Several factors appear to have contributed to this phenomenon including foreign direct investment (FDI).The study is an empirical analysis of the role of the Foreign Direct Investment (FDI) in promoting India's manufacturing exports during the period 1970-71 to 2008-09. The study assumes importance in view of the fact that manufacturing export constitutes 60 per of the total exports. FDI is found to be an important factor influencing manufacturing exports in several East Asian & South East Asian economies (the so called "Tiger Economies"). GDP as a proxy for domestic demand, world income, real effective exchange rate and unit value of export/world unit value ratio are the other important variables influencing manufacturing exports. The model presented here shows that FDI, along with domestic demand; measured by Gross Fiscal Deficit (as a % of GDP), world level of GDP as a proxy for world income, and INR-USD exchange rate, has a significant impact on the exports from manufacturing sector. Relative price of exports appear to have statistically no significant impact on export performance, although the result was later found to be quite misleading.
Economic policy of India has been known for promoting self sufficiency, which deliberately insulated its economy from external competition and exerted comprehensive controls on private enterprises. It has been tended towards protectionism for long with a strong emphasis on import substitution, state intervention in financial markets, presence of large number of Public Sector Undertakings (PSUs), in which government owns a majority of equity, and permit raj (popularly known as "Licence Raj") which required elaborative licences and regulations to set up and run business in India. In addition to the local licensing regulations, the government created laws in order to manage its foreign exchange, namely the "Foreign Exchange Regulation Act, 1973, (FERA) and related guidelines, controlling the investment and other activities of foreigners in India. The act was later replaced by the Foreign Exchange Management Act, 2000, (FEMA), with an objective of facilitating external trade. FEMA reaped benefits since its year of inception as there were significant jumps in export of manufactured goods from US$ 29714.4 million to US$ 34335.2 million (Source: handbook of statistics on Indian economy, 2008-09).
Figure Inward FDI flows (all sectors)Historically though, India's exports have grown much faster than GDP over the past few decades. For example, its exports have grown over 14% per annum while growth in GDP is about 5% during 1970-2008 periods  . Several factors appear to have contributed to this phenomenon including foreign direct investment (FDI) which has been rising consistently especially from the early 1990s after the economic reforms introduced by the then Finance Minister of India, Dr. Manmohan Singh (Figure 1). The reforms did away with the Licence RajÂ and ended many public monopolies, allowing automatic approval ofÂ foreign direct investmentÂ in many sectors.ifdi.png
According to the World Investment Report (2009), with an inflow of US $ 4.3 billion during 2003, India's FDI inflow is tiny as compared to US$ 53.5 billion worth of FDI flowing into China during the above-mentioned year. However, the experience of China & other East Asian giants cannot be generalized for India given the lower level of infrastructure, and the rigidity in both the factor as well as commodity markets. Moreover, the role of FDI in exports promotion in developing countries remains controversial and depends on the motive for investment. If the motive behind FDI is to capture domestic market, it may not contribute to export growth. On the other hand, if the motive is to tap export markets by taking advantage of countries cheap labour, then FDI may contribute to export growth. Thus, whether FDI contributes to export growth or not depends on the nature of the policy regime.
India partly opened up its market during the Rajiv Gandhi administration, which introduced reforms such as decontrolling industry, liberalize imports, reform prices, and encourage private enterprises  . Further liberalization of economy began in 1991-92 by introducing policies such as reducing tariffs from an average of 85 percent to 25 percent, encouraging foreign direct investment by increasing the maximum limit on share of foreign capital in joint ventures from 40 to 51 percent with 100 percent foreign equity permitted in priority sectors, streamlining procedures for FDI approvals, and automatically approving projects within the limits for foreign participation in some industries. Nevertheless, by any standard, India was far less open than many developing economies. Furthermore, its factor market including infrastructure sector is less efficient compared with many East and South East Asian countries with whom India competes in international market. Hence, it is possible to argue that even with the policy liberalization India may have failed to attract a significant amount of export oriented FDI and the export growth may have been brought about by factors other than FDI namely the real depreciation of Indian currency, improvements in price competitiveness and provision of export subsidies etc.
In the light of the above argument, it has been examined whether exports (particularly from the manufacturing sector) have grown in the past three decades or so only due to the spurt of FDI or whether other factors have been equally significant as well.
TRENDS OF FDI IN INDIA
During the mid-1970s and throughout early 1980s, inflows of FDI into India were negligible and the stock of FDI remained relatively low. Still, inflows had increased dramatically from an average of $78 million during 1970-91 to an average of $7897 million in the next two decades. The sectoral distribution of India's FDI also changed over this period, with FDI in plantations, mining and petroleum declining, while FDI in manufactured goods increased.
