribution of this study is to provide evidence that MNE intentions might change from exploitation (path dependence) to exploration (path creation) within a country by examining the TLF location choice in China (Luo, 2007). The sampled years in this study also shed light on the emerging position of China in the world economy during this period. Second, a growing number of studies illustrate that exploitation and exploration are a duality of fundamental learning activities for organizations (Raisch, Birkinshaw, Probst, & Tushman, 2009). Though these two approaches implicitly conflict in strategic logic and internal resource allocation, researchers believe that firms that reconcile exploitation and exploration simultaneously earn better performance and sustainable competitive capabilities (He & Wong, 2004). Yet, this balanced view often raises a fierce debate. This study, set in a downstream FDI context, offers evidence that exploitation and exploration might not be simultaneous but instead are sequential strategic intents for MNEs in the investment process. Therefore, TLFs emphasize exploitation earlier than exploration in their FDIs in China; this paper argues that MNEs tend to temporally leverage their exploitation and exploration activities in the international context. 6.2. Managerial implications This study identifies several managerial and policy implications for the host authority. For managers, while most researchers suggest the necessity for the reconciliation of exploitation and exploration, variations in the allocation of scarce resources among these two activities over time might be a feasible solution to this paradox. For policy makers who are struggling to attract internal investments, the first thing they should do is to understand the position of their home within the IDP framework as well as the position of MNEs in their FDI trajectory. Then, they will be able to offer suitable location antecedents for MNE investment. Second, as soon as MNEs begin to invest, making MNEs more willing to expand their investment is the next step. Considering that the strategic intent of MNEs toward a specific destination might change over time, policy makers need to regularly check and even adjust their location advantages. Third, based on the results of this paper, agglomeration and policy incentive always influence FDI and are worth greater attention from local authorities. 6.3. Limitations and the avenues for future research As MNEs continue to broaden their business boundaries in the global market, the location choice and the pace of their FDIs are two issues of interest to IB researchers. This study addresses these two questions and provides empirical support by arguing that the inclination of TLFs is to change their location preference and to hasten their investment pace over time. However, researchers should generalize these findings carefully. A major limitation is that this study examines TLFs’ sequential location choice and verifies the change of location preference with an exploitation-exploration path based on the assumption of motive-preference inference. For robustness, future studies should longitudinally measure the MNE's strategic intent and validate the exploitation-exploration path by directly linking motive with location choice (Brion et al., 2010 and Makino et al., 2002). The second limitation is that this study empirically examines hypotheses in a downstream FDI context using the case of TLFs in China. Whether the exploitation-exploration path remains valid in other downstream FDI flows where the home and host economies share different languages requires further investigation. Similarly, studies of upstream FDI flows and in cross-country contexts are essential as well (Galan et al., 2007). Further, due to data availability, the analyzed FDI data are all from the firms listed in the Taiwan Stock Exchange. More research should focus on FDIs of small and medium enterprises, which might provide fruitful evidence on this issue (Suh, Yi, & Houston, 2011).