Change Font: A A A A Contact Us What's New FAQs Subscribe ADB.org home
HomePublicationsForeign Direct Investment in South Asia: Policy, Trends, Impact and DeterminantsImpact and Determinants of FDI

Impact and Determinants of FDI

As discussed in Sections III and IV of the study, South Asian countries in recent years have been designing policies to attract increased FDI. Though they have received less FDI than other developing countries, there has been renewed interest in these countries since 2000 Against this backdrop, this section explores the impact of FDI on growth, investment and exports.

FDI and Economic Growth

Economic growth in any country depends upon the sustained growth of productive capacity, supported by savings and investment. Low levels of savings and investment, particularly in developing countries and least developed countries, results in a low level of capital stock and economic growth. The earlier growth models by Harrod (1939) and Domar (1946) explain that capital formation raises the standard of living, which in turn results in higher growth. Criticising the growth models proposed by Harrod and Domar on the ground of the fixed proportion of factors of production and substitubility between labor and capital, Solow (1956) argues that capital formation increases labor productivity in a dynamic process of investment growth. Some of the recent growth theories such as Lucas (1988) and Rebelo (1991) broaden the definition of capital to include human capital and the accumulation of knowledge. Similarly, Romer (1986; 1990) and Helpman and Grossman (1991) incorporate knowledge capital gained through research and development to explain growth along with other variables. Overall theoretical growth literature demonstrates the role of capital or changes in definition in capital (knowledge capital or human capital) in enhancing economic growth.

The recognition of the role of knowledge capital in economic growth creates a basis for analysing the role of FDI, which brings new technology and knowledge along with capital. In recent years, the need for FDI inflows has increased as MNCs have assumed significant importance as a source of economic growth and development (Bajpai and Sachs, 2000). Since FDI may help developing or lower income countries in South Asia by providing new knowledge and complementing domestic investment, it is important to analyse the empirical relationship between FDI and economic growth in a growth accounting framework. The FDI-Growth nexus has been mainly examined through the following ways: looking at the determinants of growth, exploring the determinants of FDI and the role of multinational firms in host countries, etc. There are a large number of macro and micro studies examining the relationship between FDI and economic growth. However the results of both country specific studies and cross-sectional studies fail to clarify the relationship between FDI and growth.

Brief review of the literature: Earlier studies examining the relationship between FDI and growth postulated a negative association for developing countries (Singer, 1950; Griffin, 1970). The logic of these studies was that FDI was concentrated on low-priced primary exports to developed countries, and had a negative impact on overall growth. However studies by Rodan (1961) and Chenery and Strout (1966) showed that FDI had a favourable impact on productivity and growth in developing countries. Further, Barro and Sala-i-Martin, (1999) and Helpman and Grossman (1991) argue that FDI has long term positive impact by generating increasing returns through technology and knowledge transfers.

Investment policy reviews by UNCTAD provide evidence of benefits of FDI in terms of employment generation, wages, and linkages with local firms, increases in technologyintensive exports, range of new products and services, etc. Overall, UNCTAD investment reviews suggest that FDI has a positive impact on growth but that it varies from country to country (UNCTAD, 2003). By and large, previous literature suggests that FDI contributes to growth through capital formation and technology transfer (Blomstrom et al. 1996 and Borensztein et al. 1995) along with accumulation of knowledge due to labor training and skill acquisition (Mello, 1999). Therefore, the most frequently cited common benefits of FDI are productivity spillovers for the host economy, resulting in higher growth. The logic is that FDI provides a stock of knowledge capital to less developed or developing economies and make factors of production, namely labor and capital, more productive. Thus, most of the previous studies show a positive impact of FDI on the host country economy (Mello, 1999; Bende- Nebende et al. 2000; Durham, 2004; Nair-Reichert and Weinhold, 2001; Xu, 2000). However, the impact varies from country to country {UNTAD, 1999; 2003; Borensztein et al., 1998; Bende-Nabende et al. 2001}. Further, a positive impact effect of FDI on improving growth and per capita growth is found in studies such as Caves, 1974; Lipsey, 1999; Globerman, 1979 and Blomstrom and Persson, 1983.

At the macro level, by and large, the previous literature finds a positive impact of FDI, but the impact varies from country to country and depending on country conditions. Blomstrom et al. (1994) find that FDI has a positive impact on growth in rich countries. Further, Borensztein et al. (1998) argue that FDI inflows are positively related to per capita GDP growth provided the host country has a highly educated workforce. Alfaro et al. (2000) find that FDI positively affects growth in sufficiently developed markets. Similarly, Balsubramanyam et al. (1996) emphasize trade reforms to create a positive impact of FDI on growth. Based on a disaggregate analysis, Wang (2002) finds that FDI in manufacturing has a significant positive impact on growth. Bende-Nebende and Ford (1998) find that the output of less developed countries responds more positively to FDI. Borensztein et al. (1995) explain that because of the transfer of technology, FDI contributes more to growth than domestic investment. Bashir (1999) demonstrates that FDI improves growth in MENA countries, though the effect varies from country to country. Chowdhury and Mavrotas (2003) find unidirectional causality running from growth to FDI in the case of Chile but find bidirectional causality for Thailand and Malaysia.

