عوامل مؤثر بر سرمایه گذاری مستقیم خارجی در بخش خدمات

This study uses industry level foreign direct investment (FDI) data from 57 countries 1989–2000, to examine the host country determinants of FDI flows in services as a whole, and in the major service industries. Institutional quality and democracy appear more important for FDI in services than general investment risk or political stability. Democracy influences FDI to developing countries only, suggesting that the absence of democracy is detrimental to investment below a certain threshold. Consistent with the observation that many services are non-tradable, we find that service FDI is market-seeking, and unaffected by trade openness. We find a strong correlation between FDI in manufacturing and FDI in producers' services such as finance and transport.

A key feature of the evolution of foreign direct investment (FDI) flows in recent years, has been the expansion of FDI in the service industries. According to UNCTAD (2003), the service industries accounted for a massive 63% of world FDI flows in 2001. FDI stocks reflect the recent dominance of service sector flows, 55% of world FDI stocks in 1999 were in the service industries (UNCTAD, 2001). As argued by Harms and Ursprung (2002), changes in the aggregate characteristics of production can have substantial effects on determinants of total FDI flows. Given the growing importance of FDI in services, eliciting its determinants is vital in keeping host country FDI policies effective. Accordingly, this study uses industry level data to estimate the determinants of FDI in the service sector as a whole, and in the major service industries. As an illustration of recent trends, Fig. 1 provides a decomposition of total FDI flows, into shares attributable to the primary, secondary and tertiary industries over the period 1986–2001. While the secondary and tertiary industries accounted for comparable shares of total FDI in the late 1980s, the figure reveals that the service industries have increased their share markedly since then. The spike in tertiary FDI in 1999–2000 is perhaps a bit exaggerated, reflecting extensive mergers and acquisitions activity in this sector, but nevertheless the trend is clear. Service sector FDI has grown more important in total FDI flows, manufacturing FDI less so. The primary sector only contributed 3–11% of total FDI in the period graphed. Most empirical studies of FDI determinants have used data on FDI flows aggregated across industries (see e.g. Noorbakhsh et al., 2001 and Blonigen, 2005, for reviews, and Bengoa and Sanchez-Robles, 2003 for a more recent study). There is, however, reason to believe that the determinants of service sector FDI might differ from those of FDI in aggregate. Though international trade in services is on the rise, the fact remains that many services are non-tradable or costly to trade. And for a sector whose products to a large extent cannot be subjected to cross-border trade, the trade openness of a host country can be expected to have less of an impact on FDI inflows in that sector. Moreover, tapping the demand for services in a host country requires a physical presence when services are difficult to trade, which implies that FDI in services is likely to be market-seeking. The idea that domestic market size affects service sector FDI, whereas trade openness does not, is tested empirically in this study. The peculiarities of service sector FDI in terms of organization and outside options, may entail a different sensitivity to public policies and institutions. Previous studies of aggregate FDI indicate that there is a relationship between broad indices of socio-political instability and institutional quality, political freedom and democracy and FDI. The basic theoretical rationale of these studies is that socio-political instability, in the form of social unrest or government upheaval, increases perceived investment risks. Institutional quality, such as the quality and corruptibility of the bureaucracy, affects the costs and/or risks of doing business in the country in question. Political freedom and democracy might reflect increased stability, and property rights might be more secure in countries ruled by impersonal laws and institutions. It is, however, an open question whether the results from studies of aggregate FDI apply to the service sector, or if FDI in this sector is influenced differently by these types of variables. This study tests empirically the impact of different political economy variables on service sector FDI, in order to identify the relevant determinants. There are also likely to be differences in locational determinants across individual service industries. Fig. 2 maps the shares of total service sector FDI attributable to the four industries in which FDI flows are the heaviest; finance; business activities (including real estate, rentals, computer and related activities, and research and development); transport (which includes post and telecommunications); and trade (wholesale and retail). For the 1986–2001 period, these four industries have contributed 86% of total service sector FDI flows. The three largest of the service industries in terms of FDI (finance, business, and transport) are commonly referred to as producer services (Nordås, 2001). Given a global trend towards a vertical disintegration of production (Feenstra, 1998 and Hummels et al., 1998), producer services are essential in binding the chain of production together. It is thus plausible that there is a positive relationship between manufacturing FDI in a country, and FDI in producers' services, whereas there is less of a rationale for such a link to other service industries. Our data set includes both FDI in manufacturing and in the four largest service industries, allowing us to test this hypothesis.Several interesting results emerge from our empirical analysis. In terms of political economy variables, institutional quality and democracy appear more important for FDI in services than general investment risk or political stability. And interestingly, when splitting the sample in high- and low-income countries we find that different political economy variables impact different groups of countries. Political risk in general, and institutional quality in particular, is important to services FDI in high-income countries, while the level of democracy is important to services FDI in developing countries. The importance of democracy in developing countries found in our study supports the findings of Harms and Ursprung (2002) for FDI in aggregate. Since richer countries tend to score high on the democracy and institutional quality variables while poorer countries usually score low, our results may reflect non-linearities in the impact of the political economy variables. Hence, it may be the case that foreign investors are discouraged by highly undemocratic countries, but for countries above a certain threshold of democracy, investors may be more concerned with the efficiency of public sector institutions. Our results also support the argument by Blonigen and Wang (2004), that pooling high- and low-income countries in empirical FDI studies may be inappropriate. Our analysis largely confirms the hypothesis that FDI in services is market seeking, and unaffected by trade openness. Moreover, we uncover a strong correlation between FDI in manufacturing and FDI in producers' services such as finance and transport, which is consistent with the idea that these types of service industries follow their clients abroad, binding vertically disintegrated production chains together. The study is structured as follows. As the review of past empirical studies in Section 2 demonstrates, there are many studies of the determinants of aggregate FDI. However, econometric studies of FDI flows in services are sparse, with the exception of financial services. We therefore conduct an econometric analysis of determinants in the service sector as a whole and in the four major service industries. Section 3 presents the data set used, which has observations from 57 countries for the period 1989–2000, and discusses the explanatory variables included. The results of our analysis are presented in Section 4. Section 5 provides concluding remarks.

