The plan of the paper is as follows: Section 2 examines FDI influence on the export-growth relationship, Methodology and Data source presents in Section 3 and 4 respectively, Section 4 analysis the empirical results and Section 5 presents a concluding summary.
2. FDI INFULENCE ON THE EXPORT-GROWTH RELATIONSHIP
The export-led growth hypothesis postulates that exports are a main determinant of overall economic growth. There are quite a few arguments that can be used to provide the theoretical rationale for this hypothesis. Firstly, export sector may generate positive externalities on non-export sectors through more efficient management styles and improved production techniques (Feder, 1982).The second argument is that export expansion will increase productivity by offering potential for scale economies (Helpman and Kruman, 1985).Thirdly, exports are likely to alleviate foreign exchange constraints and can thereby provide greater access to international market. These arguments have recently been supplemented by the literature on endogenous growth theory which emphasizes that exports are likely to increase long run growth by allowing a higher rate of technological innovation and dynamic learning from abroad (Lucas, 1988; and Edwards, 1992).
Despite of popularity of the ELG hypothesis, the empirical evidence is not clear. While a substantial literature, applying a range of cross-section type methodologies, supports an association between exports and growth, time series evidence fails to provide uniform support for the ELG hypothesis [See (Edwards (1993) and Giles and Williams (2000)].Moreover, the results obtained by cross sectional studies have been brought into question due to some limitations [(See Grossman and Helpman (1995)].A number of time series studies have conducted to examine ELG hypothesis, but the results are not uniform. Baldwin and Sbergami (2000) argue that the source the fragility of the trade and growth results may stem from the imposition by empirical researcher of a linear relation hip between openness proxies and growth.
The relation between GDP growth and openness is extremely complex and as we mentioned before, the liberalization process in developing countries has increased not only trade but also FDI flows. So, for a complete knowledge of the relation between openness and growth, one should include not only FDI but also the existence of linkages between trade and FDI. FDI can effect growth is by the generation of productivity spillovers, Blomstrom (1986) find evidence that FDI has led a significant positive spillover effects on the labour productivity of domestic firms and the rate of growth of domestic productivity in Mexico. Similarly evidence found in most of other Latin countries.
Nevertheless, the effect of FD on economic growth is an empirical question, as it seems to be dependent upon a set of condition in the host country economy. Openness can play a crucial role in the growth of both trade and FDI may encourage export promotion, or greater trade in intermediate inputs, especially between parent and affiliate producers (Goldberg and Klien (1998) However, the empirical evidence about the relationship between trade and FDI is ambiguous. This brief review of literature reveals that a full understanding of the relationship among trade in goods, FDI and output is required in order to analysis the extent and sources of international linkages between openness and economics performance in developing countries.
3. DATA
The model consists of five variables, total exports (exp), manufacturing production (mi) as proxy of domestic output, foreign direct income (fdi) foreign income(y), and real exchange rate(er).the last two variables were included to avoid the possibility of spurious association, as a result of variation in common determinants. The sample consists of annual time series observation (1972-2001).
Data for industrial production, foreign income, foreign direct investment and exchange rate were obtained from various issues of International Financial Statistics (IFS) and total export data obtained from Economic Survey of Pakistan (various issues).all the data in real term and consumer price index was used to convert them into real term.
4. METHODOLOGY
Prior to testing the long run and non-causality, it is necessary to establish the order of integration presented. To this end, an Augmented Dickey Fuller (ADF) was carried out on the time series levels and difference forms.
According to Johansen’s (1988) technique, to avoid spurious results in the causality testing we need to proceed as follows: firstly, to determine the order of integration of the series. Secondly, to identify the possible long-term relationships among the integrated variables included in the system. In the absence of cointeragtion vector ,with I(I) series, valid results in Granger causality testing are obtained by simply first differentiating the VAR model. With cointegration variables, Granger causality will further require inclusion of an error term in the stationary model in order to capture the short term deviations of series from their long-term equilibrium path.
It is important to determine the stationary properties of time series before we proceed with the multivariate cointegration analysis. In this study we employ the Augmented Dickey-Fuller (ADF) unit root test to determine the order of integration for all the series.
To find out the long run relationship among the variables, we employed the Johansen’s (1988) and Johansen and Juselius (1990) multiple cointegration test
Giles and Mirza (1999) have pointed out that the pretesting for non-stationarity and cointegration before the Granger causality test can lead to over rejection of a non-causal null; i.e. pretesting for non-stationarity can lead to the wrong conclusion of causality.
To deal with the possibility of distortion in the inference procedure, Toda and Yamamota (1995) and Dalado and Lutkepohl (1996),(after called TYDL) argue that we might test Granger’s concept of causality on an augmented VAR in levels even if analysed series are integrated or cointegrated of an arbitrary order. However, this procedure does not replace the conventional hypothesis testing of unit roots and cointegration ranks. It should be considers as complementary the pretesting method that may suffer inference biases (Toda and Yamamota (1995).