Further, FDI boosts the demand for intermediate goods from domestic firms leading to more entry of new firms, an increase in competition, industrial growth and an increase in national welfare (Markusen and Venables, 1999; Haaland and Wooton, 1999). However, in theory, externalities associated with FDI may raise or reduce national welfare. This depends on whether the positive spillover created by FDI is more than the negative externalities (such as the crowding out domestic investment by reducing their profit margins). If the impact of multinationals on the profitability of domestic firms is sufficiently negative, FDI may lower hostcountry welfare. In some conditions, where the multinational demand for labor is weaker than that of existing domestic firms, it may also lower the national welfare. Moreover, the repatriation of profit may drain capital from the host country. Thus, the impact of FDI on national welfare and economic growth can be negative. Carkovic and Levine (2002) find that FDI inflows do not have an independent influence on economic growth. Similarly Ericsson and Irandoust (2001) fail to find any relationship between FDI and growth for Denmark and Finland but find causality from FDI to GDP growth for Norway. Germidis (1977), Haddad and Aitken (1993) and Mansfield and Romeo (1980) find that FDI does not accelerate growth. Further micro level studies by Aitken, Hanson and Harrison (1997), Mello (1997), and Harrison (1996) also fail to lend support for the hypothesis that FDI accelerates overall economic growth.

The potential benefits of FDI are realized only if the local firms have the ability to absorb the foreign technologies and skills (Blomstrom and Kokko, 2003). In fact, it has been empirically proven that FDI is an important tool for development in host countries which have well-developed infrastructure and stable economic conditions (Balasubramanyam, 1998; Blomstrom et al, 1994). On the other hand, big multinational enterprises may drive out local firms because of their financial power and their technological and management superiority. Empirical evidence on the nature and extent of spillovers from FDI to domestic firms is mixed. The spillover effect depends on the technology gap between foreign and domestic firms. In the Indian context, earlier studies show that FDI has no such positive impact on growth (Dua and Rashid (1998); Chakrabarthy and Basu, 2003). Mello (1997) and Kokko (1996) find a negative relationship between FDI and total factor productivity. However, Sahoo and Maathai (2003) find a positive association between FDI and growth. There are studies finding a positive relationship between productivity growth, liberalization and foreign firms (Basant and Fikkert, 1996; Srivastava, 1991; Kathuria, 1998; 2000).

Overall, the impact of FDI on growth is far from clear and the impact varies across countries under different economic conditions. Since all these South Asian countries have labor surpluses, FDI can augment growth by providing additional employment. However, these countries are relatively closed economies with a low level of education and infrastructure facilities. Therefore, it is difficult to make inferences about the possible impact of FDI on growth without a proper empirical examination. The impact of FDI on economic growth has been estimated in a growth accounting framework as follows

Y = f (K, L,),

where Y is gross output produced in an economy using two important inputs such as capital (K) and labor (L). However, total capital consists of domestic capital (Kd) and foreign capital financed by foreign investment (Kf). Thus domestic capital and foreign capital have been taken separately. To determine the independent impact of FDI, FDI has been deducted from gross domestic capital formation (GDCF), which has been proxied for physical capital. 29 Looking at previous growth literature and empirical studies, the growth function has been augmented with human capital (H), Exports (Ex), infrastructure (INF).

Y = f (Kd, Kf, L, Ex, LIT, INF)

The proposed growth equation for the estimation is given below:

LGDPt,it = a0 + a0FDIYit + a2LGDCFit + a3LFGit + a4LEXPit + a5INFINDEXit + a6LIT + a7TRADEY + ut

where LGDP is the log of real gross domestic product, FDIY is foreign direct investment as a percentage of GDP; LGDCF is the log of gross domestic capital formation. LFG is labor force growth, LEXP is log of real exports, LIT is the literacy ratio, and TRADEY is total trade (export and import) as a percentage of GDP, proxied for openness. INFINDEX is the infrastructure index, which is constructed using different infrastructure indicators (see determinants section, V.4). The period of the study is 1970 to 2003. However, whenever infrastructure indicator is included, the period of the study is 1975 to 2003.

Data Sources: Annual data on gross domestic product, gross domestic capital formation, total exports, total trade, literacy ratio, and labor force are taken from World Development Indicators CD-ROM, World Bank, 2005. Since continuous data for Nepal on different variables such as FDI is not available, for example, the estimation is done only for four countries, India, Pakistan, Bangladesh and Sri Lanka.

Methodology: Panel data analysis has been employed to examine the relationship among variables. The reason for using this technique over time series and cross section techniques is mainly due to the higher power of the test as it combines both the cross section and time series unit. Secondly, the test takes into account the heterogeneity of variables across the industries. The general panel regression equation can be written as

For i = 1..........N cross section units, and period t = 1..........T.

A number of panel regression equations have been estimated with all relevant potential determinants of growth. Since few explanatory variables are correlated, estimations with different specifications are carried out. Many growth functions have been estimated using panel ordinary least square, panel fixed effect model and also random effect models. However, the fixed effect results are reported here because of their robust output and because the Huasman test supports fixed effect.