Figure Exports and GDP (%age growth)Many notable developments took place in the export sector in India during the past four decades. First, as mentioned earlier, exports have been growing faster than the GDP in India (Figure 2). Second, the export composition of India also showed significant changes. Apart from increasing FDI inflows, factors such as depreciation of Indian Rupee vis-a-vis US Dollar, increasing world income, low export unit value of India as compared to the unit price of world exports played a crucial role in increasing manufacturing sector exports in India.exports n gdp.png
Here, I have presented a model to test what all factors influence the manufacturing sector exports of India. Empirically, FDI and exports move in the same direction. The results could be ambiguous, though, since the impact of FDI depends the intention of the investor, as discussed above. An increase in domestic demand, captured by Gross Fiscal Deficit as a percentage of Gross Domestic Product (GDP) of India, tends to weaken export supply and channel it towards domestic consumption. Hence a negative link is expected between the two. Also, strong world demand will help in increasing the demand for exports on account of rising consumption in rest of the world. One would also expect an inverse relation between price of exports and exports volume. Likewise, real depreciation of the domestic currency has an unfavourable effect on exports. I, therefore, present the following time series model for thirty nine numbers of observations to capture the above mentioned idea (with expected sign in parenthesis):
(+) (-) (+) (-) (-)
mnex = Exports of manufacturing sector (US $ million)
ifdi = Inward FDI flows (US $ million)
gfd = Gross Fiscal Deficit of the central govt. as a percentage of GDP of India
wgdp = GDP at constant prices (2000) of rest of World (US $ million)
pxport = Relative price of exports (index number)
exrate = INR-USD exchange rate (Rs/$)
The relative price of exports is defined as the ratio of unit price of Indian exports (in US$) to the unit price of world exports (in US$). Though a better variable to measure the impact of variations in exchange rate could have been the Real Effective Exchange Rate (REER), which is the weighted average of a country's currency relative to an index of other major currencies adjusted for the effect of inflation, due to lack of data for the entire period of study, the INR-USD exchange rate has been taken into consideration. Also, Reliable and efficient infrastructure facilities are essential for reducing costs, ensuring timely supply of exports and thereby improving export performance. Thus, a positive link between improved infrastructure facilities and export supply is expected. However, due to non-availability of year-wise aggregate infrastructure investment data, this variable is not included for analysis.
The model specified above is estimated using yearly values from 1970-2008. Before proceeding for the actual regression, each variable has been checked for the absence of unit root using the Augmented Dickey -Fuller (ADF) test for stationarity so as to make sure that the regression is meaningful, i.e., there is not a problem of spurious regression, which may arise if the variables are non-stationary or I(1). Graphically, it can be seen that some of the variables seem non-stationary (Figure 3 & 4),
Choosing appropriate number of lags becomes imperative in the analysis of ADF test. For annual data, one or two lags usually suffice  . Moreover, the Akaike Information Criterion (AIC) test conducted on all variables suggests using one or two lags for all variables. Observing the above table, the null hypothesis of non-stationarity is rejected only in case of pxport. All the other variables were found to be non-stationary, even at 10% level of significance. Also, as the t-statistics are positive for 'mnex', 'ifdi' and 'wgdp', it means that the value for rho is slightly greater than one, suggesting that the time series variables are explosive. Since, (spurious) correlation may persist in non-stationary time series, one needs to check for co-integration of the I (1) variables.
Here, the Engel-Granger test has been conducted to test for co-integration, for which, all the I(1) variables have been taken together and regressed ( equation [A.1] in Appendix A.1). Since all the variables of this regression are non-stationary, there is a possibility that this regression is spurious. However, the ADF test on residuals found the residuals to be stationary (Appendix A.2). This implies that the regression is meaningful. And now, including 'pxport' in our analysis, the final regression equation becomes:
(Appendix B).The first thing to observe is that, except for price of exports, signs for all the variables came out as expected. The FDI variable is found to have a significant effect, as hypothesised by the study, implying that an increase of a million dollar of FDI inflow in India will lead to $1.19 million increase in the exports from manufacturing sector. Still, the co-efficient for FDI is quite less. One would have expected more jumps in export values with increasing FDI inflows. One reason for that not being the case here could be that a major part of FDI participation in Indian industries took place from 1993 onwards, and the period of study here is from 1970s. The effectiveness of other variables came out as expected.
Coming to the price of exports, one would expect an inverse relation between export prices and export demand. It is coming out to be positive for the period of study notwithstanding; the slope co-efficient stating that for every unit increase in the export unit value, the manufacturing sector exports tend to increase by $10087 million approx. Since the slope co-efficient is statistically not significant, we do not talk much about it. Although, there might be some issues with the Engel-Granger approach for testing co-integration as the results for the price of exports did not come as expected. One of the main problems with the multiple regression analysis in time series data is that there could be more than one co-integrating vector which further complicates the issue. To cross check the issue, conducting the "Johansen test for co-integration", it was found out that there were three co-integrating vectors. This might be the reason for getting unexpected results. Therefore, there is a scope of further analysis here.
The fit of the regression results is also good, thus leading us to conclude that collectively, all the variables account for 98.86% variation in the dependent variable. An F-value of as high as 537.11 is an evidence that all the explanatory variables have a significant impact on manufacturing sector's exports.
Using the Durbin-Watson d-statistics, we can check for the presence of autocorrelation in the error terms throughout our period of study. The d-statistic here is computed to be 1.7408, which is less than the upper critical value, of 1.870 for 6 and 37 degrees of freedom at 5% level of significance. This shows the presence of positive serial correlation in our regression. This means that the OLS estimators are no longer efficient. To rectify this, we use the Generalised Least Squares (GLS) method (Appendix C). The d-statistic for the GLS regression model was found to be 1.9745 which is greater than of 1.877 for 6 and 36 degrees of freedom at 5% level of significance, implying no autocorrelation, positive or negative.
Both, exports from the manufacturing sector as well as inward FDI flows have grown tremendously, especially since the last two decades. Much of it could be attributed to the liberalisation policy adopted by the government of India in 1991. This study examined the impact of FDI in India's manufacturing exports. While analysing the impact of FDI using the annual data from 1970 to 2008, we did found out a significant link between the two, suggesting a need to make more efforts to attract foreign capital in manufacturing sector. However, other factors such as world demand (measured by world GDP), INR-USD exchange rate, Gross Fiscal Deficit also have a significant impact on exports. Particularly, it was seen that a one rupee depreciation of rupee against the dollar led to a fall in export volumes by $324 million approximately. The results should be interpreted with caution, though, since the dependent variable here groups together export demand and export supply together. Therefore, separate effects on them have not been observed. Secondly, due to unavailability of data for infrastructural development, the impact of improvement in infrastructure on exports could not be seen. Also, due to a shorter time series data, longer lag effects could not be accounted for.