A modified Wald test for restrictions on the parameters of a VAR (k), where k is the lag length in the system, is utilized by procedure that developed by Toda Yamamoto (1995).When a VAR (k+dmax) is predicted (where dmax is the maximum order of integration to occur in the system), this test displays asymptotic chi-square distribution, it is also shown that if variables are integrated of order d, the usual selection procedure is valid whenever k ≥ d.
5. EMPIRICAL RESULTS
We employed Granger’s (1969) concept of causality to test the relationship between trade, FDI and domestic output using the annual data for Pakistan. We formulate a vector autoregressive (VAR) system, comprised of export, foreign direct investment; foreign income, exchange rate and domestic income, all of them are in the real terms. In selection of lag to be included in our model, we followed the Akaike Information Criteria (AIC).
The time series properties of the data are investigated using the Augmented Dickey Fuller (ADF) test based on inclusion of an intercept as well as a linear time trend and without the trend term. The results are reported in Table1.It is evident from the results shown in Table1 that all the variables have a unit root in their levels and are stationary in their first differences. Thus all five variables (y, mi, fdi, er and exp) are integrated of order one.
Table1: TEST OF THE UNIT ROOT HYPOTHESIS
The optimal lags (k) for conducting the ADF test were determined by AIC (Akaike information criteria). ** And * indicate significance at the 5% and 1% levels, respectively.
Given the common integrational properties of the variables under the consideration the next stage in the analysis is to test for the presence of multivariate cointegration in the five dimensional VAR model (exp, fdi, mi, er, y) by employing the Johansen (1988) and Johansen and Juselius (1990) procedure, using the an optimal lag structure for the VAR, are presented in Table 2 and indicates that there exist at most r=3 cointegration vectors Evidence of multivariate cointegration suggest that these variables are cointegrated.So there is long run relationship among the variables. This evidence of cointegration among the variables rules out spurious correlations and also implies at least one direction of Granger causality.
** And * Indicate significance at the 5%. And 1% respectively
The outcomes of Granger causality tests based on TYDL augmented lag method are shown in Table(3).According to estimation results ,Granger causality unidirectional running from export to output growth. This seems to confirm the ELG hypothesis for Pakistan .The results also show the existence of FDI-domestic output growth nexus. This suggests that domestic firms through spillover effect mechanism have got benefit from FDI. The findings dot not shows the FDI-led export growth nexus at the traditional level of significant. This would confirm the idea that most of multinational companies (MNCs) investment in Pakistan is not export-oriented investment. Causality also runs from foreign income to export.The results show that no causality run from export, output to FDI.
Table: 3 Granger Causality Test (TYDL Augmented Lags Methods)
** And * Indicate significance at the 5%. And 1% respectively. Figures parentheses are degree of freedom
5. CONCLUSIONS
Recent theoretical and empirical advancement on growth accounting and endogenous growth front has emphasized that FDI can be a catalyzed for development of development countries.FDI can contribute to the domestic stock of knowledge and its very presence generates a host of externalities enhancing productivity and competitiveness of the host country. The increasing importance of international capital flows and especially FDI seems to be another important component of outward-looking development policies that should not be ignored.FDI can contribute in growth both direct and indirect ways.First, introduction of new technology by MNCs has high skill content. This reflect by new vintages of capital, quality control and precision in production and accompanying increased training skill upgradation (World Bank 1997).Secondly, they bought with them a package of market knowledge and marketing skill accumulated from their long-standing experience and broader exposure to world wide competitive markets. The indirect contributions of FDI in enriching the over all knowledge of the host economy, these include productivity and export spillovers.
Thus, we examine the effects of openness in the Pakistan economy by taking into account both the trade and FDI growth links. In this paper, we analysis existence of causality between export, FDI and domestic output in Pakistan over the period 1972-2001.We found the long run relation among foreign direct investment, export and domestic growth.
Our results supports the export-led hypothesis but also the existence of FDI-growth nexus. In other words we have found significant spillovers effect from FDI to domestic output. Furthermore, our findings do not suggest a kind of FDI-led export growth linkage. This would confirm the idea the most of multinational firms investment in Pakistan is not an export-oriented investment.
In short, these findings suggest Pakistan’s capacity to progress on economic development will depend on her performance in attracting foreign capital. Pakistan’s outward looking development strategy should include FDI as an essential part in addition to export promotion strategy.
REFERENCES
Ahmad, Q, M; M.S.Butt and S.Alam (2000), Economic growth, Export, and External Debt Causality: The Case of Asian Countries. The Pakistan Development Review.39.4.591-608.
Akbar .M and Z. F.Naqvi (2000) Export Diversification and Structural Dynamic Growth Process, The Case of Pakistan. The Pakistan Development Review.No.4.573-589.