Since the impact of FDI varies from country to country under different country conditions, even within south Asian countries, a causality analysis has also been done to show the relationship between FDI and GDP for each of these four South Asian countries.

Determining the direction of causality – The Granger causality test: We perform a vector autoregression (VAR) procedure (for I (0) variables). Following Granger (1969), an economic time series Yt is said to be "Granger-caused" by another series Xt if the information in the past and present values of Xt helps to improve the forecasts of the Yt variable, i.e. if, MSE(Yt | Ωt) < MSE(Yt | Ωt') , where MSE is the conditional mean square error of the forecast of Yt, Ωt denotes the set of all (relevant) information up to time t, whilst Ωt' excludes the information in the past and present Xt. The conventional Granger causality test involves specifying a bivariate of pth order VAR as follows:

where υ and υ1 are constant drifts, Ut and Ut' are error terms.

Results: The estimations results show the contribution to the growth of South Asian countries made by gross domestic capital formation, export growth, infrastructure availability, foreign direct investment and literacy ratio. The coefficient of FDI is small, ranging from 0.03 to 0.08 (see Equation 1 to Equation 6 in Column 1 to Column 6 in Table 2 [ PDF 48KB | 1 pages ]) indicating that a one percent increase in the FDI to GDP ratio leads to an increase in GDP by 0.03 to 0.08. Given the amount of FDI coming to into South Asian countries, it is expected to have a small coefficient. However, the coefficients are positive and significant.

Other major contributors to growth are exports and gross domestic capital formation. The inclusion of exports in the growth function improves the overall results. Therefore, the present study supports the finding in the previous literature that FDI is beneficial for countries following an export-led growth strategy. A one percent increase in exports contributes to an approximate 0.4 percent increase in GDP in these countries (see Equations 1 to 3). Gross domestic capital formation has a positive and significant impact on GDP, with a coefficient of 0.11 and 0.13 in Equation 1 and 2, respectively. The literacy ratio, proxied for human capital, turns out to be positive, but the coefficient is negligible. Since infrastructure facilitates growth, the infrastructure index has been included in the growth function (Equation 3), which is positive and significant. The coefficient is 0.36 revealing that increase in infrastructure facility by one percent increases the growth by 0.36 percent.

From Equations 3 to 6, the openness variable (trade as a percentage of GDP) has been added, while exports have been removed. Though there is no change in the sign and coefficient of FDI, the coefficient of trade is not significantly different from 0. However, the coefficient of gross domestic capital improves. Many other relevant variables have been tried into the growth function, but failed to show any improvement in the result. The coefficient of labor is insignificant and negative. Since all the South Asian countries have labor surpluses, the labor force growth is not significant in the growth function.

Overall, the panel data results highlight the fact that FDI has a positive and significant impact on growth for four South Asian countries. Other significant factors contributing to growth are exports, gross domestic capital formation and infrastructure. Labor force growth is not significant, indicating that these countries are labor abundant countries, and it does not have any significant impact on growth. However the results of the panel estimations need be analysed with caution, since the impact of FDI varies from country to country. To substantiate the relationship between GDP and FDI inflow, a causality analysis has been done for each country, using time series data. The results of the Granger causality are reported in Table 3 [ PDF 47.4KB | 1 pages ].

Since FDI inflow as a percentage of GDI is non-stationary at levels (see Table 32 in Appendix B), Granger causality is found between the GDP growth rate and FDI growth rate, which are stationary at levels. A one-way causality is found for India and Bangladesh. In the Indian case, growth causes FDI inflow whereas in Bangladesh, FDI growth leads to GDP growth. But there is bi-directional causality in the case of Sri Lanka and Pakistan. These results support our previous panel results that FDI has a positive impact on growth in South Asia. Though it has a positive and significant impact for all four South Asian countries in the panel, FDI Granger causes growth in three countries under causality analysis.

FDI and Investment

Since FDI establishes backward and forward linkages with local industries, FDI can either complement or displace domestic investment. FDI often crowds out domestic investment due to technological superiority, better management and more efficient production process. It can also encourage domestic investment, however, by creating an enabling investment environment by transferring technologies and management techniques. The relationship between FDI and domestic investment depends, among other things, on the quality of FDI, domestic regulatory environment etc.

So far, the results of empirical studies on the impact of FDI on domestic investment are mixed. Fry (1993) finds a negative impact in India after controlling for country specific effects. Dhar and Roy (1996) confirm this finding of a negative relationship between FDI and domestic investment. A study by Bosworth and Collins (1999) examining the impact of capital inflows on domestic investment for 58 developing countries, finds that FDI has a positive and proportional impact on domestic investment. Xu (2000) finds that the technology diffusion of U.S.-affiliated MNEs is strong in developed countries but weak in less developed countries. Hanson (2001) argues that the evidence for the generation of positive spillovers for host countries by FDI is weak. In contrast, Lipsey (2002) supports the positive spillover effect of FDI from his micro studies literature review. Kokko et al. (2001) find that locally oriented FDI has a larger impact on local firms than on foreign oriented firms. In a recent study, Barrios et al. (2004) find that FDI affects the domestic firms initially but over all the positive externalities, is largely positive for the domestic industries for Ireland. Some recent studies finding a positive spillover impact of FDI are Keller and Yeaple (2003) and Haskel et al (2002). In the Indian case, Kathuria (1998, 2000) suggests that the indirect gains from FDI depend upon the local firm’s ability to learn new technologies by investing in research and development. Spillover is most commonly observed in high-tech domestic industries. This study attempts to examine the impact of FDI on investment at the macro level for South Asian countries.

