اینترنت خارجی و محل سرمایه گذاری مستقیم خارجی: مقایسه بین کشورهای توسعه یافته و در حال توسعه

This paper presents a two-stage game, which demonstrates that positive (negative) network externalities associated with Internet usage encourage (discourage) foreign direct investment (FDI). These hypotheses are tested by two major empirical methodologies—the panel data regressions and the system general method of moments estimator. The empirical findings provide strong evidence that the presence of negative network externalities in developing countries discourages inward FDI, and the presence of positive network externalities in developed countries attracts more FDI. In addition, positive network externalities are found to be more effective than negative network externalities in reducing the distance effect on FDI. As well, the evidence suggests that Internet development is likely to reduce the impact of the initial concentration of FDI, but is unlikely to reverse the self-reinforcing process of FDI.

The widespread use of the Internet in developing countries since the mid-1990s has created high expectations concerning the attraction of inflows of foreign direct investment (FDI). However, in recent years, even though the Internet traverses every single country in the world, disparities in the quality of its infrastructure in developing and developed countries have become apparent. As such, developing countries face a dilemma about increasing Internet usage, which gives rise to both positive and negative network externalities. In the present study, “positive network externalities” mean that new Internet users can add value to the value of all other users. For instance, as more users share communication network costs, the cost per user of the network decreases. Also, as more users are connected, a larger base for prospective customers and suppliers is created. On the other hand, “negative network externalities” mean that the growing number of users increase the strain on existing connections, causing Internet congestion. The purpose of the present study is to examine how the Internet—a communication network—which is characterized by the presence of positive and negative network externalities affects the locational choice of FDI1. It is worth noting that developing countries have experienced a severe Internet congestion problem and that the lack of bandwidth2 in these countries is one of its major causes. According to Sarrocco (2002, p. 23), “the availability of adequate and reliable bandwidth on international links, together with the quality of the local network, is one of the primary obstacles to universal connectivity3 of and within the less developed countries.”4 The report also points out that until a few years ago, few developing countries had more than 64 kbit/s, which means that an entire country had, on average, the same amount of bandwidth that a single user could access in Europe or the United States. The present study mainly builds on the studies by Harris, 1995, Choi, 2003 and Freund and Weinhold, 2004. In Harris (1995), the role of communication technology is to facilitate coordination within a firm, and between a firm and the suppliers of factor services. His general equilibrium model predicts that negative network externalities (i.e., Internet congestion) reduce the extent of market expansion. Choi (2003) finds a positive relationship between Internet development and FDI. Freund and Weinhold (2004) provide evidence of a positive relationship between Internet development and trade flows. This study distinguishes itself from these studies in three aspects. First, unlike Choi, and Freund and Weinhold, this study highlights the importance of positive and negative network externalities for determining the location of FDI. Second, this study extends Freund’s and Weinhold’s (2004) use of Cournot competition as a theoretical framework. This study applies a two-stage game in which firms not only compete with each other in terms of quantity but also compete to reduce coordination cost by using the Internet. Third, this study incorporates two determinants of FDI in relation to Internet development: distance barrier (or distance effect) and agglomeration effects. The term “distance barrier” means the distance between the host country and investing country, which acts as a barrier to FDI. That is, the greater the distance from the multinational corporations (MNCs) to the host countries, the higher the production costs—such as traveling costs, monitoring costs, and information costs—incurred by MNCs. Recent empirical studies, such as Brenton et al., 1999 and Buch et al., 2005, find that the geographical distance is negatively associated with inward FDI. This study models coordination cost as a function of the geographical distance and Internet expenditure, and empirically tests whether network externalities change the distance effect on FDI. Agglomeration effects emerge from the clustering of other firms and imply that the presence of past FDI attracts more new FDI. The World Investment Report (2001) acknowledges that although the Internet increases the mobility of MNCs, they tend to concentrate geographically because of agglomeration effects. This study provides empirical evidence on whether the Internet has reinforced or reduced agglomeration effects. To summarize, this study identifies the type of network externalities (positive or negative) and examines whether they reduce the distance effect and agglomeration effects on FDI in both developed and developing countries. The remainder of the present study is structured as follows. Section 2 presents a two-stage model. Section 3 discusses the data issues. Section 4 reports the results using different econometric analyses. Section 5 concludes with a summary of the results and the policy implications.

A two-stage model is developed to study the relationship among FDI, network externalities, and coordination cost (in terms of distance). If increasing Internet usage leads to positive network externalities, such as lower connectivity charges and the expansion of potential e-markets, MNCs can reduce coordination costs through their own Internet investment and thus increase foreign production. On the other hand, if the wide use of the Internet leads to negative network externalities, such as network congestion, MNCs cannot benefit form their own Internet investment and so potential foreign production is reduced. In summary, the present study’s model predicts that positive network externalities encourage FDI, while negative network externalities discourage FDI. Using the data on 106 developing countries and 30 developed countries collected for the period 1995–2002, two major empirical tests—the ordinary least-squares (OLS) regressions with year-fixed effects and the systems GMM estimator—are performed. This study extends Choi, 2003 and Freund and Weinhold, 2004 by stressing the importance of Internet infrastructure and network externalities. This study reveals the presence of negative network externalities in developing countries [columns (1) and (4) of Table 1 and columns (1) and (4) of Table 6] and positive network externalities in developed countries [columns (1) and (4) of Table 2, and columns (1) and (4) of Table 7]. The empirical evidence indicates that in the presence of positive network externalities, increasing Internet usage in developed countries considerably reduces the distance effect so as to attract more FDI [columns (2) and (5) of Table 2]. On the other hand, in the presence of negative network externalities, increasing Internet usage in developing countries does not statistically significantly reduce the distance effect on FDI [columns (2) and (5) of Table 6]. Moreover, it amplifies the negative impact of distance on FDI, as shown in columns (2) and (5) of Table 1. The findings of the present study about developing countries are complementary to Freund and Weinhold (2004):30 increased Internet penetration does not significantly alter the negative effect of distance on trade, although the present study adopts different theoretical and empirical approaches from Freund and Weinhold. In addition, the empirical findings of the present study are consistent with the study’s model prediction that positive network externalities are more likely than negative network externalities to reduce the distance effect on FDI. The present study also sheds light on the qualitative discussion of the agglomeration effects of FDI in the World Investment Report (2001). This report suggests that even with the Internet, the agglomeration effects are strong enough to encourage the concentration of FDI. As shown in columns (3) and (6) of Table 6, the results of the SYS-GMM estimator indicate that the Internet favors the self-reinforcing process of FDI (proxied by lagged FDI). However, the Internet tends to reduce the effect of the initial condition of FDI [proxied by (FDI/CAPITA)1994] on present FDI [columns (3) and (6) of Table 1 and Table 2]. All in all, to fully realize the benefits of the Internet, developing countries need not only to increase the popularity of Internet usage but also to improve their telecommunication infrastructure such as bandwidth per capita. Two benefits exist for using the Internet to attract FDI. First, the effects of geographical distance on FDI are no longer beyond the control of policy makers. In the presence of a high quality telecommunication infrastructure, the Internet can reduce the distance barrier to MNCs, creating new opportunities for poor countries that are located far away from the major foreign investors. Second, the results concerning agglomeration forces indicate that an opportunity exists for developing countries to attract FDI. Since the Internet is able to change the effect of the initial historical concentration of FDI, poor countries that improve their investment environment could bring in new FDI. Then, the self-reinforcing process is in force, and the comparative advantage starts to accumulate, thereby, attracting more FDI in the long term.