این مهم است که شما به کجا می روید: سرمایه گذاری مستقیم خارجی ظاهری و رشد اشتغال چند ملیتی در خانه

How does outward foreign direct investment (FDI) affect employment growth of the multinational corporations (MNCs) in the home country? Does the impact of outward investment differ by the level of development of the destination country of the FDI? Using a difference-in-difference approach, we assess the impact of starting to invest in less-advanced countries compared with investing in more-advanced countries. To obtain suitable control groups in each case, we use the propensity score method to select national firms that ex post did not take the investment decisions that we study even though ex ante they would have been equally likely to. We find that moving to less-advanced countries decreases a company's employment growth rate especially in the short run. On the other hand, moving to more-advanced countries does not consistently affect employment growth in any significant way. Including investment decisions of established multinationals in the estimation somewhat weakens but does not overturn this conclusion.

Multinationals have played an important role in the recent wave of globalization with its worldwide increase in exports and foreign direct investment (FDI). The public often views multinational activities with some skepticism, as it is concerned that off-shoring activities will reduce domestic employment in the firms that venture abroad. Such concerns are heard not only in the U.S. and Europe, but also in Asia. In this paper, we study the link between a multinational corporation's (MNC) employment growth rate at home and its decision to invest in either more- or less- advanced countries. With a unique dataset of South Korean firms that links the South Korean parent of an MNC with its affiliates abroad at the firm level, we can explicitly differentiate the impact of foreign direct investment by destination. Using matched sampling techniques to address self-selection and endogeneity, we compare the employment trajectories of multinationals with affiliates in either more- or less- advanced economies with the employment growth of firms that do not expand through foreign direct investment but that otherwise share all other forms of access to foreign markets. Since the mid-1980s, increasingly larger flows of foreign direct investment have found their way into China. China now tops the list of FDI recipients worldwide. China is also the predominant destination of FDI in East Asia, where the FDI flows into China and their effects on domestic production have become one of the premier policy concerns. The South Korean investment promotion agency KOTRA for example fears a “hollowing out of Korea's production base as a result of the rush into China” and suggestive data in Fig. 1 and Fig. 2 indeed show a falling share of employment in manufacturing in the 1990s as the share of trade with China as well as FDI into LDCs increase. As if to underscore the similarity with the debates surrounding NAFTA in the U.S., Ross Perot's notorious 1993 phrase “A giant sucking sound” has popped up again in the Asian context.1 The case of South Korea as an emerging economy is of particular interest. Most of the available studies of the impact of multinational activity on employment focus on advanced economies and established multinationals. Emerging economies, however, have a relatively young history of outward foreign direct investment. Before 1980, for example, only some 30 South Korean multinationals were active abroad, which is why assessing the impact of outward multinational activity on employment in emerging economies is to a large extent assessing the impact of first-time investments abroad, an aspect that has not received much attention in the literature so far.2 Moreover, as a middle-income country, South Korea's multinational activity is almost equally split between more- and less- advanced countries, which makes it ideal for comparing the impact of FDI into more- vs. less- advanced countries. Whether the particular destination country of FDI matters for the employment in the parent company is primarily an empirical question. The newer theories of multinational activity that focus on firm heterogeneity as well as the earlier literature that hinged upon the distinction between vertical and horizontal multinationals offer no conclusive answer. Horizontal multinational activity, for example, has been defined through market-seeking FDI especially to advanced economies. As Markusen (1984) and Brainard (1997) show, firms with moderate increasing returns should set up affiliates abroad to save transportation costs. Firms would relocate closer to the foreign consumer to produce the same goods that they produce at home. Going abroad would substitute for arm's-length exports and foreign labor would substitute for domestic labor. However, at the same time, moving to other markets could increase the local headquarter services that the multinational typically provides to affiliates and actually lead to more employment in the long term. The analysis of vertical FDI is similarly ambiguous. Vertical FDI is motivated by fragmentation of production, see Helpman (1984). Instead of producing the same product at different locations, firms would break up the value chain and relocate parts of their production off shore to take advantage of low labor cost in emerging economies. It is easy to see how this vertical strategy could lessen employment in the parent plants of the home country. However, nothing precludes this off-shoring strategy from being part of a long-term growth strategy. Here again, it is hard to judge a priori whether moving abroad would in the end decrease or increase employment at home. In recent years, the literature has moved beyond the distinction between horizontal and vertical FDI. Empirical work by Hanson et al. (2001) and theoretical work by Yeaple (2003) and Ekholm et al. (2007) have emphasized that the complexity of MNC integration strategies. There is likely to be both a horizontal and a vertical dimension to any multinational activity. In addition, the newer literature now explicitly includes firm heterogeneity, so that particular firm characteristics will also determine the strategy that a firm takes to enter foreign markets. As theoretical work by Grossman et al. (2006) illustrates, heterogeneity leads to a multiplicity of possible strategies that offer only limited guidance about the long-term employment effects of moving to more- or less-advanced countries. So far, the empirical evidence on the impact of multinational activities abroad is mixed. Brainard and Riker (1997) was the first study to suggest that there is no negative impact of off-shoring activities on domestic employment in the multinational. Also Desai et al. (2009) and Borga (2005) find that U.S. multinationals actually support job growth at home, which is consistent with Becker and Muendler (2008) who argue that FDI leads to less job losses when studying job separations for multinationals and non-multinationals in Germany. Brainard and Riker (2001) and Hanson et al. (2003), on the other hand, come to a different conclusion: they find that foreign employment may be a substitute for domestic employment. These mixed empirical results pose a challenge. We agree with Harrison and McMillan (2006) that they call for an empirical approach that differentiates the nature of the multinational operations at the firm level in order to assess the impact of MNC operations. Harrison and McMillan (2006), for example, differentiate the impact of multinational activity by location and by whether vertical or horizontal activities are involved. Our approach is consistent with this. Also Becker and Muendler (2009) allow for the impact of multinational activity to differ by location. To differentiate the impact of MNC operations by destination we apply difference-in-difference estimation plus propensity score matching, techniques that have been widely applied in labor economics and that are particularly well fit to study the impact of first-time investments. Among the first to apply these techniques to multinational operations were Castellani and Navaretti (2004) who studied Italian outward FDI and its effect on domestic employment and Egger and Pfaffermayr (2003) who compared the performance of multinationals and exporters.3Becker and Muendler (2008) is another recent application that focuses on job separations in Germany while comparing multinationals and national firms. We explicitly compare the employment trajectory of the parent of the MNC that goes to more- or less-advanced countries, the treatment group, with the performance of national firms, the control group. The control group that is matched with the MNCs is selected in such a way that the national firms ex ante would have been equally likely to invest abroad as the multinationals. Obviously, the quality of the results will depend on the quality of the matches between the treatment group and the control group and we go to great lengths to obtain a good match. Our results indicate that where a multinational invests matters especially in the short run for the employment growth of the multinational's parent at home. We consistently find that a move into a country that in terms of per capita GDP is less advanced than South Korea yields lower employment growth in the parent firm than in national firms that did not invest abroad. The longer the time horizon, however, the less significant that distinction becomes. On the other hand, we find in most cases no statistical difference between the employment growth rate of multinationals that open affiliates in more-advanced countries and national firms that do not. This finding is relatively robust. Our findings give some credibility to public concerns about offshoring at least in the short run. Our results indicate that it is important to know the destination of FDI in order to be able to assess its impact, which is consistent with Harrison and McMillan (2006) who distinguish by vertical and horizontal multinational activities and with Becker and Muendler (2009) who find different employment effects across European locations. Using data that do not differentiate by destination risks masking the particular impact of investing abroad. The rest of the article is structured as follows. First, we describe the estimation strategy that we follow. We then characterize the data and turn to the construction of counterfactuals. We finally discuss the estimation results before we conclude.