Analytically, FDI can improve domestic investment though positive spillovers and by creating complementary industries. However, it can also drive out domestic investment due to higher financial power, better technology and management and higher productivity. Thus, FDI has a dynamic effect on domestic investment. Considering this, OLS or panel estimates may not be appropriate. In this study, the dynamic impact of FDI on domestic investment is examined using dynamic panel data analysis developed by Arellano and Bond (1991). This method uses the first differences of the model to eliminate the individual impact and then provides estimates using two or higher period lagged dependent variables. Here, the impact of FDI on domestic investment is examined in the following way

GDCF,it = a0 + a1,GDCFit-1 + a2,GDCFit-2 + a3FDI,it-1 + a4,FDI,it-2 + a5,GDPGR,it-1 + ut

Whereas GDCF is the gross domestic capital formation taken for domestic investment and FDI foreign direct investment, GDPGR is the growth rate of GDP.

Results analysis: The results of the investment functions by panel OLS and Arellano- Bond GMM estimation are reported in Table 4 [ PDF 47.1KB | 1 pages ]. The estimation results are significant in terms of all diagnosis statistics. The estimation is done for two periods, 1970-2003 and 1990-2003. Since all these south Asian countries have improved their gross domestic capital formation and FDI inflow during the nineties, an attempt has been made to see the impact of FDI on domestic investment separately during post 1990.

The sign of FDI inflow is positive for the current period and the past one year for the whole period 1970-2003. However, it has a positive but insignificant coefficient for the current period, implying that it does not contribute significantly to domestic investment. The coefficient of FDI lagged one year is 1.42 and significant. This implies that a one percent increase in FDI to GDP in the last year leads to an increase of 1.42 percentage in domestic investment as a percentage of GDP in the current period. For the period 1990-2003, the coefficient of FDI is negative in the current period, but it is insignificant. However, the coefficient of the FDI lagged one year and FDI lagged two years is positive and significant. The coefficients are 1.13 and 1.23, respectively, implying that increasing the FDI ratio by one percent in the last year and the past two years increases the domestic investment ratio by 1.13 and 1.23 percentage points in the current period. Thus, FDI in the current period does not affect the domestic investment ratio significantly, but affects it over time through a dynamic effect. The Sargan test from a two-step estimator does not reject the null hypothesis that over-identifying restrictions are valid. The second order autocorrelation is also not significant, implying that the obtained estimates are consistent. 30

Impact of FDI on Exports

Along with the economic reforms and increased FDI inflows, South Asian countries have also experienced higher exports growth during the nineties (see Table 1). While India and Bangladesh achieved double-digit export growth in that decade, Sri Lanka improved its export performance during the nineties compared to the eighties. Though the Pakistan economy had lower export growth during the nineties due to the economic recession, it improved during recent years. The export-related success stories of PRC (UNCTAD, 2002) and East Asian countries suggest that FDI is a powerful tool for export promotion because the relative technological superiority of multinational firms helps domestic firms, directly and indirectly, in terms of technological advancement and provides market access to export markets. However, the success stories of these economies cannot be generalized to South Asian countries given the lower level of infrastructure, slow market reforms and structural rigidities (Srinivasan, 1998). The role of FDI in export promotion depends upon the motive of investment. If the motive is to capture domestic market because of high trade costs or tariffs, FDI may not improve the export growth. On the other hand, if the motive of FDI is to make use of cheap inputs or the country’s comparative advantages to tap the export market, it may contribute to export growth.

Inward FDI contributes to productivity growth, which in turn helps increase trade. This means that most of the FDI firms are concentrated in trade-intensive sectors as their trading propensity in any sector is supposedly greater than the host country firms. These are necessary prerequisites for a successful export strategy. The literature on FDI and exports reveals a positive relationship (Aitken 1997 Blomstrom, Kokko and Zejan, 1994; De Mello, 1999; UNCTAD, 1999; Lall, 2000; Lipsey 1999). It has been also debated in the literature that export-oriented industries help domestic industries and therefore crowd-in domestic investment by creating demand for intermediate demands. Multinationals firms, who bring FDI into the host country, are larger that domestic firms, pay higher wages, have higher factor productivity, are highly capital intensive and are more likely to contribute to exports due to their international exposure and competitiveness (Haddad and Harrison, 1993, Aitken et al, 1997; Aitken and Harrison, 1999).