Anwer-Sajid (1985). Export function for Pakistan: A Simultaneous Equation Approach. Pakistan .Journal of Applied Economics.39-34
Braconier and Ekholm (2000), Swedish Multinationals and Competition from High and Low –Wage Locations. Review of International Economics.8:3,448-461.
Baldwin, R.E. and F. Sbergami (2000) Non-Linearity in Openness and Growth Links: Theory and Evidence, Paper Presented At the European Trade Study Group, Second Annual Conference, Glasgow.
Blomstrom.M, (1986) Foreign Investment and Productivity Efficiency: The Case of Mexico, Journal of Industrial Economics, 15, 97-110.
Dalado, J.J and H. Lutkepokl (1996) Making Wald Tests Work for Cointegrated VAR Systems, Econometric Review 15,369-86.
Engle and C.Granger (1987), Cointegration and Error Correction: Representation, Estimation and Testing. Econometrica, Vo1.55, pp.251-76.
Edwards, S. (1992) Trade Orientation, Distortion and Growth in Developing Countries, Journal of Development Economics 39, 31-57.
Edwards, S. (1993) Openness Trade Liberalization, and Growth in Developing Countries, Journal of Economic Literature 32, 1358-93.
Engle and C.Granger (1987), “Cointegration and Error Correction: Representation, Estimation and Testing”. Econometrica, Vo1.55, pp.251-76.
Freenstra and Hanson (1997), Foreign Direct Investment and Relative Wages: Evidence from Mexico’s Maquiladoras.Journal of International Economics.Vol.42.371-393.
Feder, G (1982) On Exports and Economic Growth, Journal of Development Economics 12, 59-73.
Giles, J.A and C. L .Williams (2000) Export-Led Growth: A Survey of the Empirical Literature and Some Non-Causality Results. Part 1, Journal of International Trade and Economic Development 9,261-337.
Giles, J.A and S. Mirza (1999) Some Pretesting Issues on Testing for Granger Non-Causality .Mimeao, Department of Economics, University of Victoria.
Golberg, S and W. Klien (1999) International Trade and Factor Mobility: An Empirical Investigation, NBER Paper No.7196.
Granger (1969), “Investigating Casual Relationship by Econometric Models and Cross Spectral Methods”, Econometrica, Vol.37, 424-458.
Grossman, G and Helpman (1995): Technology and Trade .in Grossman and Rogoff, K Hand Book of International Economics, Vol.3, 1279-1337.
Helpman, E and P.R Kruman (1995) Market Structure and Foreign Trade .Cambridge: MIT Press.
Johansen (1988),”Statistical Analysis of Cointegrating Vectors,” Journal of Economic Dynamic and Control, Vol.12, pp 231-54.
Johansen and Juselius (1990),”Maximum Likelihood Estimation and Inference on Cointegration with Applications the Demand for Money, “Oxford Bulletin of Economics and Statistics, Vo1 52, No.2, 169-210.
Khan, A and Afia (1995) Export Growth and Causality: An Application of Cointegration and ECM Model. The Pakistan Development Review.No.3
Lucas, R.E (1988): On the Mechanics of the Economic Development. Journal of Monetary Economics 22, 3-42.
Masih, A.M.M., and Masih (1994) On the Robustness of Cointegration Tests of the Market Efficiency Hypothesis: Evidence from Six European Foreign Exchange Markets.Economica Internazionale.Vol.47.160-180.
Masih, A.M.M., and Masih (1995b) Temporal causality and the dynamic interactions among macroeconomic activity within multivariate cointegrated system: Evidence from Singapore and Korea.Weltwirtschaftliches Archiv.39-50.
Osterwald-Lenum, M (1992) A Note with Quantiles of the Asymptotic Distribution of the Maximum Likelihood Cointegration Rank Test Statistics, Oxford Bulletin of Economics and Statistics, Vol. 54, 461-471.
Tota, H.Y and T. Yamamoto (1995) Statistical Inference in Vector Autoregressions with possibly integrated Processes, Journal of Econometrics 66,225-50.
World Investment Report (1997).Transnational Corporations, Market Structure and Competition policy, United Nations Conference on Trade and Development.
∗ The authors are project Economists and Senior Research Economist/Associate Professor at Applied Economics Research Centre, University of Karachi, Pakistan
See Braconier and Ekholm(2000) and Freenstra and Hanson(1997)
As regards foreign income, we have used US GDP because US is major trade pattern of Pakistan.
JJ (Johansen and Juselius) procedure have been quite more popular in a multivariate context, results arrived from JJ statistics in bivariate studies have also been shown to be more robust than those arrived adopting the Engle-Granger approach(see ,by example,Masih & Masih,1994,1995b)
We have considered the foreign income and exchange rate as exogenous variables
In the case of Pakistan, Empirical studies suggest that world income is important determinant of export demand of Pakistan (See Khan (1995), Anwer Sajid (1985).