We have investigated the effect of outward FDI on home employment for South Korea, an emerging economy. For emerging economies, outward FDI is a recent phenomenon and many new multinationals have come online in recent years. The latter allows us to explicitly study the impact of these new multinationals on employment by comparing the performance of multinationals and non-multinationals, which is largely absent from the literature that has focused mostly on advanced economies with established multinationals. In addition, our particular South Korean dataset lets us directly link the South Korean parent with the particular destination country of its outward FDI. This helps us differentiate the performance of multinationals by whether they set up affiliates in countries that are more or less advanced than South Korea. In doing so, we take advantage of South Korea's position as a middle-income country that has divided its investment almost evenly across more- and less-advanced nations. To address issues of endogeneity and self-selection, we take a difference-in-difference approach with propensity score matching. We go through great lengths to guarantee the quality of the matches between investing and non-investing firms by conditioning on a long list of observables before the investment decision and by testing the quality of the match. To the extent that our matches indeed make the investment decision independent of our classification as a an investing or non-investing firm, our estimates suggest that firms that invest in less-advanced countries pay a short-term price in terms of employment growth. Including subsequent investments does not alter this conclusion. At the same time, our findings for firms that venture into more-advanced countries do not show a consistent tendency. In most instances there is no significant impact. Our findings thus support the public anxiety about multinationals only in the short term and only for investments into less-advanced countries. More importantly, our results suggest that any assessment of the impact of multinational activity on the parents' performance should differentiate by destination. If not, there can be a bias in the obtained results that depends on the destination composition of a country's FDI.