In the Indian case, Kumar (1994) finds that the export behaviour of foreign-controlled and domestic firms for 1980-91 did not differ significantly. However, the studies by Lall and Mohammad (1985) and Majumdar and Chhibar (1998) find a positive association between exports and foreign-owned firms. Sharma (2000) finds no effect of FDI on exports where as Agrawal (2001) finds weak support for the hypothesis that foreign firms perform better in exports compared to local firms. However, Kumar and Pradhan (2003) find that the export performance of foreign affiliates is better than local firms. Aitken et al (1997) show the FDI impact on exports using the example of Bangladesh, where the entry of a single Korean multinational in the garment industry led to the establishment of a number of domestic firms

exporting garments, creating a large export industry. Sharma (2000) empirically establishes that FDI does not affect export in the Indian context. PRC has succeeded in expanding manufacturing exports because MNEs and MNE affiliates account for over 80 percent of PRC’s high technology exports (see UNCTAD, 2002). In the Indian context, Pailwar (2001) argues that MNEs are more interested in the domestic market than exports. However, FDI in Sri Lanka, Bangladesh and Pakistan are relatively concentrated in a few export-oriented industries and these sectors receive the most FDI.

Here, an attempt is made to estimate a export function to examine the impact of FDI on exports. Looking at the theoretical literature and previous studies on the export function (Joshi and Little, 1994 and Srinivasan, 1998), the export function for South Asia is designed as follows:

Exit = a0 + a0 WIit + a2GCit + a3 FDIit + a4 INFINDEXit + a5 RERit + a6 GDPGR + ut,

where, Ex is exports and WI is world income. An increase or decrease in world income influences the exports of an economy accordingly; GC is government final consumption, which is proxied for domestic demand. The higher the domestic demand or consumption, the less resources or output for exports there is. INFINDEX is an infrastructure index, which facilitates exports; RER is the real exchange rate vs. the US dollar. The major trading of South Asian countries takes place in US dollars. Thus, any change in the value of domestic currencies in US dollars negatively affects exports and vice-versa. GDPGR is GDP growth rate. All variables are taken from World Development Indicators, CD-ROM, 2005. The period of the study is 1975-2003.

The panel results are reported in Table 5 [ PDF 47.1KB | 1 pages ], which shows that FDI has a significant positive impact on exports. The coefficient is around 1.4 across specifications, implying that a one percent increase in the ratio of FDI to GDP increases exports by more than 1.4 percent in exports to GDP. As the previous literature explains, FDI brings in better technology and managerial skills along with international marketing networking, which help the exports of the host country. This hypothesis seems to be working for South Asian countries. The other important factor that contributes to exports is the infrastructure index. The increased availability of infrastructure facilities like proper roads, rail, air, etc., certainly reduces trade costs and improves exports. Domestic demand has the expected negative sign, but it is insignificant. Though world income generally influences exports, it is insignificant for South Asian countries.

Conclusion

Overall, the study finds that FDI has a significantly positive impact on growth for four South Asian countries. Other significant factors contributing to growth are exports, gross domestic capital formation and infrastructure. These results support the hypothesis that FDI is more beneficial for the export-led growth economies of South Asia. Therefore, South Asian countries need to improve their domestic investment, exports and infrastructure facilities along with more foreign investment for higher growth. Further, FDI has a positive impact on export growth through its positive spillover effects for South Asian countries. Moreover, FDI influences exports along with infrastructure facility. Though FDI does not affect domestic investment in the current period, it has a positive and significant impact effect over time through dynamic effects.

Determinants of FDI

Foreign direct investment to developing countries has increased substantially in the nineties. However, the South Asian countries have lagged behind and received low FDI inflow compared to other developing countries (see Section IV). Therefore, the relevance of understanding foreign direct investment flows in the South Asian region is important. FDI flowing into any country depends upon the rate of return on investment and the certainties and uncertainties surrounding those returns. Therefore, private investors compare the potential return and risks of their investment in the context of different investment destinations. The literature on the determinants of FDI is very rich. The expectations of private investors in a host country are guided by a host of economic, institutional, and regulatory and infrastructure related factors. 31 Before making an investment, investors look at certain major economic policy issues particularly relating to trade, labor, governance and the regulatory framework, and the availability of physical and social infrastructure. Some of the fundamental determinants of FDI, such as geographical location, resource endowment and size of the market, are largely outside the control of the national policy (UNTAD, 2003). However, national economic policies to create a conducive investment environment, and particularly the investment framework, can help to make FDI inflows consistent with economic potential. Countries can also act on their economic determinants to maximize their economic potential. The East Asian FDI boom before 1997 showed that the accrual of the benefits of FDI depends largely on factors such as income, growth and appropriate infrastructure and labor policy. Sound macroeconomic fundamentals, along with other factors such as stable exchange rate policies, low inflation, and sustained growth, influence the decision of investors in a host country.

There are well-established theories explaining why foreign direct investment takes place and what the potential determining factors are, including the market imperfection hypothesis (Hymer, 1976), internalisation theory (Rugman 1986), and eclectic approach (Dunning, 1988). There can be vertical and horizontal FDI inflows. Vertical FDI take place when factor prices are not equalized across countries (Hanson, 2001; Helpman and Krugman, 1985). Higher trade costs and stronger firm level scale economies encourage FDI relative to exports (Barinard, 1997). Thus, horizontal FDI takes place because of trade costs (Markusen, 1984; Markusen and Venables, 1998).