The service sector is dominant in world foreign direct investment flows. This study suggests that political economy variables may influence FDI flows in the service industries. Institutional quality and democracy appear more important for FDI in services than general investment risk or political stability. The institutional quality variable is robust to changes in specification for the service industries as a whole, and of particular importance to FDI in the transport sector, which might be attributable to the influence of the telecommunications industry. Our results show that different political variables affect countries at different level of development, democracy is important for FDI in developing countries and general political risk and institutional quality matter in industrialized countries. These nuances in host country determinants have important policy implications, and should be pursued in more detail in further studies. This study also shows that FDI in the service sector is market seeking, and unaffected by the trade openness of host countries. The significance of the domestic market, and the insignificance of trade, are evident from the analysis of aggregate service sector FDI, and by and large reaffirmed by the analyses of the four major service industries. There are, however, also important differences in the determinants of FDI flows in different service industries. Consistent with the idea that producer services bind together a globally disintegrated chain of production, FDI in manufacturing is a robust determinant of FDI in certain producer services (finance and transport) but insignificant for FDI in other types of service industries such as trade. The use of lagged variables suggests that the producer services respond with different lags to FDI in manufacturing, the finance industry being quicker to respond than the transport industry. Given the large and growing role of services FDI in total FDI flows, eliciting the host country determinants thereof paves the way for efficient and proactive host country FDI policies. In particular, several theoretical contributions have stressed the potential gains from trade in producer services, in terms of increased productivity (Markusen, 1989 and Hodge and Nordås, 2001). Where trade is unfeasible, FDI can have similar effects. Our results indicate that a producer services sector broad in variety depends on a foreign manufacturing presence, and a large domestic market, but also on aspects of the institutional environment in the host economy.