According to Dunning (Dunning 1977, 1988; 1993), multinational firms enjoy three distinct types of advantages to producing abroad. They are: (i) ownership advantages; (ii) locational advantages; and (iii) internalization advantages. The ownership advantages are in the form of firm-specific intangible assets, such as technology, know-how in production, marketing or management, a patented process or design, or a registered framework or brand. Given these advantages, a firm may subsequently decide to internalize activities owing to a market failure associated with arm’s length transactions in intangible assets. Thus, producing abroad enables the firm to minimize transaction costs and increase productive efficiency. Locational advantages, therefore, complete what is known as the eclectic ownership, location and internalization (OLI) paradigm, which is frequently used to explain investment abroad in the form of FDI.

In the context of the supply of capital to a particular location, such as the South Asian countries, locational advantages or the absence thereof play an important role. Locational advantages cover a multitude of factors that can influence the choice of location. However, they can be grouped into five main categories: (i) macroeconomic fundamentals (ii) infrastructural facilities, (iii) availability and costs of specific inputs, (iv) market size and growth prospect, and (v) FDI and trade regulatory policies.

By now, there is a substantial literature explaining the determinants of FDI (Dunning, 1993; Globerman and Shapiro, 1999; Shapiro and Globerman, 2001; Bevan and Estrin, 2004; Campos and Kinoshita, 2003). All the determinants of FDI can be grouped under two categories (i) economic conditions and (ii) host country policies. Economic conditions include market size, growth prospect, rate of return, urbanisation/industrialization, labor cost, human capital, physical infrastructure, and macroeconomic fundamentals like inflation, tax regime, external debt, etc. Host country policies include the promotion of private ownership, efficient financial market, trade policies/free trade policy/regional trade agreements, FDI policies, perception of country risk, legal framework, and quality of bureaucracy. Empirical research suggests that FDI is sensitive to the host country’s overall economic policies, including its tax policy.

Potential Determinants of Foreign Direct Investment

Market size: The aim of FDI in emerging developing countries is to tap the domestic market, and thus market size does matter for domestic market oriented FDI. Market size is generally measured by GDP, per capita income or size of the middle class. The size of the market or per capita income are indicators of the sophistication and breath of the domestic market. Thus, an economy with a large market size (along with other factors) should attract more FDI. Market size is important for FDI as it provides potential for local sales, greater profitability of local sales to export sales and relatively diverse resources, which make local sourcing more feasible (Pfefferman and Madarassy 1992). Thus, a large market size provides more opportunities for sales and also profits to foreign firms, and therefore attracts FDI (Wang and Swain, 1995: Moore, 1993; Schneider and Frey, 1985; Frey, 1984). FDI inflow in any period is a function of market size (Wang and Swain, 1995). However, studies by Edwards (1990) and Asidu (2002) show that there is no significant impact of growth or market size on FDI inflows. Further, Loree and Guisinger (1995) and Wei (2000) find that market size and growth impact differ under different conditions.

Growth prospects and positive country conditions: Along with market size, the prospect of growth (generally measured by growth rates) also has a positive influence on FDI inflows. Countries that have high and sustained growth rates receive more FDI flows than volatile economies. There are good number of studies showing the positive impact of per capita growth or growth prospect on FDI (Schneider and Frey, 1985; Lipsey,1999; Dasgupta and Rath, 2000; and Durham, 2002).

Labor cost and availability of skilled labor: Cheap labor is another important determinant of FDI inflow to developing countries. A high wage-adjusted productivity of labor attracts efficiency-seeking FDI both aiming to produce for the host economy as well as for export from host countries. Studies by Wheeler and Mody (1992), Scneider and Frey (1985), and Loree and Guisinger (1995) show a positive impact of labor cost on FDI inflow. Countries with a large supply of skilled human capital attract more FDI, particularly in sectors that are relatively intensive in the use of skilled labor.

Infrastructure facilities: The availability of quality infrastructure, particularly electricity, water, transportation and telecommunications, is an important determinant of FDI. When developing countries compete for FDI, the country that is best prepared to address infrastructure bottlenecks will secure a greater amount of FDI. The previous literature shows the positive impact of infrastructure facilities on FDI inflows (Wheeler and Mody (1992), Kumar (1994), Loree and Guisinger (1995), Asidu (2002)). In this study, the construction of an infrastructure index has been attempted taking different infrastructure indicators.

Openness and export promotion: The key hypothesis from various theories is that gains from FDI are far higher in the export promotion (EP) regime than the import promotion regime. The theory proposes that import substitution (IS) regimes encourage FDI to enter in cases where the host country does not have advantages leading to extra profit and rentseeking activities. However in an EP regime, FDI uses low labor costs and available raw materials for export promotion, leading to overall output growth. Trade openness generally positively influences the export-oriented FDI inflow into an economy (Edwards (1990), Gastanaga et al. (1998), Housmann and Fernandez-arias (2000), Asidu (2001)). Overall, the empirical literature reveals that one of the important factors for attracting FDI is trade policy reform in the host country. The theoretical literature has explored the trade openness or restrictiveness of trade policies (Bhagwati, 1973; 1994; Brecher and Diaz-Alejandro, 1977; Brecher and Findley; 1983). Investors generally want big markets and like to investm in countries which have regional trade integration, and also in countries where there are greater investment provisions in their trade agreements.

Government finance: Government finance is an important issue that affects capital flows. A high fiscal deficit leads to more government liabilities and therefore more taxes and defaults on international debt. Therefore, fiscal stability is generally considered to be one of the indicators of macroeconomic stability. We consider the fiscal deficit for government finance.

Rate of return on investment: The profitability of investment is one of the major determinants of investment. Thus the rate of return on investment in a host economy influences the investment decision. Following previous studies (see Asiedu, 2002), the log of inverse per capita has been used as proxy for the rate of return on investment as capitalscarce countries generally have a higher rate of return, implying low per capita GDP. This implies that the lower the GDP per capita, the higher the rate of return and thus FDI inflow.

Human capital: The availability of a cheap workforce, particularly an educated one, influences investment decisions and thus is one of the determinants of FDI inflow. In the present study, we use both labor force growth and literacy rate. 32

Policy measures: The previous literature shows the impact of government policies including investment incentives on FDI inflows into a host country (Dunning, 2002, Blomsrom and Kokko, 2002, Schneider and Frey, 1985, Grubert and Mutti, 1991, Loree and Guisuinger, 1995, Taylor, 2000, Kumar, 2002. Though investment incentives are considered another determinant for FDI, the recent paper by Blomstrom and Kokko (2003) suggests that investment incentives alone are generally not an efficient way to increase national welfare.

Policies to promote FDI take a variety of forms, but the most common are partial or complete exemptions from corporate taxes and import duties. Standard policies to attract FDI include tax holidays, import duty exemptions, and different kinds of direct subsidies. FDI inflows are also affected by corporate tax rate differentiation. Subsidizing FDI helps multinational firms reduce production costs, improves incentives to create patents, trademarks, and enhances the relative attractiveness of locating production facilities in the country offering incentives and raises the economic benefits of FDI relative to exporting.

Earlier country-specific studies on the South Asian region find that FDI inflow to South Asian countries has been affected by structural factors such as market size, low level of incomes, extent of urbanization, availability of quality infrastructure, investment incentives and performance requirements. Thus, most of the relevant variables considered are based on the theories and the previous empirical literature for examining the determinants of FDI in South Asia. After reviewing all the potential determinants of FDI, we adopt the final FDI function below:

FDIY t, it = a0 + a1GDPit + a2GDPGRit + a3INFINDEXit + a4IRRit + a5INFLit + a6LITit + a7TRADEYit + a8LEXPit + + a9LFTGRit + a10RESMit + a11DBCYit + ut

Data Considerations: The data source for the dependent variable log of FDI in UD dollars (LFDIUSD), total nominal gross domestic product in US dollars (LGDPUSD), growth rate (GDPGR), trade openness (TRADEY) which is proxied by export plus import as percentage of GDP, export (EXY), inflation rate (INFLA), Labor force growth (LFTGR), literacy ratio (LIT), inverse rate of return (IRR), total reserves sufficient for number of months of imports (RESM), domestic bank credit as percentage of GDP (DBCY), and real interest rate (RIR) is World Development Indicators (WDI) CD-ROM 2005. The estimation period is 1975 to 2003.

Infrastructure Index: To examine the impact of infrastructure facilities on FDI inflow, one infrastructure index has been constructed. Given the availability of infrastructure indicators for the whole period of the study, only infrastructure indicators such as transport, telecommunication and energy infrastructure are considered for principal component analysis for making the infrastructure index. The following variables are taken for constructing infrastructure indicators.

  • Air freight transport per 1000 km. has been proxied for air transport and taken from the Center for Monitoring Indian Economy (CMIE).
  • Electricity use per capita is taken from World Development Indicators CD-ROM 2005.
  • Energy use per capita (kg. oil equivalent per capita) is taken from World Development Indicators CD-ROM 2005.
  • Total telephone lines per 1000 population (both main lines and hand phones) is taken from World Development Indicators CD-ROM 2005.

The eigenvalues, respective variance and cumulative variance are reported in Table 30, Appendix B. The first two components have values higher than one, explaining the large variance. The factor loadings for the original variables are reported in Table 31, Appendix B.

Order of integration of the variables: The Augmented Dickey Fuller (ADF) test (see Dickey and Fuller 1981, see Appendix A.1) has been used for testing the time series properties of the variables (see Appendix for Dickey-Fuller Unit Root Tests). The results of the unit root tests are reported in the Table in the appendix. It can be seen that we have a mixture of stationary (I(0)) and I(1) variables. Most of the relevant variables are I(1), but the growth variables are I(0). Given the importance of the growth variables, they are considered in the analysis since they are normalized variables. Though growth variables are conventional, the unit root tests examine the unit-root null based on the single equation method. Levin and Lin (1993) argue that applying a unit root test on a pooled cross-section data set, rather than performing separate unit-root tests for each individual series, can increase statistical power (see Appendix A.2). The results of the panel unit root tests are the same as the time series ADF unit root tests (see Table 33, Appendix B).

Panel Cointegration Test: There are different methods for testing panel cointegration. The Engle and Granger (1987) test presumes the null hypothesis of no-cointegration and uses residuals derived from the panel regression. The Pedroni (1995, 1997) and McCoskey and Kao (1998) panel cointegration tests are based on this method. All these panel cointegration tests allow for heterogeneity in the cointegrating coefficients. Initially, the Engle Granger twostep methodology was followed for panel cointegration tests, where unit root tests directly applied to the residuals. However, test statistics using this approach would be biased towards being found to be stationary (Pedroni 1995). Pedroni argues that applying panel unit root tests directly to regression residuals is inappropriate for several reasons, such as the lack of exogeneity of the regressors and the dependency of the residuals on the distribution of the estimated coefficients (see Pedroni 1995,1997 for details). For these reasons it is important to have a test procedure for cointegration which is robust to the presence of heterogeneity in the alternative. We prefer to use the cointegration test 33 proposed by Pedroni, as it allows for considerable heterogeneity. The panel cointegration test of Pedroni is as follows:

Yn = αt + βlixli,t + β2ix2i,t + ... + βmi,txmi,t + ei,t

i = 1, 2, ...,N, t = 1, 2, ......,T, m = 1, 2, ...M,

where T is the number of observations over time, N is the total number of individual units in the panel and M is the number of regression variables. In equation (8), αi implies a member specific intercept. Among the seven Pedroni tests, four are based on within dimensions (panel cointegration tests) and three on between dimensions (group mean panel cointegration tests). Both categories of tests are based on the null hypothesis of no cointegration: pi = 1∀i pi being the autoregressive coefficient on estimated residuals under the alternative hypothesis (i.e. pi is such that êit = ρiêit-l + υit) (see Pedroni, 1995 for detailed methodology).

Discussion of Results: Integrating the variables in the same order, i.e., I (I), facilitates examining their long run equilibrium relationships using Pedroni’s panel cointegration tests. The results of the cointegration analysis tests based on Equation 9 are presented in Table 6 [ PDF 51KB | 1 pages ]. The null hypothesis of non-cointegration against the alternative of cointegration is rejected in the case of the equation as both the panel-adf and group-adf statistics are significant at the 5 percent levels. Besides adf-statistics, the results also reaveal that the panel-PP statistics are significant. Overall, cointegration results of the two equations indicate that FDI and potential determinants are cointegrated in the long run. Table 7 [ PDF 57.6KB | 2 pages ] reports the results of individual determinant variables. The results show that total GDP is the most significant factor positively affecting FDI inflow into the South Asian countries. Given the huge population and emerging developing markets in the region, FDI is flowing to tap the domestic markets, which are huge and growing. The coefficient of the total GDP, which is basically the market size, is more than 0.5 percent for both Equation 1 and Equation 2, implying that a one-percent increase in total market size increases the FDI by 0.5 percent.

GDP growth, which is the indicator of the market’s prospects, is positive across specifications and both panel cointegration and panel OLS. However, the coeffcient is insignificant. The infrastructure index, which is one of the major determinants of FDI in developing countries, is statistically significant across all specifications. This reveals that improvements in infrastructure facility attract FDI inflows to South Asian countries. Another major factor that determines FDI inflow into South Asian countries is labor force growth. The region is known for the availability of a cheap and abundant labor force, which attracts FDI. The openness variable, trade as a percentage of GDP, is significant in all equations, implying that more outward oriented South Asian economies attract more FDI. The rate of return, proxied by inverse per capita, has also been incorporated but dropped later as it is strongly correlated with growth. Similarly, the literacy ratio and other macro stability factors such as total debt as a percentage of GDP and total reserves sufficient for number of months of imports, have been dropped from estimation as they turned out be insignificant.

Conclusion

The results from the panel cointegration estimation reveal that FDI and all its potential determinants have a long-run equilibrium relationship. The major determinants of FDI in South Asia are market size, labor force growth, infrastructure index and trade openness. However the most significant and influential factors are market size and labor force growth. Overall, South Asian countries need to maintain growth momentum to improve market size, frame policies to make better use of their abundant labor forces, improve infrastructure facilities and follow more open trade policies for attracting more FDI.

Download this Discussion Paper [ PDF 351.9KB| 76 pages ].




[previous chapter] [next chapter]


Post a Comment

We welcome your feedback on this publication. Post a comment. ADBI is not obliged to acknowledge or publish comments and may abridge or edit them before web posting.

Comment(s)

There are [0] comment(s) for this entry. Post a comment.

    The views expressed in this paper are the views of the authors and do not necessarily reflect the views or policies of the Asian Development Bank Institute (ADBI), the Asian Development Bank (ADB), its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms.

    Back to Top 
    © 2014 Asian Development Bank Institute.